EPA/600/R-10/038F
                                        www.epa.gov/iris
EPA's Reanalysis of Key Issues Related to
      Dioxin Toxicity and Response to
         NAS Comments, Volume 1
                (CAS No. 1746-01-6)
       In Support of Summary Information on the
       Integrated Risk Information System (IRIS)
                  February 2012
             U.S. Environmental Protection Agency
                   Washington, DC

-------
                                   DISCLAIMER
       This document has been reviewed in accordance with U.S. Environmental Protection
Agency policy and approved for publication.  Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
                                          11

-------
                                     ABSTRACT
       This document comprises the first of two EPA reports (U.S. EPA 's Reanalysis of Key
Issues Related to Dioxin Toxicity and Response to NAS Comments Volumes 1 and 2 [Reanalysis
Volumes 1 and 2]} that, together, will respond to the recommendations and comments on
2,3,7,8-Tetrachlorodibenzo-/>-Dioxin (TCDD) dose-response assessment included in the 2006
NAS report, Health Risks from Dioxin and Related Compounds: Evaluation of the EPA
Reassessment.  This document, Reanalysis Volume 1, includes (1) a systematic evaluation of the
peer-reviewed epidemiologic studies and rodent bioassays relevant to TCDD dose-response
analysis; (2) dose-response analyses using a TCDD physiologically based pharmacokinetic
model that simulates TCDD blood concentrations following oral intake; and (3) an oral reference
dose (RfD) for TCDD. An RfD of 7 x  10 10 mg/kg-day is derived based on two epidemiologic
studies: (a) a study that associated TCDD exposures with decreased sperm concentration and
sperm motility in men who were exposed during childhood and (b) a study that associated
increased thyroid-stimulating hormone levels in newborn infants born to mothers who were
exposed to TCDD.  A qualitative discussion of uncertainties in the RfD and a focused
quantitative uncertainty analysis of the choices made in the development of points of departure
for RfD derivation are also provided.
                                          in

-------
                CONTENTS - DIOXIN REANALYSIS (VOL. 1)
                          (CAS No. 1746-01-6)
LIST OF TABLES	vii
LIST OF FIGURES	ix
LIST OF ABBREVIATIONS AND ACRONYMS	xii
PREFACE	xiv
AUTHORS, CONTRIBUTORS, AND REVIEWERS	xvi
EXECUTIVE SUMMARY	xxi

1.    INTRODUCTION	1-1
     1.1.   SUMMARY OF KEY NAS (2006b) COMMENTS ON DOSE-
           RESPONSE MODELING IN THE 2003 REASSESSMENT	1-3
     1.2.   EPA's SCIENCE PLAN	1-5
     1.3.   SAB (SCIENCE ADVISORY BOARD) REVIEW OF EPA'S DRAFT
           REANALYSIS	1-6
     1.4.   SCOPE OF EPA'S REANALYSIS VOLUMES 1 AND 2	1-8
     1.5.   OVERVIEW OF EPA'S RESPONSE TO NAS (2006b)	1-9
           1.5.1.  TCDD Literature Update	1-11
           1.5.2.  EPA's 2009 Workshop on TCDD Dose Response	1-12
           1.5.3.  Organization  of EPA's Response to NAS Recommendations
                (Reanalysis Volume 1)	1-14

2.    TRANSPARENCY AND CLARITY IN THE SELECTION OF KEY DATA
     SETS FOR DOSE-RESPONSE ANALYSIS	2-1
     2.1.   SUMMARY OF NAS COMMENTS ON TRANSPARENCY AND
           CLARITY IN THE SELECTION OF KEY DATA SETS FOR DOSE-
           RESPONSE ANALYSIS	2-2
     2.2.   EPA'S RESPONSE TO NAS COMMENTS ON TRANSPARENCY
           AND CLARITY IN THE SELECTION OF KEY DATA SETS FOR
           DOSE-RESPONSE ANALYSIS	2-2
     2.3.   STUDY SELECTION PROCESS FOR TCDD DOSE-RESPONSE
           ANALYSIS	2-5
           2.3.1.  Study Inclusion Criteria for TCDD Epidemiologic Studies	2-9
           2.3.2.  Study Inclusion Criteria for TCDD In Vivo Mammalian Bioassays	2-13
     2.4.   SUMMARY OF KEY DATA SET SELECTION FOR TCDD DOSE-
           RESPONSE MODELING	2-16
           2.4.1.  Key Epidemiologic Data Sets	2-44
           2.4.2.  Key Animal Bioassay Data Sets	2-45

3.    THE USE OF TOXICOKINETICS IN THE DOSE-RESPONSE MODELING
     FOR CANCER AND NONCANCER ENDPOINTS	3-1
     3.1.   SUMMARY OF NAS COMMENTS ON THE USE OF
           TOXICOKINETICS  IN DOSE-RESPONSE MODELING
           APPROACHES FOR TCDD	3-1

                                  iv

-------
                          CONTENTS (continued)
      3.2.   OVERVIEW OF EPA'S RESPONSE TO THE NAS COMMENTS ON
            THE USE OF TOXICOKINETICS IN DOSE-RESPONSE MODELING
            APPROACHES FOR TCDD	3-3
      3.3.   PHARMACOKINETICS (PK) AND PK MODELING	3-4
            3.3.1.  Pharmacokinetics (PK) Data and Models in TCDD Dose-Response
                  Modeling: Overview and Scope	3-4
            3.3.2.  Pharmacokinetics (PK) of TCDD in Animals and Humans	3-6
            3.3.3.  Pharmacokinetics (PK) of TCDD in Humans: Interindividual
                  Variability	3-14
            3.3.4.  Dose Metrics and Pharmacokinetic Models for TCDD	3-22
            3.3.5.  Uncertainty in Dose Estimates	3-97
            3.3.6.  Use of the Emond Pysiologically Based Pharmacokinetic (PBPK)
                  Models for Dose Extrapolation from Rodents to Humans	3-103

4.     ORAL REFERENCE DOSE	4-1
      4.1.   NAS COMMENTS AND EPA'S RESPONSE ON IDENTIFYING
            NONCANCER EFFECTS OBSERVED AT LOWEST DOSES	4-1
      4.2.   NONCANCER DOSE-RESPONSE ASSESSMENT OF TCDD	4-9
            4.2.1.  Determination of Toxicologically Relevant Endpoints	4-10
            4.2.2.  Use of Toxicokinetic Modeling for TCDD Dose-Response
                  Assessment	4-11
            4.2.3.  Noncancer Dose-Response Assessment of Epidemiologic Data	4-12
            4.2.4.  Noncancer Dose-Response Assessment of Animal Bioassay Data	4-18
      4.3.   REFERENCE DOSE (RfD) DERIVATION	4-39
            4.3.1.  Toxicological Endpoints	4-46
            4.3.2.  Exposure Protocols of Points of Depature (PODs)	4-47
            4.3.3.  Uncertainty Factors	4-48
            4.3.4.  Choice of Human Studies for Reference Dose (RfD) Derivation	4-50
            4.3.5.  Derivation of the Reference Dose (RfD)	4-60
            4.3.6.  Studies Reporting Outcomes Comparable to the Principal Studies
                  Used to Derive the Reference Dose (RfD)	4-61
      4.4.   QUALITATIVE UNCERTAINTIES IN THE REFERENCE DOSE (RfD) .... 4-65
      4.5.   QUANTITATIVE UNCERTAINTY IN THE REFERENCE DOSE (RfD) .... 4-71
            4.5.1.  Development of Variable Sensitivity Trees for the Principal
                  Epidemiologic Studies that were the basis of the Reference Dose
                  (RfD) and for the NTP (2006a) Rodent Bioassay	4-71
            4.5.2.  Evaluation of Range of Alternative Points of Departure (PODs) for
                  Additional Epidemiologic Endpoints	4-90

5.     REFERENCES	5-1

APPENDIX A: SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
            COMMENTS AND DISPOSITION	A-l

-------
                      CONTENTS (continued)


APPENDIX B:DIOXIN WORKSHOP	B-l

APPENDIX C: SUMMARIES AND EVALUATIONS OF CANCER AND
          NONCANER EPIDEMIOLOGIC STUDIES FOR INCLUSION IN
          TCDD DOSE-RESPONSE ASSESSMENT	C-l

APPENDIX D: SUMMARIES AND EVALUATIONS OF CANCER AND
          NONCANCER IN VIVO ANIMAL BIO AS SAYS FOR INCLUSION IN
          TCDD DOSE-RESPONSE ASSESSMENT	D-l

APPENDIX E: RODENT BIO AS SAY KINETIC MODELING	E-l

APPENDIX F: EPIDEMIOLOGIC KINETIC MODELING	F-l

APPENDIX G: NONCANCER BENCHMARK DOSE MODELING	G-l

APPENDIX H: ENDPOINTS EXCLUDED FROM REFERENCE DOSE DERIVATION
          BASED ON TOXICOLOGICAL RELEVANCE	H-l

APPENDIX I: LITERATURE SEARCH TERMS	1-1
                                VI

-------
                                 LIST OF TABLES
2-1.   Epidemiologic studies selected for TCDD cancer dose-response modeling	2-19
2-2.   Epidemiologic studies selected for TCDD noncancer dose-response modeling	2-25
2-3.   Animal bioassays selected for cancer dose-response modeling	2-28
2-4.   Animal bioassay studies selected for noncancer dose-response modeling	2-30
3-1.   Partition coefficients, tissue volumes, and volume of distribution for TCDD in
      humans	3-7
3-2.   Blood flows, permeability factors, and resulting half lives (tVa) for perfusion
      losses for humans as represented by the TCDD PBPK model of Emond et al.
      (2006; 2005)	3-9
3-3.   Toxicokinetic conversion factors for calculating human equivalent doses from
      rodent bioassays based on first-order kinetics	3-32
3-4.   Equations used in the concentration and age-dependent model (CADM; Aylward
      et al., 2005b)a	3-37
3-5.   Parameters of the concentration and age-dependent model (CADM; Aylward et
      al., 2005b)	3-38
3-6.   Confidence in the CADMa model simulations of TCDD dose metrics 	3-41
3-7.   Equations used in the TCDD PBPK model of Emond et al. (2006)	3-46
3-8.   Parameters of the PBPK model for TCDD	3-53
3-9.   Regression analysis results for the relationship between logic serum TCDD at the
      midpoint of observations and the logic of the rate constant for decline of TCDD
      levels using Ranch  Hand data	3-64
3-10.  Dosing protocols for human and animal models	3-66
3-11.  Most sensitive variables for the rat and mouse nongestational and gestational
      models	3-67
3-12.  Most sensitive variables for the human nongestational and gestational models	3-69
3-13.  TCDD serum measurements over time for two Austrian women exposed to
      TCDDinl997a	3-74
3-14.  TCDD serum measurements over time for two Seveso males exposed to TCDD in
      1976a	3-75
3-15.  Results of Hill coefficient sensitivity analysis simulations with Emond human
      PBPK model	3-77
3-16.  Alternative CYP1A2 parameter estimates for sensitivity analysis of Emond
      human PBPK model	3-78
3-17.  Results of CYP1A2 parameter sensitivity analysis simulations with Emond
      human PBPK model	3-80
3-18.  Results of Emond human PBPK model parameter sensitivity analysis simulations.
      Comparison of modeled human oral intakes for a range of lifetime average TCDD
      serum concentrations for alternative parameter values	3-81
3-19.  Confidence in the PBPK model  simulations of TCDD dose metrics	3-83
3-20.  Overall confidence associated with alternative dose metrics for noncancer dose-
      response modeling  for TCDD using rat PBPK model	3-96
3-21.  Overall confidence associated with alternative dose metrics for noncancer dose-
      response modeling  for TCDD using mouse PBPK model	3-96
                                         vn

-------
                         LIST OF TABLES (continued)
3-22.   Contributors to the overall confidence in the selection and use of dose metrics in
       the dose-response modeling of TCDD based on rat and human PBPK modelsa	3-102
3-23.   Contributors to the overall uncertainty in the selection and use of dose metrics in
       the dose-response modeling of TCDD based on mouse and human PBPK models.... 3-102
3-24.   Comparison of human equivalent doses from the Emond human PBPK model for
       the 45-year-old and 25-year-old gestational exposure scenarios	3-105
3-25.   Impact of toxicokinetic modeling on the extrapolation of administered dose to
       HED, comparing the Emond PBPK and first-order body burden models
       (administered dose =  1 ng/kg-day)	3-108
4-1.    PODs for epidemiologic studies of TCDD	4-14
4-2.    Models run for each study/endpoint combination in the animal bioassay BMD
       modeling	4-21
4-3.    Summary of key animal study PODs (ng/kg-day) based on three different dose
       metrics: administered dose, first-order body burden HED, and blood
       concentration	4-24
4-4.    TCDD analysis (NOAEL, LOAEL, BMD, and BMDL values given as animal
       whole blood concentrations in ppt)a	4-28
4-5.    Candidate RfDs for TCDD using blood-concentration-based human equivalent
       doses	4-40
4-6.    Qualitative analysis of the strengths and limitations/uncertainties associated with
       animal bioassays providing PODs for the TCDD RfD	4-52
4-7.    Basis and derivation of the TCDD RfD	4-62
4-8.    Alternative PODs for the impact of TCDD exposure during gestation and nursing
       on semen quality of male offspring (Mocarelli et al., 2011)	4-92
4-9.    Alternative PODs for developmental endpoints other than increased neonatal TSH
       and semen quality	4-92
4-10.   Alternative PODs for adult endpoints for which critical exposure windows are
       undefined	4-93
                                         Vlll

-------
                              LIST OF FIGURES
2-1.    EPA's process to select and identify in vivo mammalian and epidemiologic
       studies for use in the dose-response analysis of TCDD	2-3
2-2.    EPA's selection process to evaluate available epidemiologic studies using study
       inclusion criteria and other epidemiologic considerations for use in the dose-
       response analysis of TCDD	2-10
2-3.    EPA's process to evaluate available animal bioassay studies using study inclusion
       criteria for use in the dose-response analysis of TCDD	2-15
2-4.    Results of EPA's process to select and identify in vivo mammalian and
       epidemiologic studies for use in the dose-response analysis of TCDD	2-18
3-1.    Liver/fat concentration ratios in relation to TCDD dose at various times after oral
       administration of TCDD to mice	3-11
3-2.    First-order elimination rate fits to 36 sets of serial TCDD sampling data from
       Seveso patients as function of initial serum lipid TCDD	3-13
3-3.    Observed relationship of fecal 2,3,7,8-TCDD clearance and estimated percent
       body fat	3-16
3-4.    Unweighted empirical relationship between percent body fat estimated from body
       mass index and TCDD elimination half-life—combined Ranch Hand and Seveso
       observations	3-17
3-5.    Relevance of candidate dose metrics for dose-response modeling, based on mode
       of action and target organ toxicity of TCDD	3-23
3-6.    Process of estimating a human-equivalent TCDD lifetime average daily oral
       exposure (dH) from an experimental animal average daily oral exposure ($4) based
       on the body-burden dose metric	3-27
3-7.    Human body burden time profiles for achieving a target body burden for different
       exposure duration scenarios	3-31
3-8.    Schematic of the CADM structure	3-34
3-9.    Comparison of observed and simulated fractions of the body burden contained in
       the liver and adipose tissues in rats	3-40
3-10.   Conceptual representation of PBPK model for rat exposed to  TCDD	3-43
3-11.   Conceptual representation of PBPK model for rat developmental exposure to
       TCDD	3-44
3-12.   TCDD distribution in the liver tissue	3-47
3-13.   Growth rates for physiological  changes occurring during gestation	3-56
3-14.   Comparisons of model predictions to experimental data using a fixed elimination
       rate model with hepatic sequestration (A) and an inducible elimination rate model
       with (B) and without (C) hepatic sequestration	3-58
3-15.   PBPK model simulation of hepatic TCDD concentration (ppb) during chronic
       exposure to TCDD at 50, 150, 500, or 1,750 ng TCDD/BW using the inducible
       elimination rate model compared with the experimental data measured at the end
       of exposure	3-59
3-16.   Model predictions of TCDD blood concentration in 10 veterans (A-J) from
       Ranch Hand Cohort	3-60
3-17.   Time course of TCDD in blood (pg/g lipid adjusted) for two highly exposed
       Austrian women (patients 1 and 2)	3-61

                                           ix

-------
                        LIST OF FIGURES (continued)
3-18.   Observed vs. Emond et al. (2005) model simulated serum TCDD concentrations
       (pg/g lipid) overtime (In = natural log) in two Austrian women	3-62
3-19.   Comparison of the dose dependency of TCDD elimination in the Emond model
       vs. observations of nine Ranch Hand veterans and two highly exposed Austrian
       patients	3-63
3-20.   Elasticities in the nongestational human model, POD dose	3-71
3-21.   Elasticities in the nongestational human model, RfD dose	3-72
3-22.   Hill coefficient sensitivity analysis	3-76
3-23.   CYP1A2 parameter sensitivity analysis	3-79
3-24.   Experimental data (symbols) and model simulations (solid lines) of (A) blood, (B)
       liver, and (C) adipose tissue concentrations of TCDD after oral exposure to 150
       ng/kg-day, 5  days/week, for 17 weeks in mice	3-84
3-25.   Comparison of PBPK model simulations with experimental data on liver
       concentrations in mice administered a single oral dose of 0.001-300 ug
       TCDD/kg	3-85
3-26.   Comparison of model simulations (solid lines) with experimental data (symbols)
       on the effect  of dose on blood (cb), liver (cli), and fat (cf) concentrations
       following repetitive exposure to  0.1-450 ng TCDD/kg, 5 days/week, for 13
       weeks in mice	3-86
3-27.   Comparison of experimental data (symbols) and model predictions (solid lines) of
       (A) blood, (B) liver, and (C) adipose tissue concentrations of TCDD after oral
       exposure to 1.5 ng/kg-day, 5 days/week, for 17 weeks in mice	3-87
3-28.   Comparison of experimental data (symbols) and model predictions (solid lines) of
       (A) blood concentration, (B) liver concentration, (C) adipose tissue concentration,
       (D) feces excretion (% dose), and (E) urinary elimination (% dose) of TCDD after
       oral exposure to 1.5 ng/kg-day, 5 days/week, for 13 weeks in mice	3-88
3-29.   Comparison of experimental data (symbols) and model predictions (solid lines) of
       (A) blood concentration, (B) liver concentration, (C) adipose tissue concentration,
       (D) feces excretion (% dose), and (E) urinary elimination (% dose) of TCDD after
       oral exposure to 150 ng/kg-day,  5 days/week, for 13 weeks in mice	3-89
3-30.   PBPK model simulations  (solid lines) vs.  experimental  data (symbols) on the
       distribution of TCDD after a single acute  oral exposure to A-B) 0.1, C-D) 1.0,
       and E-F) 10  ug of TCDD/kg of body weight in mice	3-90
3-31.   PBPK model simulation (solid lines) vs. experimental data (symbols) on the
       distribution of TCDD after a single dose of 24 ug/kgBW on GD 12 in mice	3-91
3-32.   Comparison of the near-steady-state body burden simulated with CADM and
       Emond models for a daily dose ranging from 0 to 10,000 ng/kg-day in rats and
       humans	3-93
3-33.   TCDD serum concentration-time profile for lifetime, less-than-lifetime, and
       gestational exposure scenarios, with target concentrations shown for each;  profiles
       generated with Emond human PBPK model	3-104

-------
                         LIST OF FIGURES (continued)
3-34.   TCDD serum concentration-time profile for lifetime, less-than-lifetime and
       gestational exposure scenarios, showing continuous intake levels to fixed target
       concentration; profiles generated withEmond human PBPK model	3-107
4-1.    EPA's process to identify and estimate PODs from key epidemiologic studies for
       use in noncancer dose-response analysis of TCDD	4-3
4-2.    Disposition of noncancer animal bioassays selected for TCDD dose-response
       analysis	4-4
4-3.    EPA's process to identify and estimate PODs from key animal bioassays for use
       in noncancer dose-response analysis of TCDD	4-5
4-4.    Exposure-response array for ingestion exposures to TCDD	4-44
4-5.    Candidate RfD array	4-45
4-6.    Sensitivity tree showing TCDD exposure-variable uncertainty for Mocarelli et al.
       (2008)	4-73
4-7.    Sensitivity tree showing TCDD exposure-variable uncertainty for Baccarelli et al.
       (2008)	4-74
4-8.    Sensitivity tree showing TCDD exposure-variable uncertainty for NTP (2006a)	4-75
4-9.    Alternative POD exposure-response array	4-94
                                           XI

-------
                   LIST OF ABBREVIATIONS AND ACRONYMS
Ah          aryl hydrocarbon
AhR         aryl hydrocarbon receptor
AIC         Akaike Information Criterion
ANL         Argonne National Laboratory
AUC         area under the curve
BMD        benchmark does
BMDLs      benchmark dose lower confidence bounds
BMDS       Benchmark dose software
BMI         body mass index
BMR        benchmark response
BW         body weight
CADM       concentration- and age-dependent elimination model
CYP         cytochrome P450
DLC         dioxin-like compounds
EDX         effective dose eliciting x percent response
EPA         Environmental Protection Agency
FSH         follicle stimulating hormone
GD          gestation day
GI           gastrointestinal
HED         human equivalent dose
IDD         iodine deficiency disease
ILSI         International Life Sciences Institute
IQ           intelligence quotient
IRIS         Integrated Risk  Information System
KO          knockout
LASC        lipid-adjusted serum concentration
LOAEL      lowest-observed-adverse-effect-level
LOAELHED   HED estimate based on LOAELs
MOA        mode of action
NAS         National Academy of Sciences
NCEA       National Center for Environment Assessment
NIOSH       National Insitute for Occupational Safety and Health
NOAEL      no-observed-adverse-effect-level
NTP         National Toxicology Program
OSF         oral slope factor
PA          permeability x area
PBPK        physiologically  based pharmacokinetic
PCB         polychlorinated biphenyl
PCBs        polychlorinated biphenyl s
PCDFs       polychlorinated dibenzofuran
PK          pharmacokinetics
POD         point of departure
                                         xn

-------
              LIST OF ABBREVIATIONS AND ACRONYMS (continued)
RfD         reference dose
SAB         Science Advisory Board
SD          standard deviation
TC          total cholesterol
TCDD       2,3,7,8-Tetrachlorodibenzo-p-dioxin
TEF         toxicity equivalence factors
TEQ         toxicity equivalence
TK          toxicokinetic
TSH         thyroid stimulating hormone
TWA        time-weighted average
UF          uncertainty factor
UF          uncertainty factor
UFA         interspecies extrapolation factor
             database factor
             human interindividual variability
UFL         LOAEL-to-NOAEL UF
UFs         subchronic-to-chronic UF
Vd          Volume of distribution
WHO        World Health Organization
                                         xin

-------
                                      PREFACE
       This report was developed by the U.S. Environmental Protection Agency's (EPA) Office
of Research and Development (ORD), National Center for Environmental Assessment (NCEA).
       In 2003, EPA, along with other federal agencies, asked the National Academy of
Sciences (NAS) to review aspects of the science in EPA's draft dioxin reassessment titled,
Exposure and Human Health Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and
Related Compounds ("2003 Reassessment").  In 2004, EPA sent the 2003 draft Reassessment to
the NAS for their review.  In 2006, the NAS released the report of their review titled, Health
Risks from Dioxin and Related Compounds: Evaluation of the EPA Reassessment. The NAS
identified three areas in EPA's 2003 draft Reassessment that required improvement:
(1) justification of approaches to dose-response modeling for cancer and noncancer endpoints;
(2) transparency and clarity in selection of key data sets for analysis; and (3) transparency,
thoroughness, and clarity in quantitative uncertainty analysis. The NAS provided EPA with
recommendations to address their key concerns.
       In 2008, EPA, in collaboration with the Department of Energy's Argonne National
Laboratory (ANL),  developed and published a literature database of peer-reviewed studies on
TCDD toxicity, including in vivo mammalian dose-response studies and epidemiologic studies.
EPA subsequently requested public comment on this database. EPA and ANL then convened a
scientific workshop in 2009. The workshop goals were to identify and address issues related to
the dose-response assessment of TCDD and to ensure that EPA's response to the NAS focused
on the key issues and reflected the most meaningful  science.
       In May 2010, EPA released a draft report titled EPA 's Reanalysis of Key Issues Related
to Dioxin Toxicity and Response to NAS Comments ("Reanalysis") that provided a technical
response to the 2006 NAS report.  The draft Reanalysis (1)  developed a study selection process
to evaluate studies reporting cancer and noncancer effects; (2) utilized a TCDD physiologically
based pharmacokinetic (PBPK) model in its development of dose-response analyses of TCDD
lexicological and epidemiologic literature; (3) presented new analyses of both the potential
cancer and noncancer human health effects that may result from exposures to TCDD; (4)
developed an oral reference dose (RfD) for TCDD; and (5)  developed a new cancer oral slope
factor for TCDD. Federal agencies and White House offices were provided an opportunity for
review and comment on the draft Reanalysis prior to its public release; their comments are
available at www.epa.gov/iris. The draft Reanalysis received public comments and was
provided to EPA's Science Advisory Board (SAB) for independent external peer review. The
SAB convened an expert panel composed of scientists knowledgeable about technical issues
related to dioxins and risk assessment. For their review, the SAB held public meetings in June,
July, and October 2010, and in March and June 2011.
       The SAB released their final review report on August 26, 2011. In their final report, the
SAB panel: (1) commended the comprehensive and rigorous process that was used to identify
and evaluate the TCDD literature; (2) agreed that EPA's choice of kinetic model provided the
best available basis for the dose metric calculations; (3) supported EPA's selection of
two coprincipal epidemiologic studies for the derivation of the RfD for TCDD; and (4) generally
agreed with EPA's characterization of TCDD as carcinogenic to humans in accordance with
EPA's 2005 Guidelines for Carcinogen Risk Assessment and with EPA's selection of the critical
study for the quantitative cancer assessment. However, the SAB found that the draft Reanalysis
did not respond adequately to the NAS recommendation to  adopt both linear and nonlinear

                                          xiv

-------
methods of extrapolation to account for the uncertainty in the cancer dose-response curve for
TCDD.  Also, the SAB report conveyed disagreement with EPA's position in the draft
Reanalysis that a comprehensive uncertainty analysis was infeasible and suggested a number of
methods that could be used for this purpose.
      Based on the SAB review, EPA decided to separate the dioxin Reanalysis into two
volumes. This document, Volume 1, systematically evaluates the epidemiologic studies and
rodent bioassays relevant to TCDD dose response, including studies evaluating cancer and
noncancer responses. It uses a TCDD PBPK model to simulate TCDD blood concentrations, the
dose metric used in all dose-response analyses for TCDD in this volume.  Volume  1 also
develops an oral reference dose (RfD) based on two epidemiologic studies that associated TCDD
exposures with adverse health effects.  The first study reports decreased sperm concentration and
sperm motility in men who were exposed to TCDD during childhood during the Seveso accident
(Mocarelli et al., 2008), and the second reports increased thyroid-stimulating hormone levels in
newborns born to mothers who were exposed to TCDD during the Seveso accident (Baccarelli et
al., 2008). Volume 1 also provides a focused quantitative uncertainty analysis of the decisions
made in the development of points of departure for TCDD RfD derivation.
      In Volume 2, EPA will complete the evaluation of cancer mode-of-action, cancer
dose-response modeling, including justification of the approaches used for dose-response
modeling of the cancer endpoints, and an associated quantitative uncertainty analysis.  The
information provided in Volume 1 will be used in three ways: (1) as the first of two reports that
contain EPA's response to the NAS (2006b) report, (2) as the Support Document for the TCDD
noncancer IRIS Summary and TCDD oral RfD, and (3) as technical support for the dioxin
Reanalysis Volume 2. The summaries of the cancer studies included in Volume 1 are presented
for use related to noncancer effects. These summaries are not intended to inform regulatory or
other decision-making purposes related to carcinogenesis; further, no quantitative dose-response
assessments are developed for cancer studies in Volume  1.
                                          xv

-------
                  AUTHORS, CONTRIBUTORS, AND REVIEWERS
PRIMARY AUTHORS

National Center for Environmental Assessment, U.S. Environmental Protection Agency,
Cincinnati, OH
       Belinda Hawkins
       Glenn Rice (Project Colead)
       Jeff Swartout (Project Colead)
       Linda K. Teuschler
CONTRIBUTING AUTHORS

National Center for Environmental Assessment, U.S. Environmental Protection Agency,
Cincinnati, OH
       Janet Hess-Wilson (formerly with NCEA, currently with Department of Defense)
       Scott Wesselkamper
       Michael Wright
       Bette Zwayer

National Health and Environmental Effects Research Laboratory, U.S. Environmental Protection
Agency, Research Triangle Park, NC
       Hi sham El-Masri

Argonne National Laboratory, Argonne, IL
       Margaret MacDonell

University of Montreal; BioSimulation Consulting, Newark, DE
       Claude Emond

University of Montreal, Montreal, Canada
       Kannan Krishnan
CONTRIBUTORS

National Center for Environmental Assessment, U.S. Environmental Protection Agency,
Washington, DC
       Karen Hogan
       Leonid Kopylev

Argonne National Laboratory, Argonne, IL
       Maryka H. Bhattacharyya                 Mary E. Finster
       Andrew Davidson                        David P. Peterson
                                         xvi

-------
            AUTHORS, CONTRIBUTORS, AND REVIEWERS (continued)
CONTRIBUTORS (continued)

Bruce Allen Consulting, Chapel Hill, NC
       Bruce C. Allen

Clark University, Worcester, MA
       Dale Hattis

Colorado State University, Fort Collins, CO
       Raymond Yang, Retired

Emory University, Atlanta, GA
       Kyle Steenland

ICF International, Durham, NC
       Robyn Blain
       Rebecca Boyles
       Patty Chuang
       Cara Henning
       Baxter Jones
       Penelope Kellar
       Mark Lee
       Nikki Maples-Reynolds
       Amalia Marenberg

Penn State University, University Park, PA
       Jack P.  Vanden Heuvel

Resources for the Future, Washington, DC
       Roger M. Cooke

Risk Sciences International, Ottawa, Ontario
       Jessica  Dennis
       Dan Krewski
       Greg Paoli

University of California-Berkeley, Berkeley, C A
       Brenda Eskenazi

University of California-Irvine, Irvine, CA
       Scott Bartell
Garrett Martin
Margaret McVey
Sara Mishamandani
Chandrika Moudgal
Bill Mendez
Ami Parekh
Andrew Shapiro
Courtney Skuce
Audrey Turley
Salomon Sand
Natalia Shilnikova
Paul Villenueve
                                         xvn

-------
            AUTHORS, CONTRIBUTORS, AND REVIEWERS (continued)
REVIEWERS

       This document has been provided for review to EPA scientists and interagency reviewers
from other federal agencies and White House offices.
INTERNAL REVIEWERS

National Center for Environmental Assessment, U.S. Environmental Protection Agency
     Ted Berner, Washington, DC              Matthew Lorber, Washington, DC
     Glinda Cooper, Washington, DC          Eva McLanahan, Research Triangle Park, NC
     Ila Cote, Research Triangle Park, NC      Susan Rieth, Washington, DC
     Lynn Flowers, Washington, DC           Reeder Sams, Research Triangle Park, NC
     Martin Gehlhaus, Washington, DC         Paul Schlosser, Research Triangle Park, NC
     Kate Guyton, Washington, DC            Jamie Strong, Washington, DC
     Samantha Jones, Washington, DC         John Vandenberg, Research Triangle Park, NC
U.S. ENVIRONMENTAL PROTECTION AGENCY
SCIENCE ADVISORY BOARD
DIOXIN REVIEW PANEL

Chair
Timothy Buckley, Associate Professor and Chair, Division of Environmental Health Sciences,
    College of Public Health, The Ohio State University, Columbus, OH

Members
Harvey Clewell, Director of the Center for Human Health Assessment, The Hamner Institutes for
    Health Sciences, Research Triangle Park, NC
Louis Anthony (Tony) Cox, Jr., President, Cox Associates, Denver, CO
Elaine Faustman, Professor and Director, Institute for Risk Analysis and Risk Communication,
    School of Public Health, University of Washington, Seattle, WA
Scott Person,  Senior Scientist, Applied Biomathematics, Setauket, NY
Jeffrey Fisher, Research Toxicologist, National Center for Toxicological Research, U.S. Food
    and Drug Administration, Jefferson, AR
Helen Hakansson, Professor of Toxicology, Unit of Environmental Health Risk Assessment,
    Institute of Environmental Medicine, Karolinska Institutet,  Stockholm, Sweden
Russ Hauser, Frederick Lee Hisaw Professor, Department of Environmental Health, Harvard
    School of Public Health, Boston, MA
B. Paige Lawrence, Associate Professor, Departments of Environmental Medicine and
    Microbiology and Immunology, School of Medicine and Dentistry, University of Rochester
    School of Medicine and Dentistry, Rochester, NY
                                         xvin

-------
             AUTHORS, CONTRIBUTORS, AND REVIEWERS (continued)
SCIENCE ADVISORY BOARD (continued)

Michael I. Luster, Professor, Department of Community Medicine, West Virginia University
   Health Sciences Center, Morgantown, WV
Paolo Mocarelli, Professor of Clinical Biochemistry, Department of Clinical Laboratory,
   Hospital of Desio-Nuovo Monoblous, University of Milano Bicocca, Desio-Milano, Italy
Victoria Persky, Professor, Epidemiology and Biostatistics Program, School of Public Health,
   University of Illinois at Chicago, Chicago, IL
Sandra L. Petersen, Professor, Associate Graduate Dean, Department of Veterinary and Animal
   Sciences, College of Natural Sciences, University of Massachusetts-Amherst, Amherst, MA
Karl Rozman, Professor, Pharmacology, Toxicology and Therapeutics, The University of Kansas
   Medical  Center, Kansas City, KS
Arnold Schecter, Professor, Environmental and Occupational Health Sciences, School of Public
   Health-Dallas Campus, University of Texas, Dallas, TX
Allen E. Silverstone, Professor, Department of Microbiology and Immunology, Health Science
   Center, SUNY Upstate Medical University, Syracuse, NY and Adjunct Professor of
   Environmental Medicine, University of Rochester School of Medicine and Dentistry,
   Rochester, NY
Mitchell J. Small, The H. John Heinz III Professor of Environmental Engineering, Department of
   Civil and Environmental Engineering  and Engineering and Public Policy,  Carnegie Mellon
   University, Pittsburgh, PA
Anne Sweeney, Professor of Epidemiology, Department of Epidemiology and Biostatistics,
   School of Rural Public Health, Texas A&M Health Science Center, College Station, TX
Mary K. Walker, Professor, Division of Pharmaceutical Sciences, College of Pharmacy,
   University of New Mexico,  Albuquerque, NM
ACKNOWLEDGMENTS

National Center for Environmental Assessment, U.S. Environmental Protection Agency
     Rebecca Clark, Washington, DC              Michael Troyer, Cincinnati, OH
     Jeff Frithsen, Washington, DC                Maureen Johnson, Washington, DC
     Kathleen Deener, Washington, DC            Linda Tux en, Washington, DC, Retired

Immediate Office of the Assistant Administrator of Office of Research and Development,
U.S. Environmental Protection Agency
       Peter Preuss, Washington, DC

National Risk Management Research Laboratory, U.S. Environmental Protection Agency
     Annette Gatchett, Cincinnati, OH

National Exposure Research Laboratory, U.S. Environmental Protection Agency
     Andrew Gillespie, Cincinnati, OH

                                          xix

-------
             AUTHORS, CONTRIBUTORS, AND REVIEWERS (continued)
ACKNOWLEDGMENTS (continued)

Office of Administrative and Research Support, U.S. Environmental Protection Agency
       Marie Nichols-Johnson, Cincinnati, OH

Colorado State University, Fort Collins, CO
       William H. Farland
ECFlex, Inc., Fairborn, OH
       Dan Heing
       Heidi Glick
       Debbie Kleiser
       Crystal Lewis

IntelliTech Systems, Inc., Fairborn, OH
       Cris Broyles
       Luella Kessler
       Stacey Lewis
Sandra Moore
Amy Prues
Lana Wood
Kathleen Secor
Linda Tackett
National Institute of Environmental Health Sciences, Research Triangle Park, NC
       Linda S. Birnbaum
       Christopher J. Portier

National Toxicology Program, Research Triangle Park, NC
       Nigel Walker
       Michael Devito

2009 Dioxin Workshop Participants
                                          xx

-------
                               EXECUTIVE SUMMARY

OVERVIEW
       Dioxins and dioxin-like compounds (DLCs), including polychlorinated dibenzo-dioxins,
polychlorinated dibenzofurans, and polychlorinated biphenyls, are structurally and
lexicologically related halogenated dicyclic aromatic hydrocarbons.1  Dioxins and DLCs are
released into the environment from several industrial sources such as chemical manufacturing,
combustion, and metal processing; from individual activities including the burning of household
waste; and from natural processes such as forest fires. Dioxins and DLCs are widely distributed
throughout the environment and typically occur as chemical mixtures.  They do not readily
degrade; therefore, levels persist in the environment, build up in the food chain, and accumulate
in the tissues of animals. Human exposure to these compounds occurs primarily through the
ingestion of contaminated foods (Lorber et al., 2009), although exposures to other environmental
media and by other routes and pathways do occur.
       The health effects from exposures to dioxins and DLCs have been documented
extensively in  epidemiologic and toxicological  studies. 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD) is one of the most toxic members of this class of compounds and has a robust
toxicological database.  Characterization of TCDD toxicity is critical to the risk assessment of
mixtures of dioxins and DLCs because it has been selected repeatedly as the "index chemical"
for the dioxin toxicity equivalence factors (TEF) approach. In this approach, the toxicity of
individual components of dioxin and DLC mixtures is scaled to that of TCDD.  Then, the
dose-response information for TCDD is used by the U.S. Environmental Protection Agency
(EPA) and other organizations to evaluate risks from exposure to mixtures of DLCs  (U.S. EPA,
201 Ob: Van  den Berg et al., 2006: Van den Berg et al., 1998).
       To provide guidance on the use of the TEF approach in environmental health risk
assessments, EPA published a report titled, Recommended Toxicity Equivalence Factors (TEFs)
for Human Health Risk Assessments of 2,3,7,8-Tetrachlorodibenzo-p-dioxin and Dioxin-Like
Compounds (TEF report) (U.S. EPA, 201 Ob). The TEF report describes EPA's updated
approach for evaluating the human health risks from exposures to environmental media
containing DLCs. In the TEF report, EPA recommends use of the consensus TEF values for
1 For further information on the chemical structures of these compounds, see U.S. EPA (U.S. EPA. 2010b. 2008b,
2003).
                                          xxi

-------
TCDD and DLCs published in 2005 by the World Health Organization (Van den Berg et al..
2006) for all cancer and noncancer effects mediated through aryl hydrocarbon receptor binding.
Further, EPA recommends that the TEF methodology, a component mixture method, be used to
evaluate human health risks posed by these mixtures, using TCDD as the index chemical;
therefore, it is imperative to correctly assess the dose response of TCDD and understand the
uncertainties and limitations therein.
       In 2003, EPA completed a comprehensive human health assessment external review draft
titled, Exposure and Human Health Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin
(TCDD) and Related Compounds ("2003 Reassessment"). As part of EPA's commitment to the
development of health assessment information of the highest scientific integrity, scientific peer
review is an integral component of the process EPA uses to generate high quality toxicity and
exposure assessments of environmental contaminants. To this end, EPA asked the National
Academy of Sciences (NAS) to review the 2003 draft Reassessment. In 2006, NAS released
their report titled, Health Risks from Dioxin and Related Compounds: Evaluation of the  EPA
Reassessment (NAS,  2006a). In this review, the NAS identified three key recommendations
requiring improvement to support a scientifically robust characterization of human responses to
exposures to TCDD.  These three key areas are (1) improved transparency and clarity in the
selection of key data sets for dose-response analysis, (2) further justification of approaches to
dose-response modeling for cancer and noncancer endpoints, and (3) improved transparency,
thoroughness, and clarity in quantitative uncertainty analysis. NAS also encouraged EPA to
calculate an oral noncancer reference dose (RfD), and provided specific comments on various
aspects of EPA's 2003 draft Reassessment.
       In May 2009,  EPA Administrator Lisa P. Jackson announced the Science Plan for
Activities Related to Dioxins in the Environment ("Science Plan") that addressed the need to
finish EPA's dioxin reassessment and provide a completed health assessment on this high profile
chemical to the American public.2  The Science Plan stated that EPA would release a draft report
responding to the recommendations and comments included in the NAS review of EPA's 2003
draft Reassessment.
       As outlined in the  Science Plan, in 2009, EPA developed a draft report titled EPA's
Reanalysis of Key Issues Related to Dioxin Toxicity and Response to NAS Comments
 Available online at http://www.epa.gov/dioxin/scienceplan.
                                          xxii

-------
("Reanalysis") that responded to the key comments and recommendations in the NAS report
(U.S. EPA, 2010a).  The draft Reanalysis focused on TCDD dose-response issues and included
analyses of relevant new studies and the derivation of an oral noncancer RfD and an oral slope
factor (OSF) for cancer.  The draft Reanalysis was reviewed internally by EPA scientists and was
provided for review to other federal agencies and White House offices. On May 21, 2010, the
draft Reanalysis was released for public review and comment and independent external peer
review by EPA's Science Advisory Board (SAB).
       For their review, the SAB held public meetings in June, July, and October 2010, and in
March and June 2011.  They released their final report reviewing the draft Reanalysis on August
26, 2011 (SAB, 2011).3 In their report, the SAB communicated the following  overarching
observations:
   •   They found that the draft Reanalysis was clear, logical, and responsive to many—but not
       all—of the NAS recommendations; they were impressed with the comprehensive and
       rigorous study selection process that was used to identify, review and evaluate the
       scientific literature on TCDD dose response;
   •   They agreed with the choice of the Emond physiologically based pharmacokinetic
       (PBPK) model for dose metric calculations and with the selection of whole blood as the
       dose metric;
   •   They agreed with the choice of two epidemiologic studies as coprincipal studies whose
       developmental toxicity data were used to derive the RfD for TCDD;
   •   They agreed with EPA's cancer weight of evidence classification of TCDD as
       carcinogenic to humans (with the exception of one panelist with a dissenting view);
The SAB also identified two deficiencies in EPA's draft Reanalysis with respect to the
completeness of the consideration of two critical elements:
   •   Nonlinear dose response for TCDD carcinogenicity; and
   •   Uncertainty analysis
3 Available online at
http://vosemite.epa.gov/sab/sabproduct.nsf/2A45B492EB AA8553852578F9003ECBC5/$File/SAB-ll-014-
unsigned.pdf.
                                          xxiii

-------
       The SAB recommended that EPA fully evaluate both linear and nonlinear dose-response
approaches to TCDD cancer dose-response assessment—including a discussion of carcinogenic
mode of action. The SAB also recommended a number of approaches to quantitative uncertainty
analysis that could be implemented by EPA, including the use of sensitivity analyses and
probability trees.
       In August 2011, EPA announced a plan for moving forward to complete the draft
Reanalysis.4 Per this plan, the current document is the first of two EPA reports (U.S. EPA 's
Reanalysis of Key Issues Related to Dioxin Toxicity and Response to NAS Comments Volumes 1
and 2 [Reanalysis Volumes 1 and 2]} that, together, will respond to the recommendations and
comments on TCDD dose-response assessment included in the NAS review of EPA's 2003 draft
Reassessment.  Both Volumes focus on TCDD only.  This report, Reanalysis Volume 1,
completes and  publishes EPA's study selection criteria and study selection results for both
noncancer and cancer TCDD dose-response assessment; choice of kinetic model; noncancer RfD
for TCDD; and a qualitative discussion of uncertainties in the RfD with a focused quantitative
uncertainty analysis. Reanalysis Volume 1 responds to key comments and recommendations
pertaining to noncancer TCDD dose-response assessment published by the NAS in their review
(NAS. 2006b).
       The information and analyses in this Volume have undergone revisions in response to
SAB comments and recommendations as well as comments provided by the public (see
Appendix A).  Reanalysis Volume 2 will address the two deficiencies identified by the SAB, i.e.,
nonlinear dose response for TCDD carcinogenicity and quantitative uncertainty analysis for
TCDD carcinogenicity. In Volume 2, EPA will complete the evaluation of cancer mode of
action, cancer dose-response modeling, including an updated literature search, justification of the
approaches used for dose-response modeling of the cancer endpoints, and an associated
quantitative uncertainty analysis. The information provided in Volume 1 will be used in three
ways: (1) as the first of two reports that contain EPA's response to the NAS (2006b) report,
(2) as the Support Document for the TCDD noncancer IRIS Summary and TCDD oral RfD, and
(3) as technical support for Reanalysis Volume 2. The summaries of the cancer studies included
in Volume 1 are presented for use related to non-cancer effects. They also provide information
on the complete literature review and study selection process that EPA conducted in preparing
 Available online at http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=209690.
                                         xxiv

-------
the draft Reanalysis, which included information on both cancer and noncancer effects. These

summaries are not intended to inform regulatory or other decision-making purposes related to

carcinogenesis; further, no quantitative dose-response assessments are developed for cancer

studies in Volume 1.  The final cancer analysis will be included in EPA's Reanalysis, Volume 2.

       The three key NAS recommendations specifically pertain to dose-response assessment

and uncertainty analysis. Therefore, EPA's response to the NAS in this document is focused on

these issues.

       EPA thoroughly considered the recommendations of the NAS and, in Reanalysis

Volume 1, responds with an evaluation of TCDD hazard identification and dose-response data

via the following:
       An updated literature search that identified new TCDD dose-response studies (see
       Section 2);

       A workshop that included the participation of external experts in TCDD health effects,
       toxicokinetics, dose-response assessment and quantitative uncertainty analysis; these
       experts discussed potential approaches to TCDD dose-response assessment and
       considerations for EPA's response to the NAS (U.S. EPA, 2009a) (see Appendices B and
       i);

       Development of a detailed study selection process including criteria and considerations
       for the selection of key epidemiologic and animal bioassay studies (see Section 2.3) for
       quantitative TCDD dose-response assessment (see Section 2.4.1/Appendix C and Section
       2.4.27Appendix D, respectively);

       Kinetic modeling that quantifies appropriate dose metrics for use in TCDD dose-response
       assessment (see Section 3 and Appendices E and F);

       A sensitivity analysis performed on each of the Emond animal  and human PBPK models
       that identify the most sensitive variables in each model (see Section 3.3.4);

       Dose-response modeling for all appropriate noncancer data  sets (see
       Section 4.27Appendix G);

       A thorough and transparent evaluation of the selected TCDD data for use in the
       derivation of an RfD, including justification of approaches used for dose-response
       modeling of noncancer endpoints (see Section 4.2 and Appendix H);

       The development of an RfD (see Section 4.3);
                                          xxv

-------
   •   A qualitative discussion of the uncertainty in the RfD and a focused quantitative
       uncertainty analyses of the RfD (see Sections 4.4 and 4.5, respectively); and
   •   Responses to the comments and recommendations made by the SAB in their final report
       (SAB. 2011) (see Appendix A).
       Those activities and analyses are briefly described in this Executive Summary, and they
are described in detail in the related sections of this document.
       In addition to this document, several additional EPA activities address other TCDD
issues, specifically related to the application of dioxin TEFs and to TCDD and DLC background
exposure levels. Information on the application of the dioxin  TEFs is published elsewhere by
EPA for both ecological (U.S. EPA. 2008b) and human health assessment (U.S. EPA. 201 Ob).
As a consequence, EPA does not directly address TEFs herein but makes use of the concept of
toxicity equivalence as applicable to the analysis of exposure  dose uncertainty in epidemiologic
studies and an animal bioassay.  Furthermore, this document does not address the NAS
recommendations pertaining to the assessment of human exposures to TCDD  and other dioxins.
Information on updated background levels of dioxin in the U.S. population has been recently
reported (Lorber et al., 2009). In 2006, EPA also released a report titled An Inventory of Sources
and Environmental Releases of Dioxin-Like Compounds in the United States for the Years 1987,
1995, and 2000, which presents an evaluation of sources and emissions of dioxins,
dibenzofurans, and coplanar polychlorinated biphenyls (PCBs) to the air, land and water of the
United States (U.S. EPA. 2006a).

PRELIMINARY ACTIVITIES UNDERTAKEN BY EPA TO ENSURE THAT THE
REANLAYSIS VOLUMES 1 AND 2 REFLECT THE CURRENT STATE-OF-THE-
SCIENCE
       As part of the development of this document, EPA undertook two activities that involved
the public: an updated literature search and a scientific expert workshop.  The adverse health
effects associated with TCDD exposures are documented extensively in epidemiologic and
toxicologic studies. As such, the database of relevant information pertaining to the
dose-response assessment of TCDD is vast and constantly expanding. Responding directly to the
NAS recommendation to use the most current and up-to-date  scientific information related to
TCDD, EPA, in collaboration with the Department of Energy's Argonne National Laboratory

                                         xxvi

-------
(ANL), developed an updated literature database of peer-reviewed studies on TCDD toxicity,
including in vivo mammalian dose-response studies and epidemiologic studies.  An initial
literature search for studies published since the development of the 2003 draft Reassessment was
conducted to identify studies published between January 1, 2000, and October 31, 2008.  EPA
published the initial literature search results in the Federal Register in November 2008 and
invited the public to review the list and submit additional, relevant, peer-reviewed studies.
Additional studies identified by the public and through continued work on this response were
incorporated into the final set of studies for TCDD  dose-response assessment (updated through
October 2009).  Since release of the draft Reanalysis for public comment and external peer
review in 2010, EPA has collected a limited number of additional studies that inform EPA's
derivation of an RfD for TCDD. These studies were identified by EPA scientists, the SAB, and
the public, and they have been used to further evaluate the biological significance of the
endpoints used to derive the RfD and to develop information on uncertainty in the RfD.  These
additional studies are cited in the appropriate sections of this document.  None of the data sets
collected since October 2009 was used quantitatively in the noncancer dose-response assessment
of TCDD.
       To assist in responding to the NAS, EPA, in collaboration with ANL, convened a
scientific expert workshop ("Dioxin Workshop") in February 2009 that was open to the public.
The primary goals of the Dioxin Workshop were to identify and address issues related to the
dose-response assessment of TCDD and to ensure that EPA's response to the NAS focused on
the key issues, while reflecting the most meaningful science.  EPA and ANL assembled expert
scientists and asked them to identify and discuss the technical challenges involved in addressing
the NAS comments, discuss approaches for addressing these key recommendations, and to assist
in the identification of important published and peer-reviewed literature on TCDD.  The
workshop was structured into seven scientific topic sessions as follows: (1) quantitative
dose-response modeling issues, (2) immunotoxicity, (3) neurotoxicity and nonreproductive
endocrine effects, (4) cardiovascular toxicity and hepatotoxicity, (5) cancer, (6) reproductive and
developmental toxicity, and (7) quantitative uncertainty analysis of dose response. External
cochairs (i.e., scientists who were not members of EPA or ANL) were asked to facilitate the
sessions and then prepare summaries of discussions occurring in each session. The session
                                          xxvn

-------
summaries formed the basis of a final workshop report (U.S. EPA, 2009a) (see Appendix B).
Some of the key outcomes from the workshop include the following recommendations:
       Further develop study selection criteria for evaluating the suitability of developing
       dose-response models based on animal bioassays and human epidemiologic studies;
       Use kinetic modeling to identify relevant dose metrics and dose conversions between test
       animal species and humans, and between human internal dose measures and human
       intakes;
       Consider newer human or animal bioassay (NTP, 2006a) publications when evaluating
       quantitative dose-response models for cancer;
       Consider both linear and nonlinear modeling in the cancer dose-response analysis.
The discussions held during the Dioxin Workshop helped inform, guide, and focus EPA's
response to the NAS.

EPA'S APPROACH TO CONSIDERING TRANSPARENCY AND CLARITY IN THE
SELECTION OF KEY STUDIES AND DATA SETS FOR DOSE-RESPONSE
MODELING
       One of the key NAS recommendations to EPA was to utilize a clear and transparent
process for the selection of key studies and data sets for dose-response assessment. EPA agrees
with the NAS and believes that clear delineation of the study selection process and decisions
regarding key studies and data sets will facilitate communication of critical decisions made in the
TCDD dose-response assessment. EPA developed detailed processes and TCDD-specific
criteria and considerations for the selection of key dose-response studies.  These criteria and
considerations are based on current guidance for point of departure (POD) identification and RfD
and OSF derivation (U.S. EPA. 2005a. b, 2000. 1998. 1996. 1991. 1986a. b); they also consider
issues specifically related to TCDD. These criteria reflect EPA's goal  of developing noncancer
and cancer toxicity values for TCDD through a transparent study selection process. Following
the selection of key studies, EPA employed additional processes to further select and identify
cancer and noncancer data sets from these key studies for use in dose-response analysis of
TCDD.
       Figure ES-1 presents EPA's study selection process for the evaluation of the
epidemiologic studies considered for this TCDD dose-response assessment, including specific
                                         xxviii

-------
study inclusion criteria (see Section 2.3.1).  EPA applied its TCDD-specific epidemiologic study

inclusion criteria to all studies published on TCDD. For all peer reviewed studies, EPA
                         List of available epidemiologic studies on TCDD and DLCs
                                      (All studies summarized.)
                                            Study
                                         in peer-reviewed
                                           literature?
                                           Exposure
                                        primarily to TCDD
                                        and quantifiable?
                            Long-term
                          exposures and
                         latency information
                         available forcancer
                            ssessment?
    Exposure
  windows and
latency information
 available for RfD
   ssessment?
                   Evaluate study using five considerations:
                   •  Methods used to ascertain health outcomes are clear and unbiased?
                   •  Confounding and other potential sources of bias are addressed?
                   •  Association/exposure response between TCDD and adverse effect?
                   •  Exposures based on individual-level estimates, uncertainties described?
                   •  Statistical precision, powerand study follow-up are sufficient?
                                         Considerations
                                          adequately
                                           satisfied?
                                       Key study included
                                 forTCDD cancerand/or noncancer
                                    dose-response assessment
                  Study excluded
                    from TCDD
                  dose-response
                    assessment
        Figure ES-1.  EPA's selection process to evaluate available epidemiologic
        studies using study inclusion criteria and other epidemiologic considerations
        for use in the  dose-response analysis of TCDD.
        EPA applied its TCDD-specific epidemiologic study inclusion criteria to all studies published on
        TCDD and DLCs. For all peer reviewed studies, EPA examined whether the exposures were
        primarily to TCDD and if the TCDD exposures could be quantified so that dose-response analyses
        could be conducted. Then, EPA required that the effective dose and oral exposure be estimable:
        (1) for cancer, information is required on long-term exposures, (2) for noncancer, information is
        required regarding the appropriate time window of exposure that is relevant for a specific, nonfatal
        health endpoint, and (3) for all endpoints, the latency period between TCDD exposure and the
        onset of the health endpoint is needed.  Finally, studies were evaluated using five considerations
                                                 XXIX

-------
       regarded as providing the most relevant kind of information needed for quantitative human health
       risk analyses. Only studies meeting these criteria and adequately satisfying the considerations
       were selected for EPA's TCDD dose-response analysis.

examined whether the exposures were primarily to TCDD and if the TCDD exposures could be
quantified so that dose-response analyses could be conducted. Then, EPA required that the
effective dose  and oral exposure be estimable: (1) for cancer, information is required on
long-term exposures, (2) for noncancer, information is required  on the appropriate time window
of exposure that is relevant for a specific, nonfatal health endpoint, and (3) for all endpoints,
information concerning the latency period between TCDD exposure and the onset of the effect is
needed. Finally, studies were evaluated using five considerations regarded as providing the most
relevant kind of information needed for quantitative human health risk analyses.  Only studies
meeting these  criteria and adequately satisfying the considerations were included in EPA's
TCDD dose-response analysis.
       Figure  ES-2 presents EPA's study selection process for the evaluation of mammalian
bioassays considered for TCDD dose-response assessment—including the specific study
inclusion criteria (see Section 2.3.2).  EPA evaluated all available in vivo mammalian bioassay
studies on TCDD. Studies had to be published in the peer-reviewed literature.  Studies on
genetically altered species were excluded as their direct relevance to human health is not known.
Next, EPA applied dose requirements to each study's lowest tested average daily dose, with
specific requirements for cancer (<1 ug/kg-day) and noncancer (<30 ng/kg-day) studies. EPA
also required that the animals were exposed via the oral route to only TCDD. Finally, the  studies
were evaluated for quality and summarized to ensure the most relevant information for
quantitative analyses was provided.  Only studies meeting all of the criteria were included in
EPA's TCDD  dose-response analysis.
                                           XXX

-------
        List of available in vivo mammalian bioassay studies on TCDD
                                Study
                           in peer-reviewed
                              literature?
                              Study on a
                           genetically-altered
                               species?
                Lowest
             dose tested for
           cancer endpoint<1
              ug/kg-day?
    Lowest dose
tested for noncancer
    endpoint<30
    ng/kg-day?
                                 Oral
                          exposure to TCDD
                                only?
                     Study summarized; evaluated for
                       quality andto note adequacy
                        of data needed for TCDD
                       dose-response assessment.
                           Key study included
                    forTCDD cancer and/or noncancer
                       dose-response assessment
                 V	/
                    Study excluded
                      from TCDD
                    dose-response
                      assessment
Figure ES-2. EPA's process to evaluate available animal bioassay studies using
study inclusion criteria for use in the dose-response analysis of TCDD.
EPA evaluated all available in vivo mammalian bioassay studies on TCDD.  Studies had to be
published in the peer-reviewed literature. Studies on genetically-altered species were excluded as
their relevance to human health is not known. Next, EPA applied dose requirements to each
study's lowest tested average daily dose, with requirements for cancer (<1 ug/kg-day) and
noncancer (<30 ng/kg-day) studies. EPA also required that the animals were exposed via the oral
route to only TCDD. Finally, the studies were evaluated for quality and summarized to ensure
providing the most relevant information for quantitative human health risk analyses. Only studies
meeting all of the criteria were selected for EPA's TCDD dose-response analysis.
                                        XXXI

-------
       Figure ES-3 shows the results of EPA's process to select and identify in vivo mammalian
bioassays and epidemiologic studies for quantitative TCDD dose-response assessment. A total
of 1,441 studies were examined. Of these, 637 studies were eliminated from consideration as
they were not suitable study types; these included, in vitro bioassays, review articles, PBPK
modeling studies, and studies that evaluated dioxin-like compounds (DLCs) other than TCDD.
Of the remaining studies, 49 were epidemiologic studies (7 studies contained both cancer and
noncancer endpoints), and 755 were animal bioassays (4 studies contained both cancer and
noncancer endpoints). These epidemiologic studies and animal bioassays were then evaluated
using EPA's study inclusion criteria.  Appendices C  and D detail EPA's study summaries and
evaluations for the epidemiologic studies and animal bioassays, respectively. Results  of the
study selection process for the epidemiologic studies are shown in Tables 2-1 and 2-2
(preliminary cancer studies and final noncancer studies, respectively) and for the animal
bioassays are shown in Tables 2-3  and 2-4 (preliminary cancer bioassays and final noncancer
bioassays, respectively). Through this study selection process, EPA was able to identify a group
of studies for TCDD dose-response evaluation that spanned the types of adverse health effects
associated with TCDD exposures and encompass the range of doses in the lower end of the
dose-response region most relevant to the development of an RfD.  The summaries of the cancer
studies are presented for use related to non-cancer effects in this document.  Quantitative
dose-response assessments will be developed for the cancer studies in the Reanalysis,  Volume 2.
                                         xxxn

-------
                Studies from literature search and data collection activities
                                       1,441
                                                              i_
Right study type for quantitative
TCDD dose-response analysis:
804 considered further
\

Wrong study type
TCDD dose-resp
637 exc

f
Epidemiologic (Epi) studies
49

3
Animal b
75
i
Epi cancer
studies
24
v
Failed >1 of:
Peer-review (0)
Primarily TDCC
(10)
Effective
exposure
estimable (11 )
Considerations**
(1)
v
( Epi "I
cancer
studies
Included
1 8 J






1
Epi noncancer
studies
32
v
Failed >1 of:
Peer-review (1)
Primarily TDCC
(7)
Effective
exposure
estimable (26)
Considerations**
(1)
v
c ~\
noncancer
studies
included
1 4 J
^
Cancer
bioassays
8
v
Failed >1 of:
Peer-review (0)
Genetically-
altered(1)
Dose cutoffs
(0)
TCDDonly(O)
Non-oral (1)
v
/" N
Animal
cancer
bioassays
Included
1 6 J











1
Noncancer
bioassays
751
\^
Failed >1 of:
Peer-review (0)
Genetically-
altered (66)
Dose cutoffs
(370)
TCDD only (142)
Non-oral (135)
v
Animal
noncancer
bioassays
included
1 78 J
  *Failed criteria are not mutually exclusive; more than one can fail for a given study.
  **lndicates those studies that passed allthree criteria butwere not selected based on
  study considerations.


Figure ES-3. Results of EPA's process to select and identify in vivo
mammalian and epidemiologic studies for use in the dose-response analysis
of TCDD.
Four animal studies and seven epidemiologic studies contained both cancer and noncancer
endpoints. Two epidemiologic cancer studies, Steenland et al. (1999) and Flesch-Janys et al.
(1998). passed all criteria, but were still not selected because they were superseded by other
studies on the same cohort for which an improved analysis was done.  One noncancer
epidemiologic study, Baccarelli et al. (2005). passed all criteria, but was excluded because the
health endpoint, chloracne, is  considered to be an outcome associated with high TCDD exposures.
                                       xxxin

-------
       For the selected studies, EPA conducted additional evaluations to determine which
study/endpoint data sets were the most appropriate for development of the RfD for TCDD.
During the study selection process, EPA identified four epidemiologic studies and 78 animal
bioassays that met the study inclusion criteria and adequately satisfied the considerations for
TCDD dose-response analyses.  From the epidemiologic studies, one was eliminated because
EPA could not assess the biological significance of the finding and could not establish a
LOAEL; EPA derived three candidate RfDs from the other studies. Figure ES-4 overviews the
disposition of the 78 noncancer animal bioassays selected for TCDD dose-response. Of these,
EPA eliminated those studies that contained no lexicologically relevant endpoints for RfD
derivation (see Appendix H and Section 4.2.1). EPA then identified PODs from the remaining
bioassays and eliminated from further analysis those studies with PODs above specified dose
limits.  (See additional details on POD development in the section below on Derivation of an
RfD for TCDD.) These dose limits were imposed to limit the size of the analysis yet ensure
representation of all important health effects associated with TCDD exposure. EPA derived
37 candidate RfDs from the remaining 48 animal studies, with  11 studies presented as supporting
information.
       In summary, EPA conducted a transparent study selection process to select epidemiologic
studies and animal bioassays for TCDD quantitative  dose-response analyses.  From these
selected studies, EPA identified 40 candidate RfDs, three from the epidemiologic studies and 37
from the animal bioassays.
                                         xxxiv

-------
                      Non cancer Animal Bioassays Selected for
              TCDD Dose-Response Assessment (See Tables 2-4 and D-1)
                                    78 Studies
       Eliminate Studies with No lexicologically Relevant Endpoints for RfD Derivation
                          (See Appendix H and Section 4.2.1)
                                16 Studies Eliminated
       Bu rleson et al. (1996)                 DeVito et al. (1994)
       Hassoun etal. (1998)                 Hassoun etal. (2000)
       Hassoun etal. (2002)                 Hassoun etal. (2003)
       Hongetal. (1989)                   Kitchin and Woods(1979)
       Latchoumycandaneetal. (2003)       Lucieretal. (1986)
       MallyandChipman (2002)           Sewall etal. (1993)
       Slezaketal. (2000)                  Sugita-Konishi etal. (2003)
       Tritscheret al. (1992)                Vanden  Heuvel etal. (1994)
            Identify and Estimate PODs from the 62 Remaining Animal Bioassays
                  for use in Noncancer Dose-Response Analysis of TCDD
                                  (SeeFigureES-6)
                            Eliminate Studies with Both a
       LOAELHED>1 ng/kg-d and a NOAELHED/BMDLHED> 0.32 ng/kg-d* (See Table 4.3)
                                14 Studies Eliminated
       Chu etal. (2001)                     Croutch etal. (2005)
       Fox etal. (1993)                     Ikedaet al. (2005)
       Maron pot etal. (1993)                Noharaetal.  (2000,2002)
       Simanainen etal. (2002, 2003, 2004a)  Smialowiczet. al. (2004)
       Smith etal. (1976)                   Weber et al. (1995)

       *Hochstein etal. (2001) is also not carried forward because of the
       lack of toxicokinetic information for estimation of an HED
                           Derive Candidate RfDs from the
                     48 Remaining Noncancer Animal Bioassays
                 Final Candidate RfDs from Noncancer Animal Bioassays
                    (11 Studies Presented as Supporting Information;
                                   See Table 4-5)
                                 37 Candidate RfDs
           \	s

Figure ES-4.  Disposition of animal noncancer bioassays selected for TCDD
dose-response analysis.
EPA evaluated each noncancer endpoint found in the 78 studies that passed the study inclusion
criteria.  From this evaluation, EPA eliminated 16 studies that contained no lexicologically
relevant endpoints for RfD derivation.  Then, as detailed in Figure 4-3, EPA selected and
identified PODs for use in deriving candidate RfDs. EPA then eliminated 13 studies based on
dose limits for the PODs' HEDs; one study was also not carried forward because of the lack of
toxicokinetic information for estimation of an HED. Of the remaining 48 studies, EPA derived
37 RfD candidates, with 11 studies presented as supporting information.

                                       xxxv

-------
USE OF KINETIC MODELING TO ESTIMATE TCDD HUMAN EXPOSURES AND
DOSES IN ANIMAL BIOASSAYS
       The NAS recommended that EPA utilize state-of-the-science approaches to finalize the
2003 draft Reassessment.  Although the NAS concurred with EPA's use of first-order body
burden models in the 2003 draft Reassessment, analyses of recent TCDD literature and
comments by experts at the Dioxin Workshop suggested that the understanding of TCDD
kinetics had increased significantly since the release of EPA's 2003 draft Reassessment.  These
advances led to the development of several pharmacokinetic models for TCDD (Emond et al.,
2006: Aylward et al.. 2005a: Emond et al.. 2005: Emond et al.. 2004) and resulted in EPA's
incorporation of TCDD pharmacokinetics in the dose-response assessment of TCDD.
       The evaluation of internal dose in exposed humans and other species is facilitated by an
understanding of pharmacokinetics (i.e., absorption, distribution, metabolism, and excretion).
TCDD pharmacokinetics are influenced by three distinctive features: (1) TCDD is highly
lipophilic, (2) TCDD is slowly metabolized, and (3) TCDD induces binding proteins in the liver.
The overall impact of these factors results in preferential storage of TCDD in adipose tissue, a
long half-life of TCDD in blood due to slow metabolism, and sequestration in liver tissue when
binding induction becomes significant. As these kinetic features control target tissue levels of
dioxin, they become important in relating toxicity in animals to possible effects in humans.
       Consideration of pharmacokinetic mechanisms is critical to the selection of the dose
metrics of relevance to dose-response modeling of TCDD. Earlier assessments for
TCDD—including the 2003 Reassessment—used  estimates of body burden as the dose metric
for extrapolation between animals and humans.  These body burden calculations used a simple
one-compartment kinetic model based on the assumption of a first-order decrease in the levels of
administered dose as a function of time. However, the assumption of a constant half-life value
for the clearance of TCDD from long-term  or chronic exposure is  not well-supported
biologically given the dose-dependent elimination observed in rodents and humans. The
dynamic disposition and redistribution of TCDD between blood, fat, and liver as a function of
time and dose is better described using biologically-based models. Additionally, these models
provide estimates for other dose metrics (e.g., serum, whole blood, or tissue levels) that are more
biologically relevant to response than body burden estimated based on an assumption of
first-order elimination over time.

                                         xxxvi

-------
       For extrapolation from rodents to humans, EPA considered the following possible dose
metrics for TCDD: administered dose, first-order body burden, lipid-adjusted serum
concentration (LASC), whole blood concentration, tissue concentration, and functional-related
metrics of relevance to the mode of action (MOA) (e.g., receptor occupancy) (see
Section 3.3.4.1). After evaluation of these dose metrics, EPA chose to use TCDD concentration
in whole blood, modeled as a function of administered dose, as the dose metric for assessing
TCDD dose response in this document. LASC is commonly used in the epidemiologic literature
as the metric of choice because TCDD is highly lipid-soluble and LASC accounts for individual
differences in the size of the serum lipid compartment. However, whole blood concentration was
chosen because of the structure of the Emond PBPK model, in which the liver and other tissue
compartments are connected to the whole blood compartment rather than to the serum
compartment; LASC is estimated only as a result of model  simulations by multiplying
whole-blood concentrations by a conversion constant. EPA used the time-weighted average
whole-blood concentration over the relevant exposure periods for all animal bioassay dosing
protocols, dividing the area under the time-course concentration curve (AUC) by the exposure
duration.  Because all of the epidemiologic studies evaluated by EPA reported TCDD exposures
as LASC rather than whole-blood concentrations, oral intakes were modeled using LASC as the
dose metric.  In most cases, the reported TCDD LASC was extrapolated both forward and
backward in time to simulate the actual exposure scenario.5
       Several biologically-based kinetic models for TCDD exist in the literature.  The more
recent pharmacokinetic models explicitly characterize the concentration-dependent elimination
of TCDD (Emond et al.. 2006: Aylward et al.. 2005a: Emond et al.. 2005: Emond et al.. 2004:
Carrier et al.,  1995a, b).  The biologically based pharmacokinetic models describing the
concentration-dependent elimination (i.e., the pharmacokinetic models of Aylward et al. (2005a)
and Emond et al. (2006: 2005)) are relevant for application to simulate the TCDD dose metrics
in humans and animals exposed via the oral route.  The rationale for considering the application
of the Aylward et al. (2005a) and Emond et al.  (2006: 2005: 2004) models was largely based on
the fact that both models reflect research results from recent peer-reviewed publications, and
both models are formulated with dose-dependent hepatic elimination consistent with the
5 For the Seveso cohort, which had a high single TCDD exposure followed by low-level background exposures
leading to a gradual decline in the internal TCDD concentrations, EPA estimated both peak and average exposures
over a defined critical exposure window (see Section 4.2.2).
                                         xxxvn

-------
physiological understanding of TCDD kinetics. Dose-response modeling based on body burden
of TCDD in adult animals and humans can be conducted with either of the models.6  The
predicted slope and body burden over a large dose range are quite comparable between the
two models (generally within a factor of two).
       Results of simulations of serum lipid concentrations or liver concentrations vary for the
two models to a larger extent (up to a factor of 7), particularly for simulations of short duration.
These differences reflect two characteristics of the Emond et al. (2006) model: first,
quasi-steady-state is not assumed in the Emond et al. (2006) model; second, the serum lipid
composition used in the model is not the same as the adipose tissue lipids. The Aylward et al.
(2005a) model does not account for differential solubility of TCDD in serum lipids and adipose
tissue lipids, nor does it account for the diffusion-limited uptake by adipose tissue. Based on this
evaluation, EPA determined that the Emond et al. (2006) provided more applicability than the
Aylward et al. (2005a) model with respect to  the ability to simulate serum lipid and tissue
concentrations during exposures  that do not lead to the onset of steady-state condition in the
exposed organism.  Of the two selected models, the pharmacokinetic model developed by
Emond et al. (2006) is more physiologically based, as compared to the Aylward et al. (2005a)
model.  The Emond et al. (2006) pharmacokinetic model simulates the blood compartment
directly in the rat, mouse, and human, but the Aylward et al. (2005a) model does not. Finally,
there are also gestational and life-time nongestational forms of the Emond et al. (2006) model,
but not for the Aylward et al (2005a) model.  As a result, in this document, EPA chose the
Emond rodent PBPK model to estimate blood TCDD concentrations based on administered
doses (see Section 3.3.4, Appendix E).
       To enhance the biological basis of the PBPK model of Emond et al. (2006), three minor
modifications were made before  its use in the computation of dose metrics for TCDD:
(1) recalculation of the volume of the "rest of the body compartment" after accounting for
volume of the liver and fat compartments; (2) calculation of the rate of TCDD excreted via urine
by multiplying the urinary clearance parameter by blood concentration in the equation instead of
by the  concentration in the rest of the body compartment; and (3) recalibration for the human
gastric nonabsorption constant to match oral bioavailability data in humans (Poiger and
6 The Aylward et al. (2005a) model cannot be used to estimate TCDD body burden when the duration of the rodent
bioassay is less than 1 month,
                                        xxxviii

-------
Schlatter, 1986) (see Section 3.3.4.4 for details).  The modified PBPK model was evaluated
against all published data used in the original model.  EPA assumed that the same blood TCDD
levels that led to effects in animals would also lead to effects in humans; therefore, the Emond
human PBPK model was used to estimate the lifetime average daily oral doses (consistent with
the chronic RfD) that would correspond to the blood TCDD concentrations estimated to have
occurred during the animal bioassays. EPA used the same Emond human PBPK model to
estimate the lifetime average daily doses that would correspond to the TCDD blood or tissue
concentrations reported in the epidemiologic studies (see Appendix F). These estimates are the
Human Equivalent Doses (HEDs) that are used to develop candidate RfDs for TCDD.
       A sensitivity analysis was performed on each of the animal and human Emond PBPK
models to determine the most sensitive variables  (see Section 3.3.4.3.2.5).  In each case, all input
variables in each model were included in the analysis; the sensitivity analysis was conducted by
varying each parameter one at a time. For the rat and mouse nongestational models and rat and
mouse gestational models for the low and high doses when variables were increased by +5%,
predicted TCDD blood concentrations were very  sensitive to the Hill coefficient (see h in
Eq. 3-20, Section 3.3.4.3.2.2).  Other influential PBPK model variables are associated with the
overall dioxin elimination/sequestration rate, including the CYP1A2 induction rates, the liver
weight, the binding capacity and affinity, and the gastric and intestinal excretion rates. For the
gestational  model dosing protocols, the Hill coefficient remains the most sensitive variable but
the elasticity decreases compared with the  nongestational  analysis. Otherwise, many of the most
sensitive variables remain those associated with elimination. Additional parameters related to
the adipose tissue blood flow and with the  adipose diffusional permeability fraction are also
relatively influential.  For the human gestational and nongestational models, additional variables
associated with the adipose compartment partition coefficient, the body weight, and the
fractional adipose tissue volume are also relatively influential variables at the RfD and POD dose
compared with the animal models. For all models, the elasticities are relatively similar across
the different doses evaluated.
       For variables which are optimized,  a sensitivity analysis which varies each parameter one
at a time may overestimate the model uncertainty associated with the variable. In this analysis,
the most sensitive variable in all the models is the Hill parameter.  The elasticity is high in part
because the Hill parameter is an exponent; thus, small changes in the value can lead to larger
                                         xxxix

-------
changes in the whole blood concentration.  The Hill coefficient (as it is used in the PBPK
models) can only be estimated with high confidence when optimized against in vivo hepatic
CYP1A2 induction data in response to TCDD exposure. This type of data is found in animal
experiments only. When this coefficient is optimized against human blood levels of TCDD, it is
influenced by other parameters describing the dose-dependent elimination mechanism of the
chemical; these data cannot be evaluated in vivo in humans.
       This analysis highlights several important research needs.  While the disposition of
TCDD following high exposures is reasonably understood and simulated in current models, the
current scientific understanding of disposition following TCDD exposures near current
background dietary intakes (likely the primary source of TCDD exposure for most of the U.S.
population) are not understood as well at present.  This uncertainty affects the  estimation of
TCDD intake rates corresponding to the lower blood TCDD levels associated with LOAELs and
NOAELs. The disposition of DLCs following exposures at background levels is similarly not
well understood.

DERIVATION OF AN RFD FOR TCDD
       The NAS specifically recommended that EPA derive an RfD for TCDD.  Through a
transparent study selection process, EPA identified key studies from both epidemiologic studies
and animal bioassays. EPA then identified PODs for RfD derivation from those key human
epidemiologic studies and animal bioassays.  Figure ES-5 (exposure-response  array) shows the
PODs for TCDD graphically in terms of human-equivalent intake (ng/kg-day). The human study
endpoints are shown at the far left of the figure and, to the right, the rodent endpoints are
arranged by the following study categories: less than 1 year, greater than 1 year, reproductive,
and developmental.
                                          xl

-------
   •9
   I1
   "3)
 C 4-i
 O C
 £ o
   re
   E
100 -
1 n



1 •
0.1 •



0.01 •




0.001 •


nnm •
Human
Studies








4

n
	





r


A
i *







	



Animal Studies
Lessthanlyear | l + year(s) | Reproductive | Developmental









4




i








4
'f




\

	



k



4



	

<



> *'

	

y




4






>

4j



4(
t
	
V
. A



i



	







t
J
L
	









	





t
4
1 ^
<
>



«
*
> j
	




i>

«
i




4
"f
1 1 1
1
_L
1
1
..!.
1
	 •_
1
1
|


1 1
	 1 	 | 	
1 1
1
, J 1
> 1
1
||

I 1
JL ^
i


1 >J
i

	


<


4


/.
i
i
<
—

,
1



>
1 1
| |
	
1 1
1 1
1 1
1
	
1
1
|


1

1
|

—







<

—







<
—

/.



•~T

<
i








i
i
i







. 4








	









	
, 4








	




Dosing Range
• LOAEL
A NOAEL
0 BMDL





— T




...



,,,,1.










	



~n



< •
\
J
	









	



T
oo
o
o
CM
"ro
'5
jccarelli
^


T
oo
o
o
CM
"ro
'5
jccarelli
m


T
8
CM
"ro
'5
aluusua
^


IY
O5
O5
"ro
'5




(')
CD
00
O5
"ro
'5
o
Q.
(0
O
Q


->
oo
o
o
CM
"ro
'5
iaiowicz
E
w

CT
CO
"ro
"QJ
O



•>
CD
1
"ro
"QJ
1



IY
O
O
CM
"ro
'5
O



IT
00
O5
"ro
'5
Canto ni



IY
o
o
CM
"ro
'5
Crofton



IY
1
"ro
'5
"ro
W



IT
o
o
CM
"ro
'5
o
c
ro
LL



IT
2
"ro
'5
(0
.Q
'o



01
(0
O5
O5
03

QJ
C
^J
D)
m
C
03

IT
8
o
CM
"(D
'5
QJ
03
LJ-



^
O5
O5
"(D
QJ
£



•5
CM
00
O5
CL
1 —




IT
O5
"(D
'5
03
.Q
'o



DC.
03
CD
8
CM
CL
1 	





IY
8
CM
"03
'5
W



^
'5
en"
oo
CO

anetal.
E
3
o
m
IY
s
o
CM
_§
03
T3
C
03
f~
0
03
-1
IT
O5
"(D
'5
>s
03
i



•>
o
°
"(D
'5
shihara



•>
CD
8
CM
"(D
QJ




^
Cvf
8
CM
"(D
0)
Cuchiiwa
J£_


^
03"
00
O
o
CM
Q
O
CM
"(D
QJ

—
"ol
IY
o
o
CM
"03
'5
Ohsako



IY
o
CM
03

'5
m



IY
o
o
CM
"ro
'5
irkowski
IU
^
^

IT
CM
O
O
CM
"(D
'5
O
'o'
X



IY
CD
O
O
CM
"03
'5
liettinen
^


ce
o
o
CM
"ro
'5
attainen
^


IT
10
O)
O)
"(D
'5
O
01
w



IT
CD
CJ^
O)
"03
'5
Schantz



IY
o
o
o
CM
"S
'5
c



IY
oo
o
o
CM
"(D
'5
X



IY
CD
8
CM
"(D
'5
ranczak
LL


ce
i
"ro
'5
^
o
w
03
Q.
W


G = guinea pig; H = human; M = mouse; Mk = monkey; R = rat


Figure ES-5. Exposure-response array for ingestion exposures to TCDD.

-------
       For each noncancer epidemiologic study that EPA selected, EPA evaluated the
dose-response information developed by the study authors to determine whether the study
provided noncancer effects and TCDD-relevant exposure data for a toxicologically-relevant
endpoint.  If such data were available, EPA identified a NOAEL or LOAEL as a POD.  Then,
EPA used the Emond human PBPK model to estimate the continuous oral daily intake
(ng/kg-day) that would lead to the relevant blood TCDD concentrations associated with the
POD. If all of this information was available, then the result was included as a POD.
       Through this process, EPA identified adverse health effects from the following
four epidemiologic studies to be considered as the basis for the RfD: Eskenazi et al.  (2002b)
(menstrual cycle effects) Alaluusua et al. (2004) (developmental—tooth development), Mocarelli
et al. (2008) (reproductive—decreased sperm concentrations and motility  [semen quality]), and
Baccarelli et al.  (2008) (developmental—increased thyroid-stimulating hormone levels in
neonates [neonatal TSH]).  All four studies are from the Seveso cohort, whose members were
exposed environmentally to high peak concentrations of TCDD as a consequence of an industrial
accident. For each of the menstrual cycle, tooth development, and  semen  quality endpoints, EPA
calculated a POD for derivation of a candidate RfD by estimating dose as  the mean of the peak
exposure (following the accident) and the average exposure over a  defined critical exposure
window for that endpoint. For neonatal TSH, EPA calculated the POD from estimates of
maternal exposure during pregnancy reported by the study authors (Baccarelli et al., (2008) (see
Section 4.2.3).  The PODs estimated for both menstrual cycle and tooth development were well
above those estimated for semen quality and neonatal TSH.
       Figures ES-4 and ES-6 together present the strategy EPA used to evaluate the
study/endpoint combinations found in the animal bioassays that met EPA's study inclusion
criteria, estimate PODs, and develop a final set of candidate RfDs for TCDD. Figure ES-4
overviews the disposition of the 78 animal noncancer studies selected for TCDD dose-response
analyses.  Of these studies,  16 were eliminated because EPA determined that they contained no
lexicologically relevant endpoints that could be used to derive a candidate RfD (see Appendix H
and Section 4.2.1).  EPA then identified PODs from the remaining bioassays; at this point,
Figure ES-4 refers to Figure ES-6, which is a flow chart of the iterative process used to estimate
PODs and compare them within and across studies to arrive at a final set of PODs from these
bioassays (see additional details below).  From this final set of PODs, Figure ES-4 shows that
                                         xlii

-------
            Study/endpointcombinationsfrom key noncanceranimal bioassayswith at
                  least one toxicologically relevant endpoint for RfD derivation
                                       Is the
                            'endpoint under consideration"
                                   toxicologically
                                     relevant?

                                    Yes
                                       No
            Determine NOAEL, LOAEL, and BMDL (if possible) human equivalentdose
            (HED) based on 1st-order body burden for each study/endpointcombination
                                                                         No
     Is the
endpoint observed
 ear the LOAEL?
                                                    Is the BMDL less
                                                    than the LOAEL?
                                  sthe endpomtless
                                  than the minimum
                                   LOAEL xlOO?
                                     Yes
                 Determine a NOAEL, LOAEL, and BMDL (if possible)for each
              study/endpointcombination, based on blood concentrations from the
                              Emond rodent PBPK model
                        Estimate a Human Equivalent Dose (HED)
              corresponding to each blood concentration NOAEL, LOAEL, or BMDL
                          usingthe Emond human PBPKmodel
                                Does kinetic modeling
                             suggest considering additional
                                ndpointsathigherdoses?
                 Include NOAEL/LOAEL/BMDL
                         as a POD
                                    Exclude endpoint
                                       as a POD
Figure ES-6. EPA's process to identify and estimate PODs from key animal
bioassays for use in noncancer dose-response analysis of TCDD.
For the studies with at least one toxicologically relevant endpoint, EPA first determined if each
endpoint was toxicologically relevant. If so, EPA determined the NOAEL, LOAEL, and BMDL
Human Equivalent Dose (HED) based on lst-order body burdens for each endpoint. Within each
study, these potential PODs were included when the endpoint was observed near the LOAEL and
if the BMDL was less than the LOAEL. Then, if the endpoint was less than the minimum LOAEL
x 100 across all studies, EPA calculated PODs based on blood concentrations from the Emond
rodent PBPK model and, for all of the PODs, HEDs were estimated using the Emond human
PBPK model. If the kinetic modeling results suggested considering additional endpoints at higher
doses, the process was repeated. Finally, the lowest group of the toxicologically relevant PODs
was selected for final use in derivation of candidate RfDs.
                                        xliii

-------
       EPA then eliminated 13 studies from further analysis because both of the following
conditions were met: human equivalent dose (HED) LOAELnED >1 ng/kg-day and
NOAELHED/BMDLHED >0.32 ng/kg-day (see Table 4-3). One additional study was also not
carried forward because of the lack of toxicokinetic information for estimation of an HED.
       Figure ES-6 summarizes the strategy employed for identifying and estimating PODs from
the 62 animal bioassays with at least one lexicologically relevant endpoint for RfD derivation.
For the noncancer endpoints within these studies, EPA first evaluated the toxicological relevance
of each endpoint, rejecting those judged not to be relevant for RfD derivation.  Next, initial
PODs based on the first-order body burden metric (see Section 3.3.4.2) and expressed as HEDs
(i.e., NOAELnED, LOAELnED, BMOLnEo) were determined for all relevant endpoints
(summarized in Table 4-3). Because there were very few NOAELs and BMDL modeling was
largely unsuccessful due to data limitations (see Section 4.2), the next stage of evaluation was
carried out using LOAELs only.  Within each study, effects not observed at the LOAEL (i.e.,
reported at higher doses) with BMDLnEDS greater than the LOAELnED were eliminated from
further analysis, as they would not be considered as candidates for the final POD on either a
BMDL or NOAEL/LOAEL basis (i.e., the POD would be higher than the PODs of other relevant
endpoints). In addition, all endpoints with LOAELnED estimates beyond a 100-fold range of the
lowest identified LOAELnED across all studies were (temporarily) eliminated from further
consideration, as they would not be POD candidates either (i.e., the POD would be higher than
the PODs of other relevant endpoints). For the remaining endpoints, EPA then determined final
potential  PODs based on TCDD whole-blood concentrations obtained from the Emond rodent
PBPK models. FtEDs were then estimated for each of these PODs using the Emond human
PBPK model.  At this point, if the PBPK modeling results suggested considering additional
endpoints at higher doses,  the process was repeated. From the final set of FtEDs, a POD was
selected7 for each study, to which appropriate uncertainty factors (UFs) were applied following
EPA guidance  (see Section 4.3.3).  The resulting candidate RfDs were then considered in the
final selection process for  the RfD.  Other endpoints occurring at slightly higher doses
representing additional effects associated with TCDD exposure (beyond the 100-fold LOAELHED
 In the standard order of consideration: BMDL, NOAEL, and LOAEL.
                                         xliv

-------
range) were evaluated, modeled, and included in the final candidate RfD array  to examine
endpoints not evaluated by studies with lower PODs. In addition, Benchmark Dose (BMD)
modeling based on administered dose was performed on all endpoints for comparison purposes.
       For BMD modeling, EPA used a 10% BMR for dichotomous data for all endpoints; no
developmental studies were identified with designs that incorporate litter effects, for which a
5% BMR would be used (U.S. EPA, 2000). For continuous endpoints in this document, EPA
used a BMR of 1 standard deviation from the control mean whenever a specific
toxicologically-relevant BMR could not be defined.  Importantly, the 2003 Reassessment defined
the EDoi as 1% of the maximal response for a given  endpoint, not as a 1% change from control.
Because RfD derivation is one goal  of this document, the noncancer modeling effort undertaken
here differs substantially from the modeling in the 2003 Reassessment. Evaluation of BMD
modeling performance, goodness-of-fit, dose-response data, and resulting BMD and BMDL
estimates included  statistical criteria as well as expert judgment of their statistical and
toxicological properties.  EPA has reported and  evaluated the BMD results using the standard
suite of goodness-of-fit measures from the benchmark dose modeling software (BMDS 2.1).
(see Appendix H and Section 4.2 for more information on the BMD modeling criteria and
results.)
       For selection of the POD to  serve as the basis of the RfD, EPA gave the epidemiologic
studies the highest  consideration because human data are preferred in the derivation of an RfD.
This preference for epidemiologic study data also is  consistent with recommendations of
panelists at the Dioxin Workshop (U.S. EPA, 2009a) (see Appendix B). Figure ES-7  arrays the
candidate RfDs from both the human and animal bioassays in units of human-equivalent intake
(mg/kg-day). The human studies included in Figure ES-7 (Baccarelli et al., 2008; Mocarelli et
al., 2008; Alaluusua et al., 2004) each evaluate a segment of the Seveso civilian  population (i.e.,
not an occupational cohort) exposed directly to TCDD released from an industrial accident. EPA
designated the (Baccarelli et al., 2008; Mocarelli et al., 2008; Alaluusua et al., 2004) studies as
coprincipal in deriving the RfD (see Section 4.3).  In the Seveso cohort, exposures were
primarily to TCDD, the chemical of concern, with apparently minimal DLC exposures beyond
those associated with background intake, qualifying  these studies for use in the RfD derivation
 However, studies with a lowest dose tested greater than 30 ng/kg-day were not included in the expanded
evaluation.
                                          xlv

-------
     1 .E-04
             Human
                                                                     Animal Bioassays
     1.E-05 ••—•
     1 .E-06
     1 .E-07
l\
£!~
3 ><
 i n
 I."*
*5
*%
5 E
0~
1 .E-08
                                                                                                         DUfH (human variability)
                                                                                                         D UFA (an imal-to-human)
                                                                                                         * Point of Departure
                                                                                                         • RfD
     1 .E-09
       1c -I n
      .t- I U "
       1r- A A
      .t-1 1  '
     1.E-12 +—
                                                                                                                                    |
00
0
0
CM
ro
CU
jccarelli
s

oo
o
o
CM
CO
cu
accarelli
m

o
o
CM
ro
cu
aluusua
<

CD
O
O
CM
ro
cu



CM
o
o
CM
CO
cu
uchiiwa
x.

00
o
o
CM
ro
cu
iaiowicz
E
w

"flU
en
m

nanetal
0
m
CO
oo
H
CM_
n
0
CM
"(0
CD
CD
v>
m

ro
(i)
_c


CM
0
O
CM

ro
08
_c
o
ro
	 i

.Q CD
CM 00
00 05
Cn T-
Q- ro

i


h- CD
05 00
en en
••" *-
ro co
'cu cu
Zi .5
Q.
CO
O
CD
Q

0
0
CM
ro
cu
W


o
o
CM
ro
cu
rkowski


CM
o
0
CM
CO
CD
o_
X


00
en
't~
ro
cu
Cantoni


CO
m

ro
cu
in


8
n
CM
CO
cu
liettinen
^

o
o
CM
ro
cu
c

-------
for TCDD. In addition, by using PODs derived from human data, the uncertainty of interspecies
extrapolation is eliminated.  The study subjects included newborns (exposed in utero) and adults
who were exposed when they were less than 10 years of age, identifying effects in potentially
vulnerable lifestages, accounting for at least some part of the uncertainty in extrapolation of
effect levels to sensitive human populations and lifestages.
       For Baccarelli et al. (20081 EPA defined the LOAEL (in LASC terms) as the maternal
TCDD LASC of 235 ppt corresponding to a neonatal TSH level of 5 uU/mL, determined by the
regression modeling performed by the study authors. The World Health Organization (1994)
established the 5 uU/mL standard as a benchmark indicator for medical follow-up for
investigation of potential congenital hypo-thyroidism. This benchmark was intended to address
potential iodine deficiencies, but it is equally applicable to TCDD exposure for evaluating the
equivalent effect.  Baccarelli et al. (2008) discounted iodine status in the population as a
confounder.  For TCDD, the toxicological concern is not likely to be iodine uptake inhibition,
but rather increased metabolism and clearance of the thyroid hormone, thyroxine (T4). An
increased TSH level is an indicator of a potential  decrease in circulating T4 levels, which could
eventually lead to  neurological deficiencies. TCDD has been associated with reductions in T4 in
a number of animal studies9 as discussed in Section 4.3.6.1.  Adequate levels of thyroid hormone
are essential in the newborn and young infant as this is a period of active brain development
(Zoeller and Rovet 2004; Glinoer and Delange, 2000).  Thyroid hormone disruption during
pregnancy and in the neonatal period can lead to irreversible neurological  deficiencies.
       Baccarelli  et al.  (2008) did not provide oral intakes associated with TCDD serum
concentrations. EPA estimated the maternal TCDD intake corresponding  to the LASC LOAEL
of 235 ppt (at delivery) by use of the Emond human PBPK model the continuous daily intake
from birth to age 30, the average age of the maternal cohort at delivery, that resulted in a 235 ppt
maternal LASC at delivery. The resulting modeled maternal daily intake rate of 0.020 ng/kg-day
established the LOAEL POD for the RfD.  EPA did not define a NOAEL because it is not clear
what maternal intake should be assigned to the group below 5 uU/mL.
       For Mocarelli et al. (2008), EPA defined the LOAEL as  the lowest exposed group
(lst-quartile) median TCDD LASC of 68 ppt, corresponding to decreased sperm concentrations
9Sewall et al. (1995), Seo et al. (1995), Van Birgelen et al. (1995a; 1995b), Crofton et al. (2005). and NTP (2006a)

                                          xlvii

-------
(25%) and decreased motile sperm counts (12%) in men who were 1-9 years old at the time of
the Seveso accident (initial TCDD exposure event).  There is no clear adverse effect level
indicating male fertility problems for either of these sperm effects. As sperm concentration
decreases, the probability of pregnancy from a single ejaculation also decreases; infertile
conditions arise when the number of normal sperm per ejaculate is consistently and sufficiently
low. Previously, the incidence of male infertility was considered increased at sperm
concentrations less than 20 million sperm/mL (WHO, 1980). More recently, Cooper et al.
(2010) suggested that the 5* percentile for sperm concentration (15 million/mL) could be used as
a limit by clinicians to indicate needed follow-up for potential infertility.  Skakkebaek (2010)
suggests the following two limits for human sperm concentrations: 15 million sperm/mL, based
on Cooper et al. (2010) and 40 million sperm/mL. Skakkebaek justifies the upper level of
40 million sperm/mL citing a study by Bonde et al. (1998) of couples planning to become
pregnant for the first time; in the Bonde study, pregnancy rates declined when sperm
concentrations were below 40 million sperm/mL.  Skakkebaek suggests that 15 million
sperm/mL may be too low of a cut off for normal fertility and that sperm concentrations between
15 million sperm/mL and 40 million sperm/mL may indicate a range of reduced fertility. For
fertile men, between 50% and 60% of sperm are motile (Swan et al., 2003; Slama et al., 2002;
Wijchman et al., 2001).  Any impacts on these reported levels could become functionally
significant, leading to reduced fertility.  Low sperm counts are typically accompanied by poor
sperm quality with respect to morphology and motility (Slama et al., 2002).
       EPA judged that the impact on sperm concentration and quality reported by Mocarelli
et al. (2008) is biologically significant given the potential for functional impairment.  Although a
decrease in sperm concentration of 25% likely would not have clinical significance for a typical
individual, EPA's concern with the reported decreases in sperm concentration and total number
of motile sperm (relative to the comparison group) is that such decreases associated with TCDD
exposures could lead to shifts in the distributions of these measures in the general population.
Because male fertility is  susceptible to reductions in both the number and quality of sperm
produced, such shifts in the population could result in decreased fertility in men at the low ends
of these population distributions.  Further, in the group exposed due to the Seveso accident,
individuals 1 standard deviation below the mean had sperm concentrations of 21.8 million/mL;
                                          xlviii

-------
this concentration falls at the low end of the range of reduced fertility (15 million and
40 million sperm/mL) suggested by Skakkebaek (2010).
       For Mocarelli et al. (2008), TCDD LASC levels were measured within approximately
1 year of the initial exposure event.  Because effects were only observed in men who were under
10 years of age at the time of exposure, EPA assumed a maximum 10-year critical exposure
window for elicitation of these effects.  Using the Emond human PBPK model, EPA has
estimated a continuous daily oral intake of 0.020 ng/kg-day associated with the (LASC) LOAEL
of 68 ppt (see Section 4.2.3.2).  The reference group is not designated as a NOAEL because there
is no clear zero-exposure measurement for any of these endpoints, particularly considering the
contribution of background exposure to DLCs, which further complicates the interpretation of
the reference group response as a true "control" response (see discussion in Section 4.4).
However, males less than 10 years old can be designated as a being in a sensitive lifestage as
compared to older males who were not affected.
       The two PODs based on the Baccarelli et al. (2008) and Mocarelli et al. (2008) studies,
are adjusted LOAELs with the same value of 0.020 ng/kg-day, providing mutual quantitative
support. Because these two studies define the most sensitive endpoints evaluated in the
epidemiologic literature, they are designated as coprincipal studies for the RfD. Increased TSH
in neonates (Baccarelli et al., 2008) and male reproductive effects (decreased sperm count and
motility) (Mocarelli et al., 2008) are designated as cocritical  effects. The adjusted LOAEL of
0.020 ng/kg-day is designated as the POD for the RfD. EPA used a composite UF of 30 for the
RfD.  A factor of 10 for UFL was applied  to account for lack of a NOAEL.  A factor of 3 (10°5)
for UFn was applied to account for human interindividual variability because the effects were
elicited in sensitive lifestages.  A UF of 1  was not applied because the sample sizes in these
two epidemiologic studies were relatively small, which, combined with uncertainty in exposure
estimation, may not fully capture the range of interindividual variability. In  addition, potential
chronic effects are not well defined for humans and could possibly be more sensitive. The
resulting RfD for TCDD in standard units is 7 x 10~10 mg/kg-day.
       Although the human data are preferred, Figure ES-7 presents a number of candidate RfDs
derived from animal bioassays that are lower than the human RfDs.  Two of the rat bioassays
among this group of studies—Bell et al. (2007b) and NTP (2006a)—are of particular note.  Both
studies were recently conducted and very  well designed and  conducted, using 30 or more
                                          xlix

-------
animals per dose group; both also are consistent with and, in part, have helped to define the
current state of practice in the field of toxicology. Bell et al. (2007b) evaluated several
reproductive and developmental endpoints, initiating TCDD exposures well before mating and
continuing through gestation. NTP (2006a) is the most comprehensive evaluation of TCDD
chronic toxicity in rodents to date, evaluating dozens of endpoints at several time points in all
major tissues.  Thus, proximity of the candidate RfDs derived from these two high quality, recent
studies, provide additional support for the RfD  derived from the two coprincipal epidemiologic
studies.
       EPA also developed cross-species comparison tables and figures of selected toxicological
endpoints for all the animal and human studies  that met the EPA selection criteria (see
Appendix D.3). The endpoints include male and female reproductive effects, thyroid hormone
levels and developmental dental effects, all of which have been reported for humans.  In
addition, immunological and neurological effects are shown because they are sensitive effects in
experimental animal studies, although not evident in humans.  The analysis presented in
Appendix D.3  supports the conclusion that there is a substantial amount of qualitative
concordance of effects between rodents and humans, but a much lower quantitative concordance.
       There are several animal bioassay candidate  RfDs at the lower end of the RfD range in
Figure ES-7 that are more than 10-fold below the human-based RfDs. Two of these studies
report effects that are analogous to the endpoints reported in the three human studies and support
the RfDs based on human data. Specifically, decreased sperm production in Latchoumydandane
and Mathur (2002) is consistent with the decreased sperm counts and other sperm effects in
Mocarelli et al. (2008). and missing molars in Keller et al.  (2008a; 2008b:  2007) are similar to
the dental defects seen in Alaluusua et al. (2004). Thus, because these endpoints have been
associated with TCDD exposures in humans, these animal  studies would not be selected for RfD
derivation in preference to human data showing similar effects.
       Another characteristic of the remaining  studies in the lower end of the candidate RfD
distribution is that they are dominated by mouse studies (comprising 7 of the 9 lowest
rodent-based RfDs). EPA has less confidence in the candidate RfD estimates based on mouse
data than either the rat or human candidate RfD estimates.  EPA has less confidence in the
Emond mouse PBPK model than the other Emond PBPK models used to estimate the PODs
because of the lack of key mouse-specific data,  particularly for the gestational component (see

-------
Section 3.3.4.3.2.5).  The LOAELnEoS identified in mouse bioassays are low primarily because
of the large toxicokinetic interspecies extrapolation factors used for mice, for which there is
more potential for error. In addition, each one of the mouse studies has other qualitative
limitations and uncertainties that make them less desirable candidates as the basis for the RfD
than the human studies.
       EPA conducted additional sensitivity analyses of two groups of studies.  Using variable
sensitivity trees, EPA further analyzed the impacts of some sources of uncertainty encountered in
the development of candidate RfDs based on Baccarelli et al. (2008), Mocarelli et al. (2008) and
NTP (2006a), specifically examining the sensitivity of the POD value to choices made for
estimating possible contributions associated with exposures to DLCs, exposure uncertainties and
PBPK model variables and inputs (see Section 4.5.1). In Section 4.5.2, EPA also evaluated a
number of endpoints presented in seven other Seveso cohort studies to estimate the range of
potential PODs based on uncertainties in exposure duration, exposure averaging protocols and
DLC background exposures. Included among those seven study/endpoint combinations are
two studies that satisfied all the study selection criteria and considerations—developmental
dental effects (Alaluusua et al., 2004) and duration of menstrual period (Eskenazi et al.,
2002b)—a new developmental study on semen quality (Mocarelli et al., 2011) that was
published after the study selection process was  completed but is useful in this uncertainty
analysis of the POD ranges, and four studies that did not satisfy all the study inclusion criteria
and considerations.10
       Overall, the results of these sensitivity analyses increase the confidence in the TCDD
RfD—both qualitatively and quantitatively. EPA's sensitivity analyses show some POD
estimates that are higher than the POD used to derive the RfD, while other analyses show POD
estimates lower than the POD used to derive the RfD. These sensitivity analyses also highlight
several important research needs.  They highlight that the current scientific understanding of
disposition following TCDD exposures that are closer to current background dietary intakes are
not understood as well as the disposition of high TCDD exposures at present. There is also
toxicological uncertainty regarding several of the endpoints; additional studies corroborating
10 Mocarelli (2000), Eskenazi et al. (2005), and Warner et al. (2007: 2004). See Appendix C for study
descriptions.
                                            li

-------
these outcomes and their toxicological significance would further increase their utility in refining
the TCDD RfD.
                                            Hi

-------
                                  1.  INTRODUCTION

       Dioxins and dioxin-like compounds (DLCs), including polychlorinated dibenzo-dioxins,
polychlorinated dibenzofurans, and polychlorinated biphenyls are structurally and
lexicologically related halogenated dicyclic aromatic hydrocarbons.11  Dioxins and DLCs are
released into the environment from several industrial sources such as chemical manufacturing,
combustion, and metal processing; from individual activities including the burning of household
waste; and from natural processes such as forest fires.  Dioxins and DLCs are widely distributed
throughout the environment and typically occur as chemical mixtures. They do not readily
degrade; therefore, levels persist in the environment, build up in the food chain, and accumulate
in the tissues of animals.  Human exposure to these compounds occurs primarily through the
ingestion of contaminated foods (Lorber et al., 2009), although exposures to other environmental
media and by other routes and pathways do occur.
       The health effects from exposures to dioxins and DLCs have been documented
extensively in  epidemiologic and toxicological studies. 2,3,7,8-Tetrachlorodibenzo:p-dioxin
(TCDD) is one of the most toxic members of this class of compounds  and has a robust
toxicological database.  Characterization of TCDD toxicity is critical to the  risk assessment of
mixtures of dioxins and DLCs because it has been selected repeatedly  as the "index chemical"
for the dioxin toxicity equivalence factors (TEF) approach.  In this approach, the toxicity of
individual components of dioxin and DLC mixtures is  scaled to that of TCDD. Then, the
dose-response information for TCDD is used by the U.S. Environmental Protection Agency
(EPA) and other organizations to evaluate risks from exposure to mixtures of DLCs (U.S. EPA,
201 Ob: Van  den Berg et al., 2006: 1998) (also see the World Health Organization's Web site for
thedioxinTEFs).12
       To provide guidance on the  use of the TEF approach in environmental health risk
assessments, EPA published a report titled, Recommended Toxicity Equivalence Factors (TEFs)
for Human Health Risk Assessments of 2,3,7,8-Tetrachlorodibenzo-p-dioxin and Dioxin-Like
Compounds (TEF report) (U.S. EPA, 201 Ob). The TEF report describes EPA's updated
approach for evaluating the human health risks from exposures to environmental media
containing DLCs. In the TEF report, EPA recommends use of the consensus TEF values for
11 For further information on the chemical structures of these compounds, see U.S. EPA (2010b. 2008b. 2003)
12 Available online at http://www.who.int/ipcs/assessment/tef_update/en/.
                                           1-1

-------
TCDD and DLCs published in 2005 by the World Health Organization (Van den Berg et al..
2006) for all cancer and noncancer effects mediated through aryl hydrocarbon receptor binding.
Further, EPA recommends that the TEF methodology, a component mixture method, be used to
evaluate human health risks posed by these mixtures, using TCDD as the index chemical.  The
TEFs are factors that scale individual DLC exposures to toxicity equivalence (TEQ)13 units of
TCDD. To assess health risks for a given exposure to a mixture of DLCs, the TEQ's of those
DLCs are summed, and the sum (i.e., total TEQ) is compared to dose-response information for
TCDD. Therefore, it is imperative to correctly assess the dose response of TCDD and
understand the uncertainties and limitations therein.
       In 2003, EPA produced an external review draft of the multiyear comprehensive
reassessment of dioxin exposure and human health effects titled, Exposure and Human Health
Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds (U.S.
EPA, 2003). This draft report, herein called the "2003 Reassessment," consisted of (1) a
scientific review of information relating to sources of and exposures to TCDD, other dioxins, and
DLCs in the environment; (2) detailed reviews of scientific information on the health effects of
TCDD, other dioxins, and DLCs; and (3) an integrated risk characterization for TCDD and
related compounds.
       In 2004, EPA asked the National Research Council of the National Academy of Sciences
(NAS) to review the 2003 Reassessment.  The NAS Statement of Task was as follows:
13 TEQ is the product of the concentration of an individual DLC in an environmental mixture and the corresponding
TCDD TEF for that compound. These products are summed to yield the TEQ of the mixture.
                                          1-2

-------
         The National Academies' National Research Council will convene an expert committee that will
         review EPA's 2003 draft reassessment of the risks of dioxins and dioxin-like compounds to
         assess whether EPA's risk estimates are scientifically robust and whether there is a clear
         delineation of all substantial uncertainties and variability. To the extent possible, the review will
         focus on EPA's modeling assumptions, including those associated with the dose-response curve
         and points of departure; dose ranges and associated likelihood estimates for identified human
         health outcomes; EPA's quantitative uncertainty analysis; EPA's selection of studies as a basis
         for its assessments; and gaps in scientific knowledge. The study will also address the following
         aspects of EPA's 2003 Reassessment: (1) the scientific evidence for classifying dioxin as a human
         carcinogen; and (2) the validity of the nonthreshold linear dose-response model and the cancer
         slope factor calculated by EPA through the use of this model.  The committee will also provide
         scientific judgment regarding the usefulness of toxicity equivalence factors (TEFs) in the risk
         assessment of complex mixtures of dioxins and the uncertainties associated with the use of TEFs.
         The committee will also review the uncertainty associated with the 2003 Reassessment's
         approach regarding the analysis of food sampling and human dietary intake data, and, therefore,
         human exposures, taking into consideration the Institute of Medicine's report Dioxin and Dioxin-
         Like Compounds in the Food Supply: Strategies to Decrease Exposure. The committee will focus
         particularly on the risk characterization section of EPA's 2003 Reassessment report and will
         endeavor to make the uncertainties in such risk assessments more fully understood by decision
         makers. The committee will review the breadth of the uncertainty and variability associated with
         risk assessment decisions and numerical choices, including, for example, modeling assumptions,
         including those associated with the dose-response curve and points of departure. The committee
         will also review quantitative uncertainty analyses, as feasible and appropriate.  The committee
         will identify gaps in scientific knowledge that are critical to understanding dioxin reassessment
         (NAS. 2006. p. 43. Box 1-1).
In 2006, the NAS published its review of EPA's 2003 Reassessment titled Health Risks from

Dioxin and Related Compounds: Evaluation of the EPA Reassessment (NAS, 2006b).


1.1.  SUMMARY OF KEY NAS (2006B) COMMENTS ON DOSE-RESPONSE
      MODELING IN THE 2003 REASSESSMENT

        While recognizing the effort that EPA expended to prepare the 2003 Reassessment, the

NAS committee identified three key areas  that they believed required  improvement to support a

scientifically robust health assessment.  These three key areas are



    •   Transparency and clarity in selection of key data sets for analysis;

    •   Justification of approaches to dose-response modeling for cancer and noncancer
        endpoints; and

    •   Transparency, thoroughness, and clarity in quantitative uncertainty analysis.
                                                1-2

-------
       In their Public Summary, the NAS made the following overall recommendations to aid

EPA in addressing their key concerns:
       EPA should identify the most important data sets to be used for quantitative risk
       assessment for each of the four key end points (cancer, immunotoxicity, reproductive
       effects, and developmental effects).  EPA should specify inclusion criteria for the studies
       (animal and human) used for derivation of the benchmark dose (BMD) for different
       noncancer effects and potentially for the development of RfD (reference dose) values and
       discuss the strengths and limitations of those key studies; describe and define
       (quantitatively to the extent possible) the variability and uncertainty for key assumptions
       used for each key end-point-specific risk assessment (choices of data set, POD [point of
       departure],14 model, and dose metric); incorporate probabilistic models to the extent
       possible to represent the range of plausible values; and assess goodness-of-fit of
       dose-response models for data sets and provide both upper and lower bounds on central
       estimates for all statistical estimates.  When quantitation is not possible, EPA should
       clearly state it and explain what would be required to achieve quantitation (NAS, 2006b,
       IL9).

       EPA should continue to use body burden as the preferred dose metric but should also
       consider physiologically based pharmacokinetic modeling as a means to adjust for
       differences in body fat composition and for other differences between rodents and
       humans (NAS. 2006b. p. 9).

       When selecting a BMD as a POD, EPA should provide justification for selecting a
       response level (e.g., at the 10%, 5%, or 1% level). In either case, the effects of this
       choice on the final risk assessment values should be illustrated by comparing point
       estimates and lower bounds derived from selected PODs (NAS, 2006b, p. 9).

       EPA should compare cancer risks by using nonlinear models consistent with a receptor
       mediated mechanism of action and by using epidemiologic data and the new National
       Toxicology Program (NTP) animal bioassay data (NTP, 2006a).  The  comparison should
       include upper and lower bounds, as well as central estimates of risk. EPA should clearly
       communicate this information as part of its risk characterization (NAS, 2006b, p. 9).

       Although EPA addressed many sources of variability and uncertainty qualitatively, the
       committee noted that the 2003 Reassessment would be substantially improved if its risk
       characterization included more quantitative approaches.  Failure to characterize
       variability and uncertainty thoroughly can convey a false sense of precision in the
       conclusions of the risk assessment (NAS, 2006b, p. 5).
14 Point of departure: The dose-response point that marks the beginning of a low-dose extrapolation. This point can
be the lower bound on dose for an estimated incidence or a change in response level from a dose-response model
(BMD), or a NOAEL (no-observed-adverse-effect-level) or LOAEL (lowest-observed-adverse-effect-level) for an
observed incidence, or change in level of response (available online at http://www.epa.gov/iris/help gloss.htm#p).
                                            1-4

-------
       Importantly, the NAS encouraged EPA to calculate an RfD as the 2003 Reassessment
does not contain an RfD derivation.  The committee suggested that:
       ... estimating an RfD would provide useful guidance to risk managers to help
       them (1) assess potential health risks in that portion of the population with intakes
       above the RfD, (2) assess risks to population subgroups, such as those with
       occupational exposures, and (3) estimate the contributions to risk from the major
       food sources and other environmental sources of TCDD, other dioxins, and DLCs
       for those individuals with high intakes (NAS, 2006b, p. 6).
       The NAS made many other thoughtful and specific recommendations throughout their

review; additional NAS recommendations and comments pertaining to the dose-response

assessment of TCDD will be presented and addressed in various sections throughout this

document.


1.2.  EPA'S SCIENCE PLAN
       In May 2009, EPA Administrator Lisa P. Jackson announced the "Science Plan for

Activities Related to Dioxins in the Environment" ("Science Plan") that addressed the need to

finish EPA's dioxin reassessment and provide a completed health assessment on this high profile

chemical to the American public.15

       The Science Plan outlined EPA's interim milestones for addressing several issues related

to dioxins and DLCs. With regard to EPA's response to the NAS comments on the 2003 Dioxin

Reassessment, the Science Plan stated  the following:
    1.  EPA will release a draft report that responds to the recommendations and comments
       included in the NAS 2006 review of EPA's 2003 Dioxin Reassessment.

          a.  EPA's National Center for Environment Assessment (NCEA) in the Office of
             Research and Development, will prepare a limited response to key comments and
             recommendations in the NAS report.
Research and Development, will pre
recommendations in the NAS report
          b. The draft response will focus on dose-response issues raised by the NAS and will
             include an analysis of relevant new key studies.
15 Available at http://www.epa.gov/dioxin/scienceplan.
                                          1-5

-------
   2.  EPA will provide the draft response to comments report for internal and external review.
          a.  The draft response to comments report will also undergo both internal EPA
             review and interagency review.
          b.  The draft response will be provided for public review and comment and
             independent external peer review.
   3.  The EPA Science Advisory Board (SAB) will review the science content of the response
       to comments report.
       As outlined in the Science Plan, in 2009, EPA developed a draft report titled EPA's
Reanalysis of Key Issues Related to Dioxin Toxicity and Response to NAS Comments (draft
Reanalysis) that responded to the key comments and recommendations in the NAS report (U.S.
EPA, 2010a). The draft Reanalysis focused on TCDD dose-response issues and included
analyses of relevant new studies and the derivation of an oral RfD. The draft Reanalysis was
reviewed internally by EPA scientists and externally by other federal agencies and White House
Offices. On May 21, 2010, the draft Reanalysis was released for public review and comment
and independent external peer review by EPA's SAB.

1.3.  SAB (SCIENCE ADVISORY BOARD) REVIEW OF EPA'S DRAFT REANALYSIS
       For their review, the SAB convened an expert panel composed of scientists
knowledgeable about technical issues related to dioxins and risk assessment. The SAB held
public meetings in June,  July, and October 2010 and March and June 2011. They released their
final report reviewing the draft Reanalysis on August 26, 2011 (SAB, 2011).16  In their report,
the SAB made the following overarching observations:
   •   They found that the draft Reanalysis was clear, logical and responsive to many, but not
       all, of the NAS recommendations; they were impressed with the comprehensive and
       rigorous study selection process that was used to identify, review and evaluate the
       scientific literature on TCDD dose response;
          o   .. .the SAB finds that the Report is generally clear, logical, and responsive to
              many but not all of the recommendations of the NAS. The SAB has, however,
16 Available online at
http://vosemite.epa.gov/sab/sabproduct.nsf/2A45B492EB AA8553852578F9003ECBC5/$File/SAB-ll-014-
unsigned.pdf.
                                          1-6

-------
             provided many recommendations to further improve the clarity, organization, and
             responsiveness of various parts of the Report.  The SAB was impressed with the
             process that EPA used to identify, review, and evaluate the relevant literature.
             The SAB finds that EPA's process was comprehensive and rigorous and included
             public participation. (SAB, 2011, p. 1)

   •   They agreed with the choice of the Emond physiologically based pharmacokinetic
       (PBPK) model for dose metric calculations and with whole blood as the appropriate dose
       metric;

          o  The SAB agrees with EPA's use of blood TCDD concentration as a surrogate for
             tissue exposure to TCDD. Blood TCDD concentration is a better choice than
             using body burden (as in the 2003 Reassessment) because it is more closely
             related to the biologically relevant dose metric: the free concentration of dioxin in
             the target tissues.  It is important to recognize, however, that TCDD distribution
             within tissues such as the liver can be nonuniform.  The SAB further agrees that
             the PBPK model developed by Emond et al. (2006;  2005; 2004) provides the best
             available basis for the  dose metric calculations in the assessment. (SAB, 2011, p.
             2)

   •   They agreed with the choice of two epidemiologic studies as co-critical studies whose
       developmental toxicity data were used to derive the RfD for TCDD;

          o  The SAB supports EPA's selection of the Mocarelli et al. (2008) and Baccarelli
             et al. (2008) studies for identifying "cocritical" effects for the derivation of the
             RfD.  These two human epidemiologic  studies are well designed and provide
             sufficient exposure information, including biological concentrations that could be
             used to establish acceptable lifetime daily exposure  levels. (SAB, 2011, p. 3)

   •   They agreed with EPA's evaluation of TCDD carcinogen!city (with the exception of
       one panelist with a dissenting view);

          o  The SAB agrees with EPA's conclusion that TCDD is "Carcinogenic to
             Humans:' (SAB. 2011. p. 5).
       The SAB also noted two deficiencies in EPA's draft Reanalysis with respect to the

completeness of the consideration of two critical elements:



   •   Nonlinear dose response for TCDD carcinogenicity, and

   •   Uncertainty analysis
                                          1-7

-------
       The SAB recommended that EPA fully evaluate both linear and nonlinear dose-response
approaches to TCDD cancer dose-response assessment, including a discussion of carcinogenic
mode of action. The SAB also recommended a number of approaches to quantitative uncertainty
analysis that could be implemented by EPA, including the use of sensitivity analyses and
probability trees.
   •   The SAB finds that the Report did not respond adequately to the NAS recommendation to
       adopt "both linear and nonlinear methods of risk characterization to account for the
       uncertainty of dose-response relationship shape below the ED0i (effective does
       eliciting x percent response)." EPA should present both linear and nonlinear risk
       assessment approaches. In the absence of a definitive nonlinear mode of action, the
       linear option results can serve as the baseline for comparison with other estimates. (SAB,
       2011. p. 6)

   •   .. .the SAB does not agree with EPA's argument that conducting a unified quantitative
       uncertainty analysis for TCDD toxicity is unfeasible	EPA argues that a complete
       quantitative uncertainty analysis would require data and resources not available. The
       SAB disagrees with this logic. While EPA may lack an adequate empirical basis for full
       Monte-Carlo propagation of input distributions, there are other options available. More
       limited evaluations can, and should, be implemented to inform critical issues in the
       dioxin reassessment. (SAB, 2011, p. 7)
The SAB made many additional thoughtful comments and specific recommendations throughout

their review pertaining to the dose-response assessment of TCDD (SAB, 2011).


1.4.  SCOPE OF EPA'S REANALYSIS VOLUMES 1 AND 2
       In August 2011, EPA announced a plan for moving forward to complete the draft

Reanalysis.17  Per this plan, the current document comprises the first of two EPA reports
(U.S. EPA 's Reanalysis of Key Issues Related to Dioxin Toxicity and Response to NAS

Comments Volumes 1 and 2 [Reanalysis Volumes 1 and 2J) that together will respond to the

recommendations and comments on TCDD dose-response assessment included in the NAS

review of EPA's 2003 draft Reassessment. Both Volumes focus on TCDD only. This report,

Reanalysis Volume 1, completes and publishes EPA's study selection criteria and results for both

noncancer and cancer TCDD dose-response assessment; choice of kinetic model; noncancer RfD
17 Available online at http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=209690.
                                          1-8

-------
for TCDD; and a qualitative discussion of uncertainties in the RfD with a focused quantitative
uncertainty analysis.
       These information and analyses have undergone revisions in response to SAB comments
and recommendations (see Appendix A). Reanalysis Volume 2 will address the two deficiencies
identified by the SAB, i.e., nonlinear dose response for TCDD carcinogenicity and quantitative
uncertainty analysis. In Volume 2, EPA will complete the evaluation of cancer mode-of-action,
cancer dose-response modeling, including justification of the approaches used for dose-response
modeling of the cancer endpoints, and an associated quantitative uncertainty analysis. The
information provided in Volume 1 will be used in three ways: (1) as the first of two reports that
contain EPA's response to the NAS (2006b) report, (2) as the Support Document for the TCDD
noncancer Integrated Risk Information Systems (IRIS) Summary and TCDD oral RfD, and (3) as
technical support for Reanalysis Volume 2.  The summaries of the cancer studies included in
Volume 1  are presented for use related to noncancer effects.  These summaries are not intended
to inform regulatory or other decision-making purposes related to carcinogenesis; further, no
quantitative dose-response assessments are developed for cancer studies in Volume 1.

1.5.  OVERVIEW OF EPA'S RESPONSE TO NAS (2006B)
       In their key recommendations, the NAS commented that EPA should thoroughly justify
and communicate approaches to dose-response modeling, increase transparency in the selection
of key  data sets, and improve the communication of uncertainty (particularly quantitative
uncertainty).  They also encouraged EPA to calculate an RfD. These main areas of improvement
refer to issues specifically related to TCDD dose-response assessment (and uncertainty analysis);
therefore, as noted in the Science Plan, EPA's response to the NAS is particularly focused on
these issues.
       EPA thoroughly considered the recommendations of the NAS and, in Reanalysis
Volume 1, responds with scientific and technical evaluation of TCDD dose-response data via the
following:
       An updated literature search that identified new TCDD dose-response studies (see
       Section IIAppendix I);
                                          1-9

-------
       A workshop that included the participation of external experts in TCDD health effects,
       toxicokinetics, dose-response assessment and quantitative uncertainty analysis; these
       experts discussed potential approaches to TCDD dose-response assessment and
       considerations for EPA's response to NAS (U.S. EPA, 2009a) (see Appendix B);

       Detailed study inclusion criteria and processes for the selection of key studies (see
       Section 2.3) and epidemiologic and animal bioassay data for quantitative TCDD
       dose-response assessment (see Section 2.4.1/Appendix C and Section 2.4.27Appendix D
       respectively);

       Kinetic modeling that quantifies appropriate dose metrics for use in TCDD dose-response
       assessment (see Section 3 and Appendices E and F);

       Sensitivity analyses that were performed on each  of the animal and human Emond PBPK
       models that identify the most sensitive variables in each model (see Section 3.3.4);

       Dose-response modeling for all appropriate noncancer data sets (see
       Section 4.27Appendix G);

       Thorough and transparent evaluation of the selected TCDD data for use in the derivation
       of an RfD, including justification of approaches used for dose-response modeling of
       noncancer endpoints (see Section 4.2 and Appendix H);

       The development of an RfD (see Section 4.3);

       A qualitative discussion of the uncertainty in the RfD and a focused quantitative
       uncertainty analyses of the RfD (see Sections 4.4  and 4.5, respectively); and

       Responses to the comments and recommendations made by the SAB in their final report
       (SAB. 2011) (see Appendix A).
       Each of those activities is described in detail in subsequent sections of this document.

The majority of the risk assessment terms used in this document are typically used in IRIS

documents.  Definitions can be located by referring to the IRIS online glossary, available at

http://epa.gov/iris/help_gloss.htm.  In addition to this document, it should be noted that several

additional EPA activities address other TCDD issues, specifically related to the application of

dioxin TEFs and to TCDD and DLC background exposure levels. Information on the application

of the dioxin TEFs is published elsewhere by EPA for both ecological (U.S. EPA, 2008b) and

human health risk assessment (U.S. EPA, 201 Ob).  As a consequence, EPA does not directly

address TEFs herein, but makes use of the concept of toxicity equivalence as applicable to the

analysis of exposure dose in epidemiologic studies. Furthermore, this document does not

                                          1-10

-------
address the NAS recommendations pertaining to the assessment of human exposures to TCDD
and other dioxins. Information on updated background levels of dioxin in the U.S. population
has been recently reported (Lorber et al., 2009).  In 2006, EPA also released a report titled An
Inventory of Sources and Environmental Releases of Dioxin-Like Compounds in the United
States for the Years 1987, 1995 and 2000, which presents an evaluation of sources and emissions
of dioxins,  dibenzofurans, and coplanar polychlorinated biphenyls (PCBs) to the air, land and
water of the United States (U.S. EPA. 2006b).

1.5.1.  TCDD Literature Update
       EPA has developed a literature database of peer-reviewed studies on TCDD toxicity,
including in vivo mammalian dose-response studies and epidemiologic studies for use in
quantitative TCDD dose-response assessment and supporting qualitative discussions. An initial
literature search for studies published since the 2003 Reassessment was conducted by the
U.S. Department of Energy's Argonne National Laboratory (ANL) through an Interagency
Agreement with EPA. ANL used the online National Library of Medicine database  (PubMed)
and identified studies published between the year 2000 and October 31, 2008 (see Appendix I).
Supporting references published since the release of the 2003 Reassessment were  also identified.
Supporting studies were classified as studies pertaining to TCDD kinetics, TCDD
mode-of-action, in vitro TCDD studies, and TCDD risk assessment approaches. The literature
search  strategy explicitly excluded studies addressing: (1) analytical/detection data and cellular
screening assays; (2) environmental fate, transport and concentration data; (3) dioxin-like
compounds and toxic equivalents; (4) nonmammalian dose-response data; (5) human exposure
analyses only, including body burden data; and (6) combustor or incinerator or other
facility-related assessments absent primary dose-response data.
       EPA published the initial literature  search results in the Federal Register on
November 24, 2008 (73 FR 70999; November 24, 2008) and invited the public to  review the list
and submit additional peer-reviewed in vivo mammalian dose-response studies for TCDD,
including epidemiologic studies that were absent from the list (U.S. EPA, 2008a).  Submissions
were accepted by the EPA through an electronic docket, email, and hand delivery, and they were
evaluated for use in TCDD dose-response assessment.  The literature search results and
subsequent submissions were used during a 2009 scientific workshop, which was  open to the

                                          1-11

-------
public and featured a panel of experts on TCDD toxicity and dose-response modeling (discussed
below).  Additional studies identified during the workshop, and those collected by EPA scientists
during the development of this report through October 2009, have been incorporated into the
final set of studies for TCDD quantitative dose-response assessment.
       Since release of the draft Reanalysis for public comment and external peer review in
2010, EPA has collected a limited number of additional studies published since October 2009
that also inform EPA's derivation of an RfD for TCDD. These studies were identified by EPA
scientists, the SAB, and the public, and they have been used to further evaluate the biological
significance of the endpoints used to derive the RfD and to develop information on uncertainty in
the RfD.  These additional studies are cited in the appropriate sections of this document. None
of the data sets collected since October 2009 was used quantitatively in the noncancer dose-
response assessment of TCDD.

1.5.2.  EPA's 2009 Workshop on TCDD Dose Response
       To assist EPA in  responding to the NAS, EPA and ANL convened a scientific workshop
(the "Dioxin Workshop") on February 18-20, 2009, in Cincinnati, OH.  The goals of the Dioxin
Workshop were to identify and address issues related to the dose-response assessment of TCDD
and to ensure that EPA's response to the NAS focused on the key issues and reflected the most
meaningful science. The Dioxin Workshop included seven scientific sessions: quantitative
dose-response modeling  issues, immunotoxicity, neurotoxicity and nonreproductive endocrine
effects, cardiovascular toxicity and hepatotoxicity, cancer, reproductive and developmental
toxicity, and quantitative uncertainty analysis of dose  response. During each session, EPA asked
a panel of expert scientists to perform the following tasks:
   •   Identify and discuss the technical challenges involved in addressing the NAS comments
       related to the dose-response issues within each specific session topic and the TCDD
       quantitative dose-response assessment.
   •   Discuss approaches for addressing the key NAS recommendations.
   •   Identify important published, independently peer-reviewed literature—particularly
       studies describing epidemiologic studies and in vivo mammalian bioassays expected to
       be most useful for informing EPA's response.
                                          1-12

-------
       The sessions were followed by open comment periods during which members of the

audience were invited to address the expert panels.  The session's Panel Cochairs were asked to

summarize and present the results of the panel discussions—including the open comment

periods. The summaries were intended to reflect the core of the panel discussions and

incorporated points of agreement as well as minority opinions. Final session summaries were

prepared by the session Panel Co-chairs with input from the panelists, and they formed the basis

of a final workshop report (U.S. EPA, 2009a) (Appendix B of this report). Because the sessions

were not designed to achieve consensus among the panelists, the summaries do not necessarily

represent the opinions of all the scientists that attended the meeting.  Some of the key discussion

points from the workshop that influenced EPA's development of this document are listed below

(see Appendix B for detail):
   •   In the development of study selection criteria, more relevant exposure-level decision
       points using tissue concentrations could be defined.

   •   A linear approach to body-burden estimation, which was utilized in the 2003
       Reassessment (U.S. EPA, 2003), does not fully consider key toxicokinetic issues related
       to TCDD—e.g., sequestration in the liver and fat, age-dependent elimination, and
       changing elimination rates over time. Thus, kinetic/mechanistic modeling could be used
       to quantify tissue-based metrics. In considering human data, lipid-adjusted serum levels
       may be preferable over body burden, although the assumptions used in the back
       calculation of the body burden in epidemiologic cohorts are of concern.  In considering
       rat bioassay data, lipid-adjusted body-burden estimates may be preferable.

   •   New epidemiologic studies on noncancer endpoints have been published since the
       2003 Reassessment that may need to be considered (e.g., thyroid dysfunction literature
       from Wang et al. (2005) and Baccarelli et al. (2008).

   •   The 1% of maximal response (EDoi) that was utilized in the 2003 Reassessment has not
       typically been used in dose-response assessment.  Some alternative ideas were as follows:
       (1) the POD should depend on the specific endpoint;  (2) for continuous measures, the
       benchmark response (BMR) could be based on the difference from control and consider
       the adversity level; and (3) for incidence data, the BMR should be set to a fixed-risk
       level.

   •   The quantitative dose-response modeling for cancer could be based on human or animal
       data. There are new publications in the literature for  four epidemiological cohort studies
       (Dutch  cohort, NIOSH (National Institue for Occupational Saftey and Health) cohort,
       BASF accident cohort, and Hamburg cohort).  The increase in total cancers could be
       considered for modeling human cancer data.  However, non-Hodgkin lymphoma and

                                          1-13

-------
       lung tumors are the main TCDD-related cancer types seen from human exposure. In
       reviewing the rat data, the NTP (2006a) data sets are new and can be modeled.  Although
       the liver and lungs are the main target organs, modeling all cancers,  as well as using
       tumor incidence in lieu of individual rats as a measure, should be considered.

       Both linear and nonlinear model functions should be considered in the cancer
       dose-response analysis because there are data and rationales to support use of either
       below the POD.

       For quantitative uncertainty analysis, consider the impacts of choices among plausible
       alternative data sets, dose metrics, models, and other more qualitative choices.  Issues to
       consider include how much difference these choices make and, also, how much relative
       credence should be put toward each alternative as a means to gauge  and describe the
       landscape of imperfect knowledge with respect to possibilities for the true dose response.
       This may be difficult to do quantitatively because the factors are not readily expressed as
       statistical distributions.  However, the rationale for accepting or questioning each
       alternative in terms of the available supporting evidence, contrary evidence, and needed
       assumptions, can be delineated.
1.5.3.  Organization of EPA's Response to NAS Recommendations (Reanalysis Volume 1)
       The remainder of this document, Reanalysis Volume 1, is divided into three sections that

address the three primary areas of concern resulting from the NAS (2006b) review. Section 2

describes EPA's approach to the recommendation for transparency and clarity during selection of

key data sets suitable for TCDD dose-response assessment—including criteria for the selection

of key dose-response studies and results of the evaluations of the important epidemiologic

studies and animal bioassays (Appendices C and D contain study summaries and additional

details on study evaluations for the epidemiologic and animal bioassays, respectively).

Sections 3 and 4 present EPA's response to the NAS recommendation to better justify the

approaches used in dose-response modeling of TCDD for noncancer endpoints.  Section 3

discusses the toxicokinetic modeling EPA conducted to support the dose-response analyses.
Section 4 presents EPA's noncancer data set selection, the noncancer dose-response modeling

results, the RfD derivation for TCDD, a qualitative discussion of the uncertainties associated

with the RfD, and a focused quantitative uncertainty analysis of the PODs considered for RfD
derivation.
                                          1-14

-------
   2.  TRANSPARENCY AND CLARITY IN THE SELECTION OF KEY DATA SETS
                            FOR DOSE-RESPONSE ANALYSIS
       This section addresses transparency and clarity in the study selection process and
identifies key data sets for TCDD dose-response analysis. Section 2.1 summarizes the NAS
committee's comments specifically regarding this issue.  Section 2.2 presents EPA's response to
those comments and describes EPA's approach to ensuring transparency and clarity in the
selection of studies for subsequent dose-response analyses.  Section 2.3 describes the
TCDD-specific study inclusion criteria and study quality evaluation process EPA used in this
document for determining the eligibility of both epidemiologic and experimental animal studies
for TCDD dose-response analysis.  Section 2.4 summarizes the results of applying the study
inclusion criteria to the epidemiologic studies (see Section 2.4.1, Tables 2-1 and 2-2) and the in
vivo mammalian bioassays (see Section 2.4.2, Tables 2-3  and 2-4).  These results present the key
TCDD epidemiologic and animal bioassays that were identified using the study inclusion
criteria. Additional details on this process can be found in Appendices C and D.  Appendix C
summarizes all of the available epidemiologic studies, evaluates the suitability of these studies
for TCDD dose-response analyses, and presents the study selection process results. Appendix D
summarizes only the animal bioassay data that have met the study inclusion criteria for TCDD
dose-response assessment and,  in Tables D-l  and D-2, shows the results of the study selection
process for all of the animal bioassays identified by EPA.  Study/endpoint combination data sets
for developing TCDD toxicity values for noncancer effects are further evaluated in Section 4 of
this document.  Based on the cancer studies identified in this document, study/endpoint
combination data sets for developing toxicity  values for cancer effects will be explored in a
separate document, Volume 2 of this effort. The summaries and study evaluations for the cancer
studies presented in this section and in Appendices C and  D for epidemiologic studies and
animal bioassays, respectively,  are presented for use related to noncancer effects.  These
summaries are not intended to inform regulatory or other decision-making purposes related to
carcinogenesis; further, no quantitative dose-response assessments are developed for cancer
studies in Volume 1.
                                           2-1

-------
2.1.  SUMMARY OF NAS COMMENTS ON TRANSPARENCY AND CLARITY IN
     THE SELECTION OF KEY DATA SETS FOR DOSE-RESPONSE ANALYSIS
       The NAS committee proposed that EPA develop a clear and readily understandable

methodology for evaluating and including epidemiologic and animal bioassay data sets in

dose-response evaluations.  The NAS committee recommended the development and application

of transparent initial criteria to judge whether or not specific epidemiologic or animal bioassay

studies be included in TCDD dose-response analysis.
       Specific NAS comments on the topic of study evaluation and inclusion criteria include
the following:
       EPA should specify inclusion criteria for the studies (animal and human) used for
       derivation of the benchmark dose (BMD) for different noncancer effects and
       potentially for the development of RfD values and discuss the strengths and
       limitations of those key studies (NAS. 2006b. p. 27).
       ...in its [EPA's] evaluation of the epidemiological literature of carcinogenicity, it
       did not outline eligibility requirements or otherwise provide the criteria used to
       assess the methodological quality of other included studies (NAS, 2006b, p. 56).
       With regard to EPA's review of the animal bioassay data, the committee
       recommends that EPA establish clear criteria for the inclusion of different data
       sets (NAS. 2006b.  p. 191).
       .. .the committee expects that EPA could substantially improve its assessment
       process if it more rigorously evaluated the quality of each study in the database
       (NAS. 2006b. p. 56).
       EPA could also substantially improve the clarity and presentation of the risk
       assessment process for TCDD.. .by using a summary table or a simple summary
       graphical representation of the key data sets and assumptions... (NAS, 2006b, p.
       56).
2.2.  EPA'S RESPONSE TO NAS COMMENTS ON TRANSPARENCY AND CLARITY
     IN THE SELECTION OF KEY DATA SETS FOR DOSE-RESPONSE ANALYSIS
       EPA agrees with the NAS committee regarding the need for a transparent and clear

process with criteria identified for selecting studies and key data sets for TCDD dose-response

analyses. The delineation of the study selection process and decisions regarding key data sets

will facilitate communication regarding critical decisions made in the TCDD dose-response

assessment.  In keeping with the NAS committee's recommendation to use a transparent process

and improve clarity and presentation of the health assessment process for TCDD, Figure 2-1

                                          2-2

-------
                      Literature search for in vivo mammalian bioassays and
                           epidemiologic TCDD studies (2000-2008)
                       Federal Register Notice; Web publication of literature
                           search for public comment and submissions
                           Initial TCDD-specific study inclusion criteria
                         development for in vivo mammalian bioassays
                        Dioxin workshop (2009) and expert refinement of
                   TCDD study inclusion criteria for in vivo mammalian bioassays
                  Final development of two sets of TCDD study inclusion criteria:
                      For in vivo mammalian bioassays
                      For epidemiologic studies
                            Final literature collection (October, 2009)
                  Studies screened using TCDD study inclusion criteria:
                  •   Studies cited in 2003 Reassessment
                  •   Studies identified via literature search results
                  •   Studies submitted by the public
                  •   Studies collected by EPA in 2009
                                    Yes
                   Studies Included in final list of
                    key cancer and noncancer
                   studies for quantitative dose-
                    response analysis of TCDD
Studies excluded
from quantitative
 dose-response
analysis of TCDD
Figure 2-1. EPA's process to select and identify in vivo mammalian and
epidemiologic studies for use in the dose-response analysis of TCDD.
EPA first conducted a literature search to identify studies published since the 2003 Reassessment.
Results were published, and additional study submissions were accepted from the public.  Next,
EPA developed TCDD-specific study inclusion criteria for in vivo mammalian studies and held a
Dioxin Workshop where these criteria were discussed and refined. Third, EPA developed
two final sets of study inclusion criteria, one for in vivo mammalian studies and another for
epidemiologic studies. Finally, EPA applied these two sets of criteria to all studies from the
literature search, public submissions, 2003 Reassessment, and additional studies identified by EPA
after the Dioxin Workshop through October 2009. The studies that met these criteria formed a list
of key studies for EPA's consideration in TCDD dose-response assessment.

-------
provides an overview of the approach that EPA has used in this document to develop a final list

of key cancer and noncancer studies for quantitative dose-response analysis of TCDD. The steps

in Figure 2-1 are further explained below.
       Literature search for in vivo mammalian and epidemiologic TCDD studies
       (2000-2008): EPA conducted a literature search to identify peer-reviewed, dose-response
       studies for TCDD that have been published since the 2003 Reassessment. This search
       included in vivo mammalian and epidemiologic studies of TCDD from 2000 to 2008.
       Additional details describing the conduct of this literature search are presented in
       Section 1.5.1 of this document.

       Federal Register Notice—Web publication of literature search for public comment:
       In November 2008, EPA published a list of citations from results of this literature search
       (U.S. EPA, 2008a) and invited the public to review this preliminary list of dose-response
       citations for use in TCDD dose-response assessment. EPA requested that interested
       parties identify and submit peer-reviewed studies for TCDD that were absent from this
       list. Two parties identified additional references that were not included in the 2008
       Federal Register notice and submitted additional references for EPA to consider. These
       references were included in the final TCDD literature database considered by EPA for
       TCDD dose-response analysis.

       Initial study inclusion criteria development for TCDD in vivo mammalian
       bioassays: EPA developed an initial set of draft criteria for evaluating the extensive
       TCDD database of in vivo mammalian bioassays.  These initial study inclusion criteria
       had three purposes. First, they provided a method to transparently and rigorously
       evaluate the scientific quality of each study in EPA's database, a deficiency in the 2003
       Reassessment identified by the NAS committee.  Second, their application provided an
       efficient way to initially screen the vast number of TCDD mammalian bioassays for
       consideration in TCDD dose-response analyses. Third, they served as a starting point for
       discussions of study inclusion criteria by expert panelists who were convened by EPA for
       its  scientific workshop on TCDD dose-response analysis (the Dioxin Workshop),
       described next  [also see the workshop report in Appendix B, U.S. EPA (2009a)].

       Dioxin Workshop and expert refinement of TCDD in vivo mammalian study
       inclusion criteria: In February 2009, EPA convened "A Scientific Workshop to Inform
       EPA's Response to NAS Comments on the Health Effects of Dioxin in EPA's 2003
       Dioxin Reassessment" [see workshop details in Section 1.5.2 and Appendix B (U.S.
       EPA, 2009a)].  At the workshop, EPA presented the draft set of study inclusion criteria;
       the workshop panelists evaluated the study inclusion criteria in relation to the various
       toxic endpoints that were discussed and made recommendations for their revision.

       Final development of study inclusion criteria for TCDD in vivo mammalian studies:
       Based on discussions and recommendations made at the Dioxin Workshop, the initial
                                          2-4

-------
       draft study inclusion criteria for evaluating the TCDD mammalian bioassay literature
       were revised and are presented in Section 2.3.2.

       Development of study inclusion criteria for epidemiologic studies: Following the
       Dioxin Workshop, EPA determined that an evaluation process was also needed for
       selection of epidemiologic studies for TCDD dose-response assessment.  These criteria
       were developed and are detailed in Section 2.3.1.

       Final literature collection (October 2009): Additional literature was collected as it was
       identified by EPA following the Dioxin Workshop through October 2009 to ensure the
       consideration of all recently published data for this report.

       Studies screened using study inclusion criteria: The two sets of TCDD-specific study
       inclusion criteria for epidemiologic studies and in vivo animal bioassays  presented in
       Sections 2.3.1 and 2.3.2, respectively, were used to evaluate all studies included in the
       2003 Reassessment, studies identified in the 2000-2008 literature  search, studies
       identified through public comment and submission, and studies collected in 2009 as
       identified by EPA during the development of this document.  Section 2.4 and
       Appendices C and D present results of EPA's evaluation of epidemiologic and
       mammalian bioassay literature for both cancer and noncancer endpoints.

       Final list of key noncancer studies and preliminary list  of cancer studies for
       quantitative dose-response analysis of TCDD: Application of the study inclusion
       criteria concludes in Section 2.4 with development of a final list of key noncancer studies
       and a preliminary list of cancer studies to be considered for quantitative dose-response
       analyses of TCDD. In Section 4, PODs are developed and evaluated for  all biologically
       relevant noncancer study/endpoint combinations from the final key noncancer study lists,
       and key data sets and PODs for the development of TCDD noncancer toxicity values are
       identified.  Similar  analyses will be undertaken in Volume 2 of this effort for TCDD
       cancer dose-response assessment.
2.3.  STUDY SELECTION PROCESS FOR TCDD DOSE-RESPONSE ANALYSIS
       In this section, EPA describes the study selection process that includes both

TCDD-specific study selection criteria and methodological considerations that have been

developed to evaluate epidemiologic studies and animal bioassays for quantitative TCDD dose-

response assessment. These criteria and considerations reflect EPA's goal of developing

noncancer and cancer toxicity values for TCDD through a transparent study selection process;

they are intended to be used by EPA for TCDD dose-response assessment only.  The TCDD in

vivo mammalian literature base differs from most other chemicals in magnitude and

comprehensiveness. It comprises -1,500 studies that evaluate multiple cancer and noncancer

endpoints, many species including humans, and covers an expansive dose range, including doses

                                          2-5

-------
at and below 1 nanogram per kilogram body weight per day (ng/kg-day).  Thus, the study
inclusion criteria and considerations developed in this document are specific to evaluating the
TCDD literature and cannot necessarily be generically applied to other chemicals. Further,
TCDD has a long half-life in humans (~7 years) and bioaccumulates in fat tissue, resulting in the
specification of study inclusion criteria for estimating exposures during the critical windows for
adverse health effects. In this effort, EPA sought to identify a group of studies for TCDD
dose-response evaluation that would span the types of adverse health effects associated with
TCDD exposures and encompass the range of doses in the lower end of the dose-response region
most relevant to human health protection. Detailed study inclusion criteria have been developed
that consider TCDD-specific issues and reflect EPA methods for POD identification, RfD
derivation, and oral slope factor (OSF) derivation.  (The effort in this document contrasts with
EPA's 2003 Reassessment where the focus was on individual endpoints and the goal was to
compare dose response across studies.)
       The study inclusion criteria and considerations were applied to each of the studies listed
in the "Preliminary Literature Search Results and Request for Additional Studies  on
2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) Dose-Response Studies" (U.S. EPA. 2008a):
studies identified and submitted by the public and by participants in the Dioxin Workshop (U.S.
EPA, 2009b): studies included in the 2003 Reassessment; and other relevant published studies
collected by EPA scientists through October 2009.  In this effort, the goal was to  identify the
most relevant studies for TCDD quantitative human health risk analyses.  Those that did not
qualify were not used quantitatively, but some of these were still considered relevant to the
qualitative evaluations of TCDD noncancer and cancer assessments.  Similarly, some types of
studies were not screened, i.e., studies on DLCs, mixtures toxicity, mode of action,  in vitro
toxicity, nonmammalian toxicology, and risk assessment; however, they were considered to be
important supplemental information to be used as needed, for example, in discussions of
biological significance.
       For the study selection process, EPA has focused on TCDD studies and has not included
studies on DLCs or DLC mixtures because inclusion of the DLC literature would likely increase
the uncertainty in TCDD dose response unnecessarily, given that the TCDD database is quite
robust. In addition, EPA believes that using studies evaluating information primarily or
exclusively on TCDD dose response provides the most appropriate data for the risk assessment
                                          2-6

-------
of dioxins and DLCs using the TEF approach. EPA is concerned that: (1) using the TEQ data to
estimate TCDD toxicity values would not accurately reflect TCDD dose response; and
(2) uncertainty in the underlying data used to derive the TEF values would complicate the
extrapolation of TEQ dose-response data to inform TCDD dose response.
       Because TCDD is used as the index chemical in the TEF approach, the most relevant and
accurate information that specifically addresses quantitative dose response of individual TCDD
exposures is needed.  The WHO (World Health Organization) expert panel assigned TEF values
from a conservative perspective that was intended to be health protective (Van den Berg et al.,
2006). In the development of the TEFs, the WHO expert panel considered data from Haws et al.
(2006a, b), who present summary statistics of relative potency values assembled from selected in
vivo and in vitro studies. For each individual DLC, the WHO expert panel typically assigned
TEF values using an in vivo study whose relative potency value was above the 50th percentile of
the ranges presented by Haws et al. (2006a, b).  Thus, when these TEFs are used in a dose-
response study, they produce total TEQ estimates that may be biased high for certain
combinations of DLCs. If a RfD for TCDD were derived based on TEQ dose-response data, that
RfD would likely also be biased high and, in that case, would underestimate health risk from
environmental exposures. Thus, using the TEQ data to estimate TCDD toxicity values would not
accurately reflect TCDD dose response.
       Finally, there is uncertainty in how the underlying data were used to derive the TEF
values that complicates the extrapolation of TEQ dose-response data to inform TCDD dose
response.  The kinds of information available for calculating relative potencies within a study are
highly variable across DLCs, including many types of and numbers of in vivo (including
different test species) and in vitro studies. In addition, a number of different methods are
employed to calculate the range of relative potencies presented by Haws et al. (2006a, b),
ranging from comparing dose-response curves, to developing ratios of effective doses that cause
an effect in 50% of the test units (EDsos), to estimating values from graphs of dose-response data.
The uncertainty in the TEFs can be a substantial issue for dose-response modeling when effect
levels in a study occur at doses close to background TEQ levels and TCDD is not a dominant
component of the mixture. In this case, the contribution of TCDD dose to the observed toxic
effect may not be feasible to estimate as it is confounded by other TEQ concentrations and
impacted by other TEF uncertainties.
                                          2-7

-------
       EPA has undertaken different approaches for epidemiologic versus in vivo animal
bioassay study evaluation and key data set selection.  The significant differences between animal
and human health effects data and their use in EPA health assessment support development of
separate study inclusion criteria and different approaches to study evaluation.  For example,
animal bioassays on TCDD are closely controlled experiments where dose and effect are
precisely measured and causality can be more easily inferred; thus, the animal criteria contain
precise dose limits and specific limitations on elements of the experimental design. Because
epidemiologic studies on TCDD are carried out within a population setting, these observational
studies employ statistical and other analytical techniques to estimate exposures/doses, and to
assess dose-response relationships after controlling or accounting for confounding factors and
other potential sources of bias. Thus, the epidemiologic criteria contain requirements for being
able to reasonably quantify the exposure-response relationship for the biologically-relevant
exposure window.18
       Section 2.4 and Appendices C and D present the results of the study selection process. In
Appendix C, all of the available epidemiologic studies on TCDD are summarized and evaluated
for suitability for dose-response modeling using the TCDD-specific study inclusion criteria
described in Section 2.3.1 below; only studies meeting the  study inclusion criteria and study
quality considerations are presented as key studies in Section 2.4.1 (see Tables 2-1  and 2-2 for
the cancer and noncancer endpoints, respectively).  In Appendix D, because summarizing all of
the available animal bioassays on TCDD was prohibitive, only studies first meeting the in vivo
animal bioassays study inclusion criteria described in Section 2.3.2 below are summarized;
Tables D-l  and D-2 present the results of the study selection process evaluations for the studies
that met and did not meet the study inclusion criteria, respectively.  The selected animal studies
are presented as key studies in Section 2.4.2 (see Tables 2-3 and 2-4 for cancer and noncancer
endpoints, respectively).
18 Critical exposure windows can be identified either through conceptual understanding of the timing of the affected
biological process, such as a susceptible life-stage during which the effect is manifested, or empirically, when such
critical windows are evident from the results of an epidemiological study. Note that the conceptual understanding
can be obtained independently of the epidemiologic study in question.
                                            2-8

-------
 2.3.1.  Study Inclusion Criteria for TCDD Epidemiologic Studies
       This section describes the process EPA used to select epidemiologic studies for
identifying PODs for TCDD quantitative dose-response assessment.19 This selection process
includes specific criteria based on EPA's approaches for deriving OSFs and RfDs (see
Text Box 2-1). Additional considerations used in selecting epidemiologic data for quantitative
dose-response modeling are also necessary, particularly given EPA's preference to use human
studies over animal studies whenever possible (U.S. EPA, 2005a). As described by
Hertz-Picciotto (1995), key components needed for the use of an epidemiologic study as a basis
for quantitative risk assessment include issues
regarding exposure assessment and overall
study quality. Exposure assessments need to
be well-quantified with exposures linked to
individuals. Different types of biases (e.g.,
confounding) also need to be eliminated in
these studies.  For example, biases related to
inclusion criteria for membership in the study
population and follow-up procedures need to
be ruled out or considered to have a neglible
impact on study findings. In addition,
confounding should be controlled for or at
least likely to be limited.  The strength of the
 Text Box 2-1.  EPA Risk Assessment Guidelines and
    Guidance Documents for Toxicity Assessment
Guidelines for Mutagenicity Risk Assessment (U.S. EPA.
1986a)
Guidelines for the Health Risk Assessment of Chemical
Mixtures (U.S. EPA. 1986b)
Guidelines for Developmental Toxicity Risk Assessment
(U.S. EPA. 19911
Guidelines for Reproductive Toxicity Risk Assessment
(U.S. EPA. 19961
Guidelines for Neurotoxicity Risk Assessment (U.S. EPA.
1998)
Benchmark Dose Technical Guidance Document [external
review draft] (U.S. EPA. 20001
Guidelines for Carcinogen Risk Assessment (U.S. EPA.
2005a')
Supplemental Guidance for Assessing Susceptibility from
Early-Life Exposure to Carcinogens (U.S. EPA. 2005b')
association, either within the full study or within a high exposure subgroup, can also be
considered in the evaluation of suitability for dose-response modeling (Hertz-Picciotto, 1995).
Stayner et al. (1999), however, note that even weak associations could be useful in terms of
providing an estimate of a potential upper bound for a quantitative risk estimate.
       EPA's study selection process included applying TCDD-specific study inclusion criteria
to epidemiologic data which met the five following considerations (also see Figure 2-2 for more
details):
19 In general, for these epidemiologic studies, EPA is evaluating tissue concentrations of TCDD that have been used
in conjunction with kinetic modeling to estimate previous TCDD exposures.
                                             2-9

-------
                  List of available epidemiologic studies on TCDD and DLCs
                              (All studies summarized.)
                                     Study
                                 in peer-reviewed
                                    literature?
                                    Exposure
                                 primarily to TCDD
                                 and quantifiable?
                    Long-term
                   exposures and
                 latency information
                 available forcancer
                    ssessment?
    Exposure
  windows and
latency information
 available for RfD
   ssessment?
                     Yes
                                                  Yes
           Evaluate study using five considerations:
           •  Methods used to ascertain health outcomes are clear and unbiased?
           •  Confounding and other potential sources of bias are addressed?
           •  Association/exposure response between TCDD and adverse effect?
           •  Exposures based on individual-level estimates, uncertainties described?
           •  Statistical precision, powerand study follow-up are sufficient?
                                  Considerations
                                   adequately
                                    satisfied?
                                    Yes
                                Key study included
                          forTCDD cancerand/or noncancer
                            dose-response assessment
                  Study excluded
                    from TCDD
                  dose-response
                    assessment
Figure 2-2. EPA's selection process to evaluate available epidemiologic
studies using study inclusion criteria and other epidemiologic considerations
for use in the dose-response analysis of TCDD.
EPA applied its TCDD-specific epidemiologic study inclusion criteria to all studies published on
TCDD and DLCs.  For all peer-reviewed studies, EPA examined whether the exposures were
primarily to TCDD and if the TCDD exposures could be quantified so that dose-response analyses
could be conducted. Then, EPA required that the effective dose and oral exposure be estimable:
(1) for cancer, information is required on long-term exposures, (2) for noncancer, information is
required regarding the appropriate window of exposure that is relevant for a specific, nonfatal
health endpoint, and (3) for all endpoints, the latency period between TCDD exposure and the
onset of the health endpoint is needed.  Finally, studies were evaluated using five considerations
regarded as providing the most relevant kind of information needed for quantitative human health
risk analyses. Only studies meeting these criteria and adequately satisfying the considerations
were selected for EPA's TCDD dose-response analysis.
                                          2-10

-------
    1.  The methods used to ascertain health outcomes are clearly identified and unbiased (e.g.,
       outcome classification was made "blinded" to exposure levels of the study participants).

    2.  The risk estimates generated from the study are not susceptible to important biases
       arising from an inability to control or account for confounding factors or other sources of
       bias (e.g., selection or information bias) arising from limitations of the study design, data
       collection, or statistical analysis.

    3.  The study demonstrated an association between TCDD and an adverse health endpoint
       (assuming minimal misclassification of exposure and absence of important biases) with
       some suggestion of an exposure-response relationship.

              This consideration addresses the use of null studies (i.e., studies reporting no
              association between TCDD and the health endpoint of interest) for the
              quantitative dose-response assessment used to derive an RfD; such studies are still
              used in qualitative assessments. Theoretically, a no-observed-adverse-effect level
              (NOAEL) can be identified from a null study and used to derive an RfD; that is,
              the highest available exposure dose from such a study could provide a NOAEL,
              which could serve as a basis for an RfD after appropriate uncertainty factors were
              applied. However, a NOAEL from a study in which no adverse effects have been
              observed is not usually chosen for RfD derivation when other available studies
              demonstrate lowest-observed-adverse-effect levels (LOAELs). The large and
              comprehensive database available to assess quantitative TCDD dose response
              provides many positive studies that are considered stronger candidates for
              derivation of an RfD than the studies for which only a NOAEL can be identified.
              [However, null studies are used by EPA to discuss the biological significance of
              the critical endpoint(s) used as the basis for deriving an RfD.]

    4.  The exposure assessment methodology is clearly  described and can  be expected to
       provide adequate characterization of exposure, with assignment of individual-level
       exposures within  a study (e.g., based on biomarker data, or based on a
       job-exposure-matrix approach20).  Limitations and uncertainties in the exposure
       assessment are considered.

    5.  The size and follow-up period of a cohort study are large enough and long enough,
       respectively, to yield sufficiently precise estimates for use in development of quantitative
       risk estimates and to ensure adequate statistical power to limit the possibility of not
       detecting an association that might be present.  Similar considerations regarding sample
       size and statistical precision and power apply to other study designs such as case-control
       studies.
20 A job-exposure matrix approach consists of a number of related methods for the quantification of occupational
exposures that can be used to help assess potential risk.

                                           2-11

-------
       In addition to these five study considerations, three specific study inclusion criteria were

used to select studies for further evaluation and potential TCDD quantitative dose-response

assessment:
    1.  The study is published in the peer-reviewed
       scientific literature and provides an appropriate
       discussion of data collection and analysis methods,
       as well as sufficient detail to allow consideration
       of its strengths and limitations.

    2.  The exposure is primarily to TCDD, rather than
       DLCs, and can be quantified so that dose-response
       relationships can  be assessed for non-fatal adverse
                  91
       endpoints.   Because all epidemiologic cohorts
       have background exposures to DLCs, in which
       TCDD is a minor component, only those studies
       for which TCDD  exposure is well above
       background will qualify for dose-response
       modeling. To the extent to which background
       DLC exposure becomes more significant with
       respect to TCDD  exposure,  limited quantitative
       assessment of DLC background exposures  may be
       necessary.

    3.  The effective dose and oral  exposure must be
       quantifiable.  The timing of the  measurement of
       health endpoints (i.e., the response) also must be
       consistent with current biological understanding of
       the endpoint and  its progression.

           For cancer endpoints, EPA assumes that
           cumulative TCDD dose estimates are
           lexicologically relevant measures.  Thus,
           cancer studies must provide information about
           long-term  TCDD exposure levels.  Further, for
           measures of cancer occurrence or death,
           sufficient follow-up is needed to allow  for
           examination of latency between the end of
           effective exposure and cancer detection or
           death.
            Text Box 2-2.
       Critical Exposure Window

In this document, a biologically-relevant
critical exposure window of susceptibility
("critical exposure  window" or "critical
window") is defined as an exposure period
during some specific life stage over which
an individual is susceptible to the agent
(e.g.,  TCDD)  for  a particular  health
endpoint.  In  utero and  early lifetime
exposures are often identified  as critical
exposure windows  for  many  defects in
anatomical  and physiological  processes
under development during those periods.
Critical  exposure   windows   can   be
identified   either  through   conceptual
understanding of the timing of the affected
biological process,  such as a susceptible
life-stage  during  which  the  effect is
manifested,  or empirically,  when such
critical windows are evident  from the
results of  an epidemiologic  study.  An
example of the latter is the semen quality
effects associated with early exposure to
TCDD for  boys under 10 years of age
compared to boys 10-17 years of age at
the time of TCDD exposure (Morarelli et
al. [20081:  see Appendix  C  for study
details). Identifying  such critical windows
is  important for TCDD in the practical
sense  of defining a reasonable duration
over which  to average internal exposures
that vary greatly from an initial high peak
exposure  to  a much lower terminal
exposure, as is the case  for almost of the
epidemiologic studies under consideration
for TCDD.  EPA  considers  the internal
exposures following  the  actual TCDD
exposure  incident  to  be  relevant  for
averaging because of the relatively slow
elimination  of TCDD and the  possibility
that these  concentrations  could still be
affecting the  processes leading to  the
adverse health outcome.
 1 The IRIS Program does not generally base RfDs on highly severe effects, such as mortality.
                                             2-12

-------
          For noncancer endpoints, exposure estimates and analysis must allow for examination
          of issues of latency and other issues regarding the appropriate time window of
          exposure relevant for specific endpoints. That is, there must be sufficient
          information, either in the study or elsewhere, to allow for the identification of a
          biologically-relevant "critical exposure window" of susceptibility (see Text Box 2-2).
       Those studies that satisfied these three study inclusion criteria and, in addition,
adequately satisfied the study quality provisions specified in the five considerations were
considered to be suitable for quantitative TCDD dose-response analyses (see results in
Section 2.4.1 and Appendix C).

 2.3.2.  Study Inclusion Criteria for TCDD In Vivo Mammalian Bioassays
       This section identifies the criteria EPA applied to select nonhuman in vivo mammalian
studies for defining PODs for use in TCDD dose-response modeling. These criteria are
specifically developed to evaluate the TCDD literature and are not necessarily generic, however,
they are based on EPA's approaches for deriving OSFs and RfDs from bioassay data (see
Text Box 2-1).  EPA agrees with the NAS committee regarding the utility of an oral RfD and the
need for reevaluation of the OSF for TCDD, specifically in light of data that have been published
since the 2003 Reassessment was released. RfDs and OSFs are generally derived using data sets
that demonstrate the occurrence of adverse effects, or their precursors, in the low-dose range for
that chemical. RfDs and OSFs are derived from a health-protective perspective for chronic
exposures. Thus, when a group of studies is available on a chemical for which a number of
effects are observed at various doses across those studies, the studies using the lowest doses that
show effects will typically be selected as the basis of the RfD and OSF derivations, all other
considerations being equal.  Studies conducted at higher doses relative to other available studies
are used as supporting evidence for the final RfD or OSF because they were conducted at doses
too high to impact the numeric derivations of toxicity values.
       EPA expresses RfDs and OSFs in terms of average daily doses, usually as mg/kg-day and
per mg/kg-day,  respectively. Thus, the study inclusion criteria for the animal bioassay data
presented  in this section include requirements that average daily exposures in the studies are
within a low-dose range where, relative to other studies, they could be considered for
development of a toxicity value. These low-dose requirements do not imply that TCDD studies
conducted at higher doses are of poor quality, simply that they are not quantitatively useful in the
                                          2-13

-------
development of toxicity values because other studies with lower exposures will be selected as the
basis of the RfD and OSF derivations under current EPA guidance (see Text Box 2-1). Because
EPA has identified hundreds of in vivo mammalian studies that may be considered for
quantitative TCDD dose-response assessment, the development and application of these study
inclusion criteria have been critical  to moving the health assessment process forward.
       EPA's method for applying  TCDD-specific  study inclusion criteria for mammalian
bioassays is detailed below and in Figure 2-3.  Four specific study inclusion criteria were used to
select studies for further evaluation and potential TCDD quantitative dose-response analyses and
identification of PODs:

    1.  The study is published in the peer-reviewed  scientific literature.
    2.  The study was not conducted on a genetically-altered species.
    3.  The lowest dose level tested is <1 ug/kg-day for cancer studies and <30 ng/kg-day for
       noncancer studies.
    4.  The study design consists of orally administered TCDD-only doses.

       Those studies that satisfied these four criteria (see results in Section 2.4.2 and
Appendix D) were considered suitable for quantitative TCDD dose-response analysis.
       In evaluating the selected in vivo animal studies, EPA considered study quality issues to
ensure that the study provided important information needed to assess the relevance of the
study's endpoints and to quantify the dose-response relationship.  Each study needed to test a
mammalian species and identify the strain, gender, and age of the tested animals. The study had
to clearly document its testing protocol, including dosing frequency, duration, and timing of dose
administration relative to age of the animals. For example, the control group or groups had to be
well characterized and appropriate,  given the testing protocol.  Also, clinical and pathological
examinations conducted during the  study needed to be endpoint-appropriate, particularly for
negative findings. EPA used the results of these study evaluations in drafting study summaries
for  all of the animal bioassays that met the study inclusion criteria (see Appendix D).
                                          2-14

-------
   List of available in vivo mammalian bioassay studies on TCDD
                           Study
                       in peer-reviewed
                         literature?
                         Study on a
                      genetically-altered
                          species?
           Lowest
        dose tested for
      cancerendpoint<1
          ug/kg-day?
    Lowest dose
tested for n on cancer
    endpoint<30
    ng/kg-day?
                            Oral
                      exposure to TCDD
                           only?
                Study summarized; evaluated for
                  quality and to note adequacy
                    of data needed for TCDD
                   dose-response assessment.
                      Key study included
               forTCDD cancerand/ornoncancer
                   dose-response assessment
                     Study excluded
                       from TCDD
                     dose-response
                      assessment
Figure 2-3. EPA's process to evaluate available animal bioassay studies
using study inclusion criteria for use in the dose-response analysis of TCDD.
EPA evaluated all available in vivo mammalian bioassay studies on TCDD. Studies had to be
published in the peer-reviewed literature. Studies on genetically-altered species were excluded as
their relevance to human health is not known. Next, EPA applied dose requirements to each
study's lowest tested average daily dose, with requirements for cancer (<1  ug/kg-day) and
noncancer (<30 ng/kg-day) studies. EPA also required that the animals were exposed via the oral
route to only TCDD. Finally, the studies were evaluated for quality and summarized to ensure
providing the most relevant information for quantitative human health risk analyses. Only studies
meeting all of the criteria were selected for EPA's TCDD dose-response analysis.
                                        2-15

-------
       The criteria for dose requirements are intended to be reasonable limits that restrict the
number of studies that would need to be considered while ensuring that all study/data set
combinations that could be candidates for deriving a cancer or noncancer toxicity value were
analyzed.  Thus, the dose range under consideration allows for liberal ranges of NOAELs,
LOAELs, and benchmark dose lower confidence bounds (BMDLs) for assessment of both cancer
and noncancer effects. The dose requirements for cancer and noncancer studies were set after
EPA conducted a brief review of typical dose levels in studies analyzed in the 2003
Reassessment and in some  of the more recent studies found through EPA's literature search.
       For cancer studies, the  low-dose limit was selected liberally so as not to exclude a study
that might possibly report a sensitive tumor endpoint.  Given that the limit of 1  ug/kg-day is
3 orders of magnitude higher than the lowest-tested dose in one of the most sensitive animal
bioassays (Kociba et al., 1978) evaluated in U.S. EPA (2003), it is virtually impossible that a
study with a low dose of 1  ug/kg-day or greater would ever be considered for deriving a cancer
toxicity value. Following identification of new animal cancer bioassays, no studies were
eliminated based on this limit.
       For noncancer studies,  the identification of a low-dose limit is more complicated because
of the variety of exposure protocols and endpoints and the consequent varied degree of
toxicokinetic  extrapolation to human equivalent exposures.  However, EPA is confident that the
low-dose limit of 30 ng/kg-day will not exclude any study from which a POD could be derived
that would be low enough to be considered for the RfD.  A preliminary screening of the literature
indicated that, for all study types (e.g., acute, developmental, chronic), there are many studies
with apparent effect levels well below 30 ng/kg-day.  Effects observed above 30 ng/kg-day,
therefore, would have no chance of being considered as the basis for an RfD.

2.4.  SUMMARY OF KEY DATA SET SELECTION FOR TCDD DOSE-RESPONSE
     MODELING
       To meet the NAS' concerns regarding transparency and clarity in the identification of
TCDD studies for dose-response assessment, EPA has developed and applied two sets of criteria
for epidemiologic studies and animal bioassays.  EPA collected these studies through October,
2009, including studies from the 2003 Reassessment and newer studies found via literature
searches and through  public submissions (see Section 2.2 and Figure 2-1). Based on these
activities, a total of 1,441 studies were examined for their potential to be used in TCDD
                                         2-16

-------
quantitative dose-response analysis.  Of these, Figure 2-4 shows that 637 studies were eliminated
from consideration as they were not suitable study types; these included, in vitro bioassays,
review articles, PBPK modeling studies, and studies that evaluated PCBs or other dioxin-like
compounds other than TCDD.  Of the remaining studies, 49 were epidemiologic studies
(7 studies contained both cancer and noncancer endpoints), and 755 were animal studies
(4 studies contained both cancer and noncancer endpoints).  These epidemiologic and animal
studies were then evaluated using EPA's study inclusion criteria.
       Detailed results of EPA's evaluations and study summaries are shown in Appendices C
and D for the epidemiologic studies and animal bioassays, respectively.  Final results in tabular
form are shown in this section. Tables 2-1 and 2-2 contain the preliminary list of cancer studies
and the final list of key noncancer studies, respectively, that have met EPA's study inclusion
criteria for  epidemiologic data. Tables 2-3 and 2-4 provide the preliminary list of cancer
bioassays and the final list of key noncancer bioassays, respectively, that have met EPA's study
inclusion criteria for animal bioassay data. Collectively, Tables  2-2 and 2-4 contain the final set
of key studies that EPA has selected for development of the noncancer dose-response assessment
for TCDD presented in Section 4 of this document, Reanalysis, Volume 1. Tables 2-1 and 2-3
provide preliminary lists of cancer studies that will be useful in developing the cancer
dose-response assessment to be presented in Reanalysis, Volume 2.
       Through this study selection process, EPA has identified a relevant group of studies that
spans the possible risk analytic choices for human health protection.  Each study provides
important TCDD dose-response information but also is associated with limitations and
uncertainties that must be considered and characterized during TCDD dose-response evaluations.
EPA has benefited from this effort by greatly reducing the scope of dose-response modeling and
analyses to a manageable size, and by focusing on the most important studies from the
perspective of developing cancer and noncancer toxicity values.  Results  of applying the study
inclusion criteria showed that exposure information was a primary factor in study selection (see
Figure  2-4). In the epidemiologic studies, exposure needed to be primarily to TCDD and
quantifiable on an individual level.  In addition, the identification of critical exposure windows
(see Text Box 2-2) and the availability of latency information  in the epidemiologic studies were
vital data for developing human exposure estimates. In the animal studies, dose limits were the
most important criteria.
                                          2-17

-------
                 Studies from literature search and data collection activities
                                        1,441
                                          I
Rightstudy type for quantitative
TCDD dose-response analysis:
804 considered further
\

Wrong study type for quantitative
TCDD dose-response analysis:
637 excluded

t
Epidemiologic (Epi) studies
49

si'
Animal bioassays
755
\E
Epi cancer
studies
24
v
Failed >1 of*:
Peer-review (0)
Primarily TDCC
(10)
Effective
exposure
estimable ( 11 )
Considerations**
(1)
V
Epi
cancer
studies
Included
1 8 J






^
Epi noncancer
studies
32

Failed>1 of*:
Peer-review (1)
Primarily TDCC
(7)
Effective
exposure
estimable (26)
Considerations**
(1)
I
Epi
noncancer
studies
included
1 4 J
\E
Cancer
bioassays
8



i
Noncancer
bioassays
751
V V
Failed >1 of*:
Peer-review (0)
Genetically-
altered(1)
Dose cutoffs
(0)
TCDD only (0)
Non-oral (1)








Failed >1 of*:
Peer-review (0)
Genetically-
altered (66)
Dose cutoffs
(370)
TCDD only (142)
Non-oral (135)
v v
s ~\ f ^
Animal
cancer
bioassays
Included
6





Animal
noncancer
bioassays
included
78
v y v j
   *Failed criteria are not mutually exclusive; more than one can fail for a given study.
   **lndicates those studies that passed allthree criteria butwere not selected based on
   study considerations.
Figure 2-4. Results of EPA's process to select and identify in vivo
mammalian and epidemiologic studies for use in the dose-response analysis
of TCDD.
Four animal studies and seven epidemiologic studies contained both cancer and noncancer
endpoints. Two epidemiologic cancer studies, Steenland et al. (1999) and Flesch-Janys et al.
(1998), passed all criteria, but were still not selected because they were superseded by other
studies on the same cohort for which an improved analysis was done. One noncancer
epidemiologic study, Baccarelli et al. (2005). passed all criteria, but was excluded because the
health endpoint, chloracne, is considered to be an outcome associated with high TCDD exposures.
                                        2-18

-------
          Table 2-1. Epidemiologic studies selected for TCDD cancer dose-response modeling
Reference
Akhtar et al.
(2004)

























Health
outcome
Mortality and
incidence for
all cancers
and for site-
specific
cancers
including
sro state and
melanoma


















Location,
time period
Vietnam
1962-1971

























Cohort
description
Ranch Hand (RH)
cohort including
1,1 96 U.S. military
males exposed by
spraying Agent
Orange during
Vietnam war in
Southeast Asia
(SEA); comparison
(C) cohort matched
ay age, race, and
military
occupation.














Exposure
assessment
Cumulative serum
lipid concentrations
(CSLC)ofTCDD
based on serum
levels collected
from veterans in
1987, 1992, 1997,
and a first-order
kinetic model with
a 7.6-year half-life.
CSLC estimates for
1, 009 RH cohort
and 1,429 C cohort
veterans.













Exposure
measures
CSLC
(ppt-years)
RH and C <2 yrs
in SEA:

All site
Comparison
<10
Low>10-118.5
High >1 18.5
Continuous (Log
fCDD)

Melanoma
Comparison
<10
Low>10-118.5
High >1 18.5
Continuous (Log
fCDD)
Prostate
Comparison
<10
Low>10-118.5
High >1 18.5
Continuous (Log
fCDD)
No. of
cases or
deaths





No.,%
34,5.9
28,9.8
22, 14.6
15,8.6



No., %
3,0.5
4,1.4
4,2.7
3,1.7


No., %
7,1.2
10,3.5
6,4.0
5,2.9


Effect measure/
trend tests
(p-value)





RR (95% CI)
1.0
1.44(0.82-2.53)
2.23(1.24-4.00)
2.02(1.03-3.95)
1.24(1.01-1.53)
p = 0.04


1.0
2.99(0.53-16.8)
7.42(1.34-41.04)
7.51(1.12-50.21)
2.24(1.29-3.89)
p = 0.004

1.0
1.5(0.51-4.40)
2.17(0.68-6.87)
6.04(1.48-24.61)
1.48(0.93-2.35)*
p = 0.10
Risk factors
Adjusted for age at
tour, military
occupation, smoking,
skin reaction to sun
exposure, eye color,
number of years in
SEA.

Also stratified
analyses by year of
tour of duty.
Restricted to <2 years
in SEA, white Air
Force veterans, 0%
and 100% time in
Vietnam for RH and
C Cohorts,
respectively.









Comments
Used multiplicative
Poisson regression
models to compare
cancer incidence and
cancer mortality with
national rates and
proportional
hazards models to
contrast cohorts with
regard to cancer
incidence.
















to
I


VO

-------
          Table 2-1. Epidemiologic studies selected for TCDD cancer dose-response modeling (continued)


Reference
Becher et al.
C1998)
















Cheng et al.
f2006)















Health
outcome
Mortality
from all
cancers
combined














Mortality
from all
cancers














Location,
time period
Hamburg,
Germany,
production
seriod was
1950-1984,
and mortality
follow-up
extended
through 1992









USA, 1942-
1993















Cohort
description
Boehringer cohort
including
approximately
1,189 workers
employed in the
production of
tierbicides.











NIOSH cohort
including 3,538
occupationally
exposed male
workers at 8 plants
in the United
States; 256 cancer
deaths.









Exposure
assessment
CSLC of TCDD
based on area
under curve (in
ug/kg years); back-
extrapolation to
date of last
employment took
into account age
and percentage
body fat; half-life
value was
7.2 years.






CSLC of TCDD
based on work
tiistories, job-
exposure matrix,
and concentration
and age-dependent
two-compartment
model of
elimination
kinetics.







Exposure
measures



Categorical
exposures (Cox
model)
0-<1
l-<4
4-<8
8-<16
16-<64
64+


Continuous
exposure
TCDD (ug/kg
years)
No exposure
categories
provided













No. of
cases or
deaths





124








124



256 cancer
deaths














Effect measure/
trend tests
(p-value)





RR (95% CI)
1.0
1.12(0.70-1.80)
1.42(0.70-2.85)
1.77(0.81-3.86)
1.63(0.73-3.64)
2.19(0.76-6.29)
p = 0.03

P = 0.0089,
p = 0.0047


The slope (P) was 3.3
x 10"6 for lag of 15
years excluding
upper 5% of TCDD
exposures.
The slopes ranged
wo orders of
magnitude depending
on modeling
assumption.








Risk factors



Available: year of
entry, age of entry,
duration of
employment, birth
cohort, P-HCH; TEQ
other than TCDD.

Available: year of
entry, age of entry,
duration of
employment, birth
cohort, P-HCH; TEQ
other than TCDD.


Available: age, year
of birth, and race.

Risks adjusted for:
year of birth, age, and
race.

Indirectly examined
other potential
confounders such as
smoking and other
occupational
exposures.





Comments
Included in U.S. EPA
(2003).
A large number of
models were fitted.
These included models
for 5 different latency
intervals (0, 5, 10, 15,
and 20 years), as well as
multiplicative, additive,
and power models, and
different offset variables
(person years and
expected deaths).





Confounding by
smoking was considered
indirectly by analysis of
smoking-related and
smoking-unrelated
cancers.
Other occupational
exposures were
considered indirectly by
repeated analyses
removing one plant at a
ime.
Based on indirect
evaluation, there was no
clear evidence of
confounding.
to
o

-------
          Table 2-1.  Epidemiologic studies selected for TCDD cancer dose-response modeling (continued)
Reference
Collins et al.
f2009)
























Health
outcome
Mortality
from all
cancers and
specific
cancer types





















Location,
time period
Midland, MI,
USA.
Follow-up
seriod: 1942-
2003. Serum
collection
period: 2004-
2005


















Cohort
description
Subset of NIOSH
cohort including
1,615
occupationally
exposed male
workers at 1 plant
in the United
States; 1 77 cancer
deaths.

















Exposure
assessment
CSLC of TCDD
based on work
histories,
job-exposure
matrix, and
concentration and
age-dependent two-
compartment
model of
elimination
kinetics. Serum
samples were
obtained from 280
former workers
collected during
2004-2005.










Exposure
measures
Part per billion-
year
estimates of
cumulative TCDD
exposure





















No. of
cases or
deaths
177 cancer
deaths
























Effect measure/
trend tests
(p-value)
The slope of a
sroportional hazards
regression model for
fatal soft tissue
sarcoma was 0.05872
(95% CI not
provided but for Chi-
square^ = 0.0060)
for every 1-part per
sillion-year increase
in cumulative
exposure of TCDD.
Slope estimates for
all fatal cancers
(0.00161, ;> = 0.78),
fatal lung (-0.00 173,
v = 0.89), fatal
srostate (0.01294,
v = 0.30), fatal
leukemias (-0.12822,
p = 0.34), and fatal
non-Hodgkin
[ymphomas
(0.01081, ;> = 0.68)
were not statistically
significant.
Risk factors
Hazard ratios
adjusted for age, year
of birth, and hire
year. Stratified
analyses used to
examine potential
impact of
pentachlorophenol
exposure on
mortality.
















Comments
Confounding by
smoking was not
considered directly due
:o a lack of data.
Relatively long follow-
up period (average =36
years).
Potential outcome
misclassification for
soft tissue sarcoma due
:o potential inaccuracies
on death certificates.
Data analyzed from one
slant reduces
leterogeneity associated
with multiplant
analyses. More serum
samples (n = 280)
analyzed than used to
derive TCDD estimates
for other NIOSH cohort
analyses.




to
to

-------
          Table 2-1.  Epidemiologic studies selected for TCDD cancer dose-response modeling (continued)
Reference
Michalek and
Pavuk (2008)
















Health
outcome
Cancer
incidence, all
sites
combined














Location,
time period
Vietnam
1962-1971
















Cohort
description
RH cohort
including 1,196
U.S. military males
exposed by
spraying Agent
Orange during
Vietnam war in
Southeast Asia
(SEA); C cohort
matched by age,
race, and military
occupation.






Exposure
assessment
CSLC of TCDD
based on serum
levels collected
from veterans in
1987, 1992, 1997,
2002, and a first-
order kinetic model
with a 7.6-year
half-life. CSLC
estimates for 986
RH cohort and
1,5 97 C cohort
veterans.





Exposure
measures
CSLC
(ppt-years)

Results stratified
by<1968,>30
day spre- 1967, <2
yrs in SEA:






Comparison
<10
Low>10-91
High>91

No. of
cases or
deaths
Continuou
s ex-
posure:
Log
(TCDD)
No.,%
67, 12.6



Cate-
gorical
TCDD
No., %
30,11.2
10,8.3
12,24.5
15,16.1
Effect measure/
trend tests
(p-value)






1.4(1.1-1.7)
v = 0.005





RR (95% CI)
1.0
0.5(0.2-1.1)
1.7(0.8-3.5)
2.2(1.1-4.4).
Risk factors
Cox regression
proportional hazards
models adjusted for
year of birth, eye
color, race, smoking,
body mass index at
the qualifying tour,
military occupation,
and skin reaction to
sun exposure.

Also stratified
analyses by years of
service in SEA, days
of herbicide spraying,
calendar period of
service.

Comments
Without stratification,
ihere was no significant
increase in the risk of
cancer with log (TCDD)
in the combined cohort.













to
to

-------
          Table 2-1. Epidemiologic studies selected for TCDD cancer dose-response modeling (continued)


Reference
Ott and Zober
(1996)




















Steenland et
al. (2001)










Health
outcome
Mortality and
incidence for
all cancers
combined, as
well as for
specific
cancer sites















Mortality
from all
cancers









Location,
time period
Ludwig-
shafen,
Germany,
1954-1992


















USA,
1942-1993










Cohort
description
BASF cohort, 243
men exposed from
accidental release
that occurred in
1953 during
production of
trichlorophenol, or
who were involved
in clean-up
activities.












NIOSH cohort
including 3,538
male workers, 256
cancer deaths.








Exposure
assessment
CSLC of TCDD
expressed in |ig/kg
based on TCDD
half-life of 5. 1-8. 9
years, Cox
regression model.
















CSLC of TCDD
based on work
histories, job-
exposure matrix,
and a simple one-
compartment, first-
order
pharmacokinetic
elimination model
with 8.7-year half-
life.

Exposure
measures
Internal
comparisons basec
on continuous
measure of
TCDD.






External
comparisons
exposure
categories (for
malignant
neoplasms):
0.1-0.99 1.0-1.99
>2 |ig/kg



CSLC
(ppt-years)
<335
335-520
520-1,212
1,212-2,896
2,896-7,568
7,568-20,455
>20,455


No. of
cases or
deaths
Internal
cohort
analysis

31 All
cancer
deaths

47 All
incident
cancers

External
cohort
analyses


Deaths
8
8
8
7


64
29
22
30
31
32
48


Effect measure/
trend tests
(p-value)




RR (95% CI)
1.22(95%CI:
1.00-1.50)


1.11 (95% CI:
0.91-1.35)






SMR (95% CI)
0.8(0.4-1.6)
1.2(0.5-2.3)
1.4(0.6-2.7)
2.0 (0.8-4.0)

RR (95% CI)
1.00
1.26(0.79-2.00)
1.02(0.62-1.65)
1.43(0.91-2.25)
1.46(0.93-2.30)
1.82(1.18-2,82)
1.62(1.03-2,56)




Risk factors
Available: age, BMI,
smoking status, and
history of
occupational
exposure to aromatic
amines and asbestos.
















Available: date of
birth and age.

Adjusted for date of
birth, and age was
used as time scale in
Cox model.






Comments
Included in U.S. EPA
(2003)

Positive associations
noted for digestive
cancer, but not for
respiratory cancer.

Association between
ICDD and increased
SMRs found only
among current smokers.

Last published account
of this cohort.







Included in U.S. EPA
f2003)









to

-------
           Table 2-1. Epidemiologic studies selected for TCDD cancer dose-response modeling (continued)
Reference
Warner et al.
(2002)










Health
outcome
Breast cancer
incidence









Location,
time period
Italy
1976-1998









Cohort
description
981 women from
Zones A and B
with available
archive serum
samples, 1 5 breast
cancer cases.









Exposure
assessment
CSLC of TCDD
(ppt) collected
between 1 976 and
1981. For most
samples collected
after 1977, serum
rCDD levels were
back-extrapolated
using a first-order
kinetic model with
a 9-year half-life.







Exposure
measures
Categorical
<20 ppt
20. 1-44 ppt
44. 1-100 ppt
> 100 ppt
Continuous
(Log10TCDD)








No. of
cases or
deaths
Cases
1
2
7
5
15








Effect measure/
trend tests
(p-value)
RR (95% CI)
1.0
1.0(0.1-10.8)
4.5 (0.6-36.8)
3.3 (0.4-28.0)
p = 0.07
2.1(1.0-4.6)








Risk factors
Available: gravidity,
parity, age at first
pregnancy, age at last
pregnancy, lactation,
family history of
breast cancer, age at
menarche, current
body mass index, oral
contraceptive use,
menarcheal status at
explosion, menopause
status at diagnosis,
tieight, smoking,
alcohol consumption.
Adjusted for age,
which was used as
time scale in Cox
model; other
covariates were
evaluated but were
not identified as
confounders.
Comments
Included in U.S. EPA
(2003)









to
    CI = confidence interval; CSLC = cumulative serum lipid concentration; HCH = hexachlorocyclohexane.

-------
          Table 2-2. Epidemiologic studies selected for TCDD noncancer dose-response modeling
Reference
Alaluusua et al.
(2004)























Health
outcome
Dental defects
























Location,
time period
Seveso, Italy,
Dental exams
administered
in 2001
among those
exposed to
TCDD in
1976

















Cohort
description
65 subjects
<9.5 years old
at time of
Seveso
explosion and
residing in
Zones ABR
(i.e., the most
heavily
contaminated
area in
decreasing
order); 130
subjects
recruited from
the non-ABR
region (i.e. the
unexposed).







Exposure
assessment
Serum TCDD
(ng/kg) from
1976 samples for
those who
resided in Zones
ABR; no serum
levels for non-
ABR residents
(unexposed).
TCDD exposure
represent levels
as of 1976 (after
accident).












Exposure
measures

Non-ABR
Zone
31-226
ng/kg
238-592
ng/kg
700-26,000
ng/kg

<5 years of
age at time of
accident




Non-ABR
Zone or
31-226
ng/kg serum
TCDD
238-26,000
ng/kg serum
TCDD
No. of
cases

10

1

5

9


25














Effect measure/
trend tests
(p-value)
Dental defect %
26%

10%

45%

60%
p-value = 0.0 16

33%
p-value = 0.0009


Odds Ratios (95%
CI)
(among those
<5 years of age at
time of accident)
1.0


2.4(1.3-4.5)
p-value = 0.007

Risk factors
Available: medical
history, age, sex,
education, smoking.






















Comments
Dose-response
pattern observed with
dental defects in the
ABR zone; however,
the control population
had a much higher
prevalence of dental
defects (26%) than
those in the lowest
exposure group
(10%).

Also assessed
hypodontia and other
dental and oral
aberrations, but these
were too rare to allow
modeling by ABR
zone.






to

-------
          Table 2-2. Epidemiologic studies selected for TCDD noncancer dose-response modeling (continued)


Reference
Baccarelli et
al. (2008)




















Eskenazi et al.
(20021))











Health
outcome
b-TSH
measured
72 hours after
birth from a
heel pick
(routine
screening for all
newborns in the
region)













Menstrual cycle
characteristics:
menstrual cycle
length.









Location,
time period
Italy, 1976;
children,
1994-2005



















Seveso, Italy,
follow-up
interview
conducted in
1996-1997 of
women
exposed to
TCDD in the
1976 accident




Cohort
description
Population-
based study:
1,041
singletons (56
from Zone A,
425 from
Zone B, and
533 from
reference)
born between
Jan. 1, 1994-
June 30, 2005.
Plasma dioxin
•study: 51
children born
to 38 women
of fertile age
who were part
of the Seveso
Chloracne
Study.

Women who
were <40
years from
Zones A or B
in 1976.








Exposure
assessment
Based on zone of
residence,
estimated mean
values from a
previous study.
Maternal plasma
TCDD levels
estimated at the
date of delivery
using a first-
order
pharmacokinetic
model and
elimination rate
estimated in
Seveso women
(half-life =
9.8 years).




Serum TCDD
(ng/kg) from
1976 samples.
TCDD exposure
level was back-
extrapolated to
1976 using the
Filser or the first-
order kinetic
models.



Exposure
measures
Population-
based study:




Reference


Zone A


Zone B



Plasma
dioxin study:
Continuous
maternal
plasma
TCDD
Interquartile
range was
64-322 ppt

TCDD
examined as
continuous
measure (per
10-fold
increase in
serum
levels).

No. of
cases






533
births

56 births

425
births





















Effect measure/
trend tests
(p-value)
Population-based
study
Geometric Mean
b-TSH (log-
transformed)

Reference:
0.98 (95% CI:
0.90-1.08)
ZoneB:
1.66 (95% CI:
1.19-2.31)
Zone A:
1.35 (95% CI:
1.22-1.49)

Association
setween neonatal
b-TSH with
plasma TCDD:
adjusted p = 0.75
(p< 0.001)




Lengthening of the
menstrual cycle by
0.93 days (95%
CI: -0.01, 1.86)






Risk factors
Available: gender,
birth weight, birth
order, maternal age
at delivery, hospital,
type of delivery.

There was limited
evidence of
confounding, so
mean TSH results
presented here are
unadjusted.










Interview data:
medical history,
personal habits,
work history,
reproductive history,
age, smoking, body
mass index, alcohol
and coffee
consumption,
exercise, illness,
abdominal surgeries.



Comments
An association with
serum TCDD levels
of mothers was found
with b-TSH among
the 5 1 births in the
plasma dioxin study.
















A positive association
setween menstrual
cycle length and
serum TCDD was
found among women
who were
premenarcheal at the
time of accident
(n = 134).



to

-------
           Table 2-2. Epidemiologic studies selected for TCDD noncancer dose-response modeling (continued)
Reference
Mocarelli et
al. (2008)


























Health
outcome
Sperm cone.
(million/mL)
Progressive
motility (%)
Serum E2
(pmol/L)






















Location,
time period
Italy, 1976,
1998


























Cohort
description
Among the
257 exposed
(from Zone
A), men 1-26
in 1976 with
serum levels
<2000 ppt in
1976, 135
(53%) were
included.
Among the
372
nonexposed
invitees, 184
(49%) men
aged 1-26 in
1976 were
included.










Exposure
assessment
Serum TCDD (in
ppt) from
1976-1977
samples (for
exposed men);
background
values were
assumed for
unexposed men
based on serum
analysis of
residents in
uncontaminated
areas.














Exposure
measures
Median
serum TCDD
levels (in
ppt) by
quartile for
men aged
1-9 in 1976
(68; 142;
345; 733 ppt)



















No. of
cases




























Effect measure/
trend tests
(p-value)

Men exposed
setween the ages
1-9 had reduced
semen quality
22 years later.
Reduced sperm
quality included
decreases in sperm
count (p = 0.025),
progressive sperm
motility
(p = 0.001), and
total number of
motile sperm
(p = 0.01) relative
to the comparison
group.










Risk factors
Available: age,
abstinence time,
smoking status,
education, alcohol
use, maternal
smoking during
pregnancy,
employment status,
BMI, chronic
exposure to solvents
and other toxic
substances.

Adjusted for
smoking status,
organic solvents, age
at time of tests,
BMI, alcohol use,
education,
employment status,
and abstinence
(days) for sperm
data.
Hormone data not
adjusted for
education level,
employment status,
and abstinence time.
Comments
Results stratified by
timing of exposure
(1-9 yrs old vs.
10-17 yrs old in
1976).























to
to
    b-TSH = blood thyroid-stimulating hormone; CI = confidence interval.

-------
           Table 2-3. Animal bioassays selected for cancer dose-response modeling
Reference
Delia Porta et al.
(1987)

Kociba et al.
(1978):
Goodman and
Sauer (1992)
NTP (1982c)
NTP (1982c)
Species/strain
Mouse/ B6C3FJ
Rat/Sprague-
Dawley
Mouse/ B6C3FJ
Rat/Osborne-
Mendel
Sex
exposure
route/duration
Male/female
Oral gavage once
per week; 52 weeks
Male/female
Oral-lifetime
feeding; 2 years
Male/female
Oral-gavage twice
per week; 104 weeks
Male/female
Oral-gavage twice
per week; 104 weeks
n
-40 to 50 in
each dose
group
including
controls
50 each
(86 each in
vehicle
control
group)
50 each
(75 each in
vehicle
control
group)
50 each
(75 each in
vehicle
control
group)
Average daily
dose levels
(ng/kg-day)
0,351, and 714
0, 1, 10, or 100
0, 1.4, 7.1, or 71 for
males;
0,5.7, 28.6, or 286
for females
0, 1.4, 7.1, or 71
Cancer types
Females and males:
hepatocellular
adenomas and
carcinomas
Females: liver, lung,
oral cavity
Males: adrenal, oral
cavity, tongue
Females: hematopoietic
system, liver,
subcutaneous tissue,
thyroid
Males: liver, lung
Females: adrenal, liver,
subcutaneous tissue,
thyroid
Males: adrenal, liver,
thyroid
Statistical significant tumors
(pairwise with controls or trend tests)
Liver: adenomas and carcinomas in females
and carcinomas in males (using incidental
tumor statistical test)
Adrenal cortex: adenoma
Liver: hepatocellular adenoma(s) or
carcinoma(s); hyperplastic nodules
Lung: keratinizing squamous cell carcinoma
Oral cavity: stratified squamous cell carcinoma
of hard palate or nasal turbinates
Tongue: stratified squamous cell carcinoma
Hematopoietic system: lymphoma or leukemia
Liver: hepatocellular adenoma or carcinoma
Lung: alveolar/bronchiolar adenoma or
carcinoma
Subcutaneous tissue: fibrosarcoma
Thyroid: follicular-cell adenoma
Adrenal: cortical adenoma, or carcinoma or
adenoma, NOS
Liver: neoplastic nodule or hepatocellular
carcinoma
Subcutaneous tissue: fibrosarcoma
Liver: neoplastic nodule or hepatocellular
carcinoma
Thyroid: follicular-cell adenoma or carcinoma
to

to
oo

-------
          Table 2-3. Animal bioassays selected for cancer dose-response modeling (continued)
Reference
NTP(2QQ6a)
Toth et al.
(19791
Species/strain
Rat/Harlan
Sprague-
Dawley
Mouse/
Outbred
Swiss/H/Riop
Sex
exposure
route/duration
Female
Oral-gavage
5 days per week;
2 years
Male
Gastric intubation
once per week;
1 year
n
53 or 54
43 or 44
(vehicle
control
group = 38)
Average daily
dose levels
(ng/kg-day)
0,2.14,7.14, 15.7,
32.9, or 71.4
0, 1, 100, or 1,000
Cancer types
Liver
Lung
Oral mucosa
Pancreas
Liver
Statistical significant tumors
(pairwise with controls or trend tests)
Liver: hepatocellular adenoma
Liver: cholangiocarcinoma
Lung: cystic keratinizing epithelioma
Oral mucosa: squamous cell carcinoma
Pancreas: adenoma or carcinoma
Liver: tumors
to
VO

-------
           Table 2-4. Animal bioassay studies selected for noncancer dose-response modeling
Reference
Species/
strain
Exposure
protocol
Sex
(exposure
group)
n
Average daily
dose levels
(ng/kg-day)
NOAEL
(ng/kg-day)
LOAEL
(ng/kg-day)
Endpoint(s)
examined
LOAEL/NOAEL
Endpoint(s)
Reproductive toxicity studies
Bowman et
al.Q989a; 1989b);
Schantz and
Bowman (1989):
Schantz et al.
(1992: 1986)
Franc et al. (2001)
Hochstein et al
(2001)
Hutt et al. (2008)
Monkey/
Rhesus
Rat/Sprague-
Dawley, Long-
Evans,
Han/Wistar
Mink
Rat/Sprague-
Dawley
Daily dietary
exposure in
female monkeys
(3. 5 -4 years)
Biweekly oral
gavage
(22 weeks)
Daily dietary
exposure
(132 days)
Oral gavage
(CDs 14 and 21,
postpartum days
7 and 14),
(Pups: once per
week for
3 months)
F (FO, Fl,
F2, F3)
Female
F
Female (FO
andFl)
3 to 7 (Fl)
8
12
3 (FO andFl)
0,0.12, or 0.67
0, 10, 30 or
100
0.03 (control),
0.8, 2.65, 9, or
70
0 or 7.14
None
10
None
None
0.12
30
2.65
7.14
Reproductive and
developmental
effects
Body weight,
relative liver
weight, relative
thymus weight
Reproductive
effects
Developmental
effects
Neurobehavioral
effects (e.g.,
discrimination-
reversal learning
affected)
Increased relative liver
weight in Sprague-
Dawley and Long-
Evans Rats; Increased
relative thymus weight
in Sprague-Dawley,
Han/Wistar, and
Long-Evans Rats
Reduced kit survival
Lower proportion of
morphologically
normal pre-
implantation embryos
during compaction
stage
to
I
OJ

o

-------
Table 2-4. Animal bioassay studies selected for noncancer dose-response modeling (continued)
Reference
Species/
strain
Exposure
protocol
Sex
(exposure
group)
n
Average daily
dose levels
(ng/kg-day)
NOAEL
(ng/kg-day)
LOAEL
(ng/kg-day)
Endpoint(s)
examined
LOAEL/NOAEL
Endpoint(s)
Reproductive toxicity studies (continued)
Ikeda et al. (2005)
Ishihara et al.
(2007)
Latchoumy-
candane and
Mathur (2002) and
related
Latchoumy-
candane et al.
(2003 , 2002a;
2002b)
Rat/ Holtzman
Mouse/ICR
Rat/Wistar
albino
Corn oil
gavage (initial
loading dose
followed by
weekly dose
during mating,
pregnancy, and
lactation-about
10 weeks)
Sesame oil
gavage (initial
loading dose
followed by
weekly doses
for 5 weeks)
Olive oil
gavage (daily
for 45 days)
F(FO)
FandM(Fl
andF2)
M(FO)
M
12 (FO)
Not specified
(FlandF2)
42 or 43
6
0 or 16.5
0,0.095, or 950
0, 1, 10, or 100
None
0.1
None
16.5
(maternal
exposure)
100
1
Reproductive and
developmental
effects
Reproductive
effects
Reproductive
effects
Decreased
development of the
ventral prostrate (Fl),
decreased sex ratio
(percentage of males)
(F2)
Decreased
male/female sex ratio
(percentage of males)
(Fl)
Reduced sperm
production, decreased
reproductive organ
weights
Reproductive toxicity studies (continued)
Murray et al.
(1979)
Rat/Sprague-
Dawley
Daily dietary
exposure
(3 generations)
F and M,
(FO)
F and M,
(FlandF2)
10-32 (FO)
22 (Fl)
28 (F2)
0, 1, 10, or 100
1
10
Reproductive and
developmental
effects
Decrease in fertility,
decrease in the
number of live pups,
decrease in gestational
survival; decrease in
postnatal survival,
decreased postnatal
body weight in one or
more generations

-------
           Table 2-4. Animal bioassay studies selected for noncancer dose-response modeling (continued)


Reference
Shi et al. (20071









Yang et al. (2000)







Species/
strain
Rat/Sprague-
Dawley








Rhesus
monkey/
Cynomolgus





Exposure
protocol
Maternal corn
oil gavage
(weekly on
CDs 14 and
21;PNDs7
and 14)
Offspring corn
oil gavage
(weekly for
1 1 months)
Fed gelatin
capsules
(5 days/week
for 12 months)



Sex
(exposure
group)
F(FO)
F(F1)








F








n
3(FO)
10 (Fl)








6 (treatment)
5 (controls)





Average daily
dose levels
(ng/kg-day)
0,0.14,0.71,
7.14, or 28.6








0,0.71, 3.57, or
17.86






NOAEL
(ng/kg-day)
0.14









17.86







LOAEL
(ng/kg-day)
0.71









None







Endpoint(s)
examined
Reproductive
effects








Endometriosis
effects






LOAEL/NOAEL
Endpoint(s)
Decrease serum
estradiol levels (Fl)








Increased endometrial
implant survival,
increased maximum
and minimum implant
diameters, growth
regulatory cytokine
dysregulation
Developmental toxicity studies
Amin et al. (2000)



Rat/Harlan
Sprague-
Dawley

Corn oil
gavage (GDs
10-16)

F(FO)



80-88 (Fl)



0, 25, or 100



None



25



Developmental
effects


Decreased preference
in the consumption of
0.25% saccharin
solution (Fl)
to
I
OJ

to

-------
          Table 2-4.  Animal bioassay studies selected for noncancer dose-response modeling (continued)


Reference
Bell et al. (2007b)












Franczak et al.
(2006)








Species/
strain
Rat/CRL:WI
(Han)











Rat/Sprague-
Dawley








Exposure
protocol
Maternal daily
dietary
exposure for
an estimated
20 weeks
(12 weeks
prior to mating
through
parturition)




Maternal corn
oil gavage
(CDs 14 and
21;PNDs7
and 14)
Offspring corn
oil gavage
(weekly for
8 months)
Sex
(exposure
group)
F(FO)
M(F1)











F (FO and
Fl)









n
65 (FO
treatments)
75 (FO
controls) at
study
initiation;
following
interim
sacrifice
-30 animals
were allowed
to litter; Fl on
PND 21 was
2 or 3 (FO)
7(F1)







Average daily
dose levels
(ng/kg-day)
0,2.4, 8, or 46












0,7.14, or 28.6









NOAEL
(ng/kg-day)
None












None









LOAEL
(ng/kg-day)
2.4
(maternal
exposure)










7.14









Endpoint(s)
examined
Reproductive and
developmental
effects










Developmental
effects








LOAEL/NOAEL
Endpoint(s)
Delayed EPS (Fl)












Decreased serum
estradiol levels (Fl)







Developmental toxicity studies (continued)
Hojo et al. (2002)
and related Zareba
et al. (2002)






Rat/Sprague-
Dawley







Maternal single
corn oil gavage
(GD8)

Offspring
exposed during
gestation and
lactation
(35 days)
F(FO)
FandM
(Fl)






12 (FO)
50 or 60 (Fl)







0, 20, 60, or 180








None








20 (maternal
exposure)







Developmental
effects







Abrogation of
sexually dimorphic
neuro -behavioral
responses (Fl)





to

-------
          Table 2-4.  Animal bioassay studies selected for noncancer dose-response modeling (continued)
Reference
Kattainen et al.
(2001)
Keller et al.
(2008a: 2008b:
2007)

Species/
strain
Rat/
Han/Wistar
and Long-
Evans
Mouse/
C57BL/6J,
BALB/cByJ,
A/J, CBA/J,
C3H/HeJ, and
C57BL/10J
Exposure
protocol
Maternal single
corn oil gavage
(GD 15)
Maternal single
corn oil gavage
(GD 13)
Sex
(exposure
group)
F(FO)
FandM
(Fl)
F(FO)
FandM
(Fla, b, c)
n
4 to 8 (FO)
3F/3M per
treatment
group (Fl)
Dams not
specified (FO);
23-36 (Fla);
4-5 (Fib);
107-1 10 (Flc)
Average daily
dose levels
(ng/kg-day)
0, 30, 100, 300,
or 1,000
0, 10, 100, or
1,000
NOAEL
(ng/kg-day)
None
None
LOAEL
(ng/kg-day)
30 (maternal
exposure)
10 (maternal
exposure)
Endpoint(s)
examined
Developmental
effects
Developmental
effects
LOAEL/NOAEL
Endpoint(s)
Reduced mesiodistal
length of the lower
third molar (Fl)
Variation in Ml
morphology in
C57BL/10J males and
females (Fla);
decreased mandible
shape and size in
C3H/HeJ males (Fib);
variation in molar
shape in C3H/HeJ
males (Flc)
(2008a; 2008b; 2007)
Developmental toxicity studies (continued)
Kuchiiwa et al.
(2002)
Li et al. (2006)
Markowski et al.
(2001)
Mouse/ddY
Mouse/NIH
(pregnant and
pseudo-
pregnant)
Rat/Holtzman
Maternal olive
oil gavage
(weekly for
8 weeks prior
to mating)
Maternal
sesame oil
gavage daily
for 8 days
(CDs 1-8)
Maternal single
olive oil
gavage
(GD 18)
F(FO)
M(F1)
F
F (FO and
Fl)
7(FO)
3 (Fl immuno-
cytochemical
analysis)
6 (Fl cell
number count)
10
4-7 (FO and
Fl)
0,0.7, or 70
0, 2, 50, or 100
0, 20, 60, or 180
None
None
None
0.7
(LOEL)
(maternal
exposure)
2
20
(maternal
exposure)
Neurotoxicity
Developmental
effects
Behavioral effects
Decreased serotonin-
immunoreactive
neurons in raphe
nuclei of male
offspring (Fl)
Decreased
progesterone and
increased serum
estradiol levels
Decreased training
responses (Fl)
to

-------
          Table 2-4.  Animal bioassay studies selected for noncancer dose-response modeling (continued)
Reference
Miettinen et al.
(20061
Nohara et al.
(2000)

Ohsako et al.
(2001)
Species/
strain
Rat/Line C
Rat/ Holtzman
Rat/ Holtzman
Exposure
protocol
Maternal single
corn oil gavage
(GD 15)
Maternal single
corn oil gavage
(GD 15)
Maternal single
corn oil gavage
(GD 15)
Sex
(exposure
group)
F(FO)
FandM
(Fl)
F(FO)
M(F1)
F(FO)
M(F1)
n
24-32
(treatment)
12-48
(controls)
Not specified
(FO)
5 males and
3 females (Fl)
6(FO)
5 males and
3 females (Fl)
Average daily
dose levels
(ng/kg-day)
0, 30, 100, 300,
or 1,000
0, 12.5, 50, 200,
or 800
0, 12.5, 50, 200,
or 800
NOAEL
(ng/kg-day)
None
800
(maternal
exposure)
12.5
(maternal
exposure)
LOAEL
(ng/kg-day)
30 (maternal
exposure)
None
50
(maternal
exposure)
Endpoint(s)
examined
Developmental
effects
Immunotoxicity
Developmental
effects
LOAEL/NOAEL
Endpoint(s)
Increase in dental
caries (Fl)
Decreased spleen
cellularity (Fl)
Decreased anogenital
distance (Fl)
Developmental toxicity studies (continued)
Schantz et al.
(1996)
Seo et al. (1995)
Simanainen et al.
(2004)
Sparschu et al.
(1971)
Rat/Harlan
Sprague-
Dawley
Rat/Sprague-
Dawley
Rat/TCDD-
resistant
Han/Wistar
sred with
TCDD-
sensitive Long-
Evans
Rat/Sprague-
Dawley
Maternal corn
oil gavage
(CDs 10-16
Maternal corn
oil gavage
(CDs 10-16)
Maternal corn
oil gavage
(CDs 15)
Maternal corn
oil gavage
(CDs 6-15)
F(FO)
FandM
(Fl)
F(FO)
M(F1)
F(FO)
~4 (FO);
80-88 (Fl)
-15 (FO);
5-9 (Fl)
5-8 (FO)
3 1 (controls)
10-14 (FO)
0, 25, or 100
0, 25, or 100
0, 30, 100, 300,
or 1,000
0, 30, 125, 500,
2,000, or 8,000
None
25
100
50
None
100
300
125
Developmental
effects
Developmental
effects
Reproductive
effects
Maternal toxicity;
Developmental
effects
Facilitatory effect on
radial arm maze
learning (Fl)
Decreased thymus
weight
Reduction in daily
sperm production and
cauda epididymal
sperm reserves
Decreased body
weight in dams and
male fetuses; fetal
intestinal hemorrhage
and subcutaneous
edema
to

-------
          Table 2-4.  Animal bioassay studies selected for noncancer dose-response modeling (continued)
Reference
Smith et al. (1976)
Species/
strain
Mouse/CF-1
Exposure
protocol
Maternal corn
oil gavage
(CDs 6-15)
Sex
(exposure
group)
F(FO)
n
14-41 (FO)
Average daily
dose levels
(ng/kg-day)
0, 1.0, 10, 100,
1,000, or 3,000
NOAEL
(ng/kg-day)
1,000
(maternal)
100
(fetal)
LOAEL
(ng/kg-day)
3,000
(maternal)
1,000
(fetal)
Endpoint(s)
examined
Teratogenic and
developmental
effects
LOAEL/NOAEL
Endpoint(s)
Increased relative liver
weight (FO dams);
increased incidence of
cleft palate (fetuses)
Developmental toxicity studies (continued)
Sugita-Konishi et
al. (2003)
Mouse/C57/6N
Cji
Maternal
drinking water
exposure (daily
for 17 -day
lactational
period)
F(FO)
FandM
(Fl)
8(FO)
Not specified
(Fl)
0, 1.14, or 11.3
1.14 (NOEL)
(maternal
exposure)
11.3(LOEL)
(maternal
exposure)
Immunotoxicity
Increased
susceptibility to
Listeria (Fl males and
females); increase in
thymic CD4+ cells
(Fl males); decreased
spleen weight
(Fl males)
Acute toxicity studies
Burleson et al.
(1996)
Crofton et al.
(2005)
Kitchin and Woods
(1979)
Mouse/B6C3F!
Rat/Long-
Evans
Rat/Sprague-
Dawley
Corn oil
gavage (single
exposure)
Corn oil
gavage
(4 consecutive
days)
Corn oil
gavage (single
dose)
F
F
F
20
14, 6, 12, 6, 6,
6, 6, 6, 6, and
4,
respectively,
in control and
treated groups
4 (treated);
9 (control)
0, 1, 5, 10, 50,
100, or 6,000
0,0.1,3, 10,30,
100, 300, 1,000,
3,000, or 10,000
0, 0.6, 2, 4, 20,
60, 200, 600,
2,000, 5,000, or
20,000
5
30
0.6 (NOEL)
10
100
2
(LOEL)
Immunotoxicity
Thyroid effects
Enzyme induction
Increased mortality
from influenza
infection 7 days after a
single TCDD
exposure
Reduction in serum
T4 levels
Increased
benzo(a)pyrene
hydroxylase (BPH)
Acute toxicity studies (continued)
Li et al. (1997)
Rat/Sprague-
Dawley
Corn oil dose
via oral gastric
intubation
(single dose)
F
10
0, 3, 10, 30,
100, 300, 1,000,
3,000, 10,000,
or 30,000
3
10
Hormonal effects
Increased serum FSH
(1997)
to

-------
          Table 2-4.  Animal bioassay studies selected for noncancer dose-response modeling (continued)


Reference
Lucier et al. (1986)









Nohara et al.
(2002)



Simanainen et al.
(2002)





Species/
strain
Ral/Sprague-
Dawley








Mouse/
B6C3FJ ,
BALB/c,
C57BL/6N and
DBA2
Ral/TCDD-
resistant
Han/Wistar
bred; TCDD-
sensitive Long-
Evans

Exposure
protocol
Corn oil
gavage or
TCDD-
contaminated
soil (single
dose)




Corn oil
gavage (single
dose)


Corn oil
gavage (single
dose)



Sex
(exposure
group)
F









M,F




M, F







n
6









10-40




9-11





Average daily
dose levels
(ng/kg-day)
0, 15, 40, 100,
200, 500, 1,000,
2,000, or 5,000
in corn oil

0, 15, 44, 100,
220,500,1,100,
2,000, or 5,500
in contaminated
soil
0, 5, 20, 100, or
500



30-100,000






NOAEL
(ng/kg-day)
None









500




100






LOAEL
(ng/kg-day)
15
(LOEL)








None




300






Endpoint(s)
examined
Enzyme induction









Mortality and
body-weight
changes


General
lexicological
endpoints, organ
weights, dental
defects


LOAEL/NOAEL
Endpoint(s)
Induction of aryl
hydrocarbon
hydroxylase (al low
dose in bolh trealmenl
protocols)





No increased mortality
of virus-infected mice
or Ireatmenl-relaled
changes in body
weighl
Reduction in serum
T4 levels




Acute toxicity studies (continued)
Simanainen et al.
(2003)





Smialowicz et al.
(2004)


Ral/TCDD-
resistant
Han/Wistar
sred with
TCDD-
sensitive Long-
Evans
Mouse/
C57BL/6N
CYP1A2 (+/+)
wild-type
Corn oil
gavage (single
dose)




Corn oil
gavage (single
dose)

M, F






F



5-6






Not specified



Line A:
30-3,000,000
LineB:
30-1,000,000
Line C:
30-100,000

0, 30, 100, 300,
1,000, 3,000, or
10,000

100






300



300






1,000



General
lexicological
endpoints, organ
weighls, denial
defecls


[mmunoloxicity



Decreased thymus
weighl





Decreased antibody
response to SRBCs


to

-------
           Table 2-4. Animal bioassay studies selected for noncancer dose-response modeling (continued)


Reference
Vanden Heuvel et
al. (1994b)



Species/
strain
Rat/Sprague-
Dawley



Exposure
protocol
Corn oil
gavage (single
dose)

Sex
(exposure
group)
F





n
5-15



Average daily
dose levels
(ng/kg-day)
0,0.05,0.1, 1,
10, 100, 1,000,
or 10,000


NOAEL
(ng/kg-day)
0.1 (NOEL)




LOAEL
(ng/kg-day)
1
(LOEL)



Endpoint(s)
examined
Liver effects




LOAEL/NOAEL
Endpoint(s)
Increase in hepatic
EROD activity and
CYP1A1 mRNA
levels
Acute toxicity studies (continued)
Weber et al. (1995)













Inbred Mouse/
C57BL/6





Inbred Mouse/
DBA/2





Corn oil
gavage (single
dose on Day 0)
Sacrificed on
Day8



Corn oil
gavage
(two doses on
Days -1 and 0)
Sacrificed on
Day8
M







M





4-7







4-7





0, 30, 100, 300,
1,000, 3,000,
9,400, 37,500,
75,000,
100,000,
133,00, or
235,000
0, 1,000,
10,000, 97,500,
375,000,
1,500,000,
1,950,000, or
3,295,000

1,000







10,000





3,000







97,500





Hepatic and renal
enzyme and
hormone
alterations; liver
and kidney
weight








Increased relative liver
weight












Subchronic toxicity studies
Chu et al. (20011






Chu et al. (20071


Rat/Sprague- (
Dawley (





Rat/Sprague- (
Dawley (

^orn oil gavage
daily for
!8 days)




^orn oil gavage
daily for
!8 days)
F






F


5






5


0, 2.5, 25, 250,
or 1,000





0, 2.5, 25, 250,
or 1,000

250






2.5


1,000






25


Body- and organ-
weight changes





Liver effects


Decreased body
weight, increased
relative liver weight
and related
biochemical changes,
decreased relative
thymus weight
Alterations in thyroid,
thymus, and liver
histopathology
to
I
OJ

oo

-------
           Table 2-4. Animal bioassay studies selected for noncancer dose-response modeling (continued)
Reference
Species/
strain
Exposure
protocol
Sex
(exposure
group)
n
Average daily
dose levels
(ng/kg-day)
NOAEL
(ng/kg-day)
LOAEL
(ng/kg-day)
Endpoint(s)
examined
LOAEL/NOAEL
Endpoint(s)
Subchronic toxicity studies (continued)
DeCaprio et al.
(1986)

DeVito et al.
(1994)
Fattore et al.
(2000)

Guinea pig/
Hartley
Mice/B6C3F!
Rat/Iva:SIV
50-Sprague-
Dawley
Daily dietary
exposure
(90 days)
Corn oil
gavage
(5 days/week
for 13 weeks)
Daily dietary
exposure
(13 weeks)
Daily dietary
exposure
(13 weeks)
Daily dietary
exposure
(13 weeks)
Daily dietary
exposure
(13 weeks, 26,
and 39 weeks)
M,F
F
M,F
M, F
M, F
F
10/sex
5
6
6
6
6
0,0.12,0.61,
4.9, or 26
(males); 0,0.12,
0.68, 4.86, or 31
(females)
0, 1.07, 3.21,
10.7, 32.1, or
107
0, 20, 200, or
2,000
0 or 200
0, 200, or 1,000
0 or 100
0.61
None
None
4.9
1.07 (LOEL)
20
Body- and organ-
weight changes
Body- and organ-
weight changes;
enzyme induction
Liver effects
Decreased body
weight (male and
females); increased
relative liver weights
(males); decreased
relative thymus weight
(males)
Increased EROD,
ACOH and
)hosphotyrosyl
proteins at all doses
Reduced hepatic
vitamin A levels
Subchronic toxicity studies (continued)
Fox et al. (19931
Rat/Sprague- (
Dawley r
d
\
1
javage loading/
naintenance
oses (every
days for
4 days)
M, F
6
0, 0.55, 307, or
1,607
0.57
327
Body- and liver-
weight changes;
hepatic cell
proliferation
Increased absolute and
relative liver weight
to
I
OJ

VO

-------
           Table 2-4. Animal bioassay studies selected for noncancer dose-response modeling (continued)
Reference
Hassoun et al.
(19981
Hassoun et al.
(2000)

Hassoun et al.
(2003)
Species/
strain
Mouse/
B6C3FJ
Rat/Harlan
Sprague-
Dawley
Rat/Harlan
Sprague-
Dawley
Exposure
protocol
Corn oil gavage
(5 days/week
for 13 weeks)
Corn oil gavage
(5 days/week
for 13 weeks)
Corn oil gavage
(5 days/week
for 13 weeks)
Sex
(exposure
group)
F
F
F
n
Not
specified
6
12
Average daily
dose levels
(ng/kg-day)
0, 0.32, 1.07,
10.7, or 107
0,2.14,7.14,
15.7, 32.9, or
71.4
0,7.14, 15.7, or
32.9
NOAEL
(ng/kg-day)
None
None
None
LOAEL
(ng/kg-day)
0.32 (LOEL)
2.14(LOEL)
7. 14 (LOEL)
Endpoint(s)
examined
Brain effects
Liver and brain
effects
Brain effects
LOAEL/NOAEL
Endpoint(s)
Induction of
biomarkers of
oxidative stress at all
doses
Induction of
biomarkers of
oxidative stress at all
doses in liver and
brain
Induction of
biomarkers of
oxidative stress at all
doses
Subchronic toxicity studies (continued)
Kociba et al.
(1976)
Mally and
Chipman (2002)
Slezak et al. (2000)
Rat/Sprague-
Dawley
Rat/F344
Mouse/
B6C3FJ
Corn oil gavage
(5 days/week
for 13 weeks)
Corn oil gavage
(2 days/week
for 28 days)
Corn oil gavage
(5 days/week
for 13 weeks)
M,F
F
F
12
3
Not specified
0,0.71,7.14,
71.4, or714
0,0.71, 7.14, or
71.4
0,0.11,0.32,
1.07, 10.7, or
107.14
7.14
None
1.07 (NOEL)
71.4
0.71 (LOEL)
10.7 (LOEL)
Liver effects,
body-weight
changes, and
hematologic and
clinical effects
Clinical signs and
histopathology
Liver, lung,
kidney, and
spleen effects
Reduced body weight
and food
consumption, slight
liver degeneration,
lymphoid depletion,
increased urinary
porphyrins and delta
aminolevulinic acid,
increased serum
alkaline phosphatase
and bilirubin
Decreased Cx32
plaque number and
area in the liver
Increased hepatic
superoxide anion
to
.u
o

-------
          Table 2-4.  Animal bioassay studies selected for noncancer dose-response modeling (continued)
Reference
Smialowicz et al.
(20081
Van Birgelen et al.
(1995a: 1995b)

Species/
strain
Mouse/
B6C3FJ
Rat/Sprague-
Dawley
Exposure
protocol
Corn oil gavage
(5 days/week
for 13 weeks)
TCDD in diet
(13 weeks)
Sex
(exposure
group)
F
F
n
8-15
8
Average daily
dose levels
(ng/kg-day)
0, 1.07, 10.7,
107, or 321
0, 14, 26, 47,
320, or 1,024
NOAEL
(ng/kg-day)
None
None
LOAEL
(ng/kg-day)
1.07
14
Endpoint(s)
examined
Immunotoxicity
and organ weight
Multiple end-
points
LOAEL/NOAEL
Endpoint(s)
Reduced antibody
response to SRBC,
increased relative liver
weight
Decreased absolute
and relative thymus
weights, decreased
liver retinoid levels
Subchronic toxicity studies (continued)
Vos et al. (1973)
White et al. (1986)
Guinea pig/
Hartley
Mouse/
B6C3FJ
Corn oil gavage
(weekly for
8 weeks)
Corn oil gavage
(daily for
14 days)
F
F
10
6-8
0, 1.14,5.71,
28.6, or 143
0, 10, 50, 100,
500, 1,000, or
2,000
1.14
None
5.71
10
Immunotoxicity
Immunotoxicity
Decreased total
leukocytes and
lymphocyte count,
decreased absolute
thymus and weight,
increase in primary
serum tetanus
antitoxin
Reduction of serum
complement activity
Chronic toxicity studies
Cantoni et al.
(1981)
Croutch et al.
(2005)
Rat/CD-
COBS
Rat/Sprague-
Dawley
Corn oil gavage
(weekly for
45 weeks)
Loading/
maintenance
dose (every
3 days for
different
durations up to
128 days)
F
F
4
5
0, 1.43, 14.3, or
143
0, 0.85, 3.4,
13.6, 54.3, or
217
(28-day
duration)
None
54.3
(28-day
duration)
1.43
217
(28-day
duration)
Hepatic porphyria
Body-weight
changes and
changes in
PEPCK activity
and IGF-I levels
Increased urinary
porphyrin excretion
Decreased body
weight, decreased
PEPCK activity, and
reduced IGF-I levels
to

-------
           Table 2-4. Animal bioassay studies selected for noncancer dose-response modeling (continued)
Reference
Hassoun et al.
(2002)
Species/
strain
Rat/Sprague-
Dawley
Exposure
protocol
Corn oil gavage
(5 days/week
for 30 weeks)
Sex
(exposure
group)
F
n
6
Average daily
dose levels
(ng/kg-day)
0,2.14,7.14,
15.7, 32.9, or
71.4
NOAEL
(ng/kg-day)
None
LOAEL
(ng/kg-day)
2.14(LOEL)
Endpoint(s)
examined
Brain effects
LOAEL/NOAEL
Endpoint(s)
Induction of
biomarkers of
oxidative stress at all
doses
Chronic toxicity studies (continued)
Hong et al. (1989)
Kociba et al.
(1978)

Maronpot et al.
(1993)
NTP (1982c)
NTP (2006a)
Rhesus
monkeys.
Rat/Sprague-
Dawley
Rat/Sprague-
Dawley
Mouse/
B6C3FJ;
Rat/Osborne
Mendel
Rat/Sprague-
Dawley
Daily dietary (4
years)
Daily dietary
exposure
(2 years)
Biweekly
gavage
(30 weeks)
Corn oil gavage
(2 days/week
for 104 weeks)
Corn oil gavage
(5 days/week
for 105 weeks)
F
M,F
F
M, F
F
7-8
50
9
50
53
0,0.12, or 0.67
0, 1, 10, or 100
0, 3.5, 10.7, 35,
or 125
0, 1.4, 7.1, or 71
for rats and
male mice; 0,
5.7, 28.6, or 286
for female mice
0,2.14,7.14,
15.7, 32.9, or
71.4
None
1
10.7
None
None
None
10
35
1.4
2.14
Immunotoxic
effects
Multiple
endpoints
measured
Body- and organ-
weight changes,
clinical
chemistry,
lepatocellular
proliferation
Liver and body-
weight changes
Liver and lung
effects
None
Increased urinary
porphyrins,
hepatocellular
nodules, and focal
alveolar hyperplasia
Increased relative liver
weight
Increased incidences
of liver lesions in mice
(males and females)
Increased absolute and
relative liver weights,
increased incidence of
hepatocellular
hypertrophy, increased
incidence of alveolar
to bronchiolar
epithelial metaplasia
to
.u
to

-------
            Table 2-4.  Animal bioassay studies selected for noncancer dose-response modeling (continued)
Reference
Species/
strain
Exposure
protocol
Sex
(exposure
group)
n
Average daily
dose levels
(ng/kg-day)
NOAEL
(ng/kg-day)
LOAEL
(ng/kg-day)
Endpoint(s)
examined
LOAEL/NOAEL
Endpoint(s)
Chronic toxicity studies (continued)
Sewall et al. (1993)
Sewall et al. (1995)
Toth et al. (1979)
Tritscher et al.
(1992)
Rat/Sprague-
Dawley
Rat/Sprague-
Dawley
Mouse/Swiss/
H/Riop
Rat/Sprague-
Dawley
Biweekly
gavage
(30 weeks)
Biweekly
gavage
(30 weeks)
Sunflower oil
gavage (weekly
for 1 year)
Initiated with
i.p. injection of
diethylnitrosami
ne (175 mg/kg)
or saline,
followed 2
weeks later by
biweekly TCDD
in corn oil
gavage (30
weeks)
F
F
M
F
9
9
38-44
At least 9 per
group
0, 3.5, 10.7, 35,
or 125
0,0.1,0.35, 1,
3.5, 10.7, 35, or
125
0, 1, 100, or
1,000
3.5, 10.7,35.7,
or 125
None
10.7
None
None
3.5
(LOEL)
35
1
None
EGFR kinetics
and auto-
)hosphorylation,
lepatocellular
proliferation
Thyroid function
Skin effects
CYP induction
Decrease in EGFR
maximum binding
capacity
Decreased serum T4
levels
Dermal amyloidosis
and skin lesions
None
to
     ND = not determined; ACOH = acetanilide-4-hydroxylase; EPS = balanopreputial separation; EGFR = epidermal growth factor receptor;
     EROD = 7-ethoxyresorufin-O-deethylase; FSH = follicle stimulating hormone; IGF = insulin-like growth factor; i.p. = intraperitoneal; PEPCK =
     phosphoenolpyruvate carboxykinase; PND = postnatal day.

-------
 2.4.1.  Key Epidemiologic Data Sets
       The studies listed in Tables 2-1 and 2-2, for cancer and noncancer, respectively, are those
studies that have met the epidemiologic TCDD study inclusion criteria (see Section 2.3.1).
Summaries for all of the epidemiologic studies evaluated are also provided in Appendix C and
are organized by epidemiologic cohort. Following a brief summary of each cohort, its associated
studies are then summarized chronologically, assessed for methodological considerations relative
to epidemiologic cohorts and studies, and evaluated for suitability for TCDD dose-response
assessment.  Further, Appendix C presents explicit details regarding whether the considerations
and criteria were met (see summary Tables C-2 and C-3, followed by Tables C-4 though C-57,
which provide details for each study).
       The cancer epidemiologic studies on TCDD that were subjected to the study selection
process include 24 peer-reviewed publications from 8 cohorts. An evaluation of these against
EPA's study inclusion criteria resulted in selecting 8 studies from the NIOSH, Boehringer,
BASF, Ranch Hand, and Seveso cohorts for further consideration in TCDD quantitative cancer
dose-response assessment (see Table 2-1).  All of these studies had  serum TCDD measurements
on individual study participants, used kinetic models to refine exposure estimates, and accounted
for latency or appropriate exposure windows in their analyses. As shown in Figure 2-4,  most of
the other studies were excluded because exposures were not primarily to TCDD and not
quantifiable on an individual level; many studies also failed to provide information on an
appropriate latency period or window of exposure for cancer (see Table C-2). In addition,
two studies (Steenland et al., 1999; Flesch-Janys et al.,  1998) passed all criteria but were not
selected because they were superseded by other studies on the same cohort for which an updated
analysis was done [i.e., Steenland et al. (2001) and Becher et al. (1998), respectively]. The
Baccarelli et al. (2006) study also passed all of the criteria but was not selected because of an
issue identified during evaluation of the study  considerations (i.e., lack of an obvious adverse
health endpoint).  The noncancer epidemiologic studies (see Table C-3) on TCDD  that were
subjected to the study selection process include 32 peer-reviewed publications from 10 cohorts.
An evaluation of these against EPA's study inclusion criteria resulted in selecting four studies
from the Seveso cohort for further consideration in TCDD quantitative noncancer dose-response
assessment (see Table 2-2).  The  4 Seveso cohort studies passed all criteria primarily because
TCDD serum levels were available for individuals in the studies, and the critical windows of

                                           2-44

-------
exposure were identifiable for the endpoints that served as PODs [e.g., the 9 months of
pregnancy for exposed mothers clearly defined the window of exposure for the fetus in
Baccarelli et al. (2008)1.  As shown in Figure 2-4, many of the excluded studies failed to provide
enough information on expected latency for the nonfatal endpoints or failed to provide data on
the critical period of exposure to quantitatively estimate an oral human dose.  A number of
studies also had exposures that were not primarily to TCDD.  One study, Baccarelli et al. (2005),
passed all criteria but was excluded because the health endpoint, chloracne, is considered to be
an outcome associated with high TCDD exposures; thus this study was not considered further in
RfD derivation. The Warner et al. (2004) study also passed all criteria but was not selected
because EPA could not assess the biological significance of this finding and could not establish a
LOAEL for this effect (i.e., it did not satisfy one of the study considerations).

 2.4.2.  Key Animal Bioassay Data Sets
       The studies listed in Tables 2-3 and 2-4, for cancer and noncancer,  respectively, are those
studies that have met the in vivo animal bioassay TCDD study inclusion criteria (see
Section 2.3.2 and Figure 2-3).  Appendix D provides study summaries,  is organized by
reproductive studies, developmental studies, and general toxicity studies (subdivided by
duration), and summarizes the experimental protocol, the results, and the NOAELs and LOAELs
EPA has identified for each study.  The doses shown in Tables 2-3 and 2-4 are expressed as
average daily administered intakes in units of nanograms per kilogram body weight per day
                                                          99  	
(ng/kg-day), adjusted for continuous exposure when necessary.    Tables D-l  and D-2 present
the results of the study selection evaluations for the studies that met and did not meet the study
inclusion criteria, respectively.
       A total of eight animal cancer bioassays were available for evaluation  using EPA's study
inclusion criteria (see Section 2.3.2 and Figure  2-3). Table 2-3  presents the 6 studies that met
these criteria and  are considered suitable for quantitative TCDD dose-response modeling. As
shown in Figure 2-4, only 2 of the available cancer bioassays did not meet EPA's study inclusion
criteria (and are not summarized in Appendix D). These include Eastin et  al. (1998) (genetically
22 Standard EPA guidance was applied for adjustment of intermittent gavage protocols and dietary exposures as
indicated in each specific study description in Appendix D.
                                           2-45

-------
altered mouse strain) and Rao et al. (1988) (intraperitoneal injection instead of oral route of
exposure).
       A total of 751 animal bioassays on a noncancer endpoint were available for evaluation
using EPA's study inclusion criteria (see Section 2.3.2 and Figure 2-3). As shown in Figure 2-4,
673 of the available noncancer studies were excluded based on one or more of the following
reasons: (1) 66 studies used genetically-altered animals; (2) 370 studies had a lowest tested dose
that was too high (i.e., greater than 30 ng/kg-day); (3) 142 studies tested chemicals that were not
TCDD only or used an unspecified TCDD dose; and (4) 135 studies did not use an oral dosing
method. Table D-2 of Appendix D shows these studies and identifies the study inclusion criteria
that were not met.  For many studies, more than one reason for exclusion was found and
identified.  Conversely, in some cases, at least one identified criterion was not met, and, given
the study was then excluded based on that one criterion, not all of the other criteria for exclusion
were further evaluated and articulated. Tables 2-4 and D-l of Appendix D present the 78 studies
that were selected as key data sets for TCDD noncancer dose-response analyses.
       In Section 4, additional evaluations are made to determine which study/endpoint data sets
are the most appropriate for development of the RfD for TCDD.  For further consideration in the
RfD derivation process,  only the toxicologically-relevant endpoints from the studies in Table 2-4
are carried  forward to Section 4 (see Section 4.2.1 and Appendix H for details on study/endpoint
combinations not used in RfD derivation for this reason).  For some entries in Table 2-4, there
are several  publications from the peer-reviewed literature shown in the same row of the table.  In
these  cases, the publications are grouped together because they are based on the same noncancer
animal bioassay. Additionally, in Table 2-4, the noncancer adverse effects in the animal studies
listed under the heading, "endpoints examined," are presented as general categories of effects,
such as "developmental effects," "liver effects," or "thyroid function." In  Section 4, more
detailed descriptors of the specific endpoints associated with such adverse health effects are
articulated  and evaluated to develop PODs for the derivation of an oral RfD for TCDD. Final
candidate study/endpoint data sets are selected in Section 4 based on factors such as
toxicological relevance of the endpoints (see Section 4.2.1 and Appendix H), dose-response
modeling results, and POD comparisons across studies, as illustrated in Figures 4-1 and 4-3 for
epidemiologic and toxicological data, respectively.
                                           2-46

-------
  3.  THE USE OF TOXICOKINETICS IN THE DOSE-RESPONSE MODELING FOR
                       CANCER AND NONCANCER ENDPOINTS
       A key recommendation from the NAS for improving the 2003 Reassessment was that
EPA should justify its approaches to dose-response modeling for cancer and noncancer
endpoints. Further, the NAS suggested that EPA incorporate the most up-to-date and relevant
state of the science for the TCDD dose-response assessment.
       While EPA believes that at the time of its release, the 2003 Reassessment offered a
substantial improvement over the general state-of-the-science regarding dose-response modeling,
EPA agrees with the NAS that the justification of the approaches to dose-response modeling can
be improved and the methodologies updated to reflect the most current EPA guidance (see Text
                                                                   	   ryy
Box 2-1) and science. In Section 3, EPA describes the use of toxicokinetic (TK)  information in
the dose-response modeling of TCDD.  Section 3.1 summarizes the NAS comments regarding
the use of TK in the dose-response approaches for TCDD.  Section 3.2 overviews EPA's
responses to the NAS comments. Section 3.3 discusses TCDD kinetics, including TK models
developed to simulate disposition of this compound in rodents and humans (see Section 3.3.4),
alternative measures of dose that could be used in a TCDD dose-response analysis (see
Section 3.3.4), and uncertainties in the TCDD dose estimates (see Section 3.3.5). Section 4 of
this document incorporates the TK information into noncancer dose-response modeling.

3.1. SUMMARY OF NAS COMMENTS ON THE USE OF TOXICOKINETICS IN
     DOSE-RESPONSE MODELING APPROACHES FOR TCDD
       The NAS commented on the appropriate use of TK models in dose-response modeling
for TCDD. Specifically, the committee requested that EPA consider using such  models to
provide refined estimates of dose, for example, as the underlying science and predictive
capabilities of these models improved.
       [Discussing Kinetic models].. .the committee encourages further development and
       use of these models as data become available to validate and further develop them
       (NAS. 2006b. p. 59).
23 Toxicokinetics (TK) is the branch of the pharmacokinetics (PK) that examines the disposition of toxins and
toxicants.
                                         3-1

-------
       Although the NAS agreed with EPA's use of body burden as a dose metric in the 2003
Reassessment (e.g., see NAS, 2006b, p. 7), the NAS was concerned about the limitations of
first-order kinetic models, such as the one used in the 2003 Reassessment, to estimate TCDD
body burdens.
       TCDD, other dioxins, and DLCs act as potent inducers of cytochrome P450
       (CYP), a property that can affect both the hepatic sequestration of these
       compounds and their half-lives. Hepatic sequestration of dioxin may influence
       the quantitative extrapolation of the rodent liver tumor results because the
       body-burden distribution pattern in highly dosed rats would differ from the
       corresponding distribution in humans subject to background levels of exposure.
       EPA should consider the possible quantitative influence of dose-dependent
       toxicokinetics on the interpretation of animal toxicological data (NAS, 2006b, p.
       129).
       The NAS also asked EPA to evaluate the impact of kinetic uncertainty and variability on

dose-response assessment.  The NAS committee asked EPA to use TK models to examine both

interspecies and human interindividual differences in the disposition of TCDD, which would

better justify EPA dose-response modeling choices.
       The Reassessment does not adequately consider the use of a PBPK model to
       define species differences in tissue distribution in relation to total body burden for
       either cancer or noncancer end points (NAS, 2006b, p. 62).

       EPA ... should consider physiologically based pharmacokinetic modeling as a
       means to adjust for differences in body fat composition and for other differences
       between rodents and humans (NAS, 2006b, p. 10).

       The Reassessment does not provide details about the magnitudes of the various
       uncertainties surrounding the decisions EPA makes in relation to dose metrics
       (e.g., the impact of species differences in percentage of body fat on the
       steady-state concentrations present in nonadipose tissues).  The committee
       recommends that EPA use simple PBPK models to define the magnitude of any
       differences between humans and rodents in the relationship between total body
       burden at steady-state concentrations (as calculated from the intake, half-life,
       bioavailability) and tissue concentrations.  The same model could be used to
       explore human variability in kinetics in  relation to elimination half-life. EPA
       should modify the estimated human equivalent intakes when necessary (NAS,
       2006b. p.73).

                                          3-2

-------
       Finally, the NAS asked EPA to use TK considerations to better justify its choice of dose
metric.
       EPA makes a number of assumptions about the appropriate dose metric and
       mathematical functions to use in the Reassessment's dose-response analysis but
       does not adequately comment on the extent to which each of these assumptions
       could affect the resulting risk estimates.. .EPA did not quantitatively describe how
       this particular selection affected its estimates of exposure and therefore provided
       no overall quantitative perspective on the relative importance of the selection
       (NAS. 2006b. p. 51).
3.2.  OVERVIEW OF EPA'S RESPONSE TO THE NAS COMMENTS ON THE USE OF
     TOXICOKINETICS IN DOSE-RESPONSE MODELING APPROACHES FOR
     TCDD
       In response to the NAS recommendations regarding TCDD kinetics and choice of dose
metrics, this document presents an in-depth evaluation of TCDD TK models, exploring their
differences and commonalities and their possible application for the derivation of dose metrics
relevant to TCDD. Initially, EPA discusses the application of first-order kinetics to estimate
body burden as a dose metric for TCDD. This first-order kinetic model is used to predict TCDD
body burden for all of the studies identified as Key Studies (see Section 2.4); this model uses a
constant half-life to simulate the elimination of TCDD from the body.  However, given the
observed data indicating early influence of cytochrome P450 1A2 (CYP1A2) induction and
binding to TCDD in the liver and later redistribution of TCDD to fat tissue, the use of a constant
half-life for TCDD clearance following long-term or chronic TCDD exposure is not biologically
supported.  Therefore, using half-life estimates based on observed terminal  steady state levels of
TCDD will not account for the possibility of an accelerated dose-dependent clearance of this
chemical during early stages following elevated TCDD exposures.  The biological processes
leading to dose-dependent TCDD excretion are better described using PBPK models than by
simple first-order kinetic models.  Additionally, as part of its preparation for developing this
document, EPA evaluated recent TCDD kinetic studies as NAS advocated.  Although the NAS
agreed with continued use of body burden metric as the dose metric of choice, EPA believes that
the state-of-the-practice has advanced sufficiently to justify the consideration of alternative dose
metrics (other than administered dose) based on an application of a physiologically based TK
model.

-------
       EPA identified a number of advances in the overall scientific understanding of TCDD
disposition; many of these are documented in a summary discussion introducing the section on
TCDD kinetics (see Section 3.3).  The increased understanding warranted an evaluation of
current kinetic modeling of TCDD to determine if the use of such models would improve the
dose-response assessment for TCDD.  Justification of the final PBPK model choice is detailed in
Section 3.3.  Through the choice of a published PBPK model to estimate dose metrics for dioxin,
EPA has addressed several  of the NAS concerns.  The PBPK model can be applied to estimate
dose metrics other than body burden that may be more directly related to response, e.g., tissue
levels, serum levels, blood  concentrations, or dose metrics related to TCDD-protein receptor
binding. The selected PBPK model included an explicit description of physiological and
biochemical parameters; therefore, it  can also provide an excellent tool for investigating
differences in species uptake and disposition of TCDD.  One of the criteria used to select a
PBPK model  for TCDD kinetics was the availability of both human and animal models so that
differences in species uptake and disposition of TCDD can be investigated.  Additionally, the
PBPK model  includes quantitative information that is suitable for addressing the impact of
physiological (e.g., body weight [BW] or fat tissue volume), or biochemical (e.g., induction of
CYP1A2) variability on overall risk of TCDD  between species, in response to another area of
concern in the NAS report.  The sensitivity analysis and uncertainty in dose metrics derived for
the health assessment of TCDD are also presented in Section 3.3.  A detailed discussion on the
uncertainty in choice of PBPK model-driven dose metrics is also provided in Section 3.3.

3.3.   PHARMACOKINETICS (PK) AND PK MODELING
3.3.1.  Pharmacokinetics  (PK) Data and Models in TCDD Dose-Response Modeling:
       Overview and Scope
       In general, the use of measures of internal dose in dose-response modeling is considered
to be superior to that of administered dose (or uptake) because the former is more closely related
to the response.  The evaluation of internal dose, or dose metric, in exposed humans and other
animals is facilitated by an  understanding of pharmacokinetics  (i.e., absorption, distribution,
metabolism, and excretion). When measurements of internal dose (e.g., blood concentration,
tissue concentration) are not available in animals and humans, pharmacokinetic models can be
used to estimate them.  The available data on the pharmacokinetics of TCDD in animals and
                                          5-4

-------
humans have been reviewed (NAS, 2006b: U.S. EPA, 2003; van Birgelen and van den Berg,

2000).

       It is evident based on these reviews and other analyses that three distinctive features of

TCDD play important roles in determining its pharmacokinetic behavior, as discussed below:
   •   TCDD is very highly lipophilic and thus is more soluble in fat or other relatively
       nonpolar organic media than in water.  The w-octanol/water partition coefficient is a
       commonly used measure of lipophilicity equal to the equilibrium ratio of a substance's
       concentration in w-octanol (a surrogate for biotic lipid) to the substance's concentration
       in water (Leoetal.. 1971). For TCDD, this coefficient is on the order of 10,000,000 or
       more (ATSDR, 1998).  It follows that the solubility of TCDD in the body's lipid fraction,
       i.e., the fatty portions of various tissues, including adipose, organs, and blood, is
       extremely high.

   •   TCDD is very slowly metabolized compared to many other organic compounds, with an
       elimination half-life in humans on the order of years following an initial period of
       distribution in the body (Michalek and Pavuk, 2008; Carrier et al., 1995a). Most
       laboratory animals used for toxicological testing tend to eliminate TCDD much more
       quickly than humans, although even in animals, TCDD is eliminated much more slowly
       than most other chemicals.

   •   TCDD induces binding proteins in the liver that have the effect of sequestering some
       of the TCDD.  The ability of TCDD to alter gene expression and the demonstration that
       the induction of CYP1A2 is responsible for hepatic TCDD sequestration suggest that
       both pharmacokinetic and pharmacodynamic events must be incorporated for a
       quantitative description of TCDD disposition (Santostefano et al., 1998).  The induction
       of these proteins implies that TCDD tends to be eliminated more rapidly in the early
       years following short-term, high-level exposures than it is after those initial levels have
       declined. Leung et al. (1988) and Andersen et al. (1993), in their PBPK modeling, have
       taken into consideration the issue of liver protein binding. Recent efforts of
       pharmacokinetic modeling have supported the concentration-dependent elimination of
       TCDD in animals and humans (Emond et al., 2006; Aylward et al., 2005b).
       Sections 3.3.2 and 3.3.3 present the salient features of TCDD pharmacokinetics in

animals and humans, respectively, with particular focus on mechanisms and data of relevance to

interspecies and intraspecies variability. Section 3.3.4 describes the various dose metrics for the

dose-response modeling of TCDD and the characteristics of pharmacokinetic models potentially

useful for estimating these metrics.  Finally, Sections 3.3.5 and 3.3.6 summarize uncertainty in

the dose estimate and the application of pharmacokinetic models associated with the predictions

of dose metrics used in dose-response modeling, respectively. Dose metrics derived via PBPK

                                          3-5

-------
modeling approaches are utilized in Section 4 of this document for noncancer TCDD
dose-response modeling.

3.3.2.  Pharmacokinetics (PK) of TCDD in Animals and Humans
3.3.2.1. Absorption and Bioavailability
       When administered via the oral route in the dissolved form, TCDD appears to be well
absorbed.  Animal studies indicate that oral exposure to TCDD in the diet or in an oil vehicle
results in the absorption of >50% of the administered dose (Olson et al., 1980; Nolan et al.,
1979). Human data from Poiger and Schlatter (1986) indicate that >87% of the oral dose (after
ingestion of 105 ng [3H]-2,3,7,8-TCDD [1.14 ng/kg BW] in 6 mL corn oil) was absorbed from
the gastrointestinal tract. Lakshmanan et al. (1986), investigating the oral absorption of TCDD,
suggested that it is absorbed primarily by the lymphatic route and transported predominantly by
chylomicrons.
       Oral absorption is generally less efficient when TCDD is more tightly bound in soil
matrices. Based on experiments in miniature swine, Wittsiepe et al.  (2007) reported an
approximately 70% reduction in bioavailability when TCDD was administered in the form of
contaminated soil, relative to TCDD after extraction from the same soil matrix with solvents.
Working with soil from the prominent contamination site at Times Beach, Missouri, Shu et al.
(1988) reported an oral bioavailability of approximately 43% based on experiments in rats.
Percent dose absorbed by the dermal route is reported to be less than the oral route, whereas
absorption of TCDD by the transpulmonary route appears to be efficient (Banks and Birnbaum,
1991) (see for example; Roy et al.. 2008: U.S. EPA. 2003: Diliberto  et al.. 1996: Nessel et al..
1992: Banks etal.. 1990).

3.3.2.2. Distribution
       TCDD in systemic circulation equilibrates and partitions into the tissues where it is then
accumulated,  bound, or eliminated. Whereas the bulk of the body tissues are expected to
equilibrate in a matter of hours, the adipose tissue will approach equilibrium concentrations with
blood much more slowly.  Consistent with these assertions, a number of experimental and
modeling studies in rats and humans have shown that TCDD has a large volume of distribution
(Vd), i.e., the apparent volume in which it is distributed. The Vd corresponds to the volume of

                                          3-6

-------
blood plus the product of internal tissue volumes and the corresponding tissue:blood partition
coefficients.  This parameter is a key determinant of the elimination rate of TCDD in exposed
organisms. The tissue:blood partition coefficients of TCDD, in turn, are determined by the
relative solubility of TCDD in tissue and blood components (including neutral lipids,
phospholipids, and water).
       Column 2 in Table 3-1 presents the tissue:blood partition coefficients for TCDD (Emond
et al., 2005; Wang et al., 1997). Column 3 of this table lists the physical volume of each tissue,
scaled to a person weighing 60 kg.  The last column shows the implications of the tissue volumes
and tissue:blood partition coefficients for the effective volumes of distribution for each tissue
and for the body as a whole.  It can be  seen that, purely on the basis of solubility space, the fat
should be expected to contain about 94% of the TCDD in the body, and that the body as a whole
behaves as if it is about 1,200 L in terms of blood-equivalents (i.e., approximately 22-fold larger
than its physical volume).
       Table 3-1. Partition coefficients, tissue volumes, and volume of distribution
       for TCDD in humans
Tissue
Blood
Fat
Jver
Rest of the body
Tissuerblood
partition
coefficient
1
100
6
1.5
Total
Tissue volume
(liters, for a
60-kg person)
O
11.4
1.56
38.64
54.6a
Effective volume of
distribution (Vd — liters of
blood equivalent)
O
1.140
9
58
1.210
Percent
total Vd
0.25
94.19
0.77
4.79
100.00
aThe total tissue volume presented here represents only 91% of body weight because some of the weight and volume
of the body is occupied by bone and other structures where TCDD uptake and accumulation do not occur to a
significant extent.
Source: Wang et al. (1997), Emond et al. (2006: 2005).
       Maruyama et al. (2002) have published another set of tissue:blood partition coefficients
for TCDD and other dioxin congeners based in part on observations of tissue concentrations
measured in autopsy specimens from eight Japanese people without known unusual exposures to
TCDD. Their estimates of TCDD partition coefficients seem to be rather large and variable,
                                            5-7

-------
with a fatblood value of 247 ± 78 (standard deviation [SD]), a liverblood value of 9.8 ± 5.7, and
a muscle:blood value of 18 ± 10.6. Depending on time of autopsy, tissue samples may not be an
accurate source of information on observed, in vivo partition coefficients because weight loss is
likely to occur pre and post mortem. In particular, a decline in the fat stores volume could lead
to an increased concentration of dioxin in fat in autopsy specimens relative to what would be
observed in vivo.
       The calculations shown in Table 3-1 do not include the additional amount that will be
bound to induced proteins in the liver.  That induction and binding will tend to increase the
contribution of the liver on the effective volume of distribution (Birnbaum, 1986).
       It is also of interest to point out some basic implications of the data in Table 3-1 for the
expected rates of perfusion-mediated transfer of TCDD between blood  and each of the
organ/tissues.  The rate of loss from a tissue (occurring primarily via blood flow) and the
corresponding half-life can be calculated using the following equations:
 „           „  ,    ,.    -K                  Blood flow (liters / hour)                 ,-c   ~,  ^
 Rate constant for loss (hour  ) =	-	-	  (Eq. 3-1)
                             Tissue volume (liters) x Tissue / Blood Partition Coefficent
                                             In (2)
            t1/2 for tissue perfusion loss =
                                      Rate constant for loss
                                                                                 (Eq. 3-2)
            _ ln(2) x Tissue volume (liters) x Tissue/Blood Partition Coefficent
                                 Blood flow (liters/hour)
       Because TCDD is highly lipophilic, its concentration in the aqueous portion of the blood
is very small, and TCDD tends to partition from blood components into cellular membranes and
tissues, probably in large part via diffusion. As a result, full equilibrium concentrations of
TCDD are not attained by the end of the transit time through organs from the arterial to venous
blood. For organs in which this occurs, diffusion coefficients or "permeability factors" have
been estimated to assess the fractional attainment of equilibrium concentration that occurs by the
time the blood leaving each organ reaches the venous circulation.  Table 3-2 presents the
permeability factors and implications for perfusion half-lives for TCDD, per Emond et al. (2006;
2005).

-------
       Table 3-2. Blood flows, permeability factors, and resulting half lives (tVz) for
       perfusion losses for humans as represented by the TCDD PBPK model of
       Emond et al. (2006; 2005)
Tissue
Fat
Jver
lest of the body
Permeability (fraction of
compartment blood flow)
0.12
0.03
0.35
Rate constant for
compartmental
elimination (hour-1)
0.0049
0.77
3.84
t1/! (hrs)
143
0.90
0.18
       Despite the high lipid bioconcentration potential of TCDD, the adipose tissue does not
always have the highest concentration (Abraham et al., 1988; Geyer et al., 1986; Poiger and
Schlatter, 1986). Further, the ratios of tissue:tissue concentrations of TCDD and related
compounds (e.g., the liveradipose ratio) may not remain constant during nonsteady-state
conditions.  TCDD concentrations have been observed to decrease more rapidly in the liver than
in adipose tissue.  For example, Abraham et al. (1988) found that the liveradipose tissue
concentration ratio in female Wistar rats exposed to a subcutaneous TCDD dose of 300 ng/kg
decreased from  10.3 at 1 day postexposure to 0.5 at 91 days postexposure. It should be noted
that even at a  ratio of 0.5, the amount of TCDD in the liver is greater than that based on lipid
content of the tissue alone, consistent with the presence of hepatic TCDD-binding proteins. The
liveradipose tissue concentration ratio also was dose-dependent, such that the liver TCDD
burden increased from ~11% of the administered dose at low doses (i.e., 1-10 ng/kg) to -37% of
the dose at an exposure level of 300 ng/kg. The increase in TCDD levels in liver, accompanied
by a decrease in concentration in the adipose tissue, is a particular behavior to be considered in
high-dose to low-dose extrapolations. This behavior is essentially  a result of dose-dependent
hepatic processes, as described below.

3.3.2.3. Metabolism and Protein Binding
       The metabolism of TCDD is slow, particularly in humans, and it is thought to be
mediated by the CYP1A2 enzyme that is inducible by TCDD (Weber etal.. 1997: Olson et al..
1994; Wendling et al., 1990: Ramsey et al., 1982).  The low rate of metabolism in combination
with sequestration appear to account for the retention of TCDD in liver, and these processes
collectively contribute to the long half-life for elimination of TCDD from the body.
                                          5-9

-------
       Dynamic changes in TCDD binding in liver and partitioning to adipose tissues have been
studied extensively in rats and mice (Diliberto et al., 2001; Diliberto etal., 1995). Figure 3-1
shows observations by Diliberto et al. (1995) of the ratio of liver concentrations to adipose tissue
concentrations for mice given doses spread over a 100-fold range and studied at four different
times following exposure. It can be seen that even for the lowest dose studied, the liveradipose
concentration ratio is higher than would be expected based on the lipid contents of the tissues
(i.e., 6:100, corresponding to the ratio of human liverblood and adipose:blood partition
coefficients; see Table 3-1). Moreover, the relative concentration in the liver consistently rises
with dose, with the steepest rise observed during the first 2 weeks after dosing.  If the
distribution of TCDD were governed solely by passive partitioning into  adipose, there should be
no such change in relative concentrations with dose. However, data  presented in Figure 3-1
illustrate that at longer time points,  the ratio of TCDD in the liver to  TCDD in adipose decreases,
indicating that a redistribution of the chemical occurs as time goes on for each applied dose. The
redistribution of TCDD tissue levels from liver to adipose with increasing time suggests that
binding of the  chemical in the liver (including via induction of CYP1A2) is an important kinetic
consideration at early exposure points with relatively high applied doses.
       Experiments with CYP1A2  "knock-out" mice (i.e., congenic strains differing in only a
single gene that is "knocked out" in one of the strains) indicate that the inducible binding of
TCDD is attributable to CYP1A2 (Diliberto et al.. 1999. 1997).  As noted previously, this
enzyme is believed to make an  important contribution to metabolism of TCDD. Given the
critical role of CYP1A2 induction in the kinetics of TCDD, dose-and time-dependent induction
of this protein  in rats has been examined and modeled (Emond et al., 2006, 2004; Santostefano et
al.,  1998; Wang et al., 1997). Accordingly, the amount of CYP1A2 in the liver can be computed
as the time-integrated product of inducible production and a simple first-order loss process
(Wang etal.. 1997):
                                           5-10

-------
         es
         0>


        J

        ^s

         09
         =
         es



         u
         u
         =
         o
        U

         o

        .o
        •M
         C8

        PS
                   3 -
                   2 -
D 7 Day Liver/Fat

• 14 Day Liver/Fat

A 21 Day Liver/Fat

• 35 Day Liver/Fat

                                                Dose jig/Kg
       Figure 3-1. Liver/fat concentration ratios in relation to TCDD dose at

       various times after oral administration of TCDD to mice.



       Source: Dilberto et al. (1995).
                                 dCYP.
                                      2A1 _
                                    dt
                                         = Sft)K0 -
                                                       ^^12<
                                                         (Eq. 3-3)
where CYP2A1 is the concentration of the enzyme, K2 is the rate constant for the first-order loss,


CA2t is the concentration of CYP1A2 in the liver, KQ is the basal rate of production of CYP1A2 in


the liver, and S(f) is a multiplicative stimulation factor for CYP1A2 production in the form of a


Hill-type function:
                                            5-11

-------
where IC&2 corresponds to the concentration of the aryl hydrocarbon (Ah)-TCDD complex at
which half of the maximum fold stimulation of CYP2A production is reached, and //, the Hill
exponent, determines the curvature of the stimulation in relation to concentration of the
Ah-TCDD complex at relatively low doses.  A value of 0.6 as the Hill exponent has been used by
Wang et al. (2000: 1997) and Emond et al. (2006: 2005: 2004). indicative of a negative
cooperation, i.e., the curve is convex-upward (supralinear), depicting a faster increase in the
low-dose region compared to a straight line. Additional parameters in this expression include
IriA2, the maximum fold increase in the CYP1A2 synthesis rate over the basal rate that can occur
at high levels of TCDD,  and (C-Ah-TCDD), the concentration of TCDD bound to the aryl
hydrocarbon receptor (AhR).  This concentration in turn depends on the concentration of TCDD
in the liver (C/,zy), the concentration of the AhR (Ah^,) in liver, and the dissociation constant for
the Ah-TCDD receptor complex, KDAH-


                                         Ahr xCrf
                              £      _  	u	L£_                             ,E  3_5x
                                A.H-TCDD    TS~     s~i                               \ i.'     '
3.3.2.4. Elimination
       Estimated elimination half-lives (i.e., the time taken for the concentration to be reduced
to one-half of its initial level) of TCDD range from 11 days in the hamster to 2,120 days in
humans (U.S. EPA, 2003). Hepatic metabolism and binding processes, fecal excretion, and
accumulation in adipose tissue collectively determine the dose-dependent elimination half-lives
in various species. Aylward  et al. (2005a) depicted the relationship between the elimination rate
versus initial level of lipid-corrected TCDD in serum for 36 people (see Figure 3-2). Even
though this analysis was done using the initial TCDD level, rather than the geometric mean or
midpoint level in the decline  for each person, it indicated a concentration-dependency of the
half-life and elimination of TCDD in exposed individuals.
                                          5-12

-------
                  0.3

                 C.25-

              ^ 0,2
              ^
              Uj'c.15-

                  0.1-

                 C.O'y
                     0        2rQOO      4,000      6,000      S:OOC  10,000
                           INITIAL SERUM LIPID TCDD LEVEL
                                             
-------
Lorenzen and Okev. 1991: Manchester et al.. 1987: Roberts et al.. 1986: Roberts et al.. 1985).
These qualitative similarities in pharmacokinetic determinants and outcome support the use of
animal data to infer general patterns of the pharmacokinetic behavior of TCDD in humans.
However, quantitative differences in determinants, including physiological, physicochemical,
and biochemical, need to be taken into account.  Even though species-specific physiological
parameters can be obtained from the literature, key data on species-specific biochemical
parameters (particularly binding constants, maximal capacity, induction rates, and other
parameters) are not available for humans at this time.  However, these can be inferred by using a
pharmacokinetic model fit to in vivo data on the rate of TCDD elimination from specific
compartments in humans (Emond et al., 2006: Aylward et al., 2005b: Emond et al., 2005: Emond
et al.. 2004: Carrier et al.. 1995a. b).

3.3.3.  Pharmacokinetics (PK) of TCDD in Humans: Interindividual Variability
       TCDD pharmacokinetics and tissue doses vary across the human population as a function
of the interindividual variability of the key kinetic determinants.  Because the NAS comments
focused on health effects associated with chronic, lifetime exposure, the key kinetic determinants
for such exposures include clearance, binding, and temporal changes in volume of distribution.
When considering the interindividual variability in pharmacokinetics and dose metrics of TCDD,
it is important to recognize that the elevated lipid-corrected serum concentrations in highly
exposed persons are associated with greater elimination rates, probably due to greater degrees of
induction of CYP1A2 in the liver and  possibly other related metabolic enzymes (Emond et al.,
2006: Aylward et al.. 2005b: Abraham et al.. 2002:  Grassman et al.. 2000).
       The interindividual variability  in adipose content is a critical parameter in
pharmacokinetic models given the characteristics of TCDD (see Section 3.3.2). Both metabolic
elimination and elimination via the GI tract depend on the fraction of TCDD in the body that is
available outside of adipose tissue.  As body fat content rises, a smaller portion of the total body
TCDD will be contained in the relatively available fraction outside of the adipose tissue.
Because elimination of TCDD by both metabolism and fecal excretion depends on the small
proportion of TCDD that exists outside of fat tissue, people with larger proportions of body
fat—including many older people—will tend to require longer times to reduce TCDD levels by a
                                          5-14

-------
given proportion than leaner people (Emond et al., 2006; Rohde et al., 1999; Van der Molen et
al.. 1998: Van der Molen et al.. 1996).
       The sections that follow highlight key aspects of interindividual variability in TCDD
pharmacokinetics, with an emphasis on the available data related to elimination half-lives and
volume of distribution.

3.3.3.1. Life Stage and Gender
       The influence of the variability of fat content in human population on the distribution and
clearance of TCDD has been evaluated by several investigators.  There are data showing an
inverse dependency  of TCDD  elimination rate on percent body fat. Figure 3-3 shows this
relationship in a study in which TCDD elimination via feces was measured in six people in
relation to their body fat content (Rohde et al.,  1999). Observations of TCDD elimination rates
in a small number of men and  women in the Seveso cohort (Aylward et al., 2005a) provide a
modest opportunity to compare TCDD elimination rates with actual  human data.  Based on the
partition coefficients reported  by Emond et al. (2006), the elimination rates for the men in the
sampled group are expected to be greater than the elimination rates in the women. Taking into
consideration values similar to those shown in Table 3-2, and fat proportions inferred from body
mass indices using the equations of Lean et al.  (1996), the Seveso men studied are expected to
have an overall average of about 3.92% of their TCDD body burden outside of fat, whereas the
women are expected to have an average of only 2.36% outside of fat. On this basis, the  TCDD
elimination rates in the men are expected to be 3.92/2.36 = 1.66 times faster than  the elimination
rates in the women.  By comparison, Michalek et al.  (2002) reported observed elimination rates
in men and women that result  in a slightly lower ratio:

                               111 year'1 +0.010 (std.error)
                               r> mi -,T^r.*-~i -i- n m n f^^A ^^m^\
  men:0.111 year'1 + 0.010 (std.error)
	—	= 1.56
women:0.071 year  ±0.010 (std.error)
The central estimates for the elimination rates correspond to half lives of 6.5 and 9.6 years for
men and women, respectively.
                                          5-15

-------
           0.05
           0.04 -
    u
    -s
    i_
    ta
           0.03 -
           0.02
                               y = 0.09735 - 0.00282x  RA2 = 0.752
               20
                         21
                                   22
                                              23
                                                        24
                                                                  25
                                                                            26
                                           % Body Fat
       Figure 3-3. Observed relationship of fecal 2,3,7,8-TCDD clearance and
       estimated percent body fat.
       Source: Rohde et al. (1999).
       A further point of comparison can be derived using the observed body mass index
(BMI)24 and TCDD elimination rate of each of the male Ranch Hand military veterans, whose
TCDD elimination rates were observed between 9 and 33 years after their time in Vietnam. The
average BMI over that time was 29.44 (based on 287 measurements for the 97 veterans,
tabulated in three periods by Michalek et al., 2002), and their average age was about 44.5 for the
24 The BMI is calculated as the body weight in kilograms divided by the square of the height in meters.
                                          3-16

-------
measurements.  Based on these data, the corresponding average estimated percent body fat is
29.7% using the Lean et al. (1996) formula for men.  The observed average TCDD elimination
rate constant for these men for the period was 0.092 year"1 ± 0.004 (standard error),
corresponding to a half-life of 7.5 years. This half-life is slightly longer than the central estimate
of the half-life of 6.2 years (i.e., ln(2)/0. Ill) for the smaller group of Seveso males with their
slightly smaller estimated percent body fat.  Figure 3-4 shows a simple plot of these data and a
fitted unweighted regression line characterizing the relationship between estimated fat content
and TCDD elimination rates.  Variation in metabolic enzyme activities and other routes of loss is
also likely to be important, but there is little human quantitative information available on these
issues.
a
 a
 a
'a
          Q
          Q
          U
          H
                 11.0
                 10.0
                  9.0
                  8.0 -
                  7.0
                  6.0
                  5.0
                       y =  - 1.89 + 0.314x  RA2 = 0.998
              Seveso Males
                               Error bars are ± 1 standard error
                    26
                             28
                                      30
                                              32
                                                       34
                                                                36
                                                                         38
                                       % Body Fat
       Figure 3-4.  Unweighted empirical relationship between percent body fat
       estimated from body mass index and TCDD elimination half-life—combined
       Ranch Hand and Seveso observations.
                                           5-17

-------
       More recently, Kerger et al. (2006) estimated the slope of the relationship between
half-life and age to be 0.12 years (95% confidence interval, 0.10-0.14), which corresponds to the
rate of increase in TCDD half-life for each year of age. The authors speculated that although age
explained most of the variance in the individual half-life trends, it was also correlated with
TCDD concentration, BMI, and body fat mass. The regression model developed by these
authors discriminated between the high and low TCDD exposures or concentrations.  Thus, after
accounting for the TCDD (concentration x age) term's effect on the slope of age, the final model
for TCDD concentration <700 ppt was

                                 ti / 2 = 0.35 + 0.12 x Age                         (Eq. 3-7)

       For TCDD concentration >700 ppt, the final model was

                                ti/2 = 0.35 + 0.088* Age                         (Eq. 3-8)

where ti/2 is the half-life and Age is the age at time of subsequent sampling. Pharmacokinetic
information relevant to specific age groups is presented in the sections that follow.

3.3.3.1.1.  Prenatal period
       Data to estimate TCDD elimination rates for fetuses are not available. Levels of TCDD
in fetal tissues for rats were experimentally estimated at different gestational periods and utilized
in a developmental model by Emond et al.  (2004). There is information on body composition
that is relevant to prediction of TCDD dose to fetus. These data, summarized as part of the
radiation dosimetry model of the International Commission on Radiological Protection, are
consistent with the idea that early fetuses are nearly all water and less than 1% lipid, and lipid
levels rise toward parity with protein near the time of normal delivery.
       Bell et al. (2007a) reported that the disposition of TCDD into the fetus shows dose
dependency, with a greater proportion of the dose reaching the fetus at lower doses of TCDD.
Further, both CYP1A1 and CYP1A2 are highly inducible (~103-fold) in fetal liver, whereas
CYP1A2 shows much lower induction (10-fold) in maternal liver.  It has been speculated that
this is due to the lower basal levels of CYP1A2 in fetal liver, as compared to maternal liver (Bell
                                          3-18

-------
et al., 2007a).  The greater relative disposition to the fetus at low doses may be the result of
higher bioavailability due to less hepatic sequestration and elimination in the mother.

3.3.3.1.2.  Infancy and childhood
       Hattis et al. (2003) describe the general pattern of change of body fat content with age in
children. Central tendency values for percent body fat begin at about 12% at birth and rise
steeply to reach about 26% near the middle of the first year of life.  Fat content then falls to reach
a minimum of approximately 15% at 5-8 years of age, followed by a sex-dependent "adiposity
rebound" that takes females to about 26% body fat while the males remain near 16-17% on
average by age 20. The interindividual variability distributions about these central values are
complex, as some children experience the "adiposity rebound" earlier than others, and this
creates patterns that are not simply interpretable as unimodal normal distributions. Hattis et al.
(2003) did find it possible to fit distributions of body fat content inferred from National Health
and Nutrition Examination Survey skin fold measures to mixtures  of two normal distributions for
children between age 5 and 18.
       At least two groups of authors have published PBPK modeling results indicating
generally more rapid clearance of TCDD in children than in adults, a trend that is consistent with
the generally lower fat content of children (Leung et al., 2006; Van der Molen et al., 2000;
Kreuzer et al., 1997). The rapid expansion of the adipose tissue compartment can contribute, in
part, to the reduced apparent half-life in children (Clewell et al., 2004).  This reduction may also
be due to varying rates of metabolism and/or  fecal lipid excretion (Kerger et al., 2007; Abraham
etal.. 1996).
       Furthermore, very young children have different modes and quantities of TCDD exposure
compared to adults. Lakind et al. (2000) characterize distributions of milk intake for nursing
infants to characterize distributions of TCDD exposure.  This is also a corresponding route of
loss of TCDD stores for lactating women, as described in Section 3.3.3.2 below.

3.3.3.1.3.  Adulthood and old age
       The fraction of fat in relation to body  weight in adulthood and old  age can be computed
as a function of the BMI and age (e.g.. Lean et al., 1996):
                                           5-19

-------
                    % Body Fat (males) = 1.33 x BMI + 0.236 x Age - 20.2            (Eq. 3-9)

                    %Body Fat (females) = 1.21 x BMI+ 0.262 x Age-6.1          (Eq. 3-10)

       The above equations are the result of analysis of data based on underwater weighing of
63 men and 84 women (age range 16.8-65.4).  The salient observation with respect to TCDD for
these data is that age and BMI-dependent variability in fat content have implications for the
variability in TCDD elimination rates and internal dose among adults.

3.3.3.2. Physiological States: Pregnancy  and Lactation
       Data on body fat content in pregnant women at various stages of gestation (Pipe  et al.,
1979) have potential implications for TCDD elimination rates during pregnancy, even though the
relationship between these parameters has  not been formally analyzed.
       Lactation is viewed as an additional route of elimination for some chemicals such as
TCDD. According to a recent study, a breast-feeding woman expels through lactation an
estimated 8.76 kg fat per year [^/(kg/day), 0.8 kg milk/day with an average 3% lipid], and the
partition coefficient between blood lipid and milk fat (K^u) for TCDD is 0.92 (Milbrath et al.,
2009; Wittsiepe et al., 2007). The estimated rate of elimination of TCDD due to breast-feeding
(kbfed) can then be computed as follows (Milbrath et al., 2009):
qf x At
 —
                                                 bfd
                                       KBMX

where
         fed (unitless)  = the fraction of the year during which the woman was actively
                        breast-feeding;
                      = woman's percent body fat; and
       BW            = woman's body weight in kg.

       Assuming no interaction between breast-feeding and other half-life determinants
Milbrath et al. (2009), the authors predicted a half-life of 4.3 years for TCDD in a 30-year-old,
                                          3-20

-------
nonsmoking woman with 30% body fat if she did not breast-feed that year, and a half-life of
1.8 years if she breast fed for 6 months.

3.3.3.3. Lifestyle and Habits
       One of the factors related to lifestyle and habits that could influence TCDD kinetics is
smoking.  Smoking has been reported to enhance the elimination of dioxin and dioxin-like
compounds (Ferriby et al., 2007; Flesch-Janys et al., 1996). Milbrath et al. (2009) accounted for
interindividual variation in body composition as well as smoking habits in an empirical model.
The predicted half-life (years) for an individual / as a function of age, smoking status, and
percent body fat / was as follows
where
        J(0age)
tm (age, smoke, pbf\ = [f3(Qage} + f3(age} x age, ]xSF,x
                                                                 Pbf,
       Pbf,
       PbJref(agei)
       SFt
                                                               Pbfr
  = intercept constant derived from regressed data;
  = slope constant derived from regressed data;
  = specific age /' (years);
  = individual percent body fat;
  = reference percent body fat; and
  = the unitless, multiplicative smoking factor.
                                                                               (Eq. 3-12)
3.3.3.4. Genetic Traits and Polymorphism
       One particular genetic locus that is potentially related to TCDD pharmacokinetics and
tissue dose is the gene for the AhR. Eight candidate AhR polymorphisms have been identified to
date (Connor and Aylward, 2006; Harper et al., 2002). Given the role of AhR in regulating the
induction of CYP1 isozymes (Connor and Aylward, 2006: Toide et al., 2003: Baron etal., 1998),
the polymorphism might lead to interindividual differences in metabolic clearance, the
significance of which would depend upon the dose, fat content, and exposure scenario.  In this
regard, it should be noted that the inducibility of aromatic hydrocarbon hydroxylase in human
                                          3-21

-------
tissues has been reported to be highly variable, up to 100-fold (Connor and Avlward, 2006;
Smart and Daly. 2000: Wong et al.. 1986).
       The scientific literature contains values of K& (the dissociation constant of the
TCDD-AhR complex) ranging from about 1 to much higher values (corresponding to lower
binding affinity) (reviewed in Connor and Avlward, 2006). This provides suggestive evidence
for a heterogeneous human AhR, with functionally important polymorphisms (Micka et al.,
1997; Roberts etal., 1986), even though some of the range may be attributed to experimental
procedural differences and to other factors (Connor and Avlward, 2006; Harper et al., 2002;
Lorenzen and Okey, 1991; Manchester et al., 1987).
       The various pharmacokinetic processes and determinants (see Sections 3.3.2 and 3.3.3),
individually or together, might influence the dose metrics of relevance to the dose-response
modeling of TCDD.

3.3.4.  Dose Metrics and Pharmacokinetic Models for TCDD
3.3.4.1. Dose Metrics for Dose-Response Modeling
       The dose metric related to a toxicological endpoint can range from the maximal
concentration, the area under a time-course curve (area under the curve [AUC]), or the
time-averaged concentration of the toxic moiety in the body, blood, or target tissue, to an
appropriate measure of the resulting interactions in the target tissue (e.g., receptor occupancy or
functional biomarkers related to specific effects). A single dose metric, however, is unlikely to
be sufficient for all endpoints and exposure durations. Consideration of these issues is critical to
the selection of the dose metrics of relevance to dose-response modeling of TCDD.
       Figure 3-5 lists a range of alternative dose metrics for TCDD in terms of their relevance
based on considerations of pharmacokinetic mechanisms and mode of action (MOA). The
administered dose or daily intake (ng/kg-day) is the least relevant dose metric for dose-response
modeling of TCDD. This dose adjusts only for body-weight differences between species. The
administered dose, when used with an uncertainty factor for kinetics (or kinetic adjustment
factor, such as BW3/4) and an uncertainty factor for dynamics, can also account for allometrically
predicted pharmacokinetic (clearance) and pharmacodynamic differences between species in
deriving the human equivalent dose (HED). In effect, the use of kinetic and dynamic adjustment
or uncertainty factors facilitates the computation of HED.  Such a calculation of HED is
                                          3-22

-------
                     OJ
                     u
                     c
                     CD
                      1_
                     CkO
                     c
                     U
                     c
Functional biomarkers
Receptor occupancy
Total tissue concentration
Blood or serum concentration
Body burden
Absorbed dose
Intake
       Figure 3-5. Relevance of candidate dose metrics for dose-response modeling,
       based on mode of action and target organ toxicity of TCDD.
associated with the steady-state blood concentration of parent chemical in rats by accounting for
species differences in metabolic clearance. This is generally done by relating to body surface
area or metabolic rates, with no corresponding temporal changes in the volume of distribution
(see, for example, Krishnan and Andersen, 1991).  Such calculations of HED for TCDD may not
be appropriate given that (1) steady-state was not attained in all critical toxicological studies
chosen for the assessment, (2) the clearance is mainly due to enzyme(s) and processes whose
levels/rates do not necessarily vary across species or life stages as a function of body surface
differences, and (3) there is a likelihood of change in volume of distribution over time.
Furthermore, the use of administered dose does not explicitly account for the dose-dependent
elimination of TCDD from tissues as demonstrated in multiple studies (reviewed in
Sections 3.3.2 and 3.3.4).  The use of administered dose in TCDD dose-response modeling is
unlikely to facilitate the characterization of the true relationship between the response and the
relevant measures of internal dose that are influenced by dose-dependent elimination and binding
processes. Additionally, the use of administered dose to extrapolate across species or life stages
                                          > oo
                                          5-23

-------
would not effectively take into account the differences in fat content or the demonstrated
dose-dependent and species-dependent differences in elimination half-life of TCDD.
       Dose metrics for TCDD may include absorbed dose, body burden, serum or whole blood
concentration, tissue concentration, and possibly functional-related metrics of relevance to the
MOA (e.g., receptor occupancy, change in protein levels). These measures can be calculated as
a current (terminal), average (over a defined period), or integral quantity. The applicability of
the integral measures, such as the AUC (i.e., the area under the curve of a plot of blood or
plasma concentration vs. time), traditionally used for analyzing chronic toxicity data, is
questionable in the case of TCDD.  This is because of differences in lifespan and uncertainties
regarding the appropriateness of the duration to be specified for averaging the AUC in
experimental animals and humans for certain critical effects (NAS, 2006b).
       Among the alternative dose metrics, the absorbed dose accounts for differences in body
weight as well as species-specific differences in bioavailability.  Thus, the absorbed dose is
equivalent to body burden. Body burden, or more appropriately, the body concentration,
represents the amount of TCDD per kg body weight.  TCDD body burdens, like other dose
measures, can be determined as the peak, the average over the period of the bioassays, or the
level at the end of the experiments. Thus, the terminal or average body burdens can be obtained
either using data or pharmacokinetic models and used in dose-response modeling.  The body
burden is a measure of TCDD dose that reflects the net impact of bioavailability, uptake,
distribution, and elimination processes in the organism. It is essentially a function of the volume
of distribution and clearance processes, and as such, it does take into account the temporal
changes in volume of distribution as well as the concentration-dependent clearance. These are
phenomena that are critical to the understanding of TCDD dose to the target. However, the body
burden may not accurately reflect the tissue dose (NAS, 2006b), and as such, does not allow for
analysis of species-specific differences in target organ sensitivity to TCDD. In essence, the body
burden represents only an "overall average" of TCDD concentration in the body, without regard
to the differential partitioning and  accumulation in specific tissues, including the target tissue(s).
       Serum (or blood) concentration of TCDD is a dose metric that reflects both the body
burden and the dose-to-target tissues.  Serum or blood concentration, at steady-state, would be
reflective of the impact of clearance processes  and expected to be directly proportional to the
tissue concentrations of TCDD (NAS, 2006b).  This dose metric for lipophilic chemicals such as
                                          3-24

-------
TCDD is often expressed as a lipid-normalized value, to adjust for varying serum lipid content
(Niskar et al., 2009; Patterson et al., 2009; DeKoning and Karmaus, 2000), particularly in human
biomonitoring studies, thus of relevance to dose-response modeling; however, the serum
lipid-normalized concentrations of TCDD are not routinely collected and reported in animal
toxicological studies. Serum lipid-adjusted TCDD concentration is calculated as the ratio of
serum TCDD content over serum lipid content per unit volume.  Alternatively, TCDD serum
lipid-normalized calculation can be estimated by using the formula TL = (2.27 x TC) + TG +
62.3 mg/dL where the total lipid (TL) content of each sample is estimated from its total
cholesterol (TC) and triglyceride (TG) (Patterson et al., 2009). The lipid-adjusted serum
concentration, however, would be reflective of the lipid-adjusted concentration of TCDD in other
organs (reviewed in Avlward et al., 2008) depending upon the extent of steady-state attained and
the similarity of lipid composition across tissues in each species. In essence, the serum
lipid-normalized measure is representative of the amount of TCDD per specified volume of total
lipids, whereas the whole blood measure will be reflective of the ensemble of free, lipid-bound
and protein-bound TCDD in plasma and erythrocytes, which may be species-specific. Even
though these dose metrics are thought to be more closely and directly related to the tissue
concentrations associated with an effect, a less direct association might occur at increasing doses
when nonlinear processes dominate the kinetics and distribution of TCDD into organs such as
the liver.
       Tissue concentration of TCDD, as free, bound, or total TCDD,  is a more relevant
pharmacokinetic measure of dose, given that it provides a measure of exposure of the target cells
to the chemical.  In this regard, the CYPlA2-bound fraction may be considered as a relevant
dose metric for certain toxic effects; however, the available data contain mixed results regarding
the mechanistic linkage of this dose metric to toxicity and carcinogenicity (reviewed in Budinsky
et al., 2006).  In such cases, the use of alternative dose metrics (e.g., bound concentration as well
as the serum concentration) in dose-response modeling could be considered. Other
function-related biomarkers and dose metrics could facilitate the additional  consideration of
pharmacodynamic aspects reflecting tissue- and species-specific sensitivity.  These metrics may
represent the most relevant measures of tissue exposure and sensitivity to TCDD.  For example,
receptor occupancy  and functional biomarkers as dose metrics for TCDD require a clear
                                          5-25

-------
understanding of mode of action of TCDD and availability of relevant data. In the absence of
such information, these possible dose metrics cannot be utilized at the present time.
       Empirical time-course data on the alternative dose metrics of TCDD associated with
epidemiologic and experimental (animal) studies are not available, requiring the use of
pharmacokinetic models to obtain estimates of these dose metrics. These models may be simple,
based on first-order kinetics or more complex based on physiochemical, biochemical, and
physiological parameters for simulating uptake, distribution (including sequestration to proteins),
and clearance of TCDD (see Section 3.3.4.3).

3.3.4.2. First-Order Kinetic Modeling
       Figure 3-6 illustrates the process of estimating a human-equivalent TCDD oral exposure
from an experimental animal-administered dose, based on the assumption that body burden is the
effective dose metric for TK equivalence across species. The primary assumption is that the
time-weighted average (TWA) TCDD body burden over some critical time period is the
                                                                  r\C 	
proximate toxicokinetically effective dose eliciting a toxicological effect.   The process consists
of estimating the effective average body burden in the experimental  animal over some time IA
(generally the experimental duration) using a TK model, then "back-calculating" a daily human
exposure level that would result in that average body burden over some time tH (the human
equivalent to /^).
       The following closed-form equation is the general formula used to calculate a TCDD
terminal body burden in  an experimental animal or human at time (f).
                                    = BB(G) +      e  )jfa                      (Eq. 3-13)
where
       BB(f)  = the body burden at time t (ng/kg);
       BB(0)  = the initial body burden (ng/kg);
       d      = the daily dose (ng/kg-day);
       k      = the whole-body elimination rate (days"1);
  The conversion depicted in Figure 3-6 does not account for toxicodynamic differences between species.
                                          3-26

-------
Experimental Applied Dose
                 Body Bur denRat(t} =
                                                       +

                                                                     Body Bur den Mt (t)
                                                                            A
Human
Estimated +
Exposure
?
              - <*A
                   \I2A
                             -, ^j^A
                   \I2H
                                                                    V
                                                                     Body BurdenH(t)
Figure 3-6. Process of estimating a human-equivalent TCDD lifetime average daily oral exposure (
-------
       t      = the time at which the body burden is determined (days); and
      fa     = the fraction of oral dose absorbed (unitless).
       For the experimental animal, BB(t) is BBA (t) = BBA (Q)e~kAtA +
                                                 i       du(\ -
and for humans, this parameter is BBH (t) = BBH (0)e~kHtH  + -^
                                                                 kH
Setting BBn(i) = BBA(i) obtains the following expression:
                                                                             (Eq
Rearranging and solving for du yields:
            dH =dA-UA     e        +BBA(0)e~kAtA -BBH(0)e-kHtH    (Eq. 3.15)
                     kA faH  l-e~k
1/2
Assuming that initial body burdens are very small compared to BBft) and that the fraction of
TCDD absorbed is the same for humans and experimental animals, and using the relationship
         , where ty2 is the whole-body half-life, a simplified solution for dn is obtained.
                                 *g=^7^n-g-*^                        (Eq-3'16)
                                        hl2H U   g    J
       The term l-e fais the daily fraction eliminated. Therefore, du can be seen to be the
average daily administered dose to the experimental animal times the ratio of the animal :human
half-life times the ratio of the animal :human daily fraction eliminated over the respective times,
IA and tn. For both species at (theoretical) steady state (t —> co; daily fraction eliminated —>• 1),
                                         5-28

-------
the latter ratio approaches unity, reducing the animal: human conversion factor to the ratio of the
half-lives.
       However, for less-than-lifetime exposures eliciting noncancer effects, specific values for
IA and tH must be considered. Furthermore, Eq. 3-16 computes dH on the basis of terminal body
burdens at times IA and IH.  The more representative metric for toxicokinetic equivalence based
on average body burden over the respective time periods is given in Eq. 3-17.
              i t
BB(t) = BB(0)-
                                             = BB(0)
                                                     1-*-*)
                                                        kt
                                                                      kt
(Eq. 3-17)
       Solving for Jin Eq. 3-17 by assuming minimal initial body burden (BB(0) ~ 0) and
setting d = d yields:
                     dH =dA
                              hi 2 A
                                               kAtA
                              tl/2H
                                     1-
                                         tH
                                                                            (Eq. 3-18)
where tno is the initial human exposure time.
       The value of IA is the duration of the experimental exposure period.  For some gestational
exposures,  if a critical exposure window is defined, IA will be the duration of the critical
exposure window.  The value of tH is the human-equivalent duration corresponding to IA-
However, for IA less than lifetime (less than 2 years in rodents) and no defined susceptible life
stage, tff cannot begin at 0 (because typically animal experiments do not begin at age 0), but must
end at 25,550 days (70 years) to include the terminal (pseudo) steady-state level, at which the
BB/fft): d/f ratio is highest.  Otherwise, starting tn at 0 would not be protective for
less-than-lifetime effects that could be manifest at any age in humans; the average is determined
from the terminal end of the human exposure period because the daily exposure achieving the
target blood concentration is smaller than for the same exposure period beginning at birth (i.e.,
dff would be higher for earlier exposure periods) and is health protective for effects occurring
                                           5-29

-------
                          r\r                                                	
after shorter-term exposure.   Figure 3-7 depicts the relationship of daily dose to TWA body
burden graphically for several exposure duration scenarios. For shorter durations occurring later
in life, the average body burden over the exposure period does not differ substantially from the
steady-state value. Even for half-lifetime exposures, the deviation of the average from steady
state is minimal. Only for lifetime exposures does the difference become more marked, but only
by about 15%. Note that in the 2003 Reassessment, a constant value of 3,000 was used for
BB/fft): dff, based on the relationship of continuous exposure to theoretical steady-state body
burden (t = lifetime, ty2 = 2,593 days); this approach, while conservative, does not account for
exposure scenarios of different durations and does not strictly reflect the average body burden
dose metric.
       The simulation in Figure 3-7 is based on a unit daily exposure to humans, such that the
target body burden represents BB^t^'-dn as a general scalar for calculating dH from any given
dA.  Table 3-3 shows the resulting TK conversion factors for the rodent species and strains
comprising the bulk of the experimental animals in TCDD studies.  Monkey and mink values are
not shown in this table because, for the former, only chronic exposures were evaluated and, for
the latter, no TCDD half-life information is available. Monkey (Rhesus) half-life estimates
range from about 200-500 days.  A representative value of 365 days is used for this TCDD
assessment. The d.A to dn conversion factor for the chronic monkey exposures (3.5-4 years) in
TCDD studies is 9.2-9.7 (BBA:dA = 279-263).
       Application of first-order kinetics for the health assessment of TCDD can only be used to
estimate total body burdens or back-calculate administered dose from experimental data.  Body
burden calculations using first-order kinetics is based on the assumption of a first-order decrease
in the levels of administered dose as function of time. In that sense, any loss of TCDD from the
body is described by using a rate constant that is not specific to any biological process. This
constant is usually estimated from estimates of half-life of TCDD. Assuming a constant half-life
value for the clearance for long-term or chronic TCDD exposure is not biologically supported
given the observed data indicating early influence of CYP1A2 induction and binding to TCDD
and later redistribution of TCDD to fat tissue.  Abraham et al. (1988) found that the liveradipose
tissue concentration ratio in female Wistar rats  exposed to a subcutaneous TCDD dose
of300 ng/kg decreased from 10.3 at 1 day postexposure to 0.5 at 91 days postexposure.
26 See the following (Section 3.3.4.3) for a more detailed discussion of this concept.
                                          3-30

-------
O
m
    o
    o
    o
    CO
    o
    o
    10
    CM
    o
    o
    o
    CM
    o
    o
    10
    o
    o
    o
    o
    o
    10
    o -
     targetbody burden
     chronic exposure (BB:d = 2555)
     half-chronic exposure (BB:d = 2202)
     shorter exposures (BB:d = 2185)
                        5000
10000           15000

     Exposure days
20000
25000
      Figure 3-7.  Human body burden time profiles for achieving a target body
      burden for different exposure duration scenarios.
      BB:d is BBH(tH):dH in Figure 3-6. The curve depicted using the solid line illustrates the increase in
      the human body burden over time for a hypothetical human administered a daily TCDD dose
      where the time-weighted average human body burden estimate over the lifetime is equal to the
      target body burden attained in a rodent bioassay. When compared to shorter durations (dashed
      lines), a higher average daily TCDD dose is required to yield a time-weighted average human
      body burden over a lifetime that is equal to the target body burden attained in a rodent bioassay.
      The half-chronic exposure scenario  (depicted using a dashed line) is equivalent to a 1-year
      exposure in rodents. When compared to a chronic BBH, a lower value of dH is needed to attain the
      target body burden in a rodent bioassay when the time-weighted average is over the last 35 years
      of life; the dose-to-plateau ratio is also smaller (i.e., dHtC < dH:SC to attain the target body burden in
      a rodent bioassay).  The shorter exposure scenario is equivalent to most other shorter rodent
      exposure durations, from 1 day to subchronic, which are indistinguishable with respect to the
      BB:d ratio (subchronic shown).
                                              5-31

-------
       Table 3-3.  Toxicokinetic conversion factors for calculating human equivalent
       doses from rodent bioassays based on first-order kinetics
Half-life (days)a
Exposure
duration (days)
1
7
14
28
90
180
365
730
Mouse
10
Rat (Wistar)
20
Rat (other)
25
Guinea pig
40
Conversion factor (CF)b BBA(tA)'-dA given in parentheses
3,882 (0.77)
1,107(2.71)
681 (4.41)
453 (6.62)
307 (9.76)
282(10.6)
270(11.1)
226(11.3)
3,815 (0.79)
1,020 (2.94)
587(5.11)
350(8.56)
186(16.1)
154(19.5)
141 (21.3)
115(22.2)
3,802 (0.79)
1,004 (2.99)
569 (5.27)
331 (9.06)
163 (18.4)
129(23.2)
115(26.0)
93 (27.4)
3,783 (0.79)
979(3.07)
543 (5.53)
303 (9.90)
130(23.0)
93 (32.1)
77 (38.9)
60 (42.5)
       aHalf-life for humans = 2,593 days (7.1 years).
       bdH = dA/CF; BBH(tH):dH = 2,185 (1-180 days), 2,202 (365 days), 2,555 (730 days).
Consequently, using half-life estimates based on observed steady-state levels of TCDD will not
account for the possibility of accelerated dose-dependent clearance of the chemical at the early
stages and, thus, would result in estimation of lower administered levels of the chemical. The
dynamic change in half-life due to dose-dependent elimination at the early stages of TCDD
exposure and its later redistribution to fat tissues for steady-state levels is better described using
biologically based models, such as the PBPK models and concentration- and age-dependent
elimination (CADM) models (Emond et al., 2006; Aylward et al., 2005b: Emond et al., 2005;
Emond et al., 2004; Carrier et al., 1995a, b). Additionally, these models provide estimates  for
other dose metrics (e.g., serum  or tissue levels) that are more biologically relevant to response
than administered dose or total  body burden (see Section 3.3.4.3).
3.3.4.3. Biologically Based Kinetic Models
       The development and evolution of biologically based kinetic models for TCDD have
been reviewed by EPA (2003) and Reddy et al. (2005). The initial PBPK model of Leung et al.
(1988) was developed with the consideration of TCDD binding to CYP1A2 in the liver.  The
next level of PBPK models by Andersen et al. (1993) and Wang et al. (1997) used
diffusion-limited uptake and described protein induction by interaction of DNA-binding  sites.
The models of Kohn et al. (1993) and Andersen et al. (1997)  further incorporated extensive
                                          3-32

-------
hepatic biochemistry and described zonal induction of CYP by TCDD.  TCDD PBPK models
have evolved to include detailed descriptions of gastrointestinal uptake, lipoprotein transport,
and mobilization of fat, as well as biochemical interactions of relevance to organ-level effects
(KohnetaL 1996: Rothetal.. 1994). Subsequently, developed PBPK models either used
constant hepatic clearance rate (Maruyama et al., 2002; Wang et al., 2000; Wang et al., 1997) or
implemented varying elimination rates as an empirical function of body composition or dose
(Van der Molen et al.. 2000: Van der Molen et al.. 1998: Andersen et al.. 1997: Kohn et al..
1996: Andersen et al., 1993). The more recent pharmacokinetic models explicitly characterize
the concentration-dependent elimination of TCDD (Emond et al., 2006: Aylward et al., 2005b:
Emond et al.. 2005: Emond et al.. 2004: Carrier et al.. 1995a, b). The biologically based
pharmacokinetic models describing the concentration-dependent elimination (i.e., the
pharmacokinetic models of Emond et al., 2006: Aylward et al., 2005b: Emond et al., 2005) are
relevant for application to simulate the TCDD dose metrics in humans and animals exposed via
the oral route.  The rationale for considering the Aylward et al. (2005b) and Emond et al. (2006:
2005: 2004) models for estimating dose metrics for possible application to TCDD health
assessment is based on the following considerations.
   •   Both models were developed and calibrated using research results from the more recent
       peer-reviewed publications.
   •   Both models are relatively simple and less parameterized than earlier kinetic models for
       TCDD. The Aylward et al. (2005b) model is based on two-time scale TCDD kinetics
       described by Carrier et al. (1995a), and the Emond et al. (2006: 2005: 2004) PBPK
       models are reduced versions of earlier complex PBPK models. Although simple, both
       the Aylward et al. (2005b) and Emond et al. (2006: 2005: 2004) models are inclusive of
       important kinetic determinants of TCDD disposition.
   •   Both models are uniquely formulated with dose-dependent hepatic elimination consistent
       with current understanding of TCDD toxicokinetics.
   •   Both models and extrapolated human versions were tested against human data collected
       in a variety of human exposure scenarios (Aylward et al., 2005b: Emond et al., 2005).
   •   Both models are capable of deriving one or more of the candidate dose metrics that may
       be of interest to EPA's dose-response assessment of TCDD.
                                         J-J

-------
3.3.4.3.1.  Concentration- and age-dependent model (CADM)
3.3.4.3.1.1. Model structure
       The pharmacokinetic model of Aylward et al. (2005b), referred to as CADM in this
report, is based on an earlier model developed by Carrier et al. (1995a, b) that describes the
dose-dependent elimination and half-lives of polychlorinated dibenzo-p-dioxins and furans.  This
model describes the TCDD levels in blood (body), liver, and adipose tissue. Blood itself is not
characterized physically as a separate compartment within the model, and the distribution of
TCDD to tissues other than adipose tissue and liver (usually less than 4%) is not accounted for
by the model.  The original structure of the Carrier et al. (1995a, b) model  was modified by
Aylward et al. (2005b) to include TCDD elimination through partitioning from circulating lipids
across the lumen of the large intestine into the fecal content (see Figure 3-8).  The most recent
version of the Carrier model (2008; Aylward et al., 2005b) includes fecal excretion of TCDD
from two routes: (1) elimination from circulating blood lipid through partitioning into the
intestinal lumen; and (2) elimination of unabsorbed TCDD from dietary intake.





LBSORPTION







DISTRIBUTION
Concentration-Dependant
I ]
| Liver Burden |
i_
aco=a(o*/*(Q) r
* r
-i 	 J-,
i i
	 fe| T31 1 i
H Blood |
1 1
, 	 ,
T_ J
1
| Adipose Burden j
|a(o=a(o*[i-/*(cj]r














ELIMINATION
Tissue-Specific
/'Hepatic metabolisnix
K with first-order rate J
^-^_constant ke^^^






f Fecal excretion ^\
^l with the first-order j
\^rate constant A:^^/
"^ 	 	 -^"^


^










       Figure 3-8.  Schematic of the CADM structure.
       Source: Aylward et al. (2005b).
                                           5-34

-------
       A basic assumption of this model is that metabolic elimination of TCDD is a function of
its current concentration in the liver. The current concentration of TCDD in the liver increases
with increasing body burden in a nonlinear fashion as a result of the induction of (and binding of
TCDD to) specific proteins (i.e., CYP1A2). Consequently, the fraction of TCDD body burden
contained in the liver increases nonlinearly (with a corresponding decrease in the fraction
contained in adipose tissues) with increasing body burden of TCDD (Aylward et al., 2005a:
Carrier et al.. 1995a).
       Of particular note is that the adipose tissue compartment of the model is considered to
represent the lipid contained throughout the body.  It then assumes that the concentrations of
TCDD in lipids of plasma and various organs are essentially  equivalent to that of adipose tissue,
and as such, these concentrations are included in the adipose  compartment of the model.  Even
though this approximation is fairly reasonable given the available data, there is some concern
that the adipose compartment of this model also includes the  lipid content of the liver to some
unknown extent. Because the equilibrium balance between free  and bound TCDD in the liver is
dependent on the adipose content of the tissue, removal of lipid volume from the liver would
mathematically alter total hepatic concentration and, therefore, would affect the estimated levels
of the chemical available for binding to proteins.
       Distribution  in the body is modeled to occur between  hepatic and adipose/lipid
compartments, with the fraction of body burden in liver increasing according to a function that
parallels the induction of the binding protein CYP1A2.  Elimination is modeled to occur through
hepatic metabolism  (represented as a first-order process with rate constant K that decreases with
age) and through lipid-based partitioning of unmetabolized TCDD across the intestinal lumen
into the gut, which is also modeled as a first-order process. As the body burden increases, the
amount of TCDD in the liver increases nonlinearly, resulting in an increased overall elimination
rate.

3.3.4.3.1.2.  Mathematical representation
       The CADM  model describes the distribution to tissues (including liver and adipose
tissue) based on exchange from blood at time intervals of 1 month. The model is based on
quasi-steady-state-approximation, and, thus, it is also based on the consideration that the
intertissue processes reach their equilibrium values "quasi-instantaneously."  In this regard,

                                          3-35

-------
absorption and internal distribution reflective of kinetics at the cellular level (e.g., diffusion,
receptor binding, and enzyme induction) likely occur on a relatively fast time scale (a few hours
to a few days). However, the overall body concentration (i.e., body burden) varies slowly with
time such that it remains virtually unchanged during short time intervals.
       The CADM model does not differentiate between binding to AhR and CYP1A2, and it
lacks explicit descriptions of CYP1A2 induction, a key determinant of TCDD kinetics.
However, the empirical equation in the CADM model is based on five parameters (i.e., fmin, fmax,
K, Wa, and Wi; see Tables 3-4 and 3-5) that allow the successful description of the behavior of
TCDD in liver and adipose tissue (i.e., TCDD half-lives in each compartment increase with
decreasing body burden). This observation implies that the model adequately accounts for the
ensemble of the processes. Essentially, the CADM model describes the rate of change in tissue
concentrations of TCDD as a function of total body burden such that the global elimination rate
decreases with decreasing body burden or administered  dose.

3.3.4.3.1.3. Parameter estimation
       The CADM model is characterized by its simplicity and fewer parameters compared to
physiologically based models. Reflecting this simplicity, hepatic extraction is computed with a
unified empirical equation that accounts for all relevant  processes (i.e., protein induction and
binding).
       The key parameters (fmin, fmax, K, and ke) were all obtained by fitting to species-specific
pharmacokinetic data.  The physiological parameters (such as tissue weights) used in the model
are within ranges documented in the literature. The fat content is described to vary as a function
of age, sex, and BMI.  However, the BMI of the model is not allowed to change during an
individual simulation (which can range from 20 years to 70+ years), when in reality, the
percentage of fat in humans changes over time. None of the TCDD-specific parameters were
estimated a priori or independent of the data set simulated by the model.
                                          5-36

-------
    Table 3-4. Equations used in the concentration and age-dependent model
    (CADM; Avlward et al., 2005bV
    Parameter
                                             Equation
Hepatic
Concentration
(ng/kg)

Fat
Concentration
(ng/kg)
                   s-i
                      _ z^body •      ,  ^-/ max   Jrmn-'   body
                  hepatic   -rrr
                         W
                                ( f   ,
                                V-/ rmn
                                            -   s-i
                                             + L
                                                body
                 *~i       z^body $• /"\  f r
                  adipose    -rTr    V  W min
                          Wa
                                           ^J max  J rmn /    body
                                               K + C
                                                            ))
                                                     body
Hepatic
Elimination

Excretion via
gut of
Unchanged
TCDD
(Exsorption)
                   r^    7    ^-    7.y-~»   sk/i/^
                  Exr _hepattc =ke* Qbody * (1 - (/mn
                                                    ( f   - f
                                                    V-* max   -* min
                                                                  body
                                                             ^•body
                  Exr _gut = ka*Qa
  Change of
  TCDD due to
  bodyweight
  change
                                                  BW(i)
  Amount in
  body as a
  function of
  time
                                    = Exr _hepatic + Exr _ gut + ChangeTCDD _BW
Adipose tissue
growth
                      =
                              + (0.23* Age) -10.8* sex
                                   100
  Change of
  hepatic
  elimination
  constant with
  age
aFor abbreviations and parameter descriptions, see Table 3-5.
                                         5-37

-------
       Table 3-5.  Parameters of the concentration and age-dependent model
       (CADM; Avlward et al., 2005b)
Parameter
f a
Lhmin
f a
Lhmax
Ka
ke
ke0
^-e slope
\r
^e mm
wa (adipose weight fraction)
wh (liver body weight fraction)
ka (adipose clearance factor)
Monthly dose
Estimated absorption fraction
Body weight
Sex
Time of administration
Initial Cbody
Absorbed monthly dose 1
Value
0.01
0.7
100
Calculated
0.85
0.011
0.2
Calculated
0.03
0.0025
0.15507069
0.97
70
1
840
0.2
0.150418569
Units
unitless
unitless
ng/kg
per year
per year
per year
per year
unitless
unitless
per month
ng
unitless
kg
unitless
months
ng/kg
ng
Comments/sources
Minimum body burden fraction in liver
Maximum body burden fraction in liver
Body burden at half -maximum of fraction
iver
k. = keo - ke siope * (age) with enforced
minimum of ke mm
CADM-mean hepatic elimination base rate at
ageO
Change in ke per year of age
Vfinimum hepatic elimination rate
wa = [(1.2*BMI)+0.23*Age-10.8*sex]/100
Assumed constant
Passive elimination rate from intestinal tract
)er month
7rom Moser and McLaghlan (2001)
Standard male weight
[ = male; 0 = female

Estimated background young adults UMDES
sampling
)er month
aThe values of f^, 4^, and K were obtained by best fit of the model simulations to the experimental data with the
 method of least squares (Avlward et al. 2005a: Carrier et al.. 1995a).
3.3.4.3.1.4. Model performance and degree of evaluation
       The CADM model was not evaluated for its capabilities in predicting data sets not used
in its parameterization. In other words, one or more of the key input parameters (fhmin, fhmax, ke,
K) was obtained essentially by fitting to the species-specific pharmacokinetic data,  such that
there was no "external" evaluation data set to which the model was applied. Despite the lack of
emphasis on the "external" evaluation aspect, the authors (Aylward et al., 2005a: Carrier et al.,
1995a, b) have demonstrated the ability of the model to describe multiple data sets covering a
range of doses and species.
       The visual comparison of the simulated data to experimental values suggests that the
model could, to an approximate  degree, correctly reproduce the whole set of data (e.g.,
pharmacokinetic [PK] profile over a range of dose and time) and not just part of the PK curve,
essentially with the use of a single set of equations and parameters.
                                          5-38

-------
       The pharmacokinetic data sets for TCDD that were used to calibrate the CADM model by
Aylward et al. (2005a; Carrier et al., 1995a, b) included the following:
       Adipose tissue and liver concentrations of TCDD following a single oral dose of 1 |ig/kg
       in monkeys (McNulty et al., 1982):
       Percent dose retained in liver for a total dose of 14 ng in hamsters (Van den Berg et al.,
       1986):
       Elimination kinetics of TCDD in female Wistar rats following a single subcutaneous dose
       of 300 ng/kg (data from Abraham et al.. 1988):
       Liver and adipose tissue concentrations (terminal measurements) in Sprague-Dawley rats
       given 1, 10, or 100 ng TCDD/kg bw per day for 2 years (Kociba et al., 1978): and
       Serum lipid concentrations of TCDD over a period of several years  in 54 adults (29 men
       and 25 women) from Seveso and in three Austrian patients (Avlward et al., 2005a).
       For illustration purposes, Figure 3-9 shows model simulations of rat data from Carrier
et al. (1995a).  Figure 3-2 (see Section 3.3.2.4) depicts the human data that were used by the
authors to support the concentration-dependent elimination concept; the model was
parameterized to provide adequate fit to these data (Aylward et al., 2005a).
       The authors did not report any specialized analyses that quantitatively evaluated the
uncertainty, sensitivity, and/or variability of CADM model parameters and structure.

3.3.4.3.1.5. Confidence inconentration- and age-dependent elimination (CADM) model
           predictions of dose metrics
       Using professional judgment, EPA ranked its confidence in the CADM model as low,
medium, or high (or not applicable) based on model simulations of administered dose, absorbed
dose, body burden, serum lipid concentration, total tissue (liver) concentration, and receptor
occupancy. A  qualitative level of confidence associated with the predictability and reliability of
absorbed dose and body burden for oral exposures in humans (as well as several animal species)
by this model can be ranked as high (see Table 3-6). This model, however, does not account for
                                          5-39

-------
     0
     3
    _Q
        0.5
     c
     o
    'o
     03
                              10
   100
Cb
1000
Figure 3-9. Comparison of observed and simulated fractions of the body
burden contained in the liver and adipose tissues in rats.
fb, fraction contained in liver (observation) (n);fh-sim, fraction contained in liver (simulation) (—);
/at, fraction contained in the adipose tissue (observation) (0); /rsim, fraction contained in the
adipose tissue (simulation) (—); and Cb body concentration in ng TCDD/kg body wt.

Source: Carrier et al. (1995a): data from Abraham et al. (1988) measured 7 days after dosing.
                                       5-40

-------
       Table 3-6. Confidence in the CADMa model simulations of TCDD dose
       metrics15
Dose metric
Administered dose
Absorbed dose
3ody burden
Serum lipid concentration
Total tissue (liver) concentration
Receptor occupancy (bound concentration)
Level of confidence
NA
H
H
M
L
NA
        H = high, M = medium, L = low, NA = not applicable.
        ""Concentration and age-dependent model (Aylward et al. 2005b).
        bUsing professional judgment, EPA ranked its confidence in the CADM model as low, medium,
         or high (or not applicable) based on model simulations of administered dose, absorbed dose,
         body burden, serum lipid concentration, total tissue (liver) concentration, and receptor
         occupancy.
the differential solubility of TCDD in serum lipids and adipose tissue lipids, nor does it account
for the diffusion-limited uptake by adipose tissue. Due to these limitations, the confidence
associated with the predictions of the serum lipid concentration of TCDD is considered medium,
particularly when it is not documented that steady-state is reached during the critical
toxicological studies and human exposures.  Furthermore, the CADM model does not facilitate
the computation of TCDD concentrations in specific internal organs (other than liver and adipose
tissue). The reliability of this model for simulating the liver concentration (free, bound, or total)
of TCDD at low doses is considered to be low.  This low confidence level is a result of the
uncertainty associated with the key parameter, /hmin-  This parameter needs to be recalibrated for
each study/species/population to effectively represent the free fraction of TCDD in liver and the
amount of TCDD contained in the hepatic lipids and bound to the liver proteins (whose levels
might be reflective of background exposures of various sources: see Carrier et al., 1995a). The
uncertainty related to the numerical value of this parameter in animals and humans—particularly
at very low exposures—raises concern regarding the use of this model to predict TCDD
concentration (free, bound, or total) in liver as the dose metric for dose-response modeling.
Although the use of the parameter /hmax permits the prediction of the dose to liver at high doses,
it does not  specifically facilitate the simulation of the amount bound to the protein or level of
induction in liver.  Because the CADM model is not capable of simulating enzyme induction
based on biologically relevant parameters, its reliability for predicting the concentration of
                                           3-41

-------
TCDD bound specifically to the AhR is not known.  Finally, due to the lack of parameterization
or verification with kinetic data in pregnant, lactating, or developing animals or humans, the
CADM model is unlikely to be reliable in the current form for use in predicting potential dose
metrics for these lifestages or study groups that might form the basis of PODs for the assessment.

3.3.4.3.2.  Physiologically basedpharmacokinetic (PBPK) model
3.3.4.3.2.1. Model structure
       Emond et al. (2006, 2004) simplified the eight-compartment rat model of Wang et al.
(1997) to a four-compartmental developmental model (liver, fat, rest of body, and placenta with
fetal transfer) (Emond et al., 2004), and later to a three-compartment adult model (liver, fat, rest
of the body) (Emond et al., 2006) (see Figures 3-10 and 3-11). Their rationale for simplification
of the model was based on evaluating, critiquing, and improving all earlier PBPK models by
Wang et al. (1997). In general, the main reason for the simplification was that extrapolation of a
PBPK model to  humans with these many (i.e., eight compartments) compartments would be
problematic due to the limited availability of relevant human data for validation (Emond et al.,
2004). One major difference from earlier models, repeatedly emphasized by Emond et al. (2006;
2005), was their description (included in their simplified PBPK models) of the dose-dependent,
inducible elimination of TCDD.  The rationale for including TCDD binding and induction of
CYP1A2 into the model was earlier described by Santostefano et al. (1998).
       The most recent version of the rat and human PBPK models developed by Emond et al.
(2006) describes the organism as a set of three compartments corresponding to physiological
tissues—liver, fat, and rest of the body—interconnected by systemic circulation (see
Figure 3-10).  The liver compartment includes descriptions of CYP1A2 induction, which is
critical for simulating TCDD sequestration in liver and dose-dependent elimination of TCDD. In
this model, the oral absorption of TCDD from the gastrointestinal (GI) tract accounts for both the
lymphatic (70%) and portal (30%) systems.
                                         5-42

-------
 Urinary
excretion
    L_
                                Blood systemic circulation
  GI tract
elimination
   Oral
absorption
                                                       Cellular matrices
                              	I
                                  Liver                  Blood tissue !*
                                                       Cellular matrices


                                                       Cellular matrices
                                   Fat                   Blood tissue *•
                              	1
                                                       Cellular matrices


                                                       Cellular matrices
                               Rest Of body            Blood tissue
                                                       Cellular matrices
   Figure 3-10. Conceptual representation of PBPK model for rat exposed to
   TCDD.
   Source: Emond et al. (2006).

-------
     Urinary
     excretion
           T3
            O
             Elimination Gl tract
                                     Oral absorption
                                                              Portal vein
                            Liver (AhR and CYP1A2 induction)
                                           Fat
                                       Rest of body
                                     Placenta (AhR)
                                                                             (D
                               O
                               O
                               Q.
                                          Fetus
      Figure 3-11.  Conceptual representation of PBPK model for rat
      developmental exposure to TCDD.
      Source: Emond et al. (2004).
      The biological relationship between TCDD "sequestration" by liver protein and its
"elimination" by the liver is not entirely clear. TCDD is metabolized slowly by unidentified
enzymes. CYP1A2 is known to metabolize TCDD based on studies in CYP1A2 knockout mice
(Diliberto et al., 1999, 1997), in which the metabolic profile is different compared to wild-type
mice.  However, because several metabolites appear in the feces of CYP1A2 knock out mice, it
                                        5-44

-------
is assumed that there are other enzymes involved in TCDD metabolism. TCDD binds to AhR
and induces not only CYP1A2, but also CYP1A1, CYP1B1, and several UDP-
glucuronosyltransferase and transporters (Gasiewicz et al., 2008). Both hydroxylated and
glucuronidated hydroxyl metabolites are found in the feces of animals treated with TCDD (Hakk
et al., 2009). Because the exact enzymes involved with TCDD are unknown and yet the
metabolism is induced by TCDD, an assumption of increased elimination rate of TCDD in
proportion to the induction of CYP1A2 is made. In the PBPK model, CYP1A2 is also needed
because TCDD binds to rat, mouse, and human CYP1A2 (Staskal et al., 2005; Diliberto et al.,
1999).  Thus, CYP1A2 induction is necessary to describe TCDD pharmacokinetics due to TCDD
binding. Hence, CYP1A2 can be used as a marker of Ah-receptor induction of "TCDD
metabolizing enzymes." Other models use AhR occupancy as a marker of induction of "TCDD
metabolizing enzymes" (Kohn et al., 2001; Andersen et al., 1997).
       Figure 3-11 depicts the structure of the rat developmental-exposure PBPK model (Emond
et al., 2004). This model was developed to describe the relationship between maternal TCDD
exposure and fetal TCDD concentration during critical windows of susceptibility in the rat. In
formulating this PBPK model, Emond et al. (2004) reduced the original 8-compartment model
for TCDD in adult rats by Wang et al. (1997) to a 4-compartment (i.e., liver, fat, placenta, and
rest of the body) model for maternal rat.  Activation of the placental compartment and a  separate
fetal compartment occurs during gestation (Emond et al., 2004).

3.3.4.3.2.2.  Mathematical representation
       The key equations of the PBPK model of Emond et al. (2004) are reproduced in
Text Boxes 3-1 and 3-2, whereas those from Emond et al. (2006; 2005) are listed in Table 3-7.
The rate of change of TCDD in the various tissue compartments is modeled on the basis of
diffusion limitation considerations.  Accordingly, mass balance equations are used to compute
the rate of change in the tissue (i.e., intracellular compartment) and tissue blood (i.e.,
extracellular compartment). The membrane transfer of TCDD is computed using a permeation
coefficient-surface area cross product (PA) for each tissue.  Metabolism and binding of TCDD to
the AhR and inducible hepatic protein (CYP1A2) are described in the liver.  The total mass in
the liver is then apportioned between free dioxin (Qip) and bound forms of TCDD (see
Figure  3-12).  The dose- and time-dependent induction of hepatic CYP1A2 in the liver is

                                        3-45

-------
Table 3-7. Equations used in the TCDD PBPK model of Emond et al. (2006)
Aspect
Body-weight
growth with age
Cardiac output
Blood
compartment
Equation
BW (S) = BW roxf OAlxtime }
uwtime(s) *" -J()<\u025 + timej
(BWT5
Drf-ml 1 hr\ nrTA'Rvf\(\\
UoooJ
A factor of 60 corresponds to the conversion of minutes to hours, and 1,000 is used for the
conversion of BW from grams to kilograms.
\(Qf x Cft>} + (Qre x Creb) + (Qli x Clib) + lymph! (Cb x CL URl]
f^hdiiiml 1 vnT \ —
Qc Qc
Tissue compartment (fat, rest of the body)
Tissue blood
subcompartment
Tissue cellular
matrices
Liver tissue com
Tissue blood
subcompartment
Tissue cellular
matrices
Free TCDD
concentration in
liver
Concentration
sound to AhR in
hepatic tissue
(nmnJ'mT\ Dt(Cn Cfh\ PAt\Cth
dt 1 Pt)
f^thf viwinl / wiT\ —
Wib
(vtmnl 1 TII T \ PAf\ f^t~h
\ lUtlUi 1 IllLi } — L2Ll\ Ufct/
dt J ( Pt)
At
Ct(nmollmL} = —
Wt
partment
lnmnl/mr\ nJifCn CJih\ P^TTfCJih riHrrr\ 1 inrnii
dt
WLIB
tnmrtl 1 mJ\ PAJJtClih Cliff pp\ ( KJiJJ F J 1 v Cliff pp v WJ J\
All
Cli(nmol / mL) = 	
V ' Wli
_,.- , ,. ... _.. [_..- nTT (LIBMAXxClifree} ( CYPU2 x Clifree \\
i nfvTT'i itwinf 1 wiT \ i/i infiTT' •" PT T I '
[ ^ KDLI + Clifree ) ( KDLIU2 + Clifree )\
,, rx LIBMAXxClifree
Ct (nmnl mJ\ —
^lAWboimd\'L"lul ' llij^) TST^T T m-f
KDLI + Clifree
All other induction processes and equations have been described and presented by Wang et al.
(1997X
                                5-46

-------
      Table 3-7. Equations used in the TCDD PBPK model of Emond et al. (2006)
      (continued)
Aspect
Equation
Gastrointestinal absorption and distribution
Amount of
TCDD remaining
in lumen cavity
Amount of
TCDD eliminated
in the feces
Absorption rate
of TCDD to the
blood via the
lymphatic
circulation
Absorption rate
of TCDD by the
liver via portal
circulation
dLumen [(
dt ^
Lumen in the amount of TCD
TCDD during a subchronic e:
(Ttmnl I ~hr\ ff^^T

rfLy^/2
dt
dPortal , , . , N ^ ,
(nmnl ' hr^\ — fC 4 1
dt
of TCDD to the portal lymphatic circulation
ST + KABS^ x lumen] + intake
D remaining in the GI tract (nmol); intake is the rate of intake of
•cposure (nmol/hr).
x lumen
S x lumen x 0.7
3S x lumen x 0.3
Note: Key parameters and abbreviations are defined in Table 3-8.
    Q
    C
          Tissue blood
a
Plasma
proteins
           Water
           fraction
             TCDD-Ah
                          Nonspecific
                             bound
CYP1A2-TCDD


           Tissue
                             Q
      Figure 3-12. TCDD distribution in the liver tissue.

      Source: Wang et al. (1997).
                                        5-47

-------
described per Wang et al. (1997) and Santostefano et al. (1998).  Accordingly, the amount of
CYP1A2 in the liver was computed as the time-integrated product of inducible production and a
simple first-order loss process (Wang et al., 1997):
                                 at
                                                      A2t
                                                                             (Eq. 3-19)
In this expression, CYPjA2 is the concentration of the enzyme (nmol/g), K2 is the rate constant for
the first-order loss (hour l), CA2t is the concentration of CYP1A2 in the liver (nmol/g), KQ is the
basal rate of production of CYP1A2 in the liver (nmol/g/hr), and S(f) (unitless) is a multiplicative
stimulation factor for CYP1A2 production in the form of a Hill-type function (see
Section 3.3.2.3):
                             S(t) = 1
                                         InA2(CA
                   (Eq. 3-20)
where, S(f) is the stimulation function, IriA2 is the maximum fold of CYP1A2 synthesis rate over
the basal rate, CAH-TCDD is the concentration of AhR occupied by TCDD, and ICA2 is the
Michaelis-Menten constant of CYP1A2 induction (nM). The dose-dependent or variable
elimination of TCDD was described using the relationship:
                  KBILE LI =
                              CYP\A2mduced-CYP\A2basal
                                     CYPIA2
                                             'basal
Kelv
(Eq. 3-21)
where CYPlA2indUced'^ the concentration of induced CYP1A2 (nmol/mL), CYPlA2basai is the
basal concentration of CYP1A2 (nmol/mL), and kelv is the interspecies constant adjustment for
the elimination rate (hour l).
       There are various ways of formulating the dose-dependent elimination as a function of
the level of CYP1A2, and the above equation (used by the authors) can be viewed as one means
of describing this behavior quantitatively. The numerator in the equation above will always be

                                         3-48

-------
greater than zero when there is TCDD in the system (including TCDD derived from either
background exposures or defined external sources). Consequently, the rate of elimination will
correspond to a nonzero value for situations involving TCDD exposures.
       It should be noted that CYP\A2indUced should always be greater than CYPl A2^aM/for any
CYP1A2-mediated elimination to take place in Eq. 3-21. This will always be the case whenever
TCDD is present in the liver because the induced levels of CYPl A2 are an estimate of "total"
enzyme content at any time point including basal levels.  Furthermore, Eq. 3-21 is a
mathematical representation of the induced elimination rate of TCDD by the liver that is
numerically influenced by the scalable parameter kelv. Hence, the mathematical description for
the elimination of TCDD by the liver is dominated by the level of CYPl A2 induction (as
mathematically influenced by the Hill coefficient in Eq. 3-20) and the numerical estimation of
the kelv constant.  The interrelationship between the induction Hill coefficient (h in Eq. 3-20)
and kelv becomes a critical consideration when data are used to fit both parameters as will be
illustrated in the sensitivity analysis of the PBPK model.
       The gestational model included mathematical descriptions for the changes in physiological
parameters such as body weight, cardiac output, and tissue volumes consistent with experimental
observations in pregnant rats. Additionally, this model included a fetal compartment and
considered the transfer of TCDD between the placental and fetal compartments  as a
diffusion-limited process (rather than a perfusion-limited) process (see Text Boxes 3-1 and 3-2).2?
27 Diffusion limited, sometimes also known as "membrane limited," means a chemical's movement from one side of
the membrane to the other is limited by the membrane. Thus, the membrane, in this case, is a limiting factor for
uptake. Perfusion limited, also known as "flow limited" indicates that a chemical is so rapidly taken up (e.g., by the
tissue from the blood) that the flow rate is the only limiting factor.
                                           3-49

-------
Text Box 3-1.

                                                         (  0.41 x Time
Variation of Body Weight with Age: BWTtme (g) = BWinitial  x


Cardiac Output: Qc(mL / h) = Qcc x 60
                                                         {1402.5 + Time J

                                    BWmother\
                                       1,000   J
A factor of 60 corresponds to the conversion of minutes to hours, and 1,000 is the conversion of
body weight from g to kg.
Blood Compartment:

Cb(nmol / mL) =
((Of x Qft) + (Qre x Creb) + (Qli x CUV) + (Qpla x CplaV) + Lymph}} - (Cb x  Clru)
                                      ~Qc
                                      5-50

-------
Text Box 3-2.


Placenta Tissue Compartment


(a) Tissue -blood subcompartment



 dAplab
                                       pApla(Cplab - Cplafree)
         Aplab
Cplab = — ^ -
         Wplab


(b) Tissue cellular matrices
 dApla(     ,  .,.   r,,,,^,!   ^ i  r  \   dApla  fet   dAfet  pla
 —t—(nmol/h) = PApla(Cplab - Cplafree]	—=^— + —-—=±—
  dt   V              F    F       * J   )       dt           dt


 Cpla(nmol I mL) =	
                   Wpla


Free TCDD Concentration in Placenta
Cplafree(nmol I mF) = Clpla -
(Cplafree x Ppla +
Plabmax x Cplafree

 Kdpla + Cplafree
Dioxin Transfer from Placenta to Fetuses

 dAPla   fet       .      ~.
 - -=J— (nmol I h) = Clpla  fet x Cpla
     dt


Dioxin Transfer from Fetuses to Placenta

 dAfet  Pla
      ~ - (nmol I K) = Clpla_fet x CfetV
Fetal Dioxin Concentration (Fetuses 5 = Per Litter)

 dAfet,    ....   dAPla  fet   dAfet  Pla
 — - — (nmollh) = - =^ --- - — = -
  dt                  dt           dt

                Afet
Cfet(nmoll h) =
CfetV (nmol I mL) =
                Wfet

                    Cfet
                   Pfet
                                   5-51

-------
3.3.4.3.2.3. Parameter estimation
       Table 3-8 lists the numerical values of the adult rat and human PBPK models of Emond
et al. (2006; 2005; 2004). Additionally, Table 3-8 lists the numerical values that can be used in a
mouse PBPK model.  The values for key input parameters of the rat gestational model are
summarized in Table 3-8 as well as Figure 3-13.
       The parameters for the rat model were obtained primarily from Wang et al. (1997) except
that the value of the affinity constant for CYP1A2 was slightly changed from 0.03 to
0.04 nmol/mL to get a better fit to experimental data (Emond et al., 2004), and the variable
elimination parameter (kelv) was obtained by optimization of model fit to kinetic data from
Santostefano et al. (1998) and others (Emond et al..  2006: Emond et al.. 2005: Wang et al..
1997). Wang et al. (1997) used measured tissue weights whereas the tissue blood flows and
tissue blood weights were obtained from International Life Sciences Institute (ILSI, 1994).  The
partition coefficients (which were similar to those of Leung et al.,1990: Leung et al., 1988), the
permeability x  area (PA) value for tissues, the dissociation constant for binding to CYP1A2
(ICA2), and the Hill coefficient (h) were estimated using a two-stage process of fitting to
dose-response and time-course data on TCDD tissue distribution (Wang et al., 1997). In the
initial stage, the experimental data of arterial blood concentrations were used as input to the
individual compartment to estimate the parameters;  then, with the values obtained during stage
one as initial estimates, those unknown parameters were re-estimated by solving the entire model
at once using an optimization route (Wang et al., 1997). The receptor concentrations and
dissociation constant of TCDD bound to AhR were  obtained by fitting the model to TCDD tissue
concentration combined with enzyme data reported  by Santostefano et al. (1998) whereas the
basal CYP1A2 in liver was based on literature data (Wang et al., 1997).
       The parameters for the human PBPK model  were primarily based on the rat model
(Emond et al., 2006: Emond et al., 2005: Wangetal.,  1997).  Specifically, the blood fraction in
the tissues, the tissue:blood partition coefficients, tissue permeability coefficient, the binding
affinity of TCDD to AhR and CYP,  and the  maximum binding capacity in the liver for AhR were
all set equal to  the values used in the rat model. The species-specific elimination constant, kelv,
was estimated by fitting to human data (Emond et al., 2005).
                                          5-52

-------
Table 3-8. Parameters of the PBPK model for TCDD
Parameter
description
Body weight (g)
Cardiac output (mL/hour/kg)
Symbol
BW
QCCAR
Parameter values
Human
nongestational"
Calculated
15.36c'd
Human
gestational3
Calculated
Calculated
Mouse
nongestational
23-28
275C
Mouse
gestational
23-28
275C
Rat
nongestational
125-250b
311.4e
Rat
gestational
85-190b
311.4e
Tissue (intracellular) volumes (fraction of BW)
Liver
Fat
WLIO
WFO
Calculated
Calculated
Calculated
Calculated
0.0549'
0.069e
0.0549'
Calculated
0.036e
0.069e
0.036e
Calculated
Tissue blood volumes
Liver (fraction of WLIO)
Fat (fraction of WFO)
Rest of body (fraction of WREO)
Placenta tissue fraction of tissue blood weight
(unitless)
WLIBO
WFBO
WREBO
WPLABO
0.266e
0.05e
0.03e
N/A
0.266e
0.05e
0.03e
0.5s
0.266e
0.05e
0.03e
N/A
0.266e
0.05e
0.03 e
0.5e
0.266e
0.05e
0.03e
N/A
0.266e
0.05e
0.03e
0.5e
Tissue blood flow (fraction of cardiac output)
Liver
Fat
Placenta
QLIF
QFF
QPLAF
0.26C
0.05C
N/A
0.26C
0.05C
Calculated
0.161'
0.0711
N/A
0.161'
0.0711
Calculated
0.183e
0.069e
N/A
0.183e
0.069e
Calculated
Tissue permeability (fraction of tissue blood flow)
Liver
Fat
Placenta diffusional permeability fraction
(unitless)
Rest of body
PALIF
PAFF
PAPLAF
PAREF
0.35e
0.121
N/A
0.03e
0.35e
0.121
0.3s
0.03e
0.35e
0.121
N/A
0.03e
0.35e
0.121
0.03g
0.03e
0.35e
0.091e
N/A
0.0298e
0.35e
0.091e
0.3g
0.0298e

-------
Table 3-8. Parameters of the PBPK model for TCDD (continued)
Parameter
description
Symbol
Parameter values
Human
nongestational"
Human
gestational"
Mouse
nongestational
Mouse
gestational
Rat
nongestational
Rat
gestational
Partition coefficient
Liver
Fetus/blood partition coefficient (unitless)
Placenta/blood partition coefficient (unitless)
Fat
Rest of body
PLI
PFETUS
PPLA
PF
PRE
6e
N/A
N/A
100e
1.5e
6e
4J
1.5J
100e
1.5e
6e
N/A
N/A
4001
3k
6e
4J
3g
4001
3k
6e
N/A
N/A
100e
1.5e
6e
4J
1.5J
100e
1.5e
Metabolism constants
Urinary clearance elimination (mL/hour)
Clearance — transfer from mother to fetus
(mL/hour)
Liver (biliary elimination and metabolism;
hour'1)
Interspecies constant (hour"1)
CLURI
CLPLA_FET
KBILE_LI
KELV
4.17E-081
N/A
Inducible
0.00111
4.17E-081
16e
Inducible
0.00111
0.091
N/A
Inducible
0.41
0.091
0.171
Inducible
0.41
0.01J
N/A
Inducible
0.15e
0.01J
0.171
Inducible
0.15e
AhR
Affinity constant in liver (nmol/mL)
Binding capacity in liver (nmol/mL)
Placenta binding capacity (nmol/mL)
Affinity constant protein (AhR) in placenta
(nmol/mL)
KDLI
LIBMAX
PLABMAX
KDPLA
o.r
0.35e
N/A
N/A
o.r
0.35e
0.2J
0.1J
o.ooor
0.00035e
N/A
N/A
o.ooor
0.00035e
0.0002J
0.000 1J
o.ooor
0.00035e
N/A
N/A
o.ooor
0.00035e
0.0002J
0.000 1J
CYP1A2 induction parameters
Dissociation constant CYP1A2 (nmol/mL)
Degradation process CYP1A2 (nmol/mL)
Dissociation constant during induction
(nmol/mL)
Basal concentration of CYP1 A2 (nmol/mL)
First-order rate of degradation (hour"1)
Time delay before induction process (hour)
Maximal induction of CYP1A2 (unitless)
KDLI2
CYP1A2 1OUTZ
CYP1A2_1EC50
CYP1A2 1A2
CYP1A2 1KOUT
CYP1A2 1TAU
CYP1A2 1EMAX
40>
l,600e
130e
l,600e
o.r
0.25e
9,300'
401
l,600e
130e
l,600e
o.r
0.25e
9,300'
0.021
1.6"
0.13e
1.5k
o.r
1.5k
600e
0.021
1.6e
0.13e
1.5k
o.r
1.5k
600e
0.04J
1.6e
0.13e
1.6e
o.r
0.25e
600e
0.04J
1.6e
0.13e
1.6e
o.r
0.25e
600e

-------
        Table 3-8. Parameters of the PBPK model for TCDD (continued)
Parameter
description
Symbol
Parameter values
Human
nongestational3
Human
gestational"
Mouse
nongestational
Mouse
gestational
Rat
nongestational
Rat
gestational
Other constants
Oral absorption constant (hour"1)
Gastric nonabsorption constant (hour"1)
KABS
KST
0.061
0.01m
0.061
0.01m
0.481
0.301
0.481
0.301
0.48e
0.36e
0.48e
0.36e
"Units for human nongestational parameters are L rather than mL and kg rather than g where applicable.
bBody weight varies by study (EmondetaL 2006: EmondetaL 2005: EmondetaL 2004).
°Krishnan and Andersen (1991).
dUnits are L/kg/hr.
eWang et al. (1997).
fILSI (1994).
8Fixed.
hLeung et al. (1990).
'Optimized.
JEmond et al. (2006: 2005: 2004).
kWang et al. (2000).
'Lawrence and Gobas (1997).
mCalculated to estimate 87% bioavailability of TCDD in humans (Poiger and Schlatter. 1986).

-------
                             	

                      10      15
                       Time (days)
                                     20
                                            25
     1 °12
     \  0.1
     I
     g 0.08
     | 0.06
     I 0.04
     I 0.02
                                                  800

                                                  700
                                                  600
                                                £
                                                ~ 500
400

300

200

100

 o



70
                  10      15
                    Time (days*
                                                                                 20
                                                                                         25
I5"
I "o
i

u. 20
  10
                       10      15
                        Time (days)
                                            25
                  10      15
                   Time (days)
                                                                                 20
                                                                                         25
       Figure 3-13. Growth rates for physiological changes occurring during
       gestation.
       (a) Placental growth during gestation (calculated for n = 10 placenta). Experimental data from
       Sikov (1970).  (b) Blood flow rate in placenta! compartment during gestation.  Experimental data
       from Buelke-Sam et al. (1982a: 1982b). (c) Fat fraction of body weight during gestation.
       Experimental data came from Fisher et al. (1989). and (d) Fetal growth during gestation.
       Experimental data obtained from Sikov (1970).
       For the gestational rat model, the parameters describing the growth of the placental and
fetal compartments as well as temporal change in blood flow during gestation were incorporated
based on existing data. Exponential equations for the growing compartments were used (see
Figure 3-13), except for adipose tissue, for which a linear growth increment based on literature
data was specified. All relevant physiological parameters for the pregnant rat were obtained
from the literature while remaining input parameters were set equal to that of the nonpregnant rat
                                             5-56

-------
(obtained from Wang et al., 1997): see Table 3-8. The current version of the rat gestational
model contains parameters for variable elimination from Emond et al. (2006; Table 3-8) and still
provides essentially the same predictions as the original publication (Emond et al., 2004).

3.3.4.3.2.4. Model performance and degree of evaluation
       The PBPK model of Emond et al. (2006: 2005: 2004) had parameters estimated by fitting
to dose and time-course data, so that the resulting model consistently reproduced available
kinetic data. The same model structure with a single set of species-specific parameters could
reproduce the kinetics of TCDD following various doses and exposure scenarios not only in the
rat but also in humans. The simulations of the PBPK model of Emond et al. (2006) have been
compared with two sets of previously published rat data: blood pharmacokinetics following a
single dose of 10 |ig/kg (the dose corresponding to the mean effective dose for induction of
CYP1A2) (Santostefano et al.. 1998) (see Figure 3-14); and hepatic TCDD concentrations
following chronic exposure to average daily exposures of 3.5 to 125 ng/kg (Walker et al., 1999)
(see Figure 3-15). It is relevant to note that the PBPK model of Emond et al. (2006. 2004) is
essentially a reduced version of the Wang et al. (1997) model, and it, therefore, provides
simulations of liver and fat concentrations of TCDD that deviated by not more than 10-15% of
those of Wang et al. (1997). The nongestational model of Emond et al. (2004) was calibrated
against kinetic data in liver, fat, blood, and rest of body of female Sprague-Dawley rats given a
single dose of 10 jig TCDD/kg (data from Santostefano et al., 1996) and in liver and fat of male
Wistar rats treated with a loading dose of 25 ng/kg followed by a weekly maintenance dose of
5 ng TCDDAg by gavage (data from Krowke et al., 1989).
       The gestational rat PBPK model was calibrated against the following kinetic data sets
(Emond et al.. 2004):
   •   TCDD concentration in blood, fat, liver, placenta, and fetus of female Long-Evans rats
       given 1,  10, or 30 ng/kg, 5 days/week, for 13 weeks prior to mating followed by daily
       exposure through parturition (Hurst et al., 2000b):
   •   TCDD concentration in tissues (liver, fat), blood, placenta and fetus determined on
       gestation day (GD) 16 and GD 21 following a single dose of 0.05, 0.8, or 1 ug/kg given
       on GD 15 to pregnant Long-Evans rat (Hurst et al., 2000a):
                                          5-57

-------
                             Fixed elimination rate
                             EXPBL
                        400    600
                         Time (hr)
                                     800
                              1,000
                                                     0.0001
                                                            Variable elimination rate
                                                            EXPEL
400    600
 Time(hr)
                                                                                  800
1,000
                           o
                           o
                                                                                                     0.1-
                                                                                                   0.0001
                          	 Variable elimination rate
                           •  EXPBL
                                                                                                             200
400    600
 Time (hr)
                                                                                                                                800
1,000

-------
           100.00^
            10.00 =
        g
        1=   1.00
             0.10 =
             0.01
             1,750ngTCDD/kgBW           •
  ^^JvNV^^^^^^rv^^w^x^N^Nv^^^
|^          500ngTCDD/kgBW            A
    ^vNVJ^J^(>^^^^^^^^^Js^^*^^«v^^^^^
.^sN^1^       icn nn Trnn/L/n DVA/
y^i ni^i*Ti ^i^mmi^^T^^^^^ T ^ ^r^-w '
 150ngTCDD/kgBW
s^^J^v^^^^^N^^N^^^^M^r>^^^^^
 50nqTCDD/kaBW
     ""t m. *.K Jk Phbik h_h.*^T	
                                  10      15      20
                                       Time (week)
                                     25
                                30
35
      Figure 3-15. PBPK model simulation of hepatic TCDD concentration (ppb)
      during chronic exposure to TCDD at 50,150, 500, or 1,750 ng TCDD/BW
      using the inducible elimination rate model compared with the experimental
      data measured at the end of exposure.
      Source: Emond et al. (2006).
   •  Maternal and fetal tissue concentrations on GD 9, GD 16, and GD 21 after a single dose
      of 1.15 ng TCDD/kg given to Long-Evans rats on GD 9 or GD 15 (Hurst etal.. 1998):
      and
   •  Fetal TCDD concentrations determined on GD 19 and GD 21 in rats exposed to
      5.6 ng TCDD/kg on GD 18 (Li et al.. 2006).
      Furthermore, the scaled rat model was shown to be capable of simulating human data
(see Figures 3-16 and 3-17). In this regard, it is useful to note that the computational version of
the PBPK model of Emond et al. (2006; 2005) also contained the necessary equation to
transform the model output of blood concentration into serum lipid-adjusted concentration of
TCDD. This conversion is calculated by dividing the estimated total blood TCDD levels with
the product of two constants, the serum portion of total blood and the lipid content in serum.  The
human model of Emond et al. (2005; Emond model) has advantages for improving the TCDD
dosimetry used in existing human epidemiologic studies because the model predicts the
                                       3-59

-------
1
g
u
1—
s
10000 -








I Vt09~



^— — '
^— B-
-, 0 10 20
J Time (year)
:30

40
TCDD(PPT)
|_B
10DOD -,
1DOD -
10D •
10 •
1 .
D

— CBPPTRH

~~——- -' '

S ID 15 20 25 30
Time (year)


n_
a
nj





	 CBPPTRH
• V33457




~-— - v_ •
"~~~~
D ID 20 30
t. | Time (Year)

40
10DOD

;-
D
O
"~ ID -


V

	 CBPPTRH
• V36564


—
^^"^-— i-_ _B •
	
D ID 20
|"jj~| Time (year)
30

•to
f"
p.
a
s
i-
[|







" ; — — — ___j_ •


	 CBPPTRH
• V30208






~~~~ —*-
D S 10 15 20
Time (year)
25 30

35


i
5
R
o
I—


, 	

1000-





	 CBPPTRH I


N. — ____
~ — — —_^_^^




~m~~—f-


0 5 10 15 20 25
Time (year)


a
Q 100 •
Q
1-









PPTRH
)044

.
v--_____^ •
*~
j 	 1 0 5 10 15 2D
G | Time (year)
25

30 35

10DODO
1DODO
g, 10DO
E.
a 100
o
o
•- 10
1
D -.










— CBPPTRH
. V31731



^_
~~~~~.~^




0 10 20
Time (year)
—m-~.

3D



40





Q
S loo-
i—
ID -


V
^~~~~~~~~--~^. •




CBPPTRH

• V30172






D ID 20
HTime (year"
"


30 40


100000
&
D 1000
a
0 -im





^^

	 CBPPTRH

^~~~~^~~-~m-^
^^~* 	

0 10 20
("pi TlfTK (year)

30 40
Figure 3-16. Model predictions of TCDD blood concentration in 10 veterans
(A-J) from Ranch Hand Cohort.
Source: Emond et al. (2005).
                                  5-60

-------
     1,000,000
   TJ
   I
   3
   -51
   -*  100,000
   Q.
   1
   Q.
   "O
   §   10,000
                 100    200    300     400     500    600    700     800    900    1,000   1,100   1,200
                               - Blood model predictions (pt=l) */ |
                                                        Patient 1 Vienna women
1,000,000 -,
100,000-
1,000





	 • 	 . —


_mm


'*• •


T^*


• — -•—


u


	 •














	 H

) 100 200 300 400 500 600 700 800 900 1,000 1,100 1,200


\ "^ blood
model predictions (pt #2) */ ^^^| Patient 2 Vienna women 1



       Figure 3-17. Time course of TCDD in blood (pg/g lipid adjusted) for two
       highly exposed Austrian women (patients 1 and 2).
       Symbols represent measured concentrations, and lines represent model predictions. These data
       were used as part of the model evaluation (Geusau et al.. 2002).
       Source: Emond et al. (2005).
redistribution of TCDD within the body (to stores in fat and liver) based on physiological
principles. However, because the dose-dependency of metabolic elimination in the Emond
model was not calibrated to human data, it is important to review the predictions of this model
using a database of human observations that is as extensive as possible and a spread of internal
TCDD concentrations that is as wide as possible. Thus, presented below is a juxtaposition of
modeled elimination rates from the Emond model with observations for two highly exposed
Austrian patients (severe intoxication of "unknown origin" (Geusau et al., 2001) and 9 of
                                            5-61

-------
10 Ranch Hand veterans  used for the original "validation" comparisons presented in the Emond

et al. (2005)).

       Figure 3-18 shows the time course of the declines in TCDD serum concentrations in

two highly exposed Austrian subjects compared with the Emond model results.  The comparison

in Figures 3-17 and 3-18 indicates that the Emond model adequately describes the rate of TCDD

elimination for the more highly exposed Austrian patients but predicts a somewhat faster rate of

decline than that observed for the less heavily exposed patient.
        Q
        Q
        U
        H
        MI
        ~
                                    Qy= 11.8-0.48xRA2 = 0.999
                                    • y= 11.8-0.45xRA2 = 0.772
                                    Oy= 10.5 -0.43xRA2= 1.000
                                      y= 10.0 -0.24xRA2 = 0.612
ln(ptl Em Mod Sim pg/g TCDD)
ln(ptl Obs pg/g TCDD)
In(pt2 Em Mod Sim pg/g TCDD)
In(pt2 Obs pg/g TCDD)
                                            Years After Exposure
       Figure 3-18. Observed vs. Emond et al. (2005) model simulated serum
       TCDD concentrations  (pg/g lipid) over time (In = natural log) in
       two Austrian women.

       Data from Geusau et al. (2002).
28 In preliminary comparisons, the simulation run for the 10th Ranch Hand veteran appeared anomalous and was,
therefore, excluded from this summary.

                                           3-62

-------
       Figure 3-19 shows the results of combining the simulated and observed rates of loss for a
group of Austrian and Ranch Hand subjects evaluated by Emond et al. (2005), counting only
one data point per person.  The X-axis in this figure is the TCDD serum concentration at the
midpoint of the observations for each subject. The error bars in the figure represent ±1 standard
error. The results of this figure illustrate two points: (1) the Emond model simulation (open
squares) are generally very close to the actual data (solid  circles) for the nine Ranch Hand
subjects (clustered toward lower left corner) and one of the two Austrian patients (upper right
corner); and (2) both the Emond model simulation results and the actual data show a linear trend,
and linear regression lines were plotted, respectively, as shown in Figure 3-19.
      4*
      4*
     Q
      s
     Q
     Q
     U
                     El  Emond Mod Sim Ln(Decline/Yr)
                     •  Observed Ln(Decline/Yr)
y= - 0.101+ 0.123x RA2 = 0.995
y= - 0.054+ 0.092x RA2 = 0.884
                                  Log(TCDD pg/g at Midpoint Obs)
       Figure 3-19.  Comparison of the dose dependency of TCDD elimination in the
       Emond model vs. observations of nine Ranch Hand veterans and two highly
       exposed Austrian patients.
       Circles are observed data.
                                           5-63

-------
       Table 3-9 presents the results of regression analyses of the observed rates of decline in
relation to the estimated TCDD serum levels at the midpoint of the observations for each subject
in the Ranch Hand study (see Figure 3-19). These results indicate that some appreciable dose
dependency of TCDD elimination is unequivocally supported. However, the central estimate of
the slope of the relationship between the log of the TCDD elimination rate and the log of the
TCDD level is only about 75% of that expected under the Emond et al. PBPK model
(i.e., 0.092-0.123=0.748).
       Table 3-9. Regression analysis results for the relationship between logio
       serum TCDD at the midpoint of observations and the logio of the rate
       constant for decline of TCDD levels using Ranch Hand data
Item
Summary of fit
Parameter estimates
Aspect
RSquare
RsquareAdj
Root mean square error
Mean responses
Observations (or sum weights)
Intercept
Estimate
Standard deviation
t ratio
Prob>|t|
Log(TCDDpg/g)
Estimate
Standard error
t ratio
Prob>|t|
Value
0.894
0.871
0.044
0.130
11

-0.054
0.026
-2.07
0.0679

0.092
0.011
8.28
O.0001
       Overall, the conclusion from the above analysis is that the Emond model is reasonable to
use, but the model might be improved by (1) including the two dose-independent pathways of
elimination documented in the Geusau papers (GI elimination via the feces and loss via the
sloughing of skin cells), and (2) reducing the extent of loss via the dose-dependent metabolism
pathway from the liver (Harrad et al., 2003; Geusau et al., 2002) so that overall loss rates for the
average elimination rates from the Ranch Hand veterans are maintained.
                                          5-64

-------
3.3.4.3.2.5. Sensitivity analysis of the physiologically based pharmacokinetic (PBPK) model
       A sensitivity analysis was performed on each of the animal and human Emond PBPK
models to determine the most sensitive variables.  In each case, all input variables in each model
were included in the analysis. For equations where the parameter value varies with age
according to an equation (body weight in all models, liver and adipose tissue fractions in the
human models, and fetal weight, placental weight, and placental perfusion in the gestational
models), a constant multiplier of 1.0 was included in each equation; then, for the sensitivity
analysis, this value was varied by a fixed percentage to determine the relative effect of changing
the compartmental weight fractions.
       To perform the analysis, a representative dosing protocol was selected for each model to
ensure the analysis was performed in dose ranges that were applicable to the overall health
assessment.  For each study modeled, multiple doses were used to investigate model sensitivity
across a dosing range.  Table 3-10 shows the dosing protocols selected for each model. For the
human models, doses in the range of the identified reference dose and POD dose discussed in
Section 4 were used in the analysis.
       To perform the sensitivity analysis, variable values were varied by fixed percentages one
at a time to determine the associated change in the average whole blood  concentration. The
blood concentration averages were calculated in each study in the same manner as in the  main
health assessment, as detailed in Appendix E and repeated for convenience in Table 3-10. To
determine the local sensitivity of the whole blood concentration to each variable, the variable
values were increased and decreased from the standard model configuration by 5%.  This local
analysis shows the effects of changing the variables by relatively small amounts to account for a
theoretical level of uncertainty in the input parameters. To determine a more global sensitivity of
the whole blood concentrations to each variable, the variable values were increased and
decreased by 50%. In some cases, such  a wide change may overestimate the actual uncertainty
in the variable value in the literature; however, such a change is useful in helping to determine
how the model sensitivity may change across large portions of the variable parameter space.
                                          5-65

-------
       Table 3-10.  Dosing protocols for human and animal models
Model
Rat
Mouse
Rat
gestational
Mouse
gestational
iuman
iuman
gestational
Study
NTP (2QQ6b); 105 weeks
NTP (1982a); male mouse,
2 -year duration
Vlarkowski et al. (2001)
Li et al. (2006)
Standard lifetime scenario
daily intake for 70 years)
Standard gestational
scenario (daily intake,
>regnancy at age 45)
Low dose
3 ng/kg 5 days per week
(2.14 ng/kg -day adjusted
dose)
5 ng/kg biweekly (1.4
ng/kg-day adjusted dose)
20 ng/kg, single dose
2 ng/kg-day for GDs 1-3
7 x 10'4 ng/kg-day
7 x 10'4 ng/kg-day
High dose
100 ng/kg 5 days per
week (71.4 ng/kg-day
adjusted dose)
200 ng/kg biweekly
(71 ng/kg-day adjusted
dose)
180 ng/kg, single dose
100 ng/kg-day for
GDs 1-3
0.02 ng/kg-day
0.02 ng/kg-day
Averaging
period
105 weeks
2 years
Single day
3 days
70 years
9 months of
pregnancy
       For each percentage change in the variable, the associated percentage change in the
average whole blood concentration was recorded. Then, the elasticity was calculated as the
percent change in the average whole blood concentration divided by the percent change in the
variable value. Thus, variables where the magnitude of the elasticity is greater than 1 will induce
a change  of greater than 5% in the whole blood concentration when the variable value is changed
by 5%. The sign of the elasticity indicates whether the whole blood concentration is positively
or negatively correlated with the variable.  The elasticities were examined, and a value of 0.1
was selected as a threshold to determine the most sensitive variables in each model. This  value
tended to represent a limit, with a cluster of variables having higher magnitude elasticities and
the remaining variables having much lower elasticities. Variables were then ranked according to
the magnitude of the elasticity in the case where the variables were increased by  5% for
presentation.
       Table 3-11 shows the most sensitive variables for the rat and mouse nongestational
models and rat and mouse gestational models for the low and high doses when variables were
increased by +5%.  The associated elasticities are shown in each case. The only variable with
elasticity above one is the Hill coefficient (h in Eq. 3-20).  The other most sensitive variables are
                                          5-66

-------
Table 3-11. Most sensitive variables for the rat and mouse nongestational
and gestational models
Variable
Variable description
Rat, low dose,
+5%
elasticity
Rat, high dose,
+5% elasticity
Mouse, low
dose, +5%
elasticity
Mouse, high
dose, +5%
elasticity
Nongestational
HILL
CYP1A2 1OUTZ
CYP1A2_1A2
WLIO
CYP1A2_1EMAX
KELV
LIBMAX
CYP1A2_1EC50
KDLI
KABS
KST
-[ill coefficient
nduction concentration in
degradation process
nmol/L)
nduction basal
concentration of 1A2
nmol/L)
Fractional liver weight
unitless)
Maximum induction over
msal effect (unitless)
nterspecies constant
hrA-l)
Liver binding capacity
nmol/1)
[nduction disassociation
constant for 1A2 (nmol/L)
Jver affinity proteins
AhR (nmol/L)
ntestinal excretion and
absorption constant
hrA-l)
Gastric excretion and
absorption constant
hrA-l)
3.3
-0.8
0.8
-0.6
-0.5
-0.3
-0.4
0.4
0.3
0.3
-0.3
3.0
-0.8
0.8
-0.7
-0.7
-0.7
-0.4
0.4
0.2
0.3
-0.3
3.4
-0.8
0.9
-0.6
-0.5
-0.5
-0.3
0.3
0.3
0.3
-0.3
2.8
-0.7
0.7
-0.6
-0.6
-0.6
-0.3
0.4
0.3
0.3
-0.3
Gestational
HILL
WLIO
KABS
CYP1A2 1OUTZ
KDLI2
KST
-[ill coefficient
fractional liver weight
unitless)
ntestinal excretion and
absorption constant
hrA-l)
nduction concentration in
degradation process
nmol/L)
Jver affinity proteins 1A2
nmol/L)
Gastric excretion and
absorption constant
hrA-l)
1.2
-0.4
0.4
-0.4
0.4
-0.4
1.4
-0.4
0.4
-0.4
0.4
-0.3
0.6
-0.2
0.4
-0.3
0.2
-0.3
1.4
-0.4
0.3
-0.4
0.3
-0.3
                                  5-67

-------
      Table 3-11. Most sensitive variables for the rat and mouse nongestational and
      gestational models (continued)
Variable
QCCAR
QFF
CYP1A2_1EMAX
PAFF
LIBMAX
KDLI
CYP1A2_1EC50
CYP1A2_1KOUT
Variable description
Cardiac output (1/kg-hr)
Adipose tissue blood flow
fraction of cardiac output
unitless)
Maximum induction over
msal effect (unitless)
Adipose diffusional
)ermeability fraction
unitless)
Jver binding capacity
nmol/L)
Liver affinity proteins
AhR (nmol/L)
Induction disassociation
constant for 1A2 (nmol/L)
nduction first-order rate
of degradation (hr A- 1 )
Rat, low dose,
+5%
elasticity
-0.3
-0.2
-0.2
-0.2
-0.1
0.1
0.1
-0.1
Rat, high dose,
+5% elasticity
-0.3
-0.2
-0.3
-0.2
-0.2
0.1
0.2
-0.2
Mouse, low
dose, +5%
elasticity
-0.4
-0.4
-0.1
-0.4
-0.1
0.1
0.1
0.0
Mouse, high
dose, +5%
Elasticity
-0.3
-0.2
-0.3
-0.2
-0.2
0.2
0.2
0.0
associated with the overall dioxin elimination/sequestration rate, including the CYP1A2
induction rates, the liver weight, the binding capacity and affinity, and the gastric and intestinal
excretion rates. For the gestational model dosing protocols, the Hill coefficient remains the most
sensitive variable, but the elasticity decreases compared with the nongestational analysis.
Otherwise, many of the most sensitive variables remain those associated with elimination.
Additional parameters related to the adipose tissue blood flow and with the adipose diffusional
permeability fraction are also relatively sensitive.
       Table 3-12 shows the most sensitive variables for the human nongestational and
gestational models. The additional variables associated with the adipose compartment partition
coefficient, the body weight, and the fractional adipose tissue volume are also relatively sensitive
variables at the reference dose and POD dose compared with the animal models.  For all models,
the elasticities  are relatively similar across the different doses evaluated.
                                          5-68

-------
Table 3-12. Most sensitive variables for the human nongestational and gestational models
Variable
HILL
CYP1A2_1OUTZ
CYP1A2 1A2
CYP1A2 1EMAX
SA_CHNGELI
KELV
CYP1A2 1EC50
KDLI
LIBMAX
SA_CHNGEBW
PF
SA_CHNGEF
KABS
KST
KDLI2
Variable description
Hill coefficient
Induction concentration in degradation process
(nmol/L)
Induction basal concentration of 1A2 (nmol/L)
Maximum induction over basal effect (unitless)
Fraction liver-weight multiplier for sensitivity
analysis (unitless)
Interspecies constant (hrA-l)
Induction disassociation constant for 1 A2 (nmol/L)
Liver affinity proteins AhR (nmol/L)
Liver binding capacity (nmol/L)
Body-weight multiplier for sensitivity analysis
(unitless)
Adipose tissue :blood partition coefficient (unitless)
Fraction adipose-weight multiplier for sensitivity
analysis (unitless)
Intestinal excretion and absorption constant (hrA-l)
Gastric excretion and absorption constant (hrA-l)
Liver affinity proteins 1 A2 (nmol/L)
Human
nongestational, POD
dose +50% elasticity
5.35
-0.44
0.46
-0.42
-0.43
-0.39
0.30
0.30
-0.27
0.31
-0.07
-0.06
0.07
-0.09
0.05
Human
nongestational,
POD dose +5%
elasticity
3.56
-0.58
0.53
-0.56
-0.57
-0.50
0.34
0.34
-0.31
0.01
-0.06
-0.07
0.09
-0.09
0.07
Human
gestational, POD
dose +50%
elasticity
5.75
-0.45
0.52
-0.44
-0.44
-0.43
0.32
0.31
-0.28
0.47
-0.04
-0.03
0.06
-0.09
0.03
Human
gestational, POD
dose +5%
elasticity
3.75
-0.61
0.59
-0.596
-0.59
-0.56
0.36
0.35
-0.34
0.09
-0.03
-0.03
0.09
-0.09
0.03

-------
       In order to observe the difference between the local and global elasticities, Figures 3-20
and 3-21 show the elasticities for the most sensitive variables in the human nongestational model
for the POD dose and reference dose, respectively. In general, the elasticities are similar across
the different percentage changes in variable values that were tested. Changes in variables by
-50% tend to lead to the greatest elasticities.  Changing the variable values by +5% and -5%
lead to almost the same elasticities for nearly all the variables. These same conclusions hold for
all the other models and doses as well.
       Of the variables to which the blood concentrations are most sensitive, most of the
variables are either derived from Wang et al. (1997) or are optimized (see Table 3-8). For the
human model, parameters set equal to values in the rat model may be subject to particular
uncertainty.  In  particular, the AhR and CYP1A2 induction parameters typically were based on
the rat model parameters. The exception is CYP1 A2_1EMAX, the maximum induction of
CYP1A2, which is an optimized parameter.  The variable elimination rate, kelv, and the intestinal
excretion, KST, are also both optimized against data.  For variables that are optimized, a
sensitivity analysis that varies each parameter one at a time may overestimate the associated
model uncertainty associated with the variable. A change in KST, for example, would
necessitate a commensurate change in the other optimized variables in order to suitably capture
the comparison data, and the overall  changes in the blood concentrations might be small.
       The most sensitive variable in all the models is the Hill parameter. The elasticity is high
in part because the Hill parameter is an exponent; thus, small changes in the value can lead to
larger changes in the whole blood concentration.  However, as stated above,  any change in the
Hill parameter would also necessitate changes in optimized variables in  order to maintain an
adequate fit with the data.  The next section explores the effect of changing the Hill parameter
and the effect of changing the CYP1A2 induction parameters on the model fits to literature data.
                                          5-70

-------
 CYP1A2 1OUTZ
CYP1A2_1EMAX|
 CYP1A2  1EC50
 SA CHNGEBW!
                                     • Elasticity, POD Dose, +50%
                                     Q Elasticity, POD Dose, +5%
                                     • Elasticity, POD Dose, -5%
                                     P Elasticity, POD Dose, -50%
                             -4.0
-2.0
  0.0
Elasticity
2.0
4.0
   Figure 3-20. Elasticities in the nongestational human model, POD dose.

-------
to
HILL
CYP1A2JOUTZ
CYP1A2 1EMAX

CYP1A2JA2
SA_CHNGELI
KELV
LIB MAX
-
SA_CHNGEBW
CYP1A2JEC50
KDLI
PF
SA CHNGEF
"
-4





































=
L 	

r~:~^~~~
i 	
1 — *• 	
1
L"_"_ 	
i 	


IE
jnn

	
	 .,

j


	 i_t
::::::::i..n



T~~~J
L
L













" t 1 T !
.0 -3.0 -2.0 -1.0 0.0 1.0









	 .










• Elasticity, RfD Dose, +50%
• Elasticity, RfD Dose, +5%
• Elasticity, RfD Dose, -5%
n Elasticity, RfD Dose, -50%

mmmmmmmmmmmmmmfmmmmmmmmmmmmmmmmmmmmmm^mmmmmmmmmmmmmmmmmmm
2.0 3.0
                                                                                                                        4.0
                                                                    Elasticity
           Figure 3-21. Elasticities in the nongestational human model, RfD dose.

-------
3.3.4.3.2.6. Further uncertainty analysis of the Hill coefficient and CYP1A2 induction
           parameters
       As illustrated by the sensitivity analysis of the PBPK model, the predicted TCDD blood
concentrations are very sensitive to the Hill coefficient (h) as described in Eq. 3-20. This
parameter is included in the mathematical description for the induction of the CYP1A2.
Therefore, the best type of data needed to estimate an in vivo value for this constant would be
time-course levels of hepatic CYP1A2 in response to TCDD exposure. This type of data is only
available in experiments conducted in animals. The PBPK model adopted a value of 0.6 for this
parameter based on the earlier reported models by Wang et al. (2000) and Santostefano et al.
(1998).  In both cases, the value of 0.6 used for the Hill coefficient (the model parameter Hill) in
the model was fit to describe the temporal relationship between TCDD exposure and
CYP1A2-induction levels in animals.  Note that the value of 0.6 for Hill indicates supralinear
behavior at low exposure levels, which translates to a supralinear relationship between oral
intake and blood TCDD concentrations.
       For humans, the only data available to calibrate the in vivo model parameters are blood
levels of TCDD. Predicted TCDD blood levels are influenced by the Hill coefficient when  it is
implicitly included in the description for the hepatic elimination of TCDD by induced levels of
CYP1A2 as described in Eq. 3-21. However, as was illustrated earlier, the elimination of TCDD
by the liver is also influenced by the numerical optimization of the kelv constant in the same
equation.  Therefore, estimation of the Hill coefficient using human blood data is highly
dependent on the simultaneous estimation of kelv.
       In order to estimate the interdependence of Hill and kelv and to investigate the behavior
of the Emond human PBPK model in the absence of supralinearity, EPA calibrated the model to
several human data sets after setting Hill to 1 and varying kelv.  A Hill coefficient of 1 results in
low-dose linearity, where  supralinear behavior is first eliminated.  However, EPA does not
consider a Hill value of 1 necessarily to be a plausible replacement for the model variable of 0.6;
it is just being used to  investigate the behavior of the model as a sensitivity analysis.  The data
sets are TCDD serum concentrations (lipid-adjusted serum concentration [LASC]) over time for
four individuals: two Austrian adult females (Geusau et al., 2002) (1996) and two Italian
(Seveso) males—a 6-year-old  and a 50-year-old (Needham et al.,  1997): the data are presented in
Tables 3-13 and 3-14.  The results of Hill coefficient sensitivity analysis simulations are shown
in Figure 3-22 and Table 3-15. For each data set, the simulation was run four times—once with

-------
Table 3-13. TCDD serum measurements over time for two Austrian women
exposed to TCDD in 1997a
Austrian woman 1
Day
0
63
116
126
135
147
161
168
203
240
270
295
309
316
323
330
366
389
466
500
596
700
781
904
1,054
TCDD LASC (ppt)
144,000
111,000
85,600
80,900
72,200
70,200
87,700
89,900
62,100
65,100
68,300
64,900
68,100
72,600
73,700
72,500
60,300
73,900
85,600
68,100
47,100
39,300
27,400
30,300
35,900
Austrian woman 2
Day
0
53
63
77
84
98
105
140
177
207
238
267
326
437
533
637
718
841
998
TCDD LASC (ppt)
26,000
20,500
16,100
15,900
14,300
13,200
18,500
13,300
13,700
19,300
15,700
15,200
15,700
17,700
14,100
10,500
11,000
10,100
9,500

      "Source of data: (Geusau et al., 2001).
                                 5-74

-------
Table 3-14. TCDD serum measurements over time for two Seveso males
exposed to TCDD in 1976a
Seveso male (6 years old)
Day
0
826
1,522
2,193
5,867
TCDD LASC (ppt)
15,900
4,350
2,269
580
324

Seveso male (50 years old)
Day
0
92
981
1,218
1,921
6,011
TCDD LASC (ppt)
1,770
807
1,069
809
680
807
      Source of data: Needham et al. (1997)
                                 5-75

-------
140 000 * I     ^Default parameters, no optimization      _
              — 'Default parameters, kelv optimized
120,000

100,000

 80.000

 60.000

 40,000

 20,000

     0
\   —  'Alternative parameters, kelv optimized
                          >  1
                 300
                   600
900
1,200
                   28,000

                   24,000

                   20,000

                   16,000

                   12,000

                    8,000

                    4,000
                                                      0
                      ^Default parameters, no optimization     i
                      — -Default parameters, kelv optimized
                      — -Alternative parameters, kelv optimizeu
0
300
600
900
1,200
 16,000

 14,000

 12,000

 10,000

  8,000

  6,000

  4,000

  2,000

     0
    — Default parameters, no optimization
    —  -Default parameters, kelv optimized
    —  -Alternative parameters, kelv optimized
   i
                   1,800

                   1,500

                   1,200

                    900

                    600

                    300

                      0
                        Default parameters, no optimization
                        -Default parameters, kelv optimized
                        'Alternative parameters, kelv optimized
             1,000   2.000   3,000   4,000   5,000    6,000

                          Time (days)
                                                      0
                                     2,000    3,000    4,000   5,000    6,000

                                           Time (days)
     Figure 3-22. Hill coefficient sensitivity analysis.
     Calibration of Emond human PBPK model for 2 values of Hill for four human data sets:
     (a) Austrian Woman 1, (b) Austrian Woman 2, (c) Seveso 6-year-old male, (d) Seveso 50-year-old
     male; see text for source of data.  Values for kelv other than the standard model value of 0.0011
     are optimized.
                                                      3-76

-------
       Table 3-15.  Results of Hill coefficient sensitivity analysis simulations with
       Emond human PBPK model

Hill = 0.6
kelv = default
doseiv optimized
Hill = 1
kelv = default
doseiv optimized
Hill = 0.6
kelv and doseiv
optimized
Hill = 1
kelv and doseiv
optimized
Hill

0.6
1.0
0.6
1.0
kelv
Austrian 1
Austrian 2
Seveso 6
Seveso 50
0.0011
0.0011
1.73E-03
1.79E-03
0.00300
2.94E-04
5.74E-03
4.89E-03
0.00490
4.79E-03
doseiv
Austrian 1
Austrian 2
Seveso 6
Seveso 50
7.00E+04
1.30E+04
1.10E+04
4.98E+02
1.20E+04
2.40E+03
3.48E+02
9.76E+01
8.00E+04
1.80E+04
1.10E+04
2.98E+02
1.98E+04
3.40E+03
9.98E+02
1.37E+02
the default model parameters (Hill =0.6, kelv = 0.0011), once with Hill =1.0 and kelv
unchanged, once with Hill = 0.6 and kelv optimized for best fit to the data, and once with
Hill =1.0 and kelv optimized. In each case, the initial dose (model parameter doseiv), assuming
a single instantaneous exposure at the time of first serum measurement, was optimized for best
fit; the exposure in this case would be a simulation of the body burden at the time, as the actual
exposure scenario is unknown.  In all cases, simply changing the value of Hill resulted in poor
fits.  Optimizing kelv with Hill set to either to 0.6 or 1 yields much better fits, as would be
expected, with both values fitting the data equally well when the inter-related parameter, kelv, is
optimized.
       EPA also investigated the impact of alternate values for other model parameters related to
the CYP1A2 induction algorithm. Budinsky et al. (2010) reported an in vitro temporal
relationship between CYP1A2 induction and TCDD levels  in human and rat primary
hepatocytes. Budinsky et al.  (2010) used the CYP1A2 induction data to estimate Hill function
constants, such as baseline, fold, and maximal CYP1A2 mRNA inductions.  Using their data, an
                                          5-77

-------
estimate for the human in vivo baseline, fold, and maximal response of CYP1A2 induction can
be approximated as illustrated in Eq. 3-22 and 3-23:
        CYP
        CYP
                            CYP A2
 basal animaiinvm
(Eq. 3-22)
                              CYP A2 basal
                                            human equivalent_invivo
and
        CYP A2Maxhuman.nvjtm
        CYP A2,
                           CYP A2
             'Maxanimalinvitro
Maxanimalinvivo
(Eq. 3-23)
                                             CYP A2 Max
                                                         human equivalent_invivo
The values used in these equations are shown in Table 3-16.
       Table 3-16.  Alternative CYP1A2 parameter estimates for sensitivity analysis
       of Emond human PBPK model

CYP 1A2 Basal
CYPlA2Max
EC50/KDLI
Budinsky et al. (2010)
values
Human
11.6
12,900
0.329
Rat
22.4
322
0.0628
Emond model value
Human
1,600
9,300
130
Rat
1.6
600
0.04
Alternative
scaled value"
Human
829
24,037
209
     aEmond model rat value multiplied by the ratio of the corresponding human:rat parameter values from
      Budinsky et al. (2010).
       The calculated in vivo human CYP1A2 baseline, fold, and maximal induction response,
with their corresponding minimum and maximum values, are then used in the PBPK model to
estimate mean, minimum, and maximum blood levels in comparison to data for two Austrian
cases, and the Seveso cohort. This analysis was done with Hill set to 0.6 and optimizing kelv and
doseiv for the data sets in Tables 3-13 and 3-14. Results of the simulations are shown in
Figure 3-23 and Table 3-17.
                                          5-78

-------
140,000

120,000

100,000

 80,000

 60,000

 40,000

 20,000

     0
Q
Q
    —Default parameters, no optimization
    — -Default parameters, kelv optimized
\   — -Alternative parameters, kelv optimized
                    300
                                            1,200
I i     —Default parameters, no optimization
        • 'Default parameters, kelv optimized
        • -Alternative parameters, kelv optimized
       1,000   2,000   3,000   4,000    5,000   6,000
                    Time (days)
                                                                          —Default parameters, no optimization
                                                                          — -Default parameters, kelv optimized
                                                                          — -Alternative parameters, kelv optimizeu
                                                                                                       b
                                                                          300
                                                                                                 1,200
                                                         1,800

                                                         1,500 \

                                                         1,200
                                                              300
                                                                         —Default parameters, no optimization       J
                                                                         — -Default parameters, kelv optimized
                                                                         — -Alternative parameters, kelv optimized
                                                                 0     1,000   2,000    3,000   4,000    5,000    6,000
                                                                                   Time jdaysj
        Figure 3-23.  CYP1A2 parameter sensitivity analysis.
        Calibration of Emond human PBPK model for alternate values of CYP1A2 parameters other than
        Hill for four human data sets: (a) Austrian Woman 1, (b) Austrian Woman 2, (c) Seveso
        6-year-old male, (d) Seveso 50-year-old male; see text for source of data. Alternate parameters
        were estimated from data presented in Budinsky et al. (2010).
                                                       5-79

-------
       Table 3-17.  Results of CYP1A2 parameter sensitivity analysis simulations
       with Emond human PBPK model

Hill = 0.6
kelv = default
doseiv optimized
Hill = 0.6
kelv and doseiv
optimized
Hill = 0.6, Alternative
parameters,8
kelv and doseiv optimized
kelv
Austrian 1
Austrian 2
Seveso 6
Seveso 50
0.0011
1.73E-03
1.79E-03
0.00300
2.94E-04
4.36E-04
1.67E-04
0.00030
9.68E-06
doseiv
Austrian 1
Austrian 2
Seveso 6
Seveso 50
7.00E+04
1.30E+04
1.10E+04
4.98E+02
8.00E+04
1.80E+04
1.10E+04
2.98E+02
6.98E+04
8.00E+03
5.98E+03
1.97E+02
           "Alternative scaled values from Table 3-16.

       An attempt to directly use the in vitro values of the Hill function estimated in the
Budinsky et al. (2010) in the PBPK model was not successful in simulating blood levels in
Figure 3-23. The failure in using these values directly may be a result of the usual in vitro-to-in
vivo extrapolation complications such as in vitro cellular competency to exhibit toxicological
response comparable to the in vivo ones, and TCDD media to cell sequestration. It is also
important to note that the in vitro preparations in the Budinsky et al. (2010) came from a limited
set of five female subjects. Average and standard variation levels obtained from this set of
human subjects cannot be representative of overall human population.
       It is clear from the results shown in Figures 3-22 and 3-23, that several different
combinations of CYP1A2 induction parameters can be used to simulate the data well. This
process illustrates the interdependences of these parameters  when in vivo blood levels in
humans are the only source of data to estimate them.
       The impact of varying these parameters on model predictions of human oral intakes
corresponding to a range of lifetime average serum concentrations is shown in Table 3-18.  The
range of concentrations was chosen to be representative of human intakes  of interest for the RfD
                                          5-80

-------
          Table 3-18.  Results of Emond human PBPK model parameter sensitivity analysis simulations. Comparison of
          modeled human oral intakes for a range of lifetime average TCDD serum concentrations for alternative
          parameter values.




average
TCDD
LASCa
(ppt)
30
100
300
1,000
3,000
Standard model
configuration
Hill = 0.6
kplv — 0 001 1

CYP1A2 1A1 = 1,600
CYP1A2 1EMAX = 9,300
CYP1A2 1EC50 = 130
PF = 100
l.OE-03
5.7E-03
3.0E-02
1.9E-01
9.6E-01
Alternative Hill
Hill = 1
\fg*iv — o 001 1

CYP1A2 1A1 = 1,600
CYP1A2 1EMAX = 9,300
CYP1A2 1EC50 = 130
PF = 100
3.8E-04
1.3E-03
4.2E-03
1.8E-02
8.1E-02
Standard Hill, optimized
elimination
Hill = 0.6
kplv — 0 0017

CYP1A2 1A1 = 1,600
CYP1A2 1EMAX = 9,300
CYP1A2 1EC50 = 130
PF = 100
1.3E-03
8.0E-03
4.3E-02
2.8E-01
1.4E+00
Alternative Hill, optimized
elimination
Hill = 1
Uplv — 0 OOSO

CYP1A2 1A1 = 1,600
CYP1A2 1EMAX = 9,300
CYP1A2 1EC50 = 130
PF = 100
3.9E-04
1.5E-03
5.9E-03
3.7E-02
2.3E-01
Alternative induction
parameters1"
optimized elimination
Hill = 0.6
Uplv — 0 000?

CYP1A2 1A1 = 829
CYP1A2 1EMAX = 24,037
CYP1A2 1EC50 = 209
PF = 100
7.7E-04
4.1E-03
1.9E-02
1.2E-01
5.8E-01
oo
^    aFrom lifetime female model.
     Estimated from Budinksy et al.

-------
derivation in Section 4. Comparing the optimized simulations for the alternative Hill values
shows that, for these data sets, changing Hill to 1 decreases the modeled intakes for the TCDD
serum concentrations in this range by about 70-85%. Using the alternative parameters estimated
from Budinsky et al. (2010) results in 40-60% lower intakes than for the standard parameters
(optimized kelv). Thus, it would appear that, although the Hill value of 0.6 results in a
supralinear relationship between TCDD intake and serum concentrations in the Emond model,
eliminating the supralinear behavior does not result in higher predicted intakes for lower TCDD
serum concentrations, as might be expected. However, strong conclusions cannot be made from
these  results because the data used for the optimization are not ideal in at least two respects:
(1) they only address CYP1A2 dynamics indirectly, and (2) there are only four data sets, and
they are not necessarily representative of the entire population. In Section 4.5.1.1.1, a sensitivity
analysis is presented that illustrates the predicted change in the point of departure when the Hill
value is changed to 1.

3.3.4.3.2.7. Confidence in physiologically based pharmacokinetic (PBPK) model predictions
           of dose metrics
       The PBPK model facilitates prediction of absorbed dose, body burden, and blood
concentration of TCDD for oral exposures in adult humans and rats (adult and developing) with
high confidence (see Table 3-19). The model output of blood  concentration can be normalized to
lipid content representative of the study group (species, sex, age, lifestage, and diet). However,
the PBPK model of Emond et al. (2006; 2005; 2004) does not  simulate plasma and erythrocyte
TCDD concentrations separately, and it predicts tissue concentrations on the basis of
tissue:whole blood partition coefficients and not on the basis of serum lipid-normalized values.
       The reliability of this model for simulating the liver concentration of TCDD in rats is
considered to be high, but it is considered to be medium for humans. Although empirical data on
bound or free concentrations were not used to evaluate model performance in  humans, the
biological phenomena (consistent with available data) related to the hepatic sequestration,
enzyme induction, and dose-dependent elimination are described in the model. This is one of the
situations where PBPK models are uniquely useful; that is, they permit the prediction of system
behavior based on understanding of the mechanistic determinants, even though the required data
cannot be directly obtained in the system  (e.g., bound concentrations in the liver of exposed
humans).  For these dose measures (i.e., bound concentration and total liver concentration), the
                                          3-82

-------
       Table 3-19.  Confidence in the PBPK model simulations of TCDD dose
       metrics
Dose metric
Administered dose
Absorbed dose
3ody burden
Serum (blood) concentration
Total liver concentration
Receptor occupancy (bound concentration)
Human model
N/A
H
H
H
M/L
L
Rat model
N/A
H
H
H
H
L
Mouse model
N/A
M
M
M
M
L
H = high, M = medium, L = low, N/A = not applicable.

level of confidence can be further improved or diminished by the outcome of sensitivity analysis.
In this regard, the results of a focused sensitivity analysis indicate that the most sensitive
parameters of the human model are among the most uncertain (i.e., those parameters for which
estimates were not obtained in humans) with respect to prediction of liver TCDD concentration,
contrary to the animal model  (see Section 3.3.5).
       With respect to the mouse model, however, the level of confidence is low to medium,
given that it has not been verified extensively with blood, body burden, or tissue concentration,
time-course, or dose-response data. However, the mouse PBPK model, based on the rat model
that has been evaluated with several PK data sets, has been shown to reproduce well the limited
mouse liver kinetic data (see Figures 3-24 through 3-31; Boverhoff et al., 2005). The same
model structure has been used for simulating kinetics  of TCDD in humans successfully. Overall,
the adult mouse model, given its biological basis combined with its ability to simulate TCDD
kinetics in multiple species, is considered to exhibit a medium level of confidence for simulating
dose metrics for use in high to low dose extrapolation and interspecies (mouse to human)
extrapolation.  Even though similar considerations are applicable to gestational  model  in mice,
the confidence level is considered to be low because very limited comparison with empirical data
has been conducted (see Figure 3-31). Despite the uncertainty in these predictions, the scaled rat
gestational model, given its biological and mechanistic basis, might be of use in predicting dose
metrics  in these groups that might form the basis of PODs in certain key studies.
                                          J-C

-------
A







mi in
mu^
\l



	
1 JIIU





urn :i.





UUWi





-lUlia





jjujm.





JJOJL





JLUULUJ





UJiUJL





flJJJUi




	
LIUJdJJL





JJdOL*




* 	 CB (pg/g)
•/ ^| Cb measured





                   10    20    30    40    50    60    70    80    90   100    110    120
B
                   10    20    30    40    50    60    70    80    90   100   110   120
          10,000
                                  4	!-
               0    10    20    30    40    50    60    70    80    90   100   110   120

       Figure 3-24.  Experimental data (symbols) and model simulations (solid lines)
       of (A) blood, (B) liver, and (C) adipose tissue concentrations of TCDD after
       oral exposure to 150 ng/kg-day, 5 days/week, for 17 weeks in mice.
       Y-axis represents concentration in pg/g, and X-axis represents time in days.

       Source: Experimental data were obtained from Diliberto et al. (2001).
                                             5-84

-------
   10000000

    1000000

3   100000
o
§
8
              0.001     0.01      0.1       1        10
                                    Dose (ug/kg)
100
300
  Figure 3-25. Comparison of PBPK model simulations with experimental
  data on liver concentrations in mice administered a single oral dose of
  0.001-300 jig TCDD/kg.
  The simulations and experimental data were obtained 24 hour post-exposure.

  Source: Data obtained from Boverhoff et al. (2005).
                                       5-85

-------
   100000 i
    10000 =
•=;   1000 =
c
o

-------
A
                                                                                    v — CB (pQ/g)  I
                                                                                    V J Cb measured |
            0     10     20     30    40     50    60    70     SO    90     100   110    120
B
                                                                                   - Cli (pgg) Sim
                                                                                   | Cli (pgg) measured I
                 10    20    30    40    50    60    70    80    90    100   110    120
                                      SO    60    70    SO    90    100   110    120
       Figure 3-27. Comparison of experimental data (symbols) and model
       predictions (solid lines) of (A) blood, (B) liver, and (C) adipose tissue
       concentrations of TCDD after oral exposure to 1.5 ng/kg-day, 5  days/week,
       for 17 weeks in mice.
       Y-axis represents concentration in pg/g, and X-axis represents time in days.

       Source: Experimental data were obtained from Diliberto et al. (2001).
                                              5-87

-------
A
                 0   200
                                     1,000  1,200   1,400   1,600  1,300  2,000  Z,20(
B
                      200   400
                                  SOO   1,000  1,200
                                                  1,600  1,800  2,000   2,200
                      200   400   600   800   1,000  1,200
                                                  1,600  1,800  2,000   2,200
D
                                                              - Feces eliminatio






i
1 1 i 1
f , t j r ^
i i i ill
i i i ill
i i i ill
i i ill
/4-^^^rn r~


•


00 1,800 2,000 2,2

y 	 Urinary (% dose)


00
       Figure 3-28. Comparison of experimental data (symbols) and model
       predictions (solid lines) of (A) blood concentration, (B) liver concentration,
       (C) adipose tissue concentration, (D) feces excretion (% dose), and (E)
       urinary elimination (% dose) of TCDD after oral exposure to 1.5 ng/kg-day,
       5 days/week, for 13 weeks in mice.
       Y-axis represents concentration in pg/g, and X-axis represents time in days.
       Source: Experimental data were obtained from Diliberto et al. (2001).

-------
A
B
E

30
25
20
15
S
0





to








o



JKJr



vNVJA



wj



dl



ILJ





llLJ




tj^




*4l




Wxl



i
^x

1 «• — Cb (pg/g) |




200 400 600 300 1,000 1,200 1,400 1,600 1,300 2,000 2,200
10,000
1,000
100
10
1







;==


^^~-^—




III






	
200 400 600 300 1,000 1,200 1,400 1,600 1,£

—

	
sin
00 2,000 2,200

10,000

100



- 1"

_

200 4



	 4 	 | 	

0 600 300 1,

	

	

00 1,2



	

00 !,
-------
A
B
                           20  25   30  35  40  45  EO  55  60  65  70  75


"

3

"












L=""'
x^:
/

5 1











o i











5 2











0 2











5 3











0 3











5 4











0 4











S 5











0 5











5 6











0 6











5 7











0 7

1 ^| _ t, cfngg 1
|v>H Iine2 ]








5
E
Figure 3-30.  PBPK model simulations (solid lines) vs. experimental data
(symbols) on the distribution of TCDD after a single acute oral exposure to
A-B) 0.1, C-D) 1.0, and E-F) 10 ug of TCDD/kg of body weight in mice.
Liver and adipose concentration for each dose was measured after 72 hours. Y-axis represents
the concentration in tissues (ng/g); insets A, C, and E represent liver tissue, whereas B, D, and F
correspond to adipose tissue. X-axis represents the time in hours.

Source: Experimental data were obtained from Santostefano et al. (1996).
                                      5-90

-------
A
B
                                                                                             * 	Cb (ng/g)
                                                                                             */ B experimental
            280  282  284 286  288  290  292  294 296  298  300  302  304 306 308  310  312  314 316 318  320
                                                                                              <:— Cli (ng/g)
                                                                                              1 ^| experimental
            280  282  284 286 288  290  292  294  296 298  300  302  304  306 303  310  312  314  316 318  320
                                                                                             V 	Cf (ng/g)
                                                                                             •/ ^| Experimental
            280  282  284 286 288  290  292  294  296 298  300  302  304  306 308  310  312  314  316 318  320
        Figure 3-31.  PBPK model simulation (solid lines) vs. experimental data
        (symbols) on the distribution of TCDD after a single dose of 24 jig/kg BW on
        GD 12 in mice.
        Concentrations expressed as ng TCDD/g tissue.  (A) maternal blood, (B) maternal liver, and
        (C) maternal adipose tissue. Y-axis represents the tissue concentration, whereas X-axis represents
        the time in hours.

        Source: Experimental data were obtained from Abbott et al. (1996).
                                                    3-91

-------
3.3.4.4. Applicability of Pharmacokinetics (PK) Models to Derive Dose Metrics for Dose-
        Response Modeling ofTCDD: Confidence and Limitations
       Both the CADM and PBPK models describe the kinetics ofTCDD following oral
exposure to adult animals and humans by accounting for the key processes affecting kinetics,
including hepatic sequestration phenomena, induction, and nonlinearity in elimination, and
distribution in adipose tissue and liver. Both models can be used for estimating body burdens
and serum lipid adjusted concentrations ofTCDD.  However, there are several differences
between these two models.  The PBPK model calculates the free and bound concentrations of
TCDD in the intracellular subcompartment of tissues.  The total or receptor-bound
concentrations in liver are unambiguous and more easily interpretable with the PBPK model than
with the CADM model. In addition, the PBPK model computes bound and total concentrations
as a function of the free concentration in the intracellular compartment of the tissue.  By contrast,
the CADM model simulates the total  concentration based on empirical consideration of hepatic
processes. Consequently, the amount ofTCDD bound to AhR or CYP1A2 cannot be simulated
with the CADM model. The CADM model computes only the total TCDD concentration in liver
and describes TCDD elimination through partitioning from circulating lipids across the lumen of
the large intestine into the feces, while the PBPK model accounts for this process empirically
within its hepatic elimination constant. Elimination ofTCDD via skin, a minor process, is not
described by either model.  Thus, dose-response modeling based on body burden ofTCDD in
adult animals and humans can be conducted with either of the models, provided the duration of
the experiment is at least 1 month, due to limitations in the CADM model.  As shown in
Figure 3-32, the predicted slope and body burden over a large dose range are quite comparable
(generally within a factor of two).
       Results of simulations of serum lipid concentrations or liver concentrations vary for the
two models to a larger extent (up to a factor of 7), particularly for simulations of short duration.
These differences reflect two characteristics of the PBPK model: first, quasi-steady-state is not
assumed in the PBPK model; second, the serum lipid composition used in the model  is not the
same as the adipose tissue lipids. The CADM model does not account for differential solubility
ofTCDD in serum lipids and adipose tissue lipids, nor does it account for the diffusion-limited
uptake by adipose tissue. Therefore, the PBPK model would appear to be superior to the CADM
model with respect to the ability to simulate serum lipid and tissue concentrations during
exposures that do not lead to the onset of steady-state condition in the exposed organism.
                                         3-92

-------
                                                       	CADM(rat)
                                                       	Emond (rat)
                                                       	CADM (human)
                                                           Emond (human)
                        2000       4000       6000      8000
                                        Intake (ng/kg-day)
10000
12000
       Figure 3-32. Comparison of the near-steady-state body burden simulated
       with CADM and Emond models for a daily dose ranging from 0 to
       10,000 ng/kg-day in rats and humans
       The rat model was ran for 13 weeks, and the human model was ran from ages 20 to 30. The
       time-averaged concentration was used for each.
       The CADM model is less complex than the PBPK model and has fewer parameters.
Because the CADM model is constructed by fitting to data, its performance is likely to be
reliable for the range of exposure doses, species, and life stages from which the parameter
estimates were obtained.  On the other hand, the PBPK model structure and parameters are
biologically based and can be adapted for each species and life stage. Accordingly, the PBPK
model has been adapted to simulate the kinetics of TCDD in the human fetus and in pregnant
rats, as well as in adult humans and rats (Emond et al., 2006; Emond et al., 2005; Emond et al.,
2004).  The time step for calculation and dosing in the CADM model corresponds to 1 month.
This requirement represents a constraint in terms of the use of this model to simulate a variety of
dosing protocols used in animal toxicity studies.  This requirement, however, is not a constraint
with the PBPK models.  So, either model would appear to be useful when simulating the body
burden and serum lipid concentrations following a longer duration of exposure; but the PBPK
model would be preferred for simulating alternative dose metrics of TCDD (e.g., blood
concentration, total tissue concentration, bound concentration) for various exposure  scenarios

-------
(including single dose studies), routes, and life stages in the species of relevance, to TCDD
dose-response assessment, particularly, mice, rats, and humans.
       Two minor modifications, to enhance the biological basis, were made to the PBPK model
of Emond et al. (2006), before its use in the computation of dose metrics for TCDD. The first
one involved the recalculation of the volume of the rest of the body as follows:
where
WREO

WLIO

WFO

WREBO

WLIBO

WFBO
                                                                            (Eq. 3-24)
                = weight of cellular component of rest of body compartment (as fraction of
                  body weight);
                = weight of cellular component of liver compartment (as fraction of body
                  weight);
                = weight of cellular component of fat compartment (as fraction of body
                  weight);
                = weight of the tissue blood component of the rest of body compartment (as
                  fraction of body weight);
                = weight of the tissue blood component of the liver compartment (as fraction
                  of body weight); and
                = weight of the tissue blood component of the fat compartment (as fraction of
                  body weight).
       In the original code, the weight of the rest of body compartment was calculated as the
difference between 91% of body weight and the sum total of the fractional volumes of blood,
liver tissue (intracellular component), and adipose tissue (intracellular component). The blood
compartment in the PBPK model is not explicitly characterized with a volume; as a result, the
total volume of the compartments is less than 91%.  The recalculations shown above were used
to address this problem. Given the very low affinity of TCDD for blood and rest of the body,
reparameterizing the model resulted in less than a 1% change in output compared to the
published version of the PBPK model for chronic exposure scenarios (Emond et al., 2006).
       The second minor modification related to the calculation of the rate of TCDD excreted
via urine. The  original model code computed the rate of excretion by multiplying the urinary
clearance parameter with the concentration in the rest of the body compartment. Instead, the
                                          -94

-------
code was modified to use the blood concentration in this equation.  This resulted in the
reestimation of the urinary clearance value in the rat and human models, but it did not result in
any significant change in the fit and performance of the original model. In addition to the minor
modifications in the model structure, a recalibration of the gastric nonabsorption constant of the
PBPK model was conducted to match human oral bioavailability data (Poiger and Schlatter,
1986).
       The revised parameter estimates of the rat, mouse, and human models are captured in
Table 3-8 with a footnote.

3.3.4.5. Recommended Dose Metrics for Key Studies
       The selection of dose metrics for the dose-response modeling of key studies is largely the
result of (1) the relevance of a dose metric on the basis of current knowledge of TCDD's
mechanism of action for critical endpoints and (2) the feasibility and reliability of obtaining the
dose metric with available PK models. Secondarily, the goodness-of-fit of the dose-response
models (which reflects the relationship of the selected internal dose measures to the response)
can be used to inform selection of the most appropriate dose metric for use in deriving TCDD
toxicity values.
       Body burden—even though this metric is based on mechanistic considerations—is a
somewhat distant measure of dose with respect to target tissue dose, and this metric represents
the "overall" average concentration of TCDD in the body. However, a benefit of body burden is
that this metric represents a dose measure for which the available PK models can provide highly
certain estimates.  Thus, the overall confidence associated with the use of body burden in TCDD
assessment is categorized as medium.
       The confidence in the ability of PK models to simulate blood concentration as a dose
metric is high, given that the models have been shown to consistently reproduce whole blood (or
serum lipid-normalized) TCDD concentration profiles in both humans and rats. Considering the
facts that the PBPK models simulate whole blood rather than the serum lipid-normalized
concentrations of TCDD and that the study-specific values of serum lipid content are not known
with certainty, it is preferable to rely on TCDD blood concentrations as the dose metric. The
blood concentrations, if intended, can be normalized on the basis of appropriate total lipid levels.
However, based on mechanistic considerations, the confidence in their use would be somewhat

                                         3-95

-------
lower for hepatic effects. This conclusion reflects the concern regarding the inconsistent
relationship between the two variables with increasing dose levels and the fraction of
steady-state attained at the time of observation. For other systemic effects related to tissue
concentrations, the confidence in the use of TCDD serum or blood concentration is high,
particularly for chronic exposures, given the absence of data on organ-specific nonlinear
mechanisms.  In general, the tissue concentration typically cannot be calculated as a reliable dose
metric with either the CADM or the Emond models.  One exception is the use of the Emond
PBPK models to estimate levels in liver, a metric that is relevant based on MOA considerations.
However, it is noted that the hepatic TCDD level encompasses free and bound TCDD, and it is a
highly complex entity for dose metric considerations.  Finally, the AhR-bound concentration
may be evaluated for receptor-mediated effects.  This  dose metric can be obtained by PBPK
models, although uncertainties associated with the lack of data for this dose metric render it to be
of low confidence (see Table 3-19).  The alternative dose metrics for dose-response modeling of
TCDD selected on the basis of MOA and PK modeling considerations are summarized in
Tables 3-20 and 3-21.
       Table 3-20.  Overall confidence associated with alternative dose metrics for
       noncancer dose-response modeling for TCDD using rat PBPK model
End point
Liver effects
Nonhepatic effects
Body burden
M
M
Blood or serum
concentration

H
Liver
concentration
H

Bound
concentration in
liver
M/L
M/L
H = high, M = medium, L = low.

       Table 3-21. Overall confidence associated with alternative dose metrics for
       noncancer dose-response modeling for TCDD using mouse PBPK model


End point
Liver effects
Nonhepatic effects


Body burden
M
M

Blood or serum
concentration

M

Liver
concentration
M

Bound
concentration in
liver
L
L
H = high, M = medium, L = low.
                                          5-96

-------
       These measures of internal dose can be obtained as peak, average, integral (AUC), or
terminal values.  For chronic exposures in rodents (ca. 2 years), the terminal and average values
would be fairly comparable under steady-state conditions.  For less-than lifetime exposures,
however, the terminal and average values will differ, and, therefore, an overall average or
integrated value (AUC) would be more appropriate.  Similarly, for developmental exposures,
these alternative  dose metrics can be obtained with reference to the known or hypothesized
exposure window of susceptibility.

3.3.5.  Uncertainty in Dose Estimates
3.3.5.1. Sources of Uncertainty in Dose Metric Predictions
3.3.5.1.1.  Limitations of available pharmacokinetics (PK) data
3.3.5.1.1.1.  Animal data
       The available animal data relate to blood, liver, and adipose tissue concentrations for
certain exposure doses and scenarios. Although these data are informative regarding the dose-
and time-dependency of TCDD kinetics for the range covered by the specific studies (see
Section 3.3.2), they do not provide the peak, average, terminal, or lipid-normalized values of
dose metrics associated with the key studies selected for this assessment. The limited available
animal PK data are useful, however, in the evaluation of the pharmacokinetic models (see
Section 3.3.4).

3.3.5.1.1.2.  Human data
       The human data on potential dose metrics are restricted to the serum lipid-adjusted
TCDD concentrations associated with mostly uncharacterized exposures (see Sections 3.3.2 and
3.3.3). While these data are useful in estimating half-lives in exposed human individuals, they
do not provide estimates of hepatic clearance or reflect target organ exposure.  Some autopsy
data have been used to infer the partition coefficients; however, these data were collected
without quantification of the temporal nature of TCDD uptake (see Section 3.2). Despite the
limitations associated with the available human data, there has been some success in using these
data to infer the half-lives and elimination rates in humans using pharmacokinetic models
(Emond et al.. 2006: Aylward et al.. 2005b: Carrier et al.. 1995a).
                                           5-97

-------
3.3.5.1.2.  Uncertainties associated with model specification
       Uncertainty associated with model specification should be viewed as a function of the
specific application, such as interspecies extrapolation, intraspecies variability, or
high-dose-to-low-dose extrapolation.  Because the use of pharmacokinetic models in this
assessment is limited to interspecies extrapolation and high-dose-to-low-dose extrapolation, it is
essential to evaluate the confidence in predicted dose metrics for these specific purposes. For
interspecies extrapolation, the PBPK and CADM models calculate differences in dose metric
between an average adult animal and an average adult human. Both models have a biologically
and mechanistically relevant structure along with a set of parameters with reasonable biological
basis and reproduce a variety of pharmacokinetic data on TCDD in both rodents and humans.
These models possess low uncertainty with respect to body burden, blood, and TCDD/serum
(lipid) concentration for the purpose of conducting rat to human extrapolation. However, for
other dose metrics, such  as free, total, or bound hepatic concentrations, the uncertainty is higher
in the CADM model compared to the PBPK model due to model specification differences related
to the mechanisms of sequestration and induction in the liver (see Section 3.3.3).
       For the purpose of high-dose-to-low-dose extrapolation in experimental animals,
confidence in both models is high with respect to a variety of dose metrics (see previous
discussion).  The high confidence results from the use of the PBPK models to reproduce a
number of data sets covering a wide range of dose levels in rodents (i.e., rats, mice) including the
dose ranges of most of the key toxicological studies.  Given that the TCDD levels during and at
the end of exposures were not measured in most of the key studies, use of the PBPK models is
preferred because these models account for dose-dependent elimination, induction, and
sequestration.  Despite the empirical nature of the  specification of these key processes in PBPK
models, they essentially reproduce the dose-dependent behavior in rodents, supporting their use
in deriving dose metrics for dose-response modeling of TCDD. Overall, the confidence in the
use of the alternative dose metrics  (identified in Table 3-19) is greater than the confidence in the
use of administered dose for TCDD, for relating to the concentration within tissues to produce an
effect.  The administered dose does not take into account interspecies differences in the volume
of distribution and clearance or the complex nonlinear processes determining the internal dose.
       The PBPK model of Emond et al. (2006) could benefit from further refinement and
validation, including a more explicit consideration of dose-independent elimination pathways.

                                          3-98

-------
As indicated in Section 4, there is some uncertainty associated with the way the elimination of
TCDD is described in the existing human PBPK model. The current model essentially treats all
TCDD elimination as related to dose-dependent metabolism in the liver.  In this regard, the
classical and more recent PK data on TCDD may be useful in further improving the confidence
in their predictions.  However, it is likely that there is dose-independent elimination of TCDD
via feces and, to a lesser extent, skin; juxtaposition of available elimination rate data with the
PBPK model predictions suggests that the current PBPK model modestly overestimates the
dose-dependency of overall TCDD elimination. (The central estimate  of the slope of the
relationship between the log of the TCDD elimination rate and the log  of the TCDD level is only
about three-fourths of that expected using the unmodified PBPK model). Emond et al. (2005)
acknowledge that the model did not describe the elimination of TCDD from the blood into the
intestines, but it indirectly accounted for this phenomenon with the use of the optimized
elimination rate.

3.3.5.1.3.  Impact of human interindividual variability
       The sources and extent of human variability suggested by the available data are presented
in Section 3.3.3, although there is some discussion of the impact of individual differences in
body fat content. The CADM model facilitates the simulation of body burden and serum lipid
concentrations on the basis of BMI and tissue weights of people,  and the PBPK model simulates
alternative dose metrics in the fetus and in pregnant animals in addition to adult animals and
humans.  However, neither of these models has been parameterized for simulation of population
kinetics and distribution of TCDD dose metrics. Therefore, at the present time, a quantitative
evaluation of the impact of human variability on the dose metrics of TCDD is not feasible, and
dose metric-based replacement of the default interindividual factor has not been attempted.

3.3.5.2. Qualitative Discussion of Uncertainty in Dose Metrics
       The usefulness of the CADM and PBPK models for conducting dose-response modeling
(rodent bioassays), interspecies (rodent to human) and intraspecies (high-dose to low-dose)
extrapolations  is determined by their reliability in predicting the desired dose metrics. The
confidence in the model predictions of dose metrics is dictated by the extent to which the model
has been verified with empirical data relevant to the dose metric,  supplemented by sensitivity

                                          3-99

-------
and uncertainty analyses.  Analysis of sensitivity or uncertainty has not been conducted with the
CADM model.  For the PBPK model, Emond et al. (2006) published the initial results from
sensitivity analyses of acute exposure modeling (see Section 3.3.3).  One of the objectives of a
sensitivity analysis that is of highest relevance to this assessment is the identification of the most
critical model parameters with respect to the model output (i.e., dose metric).
       If the model simulations have only been compared to entities that do not correspond to
the moiety representing the dose metric, or if the comparisons have only been done for some but
not all relevant dose levels, routes, and species, then the reliability in the predictions of dose
metric can be an issue.  The extent to which model results are uncertain will depend largely upon
the extent to which the dose metric is measurable (e.g., serum concentrations of TCDD) or
inferred (e.g., AhR-bound TCDD concentration).
       With respect to TCDD body burden, whole-liver and blood concentration predictions in
the rat model, which are well-calibrated with measured data, uncertainty is relatively low.
Therefore, the need for sensitivity and uncertainty analysis is less critical, and confidence in
these dose metrics is high. For those dose metrics that are not directly measurable or are less
easily verified by available calibration methods, such as free-liver and AhR-bound
concentrations, sensitivity and uncertainty analyses are crucial for assessing the reliability of
model predictions, and confidence is low.  For the human model, calibration is largely dependent
on blood (LASC) TCDD measurements, which are much less extensive than for the rat model.
Because the blood measurements are reported as  LASC, uncertainty and variability in
serum:blood and fatserum ratios also are a factor when evaluating the adequacy of the
whole-blood TCDD metric. Furthermore, the human data are mostly representative of much
higher exposures than the environmental exposures of interest to the EPA.  Because of these
additional uncertainties, only medium confidence can be held in the human model whole-blood
TCDD concentration predictions at higher exposures (observed effect range) and low-to-medium
confidence at lower exposures (background exposure range).
       Sensitivity analysis for the Emond rat PBPK  model predictions  of liver TCDD
concentration indicated that hepatic CYP1A2 concentration is the most sensitive parameter
(Emond et al., 2006). For the Emond human PBPK model, the absorption parameters, basal
concentration of CYP1A2, and adipose tissue:blood partition coefficients were identified as
highly sensitive parameters.
                                         3-100

-------
       Confidence in the Emond rat and human PBPK models at high exposures is medium for
the purpose of rat-to-human extrapolation based on blood concentrations, given that the key
human model parameters are both sensitive and uncertain; confidence is low for lower
exposures.  Conversely, confidence in the use of AhR-bound TCDD is low because of the large
uncertainty in the fraction of AhR-bound TCDD in the liver.
       With regard to the predictability of body burden, the absorption and excretion parameters
were among the sensitive parameters in the rat.  Several other parameters were also identified as
being sensitive in humans. Despite the sensitivity to these parameters and the uncertainty
associated with individual parameter estimates, the overall confidence in the model predictions
of body burden appears to be high given the reproducibility  of empirical data on tissue burdens
and blood concentrations of TCDD in various experiments by both models. Similar conclusions
can be drawn for blood concentration of TCDD predicted by the PBPK model, except that the
assigned value of blood (serum) lipid content will have additional impact on this  dose metric to
the extent that the calibration data were in terms of LASC. Variability of total lipid levels and
variability of the contribution of phospholipids and neutral lipids to the total lipid pool across
species, lifestage, and study groups is to be expected (Bernert et al., 2007; Poulin and Theil,
2001).
       Both conceptual (biological) relevance and prediction uncertainty are important in the
choice of dose metric for dose-response modeling and interspecies extrapolation. Conceptual
relevance has to do with how "close" the metric is to the observed effect, taking into account
both the target tissue and the MOA. In this context, a greater degree of confidence is held for
dose metrics that are more proximate to the event (i.e., specific effect).  Prediction uncertainty
reflects the lack of confidence in the model predictions of dose metrics.  Tables 3-22 and 3-23
provide a qualitative ranking of the importance and magnitude of each dose metric with respect
to these two sources of uncertainty. Conceptual relevance is low for the use of administered
dose in dose-response modeling because known (nonlinear)  physiological processes are ignored;
conversely, conceptual uncertainty is much lower for use of internal dose metrics more proximal
to the affected organs.
                                         3-101

-------
       Table 3-22.  Contributors to the overall confidence in the selection and use of
       dose metrics in the dose-response modeling of TCDD based on rat and
       human PBPK models"
Dose metric
Administered dose
3ody burden
Blood concentration
Jver concentration
Receptor (AhR)
occupancy
Conceptual relevance
L
M
M
L
H
Prediction uncertainty
NA
M
L
M
H
Overall confidence
L
M-L
M
L
L
aUsing professional judgment, EPA ranked its confidence in the CADM model as low, medium, or high (or not
 applicable) based on model simulations of administered dose, absorbed dose, body burden, serum lipid
 concentration, total tissue (liver) concentration, and receptor occupancy.
H = high, M = medium, L = low, NA = not applicable.
       Table 3-23.  Contributors to the overall uncertainty in the selection and use
       of dose metrics in the dose-response modeling of TCDD based on mouse and
       human PBPK models
Dose metric
Administered dose
Absorbed dose
Body burden
Blood or serum concentration
Tissue concentration
leceptor occupancy
Conceptual uncertainty
H
H
M
M
L
L
Prediction uncertainty
NA
L
M
M
M/R
H
H = high, M = medium, L = low, NA = not applicable

       Table 3-22 presents a cross-walk of relevance, uncertainty, and overall confidence
associated with the use of various dose metrics for dose-response modeling of TCDD. Using
professional judgment, EPA ranked its confidence in PBPK models as low, medium, or high (or
not applicable) based on model simulations of administered dose, absorbed dose, body burden,
serum lipid concentration, total tissue (liver) concentration, and receptor occupancy. As shown
in Table 3-22, blood/serum concentrations have the highest overall confidence (medium),
followed by body burden (medium to low) for application in dose-response modeling. When
using the mouse PBPK model along with the human model (see Table 3-23), the contribution of
                                         3-102

-------
the prediction uncertainty to the overall uncertainty increases due to the limited comparison of
the mouse model simulations with empirical data.

3.3.6.  Use of the Emond Pysiologically Based Pharmacokinetic (PBPK) Models for Dose
       Extrapolation from Rodents to Humans
       EPA has selected the Emond et al. (2006: 2005: 2004) PBPK models, as modified by
EPA for this assessment, for establishing toxicokinetically equivalent exposures in rodents and
       9Q 	
humans.   The 2003 Reassessment (U.S. EPA, 2003) presented a strong argument for using the
relevant tissue concentration as the effective dose metric.  However, no models exist for
estimation of all relevant tissue concentrations.  Therefore, EPA has decided to use the
concentration of TCDD in blood as a surrogate for tissue concentrations, assuming that tissue
concentrations are proportional to blood concentrations. Furthermore, because the RfD is
necessarily expressed in terms of average daily exposure, the blood concentrations are expressed
as averages over the relevant period of exposure for each endpoint.  Specifically, blood
concentrations in the model simulations are averaged from the administration of the first  dose to
the administration of the last dose plus one dosing interval (time) unit in order to capture the
peaks and valleys for each administered dose. That is, for daily dosing, 24 hours of TCDD
elimination following the last dose is included in the average (the modeling time interval is
1 hour); for a weekly dosing protocol, a full week is included. In addition, because of the
accumulation of TCDD in fat and the large differences in elimination kinetics between rodent
species and humans, exposure duration plays a much larger role in TK extrapolation across
species than for rapidly eliminated compounds.  Because of these factors, EPA is using discrete
exposure scenarios that relate human and rodent exposure durations. The use of discrete
exposure scenarios was introduced previously in Section 3.4.4.2 describing first-order kinetic
modeling and is further described in the following paragraphs.  This section concludes with a
quantitative evaluation of the impact of exposure duration on the rodent-to-human TK
extrapolation from both the human and rodent "ends" of the process.
       Figure 3-33 shows the TCDD blood concentration-time profile for continuous exposure
at 0.01 ng/kg-day, as predicted by the Emond human PBPK model, and the target TCDD
concentrations corresponding to the three discrete exposure scenarios used by EPA in this
29 The models will be referred to hereafter as the "Emond human PBPK model" and the "Emond rodent PBPK
model," with variations when referring to individual species or components (e.g., gestational).
                                         3-103

-------
document.  The target concentrations are those that would be identified in the animal bioassay
studies that correspond to a particular POD (no-observed-adverse-effect level,
lowest-observed-adverse-effect level, or benchmark dose lower confidence bound) established
for that bioassay.  That is, the target concentrations represent the toxicokinetically equivalent
internal exposure to be translated into an equivalent human intake (or HED).
       For the lifetime exposure scenario, the HED is "matched" to the lifetime average TCDD
blood concentration from a lifetime animal bioassay result by determining the continuous daily
intake that would result in that average blood concentration for humans over 70 years. A table
for converting lifetime-average blood concentrations and other internal dose metrics to daily
human TCDD intake rates is presented in Appendix E.4.
   D)
   D)
  .0
  -l-»
  TO
   
-------
       For the gestational exposure scenario, the effective TCDD blood concentration (usually
the peak) determined for the particular POD in a particular developmental study is matched to
the average TCDD blood concentration over the gestational portion of the human gestational
exposure scenario. The HED is determined as the continuous daily intake, starting from birth
that would result in that average blood concentration over the 9-month gestational period for a
pregnancy beginning at 45 years of age.  The choice of 45 years as the beginning  age of
pregnancy is conservative in that the daily exposure achieving the target blood concentration is
smaller than for pregnancies occurring earlier in life (e.g., a pregnancy beginning at 30 years of
age).  A table for converting average gestational blood concentrations and other internal dose
metrics to human intake for the 45-year-old pregnancy scenario is presented in Appendix E.4.
Also, a comparison of the 45-year old pregnancy scenario to one beginning at age 25 is presented
in Table 3-24. Using the 25 year-old pregnancy scenario increases the HED by 30 to 60% for
typical animal bioassay PODs (3 to 30 ng/kg).
       Table 3-24.  Comparison of human equivalent doses from the Emond human
       PBPK model for the 45-year-old and 25-year-old gestational exposure
       scenarios
Animal
bioassay POD
(ng/kg-day)
3
30
Species
Mouse
Rat
Mouse
Rat
TCDD
blood
concentration51
8.800E-02
1.815E-01
7.115E-01
1.367E+00
HED
45 year-old
6.79E-04
1.87E-03
1.51E-02
4.22E-02
HED
25 year-old
1.03E-03
2.98E-03
2.07E-02
5.41E-02
25-yr:45-yr
ratio
1.5
1.6
1.4
1.3
"Determined from the Emond rodent PBPK models assuming a single exposure on GD 13.
       For a less-than-lifetime exposure, the average TCDD blood concentration over the
exposure period in the animal bioassay associated with the POD is matched to the average over
the 5-year period that includes the peak concentration (58 years for an intake of 0.01 ng/kg-day).
The FLED is determined as the continuous daily intake that would result in the target
concentration over peak 5-year period. The use of the peak is analogous to the approach in the
2003 Reassessment, where the terminal steady-state body burden played the same role. The
5-year average over the peak is taken to smooth out sharp peaks and more closely approximate a
                                         3-105

-------
plateau. The choice of peak is health protective because humans of any age must be protected
for short-term exposures, and the daily intake achieving a given TCDD blood concentration is
smallest when matched to the peak exposure as opposed to an average over shorter durations.
Thus, target concentrations for any exposure duration of less-than-lifetime must be averaged
backwards from the end of the lifetime scenario, rather than from the beginning.  The only
exception would be if the short-term endpoints evaluated in the animal bioassay were associated
with a specific life stage (such as for the gestational scenario). Note that this scenario lumps all
exposures from 1 day to over 1 year in rodents into the same less-than-lifetime category.
Conceptually, duration-specific scenarios could be constructed by defining equivalent rodent and
human exposure durations. However, for the most part, defining duration equivalents across
species is  a somewhat arbitrary exercise, not generally based on physiologic or toxicological
processes, but relying primarily on fraction-of-lifetime conversions. EPA defines "lifetime"
exposure as 2 years and 70 years for rodents and humans, respectively. So, a half-lifetime
equivalence of 1 year in rodents and 35 years in humans is defined easily.  Also, considering a
subchronic exposure to be 10-15% of lifetime, leads to an equivalence of 90 days in rodents and
7-10 years in humans. However, in the practical sense with respect to the Emond human PBPK
model predictions, the differences in the dose-to-target-concentration ratios are not significantly
dissimilar from the peak 5-year average scenario, differing by less than 5%. A table for
converting less-than-lifetime average blood concentrations and other internal dose metrics to
human intake is presented in Appendix E.4.
       The net effect of using three different scenarios for estimating the HED from rodent
exposures is that, for the same target concentration, the ratio of administered dose (to the rodent)
to HED will be larger for short-term exposures than for chronic exposures. Figure 3-34 is
similar to  Figure 3-33, except that it shows the relationship of daily intake to a fixed target
TCDD blood concentration level.  Figure 3-34 shows that, for human intakes of approximately
0.01 ng/kg-day, the difference in the defined scenarios is 40% or less, with a lifetime-scenario
daily intake of 0.014 ng/kg-day required to reach the same target concentration for a shorter-term
exposure of 0.01 ng/kg-day.  The corresponding daily intake for the gestational scenario is
0.011 ng/kg-day. Because of the nonlinearities in the Emond human PBPK model, the
magnitude of the difference between the lifetime and less-than-lifetime exposure scenarios
increases at lower intake levels, but not to a substantial degree.
                                          3-106

-------
  O)
  O)
o
o
•a
o
.o
.a
Q
Q
O
      o
      ID
      CM
      o
      CM
  ™   10 -
      o
      o
      o
      ID
      O -
                Lifetime scenario
                Less-than-lifetime scenario
                Gestational scenario
                Target concentration
                               20
                                                   r
                                                  40
 r
60
                                              Year
       Figure 3-34. TCDD serum concentration-time profile for lifetime, less-than-
       lifetime and gestational exposure scenarios, showing continuous intake levels
       to fixed target concentration; profiles generated with Emond human PBPK
       model.
       The differential effect of short- and long-term exposures is much more accentuated at the
rodent end of the exposure kinetic modeling. Analogous to the processes described in the
previous section for first-order body burden (see Section 3.3.4.2), the TCDD blood concentration
for single exposures is essentially the immediate absorbed fraction of the administered dose,
which will be somewhat lower than the administered dose, while for chronic exposure, the
TCDD blood concentration will reflect the long-term accumulation from daily exposure, which
will be very much larger than the administered dose (expressed as a daily intake).  Table 3-25
shows the overall impact of TK modeling on the extrapolation of administered dose to HED,
comparing the Emond PBPK and first-order body burden models.  For comparison purposes, the
                                         3-107

-------
administered dose is fixed at 1 ng/kg-day for all model runs.  Large animal-to-human TK
extrapolation factors (!KEF) are evident for short-term mouse studies, decreasing in magnitude
with increasing exposure duration.  The only exception is the slightly lower extrapolation factor
for the mouse 1-day exposure, which is the result of the relatively short TCDD half-life (10 days)
in mice and the use of the peak TCDD blood concentration as representative of single exposures,
compared to the average TCDD blood concentration over the exposure period used for multiple
exposures.  The TKpFS are lower for rats because of the slower elimination of TCDD in rats
compared to mice.  Also, because of the nonlinear kinetics inherent in the Emond PBPK model,
the span of the HED (13-fold for mice) across these exposure durations is greater than the span
of the LASC (fourfold for mice). Because of the dose-dependence of TCDD elimination in the
Emond model, the TKnF becomes smaller with decreasing intake. The result of this nonlinearity
is that, although Table 3-25 shows much lower TR^pS for the Emond PBPK model than for the
first-order body burden metric, at much lower HED levels, the predictions of the two models are
much closer.
       Table 3-25. Impact of toxicokinetic modeling on the extrapolation of
       administered dose to HED, comparing the Emond PBPK and first-order
       body burden models (administered dose = 1 ng/kg-day)
Exposure
duration (days)
lst-order BB
HEDa
(ng/kg-day)
TKEFb
Emond PBPK
LASCC
(ng/kg)
HED
(ng/kg-day)
TKEF
Mouse
1
14
90
365
730
2.57E-4
1.47E-3
3.25E-3
3.70E-3
4.43E-3
3,882
681
307
270
226
75.5
64.4
173
248
263
9.49E-4
8.17E-4
3.83E-3
6.66E-3
1.08E-2
1,054
1,224
261
150
93
Rat
1
14
90
365
730
2.63E-4
1.76E-3
6.13E-3
8.68E-3
1.07E-2
3,802
569
163
115
93
110
208
599
811
853
1.87E-3
5.22E-3
2.81E-2
4.52E-2
6.47E-2
535
192
36
22
15
     "Human-equivalent doses.
     bRodent-to-human toxicokinetic extrapolation factor.
     °Lipid-adjusted serum concentration.
                                         3-108

-------
                            4. ORAL REFERENCE DOSE

       This section presents U.S. EPA's response to the NAS recommendations that EPA
discuss more explicitly the modeling of noncancer endpoints and develop a RfD to address
noncancer effects associated with oral 2,3,7,8-TCDD exposures.  Section 2 details the selection
of the animal bioassays with the lowest TCDD doses associated with the development of adverse
noncancer effects and the selection of relevant epidemiologic studies of adverse noncancer health
effects.  Section 3 discusses the kinetic modeling and estimation of human equivalent daily oral
doses that are used in TCDD RfD development in this section.  This section discusses the
modeling of noncancer health effects data associated with TCDD exposure and the derivation of
an RfD. Specifically, Section 4.1 summarizes the NAS comments on TCDD dose-response
modeling and EPA's response, including justification of selected noncancer effects and statistical
characterization of modeling results. Section 4.2 presents the TCDD dose-response modeling
undertaken for identification of candidate PODs for derivation of an RfD.  In Section 4.3, EPA
derives an RfD for TCDD. Section 4.4 describes the qualitative uncertainties in the RfD.
Finally, Section 4.5 presents two separate focused quantitative analyses of uncertainty for the
TCDD RfD. The first focuses on three data sets (from two epidemiologic studies and one  animal
bioassay) and quantifies the consequences of alternative decisions in the development of PODs
based on these  studies. The second develops POD estimates for several studies, some of which
did not qualify for consideration for RfD derivation in the study selection process, but could be
considered in the context of investigating uncertainty limits for the RfD.

4.1.   NAS  COMMENTS AND EPA'S RESPONSE ON IDENTIFYING NONCANCER
      EFFECTS OBSERVED AT LOWEST DOSES
       The NAS recommended that EPA identify the noncancer effects associated with
low-dose TCDD exposures and discuss its strategy for identifying and selecting PODs for
noncancer endpoints, including biological significance of the effects.
       With respect to noncancer end points, the committee notes that EPA does not use
       a rigorous approach for evaluating evidence from studies... (p. 47 NAS, 2006b)
                                          4-1

-------
       The Reassessment should describe clearly the following aspects:
          1.   The effects seen at the lowest body burdens that are the primary focus for
              any risk assessment—the "critical effects."
          2.   The modeling strategy used for each noncancer effect, paying particular
              attention to the critical effects, and the selection of a point of comparison
              based on the biological significance of the effect; if the EDoi is retained,
              then the biological significance of the response should be defined and the
              precision of the estimate given... (p. 187, NAS, 2006b).
       In this document, EPA has developed a strategy for identifying the noncancer data sets
and PODs that represent the most sensitive and toxicologically-relevant endpoints for derivation
of an RfD for TCDD.  EPA began this process by using the animal bioassays and epidemiologic
studies that met its study inclusion criteria as sources of these data sets.
       For all noncancer epidemiologic studies that were identified as suitable for further
quantitative dose-response analyses in Section 2.4.1, EPA has chosen to use NOAELs and
LOAELs to identify PODs; BMD modeling was  not feasible given the nature of the data
presented in these studies.  Figure 4-1 shows EPA's process for determination of PODs from
these key epidemiologic studies. EPA first evaluated the dose-response information in the study
to determine whether it provided an estimate of TCDD exposure and an observed health outcome
that was lexicologically relevant30 for RfD derivation. If such data were available, EPA
identified a NOAEL or LOAEL as a POD.  For each of these, EPA applied a toxicokinetic model
to estimate the continuous oral daily intake associated with the POD that could be used in the
derivation of an RfD (see Section 4.2.3). If all of this information was available, the result was
included as a POD for derivation of a candidate RfD.
       Figures 4-2 and 4-3 together present the strategy EPA used to  evaluate the study/endpoint
combinations found in the noncancer animal bioassays that met EPA's study inclusion criteria in
Section 2,4.2, estimate PODs, and develop a final set of candidate RfDs for TCDD. Figure 4-2
summarizes the disposition of the 78 animal noncancer studies selected for TCDD dose-response
analyses. Of these studies, 16 were eliminated because EPA determined that they contained no
toxicologically-relevant endpoints that could be used to derive a candidate RfD (discussed
30 RfDs are based on health endpoints that are inherently adverse or clearly linked to downstream functional or
pathological alterations (U.S. EPA. 2002).
                                           4-2

-------
             List of key noncancerepidemiologicstudiesfor
             quantitative dose-response analysis of TCDD
           No
           Does the
       study provide data
on n on cancer effects and TCDD
xposurefor determining a POD o
    a toxicologically relevant
           endpoint?
                                    Identify a study NOAELor LOAEL
                                        foruse in POD estimation
                                       Use kinetic model to estimate
                                  continuous oral daily intake (ng/kg-day)
                                      in the affected study population
 Exclude study from
   POD estimation
                          Include as POD
Figure 4-1. EPA's process to identify and estimate PODs from key
epidemiologic studies for use in noncancer dose-response analysis of TCDD.
For each noncancer study that qualified using the study inclusion criteria, EPA evaluated the
dose-response information developed by the study authors to evaluate whether the study provided
noncancer effects and TCDD dose data for a toxicologically relevant endpoint. If such data were
available, EPA identified a NOAEL or LOAEL as a POD. Then, EPA used a human kinetic
model to estimate the continuous oral daily intake (ng/kg-day) for the POD that could be used in
the derivation of a candidate RfD based on the study data. If all of this information was available,
then the result was included as a POD.

-------
                      NoncancerAnimal Bioassays Selected for
              TCDD Dose-Response Assessment (See Tables 2-4 and D-1)
                                    78 Studies
       Eliminate Studies with No lexicologically Relevant Endpoints for RfD Derivation
                         (See Appendix H and Section 4.2.1)
                               16Studies Eliminated
      Burlesonetal.(1996)                 DeVitoetal. (1994)
      Hassoun etal. (1998)                 Hassoun etal. (2000)
      Hassoun etal. (2002)                 Hassoun etal. (2003)
      Hong etal. (1989)                   Kitchin and Woods(1979)
      Latchoumycandaneetal. (2003)       Lucieretal. (1986)
      MallyandChipman (2002)            Sewall etal. (1993)
      Slezaketal. (2000)                  Sugita-Konishi etal. (2003)
      Tritscheretal. (1992)                Vanden Heuvel etal. (1994)
           Identify and Estimate PODs from the 62 Remain ing Animal Bioassays
                  for use in Noncancer Dose-Response Analysis of TCDD
                                   (See Figure 4-3)
                            Eliminate Studies with Botha
       LOAELHED>1 ng/kg-d and  NOAELHED/BMDLHED> 0.32 ng/kg-d*(See Table 4.3)
                               14 Studies Eliminated
      Ch u et al. (2001)                     Croutch et al. (2005)
      Fox etal. (1993)                     Ikedaet al. (2005)
      Maron pot etal. (1993)                Noharaetal. (2000,2002)
      Simanainen etal. (2002, 2003, 2004a)  Smialowicz et. al. (2004)
      Smith etal. (1976)                   Weberet al. (1995)

      *Hochstein etal. (2001) is also not carried forward because of the
      lack of toxicokinetic information for estimation of an HED
                           Derive Candidate RfDs from the
                     48 Remaining Noncancer Animal Bioassays
                 Final Candidate RfDs from Noncancer Animal Bioassays
                    (11 Studies Presented as Supporting Information;
                                   See Table 4-5)
                                 37 Candidate RfDs
          x	/

Figure 4-2. Disposition of noncancer animal bioassays selected for TCDD
dose-response analysis.
EPA evaluated each noncancer endpoint found in the 78 studies that passed the study inclusion
criteria.  From this evaluation, EPA eliminated 16 studies that contained no lexicologically
relevant endpoints for RfD derivation.  Then, as detailed in Figure 4-3, EPA selected and
identified PODs for use in deriving candidate RfDs. EPA then eliminated 13 studies based on
dose limits for the PODs' HEDs; one study was also not carried forward because of the lack of
toxicokinetic  information for estimation of an HED. Of the remaining 48 studies, EPA derived
37 RfD candidates, with 11 studies presented as supporting information.

                                       4-4

-------
            Study/endpointcombinationsfrom key noncanceranimal bioassayswith at
                  leastonetoxicologically relevantendpointforRfD derivation
                                       Is the
                             "endpoint under consideration""
                                    toxicologically
                                      relevant?

                                    Yes"
                                       No
            Determine NOAEL, LOAEL, and BMDL (if possible) human equivalentdose
            (HED)based on 1st-orderbody burden foreach study/endpointcombination
                                                                          No
     Is the
endpointobserved
 ear the LOAEL?
                                                    Is the BMDL less
                                                    than the LOAEL?
                                  sthe endpointless
                                  than the minimum
                                    LOAEL x 100?
                                     Yes
                 Determine a NOAEL, LOAEL, and BMDL (if possible) for each
              study/endpointcombination, based on blood concentrationsfrom the
                               Emond rodent PBPK model
                         Estimate a Human Equivalent Dose (HED)
              corresponding to each blood concentration NOAEL, LOAEL, or BMDL
                          usingthe Emond human PBPKmodel
                                Does kinetic modeling
                             suggest considering additional
                                ndpoints at higher doses?
                 Include NOAEL/LOAEL/BMDL
                         as a POD
                                    Exclude endpoint
                                        asa POD
Figure 4-3. EPA's process to identify and estimate PODs from key animal
bioassays for use in noncancer dose-response analysis of TCDD.
For the studies with at least one toxicologically relevant endpoint, EPA first determined if each
endpoint was toxicologically relevant.  If so, EPA determined the NOAEL, LOAEL, and BMDL
HED based on lst-order body burdens for each endpoint.  Within each study, these potential PODs
were included when the endpoint was observed near the LOAEL and if the BMDL was less than
the LOAEL.  Then, if the endpoint was less than the minimum LOAEL *  100 across all studies,
EPA calculated PODs based on blood concentrations from the Emond rodent PBPK model and,
for all of the PODs, HEDs were estimated using the Emond human PBPK model. If the kinetic
modeling results suggested considering additional endpoints at higher doses, the process was
repeated.  Finally, the lowest group of the toxicologically relevant PODs was selected for final use
in derivation of candidate RfDs.
                                         4-5

-------
further in Section 4.2.1).  EPA then identified PODs from the remaining bioassays; at this point,
Figure 4-2 refers to Figure 4-3, which is a flow chart of the iterative process used to estimate
PODs and compare them within and across the remaining studies to arrive at a final set of PODs
from these bioassays (see additional details below). From this final set of PODs, Figure 4-2
shows that EPA then eliminated 13 studies from further analysis because both of the following
conditions were met: HED LOAELnED (HED estimate based on LOAELs) >1 ng/kg-day and
NOAELHED or BMDLnED >0.32 ng/kg-day (see Table 4-3). One additional study was also not
carried forward because of the lack of toxicokinetic information for estimation of an FLED.
These dose limits were imposed to limit the size of the analysis yet ensure representation of all
important health effects associated with TCDD exposure. From the final list of 48 studies, EPA
derived 37 candidate RfDs, with 11 studies presented as supporting information.
       Figure 4-3 summarizes the strategy employed for identifying and estimating PODs from
the 62 animal bioassays with at least one lexicologically relevant endpoint for RfD derivation.
For the noncancer endpoints within these studies, EPA first evaluated the toxicological relevance
of each endpoint, rejecting those judged not to be relevant for RfD derivation. Next, initial
PODs based on the  first-order body burden metric (see Section 3.3.4.2) and expressed as FtEDs
(i.e., NOAELnED, LOAELnED, BMOLnEo) were determined for all relevant endpoints
(summarized in  Table 4-3). Because there were very few NOAELs and BMDL modeling was
largely unsuccessful due to data limitations (see Section 4.2), the next stage of evaluation was
carried out using LOAELs only. Within each study, effects not observed at the LOAEL (i.e.,
reported at higher doses) with BMDLnEDS greater than the LOAELnED were eliminated from
further analysis, as they would not be considered as candidates for the final POD on either a
BMDL or NOAEL/LOAEL basis (i.e., the POD would be higher than the PODs of other relevant
endpoints). In addition, all endpoints with LOAELnED estimates beyond a 100-fold range of the
lowest identified LOAELnED across all studies were (temporarily) eliminated from further
consideration, as they would not be POD candidates either (i.e., the POD would be higher than
the PODs of other relevant endpoints). For the remaining endpoints, EPA then determined final
potential PODs based on TCDD whole-blood concentrations obtained from the Emond rodent
PBPK models. FtEDs were then estimated for each of these PODs using the Emond human
PBPK model. At this point, if the PBPK modeling results suggested considering additional
endpoints at higher doses, the process was repeated.  From the final set of FtEDs, a POD was
                                         4-6

-------
selected31 for each study, to which appropriate UFs were applied following EPA guidance (see
Section 4.3.3 following). The resulting candidate RfDs were then considered in the final
selection process for the RfD.  Other endpoints occurring at slightly higher doses representing
additional effects associated with TCDD exposure (beyond the 100-fold LOAELHED range) were
                                                           ^9
evaluated, modeled, and included in the final candidate RfD array  to examine endpoints not
evaluated by studies with lower PODs.  In addition, BMD modeling based on administered dose
was performed on all endpoints for comparison purposes. The final array of selected endpoints
is shown in Table 4-4 (summary of BMD analysis) and Table 4-5 (candidate RfDs).
       The NAS recommended that EPA better justify the selection of response levels for
endpoints used to develop risk estimates.  The NAS commented on EPA's decision to estimate
an ED0i for noncancer bioassay/data set combinations as a comparative tool across studies,
suggesting that EPA identify and evaluate the levels of change associated with adverse effects to
define the BMR level for continuous noncancer endpoints.
       The committee notes that the choice of the 1% response level as the POD
       substantially affects ... the noncancer analyses.... The committee recommends
       that the Reassessment use levels of change that represent clinical adverse effects
       to define the BMR level for noncancer continuous end points as the basis for an
       appropriate POD in the assessment of noncancer effects (p. 72, NAS, 2006b).
       The committee concludes that EPA did not adequately justify the use of the
       1% response level (the EDoi) as the POD for analyzing epidemiological or animal
       bioassay data for ... noncancer effects (p.  18, NAS, 2006b).
       In the 2003 Reassessment (U.S. EPA, 2003), EPA was not attempting to derive an RfD
when it conducted TCDD dose-response modeling.  The 2003 Reassessment developed EDoi
estimates for noncancer effects in an attempt to compare disparate endpoints on a consistent
response scale.  Importantly, the 2003 Reassessment defined the EDoi as 1% of the maximal
response for a given endpoint, not as a 1% change from control. Because RfD derivation is the
primary goal of noncancer health effects assessment in this document, the noncancer modeling
effort undertaken here differs substantially from the modeling in the 2003 Reassessment.
31 In the standard order of consideration: BMDL, NOAEL, and LOAEL.
32 However, studies with a lowest dose tested greater than 30 ng/kg-day were not included in the expanded
evaluation.
                                          4-7

-------
       The NAS committee was concerned with the statistical power to determine the shape of
the dose-response curve at doses far below observed dose-response information. EPA agrees
that the shape of the dose-response curve in the low-dose region cannot be determined
confidently when based on higher-dose information.  An observed response above background
near (or below) the BMR level is needed for discrimination of the shape of the curve and for
accurate estimation of an EDX or BMDL.  Although many of the EDoiS presented in the 2003
Reassessment were near the lowest dose tested, responses at the lowest doses were often high
and much greater than a 1% response (i.e., 1% of the maximum response). The lack of an
observed response near the BMR level is often a problem in interpretation of BMD modeling
results.
       In this document, EPA has used a  10% BMR for dichotomous data for all endpoints;
there were no developmental studies that accounted for litter effects, for which a 5% BMR would
be used (U.S. EPA, 2000).  For continuous endpoints in this document, EPA has used a BMR of
1 standard deviation from the control mean whenever a specific toxicologically-relevant BMR
could not be defined.  For the vast majority of continuous endpoints, EPA could not establish
unambiguous levels of change representative of adversity, which EPA defines as "a biochemical
change, functional impairment, or pathologic lesion that affects the performance of the whole
organism, or reduces an organism's ability to respond to an additional environmental challenge"
(U.S. EPA, 2012).  For body and organ weight change, EPA has previously established a BMR
of 10% change, which also is used in this  document.
       The NAS commented on EPA's development of EDoi estimates for numerous study/data
set combinations in the 2003 Reassessment, suggesting that EPA had not  appropriately
characterized the statistical confidence around such model predictions in the low-response region
of the model.
       It is critical that the model used for determining a POD fits the data well,
       especially at the lower end of the observed responses. Whenever feasible,
       mechanistic and statistical information should be used to estimate the shape of the
       dose-response curve at lower doses.  At a minimum, EPA should use rigorous
       statistical methods to assess model fit and to control and reduce the uncertainty of
       the POD caused by a poorly fitted model. The overall quality of the study design
       is also a critical element in deciding which data sets to use for quantitative
       modeling (NAS. 2006b. p. 18).
                                          4-8

-------
       EPA should ... assess goodness-of-fit of dose-response models for data sets and
       provide both upper and lower bounds on central estimates for all statistical
       estimates. When quantitation is not possible, EPA should clearly state it and
       explain what would be required to achieve quantitation (NAS, 2006b, p. 10).
       The NAS also commented that EPA report information describing the adequacy of
dose-response model fits, particularly in the low response region.  For those cases where
biostatistical modeling was not possible, NAS recommended that EPA identify the reasons.
       The Reassessment should also explicitly address the importance of statistical
       assessment of model fit at the lower end and the difficulties in such assessments,
       particularly when using summary data from the literature instead of the raw data,
       although estimates of the impacts of different choices of models would provide
       valuable information about the role of this uncertainty in driving the risk estimates
       (NAS. 2006b. p. 73).
       To address this concern, in this document EPA has reported the standard suite of
goodness-of-fit measures from the benchmark dose modeling software (BMDS 2.1).  These
include chi-square/?-values, Akaike's Information Criterion (AIC), scaled residuals at each dose
level, and plots of the fitted models. For the multistage model, when restricted lower-order
coefficients hit the lower bound (zero), EPA used likelihood ratio tests to evaluate whether the
improvement in fit afforded by estimating successively higher-order coefficients could be
justified. Goodness-of-fit measures are reported for all key data sets in Appendix G.
(Section 4.2.4.2 discusses the BMD modeling criteria for model evaluation.)

4.2.   NONCANCER DOSE-RESPONSE ASSESSMENT OF TCDD
       This section describes EPA's evaluation of TCDD dose response for noncancer endpoints
from studies that met the study inclusion criteria.  Discussions include BMD modeling
procedures, kinetic modeling, and development of PODs for derivation of the RfD. Section 4.2.1
discusses the types of endpoints that are considered relevant by EPA for derivation of toxicity
values (U.S. EPA. 2005a. b, 1998. 1996. 1994. 1991) and lists the study/endpoint combinations
that were not considered for the TCDD RfD derivation, with supporting text in Appendix H.
Section 4.2.2 describes how EPA has used PBPK modeling to estimate effective internal
exposures as an alternative to using  administered  doses or body burdens based on first-order
                                          4-9

-------
kinetics. Section 4.2.3 details the dose-response analysis of the epidemiologic data, with
supporting information on kinetic modeling in Appendix F. Section 4.2.4 details the
dose-response analysis for the animal bioassay data, with supporting information on kinetic
modeling in Appendix E; Appendix G provides the BMDS input tables (see Section G. 1) and
output for all modeling, including blood concentrations (see Section G.2) and administered dose
(see Section G.3).

 4.2.1.  Determination of lexicologically Relevant Endpoints
       The NAS committee commented on the low-dose model predictions and the need to
discuss the biological significance of the noncancer health effects modeled in the 2003
Reassessment.  In selecting POD candidates from the animal bioassays for derivation of the
candidate RfDs, EPA considered the toxicological relevance of the identified endpoint(s) from
any given study. Some endpoints/effects may be sensitive, but lack general toxicological
significance because of lack of inherent adversity,33 being an adaptive response, or not being
clearly linked to downstream functional or pathological alterations. Endpoints not considered to
be lexicologically relevant for TCDD include CYP induction, oxidative stress measures, mRNA
induction, protein phosphorylation, certain immune system responses, gap junction disruption,
and epidermal growth factor signaling.  As an example, CYP induction alone is not considered a
significant toxicological effect given that CYPs are induced as part of the normal hepatic
metabolism of xenobiotic agents.  Additionally, the role of CYP induction in the noncancer
toxicity of TCDD is unknown, thus, due to the lack of obvious pathological significance,
TCDD-induced CYP induction is not considered a relevant endpoint for RfD derivation.
Another example is oxidative stress. As an example, TCDD has been shown to induce changes
in oxidative stress markers, but no other indicators  of brain pathology were assessed (Hassoun et
al., 2003; 2000;  1998). In this case, it is impracticable to link the markers of oxidative stress to a
toxicological outcome in the brain; thus, this endpoint is not considered relevant for RfD
derivation. Studies otherwise meeting the study inclusion criteria, but with no toxicologically-
relevant endpoints that were considered suitable for derivation of a candidate RfD are listed in
Figure 4-2, and described and discussed in Appendix H.
33 An adverse effect is defined in EPA's Integrated Risk Information System glossary as "a biochemical change,
functional impairment, or pathologic lesion that affects the performance of the whole organism, or reduces an
organism's ability to respond to an additional environmental challenge" (U.S. EPA. 2012).
                                           4-10

-------
 4.2.2.  Use of Toxicokinetic Modeling for TCDD Dose-Response Assessment
       Because relevant toxicokinetic models for TCDD disposition in rodents and humans are
available, EPA has not applied the standard uncertainty factor approach in the derivation of the
TCDD RfD. In addition, because of the much slower elimination of TCDD in rodents than in
humans, EPA has determined that the standard uncertainty factor approach can underestimate the
interspecies toxicokinetic extrapolation factor by an order of magnitude or more (U.S. EPA,
2003). The toxicokinetic models chosen by EPA are the rodent and human PBPK models
described by Emond et al. (2006; 2005; 2004)34 and modified by EPA for this assessment as
described in Section 3.3.4 (hereafter referred to as the "Emond [rodent or human] PBPK
model"). Both the rodent and human models have a gestational component, which allow for
more relevant exposure comparisons between general adult exposures and the numerous
gestational  exposure studies.  Ideally, a relevant tissue concentration for each effect would be
estimated.  However, at present, no models exist for estimation of all relevant tissue
concentrations. As virtually all TCDD is found in the adipose fraction of tissues, or bound to
specific proteins, a preferred approach to developing a dose metric would be to account for the
fat fraction of each tissue and protein binding; however, EPA has decided that the modeling of
such estimates is too uncertain, and EPA has not found sufficient data to implement this
approach.  Therefore, EPA has decided to use the concentration of TCDD in whole blood as a
surrogate for tissue concentrations, assuming that tissue concentrations are proportional to
whole-blood concentrations. Furthermore, because the RfD is necessarily expressed in terms of
average daily exposure, the blood concentrations are expressed as averages over the relevant
period of exposure for each endpoint.  For the animal bioassays, the relevant period of exposure
is the duration  of dosing, starting at the age of the animals at the beginning of the study. For
humans, the relevant period of exposure is generally a lifetime, which is defined as 70 years.
However, EPA varied the averaging time for the equivalent human blood concentrations to
correspond to the test-animal exposure duration in the following manner.
       For correspondence with animal chronic exposures,35 the human-equivalent
       TCDD blood concentration is assumed to be the 70-year average.
34 The Emond PBPK models are three-compartment dynamic models.
35 Assumed to be >75% of nominal lifetime, or about 550 days in rodents.
                                          4-11

-------
       For correspondence with animal gestational exposures, the human-equivalent
       TCDD blood concentration is assumed to be the average over 45 years for a
       female, beginning at birth, plus 9 months of gestational exposure.36  Forty five
       years of age is considered here as an upper limit on the age at which a typical
       human female can conceive and bear a child.

       For correspondence with any other animal exposure duration, the
       human-equivalent TCDD blood concentration is assumed to be the average over
       the equivalent human exposure duration calculated backward from the peak
       exposure plateau at or near the end of the 70-year scenario. The average is
       determined from the terminal  end of the human exposure period to be protective
       of less-than-lifetime exposures occurring at any time in a lifetime; the daily oral
       intake achieving the target blood concentration is smaller than for the same
       exposure period beginning at birth. The determination of equivalent exposure
       durations across species is problematic and somewhat arbitrary, so EPA uses the
       average peak blood concentration as the human equivalent for all
       less-than-chronic animal exposures (other than gestational).37  For the first-order
       kinetics model, the average peak exposure is close to the theoretical  steady-state
       asymptote (see Section 3.3.4.2). However, for the Emond human PBPK model
       used by EPA in this assessment, the timing of the peak exposure is
       dose-dependent and tends to decline after 60 years in some cases. Therefore, the
       5-year average TCDD blood concentration that includes the peak ("5-year peak")
       is used as the relevant dose-metric for the PBPK model applications (see Section
       3.3.6 and Figure 3-33).
 4.2.3.  Noncancer Dose-Response Assessment of Epidemiologic Data
       The following four epidemiologic studies describing noncancer endpoints were identified

in Section 2.4.1 as studies to be evaluated for development of PODs for derivation of candidate

RfDs: Baccarelli et al. (20081 Mocarelli et al. (20081 Alaluusua et al. (2004). and Eskenazi et al.

(2002b). Each of these studies described effects observed in the Seveso cohort (see detailed

study summaries in Appendix C and Table 2-2). Each study reported individual-level human

exposure measures and provided information from which EPA could determine a "critical

exposure window" (see Text Box 2-2) of susceptibility over which the effective TCDD

exposures could be quantified for dose-response assessment. For studies that reported grouped

data by TCDD exposure ranges, the representative values for the ranges were determined by
36 See Section 3.3.6 for a discussion of this issue, including a comparison of the 45-year old pregnancy scenario to
one beginning at age 25 in Table 3-24.
37 By comparison to a half-lifetime equivalent (1 year in rodents, 35 years in humans), in the lst-order kinetic model
the ratio of body burden to oral intake does not differ significantly from the average-peak scenario; all shorter-term
scenarios differ even less (see Section 3.3.4.2). These relationships, with respect to the 5-year peak, hold for the
PBPK model results, as well (see Section 3).
                                           4-12

-------
taking the geometric mean of the range limits, assuming that the TCDD concentration
distribution in the population is more likely to be skewed (e.g., lognormal) than symmetrical
(e.g., normal or uniform). A sufficient number of significant digits are carried through
intermediate results to avoid round-off error in the final value.  EPA used toxicokinetic modeling
(Emond human PBPK model) to estimate daily TCDD intake rates for the exposure groups
presented in these studies (see Appendix F for details). The exposure scenario in all of these
studies, except Baccarelli et al. (2008), entailed an initial high pulse exposure at the time of the
plant explosion followed by low-level background exposure over a period of several years across
the critical exposure window, resulting in internal exposure profiles characterized by a 5 to
10-fold difference in initial and final TCDD serum concentrations (as LASC).  For these
scenarios, EPA modeled both the peak TCDD LASC and the average LASC over the critical
window,  then estimated daily average continuous TCDD intakes over the critical-window
duration corresponding to each of the peak and critical-window average serum concentrations.
Estimation of LASC and intakes was accomplished using the Emond human PBPK model.  EPA
considered the critical-window average exposures to be important, although they were much
lower than the peak exposures, because the relatively slow elimination of TCDD engenders
concerns for an ongoing contribution of residual TCDD body burdens to the adverse health
outcomes during the period  of susceptibility. However, the overall average exposure does not
reflect the influence of the much higher peak exposure, which may be a significant factor in
TCDD toxicity (Kim et al., 2003).38 That is, EPA is uncertain as to whether the health outcomes,
often observed many years beyond the period of susceptibility, are a result of permanent damage
from the  initial high exposure or more gradual impairment from longer-term ongoing exposure.
For these reasons, EPA derived the PODs for RfD consideration by averaging the TCDD intakes
for the peak exposure and critical-window exposure average, essentially treating each as equally
likely. EPA focused on identifying NOAELs and LOAELs for these studies. EPA did not
conduct BMD modeling because the covariates identified by the study authors could not be
incorporated by modeling the grouped response data.  EPA's development of PODs for these
studies is described in this section, with kinetic modeling details provided in Appendix F; the
results are shown in Table 4-1.
38 Kim et al. (2003) found a significantly higher fraction of altered hepatic foci in rats treated with a single high
TCDD dose than those administered a continuous dose over 15 weeks, yielding similar terminal liver TCDD
concentrations.
                                          4-13

-------
       Table 4-1. PODs for epidemiologic studies of TCDD
Study
Alaluusuaetal. (2004)
Baccarelli et al. (2008)
Mocarelli et al. (2008)
POD (ng/kg-day)
0.0406a (NOAEL)
0.020b (LOAEL)
0.020C (LOAEL)
Critical effects
Dental effects in adults exposed to TCDD in childhood
Elevated TSH in neonates
Decreased sperm count and motility in men exposed to
TCDD in childhood
aMean of peak exposure (0.0655 ng/kg-day) and average exposure over 10-year critical window (0.0156 ng/kg-day).
bMaternal exposure corresponding to neonatal TSH concentration exceeding 5 uU/mL.
°Mean of peak exposure (0.032 ng/kg-day) and average exposure over 10-year critical window (0.0080 ng/kg-day).
4.2.3.1. Baccarelli et al. (2008)
       For Baccarelli et al. (2008), EPA was able to define a LOAEL in terms of the maternal
TCDD serum levels corresponding to neonatal TSH level above 5 |i -Units TSH per mL of serum
(5 jiU/mL) (see Appendix C, Section C.I. 2. 1.5. 7, and Table 2-2 for study details). The adversity
benchmark of 5 |iU/mL is based on the WHO (1994) indicator for follow up examination for
potential hypothyroidism (see following discussion in Section 4.3.4.1). Baccarelli et al. (2008)
performed regression modeling of neonatal TSH against maternal TCDD LASC but did not
estimate the equivalent oral intake. The regression model related the level of TSH in 3-day-old
neonates to TCDD concentrations in maternal plasma at birth (given as LASC). The authors
extrapolated maternal plasma concentrations from previous measurements using a simple
first-order pharmacokinetic model. The study authors also reported group average neonatal
TCDD serum levels for infants above and below the 5 jiU/mL limit. However, because there is
limited information regarding the relationship between maternal and neonatal  serum TCDD
levels, EPA determined that there was too much uncertainty in estimating maternal intake from
neonatal TCDD serum concentrations directly.  Therefore, EPA determined the maternal intake
at the LOAEL from the maternal serum-TCDD/TSH  regression model by finding the maternal
TCDD LASC at which neonatal TSH exceeded 5 |iU/mL.  EPA then used the Emond PBPK
model under the human gestational scenario (see Section 4.2.2) to estimate the continuous daily
oral TCDD intake that would result in a TCDD LASC corresponding to a neonatal TSH of
5 |iU/mL at the end of gestation; EPA established the resulting maternal intake
(0.020 ng/kg-day) as the LOAEL, shown in Table 4-1 as a POD for derivation of candidate RfDs
(PBPK modeling details are shown in Appendix F).
                                         4-14

-------
4.2.3.2. Mocarelli et al (2008)
       Mocarelli et al. (2008) reported decreased sperm concentrations (21-33%) and decreased
motile sperm counts (12-25%) in men who were 1-9 years old in 1976 at the time of the
accident (initial TCDD exposure event) (see Appendix C, Section C.I.2.1.5.8, and Table 2-2 for
study details).  Men who were 10-17 years old in 1976 were not adversely affected. Serum
(LASC) TCDD levels were measured within 1 year of the initial exposure.  Serum TCDD levels
and corresponding responses were reported by quartile, with a reference group of less-exposed
individuals assigned a TCDD LASC value of 15 ppt (which was the mean of individuals outside
the contaminated area). The lowest exposed group median was 68 ppt (1st quartile). Because
sperm effects were detected only among boys under the age of 10, EPA assumes there  is a
maximum 10-year critical exposure window for elicitation of these  effects.39  However, for the
exposure profile, with a high initial pulse followed by an extended period of elimination with
only background exposure, the estimation of an average exposure resulting in the effect is
somewhat complicated. EPA implemented a procedure for the estimation of the continuous daily
TCDD intake associated with the LOAEL in the Mocarelli et al. (2008) study using the following
5-step process:
    1.  Using the Emond human PBPK model, the initial (peak) serum TCDD concentrations
       (LASC) associated with the accident were back-calculated based on the time that had
       elapsed between the explosion and the serum collection. As serum measurements were
       taken within 1 year after the event, a lag time to measurement of 0.5 years was assumed.
       The group average peak serum concentration for the 1st quartile was estimated to be
       249 ppt.
    2.  The oral exposure associated with the peak serum TCDD concentration (peak exposure)
       was calculated using the Emond PBPK model.
    3.  Starting with the peak exposure and accounting for background TCDD intake, the
       average daily serum TCDD concentration experienced by a representative individual in
       the susceptible lifestage (boys under 10 years old) was estimated using the Emond PBPK
       model. The average  subject age at the time of the event was 6.2 years. Consequently, a
       critical exposure window for the cohort was estimated to be, on average, 3.8 years (i.e., a
       boy aged 6.2 years would remain in this exposure window for 3.8 more years until he was
       10 years of age).  The critical window average serum concentration for the 1st quartile
       group was estimated to be 57.7 ppt (45 ppt at 10 years).
39 Neither the study authors nor EPA assume 10 years to be the age of puberty onset; 10 years is the age that the
study authors used to divide their study population by magnitude of effect.
                                          4-15

-------
   4.  Using the Emond PBPK model, the average daily TCDD intake rate needed to attain the
       3.8-year average serum TCDD concentration in a boy 10 years old was calculated.
   5.  The LOAEL POD was calculated as the average of the peak exposure intake
       (0.032 ng/kg-day) and the 3.8-year average exposure intake (0.0080 ng/kg-day), resulting
       in LOAEL of 0.020 ng/kg-day, shown in Table 4-1 as a POD for derivation of a
       candidate RfD.
       The PBPK modeling details are shown in Appendix F.

4.2.3.3. Alaluusua et al (2004)
       For Alaluusua et al. (2004), the approach for estimation of daily oral TCDD intake is
virtually identical to the approach used for the Mocarelli et al. (2008) data, (see Appendix C,
Section C.I.2.1.5.5, and Table 2-2 for study details.)  Alaluusua et al. (2004) reported dental
effects in male and female adults who were less than 5 years of age at the time of the initial
exposure (1976). For the 75 boys and girls who were less than 5 years old at the time of the
accident, 25 (33%) were subsequently diagnosed with some form of dental enamel defect. For
the 38 individuals who were older than 5, only 2 (5.3%) suffered dental enamel defects at a later
date. In addition, the incidence of missing permanent teeth (lateral incisors and  second
premolars) was 3 times as prevalent in zone ABR subjects compared with zone non-ABR
residents.  A window of susceptibility of approximately 5 years is assumed. Serum
measurements for this cohort were taken within a year of the accident. Serum TCDD levels and
corresponding responses were reported by tertile, with a reference group of less-exposed
individuals assigned a TCDD LASC value of 15  ppt (ng/kg); the tertile group geometric means
were 72.1, 365.4, and 4,266 ppt. The incidence of dental effects for the reference group was
26% (10/39).  The incidence of dental effects in the 1st, 2n , and 3r tertile exposure groups was
10% (1/10), 45% (5/11), and 60% (9/15), respectively.  EPA judged that the NOAEL and
LOAEL were 72.1 and 365.4 ppt TCDD in serum (LASC), in the 1st tertile and 2nd tertile,
respectively. Following the same procedure used for the Mocarelli et al. (2008) study (see
Section 4.2.3.2), EPA estimated the continuous daily human oral TCDD intake associated with
each of the tertiles for both peak and average exposure across the critical exposure window,
assuming that the average age  of the susceptible cohort at the time of the accident was 2.5 years.
Separate estimates for boys and girls were developed based on both the peak intake and average
intake across the critical exposure window (PBPK modeling details are shown in Appendix F).
                                          4-16

-------
The estimated averaged daily oral intakes for the tertiles, averaged for boys and girls, are 0.0655,
1.65, and 111 ng/kg-day for the peak exposure and 0.0156, 0.149, and 4.81 ng/kg-day for the
critical exposure window average.  The LOAEL for this study was determined to be
0.897 ng/kg-day, which is the average of the peak  exposure and window average exposure for
the second tertile. A study NOAEL of 0.0406 ng/kg-day for the first tertile was determined
similarly and serves as a POD for derivation of a candidate RfD in Table 4-1.

4.2.3.4. Eskenazi et al (2002b)
       The approach used to estimate daily TCDD intake in Eskenazi et al. (2002b) combines
the approaches EPA used for Baccarelli et al. (2008), Mocarelli et al. (2008), and Alaluusua et al.
(2004).  Eskenazi et al. (2002b) reported menstrual effects in female adults who were
premenarcheal in 1976 at the time of the initial exposure (see Appendix C, Section C.I.2.1.4.1
and Table 2-2 for study details).  In Rigon et al. (2010), the median age at menarche was shown
to be 12.4 in Italian females with intergenerational decreases in age at menarche. Thus, EPA
established a window of susceptibility of approximately 13 years for this analysis.  The average
age of the premenarcheal girls at the time of the initial exposure in 1976 was 6.8 years,
establishing an average critical-window exposure duration of 6.2 years for this cohort. Serum
samples were collected within a  year of the accident from this cohort. However, serum TCDD
levels and corresponding responses were not reported by percentile, and no internal reference
group was identified. As for Baccarelli et al. (2008), Eskenazi et al. (2002b) developed a
regression model relating menstrual cycle length to plasma TCDD concentrations (LASC)
measured in 1976.  The model estimated that menstrual cycle length was increased 0.93 days for
each 10-fold increase in TCDD LASC, with a 95% confidence interval of-0.01  to 1.86 days.
The determination of a LOAEL is somewhat arbitrary, with no independent measure of an
adversity threshold to establish the toxicological significance of a given increase in menstrual
cycle length.  The study authors  did not present data for unexposed premenarcheal girls (in
1976), so an appropriate reference population is not available.  EPA did not conduct BMD
modeling because of the lack of  a reference population and the inability to include the covariates
considered by the study authors in their analysis.  However, an approximate LOAEL can be
estimated from Figure 1 in Eskenazi et al. (2002b), noting that both the length of the menstrual
cycle and its variance increases above TCDD concentrations of about 1,000 ppt.  The highest

                                          4-17

-------
measured concentration is 16,500 ppt.  Consistent with the previously established method for
determining representative values for age limits (see Sections 4.2.3.2 and 4.2.3.3), the geometric
mean of 4,060 ppt for this range is assigned as a LOAEL. The lower range of TCDD
concentrations is too large to treat as a  single group for estimating a NOAEL, but using the study
authors' regression model, a TCDD LASC of approximately 50 ppt corresponds to a menstrual
cycle length of 28  days, generally considered to be the average normal length.  These two (1976)
serum levels were  then modeled by EPA using the Emond human PBPK model in the same
manner as for Mocarelli et al. (2008) and Alaluusua et al. (2004), but with a 6.2-year exposure
window for the premenarcheal girls.  The resulting peak and window-average TCDD intakes for
the 50 ppt exposure are 0.0168 and 0.00364 ng/kg-day, respectively; the average of the two
intakes is 0.0102 ng/kg-day. The peak and window-average TCDD intakes for the LOAEL
exposure (4,060 ppt) are 60.0 and 1.52 ng/kg-day, respectively; the average of the two intakes of
30.8 ng/kg-day defines the LOAEL POD. Further details of the PBPK modeling can be found in
Appendix F. Although 0.0102 ng/kg-day could be considered to be a NOAEL, there is too much
uncertainty in the upper end of the NOAEL range, given the very large (3,000-fold) difference
between it and the  LOAEL, for using it as a NOAEL POD.  The LOAEL of 30.8 ng/kg-day, also
uncertain in magnitude and toxicological significance, is 1,540-fold higher than the LOAEL
PODs for Mocarelli et al. (2008) and Baccarelli et al.  (2008), and will not be a factor in the
derivation of the RfD. Therefore, the LOAEL for this study is  not considered further in this
assessment except in the context of the RfD uncertainty analysis presented in Section 4.5.

 4.2.4. Noncancer Dose-Response Assessment of Animal Bioassay Data
       EPA followed the strategy illustrated in Figures 4-2 and 4-3 to evaluate the animal
bioassay data for TCDD dose response. For the administered average daily doses  (ng/kg-day) in
each animal bioassay, EPA identified NOAELs and/or LOAELs based on the original data
presented by the study author. Section 2.4.2 identifies these values in Table 2-4 and in the study
summaries found in Appendix D. These became PODs for consideration in the derivation of an
RfD for TCDD.  The candidate RfD values associated with these PODs  are presented in
Table 4-5. All PODs were converted to FtEDs using the Emond PBPK models, with
whole-blood TCDD concentration as the effective dose metric. The remainder of this section
                                         4-18

-------
describes the steps in this process and concludes with the PODs from the animal bioassay data
that were considered for derivation of the RfD.

4.2.4.1. Use of Kinetic Modeling for Animal Bioassay Data
       Whole-blood TCDD concentrations corresponding to the administered doses in each
mouse or rat bioassay qualifying as a final RfD POD were estimated using the appropriate
Emond rodent PBPK model. In each case, the simulation was performed using the exposure
durations, body weights, and average daily doses from the original studies. For all
multiple-exposure protocols, the time-weighted average blood TCDD concentrations over the
exposure period were used as the relevant dose metric. For single (gestational and
nongestational) exposures, the initial peak blood TCDD concentrations were considered to be the
most relevant exposure metric.  Gestational exposures were modeled using the species-specific
gestational component of the Emond rodent PBPK model. Bioassays employing exposure
protocols spanning gestational and postpartum life stages were modeled by sequential
application of the gestational and nongestational models.
       The Emond PBPK models do not contain a lactation  component, so exposure during
lactation was not modeled explicitly.  Only one bioassay (Shi et al., 2007) considered as a POD
for RfD derivation included exposure during lactation. In Shi et al. (2007), pregnant animals
were exposed weekly to TCDD  throughout gestation and lactation. Exposure was continued in
the offspring following weaning for 10 months.  For assessment of maternal effects, the Emond
gestational model was used, terminating at parturition. For assessment of long-term exposure in
the offspring, the Emond nongestational model was used, ignoring prior gestational and
lactational exposure, with the assumption that the total exposure during these periods was small
relative to exposure in the following 10 months.  The  assumption is conservative in that effects
observed in the offspring would be attributed entirely to adult exposure, which is somewhat less
than the actual  total exposure.
       The model code, input files, and PBPK modeling results for each bioassay are reported in
Appendix E. The modeled TCDD blood concentrations were used for BMD modeling of
bioassay response data and determination of NOAELs and LOAELs.  BMD modeling was
performed, as described in Section 3.3.6, by substituting the  modeled blood concentrations for
the administered doses and calculating the corresponding BMDL.  For each of these LOAEL,

                                         4-19

-------
NOAEL, or BMDL blood-concentration equivalents, corresponding HEDs were estimated using
the Emond human PBPK model for the appropriate gestational or nongestational scenario as
described previously (see Section 4.2.2).

4.2.4.2. Benchmark Dose Modeling of the Animal Bioassay Data
       BMD modeling was performed for each study/endpoint combination using BMDS 2.1 to
determine BMDs and BMDLs. The input data tables for these noncancer studies are shown in
Appendix G,  Section G.I, including both administered doses (ng/kg-day) and blood
concentrations (ng/kg [ppt])40 and either incidence data for the dichotomous endpoints or mean
and standard deviations for the continuous endpoints (see Section 4.2.4.1 and Sections 3.3.4 and
3.3.5 for a description of the development of TCDD blood concentrations using kinetic
modeling).
       Evaluation of BMD modeling performance, goodness-of-fit, dose-response data, and
resulting BMD and BMDL estimates included statistical criteria as well as professional judgment
of their statistical and toxicological properties. For the continuous endpoints, all available
models were run separately using both the assumption of constant variance and the assumption
of modeled variance. Saturated (0 degrees of freedom) model fits were rejected from
consideration. Parameters in models with power or slope parameters were constrained to prevent
supralinear fits, which EPA considers not to be biologically plausible and which often have
undesirable statistical properties (i.e., the BMDL converges on zero).  Table 4-2 shows each
model  and any restrictions imposed.
40 Units of ng/kg will be used exclusively for oral intakes in this section.  Blood and tissue concentrations will be
expressed in ppt units.
                                          4-20

-------
       Table 4-2. Models run for each study/endpoint combination in the animal
       bioassay BMD modeling
Model
Restrictions imposed
Continuous models
exponential M2-M5, not
grouped
Hill
Linear
'olynomial
'ower
Adverse direction specified according to the response data; power >1
Adverse direction is automatic; n > 1
Adverse direction is automatic; degree of polynomial = 1
Adverse direction is automatic; degree of polynomial unrestricted; restrict the
sign of the power to nonnegative or nonpositive, depending on the direction of
the responses
Adverse direction is automatic; power >1
Dichotomous models
Gamma
Logistic
^og-Logistic
Log-Probit
Multistage
Probit
Weibull
Power > 1
None
Slope >1
None
Beta >0, 2n degree polynomial
None
Power > 1
       For the quantal/dichotomous endpoints, all primary BMDS dichotomous models were
run.  The alternative dichotomous models were fit to several data sets, but the results were very
sensitive to the assumed independent background response and the fits were not accepted. The
confidence level was set to 95%, and all initial parameter values were set to their defaults in
BMDS. For the continuous endpoints, 1 standard deviation was chosen as the default for the
BMR when a specific toxicologically-relevant BMR could not be defined. For the dichotomous
endpoints, a BMR of 10% extra risk was used for all endpoints.41
       The model output tables in Appendix G show all of the models that were run, both
restricted and unrestricted, goodness-of-fit statistics, BMD and BMDL estimates, and whether
bounds were hit for constrained parameters. After all models were run, the one giving the best
fit was selected using the selection criteria in the draft BMD Technical Guidance (U.S. EPA,
2000). Acceptable model fits were those with chi-square goodness-of-fit/^-values greater than
0.1.  For continuous endpoints, the preference was for models with an asymptote term (plateau
for high-dose response) because continuous measures do not continue to rise (or fall) with dose
forever; this phenomenon is particularly evident for TCDD. Unbounded models, such as the
41 There were no developmental studies that accounted for litter effects, for which a 5% BMR would be used.
                                         4-21

-------
power model, must account for the plateauing effect entirely in the shape parameter, generally
resulting in a supralinear fit. Also, for the continuous endpoints, the/>-value for the homogenous
variance test (Test 2) was used to determine whether constant variance (p > 0.1) or modeled
variance (p < 0.1) should be used.  As BMDS offers only one variance model, model fits for
modeled variance models were not necessarily rejected if the variance model  did not fit well
(Test 3 p-va\ue < 0.05). Within the group of models with acceptable fits, the  selected model was
generally the one with the lowest AIC. If the AICs were similar, the model with the lowest
BMDL was selected.  However, particularly for continuous models,  the fit of the model to the
control-group response and in the lower response range was assessed.  Models with higher
BMDLs or AICs but much better fit to the lower response data were often chosen over the
nominally best-fitting model.
      For many data sets, no models satisfied the acceptance criteria, and no clear
BMD/BMDL selection could be made. In these cases, model fits were examined on an
individual basis to determine the reasons for the poor fits.  On occasion, high  doses were
dropped, and the models were refit. Also, if a poor fit to the control  mean was evident, the
model was refit to the data after fixing the control mean by specifying the relevant parameter in
BMDS.  However, these techniques rarely resulted in better fits.  If the fit was still not
acceptable, the NOAEL/LOAEL approach was applied to the study/data set combination. Most
of the problems with BMD modeling were a consequence of lack of response data near the
BMR; many of the TCDD data sets failed to show a response near the BMR, whether it was a
10% dichotomous relative change  or a continuous 1  standard deviation change.  Responses at the
lowest doses were generally much higher than the BMR, resulting in a lack of "anchoring" at the
critical response levels of interest,  resulting in insufficient information for precise numerical
estimation of BMDLs.

4.2.4.3.  Points of Departure (PODs)from Animal Bioassays Based on Human Equivalent
        Dose (HED) and Benchmark Dose (BMD) Modeling Results
      Table 4-3 summarizes the PODs that EPA estimated for each key animal study included
for TCDD noncancer dose-response modeling that also contained lexicologically relevant
endpoints (see Section 4.2.1 and Appendix H for excluded studies).  After estimating the blood
TCDD concentration associated with a particular toxicity measure (NOAEL, LOAEL, or
BMDL) obtained from a rodent  bioassay, EPA estimated a corresponding HED using the Emond
                                         4-22

-------
human PBPK model (described in Section 3). Table 4-3 summarizes the NOAEL, LOAEL, or
BMDL based on the administered animal doses for each key bioassay/data set combination.
Table 4-3 also summarizes the continuous daily HED corresponding to these administered doses
as 1st order body burdens and as whole-blood concentrations.  The doses in Table 4-3 are defined
as follows, all in units of ng/kg-day:
      Administered Dose NOAEL: Average daily dose defining the NOAEL for the test species
      in the animal bioassay
      Administered Dose LOAEL: Average daily dose defining the LOAEL for the test species
      in the animal bioassay
      Administered Dose BMDL: BMDL for the test species based on modeling of the
      administered doses from the animal bioassay
      First-Order Body Burden HED NOAEL: Average daily dose defining the NOAEL for
      humans derived from the animal bioassay using the first-order kinetics body-burden
      model
      First-Order Body Burden FLED LOAEL: Average daily dose defining the LOAEL for
      humans derived from the animal bioassay using the first-order kinetics body-burden
      model
      First-Order Body Burden HED BMDL: Human-equivalent BMDL from BMD modeling
      of the animal bioassay data using first-order body burdens
      Blood Concentration HED NOAEL: Average daily dose defining the NOAEL for
      humans derived from the animal bioassay using the Emond human PBPK model
      Blood Concentration HED LOAEL: Average daily dose defining the LOAEL for humans
      derived from the animal bioassay using the Emond human PBPK model
      Blood Concentration HED BMDL: Human-equivalent BMDL from BMD modeling of
      the animal bioassay data using the Emond human PBPK model
      An evaluation of key BMD analyses is presented in Table 4-4. Tables showing the best
model fit for each study/endpoint combination and the associated BMD/BMDL are shown in
Appendix G. As described in Section 4.2.4.2, the BMD modeling was largely unsuccessful,
primarily because of a lack of response data near the BMR, poor modeled representation of
control values, or nonmonotonic responses yielding poor fits.  The comments column in
Table 4-4 lists reasons for poor results.
                                        4-23

-------
          Table 4-3.  Summary of key animal study PODs (ng/kg-day) based on three different dose metrics: administered
          dose, first-order body burden HED, and blood concentration
Study
Amin et al. (2000)
Bell et al. QOOTb)
Bowman et
al.(1989a; 1989b);
Schantz and
Bowman (1989):
Schantz et al. (1986):
Schantz etal. (1992)
Cantonietal. (1981)
Chu etal. (2001)
Chu et al. (2007)
Crofton et al. (2005)
Croutch et al. (2005)
DeCaprio et al.
(1986)
Fattore et al. (2000)
Fox etal. (1993)
Franc et al. (2001)
Franczak et al.
(2006)
Hoio etal. (2002)'
Hochstein et al.
(200 l)g
Endpoint
Saccharin preference ratio,
female
Balano-preputial separation in
male pups
Neurobehavioral effects
Urinary coproporhyrins
Tissue-weight changes
Liver lesions
Serum T4
Decreased body weight
Decreased body weight, organ-
weight changes
Decreased hepatic retinol
Increased liver weight
Organ-weight changes
Abnormal estrous cycle
DRL response per minute
Kit mortality at 6 weeks
Administered dose"
NOAEL
—


-
2.50E+02
2.50E+00
3.00E+01
5.43E+01
6.10E-01
-
5.70E-01
l.OOE+01

-
—
LOAEL
2.50E+01
2.40E+00
1.20E-01
1.43E+00
l.OOE+03
2.50E+01
l.OOE+02
2.17E+02
4.90E+00
2.00E+01
3.27E+02
3.00E+01
7.14E+00
2.00E+01
2.65E+00
BMDLd
_e
2.87E+00

	 e
-
-
	 e
-
—
-
-
1.34E+01

_e
—
lst-order body burden HEDb
NOAEL
—


-
7.55E-01
7.55E-03
1.92E-02
2.22E-01
4.11E-03
-
1.42E-03
6.62E-02

-
—
LOAEL
2.49E-02
1.26E-02
8.22E-03
1.24E-02
3.02E+00
7.55E-02
6.40E-02
8.89E-01
3.30E-02
1.23E-01
8.12E-01
1.99E-01
5.95E-02
5.26E-03
—
BMDLd
_e
1.50E-02

	 e
-
-
	 e
-
—
-
-
8.87E-02

_e
—
Blood concentration HEDC
NOAEL
—


-
7.03E+00
3.49E-02
1.69E-01
7.81E-01
—
-
8.08E-04
4.49E-01

-
—
LOAEL
1.71E-01
8.85E-02

6.37E-02
2.96E+01
5.63E-01
7.43E-01
3.57E+00
—
7.82E-01
3.05E+00
1.41E+00
3.18E-01
5.51E-02
—
BMDLd
_e
4.34E-02

	 e
-
-
	 e
-
—
-
-
2.61E-01

_e
—
to

-------
          Table 4-3.  Summary of key animal study points of departure (PODs) (ng/kg-day) based on three different dose
          metrics: administered dose, lst-order body burden human equivalent dose (HED) and blood concentration HED
          (continued)
Study
Hutt et al. (2008)
Ikeda et al. (2005)
Ishihara et al. (2007)
ICattainen et al.
(2001)
Keller et al. (2008a;
2008b: 2007)
Kociba et al. (1976)
Kociba et al. (1978)
Kuchiiwa et al.
(2002)
Latchoumycandane
and Mathur (2002)h
Lietal. (1997)
Li et al. (2006)
Markowski et al.
(2001)
Maronpot et al.
(1993)
Miettinen et al.
(2006)
Murray etal. (1979)
NTP (1982b)
NTP (2006a)
Nohara et al. (2000)
Nohara et al. (2002)
Ohsakoetal. (2001)
Endpoint
Embyrotoxicity
Sex ratio
Sex ratio
3rd molar length
Missing mandibular molars
Liver and hematologic effects
and body -weight changes
Liver and lung lesions,
increased urinary porphyrins
[mmunoreactive neurons
Sperm production
Increased serum FSH
Hormone levels (serum
sstradiol)
FR2 revolutions
Increased relative liver weight
Cariogenic lesions in pups
Fertility index in F2 generation
Liver lesions
Liver and lung lesions
Decreased spleen cellularity
Mortality from influenza
virus-A challenge
Anogenital distance in pups
Administered dose"
NOAEL
-
-
l.OOE-01
—
—
7.14E+00
l.OOE+00
—
—
3.00E+00
—
—
1.07E+01
—
l.OOE+00
-
-
8.00E+02
5.00E+02
1.25E+01
LOAEL
7.14E+00
1.65E+01
l.OOE+02
3.00E+01
l.OOE+01
7.14E+01
l.OOE+01
7.00E-01
l.OOE+00
l.OOE+01
2.00E+00
2.00E+01
3.50E+01
3.00E+01
l.OOE+01
1.39E+00
2.14E+00
-
—
5.00E+01
BMDLd
-
-
-
e
e
—
e
—
e
e
e
e
—
e
e
e
e
-
—
e
ls'-order body burden HEDb
NOAEL
-
-
3.18E-04
—
—
4.53E-02
1.07E-02
—
—
7.89E-04
—
—
8.97E-02
—
9.43E-03
-
-
2.10E-01
1.29E-01
3.29E-03
LOAEL
4.67E-02
1.05E-01
3.18E-01
7.89E-03
2.58E-03
4.53E-01
1.07E-01
3.11E-03
3.87E-03
2.63E-03
9.85E-04
6.25E-03
2.93E-01
7.89E-03
9.43E-02
6.47E-03
2.34E-02
-
—
1.32E-02
BMDLd
-
-
-
e
e
—
e
—
e
e
e
e
—
e
e
e
e
-
—
e
Blood concentration HEDC
NOAEL
-
-
4.91E-05
—
—
2.62E-01
6.33E-02
—
—
2.90E-03
—
—
5.03E-01
—
2.89E-02
-
-
5.34E+00
1.37E+00
2.74E-02
LOAEL
2.52E-01
2.75E+00
4.96E-01
9.01E-02
9.88E-03
3.03E+00
6.34E-01
2.75E-03
1.62E-02
1.67E-02
1.58E-03
5.15E-02
1.71E+00
8.95E-02
3.79E-01
2.16E-02
1.36E-01
-
—
1.78E-01
BMDLd
-
-
-
e
e
—
e
e
e
e
e
e
—
e
e
e
e
-
—
e
to

-------
          Table 4-3.  Summary of key animal study points of departure (PODs) (ng/kg-day) based on three different dose
          metrics: administered dose, lst-order body burden human equivalent dose (HED) and blood concentration HED
          (continued)
Study
Schantzetal.fi 996)
Seoetal. f!995)
Sewalletal. f!995)
Shi et al. f2007)
Simanainen et al.
f2002)
Simanainen et al.
f2003)
Simanainen et al.
f2004)
Smialowicz et al.
f2004)
Smialowicz et al.
f2008)
Smithetal.fi 976)
Sparschu et al.
f!971)
rothetal. f!979)
VanBirgelen et al.
fl995aV
Vos et al. f!973)
Weberetal.fi 995)
Whiteetal.fi 986)
Yang et al. (2000)
Endpoint
Maze errors
Decreased thymus weight
Serum T4
Serum estradiol in female pups
Decreased serum T4
Decreased thymus weight and
change in EROD activity
Decreased daily sperm
production
Decreased antibody response
to SRBCs
PFC per 10A6 cells
Cleft palate in pups
Decreased fetal body weight
Skin lesions
Decreased liver retinyl
palmitate
Decreased delayed-type
lypersensitivity response to
tuberculin
Increased liver weight
Decreased serum complement
Increased endometrial implant
survival
Administered dose"
NOAEL
-
2.50E+01
1.07E+01
1.43E-01
l.OOE+02
l.OOE+02
l.OOE+02
3.00E+02
—
l.OOE+02
3.00E+01
-
—
1.14E+00
l.OOE+03
-
1.79E+01
LOAEL
2.50E+01
l.OOE+02
3.50E+01
7.14E-01
3.00E+02
3.00E+02
3.00E+02
l.OOE+03
1.07E+00
l.OOE+03
1.25E+02
l.OOE+00
1.35E+01
5.71E+00
3.00E+03
l.OOE+01
—
BMDLd
e
-
5.16E+00
2.24E-01
—
—
—
—
e
1.84E+02
e
e
e

-
e
—
ls'-order body burden HEDb
NOAEL
-
2.49E-02
8.97E-02
1.23E-03
2.63E-02
2.63E-02
2.63E-02
7.73E-02
—
1.59E-01
5.45E-02
-
—
6.43E-03
3.51E-01
-
6.74E-01
LOAEL
e
9.96E-02
2.93E-01
6.13E-03
7.89E-02
7.89E-02
7.89E-02
2.58E-01
5.00E-03
1.59E+00
2.27E-01
3.70E-03
8.32E-02
3.22E-02
1.05E+00
2.23E-02
—
BMDLd
4.55E-02
-
4.33E-02
1.92E-03
—
—
—
—
e
2.93E-01
—
e
e

-
e
—
Blood concentration HEDC
NOAEL
-
1.67E-01
5.03E-01
4.47E-03
4.26E-01
4.26E-01
4.26E-01
7.23E-01
—
5.24E-01
3.18E-01
-
—

3.27E+00
-
—
LOAEL
1.71E-01
9.15E-01
1.71E+00
2.69E-02
1.67E+00
1.67E+00
1.67E+00
3.28E+00
6.26E-03
7.61E+00
1.73E+00
9.91E-03
5.14E-01

1.18E+01
2.77E-02
—
BMDLd
e
-
1.80E-01
4.74E-03
—
—
—
—
e
9.46E-01
e
e
e

-
e
—
to

-------
             Table 4-3. Summary of key animal study PODs (ng/kg-day) based on three different dose metrics: administered
             dose, lst-order body burden HED and blood concentration HED (continued)

     aAverage administered daily dose over the experimental exposure period.
     bHED based on lst-orderbody burden model described in Section 3.3.4.2.
     °HED based on Emond rodent and human PBPK models described in Section 3.3.6.
     dBMR = 0.1 for quanta! endpoints and 1 standard deviation control mean for continuous endpoints, except for body and organ weights, where BMR = 10%
      relative deviation from control mean.
     eBMD modeling unsuccessful (see Table 4-4 and Appendix G for details).
     fZareba et al. (2002) is considered to be the same study but report effects at doses above the LOAEL that are not considered further; this study is not carried
       forward for determination of an RfD POD but is included in the RfD uncertainty analysis presented in Section 4.4.
     8Hochstein et al. (2001) is not carried forward because of the lack of toxicokinetic information for estimation of an HED.
     hLatchoumycandane et al. (2002a: 2002b) are considered to be the same study but report effects (not lexicologically relevant) at doses above the LOAEL that are
       not considered further; these two studies are not carried forward.
     'Van Birgelen et al. (1995b) is considered to be the same study but reports effects at doses above the LOAEL that are not considered further; this study in not
       carried forward for determination of an RfD POD but is included in the RfD uncertainty analysis presented in Section 4.4.

     - value not established or not modeled; DRL = differential reinforcement of low rate.
to

-------
          Table 4-4.  TCDD analysis (NOAEL, LOAEL, BMD, and BMDL values given as animal whole blood

          concentrations in ppt)a
Studyb'c
Amin et al.
(2000)
(rat)
NOAEL/
LOAEL
3.38E+00
Endpoint
Saccharin consumed,
female, (0.25%) (n =
10)

Saccharin consumed,
female (0.50%) (n =
10)

Saccharin preference
ratio, female (0.25%)
(n = 10)
Saccharin preference
ratio, female (0.50%)
(n = 10)

Control
response







First
response"1
22% |
(0.3 SD)

49% |
(0.7 SD)

29% |
(1.8 SD)
39% |
(1.1 SD)

Max
response6
66% |

80% |

33% |
54% |

Model fit detail
Continuous linear,
modeled variance
(p=0.55)
Continuous power,
modeled variance,
unrestricted
(p=NA)
Continuous linear,
modeled variance
(p = 0.06)
Continuous power,
modeled variance,
unrestricted
(p=NA)
Continuous linear,
modeled variance
(p = 0.002)
Continuous linear,
constant variance
(p=0.14)
Continuous power,
constant variance,
unrestricted
(p=NA)
BMD/
BMDL
9.15E+00
6.09E+00
8.37E+00
3.42E+00
1.02E+01
6.57E+00
6.57E+00
1.15E+00
1.16E+01
5.57E+00
8.14E+00
5.11E+00
2.60E+00
1.06E-14
Comments
BMDL > LOAEL; restricted power
model, constrained parameter hit
ower bound
Saturated model; supralinear fit
power = 0.74)
Restricted power model,
constrained parameter hit lower
)ound
Saturated model; supralinear fit
power = 0.40)
BMDL > LOAEL; no response
near BMR; near maximal response
at LOAEL
BMDL > LOAEL; near maximal
response at LOAEL; restricted
)ower model, constrained
parameter hit lower bound
Saturated model; supralinear fit
(power = 0.28)
to
oo

-------
          Table 4-4. TCDD BMD analysis (NOAEL, LOAEL, BMD, and BMDL values given as animal whole blood
          concentrations in ppt)a (continued)
Studyb'c
Bell et al.
(2007b)
(rat)
Cantoni et al.
(1981)
(rat)
Crofton et al.
(2005)
(rat)
NOAEL/
LOAEL
2.20E+00
1.85E+00
3.46E+00
9.26E+00
Endpoint
Balano-preputial
separation in male
pups
(n = 30 [dams])
Urinary uroporhyrins
(« = 4)
Urinary
coproporhyrins
(» = 4)
Serum T4,
(n = 4-14)
Control
response
1/30



First
response"1
5/30
2.4-fold t
(5.7 SD)
2.4-fold t
(3.1 SD)
29% |
(1.9 SD)
Max
response6
15/30
87-fold t
4.0-fold t
51% 4
Model fit detail
Dichotomous log-
logistic, restricted
(p = 0.78)
Dichotomous log-
logistic, unrestricted
(p = 0.50)
Continuous
exponential (M2),
modeled variance
(p = 0.0003)
Continuous
exponential (M4),
modeled variance
(p = 0.49)
Continuous power,
modeled variance,
unrestricted
(p = 0.61)
Continuous
exponential (M4),
constant variance
(p = 0.94)
BMD/
BMDL
2.25E+00
1.39E+00
2.00E+00
2.80E-01
3.76E+00
2.76E+00
5.34E-01
1.80E-01
2.77E-02
2.03E-05
5.19E+00
3.03E+00
Comments
Adequate fit; constrained
parameter bound hit; not litter
sased; selected
Supralinear fit
(slope = 0.93); selected
No response near BMR; poor fits
for all modeled variance models;
constant variance poor
representation of control SD;
BMDL > LOAEL
No response near BMR
Supralinear fit (n = 0.30); poor
model choice for plateau effect
No response near BMR
to
VO

-------
          Table 4-4. TCDD BMD analysis (NOAEL, LOAEL, BMD, and BMDL values given as animal whole blood
          concentrations in ppt)a (continued)
Studyb'c
Franc et al.
(2001)
(rat)
NOAEL/
LOAEL
6.59E+00
1.45E+01
Endpoint
S-D Rats, Relative
Liver Weight
L-E Rats, Relative
Liver Weight

S-D Rats, Relative
Thymus Weight

L-E Rats, Relative
Thymus Weight
H-W Rats, Relative
Thymus Weight
Control
response







First
response"1
8.1%t
(0.58 SD)
6.3% t
(0.63 SD)

9.0% |
(0.11SD)

7.7% |
(0.15SD)
3.1% \,
(0.10SD)
Max
response6
55% t
22% t

77% 4

66% 4
51% 4
Model fit detail
Continuous power,
constant variance
(p = 0.84)
Continuous Hill,
modeled variance,
restricted
(p = 0.83)
Continuous Hill,
modeled variance,
unrestricted
(p = N/A)
Continuous
exponential (M4),
modeled variance
(P = 0.72)
Continuous
polynomial, modeled
variance
(p = 0.40)
Continuous
exponential (M4),
constant variance
(p = 0.23)
Continuous
exponential (M2),
constant variance
(p = 0.70)
BMD/
BMDL
9.47E+00
4.59E+00
7.72E+00
1.22E+00
7.22E+00
1.15E+00
1.88E+00
9.22E-01
4.78E+00
3.89E+00
2.08E+00
5.93E-01
5.09E+00
3.13E+00
Comments
Acceptable fit; selected
Constrained parameter hit lower
sound; poor fit for variance model
Supralinear fit (power = 0.55)
Poor fit for responses in controls
and lowest exposure group
No response near BMR; otherwise
acceptable fit
Poor fit for responses in controls
and lowest exposure group;
dose-response relationship not
significant
No response near BMR; otherwise
acceptable fit
-^
o

-------
Table 4-4. TCDD BMD analysis (NOAEL, LOAEL, BMD, and BMDL values given as animal whole blood
concentrations in ppt)a (continued)
Studyb'c
Hojo et al.
(2002)
(rat)
Kattainen et al.
(2001)
(rat)
Keller et al.
(2008a; 2008b;
2007)
(mouse)
NOAEL/
LOAEL
1.62E+00
2.23E+00
5.37E-01
Endpoint
DRL reinforce per
minute
(n = 12)
DRL response per
minute
(n = 12)
3rd molar length in
pups
(n = 4-8)

3rd molar eruption in
pups
(n = 4-8)

Missing molars
(n = 23-36)
Control
response




1/16

0/29
First
response"1
55% t
(1.0 SD)
105% |
(2.4 SD)
15% |
(4.2 SD)

3/17

2/23
Max
response6
80% t
105% |
27% |

13/19

30/30
Model fit detail
Continuous
exponential (M4),
constant variance
(p = 0.054)
Continuous
exponential (M4),
constant variance
(p = 0.48)
Continuous Hill,
modeled variance,
restricted
(p = 0.02)
Continuous Hill,
modeled variance,
unrestricted
(p< 0.001)
Dichotomous log-
logistic, restricted
(p = 0.98)
Dichotomous log-
logistic, unrestricted
(p = 0.95)
Dichotomous 1°
multistage
(p = 0.26)
BMD/
BMDL
1.32E+00
2.37E-03
3.81E-01
1.55E-02
3.13E-01
1.68E-01
1.21E-02
2.40E+00
1.33E+00
1.93E+00
1.84E-01
1.09E+00
7.62E-01
Comments
Poor fit; near maximal response at
lowest dose, BMD/BMDL ratio
>100
No response data near BMR;
maximal response at lowest dose,
BMD/BMDL ratio »20
No response data near BMR;
Constrained parameter lower
sound hit
BMDL could not be calculated
Constrained parameter lower
sound hit
Supralinear fit (slope = 0.91)
Poor fit at first response level; not
most sensitive endpoint; other
endpoints not amenable to BMD
modeling

-------
          Table 4-4. TCDD BMD analysis (NOAEL, LOAEL, BMD, and BMDL values given as animal whole blood
          concentrations in ppt)a (continued)
Studyb'c
Kociba et al.
(1978)
(rat)
Kuchiiwa et al.
(2002) (mouse)
NOAEL/
LOAEL
1.55E+00
7.15E+00
1.42E+02
Endpoint
Uroporphyrin per
creatinine, females
(» = 5)
Urinary
coproporphyrins,
females
(» = 5)
Liver lesions
(n = 50)
Lung lesions
(n = 50)
Immunoreactive
Neurons in Dorsalis,
males
(« = 6)
Immunoreactive
Neurons in
Medianus, males
(» = 6)
Immunoreactive
Neurons inB9,
males
(« = 6)
Immunoreactive
Neurons in Magnus,
males
(« = 6)
Control
response








First
response"1
15% t
(0.48 SD)
67% t
(5.1 SD)


42% |
(3.5 SD)
63% |
(4.8 SD)
69% |
(6.6 SD)
55% |
(7.0 SD)
Max
response6
89% t
78% t


64% |
75% |
87% |
75% |
Model fit detail
Continuous linear,
constant variance
(p = 0.79)
Continuous
exponential (M4),
modeled variance
(p = 0.01)


Continuous linear,
constant variance
(p = NA, insufficient
degrees of freedom)
Continuous linear,
modeled variance
(p = NA, insufficient
degrees of freedom)
Continuous linear,
constant variance
(p = NA, insufficient
degrees of freedom)
Continuous linear,
modeled variance
(p = NA, insufficient
degrees of freedom)
BMD/
BMDL
1.31E+01
9.29E+00
1.57E+00
7.18E-01


6.04E-02
4.27E-02
4.93E-02
3.23E-02
4.17E-02
3.01E-02
3.35E-02
2.05E-02
Comments
BMDL > LOAEL; otherwise
adequate fit
Poor fit; no response near BMR
No data presented
No data presented
No response near BMR
No response near BMR
No response near BMR
No response near BMR
-^
to

-------
Table 4-4. TCDD BMD analysis (NOAEL, LOAEL, BMD, and BMDL values given as animal whole blood
concentrations in ppt)a (continued)
Studyb'c
Latchoumy-
candane and
Mathur (2002)
(rat)
Li et al. (1997)
(rat)
Li et al. (2006)
(mouse)
NOAEL/
LOAEL
7.85E-01
2.66E-01
7.99E-01
1.59E-01
Endpoint
Daily sperm
production
(« = 6)
FSH in female rats
(n = 10)

Serum estradiol
(n = 10)
Serum progesterone
(n = 10)
Control
response





First
response"1
29% |
(1.0 SD)
3.6-fold t
(2.0 SD)

2.0-fold t
(0.8 SD)
33% |
(2.0 SD)
Max
response6
41% 4
19-fold t

2.4-fold t
61% 4
Model fit detail
Continuous Hill,
constant variance,
restricted
(p = 0.96)
Continuous Hill,
constant variance,
unrestricted
(p = N/A)
Continuous power,
modeled variance,
restricted
(p<0.01)
Continuous power,
modeled variance,
unrestricted
(p = 0.003)
Continuous linear,
constant variance
(p=0.16)
Continuous Hill,
modeled variance
(p=0.39)
BMD/
BMDL
1.17E-01
1.32E-02
9.96E-02
1.23E-09
2.00E+02
1.36E+02
1.96E-01
2.48E-02
1.61E+01
5.38E+00
9.46E-04
8.01E-11
Comments
Near maximal response at LOAEL;
constrained parameter bound hit;
standard deviations given in paper
interpreted as standard errors
Slightly supralinear fit (n =0.92)
Power hit lower bound
Supralinear fit (power = 0.31)
BMDL > LOAEL; high control
coefficient variation (CV) (1.25);
near maximal response at low
dose; nonmonotonic response;
other model fits are step-function-
like
No response data near BMR; large
CVs (>1) for treatment groups;
poor fit for variance model; Hill
coefficient at lower bound (step-
function)

-------
Table 4-4. TCDD BMD analysis (NOAEL, LOAEL, BMD, and BMDL values given as animal whole blood
concentrations in ppt)a (continued)
Studyb'c
Markowski et al.
(2001)
(rat)
Miettinen et al.
(2006)
(rat)
Murray et al.
(1979)
(rat)
NTP (1982b)
(mouse)
NOAEL/
LOAEL
1.56E+00
2.22E+00
1.12E+00
5.88E+00
7.67E-01
Endpoint
FR5 run
opportunities
(n = 4-7)
FR2 revolutions
(n = 4-7)
FR10 run
opportunities
(n = 4-7)
Cariogenic lesions in
pups
(n = 4-8)

Fertility in F2 gen.
(no litters)
(n = 20)
Toxic hepatitis; males
(n = 50)
Control
response



25/42

4/32
1/73
First
response"1
10% |
(0.21 SD)
9% 4
(0.15SD)
15% 4
(0.24 SD)
23/29

0/20
5/49
Max
response6
51% 4
43% 4
57% 4
29/32

9/20
44/50
Model fit detail
Continuous Hill,
constant variance
(p = 0.94)
Continuous power,
constant variance,
unrestricted
(p=0.13)
Continuous Hill,
constant variance
(p = 0.65)
Continuous power,
constant variance,
unrestricted
(p=0.16)
Continuous
exponential (M2) ,
constant variance
(p=0.30)
Dichotomous log-
logistic, restricted
(p = 0.60)
Dichotomous log-
logistic, unrestricted
(p=0.73)
Dichotomous
multistage
(p = 0.08)
Dichotomous
multistage
(p = 0.04)
BMD/
BMDL
1.72E+00
9.08E-01
2.67E+00
1.03E-14
1.84E+00
5.99E-01
5.74E+00
1.03E-14
8.57E+00
2.89E+00
1.43E+00
5.17E-01
4.94E-02
2.73E+00
1.37E+00
2.78E+00
1.34E+00
Comments
Constrained parameter upper
sound hit
Saturated model; supralinear fit
(power = 0.39); BMD/BMDL ratio
»100
Constrained parameter bound hit
(upper bound)
Supralinear fit (power = 0.32)
BMDL > LOAEL
Constrained parameter lower
sound hit; near maximal response
at LOAEL; high control response
Supralinear fit (slope = 0.47);
BMDL could not be calculated
Poor fit; nonmonotonic response;
no response data near BMR
No acceptable model fits; lowest
BMDL shown

-------
Table 4-4. TCDD BMD analysis (NOAEL, LOAEL, BMD, and BMDL values given as animal whole blood
concentrations in ppt)a (continued)
Studyb'c
NTP(2QQ6a)
(rat)
NTP(2006a)
(rat) (continued)
NOAEL/
LOAEL
2.56E+00
2.56E+00
(continued)
Endpoint
Hepatocyte
hypertrophy
(n = 53-54)
Alveolar metaplasia
(n = 52-54)
Oval cell hyperplasia
(n = 53-54)
Gingival hyperplasia
(n = 53-54)
Eosinophilic focus,
multiple
(n = 53-54)
Liver fatty change,
diffuse
(n = 53-54)
Liver necrosis
(n = 53-54)
Liver pigmentation
(n = 53-54)
Toxic hepatopathy
(n = 53-54)
Control
response
0/53
2/53
0/53
1/53
3/53
0/53
1/53
4/53
0/53
First
response"1
19/54
19/54
4/54
7/54
8/54
2/54
4/54
9/54
2/54
Max
response6
52/53
46/52
53/53
16/53
42/53
48/53
17/53
53/53
53/53
Model fit detail
Dichotomous
multistage
(p = 0.02)
Dichotomous log-
logistic
(P = 0.72)
Dichotomous probit
(p=0.23)
Dichotomous Weibull
(p = 0.08)
Dichotomous log-
logistic, restricted
(p = 0.06)
Dichotomous log-
logistic, unrestricted
(p = 0.66)
Dichotomous probit
(p = 0.46)
Dichotomous Weibull
(P = 0.72)
Dichotomous log-
probit, unrestricted
(p = 0.80)
Dichotomous log-
probit
(p = 0.96)
Dichotomous
multistage
(p = 0.69)
BMD/
BMDL
9.27E-01
7.91E-01
6.50E-01
3.75E-01
5.67E+00
4.79E+00
5.72E+00
4.09E+00
5.85E+00
3.73E+00
7.05E-01
1.26E-05
5.58E+00
4.86E+00
3.92E+00
2.86E+00
7.50E+00
3.50E+00
2.46E+00
1.89E+00
3.98E+00
3.06E+00
Comments
Poor fits for all models
No response near BMR
Relatively poor fit for control and
low-dose groups; negative
response intercept (same for
logistic); BMDL > LOAEL
Marginal fit; BMDL > LOAEL
Poor fit; constrained parameter
bound hit; BMDL > LOAEL
Supralinear fit (slope = 0.37)
Relatively poor fit to control
response; BMDL > LOAEL
BMDL > LOAEL; otherwise
adequate fit
Adequate fit; slightly supralinear;
BMDL > LOAEL
Adequate fit
BMDL > LOAEL; otherwise
adequate fit

-------
          Table 4-4. TCDD BMD analysis (NOAEL, LOAEL, BMD, and BMDL values given as animal whole blood
          concentrations in ppt)a (continued)
Studyb'c
Ohsako et al.
(2001)
(rat)
Schantz et al.
(1996)
Sewall et al.
(1995)
(rat)
Shi et al. (2007)
(rat)
Smialowicz et al.
(2008)
(mouse)
NOAEL/
LOAEL
1.04E+00
3.47E+00
3.38E+00
7.11E+00
1.66E+01
3.42E-01
1.07E+00
4.38E-01
Endpoint
Anogenital distance
in male pups
(« = 5)

Facilitory effect on
radial arm maze
learning
(n = 10)
Serum T4
(» = 9)
Serum estradiol in
female pups
(n = 10)
PFC per spleen
(n = 15)
PFC per 10A6 cells
(n = 8-15)
Control
response








First
response"1
12% |
(1.0 SD)

22% |
(1.2 SD)
9.1%|
(0.6 SD)

38% |
(0.4 SD)
24% |
(0.5 SD)
24% |
(0.5 SD)
Max
response6
17% |

34% |
40% |

62% |
89% |
9.3-fold |
Model fit detail
Continuous Hill,
constant variance,
restricted
(p=0.15)
Continuous Hill,
constant variance,
unrestricted
(p = 0.056)
Continuous linear,
constant variance
(p = 0.16)
Continuous Hill,
constant variance,
restricted
(p = 0.90)
Continuous Hill,
constant variance,
unrestricted
(p = 0.86)
Continuous
exponential (M4),
modeled variance
(p = 0.69)
Continuous power,
unrestricted, modeled
variance
(P = 0.27)
Continuous power
unrestricted, constant
variance
(p = 0.48)
BMD/
BMDL
2.88E+00
8.03E-01
3.49E+00
3.05E-01
7.00E+00
4.60E+00
1.03E+01
3.60E+00
9.71E+00
1.97E+00
8.07E-01
3.54E-01
1.19E+01
3.76E+00
1.90E+00
2.16E-01
Comments
Constrained parameter lower
sound hit; near maximal response
at LOAEL
Supralinear fit (n = 0.59)
BMDL > LOAEL; otherwise
adequate fit
Constrained parameter hit lower
sound; otherwise acceptable fit;
selected
Supralinear fit (power = 0.57)
Adequate fit; selected
BMDL > LOAEL; fit at control
and low dose inconsistent with
data; constrained parameters in
other models hit lower bounds
Constant variance test failed;
observed control variance
underestimated by 35%; poor fits
for all modeled variance models
-^
ON

-------
Table 4-4. TCDD BMD analysis (NOAEL, LOAEL, BMD, and BMDL values given as animal whole blood
concentrations in ppt)a (continued)
Studyb'c
Smith et al.
(1976)
(mouse)
Sparschu et al.
(2008: 1971)
(rats)
Toth et al.
(1979)
(mouse)
Van Birgelen
et al. (1995a)
(rat)
NOAEL/
LOAEL
7.11E+00
5.06E+01
5.09E+00
1.63E+01
5.73E-01
5.73E-01
(cont.)
7.20E+00
Endpoint
Cleft palate in pups (n
= 14-41)
Male fetus weight
(w = 3-117)
Female fetus weight
(n = 4-129)
Skin lesions
(n = 38-44)
Dermal amyloidosis
(n = 38-44)
Hepatitic retinol
(« = 8)

Control
response
0/34


0/38
0/38


First
response"1
2/41
2.7% t
(0.1 SD)
2.3% t
(0.06 SD)
5/44
5/44
44% |
(0.74 SD)

Max
response6
10/14
33% |
30% |
25/43
17/43
96% |

Model fit detail
Dichotomous log-
logistic, restricted
(p = 0.42)
Continuous
exponential (M5),
modeled variance
(p < 0.0001)
Continuous
exponential (M2),
modeled variance
(p < 0.028)
Dichotomous log-
logistic, restricted
(p = 0.08)
Dichotomous
log-logistic,
unrestricted
(p = 0.74)
Dichotomous log-
logistic, restricted
(p = 0.05)
Dichotomous log-
logistic, unrestricted
(p = 0.90)
Continuous
exponential (M4),
modeled variance
(p<0.01)
Continuous power,
modeled variance,
unrestricted
(p = 0.01)
BMD/
BMDL
3.52E+01
1.06E+01
5.46E+02
1.30E+02
1.03E+03
6.48E+02
6.41E+00
4.02E+00
5.97E-01
6.77E-02
1.50E+01
8.75E+00
4.84E-01
5.31E-03
2.49E+01
3.36E+00
3.80E-01
1.39E-02
Comments
Adequate fit; selected
BMDL > LOAEL; variance not
captured by either variance model;
poor fit in region surrounding
NOAEL and LOAEL
BMDL > LOAEL; variance not
captured by either variance model;
poor fit in region surrounding
NOAEL and LOAEL
Constrained parameter lower
sound hit
Supralinear fit (slope = 0.48)
Poor fit; constrained parameter
lower bound hit; BMDL > LOAEL
Supralinear fit (slope = 0.33)
Poor fit
Poor fit; Supralinear fit
(power = 0.14)

-------
             Table 4-4. TCDD BMD analysis (NOAEL, LOAEL, BMD, and BMDL values given as animal whole blood
             concentrations in ppt)a (continued)
Studyb'c

White et al.
(1986)
(mouse)
NOAEL/
LOAEL

1.09E+00
Endpoint
Hepatitic retinyl
palmitate (n = 8)

Total hemolytic
complement activity
(CH50)
(« = 8)
Control
response



First
response"1
80% |
(1.4 SD)

41% 4
(2.6 SD)
Max
response6
99% 4

81% 4
Model fit detail
Continuous
exponential (M4),
modeled variance
(p<0.01)
Continuous power,
modeled variance,
unrestricted
(p = 0.24)
Continuous
Hill, modeled variance,
restricted
(p = 0.002)
Continuous Hill,
modeled variance,
unrestricted
(p = 0.07)
BMD/
BMDL
1.42E+02
3.65E+01
5.26E-02
5.89E-05
8.63E+00
1.50E+00
1.48E-01
4.35E-03
Comments
Poor fit; no response near BMR
Supralinear fit (power = 0.06)
Poor fit; no response near BMR;
constrained parameter bound hit;
BMDL > LOAEL
Supralinear fit (n = 0.25)
-^
oo
     aAnimal whole blood concentrations were used to determine the HEDs in Table 4-3 and Table 4-5.
     bThe following studies previously presented in Table 4-3 are not presented in Table 4-4 because toxicokinetic models for guinea pigs, minks, or monkeys, and
     were not found: DeCaprio et al. (1986); Hochstein et al  (2001); Vos et al. (1973); Yang et al. (2000).
     °The following studies previously presented in Table 4-3 are not presented in Table 4-4 because the data were not amenable to BMD modeling:  Chu et al. (2001)
     Chu et al. (2007); Croutch et al. (2005); Fattore et al. (2000); Fox et al. (1993); Franczak et al. (2006); Hutt et al. (2008); Ikeda et al. (2005); Ishihara et al.
     (2007); Kociba et al. (1976); Maronpot et al. (1993); Nohara et al. (2000); Nohara et al. (2002); Seo et al. (1995); Simanainen et al. (2002); Simanainen et al.
     (2003); Simanainen et al. (2004); Smialowicz et al. (2004); Weber et al. (1995).
     dMagnitude of response at first dose where response differs from control value (in the adverse direction); continuous response magnitudes given as relative to
      control plus change relative to control standard deviation; quantal response given as number affected/total number.
     TVIagnitude of response maximally differing from control value (in the adverse direction).
     SD = standard deviation; S-D = Sprague-Dawley; L-E = Long-Evans; H-W = Han-Wistar; DRL = differential reinforcement of low rate.

-------
4.3.  REFERENCE DOSE (RfD) DERIVATION
       Table 4-5 lists all the studies and endpoints considered for derivation of the RfD in order
of candidate RfD from lowest to highest (The selection process was previously described in
Section 4.1). The range of studies includes three of the four human studies.42  Figure 4-4
(exposure-response array) shows all of the endpoints listed in Table 4-5 graphically in terms of
PODs in human-equivalent intake units (ng/kg-day). The human study endpoints are shown at
the far left of the figure, and the animal bioassay endpoints are arranged by category to the right.
Figure 4-5 demonstrates the same endpoints, arrayed by RfD value, showing the POD,  applicable
UFs, and candidate RfD.
       Table 4-5 illustrates the study, species, strain and sex, study protocol, and toxicological
endpoints observed at the lowest TCDD doses.  The table also identifies the human-equivalent
BMDLs (when applicable), NOAELs, and LOAELs, as well as the composite uncertainty factor
(UF) that applies to the specific endpoint and the corresponding candidate RfD.43 The NOAELs,
LOAELs, and BMDLs are presented as FtEDs, based on the assumption that whole-blood
concentration is the toxicokinetically equivalent TCDD dose metric across species and serves as
a surrogate for tissue concentration.44 For rats and mice, these estimates relied on the two
Emond PBPK models—one for the relevant rodent species and one for the human—as  described
previously (see Section 3.3.4.3). The guinea pig and monkey studies that are included in
Table 4-5 are given in FED units based on the first-order body burden model (described in
Section 3.3.4.2) because there are no published PBPK models to estimate TCDD disposition in
guinea pigs and monkeys.  The values listed for guinea pigs and monkeys are not directly
comparable to those for rats and mice but are probably biased low, as first-order body burden
FLED estimates for rats and mice are generally two to fivefold lower than the corresponding
PBPK  model estimates. The LOAELs for the human studies also rely on the Emond PBPK
model, as described in Sections 4.2.2 and 4.2.3.
42 The RfD derived from the study of Eskenazi et al. (2002b) was outside the RfD range presented in Table 4-5.
43 Extra digits are retained for transparency and comparison prior to rounding to one significant digit for the final
RfD.
44 The procedures for estimating HEDs based on TCDD blood concentration are described in the preceding section.
                                          4-39

-------
          Table 4-5. Candidate RfDs for TCDD using blood-concentration-based human equivalent doses
Study
Li et al. (2006)
Kuchiiwa et al.
(2002)

Smialowicz
et al. (2008)
Bowman
etal.(1989a;
1989b); others'3
Keller et al.
(2008a; 2008b;
2007)d
Toth et al.
(1979)
Latchoumy-
candane and
Mathur (2002):
others6
NTP (1982b)
White et al.
(1986)
Li et al. (1997)
DeCaprio et al.
(1986)
Shi et al. (2007)
Markowski
etal. (2001)
Species, strain
(sex, if not
both)
Mouse, NIH (F)
Mouse, ddY
Mouse, B6C3FJ
(F)
Rhesus Monkey
(F)
Mouse, CBA/J
and C3H/HeJ
Mouse, Swiss/
H/Riop (M)
Rat, Wistar (M)
Mouse, B6C3FJ
(M)
Mouse, B6C3FJ
(F)
Rat, S-D
(F, 22 day-old)
Guinea pig,
Hartley
Rat, S-D (F)
Rat, Holtzman
Protocol
Gavage CDs 1-3;
n = 10
Maternal 8 week-
gavage prior to
mating; n = 3
90-day gavage;
n = 8-15
Daily dietary
exposure, 3.5-4
years
w = 3-7
Gavage GD 13;
n = 23-36 (pups)
1 -year gavage;
n = 38-44
45-day oral
pipetting; n = 6
2-year gavage;
w = 50
14-day gavage;
w = 6-8
Single gavage;
n= 10
90-day dietary;
w = 10
11 -month gavage;
n = 10
Gavage GD 18;
w = 4-7
Endpoint
Hormone levels in pregnant dams (decreased
progesterone, increased estradiol)
Decreased serotonin-immunoreactive neurons
in raphe nuclei of male offspring (Fl)
Decreased SRBC response
Neurobehavioral effects
Missing molars, mandibular shape changes in
pups
Dermal amyloidosis, skin lesions
Decreased sperm production
Liver lesions
Decreased serum complement
Increased serum FSH
Decreased body weight, organ weight
changes (liver, kidney, thymus, brain)
Decreased serum estradiol
Neurobehavioral effects in pups (running,
lever press, wheel spinning)
NOAELHED(N)or
BMDLHED (B)
(ng/kg-day)
—

—


—



2.9E-03 (N)
4.1E-03C(N)
4.5E-03 (N)
4.7E-03 (B)
—
LOAELsED
(ng/kg-day)
1.6E-03
2.7E-03
6.3E-03
8.2E-03C
9.9E-03
9.9E-03
1.6E-02
2.2E-02
2.8E-02
1.7E-02
3.3E-02C
2.7E-02
5.2E-02
UFa
300
300
300
300
300
300
300
300
300
301
30'
301
300
RfD
(mg/kg-day)
5.3E-12
9.2E-12
2.1E-11
2.7E-11
3.3E-11
3.3E-11
5.4E-11
7.2E-11
9.2E-11
9.7E-11
1.4E-10
1.6E-10
1.7E-10
-^
o

-------
Table 4-5. Candidate RfDs for TCDD using blood-concentration-based human equivalent doses (continued)
Study
Hojo et al.
(2002); Zareba
et al. (2002)
Cantoni et al.
(1981)
Vos et al.
(1973)
Miettinen et al.
(2006)
Kattainen et al.
(2001)
NTP (2006a)
Amin et al.
(2000)
Schantz et al.
(1996)
Mocarelli et al.
(2008)
Baccarelli
et al. (2008)
Hutt et al.
(2008)
Ohsako et al.
(2001)
Murray et al.
(1979)
Franczak et al.
(2006)
Chu et al.
(2007)
Bell et al.
(2007b)
Species, strain
(sex, if not
both)
Rat, S-D
Rat, CD-COBS
(F)
Guinea pig,
Hartley (F)
Rat, Line C
Rat, Line C
Rat, S-D (F)
Rat, S-D
Rat, S-D (F)
Human (M)
Human infants
Rat, S-D (F)
Rat, Holtzman
Rat, S-D
Rat, S-D (F)
Rat, S-D (F)
Rat, CRL:WI
(Han) (M)
Protocol
Gavage GD 8;
n = 12
45-week gavage;
« = 4
8-week gavage;
n = 10
Gavage GD 15;
n = 3-10
Gavage GD 15;
w=4-8
2-year gavage;
w = 53
Gavage CDs 10-16;
n = 10
Gavage CDs 10-16;
n = 80-88
Childhood
exposure; n = 157
Gestational
exposure; n = 51
13 -week dietary;
« = 3
Gavage GD 15;
n = 5
3 -generation dietary
Gavage GD 14, 21,
PND 7, 14; n = 7
28-day gavage,
n = 5
17-week dietary;
w = 30
Endpoint
Food-reinforced operant behavior in pups
Increased urinary porhyrins
Decreased delayed-type hypersensitivity
response to tuberculin
Cariogenic lesions in pups
Inhibited molar development in pups
Liver and lung lesions
Reduced saccharin consumption and
preference
Maze errors (facilitatory effect)
Decreased sperm concentration and sperm
motility, as adults
Increased TSH in newborn infants
Embryotoxicity
Decreased anogenital distance in male pups
Reduced fertility and neonatal survival (FO
andFl)
Abnormal estrous cycle
Liver lesions
Delay in onset of puberty
NOAELHED(N)or
BMDLHED (B)
(ng/kg-day)

—
6.4E-03C (N)
—
—
—
—
—
—
—
—
2.7E-02 (N)
2.9E-02 (N)
—
3.5E-02 (N)
4.3E-02 (B)
LOAELaED
(ng/kg-day)
5.5E-02
6.4E-02
3.2E-02C
8.9E-02
9.0E-02
1.4E-01
1.7E-01
1.7E-01
2.0E-028
2.0E-021
2.5E-01
1.8E-01
3.8E-01
3.2E-01
5.6E-01
8.9E-02
UFa
300
300
301
300
300
300
300
300
30h
30h
300
301
301
300
301
301
RfD
(mg/kg-day)
1.8E-10
2.1E-10
2.1E-10
3.0E-10
3.0E-10
4.5E-10
5.7E-10
5.7E-10
6.7E-10
6.7E-10
8.4E-10
9.1E-10
9.6E-10
1.1E-09
1.2E-09
1.4E-09

-------
            Table 4-5. Candidate RfDs for TCDD using blood-concentration-based human equivalent doses (continued)
Study
Ishihara et al.,
(2007)
VanBirgelen
etal. (1995a)k
Kociba et al.
(1978)
Fattore et al.
(2000)
Seo et al.
(1995)
Crofton et al.
(2005)
Sewall et al.
(1995)
Franc et al.
(2001)
Kociba et al.
(1976)
Sparschu et al.
(1971)
Alaluusua et al.
(2004)
Species, strain
(sex, if not
both)
Mouse, ICR (M)
Rat, S-D (F)
Rat, S-D (F)
Rat, S-D
Rat, S-D
Rat, Long-Evans
(F)
Rat, S-D (F)
Rat, Long-Evans
(F)
Rat, S-D
Rat, S-D (F)
Human
Protocol
Weekly gavage for
5 weeks; n = 42-43
13 -week dietary;
w=8
2-year dietary;
w = 50
13 -week dietary;
« = 6
Gavage CDs 10-16;
n = 10
4-day gavage;
n = 4-14
30-week gavage;
« = 9
22-week gavage;
w = 8
5 -days/week gavage
for 13 weeks; n = 12
Gavage GD 6-15;
n = 4-129
Childhood exposure;
n =48
Endpoint
Decreased male/female sex ratio
Decreased liver retinyl palmitate
Liver and lung lesions, increased urinary
porhyrins
Decreased hepatic retinol
Decreased serum T4 and thymus weight
Decreased serum T4
Decreased serum T4
Increased relative liver weight; decreased
relative thymus weight
Liver and lung lesions, increased urinary
porphyrins
Decreased fetal body weight
Dental defects
NOAELHED(N)or
BMDLHED (B)
(ng/kg-day)
_i
—
6.3E-02 (N)
—
1.7E-01 (N)
1.7E-01 (N)
5.0E-01 (N)
1.8E-01 (B)
4.5E-01 (N)
2.6E-01 (B)
2.6E-01 (N)
3.2E-01 (N)
4.1E-02'(N)
LOAELaED
(ng/kg-day)
5.0E-01
5.1E-01
6.3E-01
7.8E-01
9.1E-01
7.4E-01
1.7E+00
1.4E+00
3.0E+00
1.7E+00
9.0E-01m
UFa
300
300
301
300
301
301
301
301
301
30'
on
J
RfD
(mg/kg-day)
1.7E-09
1.7E-09
2.1E-09
2.6E-09
5.6E-09
5.6E-09
6.0E-09
8.7E-09
8.7E-09
LIE-OS
1.4E-08
-^
to
     aExcept where indicated, UFA = 3 (for dynamics), UFH = 10, UFL = 10.
     bSchantz and Bowman (1989); Schantz et al. (1986); Schantz et al. (1986).
     °HED determined from lst-orderbody burden model; no PBPK model available for guinea pigs or monkeys; Hochstein et al. (2001) was not presented in the
     table because no PBPK model exists for minks and lst-orderbody burden could not be calculated because a TCDD half-life could not be determined.
     dResults from three separate studies with identical designs combined.
     "Latchoumycandane et al. (2002a; 2002b).
     fUFL = 1 (NOAEL or BMDL).
     gMean of peak exposure (0.0321 ng/kg-day) and average exposure over 10-year critical window (0.0080 ng/kg-day).
     hUFH =  3, UFL = 10.
     'Maternal exposure corresponding to neonatal TSH concentration exceeding 5 uU/mL.
     JThe NOAEL of 4.9E-5 was excluded from consideration because of the large dose spacing in the study.

-------
        Table 4-5. Candidate RfDs for TCDD using blood-concentration-based human equivalent doses (continued)

kVan Birgelen et al. Q995b) is considered to be the same study but reports effects at doses above the LOAEL that are not considered further; this study in not
carried forward for determination of an RfD POD but is included in the RfD uncertainty analysis presented in Section 4.4.
'Mean of peak exposure (0.0655 ng/kg-day) and average exposure over 10-year critical window (0.0156 ng/kg-day).
"Mean of peak exposure (1.65 ng/kg-day) and average exposure over 10-year critical window (0.149 ng/kg-day).
nUFH = 3.
S-D = Sprague-Dawley.

-------
  •9
  O>
I
O C
~ «



1!
  9
  c
  ro
  E
Human

Studies

10 •


1 •


r




0.1 f









001 •


4










r



A
> *
f. 	





	 r 	 r 	

Animal Studies





Less than 1 year | n-year(s) Reproductive Developmental











4


i












,










4









> i










4









t


1



4


)>



i




	 r 	 r 	 r 	 r 	 r 	 r 	 r_.





;
i
L <











— r~



«
• i
4
>
1 1



4
>
\
> ,





\ \
M \

\
\
\
I *i
i


i


1

i
i
f l\
0 |
1
*

1 1
1

1 1
1 1
i

•
I

1








<
>
\ \
\ \
\ \
\ \
\ \




<


<




1
1
<




t\






..r 	 r 	 r_.
J
1


1


1
1

1
M









4









<








4

4
/_i





1
1
1









4
.'







..,..









, <








mmfmm

Dosing Range

a NOAEL
• BMDL







4


v> 1
I!!.'

i *r *









..r....r....r....r....r....r....r...









	











r»»
T
00
o
o
C\l
ro

X



Ol
to
0
0
C\l
ro
(U
ranczak



01
O)
ro

-------
       1 .E-04
               Human
Animal Bioassays
       1.E-05 	
       1.E-06 •
  2 i.
  o
       1.E-11  ••-
       1.E-12 +—r—r-
oo
o
o
CM
"03
S
—
m
03
8
2

oo
o
o
CM
"(D
'5
™
"2
03
8
m

o
°
"ro
'5
03
3
«)
3
_g
<

CD
8
CM
"ro
S
— •



S
o
CM
"(D
S
03
3
IE
^

oo
o
°
"ro
'5
N
o
o
J

01 00 O5
O5 'N. 03
^T O 0)
-— O (-
2 CM i
c ^3 *~
ro a
3 -52
m |
CM
O
O
CM
£
03

08
si
Jj

_Q
CM
oo
O5
Q.
•^




CD
00
CO
ro

QJ
i


r^
1
ro

•—



CD
00
O5
03

o
Q.
03
O
01
Q

O
O
CM
03
S
—
w


8
CM
03
S
' V
(/)
O
^:
ro
2

CM
8
CM
03
01
O
'cT


ro
C55
03

^
5
c


CO
O5
03
01
in
>


CD
o
8
ro

^
1
5

o
8
03
01
c
(D
c
'03
e
03
XL

<§
o
Q_
~z_




o
o
o
CM
03
'5
c
<


O5
C55
03
01
N
1


o
o
CM
03
01

5


8
CM
03
s
Q
6


1^
C55
03
S
>,
|


O
o
CM
03
01
V
1
LJ-

o
o
CM
03
01
^
5


r^
o
o
CM


QJ
QJ
ffl


o
°
03
01
ro
1


03
8
O5
—

QJ
C
01
01
D)
m
c
^
ro
03
01
03
J2
'o
5


o
o
o
CM
03
01

1


l£>
C55
O5
03
S
O
0)


l£>
O
°
03
'5
^
^
O


U5
C55
O5
03
'5
—
03


O
O
CM
03

O
LL


CD
03
01
03
.Q
'o
5


C55
03
S
3
1
03
Q.
W

Dec = decreasing effect; Inc = increasing effect




Figure 4-5. Candidate RfD array.

-------
       As is evident from Table 4-5, very few NOAELs and even fewer BMDLs have been
established for low-dose TCDD studies. BMD modeling was unsuccessful for all of the
endpoints without a NOAEL, primarily because of the lack of dose-response data near the BMR
(see discussion in Section 4.2). Therefore, the RfD assessment rests largely on evaluation of
LOAELs to determine the POD.

 4.3.1.  Toxicological Endpoints
       As can be seen in Table 4-5, a wide array of toxicological endpoints has been observed
following TCDD exposure, ranging from subtle developmental effects to overt toxicity.
Developmental effects in rodents include embryotoxicity, neonatal mortality, dental defects,
delayed puberty in males, and several neurobehavioral effects. Reproductive effects reported in
rodents include altered hormone levels in females and decreased sperm production in males.
Immunotoxicity endpoints, such as decreased response to SRBC challenge in mice and decreased
delayed-type hypersensitivity response in guinea pigs, are also observed. Longer durations of
TCDD exposure in rodents are associated with organ and body weight changes, renal toxicity,
hepatotoxicity, and lung lesions. Adverse effects in human studies are also observed, which
include both male and female reproductive effects, increased TSH in neonates, and dental defects
in children. Other outcomes including diabetes (Michalek and Pavuk, 2008) and hepatic effects
(Michalek et al.,  2001b) have also been associated with  adult human TCDD exposures, but EPA
was unable to quantify the exposure-response relationship (see Appendix C). All but three of the
study/endpoint combinations from animal bioassays listed in Table 4-5 are on TCDD-induced
toxicity observed in mice and rats; the other three study/endpoint combinations are effects in
guinea pigs and monkeys. Although the effects of TCDD also have been investigated in
hamsters and mink, those studies were not included for final POD consideration  because the
effect levels were greater than those in Table 4-5, or because effective oral intakes could not be
estimated.
       Three human studies were also included for final POD consideration in the derivation of
an RfD and are presented in Table 4-5 as candidate RfDs.  All three human study/endpoint
combinations are from studies on the Seveso cohort. The developmental effects  observed in
these studies were associated with TCDD exposures either in utero or in early childhood between
1 and 10 years of age.  Baccarelli et al. (2008) reported increased levels of TSH in newborns

                                          4-46

-------
exposed to TCDD in utero, indicating a possible dysregulation of thyroid hormone metabolism.
Mocarelli et al. (2008) reported decreased sperm concentrations and decreased motile sperm
counts in men who were 1-9 years of age in 1976 at the time of the Seveso accident (initial
TCDD exposure event). Alaluusua et al. (2004) reported dental effects in adults who were less
than 5 years of age at the time of the initial exposure (1976).

 4.3.2.  Exposure Protocols of Points of Depature (PODs)
       The studies in Table 4-5 represent a wide variety of exposure protocols, involving
different methods of administration and exposure patterns across virtually all exposure durations
and life stages. Both dietary and gavage administration have been used in rodent studies, with
gavage being the predominant method. Gavage dosing protocols vary quite widely and include
single gestational exposures, multiple daily exposures (for up to 2 weeks, intermittent schedules
that include  5 days/week, once weekly, or once every 2 weeks), and loading/maintenance dose
protocols, in which a relatively high dose is initially administered followed by lower weekly
doses. The intermittent dosing schedules require dose-averaging over time periods as long as
2 weeks, which introduces uncertainty in the effective exposures. In other words, the high unit
dose may be more of a factor in eliciting the effect than the average TCDD tissue levels over
time. Although the loading/maintenance dose protocols are designed to maintain a constant
internal exposure, these protocols are somewhat inconsistent with the constant daily TCDD
dietary exposures associated with human ingestion patterns.
       The epidemiologic studies  conducted in the Seveso cohort represent exposures over
different life stages including gestation, childhood, and young adulthood.  The Seveso exposure
profile is essentially a high initial pulse TCDD exposure followed by a 20-30 year period of
elimination with only background  exposures to TCDD and DLCs.45  While the exposures were
measured  soon after the initial pulse, health outcomes were realized, or measured,  10-20 years
following the initial exposure; the  biologically-relevant critical exposure window for
susceptibility varies with effect and may  be unknown. Therefore, the effective exposure profiles
for the Seveso cohort studies vary  considerably. For the Mocarelli et al. (2008) and Alaluusua
et al. (2004) studies, where early childhood exposures proximate to the initial event are
associated with the outcomes, there is some uncertainty as to the magnitude of the  effective
 1 In Section 4 the DLC term is exclusive of TCDD.
                                          4-47

-------
doses.  Although the effects are associated with TCDD exposure in the first 10 years of life, it is
not clear to what extent the initial peak exposure is primarily responsible for the effects. It is
also not clear if averaging exposure over the critical window is appropriate given the fairly large
(sixfold) difference between initial TCDD body burden and body burden at the end of the critical
exposure window. Because of the uncertainty in the influence of the peak exposure relative to
the average exposure over the entire window of susceptibility, the LOAELs for both Mocarelli
et al. (2008) and Alaluusua et al. (2004) are calculated as the average of the peak exposure and
average exposure across the critical exposure window (see Section 4.2 for details).
       For the gestational exposure study (Baccarelli et al., 2008), the critical exposure window
is strictly defined and relatively short (9 months) and occurs long after the initial maternal
exposure (18-29 years).46 The maternal serum TCDD concentrations were measured
16-22 years after the initial exposure when internal exposures were falling off less steeply;
consequently, there is less uncertainty in the toxicokinetic extrapolation between time of
measurement and time  of birth. The narrow critical exposure window at a much later time than
the initial exposure (where the TCDD elimination  curve is flattening) is assumed to lead to a
relatively steady-state exposure over the critical time period with much less uncertainty in the
magnitude of the effective dose.  With the exception of Eskenazi et al. (2002b) (see
Section 4.2.4), the effective exposures for other effects reported for the Seveso cohort (see
Section C. 1.1.1.4) have not been quantified for consideration as an RfD POD.  These exposures
and effects are not represented in Table 4-5  because either critical exposure windows cannot be
identified, unequivocal adverse effect levels cannot be determined, or individual exposure
estimates were not reported.  Several of these studies, however, are included in the uncertainty
analysis presented in Section 4.5.

 4.3.3.  Uncertainty Factors
       Based on U.S. EPA (2002), UFs address five areas of uncertainty. Table 4-5  summarizes
the composite (total) UF applied to the POD for each endpoint.
       For the PODs based on animal bioassays, the following UFs were applied:
46 The Sevesso accident occurred on July 10, 1976 and the subjects evaluated in the Baccarelli et al., (2008) study
were born between January 1, 1994 and June 30, 2005.
                                           4-48

-------
•  Inter species extrapolation (UFA).  A factor of 3 (10°'5) was applied for interspecies
   extrapolation. The factor of 3 represents the residual uncertainty for toxicodynamics
   after accounting for toxicokinetic differences with kinetic modeling. Although there are
   in vitro studies (Budinsky et al., 2010; Silkworth et al., 2005) that report higher rodent
   sensitivities than humans for AhR-dependent enzyme induction, EPA believes that there
   is insufficient information on subsequent toxicological processes to conclude that rodents
   are more sensitive than humans for downstream adverse effects.

•  Human interindividual variability (UFn).  A factor of 10 was applied to account for
   human interindividual variability in susceptibility to TCDD because there is insufficient
   information on sensitive populations to justify a lower value.

•   LOAEL-to-NOAEL (UFi).  For all PODs based on the animal bioassay endpoints lacking
   a NOAEL, a factor of 10 was applied to account for LOAEL-to-NOAEL uncertainty.
   The factor of 10 is the standard value in the absence of information suggesting a lower
   value; the magnitude of the effects for most of the LOAELs is relatively high compared
   to controls.

•  Subchronic-to-chronic (UF$).  A UF for study duration was not applied, because chronic
   effects for animal bioassays are well represented in the database.

•  Database factor (UFr>).  A UF  for database deficiencies was not applied because the
   database for TCDD contains an extensive range of human and animal studies that
   examine a comprehensive set of endpoints. There is no evidence to suggest that
   additional data would result in  a lower RfD.
   For the PODs based on epidemiologic studies, the following UFs were applied:
•  UFA. A UF for interspecies extrapolation was not applied because human data were
   utilized for derivation of the RfD.
         A factor of 3 was selected for interindividual variability to account for human-to-
   human variability in susceptibility.  The individuals evaluated in the two principal studies
   included infants (exposed in utero) and adults who were exposed when they were less
   than 10 years of age, groups that are considered to represent sensitive lifestages.  These
   studies considered together associate TCDD exposures with health effects in potentially
   vulnerable lifestage subgroups. A UF of 1 was not applied because the sample sizes for
   the lifestages studied were relatively small, which, combined with uncertainty in
   exposure estimation, may not fully capture the range of interindividual variability.  In
   addition, potential chronic effects were not fully elucidated for humans and could
   possibly be more sensitive.
         A factor of 10 was applied to account for LOAEL-to-NOAEL uncertainty. The
   factor of 10 for UFL is the standard value in the absence of information suggesting a
   lower value.
                                      4-49

-------
   •   UFs. A UF for study duration was not applied, because, although chronic effect levels
       are not well defined for humans, animal bioassays indicate that duration of exposure is
       not likely to be a determining factor in toxicological outcomes. Developmental effects
       and other short-term effects occur at doses similar to effects noted in chronic studies.
           -  A UF for database deficiencies was not applied because the database for TCDD
       contains an extensive range of human and animal studies that examine a comprehensive
       set of endpoints. There is no evidence to suggest that additional data would result in a
       lower RfD.
 4.3.4.  Choice of Human Studies for Reference Dose (RfD) Derivation
       For selection of the POD, the human studies are preferred, as EPA favors human data
over animal data of comparable quality. The human studies included in Table 4-5 (Baccarelli et
al., 2008; Mocarelli et al., 2008; Alaluusua et al., 2004) each evaluate a segment of the Seveso
civilian population (i.e., not an occupational cohort) exposed directly to TCDD released from  an
industrial accident. (The identification of PODs from these studies is detailed in
Sections 4.3.4.1, 4.3.4.2, and 4.3.4.3.)  Thus, exposures were primarily to TCDD, with
apparently minimal DLC exposures beyond those associated with background intake,47
qualifying these studies for use in RfD derivation for TCDD. In addition, health effects
associated with TCDD exposures were observed in humans, eliminating the uncertainty
associated with interspecies extrapolation.  The cohort members who were evaluated included
infants (exposed in utero) and adults who were exposed when they were less than 10 years of
age.  These studies considered together associate TCDD exposures with health effects in
potentially vulnerable lifestages.  Finally, the two virtually identical RfDs from different
endpoints in different studies provide an additional level of confidence in the use of these data
for derivation of the RfD for TCDD.
       Although the human data are preferred, Table 4-5 presents a number of animal studies
with RfDs that are lower than the human RfDs. Two of the rat bioassays among this group of
studies—Bell et al. (2007b) (RfD = 1.4E-9 mg/kg-day based on delay in the onset of puberty)
and NTP (2006a) (RfD = 4.5E-10 mg/kg-day based on liver and lung lesions)—are of particular
note.  Both studies were recently conducted. Both were very well designed and conducted, using
30 or more animals per dose group (see Table 4-6 for a discussion of these studies' strengths and
47 As an example, note the lack of statistically significant effects reported by Baccarelli et al. (2008) (Figures 2C and
D) in regression models based on either maternal plasma levels of noncoplanar PCBs or total TEQ on neonatal TSH
levels.
                                          4-50

-------
weaknesses); both also are consistent with and, in part, have helped to define the current state of
practice in the field. Bell et al. (2007b) evaluated several reproductive and developmental
endpoints, initiating TCDD exposures well before mating and continuing through gestation.
NTP (2006a) is the most comprehensive evaluation of TCDD chronic toxicity in rodents to date,
evaluating dozens of endpoints at several time points in all major tissues.  Thus, proximity of the
RfDs derived from these two recent high-quality studies provides additional support for the use
of the human data for RfD derivation.
       There are several animal bioassay candidate RfDs at the lower end of the RfD range in
Table 4-5 that are more than 10-fold below the human-based RfDs. Two of these studies report
effects that are analogous to the endpoints reported in the three human studies and support the
RfDs based on human data.  Specifically, decreased sperm production in Latchoumycandane and
Mathur (2002) is consistent with the decreased sperm counts and other sperm effects in
Baccarelli et al. (2008). and missing molars in Keller et al. (2008a: 2008b: 2007) are similar to
the dental defects seen in Alaluusua et al. (2004). Thus, because these endpoints have been
associated with TCDD exposures in humans, these animal studies were not selected for RfD
derivation in preference to human data showing the same effects.
       Another characteristic of the remaining studies in the lower end of the candidate RfD
distribution is that they are dominated by mouse studies (comprising 7 of the 9 lowest candidate
RfDs). EPA has less confidence in the candidate RfD estimates based on mouse data than those
based on either the rat or human data.  EPA has less confidence in the use of the Emond mouse
PBPK model to estimate the PODs because of the lack of key mouse-specific data,  particularly
for the gestational component (see Section 3.3.4.3.2.5).  The toxicokinetic interspecies
extrapolation factors used for mice are very large, introducing a potential for large errors. The
ratio of administered dose to HED (Da:HED) ranges from 65 to 1,227 depending on the duration
of exposure. The Da:HED for mice is, on average, about four times larger than that used for rats.
                                          4-51

-------
             Table 4-6. Qualitative analysis of the strengths and limitations/uncertainties associated with animal bioassays
             providing PODs for the TCDD RfD
           Study
                   Strengths
                 Limitations
                                                                                                                            Remarks
                      • Large sample size of both rat dams and
                        offspring/dose employed
                      • Several developmental effects tested
                                                 Batch-to-batch variation of up to 30% in TCDD
                                                 concentration in the diet
                                                 Longer-term dosing of dams does not accurately
                                                 define gestational period when fetus is especially
                                                 sensitive to TCDD-induced toxicity
                                                Study is a significant addition to
                                                a substantial database on the
                                                developmental toxicity of
                                                TCDD in laboratory animals
     Cantoni et al.
     (1981)
•  Experiments were designed to test qualitative and
   quantitative composition and the course of
   urinary excretion in TCDD-induced porphyria
Small sample size of rats/dose employed (n = 4)
Concurrent histological changes with tissue porphyrin
levels were not examined
TCDD used for dosing was of unknown purity
                                                                                                                  Early study on porphyrogenic
                                                                                                                  effects of TCDD
     DeCaprio et al.
     (19861
•  Subchronic oral dosing duration up to 90 days
•  Male and female guinea pigs tested
Relatively small sample size of guinea pigs/dose
employed (n = 10)
No histopathological analyses performed
TCDD used for dosing was of unknown purity
                                                                                                                  Limited subchronic study;
                                                                                                                  PBPK model not available for
                                                                                                                  estimation of HED
-^
to
     Franc et al.
   Three different rat strains with varying
   sensitivities to TCDD were utilized (Sprague-
   Dawley, Long Evans, Han/Wistar)
   Longer-term oral dosing up to 22 weeks
Relatively small sample size of rats/dose employed («|Limited subchronic study
= 8)
Only female rats were tested
Concurrent liver histopathological changes with liver-
weight changes were not examined
Gavage exposure was only biweekly	
Rojo et al. (2002)
                        Low TCDD dose levels used allowed for subtle
                        behavioral deficits to be identified in rat offspring
                        Preliminary training sessions in operant chamber
                        apparatuses were extensive
                        Neurobehavioral effects are exposure-related and
                        cannot be attributed to presence of learning or
                        discrimination deficits
                                                 Relatively small sample size of rat dams/dose
                                                 employed (n = 12)
                                                 Small sample size of rat offspring/dose evaluated (n =
                                                 5-6)
                                                 Neurobehavioral effects induced by TCDD at earlier
                                                 or later gestational dosing dates are unknown because
                                                 of single gavage administration on GD 8
                                                 Although BMD analysis was conducted, the model
                                                 parameters were not constrained according to EPA
                                                 guidance, so the results cannot be used	
                                                One of a few neurobehavioral
                                                toxicity studies; somewhat
                                                limited study size

-------
        Table 4-6. Qualitative analysis of the strengths and limitations/uncertainties associated with animal bioassays
        possessing candidate PODs for the TCDD RfD (continued)
      Study
                   Strengths
                 Limitations
          Remarks
Keller et al.
(2QQ8a; 2QQ8b;
20071
   Six different inbred mouse strains were utilized
   Large sample size of mouse offspring/dose/strain
   evaluated
   Low TCDD dose levels used compared to typical
   mouse studies allowed for identification of subtle
   sensitivity differences in presence of absence of
   third molars, variant molar morphology, and
   mandible structure in offspring
Unknown sample size of mouse dams/dose/strain
employed
All inbred strains possessed sensitive b allele at the
Ahr locus (i.e., a potentially resistant subpopulation
was not evaluated for comparison purposes)
Morphological dental and mandibular changes
induced by TCDD at earlier or later gestational
dosing dates are unknown because of single gavage
administration on GD 13
Difficulties breeding A/J mice led to abandonment of
that strain in the analysis (Keller et al.. 2008a: 2008b)
Endpoint similar to effects
observed at higher exposure
levels in humans; HED highly
uncertain using mouse PBPK
model
Latchoumy-
candane and
Mathur (2002)
•  Compared to epididymal sperm counts, the
   testicular spermatid head count provides better
   quantitation of acute changes in sperm production
   and can indicate pathology
Small sample size of rats/dose employed (n = 6)
Oral pipette administration of TCDD may be a less
efficient dosing method than gavage
Endpoint has human relevance,
similar to critical effects in
principal human study for RfD
                   Female reproductive effects (i.e., early embryo
                   loss and changes in serum progesterone and
                   estradiol) were tested at multiple exposure
                   times—early gestation, preimplantation, and peri-
                   to postimplantation	
                                                 Small sample size of dams/dose (n = 10)
                                                 Large dose-spacing interval (25-fold at lowest
                                                 2 doses)
                                                Endpoint has human relevance
                                                but HED highly uncertain using
                                                mouse PBPK model
Markowski et al.
2QQD
   Low TCDD dose levels used allowed for subtle
   behavioral deficits to be identified in rat offspring
   Several training sessions on wheel apparatuses
   were extensive
   Neurobehavioral effects are exposure-related and
   cannot be attributed to motor or sensory deficits
Unknown sample size of rat dams/dose employed
Small sample size of rat offspring/dose evaluated (n =
4-7)
TCDD used for dosing was of unknown purity and
origin
Only two treatment levels
Neurobehavioral effects induced by TCDD at earlier
or later gestational dosing dates are unknown because
of single gavage administration on GD 18	
One of a few neurobehavioral
toxicity studies; somewhat
limited study size
                   Large sample size of mice and rats/dose
                   employed
                   Comprehensive 2-year bioassay that assessed
                   body weights, clinical signs, and pathological
                   changes in multiple tissues and organs	
                                                 Elevated background levels of hepatocellular tumors
                                                 in untreated male mice
                                                 Gavage exposure was only 2 days/week
                                                 Only two treatment levels
                                                Comprehensive chronic toxicity
                                                valuations of TCDD in
                                                rodents; HED highly uncertain
                                                using mouse PBPK model

-------
        Table 4-6.  Qualitative analysis of the strengths and limitations/uncertainties associated with animal bioassays
        possessing candidate PODs for the TCDD RfD (continued)
     Study
                   Strengths
                 Limitations
          Remarks
                   Chronic exposure duration with several interim
                   sacrifices
                   Large number of dose groups with close spacing
                   Large number of animals per dose group
                   Comprehensive suite of endpoints evaluated
                   Comprehensive biochemical, clinical, and
                   histopathological tests and measures
                   Detailed reporting of results, with individual
                   animal data presented as well as group summaries
                                                 Single species, strain, and sex
                                                 Lowest dose tested too high for establishing NOAEL
                                               Study is the most
                                               comprehensive chronic TCDD
                                               toxicity evaluation in rats to
                                               date
Shi et al.
   Study design evaluated TCDD effects on aging
   female reproductive system (i.e., exposure began
   in utero and spanned across reproductive
   lifespan)
   Several female reproductive endpoints were
   evaluated, including cyclicity, endocrinology,
   serum hormone levels, and follicular reserves
Relatively small sample size of rats/dose employed
(n = 10)
Endpoint similar to effects
observed at higher exposure
levels in humans
Smialowicz et al.
 2008)
   SRBC plaque forming cell assay is highly
   sensitive and reproducible across laboratories
   when examining TCDD
Small sample size of animals/dose (n = 8)
Only female mice were tested
Thymus and spleen weights were only other immune
response-related endpoints tested	
Limited immunotoxicity study
Toth et al.
   Large sample size of mice/dose employed
   Chronic exposure duration
Reporting of findings is terse and lacks sufficient
detail (e.g., materials and methods, thorough
description of pathological findings, etc.)
Limited number of endpoints examined
Only male mice were tested
Limited chronic study; HED
highly uncertain using mouse
PBPK model
Vos et al. (1973)
•  Three different animal species tested (guinea
   pigs, mice, and rats)
•  Effects of TCDD tested on both cell-mediated
   and humoral immunity
Small sample size of animals/dose employed in each
experiment (n = 5-10)
Only female guinea pigs and rats were tested, and
only male mice were tested
Only one experimental assay was utilized to assess
cell-mediated or humoral immunity; humoral
immunity was only investigated in guinea pigs
TCDD used for dosing was of unknown purity	
Endpoints relevant to humans
but study size limited; PBPK
model not available for
estimation of HED

-------
Table 4-6. Qualitative analysis of the strengths and limitations/uncertainties associated with animal bioassays
possessing candidate PODs for the TCDD RfD (continued)
Study
White et al. (19861
Strengths
• Total hemolytic complement (CH50) is
representative functional assay of the complete
complement sequence
Limitations
• Small sample size of rats/dose employed (n = 6-8)
• Individual complement factors may be significantly
depleted without affecting CH50 activity (only C3 is
measured)
• TCDD used for dosing was of unknown purity
Remarks
Endpoint similar to effects
observed at higher exposure
levels in humans; HED highly
uncertain using mouse PBPK
model

-------
In addition, each one of the mouse studies has other qualitative limitations and uncertainties
(discussed above and in Table 4-6) that further reduce confidence in using them as the basis for
theRfD.

4.3.4.1. Identification of Point of Departure (POD) from Baccarelli et al. (2008)
       Baccarelli et al. (2008) reported increased levels of TSH in newborns exposed to TCDD
in utero, indicating a possible dysregulation of thyroid hormone metabolism. The study authors
related TCDD concentrations in maternal plasma to neonatal TSH levels using a multivariate
linear regression model adjusting for a number of covariates (gender, birth weight, birth order,
maternal age, hospital, and type of delivery). Based on this regression modeling, EPA has
defined the LOAEL for Baccarelli et al. (2008) to be the maternal TCDD LASC of 235 ppt (at
delivery) corresponding to a neonatal TSH level of 5 |iU/mL.
       The WHO (1994) established the 5 jiU/mL standard as an indicator of potential iodine
deficiency and potential thyroid problems in neonates. Increased TSH levels are indicative of
decreased thyroid hormone (T4 and/or T3) levels. The 5 |iU/mL limit for TSH measurements in
neonates was recommended by WHO (1994) for use in population surveillance programs as an
indicator of iodine deficiency disease (IDD). In explaining this recommendation, WHO (1994)
stated that
       While further study of iodine replete populations is needed, a limit of 5|iU/ml
       whole blood... may be appropriate for epidemiological studies of IDD [iodine
       deficiency disease.]  Populations with a substantial number of newborns with
       TSH levels above the limit could indicate a significant IDD problem.
       For TCDD, the toxicological concern is not likely to be iodine uptake inhibition, but
rather increased metabolism and clearance of T4, as evidenced in a number of animal studies
(see discussion in Section 4.3.6.1).  Baccarelli et al. (2008) discount iodine status in the
population as a confounder, as exposed and referent populations all lived in a relatively small
geographical area.  It is unlikely that there was iodine deficiency in one population and not in the
other population based on iodine levels in the soil.
       Clinically, a TSH level of >4 |iU/mL in a pregnant woman is followed up by an
assessment of free T4, and treatment with L-thyroxine is prescribed if T4 levels are low (Glinoer
                                          4-56

-------
and Delange, 2000). This is to ensure a sufficient supply of T4 for the fetus, which relies on
maternal T4 exclusively during the 1st half of pregnancy (Chan et al., 2005): (Calvo et al., 2002;
Morreale de Escobar et al., 2000). Adequate levels of thyroid hormone also are essential in the
newborn and young infant as this is a period of active brain development (Zoeller and Rovet
2004; Glinoer and Delange, 2000). Smaller reserves, higher demand, and shorter half-life of
thyroid hormones in newborns and young infants also could make this lifestage more susceptible
to the impact of insufficient levels of T4 (Savin et al., 2003; Greer et al., 2002; Van Den Hove et
al.,  1999). Thyroid hormone disruption during pregnancy and in the neonatal period can lead to
neurological deficiencies, particularly in the attention and memory domains (Oerbeck et al.,
2005). While such altered hormone levels are associated with decreased intelligence quotient
(IQ) scores (e.g., 2009) report such associations among adolescents), the exact relationship
between TSH increases and adverse neurodevelopmental outcome is not well defined. A TSH
level above 20 uU/L in a newborn infant is cause for immediate intervention to prevent mental
retardation, often caused by a malformed or ectopic thyroid gland in the newborn (WHO, 2007;
Rovet 2002; Glinoer and Delange, 2000). Recent epidemiologic data indicate concern  for even
lower level thyroid hormone perturbations during pregnancy.  For example, Haddow et  al. (1999)
reported that women with subclinical hypothyroidism, with a mean TSH of 13.2 uU/L had
children with IQ deficits of up to 4 IQ points on the Wechsler IQ scale. Neonatal TSH within the
first 72 hours of birth [as was evaluated by Baccarelli et al. (2008)1 is a sensitive indicator of
both neonatal and maternal thyroid status (Delange etal., 1983). Animal models have recently
indicated that very modest perturbations in thyroid status for even a relatively short period of
time can lead to altered brain development (Sharlin et al., 2010; Royland et al., 2008; Sharlin et
al., 2008; Auso et al., 2004; Lavado-Autric et al., 2003).  Rodent bioassay results also suggest
that elevated TSH levels in neonates can affect sperm development as adults (Anbalagan et al.,
2010): this study also reported reduced fertility among adult males and females with increased
neonatal TSH levels.
      EPA has defined the LOAEL for Baccarelli et al. (2008) to be the maternal  TCDD LASC
of 235 ppt corresponding to a neonatal TSH level of 5 |iU/mL, determined by the regression
modeling performed by the study authors. Using the Emond human PBPK model, the daily oral
intake at the LOAEL is estimated to be 0.020 ng/kg-day (see Section 4.2.3.1).  A NOAEL is not
                                          4-57

-------
defined because it is not clear what maternal intake should be assigned to the group below
5 |iU/mL.

4.3.4.2. Identification of Point of Departure (POD) from Mocaretti et al (2008)
       Mocarelli et al. (2008) reported decreased sperm concentrations and decreased motile
sperm counts in men who were 1-9 years old in 1976 at the time of the Seveso accident (initial
TCDD exposure event).  The sperm concentrations and motile sperm counts of men who were
10-17 years old in 1976 were not decreased. Serum (LASC) TCDD levels were measured in
samples collected within 1 year of the initial exposure.  Serum TCDD levels and corresponding
responses were reported by quartile, with a reference group of less-exposed individuals assigned
a TCDD LASC value of 15 ppt (which was the mean of the TCDD LASC reported in individuals
outside the contaminated area). In the reference group, mean sperm concentrations and percent
motile sperm counts were approximately 73 million sperm/mL and 41%, respectively. The
lowest exposed group (lst-quartile) TCDD LASC median was 68 ppt. In the 1st quartile,  mean
                                                       4R
sperm concentrations of approximately 55 million sperm/mL and motile sperm counts of
approximately 36% were reduced about 24 and 12%, respectively, from the reference group.
Further decrease in these measures in the groups exposed to more than 68  ppt was minimal.
Relative to the reference population, the percent decreases in sperm concentrations were
approximately 25, 21, and 33% in the 2nd, 3rd,  and 4th quartiles, respectively, and the percent
decreases in progressive sperm motility were approximately 20,  25, and 22% in the 2n , 3r , and
4* quartiles, respectively.
       Mocarelli et al. (2008) also conducted a separate analysis of all the 22-31 year-old men
(combining all quartiles of the men exposed when they were 1-9 years of age).  In the exposed
men, the mean total sperm concentration was reported by Mocarelli et al. (2008) to be
53.6 million/mL, with a value of 21.8 million/mL at 1  standard deviation below the mean.  In the
comparison group that consisted of men not exposed to TCDD by the Seveso explosion and of
the same age as the exposed men, the mean total sperm concentration was 72.5 million/mL
(31.7 million/mL at 1 standard deviation below the mean).
       There is no clear adverse effect level indicating male fertility problems for either  of these
sperm effects.  As sperm concentration decreases, the probability of pregnancy from a single
48 This estimate is based on Figure 3 in Mocarelli et al. (2008)
                                         4-58

-------
ejaculation also decreases; infertile conditions arise when the number of normal sperm per
ejaculate is consistently and sufficiently low. Previously, the incidence of male infertility was
considered increased at sperm concentrations less than 20 million sperm/mL (WHO, 1980).
More recently, Cooper et al. (2010) suggested that the 5th percentile for sperm concentration
(15 million/mL) could be used as a limit by clinicians to indicate needed follow-up for potential
infertility.  Skakkeback (2010) suggests the following two limits for human sperm
concentrations:  15 million sperm/mL, based on Cooper et al. (2010) and 40 million sperm/mL.
Skakkeback justifies the  upper level of 40 million sperm/mL citing a study by Bonde et al.
(1998) of couples planning to become pregnant for the first time; in the Bonde study, pregnancy
rates declined when sperm concentrations were below 40 million sperm/mL. Skakkeback
suggests that 15 million sperm/mL may be too low of a limit off for normal fertility and that
sperm concentrations between 15 million sperm/mL and 40 million sperm/mL may indicate a
range of reduced fertility. For fertile men, between 50% and 60% of sperm are motile  (Swan et
al., 2003; Slama et al., 2002; Wijchman et al., 2001). Any impacts on these reported levels could
become functionally significant, leading to reduced fertility. Low sperm counts are typically
accompanied by poor sperm quality with respect to morphology and motility (Slama et al.,
2002).
       EPA judged  that the impact on sperm concentration and quality reported by Mocarelli
et al. (2008) is biologically significant given the potential for functional impairment. Although a
decrease in sperm concentration of 25% likely would not have clinical significance for a typical
individual, EPA's concern with the reported decreases in sperm concentration and total number
of motile sperm (relative to the comparison group) is that such decreases associated with  TCDD
exposures could lead to shifts in the distributions of these measures in the general population.
Because male fertility is  susceptible to reductions in both the number and quality of  sperm
produced, such shifts in the population could result in decreased fertility in men at the low ends
of these population distributions. Further, in the group exposed due to the Seveso  accident,
individuals 1 standard deviation below the mean had sperm concentrations of 21.8 million/mL;
this concentration falls near the low end of the range of reduced fertility (15 million  and
40 million sperm/mL) suggested by (Skakkebaek, 2010); the corresponding concentration of
31.7 million/mL for the comparison group at one standard deviation below the mean is slightly
more than twice the  lower end of that range.
                                          4-59

-------
       EPA has designated the lowest exposure group (68 ppt) as a LOAEL, which translates to
a continuous daily oral intake of 0.020 ng/kg-day (see Section 4.2.3.2).  The reference group is
not designated as a NOAEL because the serum levels were not measured for this group, directly,
and background exposures to DLCs are relatively large by comparison to TCDD in this group,
introducing too much uncertainty in quantifying the full NOAEL exposure (see discussion in
Section 4.5). Also, there is no clear zero-exposure measurement for any of these endpoints,
complicating the interpretation of the reference group response as a true "control" response (see
discussion in Section 4.4).  However, males less than 10 years old can be designated as a
sensitive lifestage as compared to older males who were not affected.

4.3.4.3. Identification of Point of Departure (POD) from Alaluusua et al. (2004)
       Alaluusua et al. (2004) reported dental enamel defects and missing permanent teeth in
male and female adults who were less than 5 years of age, but not older, at the time of the initial
exposure (1976) in Seveso. EPA used the same approach to estimate daily TCDD intake as was
used for the  Mocarelli et al. (2008) data; a window of susceptibility of about 5 years was
established.  Serum measurements for this cohort were taken within a year of the accident.
Serum TCDD levels and corresponding responses were  reported by tertile, with a reference
group of less-exposed individuals assigned a TCDD LASC value of 15 ppt (ng/kg); the tertile
group means were 130, 383, and 1,830 ppt.  Both a NOAEL and LOAEL can be defined for this
study.  The NOAEL is 0.12 ng/kg-day, corresponding to the TCDD LASC of 130 ppt at the first
tertile. The LOAEL is 0.93 ng/kg-day at the second tertile. The children in this cohort less than
5 years old can be designated as a sensitive lifestage  as compared to older individuals who were
not affected  relative to the reference group.

 4.3.5.  Derivation of the  Reference Dose (RfD)
       The two human studies, Baccarelli et al.  (2008) and Mocarelli et al. (2008), have identical
LOAELs of 0.020 ng/kg-day.  Together, these two studies define the most sensitive health
effects in the epidemiologic literature and constitute the best foundation for establishing a POD
for the RfD, and are designated as coprincipal studies.  Therefore, increased neonatal TSH levels
in Baccarelli et al. (2008) and male reproductive effects (decreased sperm count and motility) in
Mocarelli et al. (2008) are designated as cocritical effects. A  composite UF of 30 is applied to

                                          4-60

-------
the LOAEL of 0.020 ng/kg-day to account for lack of a NOAEL (UFL = 10) and human
interindividual variability (UFn = 3); the resulting RfD in standard units is 7 x 1CT10 mg/kg-day.
Table 4-7 presents the details of the RfD derivation.

 4.3.6.  Studies Reporting Outcomes Comparable to the Principal Studies Used to Derive
        the Reference Dose (RfD)
       Other animal and human epidemiologic studies report associations between TCDD
exposures and effects similar to those reported by Baccarelli et al. (2008) and Mocarelli et al.
(2008).
4.3.6.1. Dysregulation of Thyroid Hormone Metabolism Associated with Dioxin Exposure in
        Neonates
       One of the principal studies for the dioxin noncancer RfD, Baccarelli et al. (2008),
reported increased levels of TSH in newborns exposed to TCDD in utero, indicating a possible
dysregulation of thyroid hormone metabolism. No other human studies that met the selection
criteria of this analysis reported similar effects.
       However, based on an analysis of over 20 epidemiology studies, Goodman et al. (2010)
concluded that DLC exposures were not clearly or consistently correlated with differences in
thyroid hormone levels in neonates and children less than 12 years of age. Focusing on neonatal
TSH for direct comparison to Baccarelli et al. (2008), Goodman et al. (2010), in Table 3 of their
analysis, identify 13 different studies, including Baccarelli et al. (2008), which measured  infant
TSH levels within 1 week of birth.  Of these studies, only Baccarelli et al. (2008) was
TCDD-specific and evaluated exposures well above ambient exposure levels. The other studies
examined total TEQ or individual DLCs near background exposure levels. The LOAEL derived
by EPA from Baccarelli et al. (2008) is approximately sixfold  higher than the ambient total TEQ
exposure levels at the time of the exposures for the general Seveso population49 and more than
30-fold above an estimate of current TEQ levels (Lorber et al., 2009). In the other studies, the
exposures appear to have been largely to DLCs, with TCDD as a minor component. Because the
equivalent TCDD exposure for DLCs is derived from TEF methodology, which is conservative
in nature (TEFs  are higher than the median), the total TEQ concentrations would likely be over-
estimated (relative to TCDD) and uncertain. In addition, only 2 of the other  12 studies evaluated
  Estimated byEPAtobe3.5 * 10 3 ng/kg-day on a total TEQ basis (see Section 4.5.1.1.1 and Appendix F).
                                         4-61

-------
Table 4-7. Basis and derivation of the TCDD RfD
Principal study detail
Study POD (ng/kg-day) Critical effects
Mocarelli et al. (2008) 0.020 (LOAEL)
Baccarelli et al. (2008) 0.020 (LOAEL)

Decreased sperm count (20%) and motility (1 1%) in
men exposed to TCDD during childhood
Elevated TSH (>5 jiU/mL) in neonates
RfD derivation
POD 0.020 ng/kg-day (2.0E-8 mg/kg-day)
UF 30(UFL=10, UFH =
3)
RfD 7 x 1Q~1U (7E-10) mg/kg-day (2.0E-8 - 30)
Uncertainty factors
LOAEL-to-NOAEL
(UFL)
rluman interindividual variability
(UFH)
Interspecies extrapolation
(UFA)
Subchronic-to-chronic
(UFs)
Database sufficiency
(UFD)
10
3
1
1
1
No NOAEL established; cannot quantify lower exposure
group in Baccarelli et al. (2008): magnitude of effects at
LOAEL sufficient to require a 10-fold factor.
A factor of 3 (10°'5) is used because the effects were
elicited in sensitive lifestages. A further reduction to 1
was not made because the sample sizes were relatively
small, which, combined with uncertainty in exposure
estimation, may not fully capture the range of
interindividual variability. In addition, chronic effects
are levels are not fully elucidated for humans and could
possibly be more sensitive.
Human study.
Chronic effect levels are not well defined for humans;
however, animal bioassays indicate that duration of
exposure does not seem to be a determining factor in
toxicological outcomes. Developmental effects and other
short-term effects occur at doses similar to effects noted
in chronic studies. Considering that exposure in the
principal studies encompasses the critical window of
susceptibility associated with development, a UF to
account for exposure duration is not used.
The database for TCDD contains an extensive range of
human and animal studies that examine a comprehensive
set of endpoints. There is no evidence to suggest that
additional data would result in a lower RfD.
                                4-62

-------
by Goodman et al. (2010) reported TSH measures 3 days after birth, which is an international
standard and would be most comparable to those in Baccarelli et al. (2008).  TSH levels
generally peak about 2 hours after birth then decline rapidly to typical long-term levels over the
next few days (Steinmaus et al., 2010). Several of the studies included in Table 3 of Goodman et
al. (2010) evaluated cord-blood TSH measurements, which represent early high TSH
concentrations and are not directly comparable to 3-day measurements.  Given these
considerations, particularly the relatively low ambient exposures and differences in the timing of
TSH measures,  it would be unlikely that any consistent pattern would be detected across these
studies.
       Several animal studies that met the selection criteria evaluated the effects of TCDD on
the thyroid or thyroid hormone levels.  Overall, this set of studies show that TCDD affects
thyroid hormone levels and the thyroid gland.  The studies of Sewall et al. (1995), Seo et al.
(1995). Van Birgelen et al. (1995a: 1995b), Crofton et al. (2005). and NTP (2006a) each reported
decreases in T4 levels.  In response to TCDD treatment, NTP (2006a) reported increases in total
T3 concentrations, and both NTP (2006a) and Sewall et al. (1995) reported increased TSH
concentrations.  Sewall et al. (1995) and Chu et al.  (2007) reported reductions in thyroid
follicles, with Chu et al. (2007) noting that, of the health effects observed in their study, thyroid
effects were the most sensitive to TCDD exposures. Although none of these studies address in
utero or neonatal exposure, they show that TCDD can affect the level of thyroid hormones and
the thyroid organ in adult animals.

4.3.6.2. Male Reproductive Effects associated with Dioxin Exposures
       The other principal study for the dioxin noncancer RfD, Mocarelli et al. (2008), reported
decreased sperm concentrations and decreased motile sperm counts in men who were aged
1-9 years at the time of the Seveso accident (initial TCDD exposure event).  The sperm
concentrations and motile sperm counts of men who were 10-17 years old in 1976 were not
adversely affected. While no other human studies that met the selection criteria of this analysis
reported similar effects, a newly published study, Mocarelli et al.  (2011), also reports male
reproductive effects.  Several animal studies that met the study selection criteria also reported
male reproductive effects.
                                          4-63

-------
       Mocarelli et al. (2011) examined the relationship between maternal serum TCDD levels
and semen quality in male offspring. Analyses were based on 39 of the 78 men aged
18-26 years born to women residing in the areas most heavily polluted by dioxin after the
explosion in Seveso, Italy, in 1976 and age-matched controls (58 out of 123 recruited) born to
women residing in noncontaminated areas of Italy.  In the exposed group of women, pregnancies
occurred between 9 months and 6 years after the accident (March 1977-January 1984). The
male offspring of these women were categorized based on whether they were breastfed (n = 21,
born to 20 mothers) or formula-fed (n = 18, born to 17 mothers) as infants. In the comparison
group, 36 were breastfed, and 22 were formula-fed.  Sons born to dioxin-exposed women whose
spouses were also exposed to TCDD, as well as all men with reported diseases,  were excluded.
       TCDD exposures were based on estimated maternal serum concentration at conception.
To estimate these levels in the exposed group, the authors relied on maternal serum measures,
all of which were collected shortly after the accident in 1976-1977, and a biokinetic model
(Kreuzer et al.,  1997) that estimated TCDD elimination from the time of the accident to
conception for individual women (average half-life = 4 years).  Mothers of sons in the
comparison group were assumed to be exposed to average background TCDD levels of 10  ppt
based on measurements reported in Eskenazi et al. (2004).
       Semen samples were collected from all participants. These samples were maintained at
37°C and examined within an hour of ejaculation.  For serum inhibin B and follicle stimulating
hormone (FSH) analyses, fasting blood samples were obtained the morning of semen collection.
Statistical analyses were performed on sperm properties,  serum hormone levels, and TCDD
levels using a "general linear model" (Mocarelli et al., 2011).  Model covariates included age,
duration of abstinence prior to semen collection, smoking status, exposures to organic  solvents,
adhesives or paints, BMI, alcohol use, educational level, and employment status.
       Relative to the comparison group, men born to exposed mothers had decreased sperm
concentration (46 million vs. 81 million sperm/mL;/> = 0.01), total sperm count (144 million vs.
231 million sperm; p = 0.03), and total number of motile  sperm (51 million vs. 91 million;
p = 0.05). Relative to the breastfed comparison group, breastfed sons born to exposed mothers
exhibited decreased sperm concentrations (36 million vs.  86 million sperm/mL;/? = 0.002), total
sperm  counts (117 million vs. 231 million sperm; p = 0.02), and motile sperm counts (39 million
vs. 98 million; p = 0.01).  Relative to the breastfed comparison group, breastfed sons born to
                                          4-64

-------
exposed mothers also exhibited increased FSH concentrations (4.1 vs. 2.6 IU/L;/> = 0.03) and
decreased inhibin B levels (70.2 million vs. 101.8 pg/mL;p = 0.01).  The formula-fed exposed
and comparison groups were not significantly different by any of these measures.
       This study was well-designed with well-characterized exposures (for the exposed group),
which relied on measured sera TCDD concentrations and a peer-reviewed TCDD elimination
model to estimate maternal serum TCDD levels at the time of conception.  Exposures in the
comparison group relied on estimates from other studies.  The study excluded sons of fathers that
were likely highly exposed to TCDD, to limit potential influences from highly exposed fathers.
The study relies on self-reported recollection of infant feeding (i.e., breastfed vs. formula-fed),
which may lead to some misclassification based on recall error.  Statistically significant
associations were evident for both the exposed men and their comparison group and breastfed
men and the breastfed comparison group.
       In this study, elevated TCDD exposures during and after pregnancy (via breast-feeding)
led to long-term decrements in male reproductive endpoints. These effects included changes in
levels of hormones that affect spermatogenesis; they also include decreases in sperm
concentration and sperm motility.
       In addition, two rodent bioassays also report sperm effects associated with dioxin
treatment. Latchoumycandane and Mathur (2002) reported decreased daily sperm production
and decreased reproductive organ weights in male albino Wistar rats given daily oral doses of
TCDD for 45 days.  The LOAEL was 1.0 ng/kg-day,  which corresponds to a LOAELHED of
0.016 ng/kg-day (see Table 4-5); a NOAEL was not identified. Simanainen et al. (2004)
reported a reduction in daily sperm production and cauda epididymal sperm reserves in male rat
pups born to dams exposed to 300 ng/kg TCDD or higher on GD 15 by oral gavage.  In this case
a NOAEL of 100 ng/kg was identified, which corresponds to a NO ABIDED of 0.426 ng/kg-day,
with a LOAELnED of 1.7 ng/kg-day (see Table 4-3). Detailed descriptions of these studies can
be found in Appendix D.

4.4.  QUALITATIVE UNCERTAINTIES IN THE REFERENCE DOSE (RfD)
       Exposure assessment is a key limitation of the epidemiologic studies (of the Seveso
cohort) used to derive the RfD.  The Seveso cohort exposure profile consists of an initial high
                                         4-65

-------
TCDD exposure50 followed by a drop in body burden to background levels over a period of
about 20 years, at which time the effects were observed. This exposure scenario is inconsistent
with the constant daily intake scenario addressed by the RfD methodology. The determination of
an effective average daily dose from the Seveso exposure scenario requires a consideration of the
biologically-relevant critical time-window of susceptibility and the influence of the peak
exposure on the occurrence of the observed effects, particularly when the peak exposure is high
relative to the average exposure over the critical exposure window (see Text Box 2-2). For one
of the principal studies (Mocarelli et al., 2008), a maximum susceptibility exposure window can
be identified based on the age of the population at risk.  However, the influence of the peak
exposure on the effects observed 20 years later is unknown, and the biological significance of
averaging the exposure over several years, with internal exposure measures spanning a 5.5-fold
range, is unknown. EPA has not developed guidance for large interval averaging.  Furthermore,
because there is an assumption of a threshold level of exposure below which noncancer effects
are not expected to occur, averaging over large intervals could include exposures that are below a
threshold.  The process used by EPA to estimate the LOAEL exposure for the Mocarelli et al.
(2008) study  is a compromise between the most- and least-conservative alternatives; as such,
there is some uncertainty in the estimate, perhaps in the range of 3- to 10-fold in either direction.
This uncertainty also applies to the LOAEL determined for the developmental dental effects
reported in Alaluusua et al. (2004) and the increased menstrual cycle length reported in Eskenazi
et al. (2002b) (see Section 4.2.3.4); in both of those  studies, the uncertainty is greater,  as the
difference between peak and average internal exposures is an order of magnitude or more. The
LOAEL for increased TSH in neonates (Baccarelli et al., 2008), however, is less uncertain
because the critical exposure window is much narrower (9 months), and the developmental
exposures occurred 20 to 30 years after the initial exposure, when internal TCDD concentrations
for the pregnant women likely were leveling off; that is, exposure over the critical window was
more constant and estimation of the relevant exposures was less uncertain.  However, there is
some uncertainty in the magnitude of the exposures  because they were estimated from
50 Mocarelli (2001) reported the release from the Seveso plant to contain a mixture of TCDD, ethylene glycol, and
sodium hydroxide. Because these chemicals are not thought to persist in the environment or in the body, coexposure
to these additional contaminants along with TCDD would not have a significant impact on longer-term TCDD
dose-response. For acute exposure, male reproductive or thyroid hormone effects are not evident for ethylene glycol
(U.S. EPA. 2012). It is unlikely that sodium hydroxide, being primarily a caustic agent, would cause these effects.
                                           4-66

-------
measurements in sera taken several years prior to pregnancy and do not take into account
changing patterns of exposure during pregnancy.
       Another source of uncertainty using human epidemiologic data is the lack of completely
unexposed populations. The available TCDD epidemiologic data were obtained by comparing
populations that experienced elevated TCDD exposures to populations that experienced lower
exposures, rather than to a population with no TCDD exposure. An additional complicating
factor is coexposure to DLCs, which can act lexicologically in the same way as TCDD.
Although the accidental exposure to the Seveso women's cohort was primarily to TCDD,
background exposure was largely to DLCs. Eskenazi et al. (2004) reported that TCDD
comprised only 20% of the total TEQ in the serum of the reference group that was not exposed
as a result of the Seveso factory explosion, which implies that the effective background TEQ
exposure was approximately fivefold higher than exposure to TCDD. WHO (1998) estimated
that TCDD may comprise only 5-20% of background exposures to dioxin and DLCs.  The
higher background exposure could be significant at the lower TCDD exposure levels, with the
effect diminishing as TCDD exposure increased. For dose-response modeling, the effect of a
higher background dose (i.e., total TEQ), if included, would be to shift the response curve to the
right, with responses now being associated with higher exposures.  Adding a constant to all
exposures would also reduce the proportional spread of the exposures, which would tend to alter
the shape of the dose-response curve towards sublinear. Both the right shift and the more
sublinear shape would result in higher POD estimates. In addition, the response in the reference
population is not a true zero-exposure (TEQ-free)  response.  The actual magnitude of the impact
of the DLC background exposure is impossible to assess without knowing the zero-exposure
background response. The (TEQ-free) background response cannot be assessed as no TEQ-free
population exists. Ideally, an independent absolute measure of adversity in terms of the response
variable, such as the 5 uU/mL neonatal TSH benchmark, is needed for dose-response modeling.
       As part of the uncertainty analysis for the TCDD RfD, the possible influence of different
background DLC exposure assumptions on the POD estimates derived from the two principal
studies, Baccarelli et al. (2008) and Mocarelli et al. (2008), and one comprehensive animal
bioassay, NTP (2006a), is examined quantitatively in Section 4.5. In addition, the range of
possible PODs for other epidemiologic studies that did not pass all the selection criteria in
comparison to the principal studies is presented in Section 4.5.
                                         4-67

-------
       A primary strength of the TCDD database is that analogous effects have been observed in
animal bioassays for most of the human endpoints, increasing the overall confidence in the
relevance to humans of the effects reported in rodents and the association of TCDD exposure
with the health outcomes reported in humans. Table 4-5 shows that low-dose TCDD exposures
are associated with a wide array of toxicological endpoints in rodents including developmental
effects, reproductive effects, immunotoxicity, and chronic toxicity. Effects reported in human
studies are similar, including male reproductive effects, increased TSH in neonates, and dental
defects in children; other human health effects such as female reproductive effects and chloracne
have been observed at higher exposures (see Appendix C).  Severe liver toxicity, which is a
consistently reported effect in rodents, has not been observed in humans; Michalek et al.(2001c),
however, reported slightly elevated liver enzyme levels in serum and other nonspecific liver
effects for the Ranch Hand cohort, suggestive of mild liver toxicity.  Overt immunological
endpoints, reported in the rodent bioassays, also have not been reported in human studies.
However, with respect to immunological effects, Baccarelli et al. (2004; 2002) evaluated
immunoglobin and complement levels in the sera of TCDD-exposed individuals from the Seveso
cohort and found reduced immunoglobulin in the highest exposure groups but no effect on other
immunoglobulins or on C3 or C4 complement levels and no indication of compromised immune
response. The latter finding indicates that at least one immunological measure in humans is not a
sensitive endpoint, as it is for mice, with large reductions in serum complement at low exposure
levels (White et aL 1986).
       Although there is a substantial amount of qualitative concordance of effects between
rodents and humans, quantitative concordance is not as strong, with reference to Table 4-5.  The
differential sensitivity of mice and humans for the serum complement endpoint is one example.
Other examples of differential  sensitivity are developmental dental effects and thyroid hormonal
dysregulation.  Developmental dental defects are relatively sensitive effects in rodents, appearing
at exposure levels in mice (Keller et al., 2008a; 2008b; 2007) more than an order of magnitude
lower than effect levels in humans (Alaluusua et al., 2004).  In contrast, thyroid hormone effects
are seen in rats (Crofton et al.,  2005) at 30-fold higher exposures than for humans (Baccarelli et
al., 2008). Male reproductive effects (sperm production) occur in rats (Latchoumycandane and
Mathur, 2002) and humans (Mocarelli et al., 2008) at about the same dose. To what extent these
differential sensitivities depend on specifics of the comparison, such as species (mouse vs. rat),
                                         4-68

-------
life-stage (e.g., fetal vs. adult), endpoint measure (e.g., T4 vs. TSH), or magnitude of the lowest
dose tested, cannot be determined, so strong conclusions about quantitative concordance cannot
be made.
       A more detailed tabular and graphical presentation of qualitative and quantitative
cross-species comparisons of selected toxicological endpoints  for all the animal and human
studies that met the EPA selection criteria is given in Appendix D.3.  The endpoints include male
and female reproductive effects, thyroid hormone levels, and developmental dental effects, all of
which have been reported for humans.  In addition, immunological and neurological effects are
shown because they are sensitive effects in experimental animal studies, although not evident in
humans. Hepatic effects, which are not shown in Appendix D.3, are evident in virtually all
rodent studies that looked for them and are often severe, but are not severe in humans. The
analysis presented in Appendix D.3 supports the conclusion that there is a substantial amount of
qualitative concordance of effects between rodents and humans, but a much lower quantitative
concordance. However, there are no endpoints in the selected  animal  bioassays that address
diabetes or glucose metabolism.  There may be other animal studies showing effects of interest at
higher doses in those studies that did not meet the dose limit selection criterion.
       A number of qualitative strengths and limitations/uncertainties are associated with the
animal bioassays listed in Table 4-5, as articulated in Table 4-6.  Considering the issue of lowest
tested dose, the general lack of NOAELs and acceptable BMDLs is a primary weakness of the
rodent bioassay database.  None of the eight most sensitive rodent studies in Table 4-5, spanning
an 18-fold range of LOAELs, had defined NOAELs or BMDLs. NOAELs or BMDLs were
established for only 4 of the next 13 rodent  studies.  In addition, many of these LOAELs are
characterized by relatively high responses with respect to the control population, so it is not
certain that a 10-fold lower dose (based on the application of UFL of 10) would be approximately
equivalent to a NOAEL. A major reason for the failure of BMD modeling was that the responses
were not "anchored" at the low end (i.e., first response levels were far from the BMR [see
Table 4-4]).  Another major problem with the animal bioassay data was nonmonotone and flat
response profiles. The small dose-group sizes and large dose intervals probably contributed to
many of these response characteristics that prevented successful BMD modeling. Larger study
sizes with narrower dose intervals at lower doses are still needed to clarify rodent response to
TCDD.

                                          4-69

-------
        Lower TCDD doses have been tested in rodents but almost entirely for investigation of

specialized biochemical endpoints51 that EPA does not consider to be lexicologically relevant for

the derivation of a noncancer RfD (see Appendix H). There is, however, a fundamental limit to

the lowest dose of TCDD that can be tested meaningfully, as TCDD is present in feed stock and

accumulates in unexposed animals prior to the start of any study.  This issue is illustrated by the

presence of TCDD in tissues of unexposed control animals, often at significant levels relative  to

the lowest tested dose in low-dose studies (Bell et al., 2007b: Ohsako et al., 2001; Vanden

Heuvel et al., 1994a: 1994b) (see Text Box 4-1).  Some DLCs also have been measured in

animal feeds (Bell et al., 2007b: NTP, 2006a) and are anticipated to accumulate in unexposed

test animals, further complicating the interpretation of low-dose studies.
 Text Box 4-1. Background levels of TCDD in Control Group Animals

     TCDD tissue levels in control animals are rarely reported either explicitly or implicitly.  Vanden Heuvel et al.
 £1994}however, reported TCDD concentrations in livers of control animals (10-week-old female Sprague-Dawley rats)
 of 0.43 ppt (ng/kg) compared to 0.49 ppt in the livers of animals given a single oral TCDD dose of 0.1 ng/kg.
 Assuming proportionality of liver concentration to total body burden, the body burden of untreated animals was 87.8%
 of that of treated animals at the lowest dose. The equivalent (single) administered dose for untreated animals (d0) can
 be calculated as equal to 0.878 x (0.1 + d0), assuming proportionality of body burden to administered dose and that all
 animals started with the same TCDD body burdens. The calculation yields a value of 0.72 ng/kg for d0, which
 represents the accumulated TCDD from all sources in these animals prior to being put on and during test.  This value
 would raise the nominal 0.1 ng/kg TCDD dose 8-fold to 0.82 ng/kg. The  next higher dose of 1 ng/kg would be nearly
 doubled to 1.72 g/kg. The impact on higher doses would be negligible, because the ratio of treatment dose to apparent
 background exposure levels increases with higher treatment levels.  Bell et al. (2007) reported slightly higher levels
 (0.66 ppt) in the livers of slightly older untreated pregnant female Sprague-Dawley rats (mated at 16-18 weeks of age
 and tested 17 days later).
     Ohsako et al. (2001) reported TCDD concentrations in the fat of offspring of untreated pregnant Holtzman rats
 that were 46% of the TCDD fat concentrations in animals exposed in utero to 12.5 ng/kg (single exposure on GD 15).
 This level of TCDD would imply a very large background exposure, but quantitation based on simple kinetic
 assumptions probably would not reflect the more complicated indirect exposure scenario.
     Bell et al. (2007) also reported concentrations of 0.1 and 0.6 ppt TCDD measured in two samples of feed stock.
 Assuming that the average of 0.35  ppt is representative of the entire supply of feed stock and a food consumption
 factor of 10% of body weight per day, the average daily oral exposure from feed to these animals would be
 0.035 ng/kg.  Discrimination of outcomes from longer-term repeated exposures  might be problematic at exposure
 levels around 0.1 ng/kg-day. Background exposure was not much of an issue for Bell et al. (2007). as the lowest
 TCDD exposure level was 2.4 ng/kg-day (28-day dietary exposure).
     NTP (2006b) reported TCDD  concentrations in the liver and fat of untreated female Sprague-Dawley rats after
 2 years on test that were 1% and 2.5% of the levels in the liver and fat of the low-dose TCDD treatment group
 (2.14 ng/kg-day) (NTP. 2006a). respectively. Assuming proportionality of fat concentration and oral intake, control
 animal exposure would have been approximately 0.05  ng/kg-day, similar to the  estimate from Bell et al. (2007).  As
 for the latter study, background intake for the NTP (2006a) study animals would not have a large effect on the
 dose-response assessment given the lowest exposure level of 2.14 ng/kg-day.
     In all of these studies, except the 28-day exposure  in Bell et al. (2007). control animals were gavaged with corn oil
 vehicle.  TCDD concentrations in corn oil were not reported in any of the  studies.
51 Enzyme induction, oxidative stress indicators, mRNA levels, etc.

                                               4-70

-------
4.5.  QUANTITATIVE UNCERTAINTY IN THE REFERENCE DOSE (RfD)
       The development of each candidate RfD in Sections 4.1 through 4.3 required the analysis
of numerous kinetic, toxicologic, and epidemiologic data sets.  These analyses included
interpretive decisions that were made considering different sources of uncertainty in each study
and EPA's methods for developing RfDs.  This section quantifies the impacts of some sources of
uncertainty encountered in the development of candidate RfDs (Sections 1.1 and 1.3 describe the
NAS and SAB comments pertaining to uncertainty analysis for the RfD). In Section 4.5.1, the
impacts of some sources of uncertainty encountered in the development of candidate RfDs based
on Baccarelli et al. (2008). Mocarelli et al. (2008) and NTP (2006a) are elucidated using
"variable sensitivity" trees depicting the sensitivity of the POD value to choices made for PBPK
model variables and inputs. In Section 4.5.2, an additional range of potential PODs is presented
as a bounding analysis considering background DLC exposures and several epidemiologic
studies, some of which did not qualify for RfD consideration, but for which limiting NOAEL and
LOAEL values can be estimated for purposes of comparison. All modeling for the analyses in
Sections 4.5.1.1 and 4.5.2 was carried out using the Emond human PBPK model (see
Appendix F). Modeling of the NTP (2006a) data in Section 4.5.1.2 was carried out using the
Emond and CADM rodent PBPK models and the Emond human  PBPK model (see Appendix E).
       In the analyses in Sections 4.5.1 and 4.5.2, EPA has terminated the sensitivity analysis
results at the POD level (human daily oral intake in ng/kg-day), as the PODs provide a
comparable measure across interpretive decisions. To extend these analyses further, candidate
RfDs can be estimated by converting the POD values EPA has generated to mg/kg-day and then
dividing by the appropriate uncertainty factors.

 4.5.1.  Development of Variable Sensitivity Trees for the Principal Epidemiologic Studies
        that were the basis of the Reference Dose (RfD) and for the NTP (2006a) Rodent
        Bioassay
       In this section, the impacts of some sources of uncertainty encountered in the
development of candidate RfDs based on Baccarelli et al. (2008), Mocarelli et al. (2008) and
NTP (2006a) are  elucidated using "variable sensitivity" trees depicting the sensitivity of the POD
value to choices made for PBPK model variables and inputs. These studies were chosen for this
analysis because Baccarelli et al. (2008) and Mocarelli et al. (2008) are the principal studies used
to develop the RfD, and NTP  (2006a) is among the most recent and comprehensive rodent
                                         4-71

-------
bioassay studies of TCDD. For each of the three PODs used to develop candidate RfDs from
these studies, EPA generated plausible alternative interpretations of the information used to
define judgment-based inputs for specific model variables.  The goal  of this analysis is to provide
quantitative insights on critical uncertainties encountered in the development of the RfD by
illustrating the consequences (quantified as alternative PODs at the end of each branch in each
tree) of plausible alternative interpretations of these key data sets.
       Previously, in their examination of low-dose carcinogenicity associated with
formaldehyde and chloroform exposures, Evans et al. (1994a: 1994b) assigned subjective
weights to each branch of a probability tree and calculated probability masses for population
risks associated with alternate interpretations of toxicological and pharmacokinetic data and
exposure information.52 In the examination of uncertainty undertaken in this section, EPA
utilizes the development of sensitivity trees; subjective probability weights are not developed for
any of the branches, and there is no propagation of probabilities across branches. Further, these
trees do not present a comprehensive analysis of quantitative uncertainty of the three candidate
RfDs; rather, EPA has focused on the impacts of key interpretive decisions largely dealing with
exposure and kinetic modeling uncertainties.  However, it should be noted that because POD
values do not vary greatly across each of the three trees (less than a factor of 3 or 4 in either
direction; see Figures 4-6 through 4-8), it is unlikely that the distribution of probability mass
resulting from specific probability assignments would result in a significant amount of mass
away from the chosen PODs.  In this analysis, the structure of the decisions and the resulting
POD estimates are presented as sensitivity trees in graphical form (see Figures 4-6 through 4-8).
In these figures,  the left-hand columns depict the variables considered in the sensitivity analysis.
For each variable in a column, alternative values are presented in the  row to its right.  Beginning
with the top row of a tree, the pathway for a single POD calculation is represented by the series
of lines that moves down through specific values on subsequent rows and ends with a POD.  The
series of bolded lines in each figure represents the primary POD estimation that was used to
develop the RfD for that study in Section 4.3, termed hereafter the "standard pathway". For all
other POD calculations, alternative values for each variable were assessed one at a time, while
52 Small (2008) discusses other studies of distributional approaches in risk assessment by Sielken and collaborators
that are similar to those of Evans and colleagues. These include the following: Sielken (1993.1990), Holland and
Sielken (1993). Sielken and Valdez Flores (1999. 1996). and Sielken et al. (1995).
                                           4-72

-------
    Reported Exposure
      Background
       Exposure
                                                                        68pptTCDD
                                                      "Needham'i
                                                                                                                                 "Eskenazi"
                                             TCDD Only
                                             Scenario 1
                                                15 ppt
                                             3.5x10-" ng/kg-d
                                                                                       Modeled
                                                                                       Scenario 3
                                                                                      8.9xlO-3ng/kg-d
                                                                                        80.6ppt
                                                                    DLC-TEQ added
                                                                       Scenario 5
                                                                      3.15xlO-3ng/kg-d
                                                                                     40.5 ppt
                                                                                  4.22xlO-3ng/kg-d.
      Exposure
      Duration
   Measurement
       Lag
                        2 4 hours
                                                             48 hours
                                                  Modeled
                                                 Scenario 4
                                                  116.6 ppt
                                                0.0180 ng/kg-d
                                                                                                                         DLC -TEQ added
                                                                                                                            Scenario 6
                                                                                                                          7.93xlO-3 ng/kg-d
                                                                                                   24 hours
   1 month
                        6 months
                                                  1 year
                                                                                            6  months
     Age at
    Exposure
   6.2 years
1 year    6.2 years    9 years
                         6.2 years
     POD
     (ng/kg-
     day)
 P = 0.0267
 W= 0.00709
AVG = 0.0169
  P:W=3.8
 P = 0.0349
 W= 0.00362
AVG = 0.0193
  P:W=9.6
           P= 0.0321
         W= 0.00797
         AVG = 0.0201
           P:W = 4.0
                   P = 0.0294
                   W= 0.0101
                  AVG = 0.0198
                   P:W=2.9
     I
 P = 0.0369
 W= 0.00872
AVG = 0.0228
 P:W=4.2
  P = 0.0225
 W= 0.00796
P = 0.0666
W= 0.0256
    I

P = 0.0353
W=0.0111
P = 0.0134
W= 0.0104
AVG = 0.0152  AVG = 0.0461  AVG = 0.0232 AVG = 0.0118
                                                                           P:W=2.8
                                                                                         P:W=2.6
                                                                                                     P:W=3.2
                                                                                                                  P:W= 1.3
    I

P = 0.0215
W= 0.0317
AVG=NA
P:W=0.7
     r
  P = 0.0213
 W= 0.0183
AVG = 0.0197
  P:W=1.2
W = critical window average, P = peak exposure, AVG = average of P and W, P:W = ratio of peak to window-average exposure
NA = not applicable (see description of Scenario 4 in text)
Figure 4-6. Sensitivity tree showing TCDD exposure-variable uncertainty for Mocarelli et al. (2008)

-------
     Effect
        Elevated Neonatal TSH Levels
POD
Basis
Background
Exposure
POD Method
Maternal LASC
Age at
Conception
NO
(Equivalei
TCDI
AEL
tit LOAEL)
)Only

«
TCDI
/\
/
Graphical Graphical Grap
Method Method Met
40 ppt 90 ppt 312
3
ye
0 3(
ars yea
3 3
rs ye
^^
)Only
\
hical
hod
ppt
0
ars
LOAEL
^^^^^^

^^^ ^^^
Total TEQ
excluding Non- Totan
CoplanarPCBs
x.
\
Regression Regn
Model Me
235 ppt 219
A
A
30 45 3
jssion
del
ppt
0
years years years
/ \
Regr
M
48!
3
ye
:EQ
ession
adel
5 ppt
0
ars
POD (ng/kg-day)
0.00161   0.00514
 (0.0161)
0.0303  0.0196 0.0162
                                                                                0.0180
0.0593
  Figure 4-7. Sensitivity tree showing TCDD exposure-variable uncertainty for Baccarelli et al. (2008)

-------
     Effect
           Elevated Neonatal TSH Levels
    POD
    Basis
  Background
POD (ng/kg-day)
    NOAEL
(Equivalent LOAEL)
LOAEL
                        TCDDOnly
                       TCDD Only
    Total TEQ
  excluding Non-TotalTEQ
Exposure
POD Method
Maternal LASC
Age at
Conception

j
Graphical
Method
40 ppt
3
ye
0
ars
./
/
Graphical
Method
90 ppt
3(
yea
3
rs
/
Grap
Met
312
3
ye
j
\
hical
hod
ppt
0
ars
V.
^X
Regression
Model
235 ppt
A
30 45
years years
/ \
<_
Coplanar
Regn
Me
219
3
ye
PCBs
jssion
del
ppt
0
ars

Regr
M
48!
3
ye

ession
adel
5 ppt
0
ars
   0.00161   0.00514
    (0.0161)
0.0303  0.0196 0.0162
                                                                          0.0180
                   0.0593
  Figure 4-8. Sensitivity tree showing TCDD exposure-variable uncertainty for NTP (2006a).

-------
fixing all the other variables at the values used in standard pathway. The values used for these
variables were either directly specified in the literature or were based on judgment using
exposure information provided in related papers. Up to three significant digits are shown for the
PODs that are presented so that differences among the PODs across analytic choices can be
readily discerned.

4.5.1.1. Epidemiologic Sensitivity Analyses
       In estimating the PODs for the principal studies for the RfD (Baccarelli et al., 2008;
Mocarelli et al., 2008),  a series of assumptions were made to model the exposure history of the
cohorts and to estimate an intake leading to the observed effect.  In this section, variable
sensitivity trees highlight the effects of choosing alternative assumptions on the POD estimates
for these two principal studies.

4.5.1.1.1. Mocarelli et al (2008)
       Mocarelli et al (2008) evaluated  sperm endpoints in adult males who were exposed as
children, between the ages of 1 and 9, to TCDD during the Seveso accident, which included an
initial peak exposure and subsequent longer-term exposure to ambient levels (see
Section C.I.2.1.5.8 for study details).  To examine the impacts of potential uncertainties
associated with the assumptions  made in estimating the standard pathway LOAEL POD in
Mocarelli et al. (2008) (see Section 4.2.3.2), EPA  evaluated the impact of several alternate
exposure assumptions on the oral intakes associated with the POD, as shown in Figure 4-6.  The
left side of the figure depicts the variables of the exposure analysis considered in the sensitivity
analysis (i.e., background exposure, exposure duration, measurement lag, and age at exposure).
As detailed below, the values used for these variables were not directly specified in the literature
but were based on judgment of the exposure information provided  in Mocarelli et al. (2008) and
related papers. In addition to the variables in Figure 4-6, a discussion is also presented of the
impact on the POD and RfD of changing the value of the Hill coefficient in the Emond PBPK
model to 1 instead of 0.6 (see Section 3.3.4.3.2.5 for modeling details).
       All of these variables are inputs to the Emond human PBPK model (see modeling code
and details in Appendix F), which was used to estimate the actual exposures to the affected
population and the corresponding continuous intakes for determining the RfD POD. The

                                          4-76

-------
sensitivity analysis begins with the reported LASC of 68 ppt TCDD in the LOAEL group. The
terminal nodes at the bottom of the figure show the PODs as daily oral intakes (ng/kg-day)
resulting from each alternative value for the variables examined.  To address the nature of the
Seveso TCDD exposures, the PODs are expressed using three different metrics as described
below.
       In Figure 4-6 and in the text that follows, the following abbreviations for the PODs are
used:
       "P" identifies the intake associated with the initial peak LASC exposure estimates.
       "W" identifies the intake associated with the average LASC over the actual exposure
       window.
       "AVG" is the average of the intakes associated with "P" and "W." Intakes associated
       with either "P" or "W" conceivably could have been selected as the primary POD.
       P:Wis the ratio of the peak intake to the window-average intake.
       In the standard pathway analysis, EPA elected to use the average of the peak exposure
intake (P) and the critical-window exposure average intake (W) as the basis for the POD, giving
equal weight to both (see discussion in Section 4.2.3); these values are labeled as "AVG" across
all terminal nodes in the tree. This was done because of the relatively large differences between
peak exposures and average exposures decreasing over a relatively long time span,53 and the
uncertainty of the relative influence of acute high exposures vs. lower longer-term averages on
the toxicological outcome.
       Background Exposure
       For Figure 4-6, background exposures in the population (labeled "Background
Exposure") were estimated using six different scenarios, based on data from two different
epidemiologic  studies. The scenarios take into account background exposures of TCDD only, or
TCDD in the presence of DLCs (i.e., total TEQ)54.  Because DLCs are presumed to act in the
same manner as TCDD (for AhR induction and subsequent effects), the magnitude of the
background DLC exposure is an important concern in establishing the POD.  The Emond human
PBPK model was used to estimate background intakes by assuming a constant exposure from
53 The modeled TCDD LASC decreased by a factor of 5.5 from peak exposure to the terminal value at 10 years.
54 DLC-TEQ = non-TCDD TEQ
                                         4-77

-------
birth to time of serum-TCDD measurement55 for each scenario (see Appendix F for modeling
details).
       Scenarios 1 and 2 consider background TCDD only, with Scenario 1 being the standard
pathway defining the RfD.  Scenario 2 uses a higher TCDD background estimate from a different
publication than the one used by Mocarelli et al. (2008). For the remaining scenarios, the
background TEQ exposures were estimated using two different methods. The first method was
to model the total TEQ LASC values directly with the Emond human PBPK model, assuming
that all DLCs are kinetically equivalent to TCDD.  This method ("modeled TEQ") accounts for
the magnitude of background DLC serum concentrations in the dose-dependent elimination
mechanism in the Emond PBPK model.  For the modeled-TEQ method, background DLC-TEQ
LASC values at the time of blood collection (i.e., "measurement time") were estimated by EPA
using measured data or by modeling with assumptions of the ratio of total TEQ to TCDD in
background exposures.  Total TEQ LASC values at measurement were estimated by adding the
resulting DLC-TEQ LASC to the measured TCDD LASC  of 68 ppt. The Emond model was
then run to compute the corresponding peak and critical-window intakes, with all other model
variables set to the standard-pathway values. EPA also applied a simple additive model, in
which background DLC-TEQ intakes were estimated by assuming a ratio of DLC intake to
TCDD intake from background sources.  The background  DLC intakes were then added to the
modeled TCDD intakes from the first two scenarios. The  DLC-TEQ intake addition method
does not account for the influence of DLCs on  dose-dependent elimination, but is less
complicated to apply and requires fewer assumptions than the modeled-TEQ method. A
limitation of both approaches, but more so for the modeled-TEQ method, is the assumption of
toxicokinetic equivalence of DLCs and TCDD. The reported TEQ values are based on serum
concentrations, while the TEFs, on which the TEQ values  are calculated, are largely derived
from oral dosing studies. The outcomes from such studies implicitly account for DLC
toxicokinetics (i.e., absorption, distribution, metabolism, and elimination).  Applications of TEFs
to DLC serum concentrations do not account for toxicokinetics, which could be very different
across DLCs.56 In addition, because both methods use TEQ values based on nominal TEFs, the
55 "Measurement time" is defined here as the average age (6.7 years) of the subjects studied by Mocarelli et al.
(2008) when serum samples were collected, which EPA estimated as 6 months following exposure.
56 As an example, whole body half-life estimates for the DLCs vary from about 6 months to 20 years (Oguraetal..
2004: Flesch-Janys et al.. 1996). Currently, there is no human PBPK model capable of addressing toxicokinetics for
                                         4-78

-------
DLC contribution to total TEQ will be overestimated. The TEF methodology is designed to be

health protective, in that the TEFs are not central tendency estimates but biased high by design

(Van den Berg et al., 2006). Therefore, exposure estimates based on nominal TEQ values are

expected to be slightly higher than actual exposure.

       The following descriptions apply to the scenarios depicted in Table 4-6. Additional detail
can be found in Appendix F.
   •   Scenario 1 (Needham TCDD scenario). The TCDD only background value used in the
       standard pathway analysis was based on an LASC of 15 ppt used by Mocarelli et al.
       (2008) in their analysis as the TCDD level in the comparison group; this value was
       reported by Needham et al. (1997) to be the median TCDD concentration in an
       unexposed reference adult population (25 years or older) (designated "Needham" in
       Figure 4-6). Using the Emond PBPK model, EPA estimated a corresponding daily
       TCDD intake of 3.5 x 10 4 ng/kg-day from birth, assuming that 15 ppt was obtained at
       age 35 (see Appendix F.I.I).

   •   Scenario 2 (Eskenazi TCDD scenario).  The alternative TCDD-only value is an
       age-specific background intake based on an average TCDD  concentration of 40.5 ppt for
       girls less than 12 years of age (designated "Eskenazi" in Figure 4-6) from Table 3 in
       (Eskenazi et al., 2004).5? Assuming that background TCDD serum concentrations were
       similar for boys and girls in the Seveso cohort, EPA estimated an average TCDD intake
       of 4.22 x 10"  ng/kg-day corresponding to the same average 40.5 ppt LASC for boys of
       similar age (see Appendix F.I.2).

   •   Scenario 3 (Needham modeled-TEQ scenario).  This method models the exposure
       directly, by matching the "target" total TEQ (as LASC ppt, TCDD included) at the time
       of measurement with the corresponding intake using the Emond model.  The target
       total-TEQ for the Ist-quartile boys aged 6.7 years at measurement time was estimated to
       be 140.5 ppt TEQ. This value was obtained by  adding a modeled estimate of 72.5 ppt
       background DLC-TEQ LASC at 6.7 years to the measured TCDD LASC of 68 ppt in
       Mocarelli et al. (2008). The DLC-TEQ estimate was obtained by first assuming that
       TCDD comprises 10% of the total background TEQ, which  is approximately the
       proportion of TCDD to total TEQ in adult serum as reported by (Eskenazi et al., 2004)
       and as estimated by WHO (1998).58  The Needham scenario TCDD background of 15 ppt
       was multiplied by 10 obtaining an estimate of 150 ppt total background TEQ at age 35,
       for which a corresponding average daily background intake  from birth of 0.0180 ng/kg-
all the DLC congeners, although both EPA (U.S. EPA. 2003) and Lorber (2002) have used DLC half-life estimates
and tissue concentrations to estimate intake rates for some DLCs (excluding dioxin-like PCBs) in humans.
57 Table 3 in Eskenazi et al. (2004) reports the results of two pools of sera collected from girls aged 0-12 years, who
did not reside in areas affected by the Seveso accident and were presumably exposed only to background levels of
TCDD. The 40.5 ppt estimate is the mean of the two pools (47.6 and 33.4 ppt).
58 TCDD also is approximately 10% of the total serum TEQ as calculated by EPA from the NHANES (2001/2002)
data reported by Lorber et al.  (2009).
                                          4-79

-------
       day was estimated using the Emond PBPK model. Using the background intake of 8.9
       xlO"3 ng/kg-day in the Emond model, a concentration of 80.6 ppt total TEQ LASC at age
       6.7 was modeled, 90% of which, or 72.5 ppt, is assumed to be DLC-TEQ.  (see Appendix
       F.3.6 for modeling details).

   •   Scenario 4 (Eskenazi modeled-TEQ scenario).  The method is the same as for Scenario 3.
       The target total TEQ for the Ist-quartile at measurement time was estimated to be
       144.1 ppt TEQ, which was obtained by adding a measured value of 76.1 ppt background
       DLC-TEQ at 6.7 years to the measured TCDD value of 68 ppt in Mocarelli et al. (2008).
       The DLC-TEQ estimate was obtained by averaging the non-TCDD TEQ for the
       0-12 year age group (girls) reported by Eskenazi et al. (2004): the total measured
       background TEQ for that group was  116.6 ppt (Table 3 in Eskenazi et al., (2004): the
       corresponding modeled background total TEQ intake was 0.0180 ng/kg-day. Lacking
       specific measurements for boys, EPA assumed that the averages for boys were the same
       as for girls.

   •   Scenario 5 (Needham DLC-TEQ intake added scenario).  This method adds DLC-TEQ
       intakes, which are estimated by scaling the modeled TCDD  intakes by the ratio of
       DLC:TCDD in serum for background exposures, assuming that the ratio is the same for
       oral intakes and serum concentrations. For Scenario 5, EPA assumes that TCDD
       comprises 10% of the total background TEQ, as in Scenario 3, which results in a 9:1 ratio
       for DLC:TCDD for background exposures. The resulting DLC-TEQ intake is
       3.15 x 1Q~3 ng/kg-day (9 x 3.5 x 10~4 ng/kg-day). The estimated DLC-TEQ intake is
       then added to the P, W, andAVG values for the standard pathway (Scenario 1).

   •   Scenario 6 (Eskenazi DLC-TEQ intake added scenario).  The method is the same as for
       Scenario 5. The DLC:TCDD LASC ratio is calculated from the measured serum
       concentrations (TCDD = 40.5 ppt; DLC-TEQ = 76.1 ppt) reported by Eskenazi et al.
       (2004). The resulting DLC:TCDD LASC ratio is 1.88 (76.1 - 40.5).  Multiplying the
       corresponding TCDD background intake of 4.22 x 10~3 ng/kg-day (Scenario 2) by this
       factor gives a background DLC-TEQ intake of 7.93 x 10  ng/kg-day. The total
       background TEQ intake is 0.0122 ng/kg-day (7.93 x 10~3 +  4.22 x 10~3). The estimated
       DLC-TEQ intake is then added to the P, W, andAVG values for Scenario 2.
      Exposure Duration

      "Exposure duration" refers to the duration of the elevated (external) TCDD exposures

immediately following the Seveso accident, which is not known with certainty. In the standard

pathway analysis, the "exposure duration" of the TCDD exposures due to the Seveso accident

was modeled using the Emond model as a single pulse on 1 day (i.e., 24 hours).  The alternative

also uses the Emond model but models the exposures following the Seveso accident using pulse

doses on two consecutive days (i.e., 48 hours).
                                        4-80

-------
      Measure Lag
      "Measurement lag" refers to the period of time between TCDD exposure following the
Seveso accident and the collection of blood for future TCDD analyses. Within the Seveso
cohort, serum samples were collected in 1976 and 1977, so in the standard pathway analysis, an
average measurement lag time of 6 months was assumed for exposure to TCDD.  The alternative
analyses simulate lag times of 1 month and 1 year.
      Age at Exposure
      "Age at exposure" is the average age of the susceptible lifestage (boys,  1-9 years old) at
the time of the Seveso accident. Within the cohort, the average age at exposure was reported to
be 6.2 years, which was used in standard pathway analysis.  The alternative analysis considers
individuals who would have been 1 year or 9 years of age at the time of the Seveso accident,
representing the bounds of the susceptible age range. This category is included to show the
potential range of exposures across the cohort for the reported age range rather than to evaluate
plausible alternatives to the mean age of 6.2 years.  That is, the  intakes associated with ages 1 or
9 would not be considered as PODs.
      Hill Coefficient
      Because the Hill coefficient is the most influential variable in the Emond PBPK model
(see Section 3.3.4.3.2.5) and the value of 0.6 results in a supralinear relationship between intake
and blood  concentrations at low doses, EPA also evaluated the impact of changing the Hill
coefficient. Based on the results of the expanded sensitivity analysis in Section 3.3.4.3.2.6, a
Hill coefficient of 1 and the corresponding optimized CYP1A1  elimination constant (kelv) of
0.005 were evaluated for impact on the POD. A value of 1 was chosen because that is the lowest
value where the model is no longer supralinear; otherwise the value of 1 has no biological or
empirical basis.  Because the relationship between TCDD serum concentrations and intake was
changed for the alternative parameter specifications, a revised TCDD background exposure was
modeled based on the Needham scenario.  Using the revised background TCDD intake of
1.9 x icf4 ng/kg-day, the modeled peak and window-average (TCDD-only) exposures at the
LOAEL are 7.6  x icf3 and 3.7 x icf3 ng/kg-day, respectively. The average (i.e., AVG) of the
peak and window intakes is 5.7 x 10 3 ng/kg-day, which is 3.5-fold lower than the LOAEL POD
for the RfD.
                                         4-81

-------
      Mocarelli et al. Sensitivity Tree Results
      Overall, excluding the age-at-exposure and Hill coefficient variables, neither of which are
considered to have plausible alternative values, the daily intakes (TCDD or total TEQ) based on
the alternative assumptions in this tree vary between 0.0071 ng/kg-day (Wfor 1-month
measurement lag) and 0.0666 ng/kg-day (P for modeled total TEQ, Needham background).  This
range spans the LOAEL POD for the standard pathway analysis of 0.020 ng/kg-day by about a
factor of three on each side (2.8-fold below to 3.3-fold above). The AVG values, which factor in
both peak and window-average exposures and are the preferred POD values59, vary over a
smaller range from 0.0118 ng/kg-day (Scenario 2: TCDD-only, Eskenazi background) to
0.0461 ng/kg-day (Scenario 3: modeled total TEQ, Needham background), bracketing the
LOAEL POD for the standard pathway by about a factor of two (1.7-fold below to 2.3-fold
above).
      The ratio of peak intake to window-average intake (P: Wratio) is of interest in evaluating
the range of exposures over which an average is taken. The P: W ratio is 4 for the standard
pathway POD. In general, the higher the background exposure, the lower the peak intake and the
lower the P: W ratio and the lower the impact of averaging P and W.  The P: W ratio is lowest for
all the Eskenazi background scenarios, decreasing to about a factor of 1 for the TEQ analyses.
For the Eskenazi modeled TEQ scenario, Wis larger than P because the background intake is
high enough to result in a higher terminal (10-year) LASC for the target population than was
experienced by the exposed population in the Seveso cohort; in this case, with a higher peak
realized for the average exposure over the critical window, neither/1 nor AVG would be relevant
and the higher W value would be used as the POD.
      The most  influential variable in either direction (above or below the standard pathway
RfD  LOAEL POD) is background exposure.  The higher Eskenazi background exposure scenario
had the largest impact on the TCDD-only intake estimates, with a 41% lower AVG than for the
standard pathway RfD LOAEL POD, primarily because of the lower peak exposure. The 12-fold
higher value for the Eskenazi TCDD background than for the Needham adult background is
likely a result of higher food consumption in children and a higher average environmental
concentration for the relevant childhood exposure period (1964-1976) than for the adult
59 The A VG for Scenario 1 was chosen as the POD for the RfD because it accounts for both peak and window-
average exposures.
                                         4-82

-------
exposures (ca. 1941-1976) (Lorber, 2002; Pinsky and Lorber, 1998). Also, the higher ratio of
TCDD to total TEQ in children may reflect the lack of attainment of steady state for many of the
DLCs relative to TCDD. The next most influential variable was exposure time, with a 24%
lower AVG for the 48-hour exposure time than for the 24-hour scenario. However, the modeled
exposures on each of the 2 days within the 48 hour period were equal when, in reality, they
would be decreasing with time, such that the peak is somewhat underestimated in this analysis;
longer exposure scenarios assuming constant levels would not be realistic.  The largest
differences in the other direction (i.e., exceeding the standard pathway RfD POD) were obtained
for the modeled total TEQ scenarios, with a 2.3-fold higher AVG and 3.3-fold higher peak (P) for
Scenario 3 (Needham) and a 1.6-fold higher window-average for Scenario 4 (Eskenazi).  Note
that any DLC background exposure estimate based on TEQ will be an over-estimate because of
the conservative nature of the TEF methodology. Further, there is additional uncertainty when
applying the TEF method to tissue concentrations such as LASC. All the other alternative
assumptions resulted in a 16% or lower change in the AVG values.  Although not a consideration
for defining the  POD, the TCDD AVG intakes across the susceptible age range (1-9 years) were
within 5% of the standard pathway RfD POD, but with a large P'.W ratio (9.6) for 1-year-olds.
       In summary, the quantitative uncertainties evaluated here for the RfD LOAEL POD
based on Mocarelli et al. (2008) span about a threefold range in either direction.  The largest
differences are those between peak and window-average exposures, which decrease when
considering the alternative Eskenazi background estimates. Using the latter, the AVG POD is
about half of the RfD POD for TCDD only (Scenario 2), but, when considering the TEQ
contribution, rises to about the same value as the RfD POD with additive background DLC
(Scenario 6) and to 60% higher than the RfD POD with modeled TEQ background (Scenario 4).
Using the modeled-TEQ method, the Needham background DLC exposure has a larger impact
on the standard RfD POD, increasing it by a factor of 2.3 (Scenario 3), but is only 16% higher
than the RfD POD for the additive method (Scenario 5).  Because of (1) the lack of background
TEQ measures in populations from the 1970's that are directly relevant to the Mocarelli et al.
(2008) study population, (2) the conservative nature of the TEF method, and (3) uncertainty in
the application of the TEF method to reported human tissue concentrations, EPA cannot
recommend, at this time, any particular approach for incorporating background DLC exposure
directly into the POD for the RfD. Overall, given the bidirectional nature and relatively small
                                         4-83

-------
magnitude of the uncertainties, EPA believes that this sensitivity analysis provides support for
the magnitude of the RfD.

4.5.1.1.2.  Baccaretti et al (2008)
       Baccarelli et al evaluated thyroid-stimulating hormone levels in newborns whose mothers
were exposed to TCDD during the Seveso accident (see Section C.I.2.1.5.7 for study details).
To examine the impacts of potential uncertainties associated with the assumptions made in
estimating the standard pathway POD for Baccarelli et al. (2008) (see Section 4.2.3.2), EPA
analyzed alternate assumptions about exposure and the level of change in neonatal TSH levels
associated with the designation of a LOAEL or a NOAEL from this study, as shown in
Figure 4-7. The sensitivity analysis begins with elevated neonatal TSH levels.  The terminal
nodes at the bottom of the figure show the PODs as daily oral intakes (ng/kg-day) resulting from
each alternative value for the variables examined. The left side of the figure depicts the variables
considered in the sensitivity analysis (i.e., basis of the POD, background exposure, POD method
of estimating material LASC, and maternal age at conception).  Values for these variables are
inputs to the Emond PBPK model under the human gestational  scenario (see Section 4.2.2),
which was used to estimate the PODs in Figure 4-7. Each POD is a continuous daily oral TCDD
or TEQ intake that would result in a specified TCDD maternal LASC corresponding to a
neonatal TSH of 5 jiU/mL at the end of gestation (see modeling code and details in Appendix F).
       POD Basis
       In the standard pathway analysis, the neonatal TSH of 5 jiU/mL at the end of gestation is
determined to be a LOAEL. The alternative assumption evaluated in Figure 4-7 is that this value
is a NOAEL. For the NOAEL in Figure 4-7,  the equivalent LOAEL (by multiplying by 10)60 is
also shown for direct comparison to the LOAEL estimates. The choice of the maternal LASC
value for the NOAEL is discussed below.
       POD Method of Determining Maternal LASC for TCDD Only
       There are several ways in which a POD could be derived from the Baccarelli et al. (2008)
study.  In the standard pathway RfD analysis, EPA used the study authors' regression model
results from their Figure 2A (designated the "Regression Model") to determine a LOAEL based
60 A tenfold factor is used because the LOAEL POD is divided by a UFL of 10 in the RfD derivation. The
"equivalent" LOAEL is not meant to be an alternative LOAEL but is used strictly for comparison.
                                         4-84

-------
on the maternal plasma concentration corresponding to neonatal TSH levels of 5 uU/mL.  The
advantage in using the regression model is that it was used to account for covariates that
influenced the dose-response relationship. Three alternative values are examined by selecting
specific points or ranges from the figures in the Baccarelli paper, without consideration of the
regression modeling results (the "graphical method"). The alternative values, therefore, do not
account for the covariates.  The first assumes a NOAEL of 40 ppt maternal LASC, which is
essentially the highest TCDD concentration above which neonatal TSH levels are consistently
above 5 uU/mL [see Figure 2A in Baccarelli et al. (2008)1.  The figure (2A) shows that 5 of the
6 neonates born to women who had TCDD concentrations above 40 ppt had TSH levels above
5 uU/mL; among the 45 women who had TCDD concentrations below 40 ppt, only two had
babies with TSH levels above 5 uU/mL.  The second alternative assumes that the 6 neonates
born to women with TCDD LASC  above 40 ppt comprise a LOAEL group, with a median
maternal LASC of 90 ppt.  The third alternative assumes a LOAEL at the highest neonatal TSH
level (8.5 uU/mL) shown in Figure 2A, which corresponds to a maternal TCDD LASC of
312 ppt.
       Background Exposure
       Background exposures in the population were estimated in several ways. The
background TCDD exposure used in the standard pathway RfD analysis was based on
continuous intake necessary to obtain 15 ppt at 30 years for females (the "Needham" TCDD
Only background in Figure 4-6); the modeled TCDD intake was 3.9 x 10~4 ng/kg-day, slightly
higher than that for males.  To examine the maternal TEQ exposures associated with a LOAEL
based on a neonatal TSH level of 5 uU/mL, EPA relied on the regression results reported in
Baccarelli et al. (2008).  Baccarelli et al. (2008) reported maternal plasma TEQ concentrations in
the following two ways: (1) polychlorinated dibenzo-p-dioxins (PCDDs), PCDFs, coplanar
PCBs, without noncoplanar PCBs (see Figure 2B) and (2) PCDDs, PCDFs, coplanar PCBs, and
noncoplanar PCBs, termed total TEQ (see Figure 2D).  The concentrations in their Figures 2B
and 2D are reported as TEQs and were modeled as TCDD for this analysis.  Excluding the
noncoplanar PCBs, maternal  TEQ levels of 219 ppt in serum are associated with neonatal TSH
level of 5 uU/mL.  For the total TEQ, maternal TEQ levels of 485 ppt in serum are associated
with a neonatal TSH  level of 5 uU/mL. Confidence in the total TEQ estimate is lower than that
                                        4-85

-------
for the one without the noncoplanar PCBs because of the lower significance of the total TEQ
regression coefficient (p = 0.14) than the one without the noncoplanar PCBs (p = 0.005).
      Age at Conception
      For the standard pathway RfD analysis, the maternal "age at conception" was set at
30 years, which was the average reported in Baccarelli et al. (2008). The alternative assumes the
maternal age at conception to be 45 years of age; this is the standard gestational scenario used in
estimating the human equivalent doses for the animal bioassays reporting reproductive or
developmental effects and is considered to be a reasonable upper end of female fertility.
      Baccarelli et al. Sensitivity Tree Results
      The alternative LOAEL PODs based on this analysis of Baccarelli et al. (2008) vary
between 0.005 and 0.059 ng/kg-day. These two values are roughly a factor of 4 lower and a
factor of 3 larger, respectively, than the LOAEL estimate of 0.020 ng/kg-day that was the  basis
of the standard pathway RfD.  The TCDD intake of 0.0016 ng/kg-day corresponding to the
alternative NOAEL is slightly more than an order of magnitude lower than the standard pathway
RfD LOAEL POD and would yield a slightly lower RfD estimate than the current RfD after
eliminating the 10-fold UFL factor.  EPA has much less confidence in the NOAEL estimate than
in the selected LOAEL because the NOAEL does not take into account the covariates and falls in
a lower concentration range where the background DLC exposures are a much more significant
component. The largest downward impact on the standard pathway LOAEL POD results  from
grouping the highest  exposures independent of the modeling results (POD = 0.005), which
decreases the LOAEL by a factor of four; however, analogous to the NOAEL alternative, the
approach ignores the contribution of covariates. Using the alternative age of conception of
45 years yielded a POD of 0.0162, which is virtually the same as the standard pathway LOAEL
POD of 0.0196.
      The largest upward impact on the standard pathway LOAEL POD is the inclusion  of
modeled total TEQ (POD = 0.059), which increases the LOAEL by a factor of three. However,
the model fit is poor, and the result can be compared with an analogous calculation to the
additive DLC approach used for the Mocarelli analysis in Figure 4-6. An additive DLC-TEQ
background of 3.5 x  10 3 ng/kg-day can be estimated for the women in the Baccarelli analysis by
multiplying the TCDD background intake of 3.9 x 10 4 ng/kg-day by 9 (not shown in
Figure 4-7). Adding the estimated DLC background to the standard pathway RfD LOAEL POD
                                         4-86

-------
of 0.0196 gives a corresponding total-TEQ intake of 0.0231 ng/kg-day.  This is 1.2-fold higher
than the standard pathway RfD POD but 2.6-fold lower than the modeled total-TEQ POD.
Leaving out the noncoplanar PCBs greatly improves the significance of the slope, which could
suggest that the noncoplanar PCBs do not contribute to the effect as much as the PCDDs and
PCDFs or that there is greater uncertainty in the TEQ estimates for the noncoplanar PCBs.  In
either case, as for the Mocarelli analysis, any estimate of background DLC exposure based on
TEQ is likely an over-estimate because of the conservative nature of TEFs; there also is
uncertainty in the application of the TEF method to reported human tissue concentrations.
Overall, although background DLC exposures will effectively increase the POD to  some degree,
EPA believes that the effect is  relatively small and is in the range of the estimated standard
pathway TCDD LOAEL.
       In summary, the quantitative uncertainties evaluated here for the RfD POD based on
Baccarelli et al. (2008) span a three to fourfold range in either direction. The alternative
LOAELs at either extreme are  not strong POD candidates; the lowest value (from the graphical
method) does not account for covariates and there is greater uncertainty in the (total TEQ)
regression model for the highest value than for the other regression models.  All the other
alternative LOAELs are within a factor of 1.5 of the RfD POD.  Overall, as for Mocarelli et al.
(2008) analysis, EPA believes that this sensitivity analysis also supports the magnitude of the
RfD.

4.5.1.2. NTP (2006a) Sensitivity Analysis
       The NTP  (2006a) bioassay is a comprehensive evaluation of TCDD chronic toxicity in
female Sprague-Dawley rats, evaluating dozens of endpoints at several time points  in all major
tissues (see Section D. 1.5.8 for study details). To examine the impacts of some of the
uncertainties associated with estimating the POD from the NTP (2006a) study (see  Section 4.2),
EPA analyzed two different approaches for estimating dose and alternate choices of rodent
kinetic model and background.  Figure 4-8 depicts this analysis, which relied on an approach
similar to those used in characterizing some of the uncertainties in the RfDs derived from
Mocarelli et al. (2008) and Baccarelli et al. (2008).  The sensitivity analysis begins  with the
administered dose or  measured tissue concentrations.  The terminal nodes at the bottom of the
figure show the LOAEL PODs as daily oral intakes (ng/kg-day) resulting from each alternative

                                         4-87

-------
value for the variables examined. The left side of the figure depicts the variables considered in
the sensitivity analysis (i.e., rodent kinetic model, dose metric, background exposure, and human
kinetic model). Values for these variables are inputs to the Emond or CADM rodent PBPK
models and the Emond human PBPK model, which were used to estimate the PODs in
Figure 4-8 (see modeling code and details in Appendix E).
       The lowest administered dose of 2.14 ng/kg-day was determined to be the animal
LOAEL based on liver and lung lesions in the rats. In the standard pathway candidate RfD
analysis, the LOAELnED was the POD.
       Exposures were estimated either based on a kinetic model of the administered TCDD
dose or on the measured concentrations of TCDD and DLCs in the rat adipose tissue after
terminal sacrifice.  NTP reported concentrations of TCDD, 2,3,4,7,8-pentachlorodibenzofuran
(PeCDF), and 3,3N,4,4N,5-pentachlorobiphenyl (PCB-126) in the adipose and liver tissues
obtained from the rats after terminal sacrifice. The 2005 WHO TEF values for PeCDF and
PCB-126 are 0.3 and 0.1, respectively (Van den Berg et al.. 2006).
      Rodent Kinetic Models
       To predict average tissue concentrations based on the administered TCDD dose, EPA
used both the Emond and CADM kinetic models; the Emond model was used in the standard
pathway analysis.  EPA also used the first-order body burden model to predict whole body
TCDD concentrations; this model uses a constant half-life to simulate the elimination of TCDD
from the body. Section 3 describes all of these models.
      Dose Metric
       EPA used several alternative dose metrics based on the modeling approach and measured
tissue concentrations.  The first-order body burden model estimates the TCDD concentration in
the whole body. When using the Emond model to  evaluate the disposition of TCDD, EPA
evaluated both the whole-blood TCDD concentrations used in the standard pathway analysis and
LASC. For the CADM model, EPA simulated TCDD concentrations in the adipose
compartment following the administered TCDD dose. EPA also used the TCDD (see Table 13
in the NTP report) or DLC concentrations (see Tables 10 and 11 in the NTP (2006c) report)
measured in the adipose tissue collected at study termination.

-------
      Background Exposure
      Using the DLC concentration information, EPA estimated TEQ in two ways. In the first
approach, based on an analysis of DLCs in the adipose tissue that was reported in another NTP
study on DLC mixtures (NTP, 2006c), EPA initially estimated the ratio of the adipose tissue
TEQ concentration to the adipose tissue TCDD concentration, then applied this ratio to the
Emond whole-blood TCDD estimates assuming proportionality (resulting in a LOAEL whole
blood concentration of 2.75 ppt instead of the TCDD-only concentration of 2.56 ppt used in the
standard pathway analysis).
      In the second approach, EPA estimated TEQ dose based on adipose tissue TCDD levels
reported by NTP; the reported TCDD concentration in the fat given in the study at the lowest
dose was used to estimate a LOAEL using the Emond model. Finally, using the 2005 WHO TEF
values (Van den Berg et al., 2006), EPA converted the reported concentrations of TCDD,
PeCDF, and PCB-126 measured in the fat of the control rats in the NTP mixtures study (NTP.
      to TEQ using eq. 4-1.
                                        Chemical ( fatMr ) x TEF
                          Chemical,(B} =	    'X™' .    ' *DoseTCDD     (Eq. 4-1)
                                            TCDD(fatTCDD)
where
   Chemicali(B)    = estimate of background exposure to Chemical /' in ppt units of TCDD
                     blood concentrations at 105 weeks, for i = TCDD, PeCDF, and PCB126.
   Chemicali(fatMc) = mean ppt (pg/g) of Chemical / in the fat tissues of the control animals at
                     105 weeks in mixtures study (NTP, 2006c).
   TCDD(fatxcDD)  = mean pg/g of TCDD in the fat tissues of the 3 ng/kg dose group at
                     105 weeks in the TCDD study (NTP. 2006a).
   DoseicoD       = 2.56 ng/kg TCDD blood concentration for the 3 ng/kg dose group in the
                     TCDD study (NTP. 2006a).
   TEF;            = Toxicity Equivalence Factor for Chemical /'  [from Van den berg et al.
                     (2006)].
       Assuming simple proportionality of blood TCDD concentrations between controls and
low-dose (2.14 ng/kg-day) animals, the TEF-adjusted ratio of each congener (Chemical /') in
control animal fat to low-dose-animal fat is multiplied by the modeled TCDD blood
concentration for the low-dose animals to obtain an equivalent background exposure in the dose
                                        4-89

-------
metric (ppt whole blood). For total TEQ, the estimates of all three congeners are summed. Total
TEQ estimates likely are biased somewhat high because they are based on terminal (2-year)
measurements rather than representing lifetime averages.
      Human Kinetic Models
      To estimate the final human intake LOAEL PODs in Figure 4-8, EPA used the Emond
human kinetic model that was used in the standard pathway analysis; CADM does not cover all
life stages needed for comparison. EPA also used first-order kinetics to estimate the LOAEL
POD under the scenario that begins with first order body burden NTP Variable Sensitivity Tree
Results
      Overall, the alternative LOAEL POD estimates in this tree (see Figure 4-8) vary between
0.023 and 0.44 ng/kg-day. This range is approximately sixfold lower to threefold higher than the
LOAEL POD for the standard pathway RfD of 0.14 ng/kg-day.  The alternative LOAEL based
on first order body burden (0.023 ng/kg-day) is the lowest value in the range, approximately 85%
lower than the LOAEL based on the standard pathway approach. The difference between these
two estimates is consistent with the more conservative approach used in modeling first-order
TCDD body burdens.  The alternative LOAEL based on the TEQ in whole blood is less than
10% greater than the LOAEL from the standard pathway RfD. The alternative candidate
LOAEL based on the TCDD in lipid-adjusted serum is approximately 120% greater than the
LOAEL for the standard pathway RfD.  The use of the CADM model to estimate adipose tissue
concentration based on administered dose resulted in a 35% increase in the LOAEL estimate
relative to the LOAEL based on the standard pathway approach. The LOAELs based on
measured TCDD or TEQ levels in rodent adipose tissue were greater than the LOAEL from the
standard pathway RfD by approximately a factor of three. EPA believes that this sensitivity
analysis is supportive of the modeling choices EPA has made in the derivation of PODs for
TCDD RfD derivation.
 4.5.2. Evaluation of Range of Alternative Points of Departure (PODs) for Additional
       Epidemiologic Endpoints
       In addition to the principal studies depicted in Figures 4-6 and 4-7, EPA evaluated a
number of endpoints presented in seven other Seveso cohort studies to estimate the range of
potential PODs based on uncertainties in exposure duration, exposure averaging protocols, and
                                        4-90

-------
DLC background exposures.  Included in those study/endpoint combinations are the following:
two that passed all the selection criteria, developmental dental effects (Alaluusua et al., 2004)
and duration of menstrual period (Eskenazi et al., 2002b): a new developmental study on semen
quality (Mocarelli et al., 2011) that was published after the study selection process was
completed but is  useful in this uncertainty analysis of the POD ranges; and four studies that did
not pass all the criteria for qualification as POD candidates (Warner et al., 2007; Eskenazi et al.,
2005; Warner et al., 2004; Mocarelli, 2000) that analyzed ovarian function/progesterone, age at
menopause, age at menarche, and sex ratio, respectively,  but for which limiting NOAEL and
LOAEL values can be estimated.  Descriptions and evaluations for all of these studies, except
Mocarelli et al. (2011), can be found in Appendix C. Mocarelli et al.  (2011) is described earlier
in this section (4.3.6.2). Tables 4-8 through 4-10 and Figure 4-9 present the exposure values
modeled using the Emond human PBPK model for potential POD ranges for these 7 additional
endpoints studied in the Seveso cohort. The details of the kinetic modeling for these endpoints
and the corresponding background exposures can be found in Appendix F.
       For most  of the studies that did not pass all the criteria, the major uncertainties are the
definition of the critical exposure window (see Text Box  2-2) and the corresponding relevant
exposure-averaging time,  and the determination of adverse effect levels. Alaluusua et al. (2004)
and Eskenazi  et al. (2002b) passed the selection criteria because a critical exposure window
could be identified for each. Alaluusua et al. is included  among the candidate RfDs in Table 4-5,
but Eskenazi et al. was not carried forward because the determination of an adverse effect level
for length of menstrual cycle was considered to be too arbitrary. A critical exposure window can
be identified also for Warner et al. (2004) (age at menarche), but no TCDD-related adverse
health outcomes were observed. However, for each of the studies considered here, with some
additional assumptions, NOAELs and LOAELs at nominal group-exposure levels can be
determined. When a critical window cannot be identified, the critical exposure window is
assumed to be the entire duration from exposure in 1976 to time of interview (i.e., end of
follow-up period).  Tentative NOAELs and LOAELs are  designated for those endpoints where
adversity levels are difficult to define. Given these assumptions and limitations, TCDD and total
TEQ intakes can be modeled but must be considered to be lower bounds on the effective
exposures, given the conservative nature of the assumptions; EPA does not consider these
estimates suitable for use in the derivation  of the TCDD RfD.
                                          4-91

-------
       Table 4-8.  Alternative PODs for the impact of TCDD exposure during
       gestation and nursing on semen quality of male offspring (Mocarelli et al.,
       2011)
POD type
NOAEL
LOAEL
NOAEL
LOAEL
Age-at-conception
scenario
30 years
45 years
Averaging
protocol"
Cont. avg.
Cont. avg.
Maternal intake (ng/kg-day)
TCDD only
2.9 x 1(T4
1.50 x 10~3
2.9 x 10~4
1.04 x 10~3
TCDD + DLCb
2.90 x 10~3
4.11 x 1Q~3
2.90 x 10~3
3.65 x 1Q~3
aCont. avg. = average continuous exposure over the specified duration.
bAdded background DLC = 2.61 x l(T3 ng/kg-day (9 x TCDD background intake at NOAEL)
       Table 4-9.  Alternative PODs for developmental endpoints other than
       increased neonatal TSH and semen quality
Population, endpoint
(cite)
Girls, duration of menstrual
cycle as women
(Eskenazi et al.. 2002b)

Girls and boys, developmental
dental effects
(Alaluusua et al.. 2004)

Girls, age at menarche
( Warner etal. 2004)

POD type
NOAEL
LOAEL
NOAEL
LOAEL
NOAEL
Averaging
protocol"
Cont. avg.
Peak
Window
P/W avg.
Peak
Window
P/W avg.
Peak
Window
P/W avg.
Peak
Window
P/W avg.
TCDD only (ng/kg-day)
Needham
0.0102
61
1.5
31
0.0655
0.0157
0.0406
1.65
0.149
0.897
0.604
0.0394
0.322
Eskenazi
3.1 x IQ-i
60
1.5
31
0.0437
0.0175
0.0306
1.51
0.151
0.841
0.517
0.0424
0.280
TCDD + DLC (ng/kg-day)
Needhamb
0.0137
61
1.5
31
0.0688
0.0190
0.0439
1.65
0.152
0.900
0.607
0.0427
0.325
Eskenazi0
0.0112
60
1.51
31
0.0517
0.0255
0.0386
1.52
0.159
0.849
0.525
0.0505
0.288
aCont. avg. = average continuous daily intake over the specified duration; P = average intake for peak
 exposure; W = average intake for critical-window exposure; P/W avg. = average of "Peak" and
 "Window" intakes.
"Added DLC = 3.51
"Added DLC = 8.1 x
< 10  ng/kg-day for girls, 3.33 x 10  ng/kg-day for boy/girl average.
10~3 ng/kg-day for girls, 8.0 x 10~3 ng/kg-day for boy/girl average.
                                             4-92

-------
        Table 4-10. Alternative PODs for adult endpoints for which critical exposure
        windows are undefined
Population, endpoint
(cite)
Men, sex ratio of offspring
(MocarellietaL. 2000)




Women, age at menopause
(Eskenazi et al.. 2005)




Women, ovarian function,
progesterone
( Warner etal. 2007)
POD type

NOAEL


LOAEL


NOAEL


LOAEL


NOAEL

Averaging
protocol"
Peak
Window
P/W avg.
Peak
Window
P/W avg.
Peak
Window
P/W avg.
Peak
Window
P/W avg.
Peak
Window
P/W avg.
TCDD only
(ng/kg-day)
0.0341
1.58 x 10~3
0.0178
0.162
4.69 x 10~3
0.0831
1.6 x 10~4-3.4 x 10~3
1.6 x 10~4-1.0 x 10~3
1.6 x 10~4-2.2 x 10~3
0.013-0.052
1.7 x 10~3-3.4 x 10~3
7.3 x 10~3-0.028
0.204
3.00 x 10~3
0.104
TCDD + DLCb
(ng/kg-day)
0.0373
4.73 x 10~3
0.0210
0.165
7.84 x 10~3
0.0863
1.6 x 10~3-6.9 x 10~3
1.6 x 10~3-4.5 x 10~3
1.6 x 10~3-5.7 x 10~3
0.016-0.055
5.2 x 10~3-7.0 x 10~3
0.011-0.031
0.208
6.51 x 10~3
0.108
aCont. avg. = average continuous daily intake over the specified duration; Peak = average intake for peak exposure;
 Window = average intake for critical-window exposure; P/W avg. = average of "Peak" and "Window" intakes.
bAddedDLC = 3.15  x 10~3 ng/kg-day for males, 3.51 x  HT3 ng/kg-day for females, 3.33 x 10~3 ng/kg-day for
male/female average.
                                              4-93

-------
VO








l.E+01
"O
OX)
H 1 .E+00
g l.E-01
^^^— TEO Exnosure Ranse ^^
^— TCDD Exposure Range
A NOAEL 0 LOAEL


Dental Menstrual |
Defects


Semen

Quality

< • < -
Neonatal
TSH



s- _± f
td
l.E-02

l.E-03


1 E-05 -
,r- 3 ,

^ ^
-' < .
j [





n r> Q a Q a
8 e 8 H 8 H
H oo H « H S
II § i § -
CS (N ._  t | Ratio





I

1 Semen

_/_, Quality '


j
^ 1 t
_ J

	
1







Age at
Men arc
Age at * A
Menopause


P < • ' ' P /
k t , Y -
4 ' |7
p < > r

' * wi ^'w
ie „
Ovarian
Function
V V


•


,
j 1
i W * ' Jt

J .
1
I




QO QO QO QO QO QO
Q^ QW QW QW QP
q Q W
CJ H
H-Q ^^ HO H"^> H •* H1^
-9o ^Ho °o "oo ^o f; o
s° s^ §^ §^ §^ §^
O cS ^' fN ^^' fN ^^' 
-------
       Additional endpoints reported in the epidemiologic literature were considered in the
context of this uncertainty analysis but were excluded based on large uncertainties in defining
adversity or plausible exposure profiles over time. All the Ranch Hand studies61 were excluded
because of the inability to construct effective exposure profiles with any confidence, given the
20-year lag between the actual TCDD exposures and measurement of serum levels. For the
                           rr\
Seveso cohort, several studies  were eliminated from consideration because uncertainties in
defining plausible NOAELs or LOAELs were too large.
       For modeling of the endpoints in Tables 4-8 to 4-10, grouped exposure ranges were
represented by the geometric mean of the range limits. The average daily intakes for exposures
(LASC) in the background range were estimated as the continuous exposure from birth resulting
in the reported serum concentrations (TCDD or total TEQ) at the average subject age at time of
measurement.  Peak and critical-window average exposures (as LASC) were modeled for
measured LASC values greater than background using the actual exposure scenarios. Because
all exposure durations were less than lifetime, average daily intakes for all modeled peak and
window-average LASC were estimated using the terminal 5-year-peak average as described in
Section 3.3.6. Precision is expressed to  the nearest 10 5 ng/kg-day for all intake estimates to
avoid rounding errors when adding DLC background intakes.  DLC background intakes are the
same as those discussed previously in this section (4.5.1.1.1).  Values less than or equal to 10~3
are shown in scientific notation for readability.
       Figure 4-9 shows the range of NOAELs and LOAELs and exposures for all of the
endpoints considered in this uncertainty analysis, the endpoints on which they are based, and the
study citation.  The study/endpoint combinations are separated into two groups representing
either those chosen for RfD POD consideration ("Candidate RfD") or those not otherwise
qualifying ("Uncertainty Analysis Only"). The NOAELS and LOAELS are indicated for each
study, as appropriate, and the vertical lines through these PODs represent the range of possible
PODs based on Emond PBPK results using alternative exposure  scenarios  (see Appendix F).
The limits across studies—indicated by  symbols of the same type—for each POD type (NOAEL
or LOAEL) for each endpoint cover the  full range of alternative PODs in Tables 4-8 to 4-10,
61 (Michalek and Pavuk. 2008: PavuketaL 2003: Michalek et al.. 200la: MichaleketaL 200Ib: MichaleketaL
200Ic: Longnecker and Michalek. 2000)
62 (Eskenazietal.. 2007: Baccarelli et aL 2005: Baccarelli et al.. 2004: Eskenazietal.. 2003: LandietaL 2003:
Baccarelli et al.. 2002: Eskenazi et al.. 2002a)
                                          4-95

-------
without distinction of the relative plausibility of each one.  That is, all the PODs are treated
equally without considering the relative confidence held in each one, individually.  The low end
of most of the ranges is the critical-window average exposure, which does not take into account
the influence of the much higher peak exposure. Conversely, the upper end of the range is
generally the peak exposure, which does not account for the potential effect of longer-term
continuous exposure.  On the "uncertainty analysis only" side of Figure 4-9, most of the
NOAELs and many of the LOAELs are somewhat speculative and would not be considered as
candidates for the RfD POD. The range limits are themselves uncertain. The same DLC
modeling issues presented in Section 4.5.1 apply to all the TEQ results here, so the TEQ results
are approximations and are unlikely to be very accurate. Also, the lowest POD estimates are
more affected by background DLC exposure than are the PODs closer to the RfD POD;
generally, TCDD is a minor component of the total TEQ for the lower PODs, subjecting the
lowest alternative PODs to the greatest uncertainty. The RfD LOAEL POD (0.02 ng/kg-day)
and its RfD NOAEL Equivalent estimate (0.002 ng/kg-day, with the 10-fold UF),  along with the
RfD (7 x 10 4 ng/kg-day), are shown on the figure for comparison to the alternative POD ranges.
       The LOAEL ranges for the two principal studies (Baccarelli et al., 2008; Mocarelli et al.,
2008) span the RfD LOAEL POD, whether based on TCDD alone or total TEQ. The
TCDD-only NOAEL estimate for Baccarelli et al. (2008) is only slightly below the RfD NOAEL
Equivalent POD.  The NOAEL and the lowest alternative LOAELs for Baccarelli  et al. (2008)
are not strong POD candidates because they are based on the raw observations and do not take
into account the covariates that affect the exposure-response relationship, as does the regression
model on which the RfD LOAEL POD is based. The ranges for the total TEQ LOAEL PODS
for the coprincipal studies straddle the RfD LOAEL POD benchmark, in the range of twofold
below to threefold above.63 The POD ranges for the other candidate RfD endpoints are well
above their respective comparison NOAEL/LOAEL benchmarks (i.e., RfD NOAEL Equivalent
and RfD LOAEL). The NOAEL for Eskenazi et al. (2002b) is somewhat arbitrary, based simply
on a continuous average exposure over a 13-year window corresponding to a normal 28-day
menstrual cycle, without considering the possible range of normal durations.
       Of the endpoints that were not selected as RfD POD candidates, there are three whose
LOAEL ranges are wholly or mostly below the RfD LOAEL POD. The sperm effects in men
63 See Sections 4.5.1.1.1 and 4.5.1.1.2 for more details
                                         4-96

-------
who were exposed in utero and by lactation reported by Mocarelli et al. (2011) are very similar
to those in men exposed as boys in one of the principal studies (Mocarelli et al., 2008). The
maternal exposures associated with the effects reported by Mocarelli et al. (2011) are very low
with the TCDD-only LOAEL being 12-fold lower than the RfD LOAEL POD for the 30-year
exposure scenario. For this study, a TCDD-only NOAEL can be established at
2.9 x 10 4 ng/kg-day (for the reference population), which is sevenfold below the equivalent RfD
NOAEL POD. Both the TCDD-only NOAEL and LOAEL are much lower than the estimated
DLC background exposure; however, assuming a simple TEQ  additive model, and with the
aforementioned uncertainties concerning DLC-TEQ estimation, a TEQ NOAEL and LOAEL of
2.9 x io~3 and 4.11 x 10~3 ng/kg-day can be estimated (see Table 4-8 and Appendix F.3.7).
Although the TEQ LOAEL is still well below that for the RfD  POD, the TEQ NOAEL is in the
range of the RfD NOAEL Equivalent POD. Given the large amount of uncertainty in the
modeled NOAEL and LOAEL for this endpoint, EPA elected not to consider either as a POD.
       The second endpoint with lower LOAELs than the RfD POD is age at menopause
reported by Eskenazi et al. (2005). The figure for this endpoint includes two separate LOAEL
candidates because of uncertainty in determining adversity at the lower exposure level in
question (3rd quintile). For that reason, the daily intakes associated with the critical-window
average and peak exposures are labeled ("W" and "P," respectively). The intakes associated
with the peak  are in the range of the RfD LOAEL benchmark, while the window-average TCDD
intakes are closer to the NOAEL benchmark. Considering background DLC intake, the
window-average TEQ intakes are considerably higher, the DLC exposures being larger than the
TCDD intakes, themselves, but still below the LOAEL benchmark. The range of the TEQ P/W
average of 0.01-0.031 ng/kg-day (see Table 4-10), however, straddles the RfD LOAEL
benchmark of 0.02 ng/kg-day. Uncertainty in the NOAEL is similar to that for the LOAEL,
depending on  whether the 1st or 2nd quintile can be called a NOAEL. Although the response in
the 2nd quintile is not significant compared to the 1st quintile, the NOAEL determination is
complicated by the lack of an absolute measure of "normal."
       The NOAELs and LOAELs for altered sex ratio reported by Mocarelli et al. (2000) span
their respective RfD POD benchmarks and are above the benchmarks when considering the
peak/window  exposure averages or background DLC exposures. The uncertainties for lack of an
identifiable critical exposure window also apply to this endpoint. The other two endpoints, age
                                        4-97

-------
at menarche (Warner et al., 2004) and ovarian function (Warner et al., 2007), are unbounded
NOAELs at the highest exposures.  The ovarian function endpoint also is uncertain for lack of an
identifiable critical exposure window.
       Additional uncertainties not covered explicitly in this analysis include exposure to other
AhR agonists, either naturally occurring in food-stuffs (Connor et al., 2008) or by-products of
combustion or manufacturing processes (e.g., poly-aromatic hydrocarbons), and choice of
uncertainty factor. As a final note on background DLC exposure, the background DLC intake
estimates for the standard scenario (Needham) used in this assessment are somewhat crude, in
that they are simple multiples of modeled TCDD intake based on an approximation of the
proportion of TCDD to total TEQ.  TCDD exposures are modeled over durations of up to
35 years (1941-1976) using a single fixed background intake term (a model limitation).
However, background TCDD/TEQ exposures are thought to have varied widely over that time
period, increasing gradually in the United States from the early 20* century to a peak in 1965,
then decreasing rapidly to near current levels  in the early 1980s (Lorber, 2002). Based on a
digitization of Figure 6 in Lorber (2002), depicting the estimated TEQ intake over the course of
the 20*  century, a time-weighted average total TEQ intake for the period 1941-1976 of
4.6 x io~3 ng/kg-day can be estimated.  Adjusting the TEFgg-based Lorber (2002) TEQ intakes to
TEFos-based values, assuming  a 10% TCDD fraction and adjusting the TEFs from 1998 to 2005
(see Appendix F, Section F.I.2.1), yields a DLC-TEQ intake estimate of 3.4 x 10~3 ng/kg-day for
that time period, which is similar to the estimated DLC background intake of
3.33 x  10 3 ng/kg-day for the standard scenario using the simple scaling model.
       However, the DLC intake estimate based on Lorber (2002) is somewhat of an
underestimate because it does not include dioxin-like PCBs. Pinsky and Lorber (1998) estimated
a TCDD intake of 4 x 10 4 ng/kg-day for the U.S. population in the 1970s, which is almost the
same as the modeled TCDD background intake for the Seveso population.  However, there is no
information on comparative environmental exposures for the United States and  Italy during this
period, and TCDD exposures before 1970 for these populations were not necessarily the same,
on average. Higher TCDD background exposures have been estimated by others. Pinsky and
Lorber (1998) estimated an average TCDD-only intake of 1.4 x 10"3 to  1.9 x 10~3 ng/kg-day for
the U.S. population in the late 1960s and early 1970s using a lst-order kinetics model with a
variable intake term and a TCDD half-life of 7.1 years. Aylward and Hays (2002) estimated a
                                         4-98

-------
TCDD intake of at least 1.3 x 10 3 ng/kg-day for the United States, Canada, Germany, and
France prior to 1972 using a lst-order kinetics model assuming a TCDD half-life of 7.5 years.
These estimates are 3.5-5 times higher than the background TCDD intake estimated by EPA
using the Emond PBPK model for this assessment.  Total TEQ background would increase
proportionally.  However, none of these estimates, including EPA's, is based on actual intake
measurements and are all dependent on modeling assumptions. Raising the background DLC
exposure would obviously increase the effective PODs. However, increasing the background
TCDD intake for modeling purposes would decrease the contribution of the actual TCDD
exposures experienced by the Seveso population in 1976, resulting in a lower TCDD POD, as
can be seen in the Eskenazi background scenario for Mocarelli et al. (2008) (see Figure 4-6).
       This analysis highlights several important research needs. While the disposition of
TCDD following high exposures is reasonably understood and simulated in current models, the
current scientific understanding of disposition following TCDD exposures that are closer to
current background dietary intakes, likely the primary source of TCDD exposure  for most of the
U.S. population, is not understood as well at present. This uncertainty affects the estimation of
TCDD intake rates corresponding to the lower blood TCDD levels associated with LOAELs and
NOAELs. The disposition of DLCs following exposures at background levels is  similarly not
well understood. Furthermore, there is uncertainty in the relationship of DLC tissue
concentrations to oral intakes in the current TEF approach. Finally, there is toxicological
uncertainty regarding several of the endpoints. Additional studies corroborating these outcomes
and their toxicological significance would further increase their utility in refining the TCDD
RfD.
       Overall,  EPA believes that the results of this analysis of alternative endpoints and PODs
increase the confidence in the TCDD RfD, both qualitatively and quantitatively. EPA's analyses
of some studies  show POD estimates higher than the RfD PODs—primarily those analyses that
consider background DLCs. Other analyses show POD estimates lower than the RfD POD, such
as the use of alternative age-adjusted background TCDD/DLC intake rates and some evaluations
of more uncertain endpoints (e.g., age at menopause endpoint  in Eskenazi et al. (2005)). The
more extreme values on the lower end are also the most uncertain, particularly with respect to the
contribution of TCDD relative to total TEQ.  In addition,  except for the male reproductive effects
in Mocarelli et al. (2011), determination of adversity for the lower LOAELs is problematic,
                                         4-99

-------
leading to lower confidence in the PODs.  The TCDD and TEQ LOAELs for semen quality in
males exposed in utero and by lactation (Mocarelli et al., 2011) are much lower than the
corresponding LOAELs for males exposed between ages 1 and 10 years (Mocarelli et al., 2008).
However, the NOAEL established for in utero and lactational exposure is fairly strong in the
qualitative sense; that is, there is fairly clear indication that semen quality is unaffected at the
corresponding dioxin exposure level.  Quantitatively, there is more uncertainty, but considering
background DLC exposure, the NOAEL is close to the RfD NOAEL benchmark.
                                        4-100

-------
                                  5.  REFERENCES
Abbott BD: Birnbaum. LS: Diliberto. JJ. (1996). Rapid Distribution of 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin (TCDD) to Embryonic Tissues in C57BL/6N Mice and
       Correlation with Palatal Uptakein Vitro. Toxicol Appl Pharmacol 141: 256-263.
Abraham, K; Krowke, R: Neubert D. (1988). Pharmacokinetics and biological activity of
       2,3,7,8-tetrachlorodibenzo-p-dioxin. 1. Dose-dependent tissue distribution and induction
       of hepatic ethoxyresorufm O-deethylase in rats following a single injection. Arch Toxicol
       62: 359-368.
Abraham, K; Knoll, A; Ende, M; Papke, O: Helge, H. (1996). Intake, fecal excretion, and body
       burden of polychlorinated dibenzo-p-dioxins and dibenzofurans in breast-fed and
       formula-fed infants. Pediatr Res 40: 671-679.
Abraham. K: Geusau. A: Tosun. Y: Helge. H: Bauer. S: Brockmoller. J.  (2002). Severe 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) intoxication: insights into the measurement of
       hepatic cytochrome P450 1A2 induction. Clin Pharmacol Ther 72: 163-174.
       http://dx.doi.org/10.1067/mcp.2002.126408.
Akhtar. FZ: Garabrant DH: Ketchum. NS: Michalek. JE. (2004).  Cancer in US Air Force
       veterans of the Vietnam war. J Occup Environ Med 46:  123.
Alaluusua, S:  Calderara, P; Gerthoux, PM; Lukinmaa, PL; Kovero, O: Needham, L; Patterson, J,
       r, D. G.;  Tuomisto, J: Mocarelli, P.  (2004). Developmental dental aberrations after the
       dioxin accident in Seveso. Environ  Health Perspect 112: 1313-1318.
Amin,  S: Moore, RW: Peterson, RE; Schantz, SL. (2000). Gestational and lactational exposure to
       TCDD or coplanar PCBs  alters adult expression of saccharin preference behavior in
       female rats. Neurotoxicol Teratol 22: 675-682. http://dx.doi.org/10.1016/S0892-
       0362(00)00094-5.
Anbalagan, J: Sashi, AM; Vengatesh, G: Stanley, JA; Neelamohan, R: Aruldhas, MM. (2010).
       Mechanism underlying transient gestational-onset hypothyroidism-induced impairment of
       posttesticular sperm maturation in adult rats. Fertil Steril 93: 2491-2497.
Andersen, ME; Mills, JJ: Gargas, ML; Kedderis, L; Birnbaum,  LS: Neubert, D: Greenlee, WF.
       (1993). Modeling receptor-mediated processes with dioxin: Implications for
       pharmacokinetics and risk assessment. Risk Anal 13: 25-36.
       http://dx.doi.0rg/10.llll/i.1539-6924.1993.tb00726.
Andersen. ME: Birnbaum. LS: Barton. HA: Eklund. CR. (1997). Regional hepatic CYP1A1 and
       CYP1A2 induction with 2,3,7,8-tetrachlorodibenzo-p-dioxin evaluated with a
       multicompartment geometric model of hepatic zonation. Toxicol Appl Pharmacol 144:
       145-155. http://dx.doi.org/10.1006/taap.1996.8067.
AT SDR (Agency for Toxic Substances and Disease Registry). (1998). Toxicol ogical profile for
       chlorinated dibenzo-p-dioxins (CDDs) [ATSDR Tox Profile]. Atlanta, GA: U.S.
       Department of Health and Human Services, Public Health Service.
       http://www.atsdr.cdc.gov/toxprofiles/tpl04.pdf
Auso. E: Lavado-Autric. R: Cuevas. E: Del Rev. FE: Morreale De Escobar. G: Serb el. P. (2004).
       A moderate and transient deficiency of maternal thyroid function at the beginning of fetal
       neocorticogenesis alters neuronal migration. Endocrinology  145: 4037-4047.
       http://dx.doi.org/10.1210/en.2004-0274.
                                          5-1

-------
Aylward, LL: Hays, SM. (2002). Temporal trends in human TCDD body burden: Decreases over
       three decades and implications for exposure levels. J Expo Anal Environ Epidemiol 12:
       319-328.
Aylward, LL: Brunei RC: Carrier, G: Hays,  SM: Gushing, CA: Needham, LL: Patterson, DG:
       Gerthoux, PM: Brambilla, P: Mocarelli, P. (2005a).  Concentration-dependent TCDD
       elimination kinetics in humans: Toxicokinetic modeling for moderately to highly exposed
       adults from Seveso, Italy, and Vienna, Austria, and impact on dose estimates for the
       NIOSH cohort. J Expo Anal Environ Epidemiol 15: 51-65.
       http://dx.doi.org/10.1038/si.iea.7500370.
Aylward, LL: Brunei RC: Starr. TB: Carrier. G: Delzell. E: Cheng. H: Beall. C. (2005b).
       Exposure reconstruction for the TCDD-exposed NIOSH cohort using a concentration-
       and age-dependent model of elimination. Risk Anal 25: 945-956.
       http://dx.doi.0rg/10.llll/i.1539-6924.2005.00645.x.
Aylward, LL: Goodman, JE: Charnley, G: Rhomberg, LR. (2008). A margin-of-exposure
       approach to assessment of noncancer risks of dioxins based on human exposure and
       response data. Environ Health Perspect 116: 1344-1351.
       http://dx.doi.org/10.1289/ehp.11514.
Baccarelli, A; Mocarelli, P: Patterson, DG, Jr: Bonzini, M: Pesatori, AC: Caporaso, N; Landi,
       MT. (2002). Immunologic effects of dioxin: new results from Seveso and comparison
       with other studies. Environ Health Perspect  110: 1169-1173.
Baccarelli, A; Pesatori, AC: Masten, SA: Patterson, DG, Jr: Needham, LL: Mocarelli, P:
       Caporaso, NE: Consonni, D: Grassman, JA: Bertazzi, PA; MT, L. (2004). Aryl-
       hydrocarbon receptor-dependent pathway and toxic  effects of TCDD in humans: a
       population-based study in Seveso, Italy. Toxicol Lett 149: 287-293.
       http://dx.doi.0rg/10.1016/i.toxlet.2003.12.062.
Baccarelli, A; Pesatori, AC: Consonni, D: Mocarelli, P: Patterson, DG, Jr: Caporaso, NE:
       Bertazzi, PA; Landi, MT. (2005). Health status and plasma dioxin levels in chloracne
       cases 20 years after the Seveso, Italy  accident. Br J Dermatol 152: 459-465.
       http://dx.doi.org/10.1111/J. 1365-2133.2005.06444.X.
Baccarelli, A; Hirt, C: Pesatori, AC: Consonni, D: Patterson, DG, Jr: Bertazzi, PA; Dolken, G:
       Landi, MT. (2006).  t(14;18) translocations in lymphocytes of healthy dioxin-exposed
       individuals from Seveso, Italy. Carcinogenesis 27: 2001-2007.
       http://dx.doi.org/10.1093/carcin/bgl011.
Baccarelli, A; Giacomini, SM: Corbetta, C: Landi, MT: Bonzini, M: Consonni, D: Grillo, P:
       Patterson, DG: Pesatori, AC: Bertazzi, PA. (2008). Neonatal thyroid  function in  Seveso
       25 years after maternal exposure to dioxin. PLoS Med 5: e!61.
Banks, YB: Brewster, DW: Birnbaum, LS. (1990). Age-related changes in dermal absorption of
       2,3,7,8-tetrachlorodibenzo-p-dioxin and 2,3,4,7,8-pentachlorodibenzofuran. Fundam
       Appl Toxicol 15:  163-173.
Banks. YB: Birnbaum. LS. (1991). Absorption of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
       after low dose dermal exposure. Toxicol Appl Pharmacol  107: 302-310.
       http://dx.doi.org/10.1016/0041-008X(91)90210-6.
Baron, JM: Zwadio-Klarwasser, G: Jugert, F; Hamann, W; Riibben, A; Mukhtar, H; Merk, HF.
       (1998). Cytochrome P450 1B1: A major P450 isoenzyme in human blood monocytes and
       macrophage subsets. Biochem Pharmacol 56: 1105-1110.
       http://dx.doi.org/10.1016/50006-2952(98)00105-1.

                                         5-2

-------
Becher, H; Flesch-Janys, D. (1998). Dioxins and furans: Epidemiologic assessment of cancer
       risks and other human health effects. Environ Health Perspect 106: 623-624.
Becher, H; Steindorf, K; Flesch-Janys, D. (1998). Quantitative cancer risk assessment for dioxins
       using an occupational cohort. Environ Health Perspect 106: 663-670.
Bell PR: Clode, S: Fan. MO: Fernandes, A: Foster. PM: Jiang. T: Loizou, G: MacNicoll A:
       Miller, BG: Rose, M; Tran, L; White, S. (2007a). Relationships between tissue levels of
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), mRNAs and toxicity in the developing
       male Wistar(Han) rat. Toxicol Sci 99: 591-604. http://dx.doi.org/10.1093/toxsci/kfml79.
Bell, DR; Clode, S; Fan, MQ; Fernandes, A; Foster, PM; Jiang, T; Loizou, G; MacNicoll, A;
       Miller, BG; Rose, M; Tran, L; White, S. (2007b). Toxicity of 2,3,7,8-tetrachlorodibenzo-
       p-dioxin in the developing male Wistar(Han) rat. II: Chronic dosing causes
       developmental delay. Toxicol Sci 99: 224-233. http://dx.doi.org/10.1093/toxsci/kfml41.
Bernert, JT;  Turner, WE; Patterson, DG; Needham, LL. (2007). Calculation of serum total lipid
       concentrations for the adjustment of persistent organohalogen toxicant measurements in
       human samples. Chemosphere 68: 824-831.
       http://dx.doi.0rg/10.1016/j.chemosphere.2007.02.043.
Birnbaum, LS. (1986). Distribution and excretion of 2,3,7,8-tetrachlorodibenzo-p-dioxin in
       congenic strains of mice which differ at the Ah locus. Drug Metab Dispos 14: 34-40.
Bonde, J, . P. E. Ernst E.,. Jensen. T... K. Hiollund, N... H. I. Kolstad, H... Scheike, T...
       Giwercman, A.,. Skakkebaek, N.,. E. Henriksen, T.,. B. Olsen, J.,. (1998). Relation
       between semen quality and fertility: A population-based study of 430 first-pregnancy
       planners. 352: 1172-1177. http://dx.doi.org/10.1016/s0140-6736(97)10514-1.
Boverhoff DR; Burgoon, LD; Tashiro, C; Chittim, B; Harkema, JR; Jump, DB; Zacharewski,
       TR. (2005). Temporal and dose-dependent hepatic gene expression patterns in mice
       provide new insights into TCDD-mediated hepatotoxicity. Toxicol Sci 85:  1048-1063.
       http://dx.doi.org/10.1093/toxsci/kfil62.
Bowman, RE; Schantz, SL;  Gross, ML; SA, F. (1989a). Behavioral effects in monkeys exposed
       to 2,3,7,8-TCDD transmitted maternally during gestation and for four months of nursing.
       Chemosphere 18: 235-242.
Bowman. RE: Schantz. SL:  Weerasinghe, NCA: Gross. ML: Barsotti, DA. Q989b). Chronic
       dietary intake of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) at 5 or 25 parts per trillion
       in the monkey: TCDD kinetics and dose-effect estimate of reproductive toxicity.
       Chemosphere 18: 243-252.
Budinsky, RA; Paustenbach, D; Fontaine, D; Landenberger, B; Starr, TB. (2006). Recommended
       relative potency factors for 2,3,4,7,8 pentachlorodibenzofuran: The impact of different
       dose metrics. Toxicol Sci 91: 275-285. http://dx.doi.org/10.1093/toxsci/kfj 125.
Budinsky, RA; LeCluyse, EL; Ferguson, SS; Rowlands, JC; Simon, T. (2010). Human and rat
       primary hepatocyte CYP1A1 and 1A2 induction with 2,3,7,8-tetrachlorodibenzo-p-
       dioxin, 2,3,7,8-tetrachlorodibenzofuran, and 2,3,4,7,8-pentachlorodibenzofuran. Toxicol
       Sci 118: 224-235. http://dx.doi.org/10.1093/toxsci/kfq238.
Buelke-Sam, J; Nelson, CJ;  Byrd, RA; Holson, JF. (1982a). Blood flow during pregnancy in the
       rat: I Flow patterns to maternal organs. Teratology 26: 269-277.
Buelke-Sam, J; Holson, JF; Nelson, CJ. (1982b). Blood flow during pregnancy in the rat: II
       Dynamics of and litter variability in uterine flow. Teratology 26: 279-288.
                                          5-2

-------
Burleson. GR: Lebrec. H: Yang. YG: Ibanes. JD: Pennington. KN: Birnbaum. LS. (1996). Effect
       of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on influenza virus host resistance in
       mice. Fundam Appl Toxicol 29: 40-47.
Calvo, RM; Jauniaux, E; Gulbis, B; Asuncion, M; Gervy, C: Contempre, B; Morreale De
       Escobar, G. (2002). Fetal tissues are exposed to biologically relevant free thyroxine
       concentrations during early phases of development. J Clin Endocrinol Metab 87: 1768-
       1777.
Cantoni, L; Salmona, M; Rizzardini, M. (1981). Porphyrogenic effect of chronic treatment with
       2,3,7,8-tetrachlorodibenzo-p-dioxin in female rats. Dose-effect relationship following
       urinary excretion of porphyrins. Toxicol Appl Pharmacol 57:  156-163.
Carrier, G: Brunei RC: Brodeur, J. (1995a). Modeling of the toxicokinetics of polychlorinated
       dibenzo-p-dioxins and dibenzofurans  in mammalians, including humans. I. Nonlinear
       distribution of PCDD/PCDF body burden between liver and adipose tissues. Toxicol
       Appl Pharmacol 131: 253-266. http://dx.doi.org/10.1006/taap.1995.1068.
Carrier, G: Brunei RC: Brodeur, J. (1995b). Modeling of the toxicokinetics of polychlorinated
       dibenzo-p-dioxins and dibenzofurans  in mammalians, including humans: II. Kinetics of
       absorption and disposition of PCDDs/PCDFs. Toxicol Appl Pharmacol 131: 267-276.
       http://dx.doi.org/10.1006/taap.1995.1069.
Chan, SY; Franklyn, JA; Kilby, MD.  (2005).  Maternal thyroid hormones and fetal brain
       development. Curr Opinion Endocrinol Diab 12: 23-30.
Cheng. H: Aylward. L: Beall C: Starr. TB: Brunei RC: Carrier. G: Delzell E. (2006). TCDD
       exposure-response analysis and risk assessment. Risk Anal 26: 1059-1071.
       http://dx.doi.0rg/10.llll/i.1539-6924.2006.00800.x.
Chu, I: Lecavalier, P; Hakansson,  H: Yagminas,  A; Valli, VE: Poon,  P: M, F. (2001). Mixture
       effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin and polychlorinated biphenyl congeners in
       rats. Chemosphere 43: 807-814. http://dx.doi.org/10.1016/S0045-6535(00)00437-9.
Chu. I: Valli.  VE:  Rousseaux. CG. (2007). Combined  effects of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin and polychlorinated biphenyl congeners in  rats. Toxicol Environ Chem 89: 71-87.
       http://dx.doi.org/10.1080/02772240600942548.
Clewell, HJ: Gentry, PR: Covington, TR: Sarangapani, R; Teeguarden, JG. (2004). Evaluation of
       the potential impact of age- and gender-specific pharmacokinetic differences on tissue
       dosimetry. Toxicol Sci 79: 381-393. http://dx.doi.org/10.1093/toxsci/kfhl09.
Collins. JJ: Bodner. K: Aylward. LL: Wilken. M: Bodnan CM. (2009). Mortality rates among
       trichlorophenol workers with exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Am J
       Epidemiol 170: 501-506. http://dx.doi.org/10.1093/aie/kwpl53.
Connor, KT; Aylward, LL. (2006). Human response to dioxin: Aryl hydrocarbon receptor (AhR)
       molecular structure, function,  and dose-response data for enzyme induction indicate an
       impaired human AhR. J Toxicol Environ Health B Crit Rev 9: 147-171.
       http://dx.doi.org/10.1080/15287390500196487.
Connor. KT: Harris. MA: Edwards. MR: Budinsky. RA: Clark. GC: Chu. AC: Finlev. BL:
       Rowlands, JC. (2008). AH receptor agonist activity in human blood measured with a cell-
       based  bioassay: evidence for naturally occurring AH receptor ligands in vivo. J Expo Sci
       Environ Epidemiol  18: 369-380. http://dx.doi.org/10.1038/si.ies.7500607.
Cooper, TG: Noonan, E: von Eckardstein, S:  Auger, J: Baker, HW: Behre, HM; Haugen, TB:
       Kruger. T: Wang. C: Mbizvo. MT: Vogelsong. KM. (2010). World Health Organization
                                          5-4

-------
       reference values for human semen characteristics. Hum Reprod Update 16: 231-245.
       http://dx.doi.org/10.1093/humupd/dmp048.
Crofton, KM; Craft ES: Hedge, JM; Gennings, C: Simmons, JE; Carchman, RA; Carter, WH, Jr;
       DeVito, MJ. (2005). Thyroid-hormone-disrupting chemicals: Evidence for dose-
       dependent additivity or synergism. Environ Health Perspect 113:  1549-1554.
Croutch. CR: Lebofsky. M: Schramm. KW: Terranova. PF: Rozman. KK. (2005). 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin (TCDD) and 1,2,3,4,7,8-hexachlorodibenzo-p-dioxin
       (HxCDD) alter body weight by decreasing insulin-like growth factor I (IGF-I) signaling.
       Toxicol Sci 85: 560-571. http://dx.doi.org/10.1093/toxsci/kfil06.
DeCaprio. AP: McMartin. DN: O'Keefe. PW: Rej. R: Silkworm. JB: Kaminsky. LS. (1986).
       Subchronic oral toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin in the guinea pig:
       Comparisons with a PCB-containing transformer fluid pyrolysate. Fundam Appl Toxicol
       6: 454-463. http://dx.doi.Org/10.1093/toxsci/6.3.454.
DeKoning, EP; Karmaus, W. (2000). PCB exposure in utero and via breast milk. A review
       [Review]. J Expo Anal Environ Epidemiol 10: 285-293.
Delange, F; Bourdoux, P; Ketelbant-Balasse, P; Humskerken, AV: Glinoer, D; Ermans, AM.
       (1983). Transient primary hypothyroidism in the newborn. In JH Dussault; P Walker
       (Eds.), Congenital hypothyroidism (pp. 275-301). New York, NY: Marcel Dekker.
Delia Porta, G: Dragani, TA; Sozzi, G. (1987). Carcinogenic effects of infantile and long-term
       2,3,7,8-tetrachlorodibenzo-p-dioxin treatment in the mouse. Tumori 73: 99-107.
DeVito, MJ: Ma, X; Babish, JG: Menache, M; Birnbaum, LS. (1994). Dose-response
       relationships in mice following subchronic exposure to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin: CYP1 Al, CYP1A2, estrogen receptor, and protein tyrosine phosphorylation.
       Toxicol Appl Pharmacol 124: 82-90.
Diliberto, JJ: Akubue, PI; Luebke, RW: Birnbaum, LS. (1995). Dose-response relationships of
       tissue distribution and induction of CYP1A1 and CYP1A2 enzymatic activities following
       acute exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in mice. Toxicol Appl
       Pharmacol 130: 197-208. http://dx.doi.org/lo'1006/taap. 1995.1025.
Diliberto, JJ: Jackson, JA; Birnbaum, LS. (1996). Comparison of 2,3,7,8-Tetrachlorodibenzo-p-
       dioxin (TCDD) Disposition Following Pulmonary, Oral, Dermal, and Parenteral
       Exposures to Rats. Toxicol Appl Pharmacol 138: 158-168.
Diliberto, JJ: Burgin, DE;  Birnbaum, LS. (1997). Role of CYP1A2 in hepatic sequestration of
       dioxin: Studies using CYP1A2 knock-out mice. Biochem Biophys Res Commun 236:
       431-433. http://dx.doi.org/10.1006/bbrc.1997.6973.
Diliberto. JJ: Burgin. DE:  Birnbaum. LS. (1999). Effects of CYP1A2 on Disposition of 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin, 2,3,4,7,8-Pentachlorodibenzofuran, and 2,2',4,4',5,5'-
       Hexachlorobiphenyl in CYP1A2 Knockout and Parental (C57BL/6N and 129/Sv) Strains
       of Mice.  Toxicol Appl Pharmacol 159: 52-64. http://dx.doi.org/10.1006/taap.1999.8720.
Diliberto. JJ: DeVito. MJ: Ross. DG: Birnbaum. LS. (2001). Subchronic Exposure of [3H]-
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in female B6C3F1 mice: Relationship of
       steady-state levels to disposition and metabolism. Toxicol  Sci 61: 241-255.
Dimitropoulos, A; Molinari, L; Etter, K; Lang-Muritano, M: Jenni, OG: Largo, RH; Latal, B.
       (2009). Children with congenital hypothyroidism: Long-term intellectual outcome after
       early high-dose treatment.  Pediatr Res 65: 242-248.
                                          5-5

-------
Eastin. WC: Haseman. JK: Mahler. JF: Bucher. JR. (1998). The National Toxicology Program
       evaluation of genetically altered mice as predictive models for identifying carcinogens.
       Toxicol Pathol 26: 461-473.
Emond, C: Birnbaum, LS: DeVito, MJ.  (2004). Physiologically based pharmacokinetic model for
       developmental exposures to TCDD in the rat. Toxicol Sci 80: 115-133.
       http://dx.doi.org/10.1093/toxsci/kfhll7.
Emond, C: Michalek, JE; Birnbaum, LS: DeVito, MJ. (2005). Comparison of the use of a
       physiologically based pharmacokinetic model and a classical pharmacokinetic model for
       dioxin exposure assessments. Environ Health Perspect 113: 1666-1668.
Emond, C: Birnbaum, LS: DeVito, MJ.  (2006). Use of a physiologically based pharmacokinetic
       model for rats to study the influence of body fat mass and induction of CYP1A2 on the
       pharmacokinetics of TCDD. Environ Health Perspect 114: 1394-1400.
Eskenazi, B; Mocarelli, P; Warner, M; Samuels, S: Vercellini, P; Olive, D: Needham, LL:
       Patterson DG, J, r; Brambilla, P; Gavoni, N; Casalini, S; Panazza, S; Turner, W;
       Gerthoux, PM. (2002a). Serum dioxin concentrations and endometriosis: A cohort study
       in Seveso, Italy. Environ Health Perspect 110: 629-634.
Eskenazi, B; Warner, M; Mocarelli, P: Samuels, S: Needham, LL: Patterson, DG, Jr: Lippman,
       S: Vercellini. P: Gerthoux. PM: Brambilla. P: Olive.  D. (2002b). Serum dioxin
       concentrations and menstrual cycle characteristics. Am J Epidemiol 156: 383-392.
Eskenazi, B: Mocarelli, P: Warner, M: Chee, WY: Gerthoux, PM: Samuels,  S: Needham, LL:
       Patterson, DG, Jr. (2003). Maternal serum dioxin levels and birth outcomes in women of
       Seveso, Italy. Environ Health Perspect  111: 947-953.
Eskenazi, B: Mocarelli, P: Warner, M: Needham, L: Patterson, DG, Jr: Samuels, S: Turner, W:
       Gerthoux, PM: Brambilla,  P. (2004). Relationship of serum TCDD concentrations and
       age at exposure of female residents of Seveso, Italy. Environ Health Perspect 112: 22-27'.
       http://dx.doi.org/10.1289/ehp.6573.
Eskenazi, B: Warner, M: Marks, AR: Samuels, S: Gerthoux, PM: Vercellini, P: Olive, PL;
       Needham, L; Patterson, D, Jr: Mocarelli, P.  (2005). Serum dioxin concentrations  and age
       at menopause. Environ Health Perspect 113: 858-862.
Eskenazi, B: Warner, M: Samuels, S: Young, J: Gerthoux, PM: Needham, L: Patterson, D: Olive,
       D: Gavoni, N: Vercellini, P: Mocarelli, P. (2007). Serum dioxin concentrations and risk
       of uterine leiomyoma in the Seveso Women's Health Study. Am J Epidemiol 166: 79-87.
       http://dx.doi.org/10.1093/aie/kwm048.
Evans, JS: Graham, JD: Gray, GM: Sielken, RL, Jr. (1994a). A distributional approach to
       characterizing low-dose cancer risk. Risk Anal 14: 25-34.
Evans. JS: Gray. GM: Sielken. RL: Smith. AE: Valdez-Flores, C: Graham. JD. (1994b).  Use of
       probabilistic expert judgment in uncertainty analysis of carcinogenic potency. Regul
       Toxicol Pharmacol 20: 15-36.
Fattore, E; Trossvik, C: Hakansson, H. (2000). Relative potency values derived from hepatic
       vitamin A reduction in male and female Sprague-Dawley rats following subchronic
       dietary exposure to individual polychlorinated dibenzo-p-dioxin and dibenzofuran
       congeners and a mixture therof. Toxicol Appl Pharmacol 165: 184-194.
       http://dx.doi.org/10.1006/taap.2000.8943.
Ferribv, LL: Knutsen, JS: Harris. M: Unice, KM: Scott P: Nony, P: Haws. LC: Paustenbach, D.
       (2007). Evaluation of PCDD/F and dioxin-like PCB serum concentration data from the
                                          5-6

-------
       2001-2002 National Health and Nutrition Examination Survey of the United States
       population. J Expo Sci Environ Epidemiol 17: 358-371.
Fisher. JW: Whittaker. TA: Taylor. DH: Clewell. HJ. Ill: Andersen. ME. (1989). Physiologically
       based pharmacokinetic modeling of the pregnant rat: A multiroute exposure model for
       trichloroethylene and its metabolite, trichloroacetic acid. Toxicol Appl Pharmacol 99:
       395-414. http://dx.doi.org/10.1016/0041-008X(89)90149-X.
Flesch-Janvs. D: Becher. H: Gurn. P: Jung. D: Konietzko. J: Manz. A: Papke. O. (1996).
       Elimination of polychlorinated dibenzo-p-dioxins and dibenzofurans in occupationally
       exposed persons. J Toxicol Environ Health 47: 363-378.
Flesch-Janys, D: Steindorf, K; Gurn, P; Becher, H. (1998). Estimation of the cumulated exposure
       to polychlorinated dibenzo-p-dioxins/furans and standardized mortality ratio analysis of
       cancer mortality by dose in an occupationally exposed cohort. Environ Health Perspect
       106: 655-662.
Fox, TR; Best LL: Goldsworthy, SM; Mills, JJ: Goldsworthy, TL.  (1993). Gene expression and
       cell proliferation in rat liver after 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure. Cancer
       Res 53: 2265-2271.
Franc. MA: Pohianvirta, R: Tuomisto, J: Okev, AB. (2001). Persistent, low-dose 2,3,7,8-
       tetrachlorodibenzo-p-dioxin exposure: effect on aryl hydrocarbon receptor expression in
       a dioxin-resistance model. Toxicol Appl Pharmacol 175: 43-53.
       http://dx.doi.org/10.1006/taap.2001.9222.
Franczak, A; Nynca, A; Valdez, KE; Mizinga, KM; Petroff, BK. (2006). Effects of acute and
       chronic exposure to the aryl hydrocarbon receptor agonist 2,3,7,8-tetrachlorodibenzo-p-
       dioxin on the transition to reproductive senescence in female Sprague-Dawley rats. Biol
       Reprod 74: 125-130. http://dx.doi.org/10.1095/biolreprod.105.044396.
Fujii-Kuriyama, Y; Etna, M; Mimura, J: Matsushita, N; Sogawa, K. (1995). Polymorphic forms
       of the Ah receptor and induction of the CYP1A1 gene. Pharmacogenetics 5 (S): 149-153.
Gasiewicz, TA: Henry, EC: Collins, LL. (2008). Expression and activity of aryl hydrocarbon
       receptors in development and cancer. Crit Rev Eukaryot Gene Expr 18: 279-321.
Geusau, A; Abraham, K: Geissler, K: Sator, MO: Stingl,  G: Tschachler, E. (2001). Severe
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) intoxication: Clinical and laboratory effects.
       Environ Health Perspect 109: 865-869.  http://dx.doi.org/10.1067/mcp.2002.126408.
Geusau, A; Schmaldienst S: Derfler, K. (2002). Severe 2,3,7,8-tetrachlorodibenzo-p-dioxin
       (TCDD) intoxication: Kinetics and trials to enhance elimination in two patients. Arch
       Toxicol 76: 316-325. http://dx.doi.org/10.1007/s00204-002-0345-7.
Geyer, H:  Scheunert, I; Korte, F. (1986). Bioconcentration potential of organic environmental
       chemicals in humans. Regul Toxicol Pharmacol 6: 313-347.
       http://dx.doi.org/10.1016/0273-2300(86)90002-4.
Glinoer, D: Delange, F. (2000). The potential repercussions of maternal, fetal, and neonatal
       hypothyroxinemia on the progeny. Thyroid 10: 871-887.
       http://dx.doi.org/10.1089/thy.2000.10.871.
Goodman, DG: Sauer, RM. (1992). Hepatotoxicity and carcinogenicity in female Sprague-
       Dawley rats treated with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD): a pathology
       working group reevaluation. Regul Toxicol Pharmacol 15: 245-252.
Goodman, M; Squibb, K: Youngstrom, E; Anthony, LG:  Kenworthy, L; Lipkin, PH; Mattison,
       PR; Lakind, JS. (2010). Using systematic reviews and meta-analyses to support
                                          5-7

-------
       regulatory decision making for neurotoxicants: Lessons learned from a case study of
       PCBs. Environ Health Perspect 118: 727-734. http://dx.doi.org/10.1289/ehp.0901835.
Grassman, JA; Needham, LL; Masten, SA; Patterson, D; Portier, CJ: Lucier, GW: Walker, NJ.
       (2000). Evidence of hepatic sequestration of dioxin in humans? An examination of tissue
       levels and CYP1A2 expression. Organohalogen Compounds 48: 87-90.
Greer, MA; Goodman, G:  Pleus, RC: Greer, SE. (2002). Health effects assessment for
       environmental perchlorate contamination: The dose response for inhibition of thyroidal
       radioiodine uptake in humans. Environ Health Perspect 110: 927-937.
Haddow, JE: Palomaki, GE: Allan. WC: Williams. JR: Knight GJ: Gagnon, J: O'Heir, CE:
       Mitchell ML: Hermos, RJ: Waisbren, SE: Faix, JD: Klein. RZ. (1999). Maternal thyroid
       deficiency during pregnancy and subsequent neuropsychological development of the
       child. NEnglJMed 341: 549-555. http://dx.doi.org/10.1056/NEJM199908193410801.
Hakk, H: Diliberto, JJ: Birnbaum, LS. (2009). The effect of dose on 2,3,7,8-TCDD tissue
       distribution, metabolism and elimination in CYP1A2 (-/-) knockout and C57BL/6N
       parental strains of mice. Toxicol Appl Pharmacol 241:  119-126.
Harper, PA; Wong, JY; Lam, MS: Okey, AB. (2002). Polymorphisms in the human AH receptor.
       Chem Biol Interact 141: 161-187.
Harrad, S: Wang, Y; Sandaradura, S: Leeds, A. (2003). Human dietary intake and excretion of
       dioxin-like compounds. J Environ Monit 5: 224-228. http://dx.doi.org/10.1039/b211406b.
Hassoun, EA: Wilt SC: Devito, MJ: Van Birgelen, A: Alsharif NZ: Birnbaum, LS: Stohs, SJ.
       (1998). Induction of oxidative stress in brain tissues of mice after subchronic exposure to
       2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Sci 42: 23-27.
       http://dx.doi.0rg/10.1093/toxsci/42.l.23.
Hassoun, EA: Li, F; Abushaban, A; Stohs, SJ.  (2000). The relative abilities of TCDD and its
       congeners to induce oxidative stress in the hepatic and brain tissues of rats after
       subchronic exposure. Toxicology 145: 103-113.
Hassoun, EA: Wang, H; Abushaban, A; Stohs, SJ. (2002). Induction of oxidative stress
       following chronic exposure to TCDD, 2,3,4,7,8-pentachlorodibenzofuran, and
       2,3',4,4',5-pentachlorobiphenyl. J Toxicol Environ Health A 65: 825-842.
Hassoun, EA: Al-Ghafri, M; Abushaban, A. (2003). The role of antioxidant enzymes in TCDD-
       induced oxidative stress in various brain regions of rats after subchronic exposure. Free
       Radic Biol Med 35: 1028-1036. http://dx.doi.org/10.1016/S0891-5849(03)00458-l.
Hattis,  D: Ginsberg. G:  Sonawane, B: Smolensk!.  S: Russ, A: Kozlak, M: Goble. R. (2003).
       Differences in pharmacokinetics between children and adults: II. Childrens variability in
       drug elimination half-lives and in some parameters needed for physiologically-based
       pharmacokinetic modeling. Risk Anal 23: 117-142. http://dx.doi.org/10.1111/1539-
       6924.00295.
Haws. LC: Su. SH: Harris. M:  Devito. MJ: Walker. NJ: Farland. WH: Finlev. B: Birnbaum. LS.
       (2006a). Development of a refined database of mammalian relative potency estimates for
       dioxin-like compounds. Toxicol Sci 89: 4-30.  http://dx.doi.org/10.1093/toxsci/kfi294.
Haws. LC: Su. SH: Harris. M:  DeVito. MJ: Walker. NJ: Farland. WH: Finlev. B: Birnbaum. LS.
       (2006b). Supplementary database II to article: Development of a refined database of
       mammalian relative potency estimates for dioxin-like compounds. Toxicol Sci 89: Sl-
       S48.
Hertz-Picciotto, I. (1995).  Epidemiology and quantitative risk assessment:  a bridge from science
       to policy. Am J Public Health 85: 484-491.

                                          5-8

-------
Hochstein, MS, Jr: Render, JA; Bursian, SJ: Aulerich, RJ. (2001). Chronic toxicity of dietary
       2,3,7,8-tetrachlorodibenzo-p-dioxin to mink. Vet Hum Toxicol 43: 134-139.
Hoio. R: Stern. S: Zareba. G: Markowskl VP: Cox. C: Kost JT: Weiss. B. (2002). Sexually
       dimorphic behavioral responses to prenatal dioxin exposure. Environ Health Perspect
       110:247-254.
Holland, CD: Sielken, RL, Jr. (1993). Quantitative cancer modeling and risk assessment.
Hong, R: Taylor, K; Abonour, R. (1989). Immune abnormalities associated with chronic TCDD
       exposure in rhesus. Chemosphere 18: 313-320. http://dx.doi.org/10.1016/0045-
       6535(89)90136-7.
Hurst CH: Abbott BD: DeVito. MJ: Birnbaum. LS. (1998). 2,3,7,8-Tetrachlorodibenzo-p-
       dioxin in Pregnant Long Evans Rats: Disposition to Maternal and Embryo/Fetal Tissues.
       45: 129-136.
Hurst CH: DeVito. MJ: Setzer. RW: Birnbaum. LS. (2000a). Acute administration of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) in pregnant Long Evans rats: association of
       measured tissue concentrations with developmental effects. Toxicol Sci 53: 411-420.
Hurst CH: DeVito. MJ: Birnbaum. LS. (2000b). Tissue disposition of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) in maternal and developing long-evans rats
       following subchronic exposure.  Toxicol Sci 57: 275-283.
Hurt KJ: Shi. Z: Albertini. DF: Petroff. BK. (2008). The environmental toxicant 2,3,7,8-
       tetrachlorodibenzo-p-dioxin disrupts morphogenesis of the rat pre-implantation embryo.
       BMC Dev Biol 8:  1-12. http://dx.doi.org/10.1186/1471-213X-8-l.
Ikeda, M; Tamura, M; Yamashita, J: Suzuki, C: Tomita, T. (2005). Repeated in utero and
       lactational 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure affects male gonads in offspring,
       leading to sex ratio changes in F2 progeny. Toxicol Appl Pharmacol 206: 351-355.
       http://dx.doi.0rg/10.1016/i.taap.2004.ll.019.
ILSI (International Life Sciences Institute). (1994). Physiological parameter values for PBPK
       models. Washington, DC: U.S. Environmental Protection Agency.
Ishihara, K; Warita, K; Tanida, T; Sugawara, T; Kitagawa, H; Hoshi, N. (2007). Does paternal
       exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) affect the sex ratio of offspring.
       J Vet Med Sci 69: 347-352.
Kattainen, H; Tuukkanen, J: Simanainen, U; Tuomisto, JT: Kovero, O: Lukinmaa, PL;
       Alaluusua. S: Tuomisto. J: Viluksela. M. (2001). In Utero/Lactational 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin Exposure Impairs Molar Tooth Development in Rats.
       Toxicol Appl Pharmacol 174: 216-224. http://dx.doi.org/10.1006/taap.2001.9216.
Keller, JM; Huet-Hudson, YM; Leamy, LJ. (2007). Qualitative effects of dioxin on molars vary
       among inbred mouse strains. Arch Oral Biol 52: 450-454.
       http://dx.doi.0rg/10.1016/i.archoralbio.2006.10.017.
Keller, JM: Huet-Hudson, Y; Leamy, LJ. (2008a). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       on molar development among non-resistant inbred strains of mice: A geometric
       morphometric analysis. Growth Development and Aging 71:  3-16.
Keller, JM: Zelditch, ML; Huet YM: Leamy, LJ. (2008b). Genetic differences in sensitivity to
       alterations of mandible structure caused by the teratogen 2,3,7,8-tetrachlorodibenzo-p-
       dioxin. Toxicol Pathol 36: 1006-1013. http://dx.doi.org/10.1177/0192623308327409.
Kerger. BD:  Leung. HW: Scott. P: Paustenbach. DJ: Needham. LL: Patterson. DG. Jr: Gerthoux.
       PM; Mocarelli, P. (2006). Age-  and concentration-dependent elimination half-life of
                                          5-9

-------
       2,3,7,8-tetrachlorodibenzo-p-dioxin in Seveso children. Environ Health Perspect 114:
       1596-1602. http://dx.doi.org/10.1289/ehp.8884.
Kerger, BD; Leung, HW: Scott PK; Paustenbach, DJ. (2007). Refinements on the age-dependent
       half-life model for estimating child body burdens of polychlorodibenzodioxins and
       dibenzofurans. Chemosphere 67: S272-S278.
       http://dx.doi.0rg/10.1016/j.chemosphere.2006.05.108.
Kim, AH; Kohn, MC: Nyska, A; Walker, NJ. (2003). Area under the curve as a dose metric for
       promotional responses following 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure. Toxicol
       Appl Pharmacol 191:  12-21. http://dx.doi.org/10.1016/S0041-008X0)3)00225-4.
Kitchin. KT: Woods.  JS. (1979). 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) effects on hepatic
       microsomal cytochrome P-448-mediated enzyme activities. Toxicol Appl Pharmacol 47:
       537-546. http://dx.doi.org/10.1016/0041-008X(79)90524-6.
Kociba. RJ: Keeler. PA: Park. CN: Gehring. PJ. (1976). 2,3,7,8-Tetrachlorodibenzo-p-dioxin
       (TCDD): Results of a 13-week oral toxicity study in rats. Toxicol Appl Pharmacol 35:
       553-574. http://dx.doi.org/10.1016/0041-008X(76)90078-8.
Kociba. RJ: Keves. DG: Beyer. JE: Carreon. RM: Wade. CE: Dittenber. DA: Kalnins. RP:
       Frauson. LE: Park. CN: Barnard. SD: Hummel RA: Humiston. CG. (1978). Results of a
       two-year chronic toxicity and oncogenicity study of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       in rats. Toxicol Appl Pharmacol 46: 279-303. http://dx.doi.org/10.1016/0041-
       008X(78)90075-3.
Kohn. MC: Lucier. GW: Clark. GC: Sewall. C:  Tritscher. AM: Portier. CJ. (1993). A
       mechanistic model of effects of dioxin on gene expression in the rat liver. Toxicol Appl
       Pharmacol 120: 138-154. http://dx.doi.org/10.1006/taap.1993.1096.
Kohn. MC: Sewall CH: Lucier. GW: Portier. CJ.  (1996). A mechanistic model of effects of
       dioxin on thyroid hormones in the rat. Toxicol Appl Pharmacol 165: 29-48.
       http://dx.doi.org/10.1006/taap.1996.0004.
Kohn, MC: Walker, NJ: Kim, AH; Portier, CJ. (2001). Physiological modeling of a proposed
       mechanism of enzyme induction by TCDD. Toxicology 162: 193-208.
       http://dx.doi.org/10.1016/S0300-483X(01)00363-8.
Kreuzer. PE: Csanadv. GA: Baur.  C: Kessler. W: Papke. O: Greim. H: Filser. JG. (1997).
       2,3,7,8-Tetrachlorodibenzo-p -dioxin (TCDD) and congeners in infants. A toxicokinetic
       model of human lifetime body burden by TCDD with special emphasis on its uptake by
       nutrition. Arch Toxicol 71: 383-400.
Krishnan, K; Andersen, ME. (1991). Interspecies scaling in pharmacokinetics. In A Rescingo; A
       Thakkur (Eds.), New trends in pharmacokinetics (pp. 203-226). New York, NY: Plenum
       Press.
Krowke, R: Chahoud, I; Baumann-Wilschke, I;  Neubert, D. (1989).  Pharmacokinetics and
       biological activity of 2,3,7,8-tetrachlorodibenzo-p-dioxin 2: pharmacokinetics in rats
       using a loading-dose/maintenance-dose regime with high doses. Arch Toxicol 63: 356-
       360. http://dx.doi.org/10.1007/BF00303123.
Kuchiiwa, S; Cheng, SB; Nagatomo, I; Akasaki, Y; Uchida, M; Tominaga, M; Hashiguchi, W:
       Kuchiiwa,  T. (2002). In utero and lactational exposure to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin decreases serotonin-immunoreactive neurons in raphe nuclei of male mouse
       offspring. Neurosci Lett 317: 73-76. http://dx.doi.org/10.1016/S0304-3940(01)02434-X.
LaKind. JS: Berlin. CM: Park. CN: Naiman. DO: Gudka. NJ. (2000). Methodology for
       characterizing distributions of incremental body burdens of 2,3,7,8-TCDD and DDE from

                                         5-10

-------
       breast milk in North American nursing infants. J Toxicol Environ Health A 59: 605-639.
       http://dx.doi.org/10.1080/009841000156628.
Lakshmanan. MR: Campbell BS: Chirtel SJ: Ekarohita. N: M. E. (1986). Studies on the
       mechanism of absorption and distribution of 2,3,7,8-tetrachlorodibenzo-p-dioxin in the
       rat. J Pharmacol Exp Ther 239: 673 -677'.
Landi, MT; Bertazzi, PA; Baccarelli, A; Consonni, D; Masten, S: Lucier, G: Mocarelli, P;
       Needham, L; Caporaso, N: Grassman, J. (2003). TCDD-mediated alterations in the AhR-
       dependent pathway in Seveso, Italy,  20 years after the accident. Carcinogenesis 24: 673-
       680. http://dx.doi.org/10.1093/carcin/bgg002.
Latchoumvcandane. C: Chitra. KC: Mathur. PP. (2002a). The effect of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin on the antioxidant system in mitochondrial and microsomal
       fractions of rattestis. Toxicology 171: 127-135. http://dx.doi.org/10.1016/S0300-
       483X0)1)00563-7.
Latchoumycandane, C: Mathur, PP.  (2002). Effects of vitamin E on reactive oxygen species-
       mediated 2,3,7,8-tetrachlorodibenzo-p-dioxin toxicity in rat testis. J Appl Toxicol 22:
       345-351. http://dx.doi.org/10.1002/jat.866.
Latchoumycandane, C: Chitra, C: Mathur, P. (2002b). Induction of oxidative stress in rat
       epididymal sperm after exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Arch Toxicol
       76: 113-118. http://dx.doi.org/10.1007/s00204-001-0308-4.
Latchoumycandane, C: Chitra, KC: Mathur, PP. (2003). 2,3,7,8-Tetrachlorodibenzo-p-dioxin
       (TCDD) induces oxidative stress in the epididymis and epididymal sperm of adult rats.
       Arch Toxicol 77: 280-284. http://dx.doi.org/10.1007/s00204-003-0439-x.
Lavado-Autric, R: Auso, E: Garcia-Velasco, JV: Del Carmen Arufe, M; Escobar Del Rev, F;
       Serb el, P: Morreale De Escobar, G. (2003). Early maternal hypothyroxinemia alters
       histogenesis and cerebral cortex cytoarchitecture of the progeny. J Clin Invest 111: 1073-
       1082. http://dx.doi.org/10.1172/JCI16262.
Lawrence, GS: Gobas, FAP, C. (1997). A pharmacokinetic analysis of interspecies extrapolation
       in dioxin risk assessment. Chemosphere 35:  427-452. http://dx.doi.org/10.1016/S0045-
       6535(97)00108-2.
Lean, MEJ: Han, TS: Deurenberg, P. (1996). Predicting body composition by densitometry from
       simple anthropometric measurements. Am J Clin Nutr 63: 4-14.
Leo, A; Hansch, C: Elkins, D. (1971). Partition coefficients and their uses [Review]. Chem Rev
       71: 525-616. http://dx.doi.org/10.1021/cr60274a001.
Leung, HW: Ku, RH; Paustenbach, DJ: Andersen, ME. (1988). A physiologically based
       pharmacokinetic model for 2,3,7,8-tetrachlorodibenzo-p-dioxin in C57BL/6J and DBA/2J
       mice. Toxicol Lett 42: 15-28.
Leung, HW: Poland, A; Paustenbach, DJ: Murray, FJ: Andersen, ME. (1990). Pharmacokinetics
       of [125I]-2-iodo-3,7,8-trichlorodibenzo-p-dioxin in mice: analysis with a physiological
       modeling approach. Toxicol Appl Pharmacol 103: 411-419.
Leung, HW: Kerger, BD; Paustenbach, DJ. (2006). Elimination half-lives of selected
       polychlorinated dibenzodioxins and dibenzofurans in breast-fed human infants. J Toxicol
       Environ Health A 69: 437-443. http://dx.doi.org/10.1080/15287390500246886.
Li, B; Liu, HY; Dai, LJ: Lu, JC: Yang, ZM; Huang, L. (2006). The early embryo loss caused by
       2,3,7,8-tetrachlorodibenzo-p-dioxin may be related to the accumulation of this compound
       in the uterus. Reprod Toxicol 21: 301-306.
       http://dx.doi.0rg/10.1016/j.reprotox.2005.09.008.

                                         5-11

-------
Li. X: Johnson. DC: Roztnan. KK. (1997). 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
       increases release of luteinizing hormone and follicle-stimulating hormone from the
       pituitary of immature female rats in vivo and in vitro. Toxicol Appl Pharmacol 142: 264-
       269. http://dx.doi.org/10.1006/taap.1996.8044.
Longnecker, MP: Michalek, JE. (2000). Serum Dioxin level in relation to Diabetes Mellitus
       among Air Force veterans with background levels of exposure. Epidemiology 11: 44-48.
Lorber, M. (2002). A pharmacokinetic model for estimating exposure of Americans to dioxin-
       like compounds in the past, present, and future. Sci  Total Environ 288: 81-95.
Lorber, M; Patterson, D; Huwe, J: Kahn, H. (2009). Evaluation of background exposures of
       Americans to dioxin-like compounds in the 1990s and the 2000s. Chemosphere 77: 640-
       651. http://dx.doi.0rg/10.1016/i.chemosphere.2009.08.016.
Lorenzen, A; Okey, AB. (1991). Detection and characterization of Ah receptor in tissue  and cells
       from human tonsils. Toxicol  Appl Pharmacol 107: 203-214.
Lucier, GW: Rumbaugh, RC: McCoy. Z: Hass, R: Harvan,  D: Albro, P. (1986). Ingestion of soil
       contaminated with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) alters hepatic enzyme
       activities in rats. Fundam Appl Toxicol 6: 364-371.
Mally, A; Chipman, JK. (2002). Non-genotoxic carcinogens: Early effects on gap junctions, cell
       proliferation and apoptosis in the rat. Toxicology 180: 233-248.
Manchester.  DK: Gordon. SK: Golas, CL: Roberts. EA: Okev, AB. (1987). Ah receptor in
       human placenta: stabilization by molydate and characterization of binding of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin,  3-methylcholanthrene, and benzo(a)pyrene. Cancer Res 47:
       4861-4868.
Markowski, VP; Zareba, G: Stern, S; Cox, C; Weiss, B. (2001). Altered operant responding for
       motor reinforcement and the  determination of benchmark doses following perinatal
       exposure to low-level 2,3,7,8-tetrachlorodibenzo-p-dioxin. Environ Health Perspect 109:
       621-627.
Maronpot RR: Foley, JF;  Takahashi, K; Goldsworthy, T; Clark, G: Tritscher, A; Portier, C:
       Lucier, G. (1993). Dose response for TCDD promotion of hepatocarcinogenesis in rats
       initiated with DEN: histologic, biochemical, and cell proliferation endpoints. Environ
       Health Perspect 101: 643-642.
Maruyama, W; Yoshida, K; Tanaka, T; Nakanishi, J. (2002). Determination of tissue-blood
       partition coefficients for a physiological model for humans, and estimation of dioxin
       concentration in tissues. Chemosphere 46: 975-985.
McNulty, WP: Nielsen-Smith, KA; Lay, JO, Jr; Lippstreu,  PL; Kangas, NL; Lyon, PA; Gross,
       ML. (1982). Persistence of TCDD in monkey adipose tissue. Food Chem Toxicol 20:
       985-986.
Michalek. JE: Akhtar. FZ: Longnecker.  MP: Burton. JE.  (2001a). Relation of serum 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) level to hematological examination results in
       veterans of Operation Ranch  Hand.  Arch Environ Health 56: 396-405.
       http://dx.doi.org/10.1080/00039890109604474.
Michalek, JE: Ketchum, NS: Longnecker, MP. (200Ib). Serum dioxin and hepatic abnormalities
       in veterans of Operation Ranch Hand. Ann Epidemiol 11: 304-311.
       http://dx.doi. org/10.1016/51047-2797(00)00218-0.
Michalek, JE: Akhtar, FZ: Arezzo, JC: Garabrant, DH; Albers, JW. (200Ic). Serum dioxin and
       peripheral neuropathy in veterans of Operation Ranch Hand. Neurotoxicology 22: 479-
       490.

                                         5-12

-------
Michalek, JE: Pirkle, JL: Needham, LL: Patterson. DG, Jr: Caudill. SP: Tripathi, RC: Mocarelli,
       R (2002). Pharmacokinetics of 2,3,7,8-tetrachlorodibenzo-p-dioxin in Seveso adults and
       veterans of operation Ranch Hand. J Expo Anal Environ Epidemiol 12: 44-53.
Michalek, JE: Pavuk, M. (2008). Diabetes and cancer in veterans of Operation Ranch Hand after
       adjustment for calendar period, days of spraying, and time spent in Southeast Asia.  J
       Occup Environ Med 50: 330-340. http://dx.doi.org/10.1097/JOM.Ob013e31815f889b.
Micka, J: Milatovich, A; Menon, A; Grabowski, GA: Puga, A; Nebert, DW.  (1997). Human Ah
       receptor (AHR) gene: Localization to 7pl5 and suggestive correlation of polymorphism
       with CYP1A1 inducibility. Pharmacogenetics 7:  95-101.
Miettinen, HM; Sorvari, R Alaluusua, S; Murtomaa, M; Tuukkanen, J: Viluksela, M. (2006).
       The Effect of Perinatal TCDD exposure on caries susceptibility in rats. Toxicol Sci  91:
       568-575. http://dx.doi.org/10.1093/toxsci/kfil58.
Milbrath, MO: Wenger, Y: Chang. CW: Emond, C: Garabrant D: Gillespie,  BW: Jolliet O.
       (2009). Apparent half-lives of dioxins, furans, and polychlorinated biphenyls as a
       function of age, body fat, smoking status, and breast-feeding. Environ Health Perspect
       117: 417-425. http://dx.doi.org/10.1289/ehp.11781.
Mocarelli, P. (2000). Dioxin and human sex ratio: The Seveso case. Milan, Italy: Department of
       Laboratory Medicine Hospital of Desio.
Mocarelli, P; Gerthoux, PM: Ferrari,  E: Patterson Jr, DG: Kieszak, SM: Brambilla, P; Vincoli,
       N; Signorini, S: Tramacere, P; Carreri, V: Sampson, EJ:  Turner, WE. (2000). Paternal
       concentrations of dioxin and sex ratio of offspring. Lancet 355: 1858-1863.
       http://dx.doi.org/10.1016/S0140-6736(00)02290-X.
Mocarelli, P. (2001). Seveso: A teaching story. Chemosphere 43: 391-402.
       http://dx.doi. org/10.1016/S0045-6535(00)00386-6.
Mocarelli, P: Gerthoux, PM: Patterson, DG, Jr: Milani, S: Limonata, G: Bertona, M: Signorini,
       S: Tramacere, P: Colombo, L: Crespi, C: Brambilla, P: Sarto, C: Carreri, V: Sampson,
       EJ: Turner, WE; Needham, LL. (2008). Dioxin exposure, from infancy through puberty,
       produces endocrine disruption and affects human semen  quality. Environ Health Perspect
       116: 70-77. http://dx.doi.org/10.1289/ehp.10399.
Mocarelli, P: Gerthoux, PM: Needham, LL: Patterson, DG,  Jr: Limonta, G: Falbo, R: Signorini,
       S: Bertona, M:  Crespi, C: Sarto, C: Scott, PK: Turner, WE: Brambilla, P. (2011).
       Perinatal exposure to low doses of dioxin can permanently impair human semen quality.
       Environ Health Perspect 119: 713-718.
Morreale de Escobar, G: Obregon, MJ: Escobar del Ray, F.  (2000). Is neuropsychological
       development related to maternal hypothyroidism or to maternal hypothyroxinemia? J Clin
       Endocrinol Metab 85: 3975-3987. http://dx.doi.org/10.1210/ic.85.ll.3975.
Moser, GA: McLachlan, MS. (2001). The influence of dietary concentration on the absorption
       and excretion of persistent lipophilic organic pollutants in the human intestinal tract.
       Chemosphere 45:201-211. http://dx.doi.org/10.1016/50045-6535(00)00551-8.
Murray, FJ: Smith, FA: Nitschke, KD: Humiston, CG: Kociba, RJ: Schwetz, BA. (1979). Three-
       generation reproduction study of rats given 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
       in the diet. Toxicol Appl Pharmacol 50: 241-252. http://dx.doi.org/10.1016/0041-
       008X(79)90149-2.
NAS (National Academy of Sciences). (2006a). Health risks from dioxin and related compounds.
       Available online at http://www.nap.edu/webcast/webcast detail.php?webcastid=328.
       (accessed February 9, 2010).

                                          5-13

-------
NAS (National Academy of Sciences). (2006b). Health risks from dioxin and related
       compounds: Evaluation of the EPA reassessment. Washington, DC: National Academy
       Press, http://www.nap.edu/catalog.php7record id=l 1688.
Nebert, DW: Peterson, DP; Puga, A. (1991). Human AH locus polymorphism and cancer:
       Inducibility of CYP1A1 and other genes by combustion products and dioxin.
       Pharmacogenetics 1: 68-78.
Needham. LL: Gerthoux. PM: Patterson. DG: Brambilla. P: Turner. WE: Beretta. C: Pirkle. JL:
       Colombo, L; Sampson, EJ: Tramacere, PL; Signorini, S: Meazza, L; Carreri, V: Jackson,
       RJ: Mocarelli, P. (1997). Serum dioxin levels in Seveso, Italy, population in 1976.
       Teratog Carcinog Mutagen 17: 225-240. http://dx.doi.org/10.1002/(SICI)1520-
       6866Q997)17:4/5<225::AID-TCM5>3.0.CO:2-K.
Nessel, CS: Amoruso, MA; Umbreit TH; Meeker, RJ: Gallo, MA. (1992). Transpulmonary
       uptake and bioavailability of 2,3,7,8-TCDD from respirable soil particles. Chemosphere
       25: 29-32.
Niskar, A; Needham, LL: Rubin, C: Turner, WE; Martin, CA: Patterson, DG, Jr: Hasty, L:
       Wong, LY; Marcus, M. (2009). Serum dioxin, polychlorinated biphenyls, and
       endometriosis: A case-control study in Atlanta. Chemosphere 74: 944-949.
       http://dx.doi.0rg/10.1016/j.chemosphere.2008.10.005.
Nohara, K: Fujimaki, H: Tsukumo, S: Ushio, H: Miyabara, Y; Kijima, M: Tohyama, C:
       Yonemoto, J. (2000). The effects of perinatal exposure to low  doses of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin on immune organs in rats. Toxicology 154: 123-133.
       http://dx.doi.org/10.1016/S0300-483X(00)00323-1.
Nohara. K: Izumi. H: Tamura. S: Nagata. R: Tohyama. C. (2002). Effect of low-dose 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) on influenza A virus-induced mortality in mice.
       Toxicology 170: 131-138.
Nolan. KJ: Smith. FA: Hefner. JG. (1979). Elimination and tissue distribution of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) in female guinea pigs following a single oral dose
       [Abstract]. Toxicol Appl Pharmacol 48: A162. http://dx.doi.org/10.1016/0041-
       008X(79)90509-X.
NTP (National Toxicology Program). (1982a). Carcinogenesis bioassay of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin in Swiss-Webster mice (dermal study). RTF, NC: U.S.
       Department of Health and Human Services.
NTP (National Toxicology Program). (1982b). Carcinogenesis bioassay of BIS(2-chloro-l-
       methylethyl) ether ( 70%) (CAS no. 108-60-1) containing 2-chloro-l-methylethyl(2-
       chloropropyl) ether ( 30%) (CAS no. 83270-31-9) in B6C3F1 mice (gavage study).
       (NTP-81-55). Research Triangle Park, NC and Bethesda, MD.
NTP (National Toxicology Program). (1982c). NTP Technical Report on Carcinogenesis
       bioassay of 2,3,7,8-tetrachlorodibenzo-p-dioxin in Osborne-Mendel rats and B6C3F1
       mice (gavage study). Research Triangle Park, NC.
       http://ntp.niehs.nih.gov/?obiectid=07060172-DEB2-6542-D7CD537BAB5B2ACD.
NTP (National Toxicology Program). (2006a). NTP technical report on the toxicology and
       Carcinogenesis studies of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (CAS no. 1746-
       01-6) in female harlan Sprague-Dawley rats (gavage studies). (NTP TR 521; NIH
       Publication No. 06-4468). Research Triangle Park, NC.
NTP (National Toxicology Program). (2006b). NTP toxicology and Carcinogenesis studies of a
       binary mixture of 3,3',4,4',5-Pentachlorobiphenyl (PCB 126) (CAS No.  57465-28-8) and

                                         5-14

-------
       2,2',4,4',5,5'-Hexachlorobiphenyl (PCB 153) (CAS No. 35065-27-1) in female Harlan
       Sprague-Dawley rats (Gavage studies) (pp. 1-258). (NTP TR 530). Research Triangle
       Park, NC. http://ntp.niehs.nih.gov/files/TR530 Webl.pdf.
NTP (National Toxicology Program). (2006c). Toxicology and carcinogenesis studies of a
       mixture of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (CAS No. 1746-01-6), 2,3,4,7,8-
       pentachlorodibenzofuran (PeCDF) (CAS No. 57117-31-4), and 3,3',4,4',5-
       pentachlorobiphenyl (PCB 126) (CAS No. 57465-28-8) in female Harlan Sprague-
       Dawley rats (gavage studies). (NTP TR 526). Research Triangle Park, NC.
       http://ntp.niehs.nih.gov/index.cfm?obiectid=070B7300-OE62-BF12-
       F4C3E3B5B645A92B.
Oerbeck, B; Sundet K; Kase, BF; Heyerdahl, S. (2005). Congenital hypothyroidism: No adverse
       effects of high dose thyroxine treatment on adult memory, attention, and behaviour. Arch
       Dis Child 90: 132-137. http://dx.doi.org/10.1136/adc.2003.043935.
Ogura, I: Masunaga, S: Nakanishi, J. (2004). Quantitative source identification of dioxin-like
       PCBs in Yokohama, Japan, by temperature dependence of their atmospheric
       concentrations. Environ Sci Technol 38: 3279-3285.
       http://dx.doi.org/10.1021/es0354622.
Ohsako, S: Miyabara, Y; Nishimura, N; Kurosawa, S: Sakaue, M; Ishimura, R: Sato, M; Takeda,
       K: Aoki, Y; Sone, H; Tohyama, C: Yonemoto, J. (2001). Maternal exposure to a low
       dose of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) suppressed the development of
       reproductive organs of male  rats: Dose-dependent increase of mRNA levels of 5a-
       reductase type 2 in contrast to descrease of androgen receptor in the pubertal ventral
       prostate. Toxicol Sci 60: 132-143.
Okev. AB: Riddick. PS: Harper. PA. (1994). The Ah receptor: Mediator of the toxicity of
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and related compounds. Toxicol Lett 70: 1-
       22. http://dx.doi.org/10.1016/0378-4274(94)90139-2.
Olson, JR; Holscher, MA; Neal, RA. (1980). Toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin in
       the Golden Syrian hamster. Toxicol Appl Pharmacol 55: 67-78.
       http://dx.doi. org/10.1016/0041-008X(80)90221-5.
Olson. JR: McGarrigle. BP: Gigliottl PJ: Kumar. S: McRevnolds. JH. (1994). Hepatic uptake
       and metabolism of 2,3,7,8-tetrachlorodibenzo-p-dioxin and 2,3,7,8-
       tetrachlorodibenzofuran. Fundam Appl Toxicol 22: 631-640.
       http://dx.doi.org/10.1006/faat.1994.1069.
Ott MG: Zober, A. (1996). Morbidity study of extruder personnel with potential exposure to
       brominated dioxins and furans. II. Results of clinical laboratory studies. Occup Environ
       Med53: 844-846.
Patterson. DG. Jr: Wong. LY: Turner. WE: Caudill. SP: Dipietro. ES: McClure. PC: Cash. TP:
       Osterloh. JD: Pirkle. JL: Sampson. EJ: Needham. LL. (2009). Levels in the U.S.
       population of those persistent organic pollutants (2003-2004) included in the Stockholm
       Convention or in other long range transboundary air pollution agreements. Environ Sci
       Technol 43: 1211-1218. http://dx.doi.org/10.1021/es801966w.
Pavuk. M: Schecter. AJ: Akhtar. FZ: Michalek. JE. (2003). Serum 2,3,7,8-tetrachlorodibenzo-p-
       dioxin (TCDD) levels and thyroid function in Air Force veterans of the Vietnam War.
       Ann Epidemiol 13: 335-343. http://dx.doi.org/10.1016/S1047-2797(02)00422-2.
Pinsky. PF: Lorber. MN. (1998). A model to evaluate past exposure to 2,3,7,8-TCDD. J Expo
       Anal Environ Epidemiol 8: 187-206.

                                          5-15

-------
Pipe. NG: Smith. T: Hallidav. D: Edmonds. CJ: Williams. C: Coltart TM. (1979). Changes in
       fat, fat-free mass and body water in human normal pregnancy. Br J Obstet Gynaecol 86:
       929-940. http://dx.doi.0rg/10.llll/i.1471-0528.1979.tbll240.x.
Poiger. H: Schlatter. C.  (1986). Pharmacokinetics of 2,3,7,8-TCDD in man.  Chemosphere 15:
       1489-1494. http://dx.doi.org/10.1016/0045-6535(86)90429-7.
Poulin, P; Theil, FP. (2001). Prediction of pharmicokinetics prior to in vivo  studies. 1.
       mechanism-based prediction of volume of distribution. J Pharm Sci 91: 129-156.
       http://dx.doi.org/10.1002/jps.10005.
Ramsey. JC: Hefner. JG: Karbowskl RJ: Braun. WH: Gehring. PJ. (1982). The in vivo
       biotransformation of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in the rat. Toxicol
       Appl Pharmacol 65: 180-184. http://dx.doi.org/10.1016/0041-008X(82)90377-5.
Rao. MS:  Subbarao. V:  Prasad. JD: Scarpelll DG. (1988). Carcinogenicity of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin in the Syrian golden hamster. Carcinogenesis 6: 1677-1679.
Reddy, M; Yang, R; Clewell, HJ: Andersen, ME. (2005). Physiologically based pharmacokinetic
       modeling:  Science and applications. Hoboken, New Jersey: John Wiley & Sons.
Rigon, F; Bianchin, L; Bernasconi, S: Bona, G: Bozzola, M; Buzi, F; Cicognani, A; De Sanctis,
       C: De Sanctis, V: Radetti, G: Tato, L; Tonini, G: Perissinotto, E. (2010). Update on age
       at menarche in Italy: Toward the leveling off of the secular trend. J Adolesc Health 46:
       238-244. http://dx.doi.0rg/10.1016/i.iadohealth.2009.07.009.
Roberts, EA: Shear, NH; Okey, AB: Manchester, DK. (1985). The Ah receptor and dioxin
       toxicity: From rodent to human tissues. Chemosphere 14: 661-674.
       http://dx.doi.org/10.1016/0045-6535(85)90174-2.
Roberts, EA: Golas, CL: Okey, AB.  (1986). Ah receptor mediating induction of aryl
       hydrocarbon hydroxylase: Detection in human lung by binding of 2,3,7,8-
       [H]tetrachlorodibenzo-p-dioxin. Cancer Res 46: 3739-3743.
Rohde. S: Moser. GA: Papke. O: McLachlan. MS. (1999). Clearance of PCDD/Fs via the
       gastrointestinal tract in occupationally exposed persons. Chemosphere 38: 3397-3410.
       http://dx.doi.org/10.1016/50045-6535(98)00551-7.
Roth, WL; Ernst S: Weber, LWD: Kerescen, L: Rozman, KK. (1994). A pharmacodynamically
       responsive model of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) transfer between liver
       and fat at low and high doses. Toxicol Appl Pharmacol 127: 151-162.
       http://dx.doi.org/10.1006/taap.1994.1149.
Rovet JF. (2002). Congenital hypothyroidism: An analysis of persisting deficits and associated
       factors. ChildNeuropsychol 8:  150-162. http://dx.doi.Org/10.1076/chin.8.3.150.13501.
Roy, T: Hammerstrom,  K; Schaum, J. (2008). Percutaneous absorption of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) from soil. J Toxicol Environ Health A 71: 1509-
       1515. http://dx.doi.org/10.1080/15287390802349875.
Royland, JE; Parker, JS: Gilbert, ME. (2008). A genomic analysis of subclinical hypothyroidism
       in hippocampus  and neocortex of the developing rat brain. J Neuroendocrinol 20:  1319-
       1338. http://dx.doi.0rg/10.llll/i.1365-2826.2008.01793.x.
SAB (U.S. EPA Science Advisory Board).  (2011). SAB review of EPA's reanalysis of key issues
       related to dioxin toxicity and response to NAS comments. (EPA-SAB-011-014).
       Washington, DC: U.S. Environmental Protection Agency.
       http://vosemite.epa.gov/sab/sabproduct.nsf/2A45B492EB AA8553852578F9003ECBC5/$
       File/EPA-SAB-1 l-014-unsigned.pdf.
                                         5-16

-------
Santostefano, MJ: Johnson, KL; Whisnant NA; Richardson, VM: Devito, MJ: Birnbaum, LS.
       (1996). Subcellular localization of TCDD differs between the liver, lungs, and kidneys
       after acute and subchronic exposure: Species/dose comparison and possible mechanism.
       Fundam Appl Toxicol 34: 365-375.
Santostefano. MJ: Wang. X: Richardson. VM: Ross. DG: DeVito. MJ: Birnbaum. LF. (1998). A
       pharmacodynamic analysis of TCDD-Induced Cytochrome 450 gene expression in
       multiple tissues: Dose and time-dependent effects. Toxicol Appl Pharmacol 151: 294-
       310.
Savin, S: Cvejic, D: Nedic, O: Radosavljevic, R. (2003). Thyroid hormone synthesis and storage
       in the thyroid gland of human neonates. J Pediatr Endocrinol Metab 16: 521-528.
Schantz, SL: Laughlin, NK: Van Valkenberg, HC: Bowman, RE. (1986). Maternal care by
       rhesus monkeys of infant monkeys exposed to either lead or 2,3,7,8-tetrachlorodibenzo-
       P-dioxin. Neurotoxicology 7: 637-650.
Schantz, SL: Bowman, RE. (1989). Learning in monkeys exposed perinatally to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD). Neurotoxicol Teratol 11: 13-19.
       http://dx.doi.org/10.1016/0892-0362(89)90080-9.
Schantz, SL: Ferguson, SA: Bowman, RE. (1992). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       on behavior of monkeys in peer groups. Neurotoxicol Teratol 14: 433-446.
Schantz. SL: Seo. BW: Moshtaghian. J: Peterson. RE: Moore. RW. (1996). Effects of gestational
       and lactational exposure to TCDD or coplanar PCBs on spatial learning. Neurotoxicol
       Teratol 18: 305-313.
Seo. BW: Li. MH: Hansen. LG: Moore. RW: Peterson. RE: Schantz. SL. (1995). Effects of
       gestational and lactational exposure to coplanar polychlorinated biphenyl (PCB)
       congeners or 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on thyroid hormone
       concentrations in weanling rats. Toxicol Lett 78: 253-262.
Sewall, C: Lucier, G: Tritscher, A; Clark, G. (1993). TCDD-mediated changes in hepatic
       epidermal growth factor receptor  may be a critical event in the hepatocarcinogenic action
       of TCDD. Carcinogenesis 14: 1885-1893.
Sewall CH: Flagler, N: Vanden Heuvel  JP: Clark. GC: Tritscher. AM: Maronpot RM: Lucier.
       GW. (1995). Alterations in thyroid function in female Sprague-Dawley rats following
       chronic treatment with 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Appl Pharmacol
       132: 237-244.
Sharlin, DS: Tighe, D: Gilbert, ME; Zoeller, RT. (2008). The balance between oligodendrocyte
       and astrocyte production in major white matter tracts is linearly related to serum total
       thyroxine. Endocrinology 149: 2527-2536. http://dx.doi.org/10.1210/en.2007-1431.
Sharlin. DS: Gilbert. ME: Taylor. MA: Ferguson. DC: Zoeller. RT. (2010). The nature of the
       compensatory response to low thyroid hormone in the developing brain. J
       Neuroendocrinol 22: 153-165. http://dx.doi.Org/10.llll/i.1365-2826.2009.01947.x.
Shi. Z: Valdez, KE: Ting. AY: Franczak, A: Gum. SL: Petroff, BK. (2007). Ovarian endocrine
       disruption underlies premature reproductive senescence following environmentally
       relevant chronic exposure to the aryl hydrocarbon receptor agonist 2,3,7,8-
       tetrachlorodibenzo-p-dioxin. Biol Reprod 76: 198-202.
       http://dx.doi.org/10.1095/biolreprod.106.053991.
Shu. H: Teitelbaum, P: Webb. AS: Marple, L: Brunck, B: Dei Rossi. D: Murray. FJ:
       Paustenbach, D. (1988). Bioavailability of soil-bound TCDD: Dermal bioavailability in
                                         5-17

-------
       the rat. Fundam Appl Toxicol 2: 335-343. http://dx.doi.org/10.1016/0272-
       0590(88)90319-3.
Sielken, RL, Jr.  (1993). Evaluation of chloroform risk to humans.The Toxicology Forum, 1993
       Annual Winter Meeting, February 15-17, 1993, The Capitol Hilton, Washington, DC.
Sielken, RL, Jr:  Valdez-Flores, C. (1996). Comprehensive realism's weight-of-evidence based
       distributional dose response characterization. Hum Ecol Risk Assess 2: 175-193.
       http://dx.doi.org/10.1080/10807039.1996.10387467.
Sielken, RL, Jr:  Valdez-Flores, C. (1999). Probabilistic risk assessment's use of trees and
       distributions to reflect uncertainty and variability and to overcome the limitations of
       default assumptions. Environ Int 25: 755-772. http://dx.doi.org/10.1016/S0160-
       4120(99)00053-7.
Sielken, RL, Jr.  (1990). A weight-of-evidence approach to quantitative cancer risk assessment:
       Information analysis. In G Schettler; D Schmahl; T Klenner (Eds.), Risk assessment in
       chemical carcinogenesis. New York, NY: Springer-Verlag.
Sielken, RL, Jr.: Bretzlaff, RS: Stevenson, DE. (1995). Challenges to default assumptions
       stimulate comprehensive realism as a new tier in quantitative cancer risk assessment.
       Regul Toxicol Pharmacol 21: 270-280.
Sikov, M. (1970). Radiation biology of the fetal and juvenile mammal. Science 167:  1640-1641.
Silkworm, JB; Koganti, A; Illouz, K; Possolo, A; Zhao, M; Hamilton, SB. (2005).  Comparison
       of TCDD and PCB CYP1A induction sensitivities in fresh hepatocytes from human
       donors, sprague-dawley rats, and rhesus monkeys and HepG2 cells. Toxicol Sci 87: 508-
       519. http://dx.doi.org/10.1093/toxsci/kfi261.
Simanainen, U;  Tuomisto, JT: Tuomisto, J: Viluksela, M. (2002). Structure-Activity
       relationships and  dose responses of Polychlorinated Dibenzo-p-dioxins for short-term
       effects in 2,3,7,8- Tetrachlorodibenzo-p-dioxin-Resistant and  sensitive rat strains.
       Toxicol  Appl Pharmacol 181: 38-47. http://dx.doi.org/10.1006/taap.2002.9386.
Simanainen, U:  Tuomisto, JT: Tuomisto, J: Viluksela, M. (2003). Dose-response analysis of
       short-term effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin  in three differentially
       susceptible rat lines. Toxicol Appl Pharmacol 187: 128-136.
       http://dx.doi.org/10.1016/S0041-008X(02)00068-6.
Simanainen, U:  Haavisto, T; Tuomisto, JT: Paranko, J: Toppari, J: Tuomisto, J: Peterson, RE;
       Viluksela, M. (2004). Pattern of male reproductive system effects after in utero and
       lactational 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) exposure in three differentially
       TCDD-sensitive rat lines Pattern of male reproductive system effects after in utero and
       lactational 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) exposure in three differentially
       TCDD-sensitive rat lines. Toxicol Sci 80: 101-108.
       http://dx.doi.org/10.1093/toxsci/kfhl42.
Skakkebaek, NE. (2010). reference ranges for semen quality and their relations to fecundity.
       Asian J Androl 12: 95-98.
Slama, R: Eustache, F; Ducot B: Jensen, TK: J0rgensen, N; Horte, A; Irvine, S; Suominen, J:
       Andersen, AG: Auger, J: Vierula, M: Toppari, J: Andersen, AN: Keiding, N;
       Skakkebaek, NE: Spira, A; Jouannet P. (2002). Time to pregnancy and semen
       parameters: A cross-sectional study among fertile couples from four European cities.
       Hum Reprod 17:  503-515. http://dx.doi.Org/10.1093/humrep/17.2.503.
Slezak, BP: Hatch. GE: DeVito, MJ: Diliberto, JJ: Slade, R: Crissman, K: Hassoun, E:
       Birnbaum, LS.  (2000). Oxidative stress in female B6C3F1 mice following acute and

                                          5-18

-------
       subchronic exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Toxicol Sci 54:
       390-398.
Small MJ. (2008). Methods for assessing uncertainty in fundamental assumptions and associated
       models for cancer risk assessment. Risk Anal 28: 1289 - 1308.
       http://dx.doi.org/10.ll 11/i. 1539-6924.2008.01134.x.
Smart. J: Daly. A. (2000). Variation in induced CYP1A1 levels: Relationship to CYP1A1, Ah
       receptor, and GSTM1 polymorphisms. Pharmacogenetics 10: 11-24.
Smialowicz. RJ: Burgin. DE: Williams. WC: Diliberto. JJ: Setzer. RW: Birnbaum. LS. (2004).
       CYP1A2 is not required for 2,3,7,8-tetrachlorodibenzo-p-dioxin-induced
       immunosuppression. Toxicology 197: 15-22. http://dx.doi.Org/10.1016/j.tox.2003.ll.016.
Smialowicz, RJ: DeVito, MJ: Williams, WC: Birnbaum, LS. (2008). Relative potency based on
       hepatic enzyme induction predicts immunosuppressive effects of a mixture of
       PCDDS/PCDFS and PCBS. Toxicol Appl Pharmacol 227: 477-484.
       http://dx.doi.0rg/10.1016/i.taap.2007.ll.018.
Smith, FA; Schwetz, BA; Nitschke, KD. (1976). Teratogenicity of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin in CF-1 mice. Toxicol Appl Pharmacol 38: 517-523.
       http://dx.doi.org/10.1016/0041-008X(76)90183-6.
Sparschu. G. . L.: Dunn. F. . L.: Rowe. V. . K. (1971). Study of the teratogenicity of 2,3,7,8-
       tetrachiorodibenzo-p-dioxin in the rat. Food Cosmet Toxicol 9: 405-412.
       http://dx.doi.org/10.1016/0015-6264(71)90045-9.
Staskal DF: Diliberto. JJ: DeVito. MJ: Birnbaum. LS. (2005). Inhibition of human and rat
       CYP1A2 by TCDD and dioxin-like chemicals. Toxicol  Sci 84: 225-231.
       http://dx.doi.org/10.1093/toxsci/kfi090.
Stayner, L; Bailer, AJ: Smith, R; Gilbert, S; Rice, F; Kuempel  E. (1999). Sources of uncertainty
       in dose-response modeling of epidemiological data for cancer risk assessment. Ann N Y
       Acad Sci 895: 212-222. http://dx.doi.Org/10.llll/i.1749-6632.1999.tb08087.x.
Steenland, K; Piacitelli, L: Deddens, J: Fingerhut, M; Chang, LI. (1999). Cancer, heart disease,
       and diabetes in workers exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin. J Natl Cancer
       Inst 91: 779-786.
Steenland, K: Deddens, J: Piacitelli, L. (2001). Risk assessment for 2,3,7,8-tetrachlorodibenzo-p-
       dioxin (TCDD) based on an epidemiologic study. Am J Epidemiol 154: 451-458.
Steinmaus, C: Miller, MD; Smith, AH. (2010). Perchlorate in drinking water during pregnancy
       and neonatal thyroid hormone levels in California. J Occup Environ Med 52: 1217-1524.
       http://dx.doi.org/10.1097/JOM.Ob013e3181fd6fa7.
Sugita-Konishi, Y: Kobavashi, K: Naito. H: Miura, K: Suzuki. Y. (2003). Effect of lactational
       exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin on the susceptibility to Listeria infection.
       Biosci Biotechnol Biochem 67: 89-93.
Swan. SH: Brazil C: Drobnis,  EZ: Liu. F: Kruse, RL: Hatch. M: Redmon, JB: Wang. C:
       Over street, JW. (2003). Geographic differences in semen quality of fertile U.S. males.
       Environ Health Perspect 111: 414-420. http://dx.doi.org/10.1289/ehp.5927.
Toide, K: Yamazaki, JH; Nagashima, R: Itoh, K: Iwano, S: Takahashi, Y: Watanabe, S:
       Kamataki, T. (2003). Aryl hydrocarbon hydroxylase represents CYP1B1 and not
       CYP1 Al, in human freshly isolated white cells: Trimodal distribution of Japanese
       population according to induction of CYP1B1 mRNA by environmental dioxins. Cancer
       Epidemiol Biomarkers  Prev 12: 219-222.
                                         5-19

-------
Toth, K; Somfai-Relle, S: Sugar, J: Bence, J. (1979). Carcinogenicity testing of herbicide 2,4,5-
       trichlorophenoxyethanol containing dioxin and of pure dioxin in Swiss mice. Nature 278:
       548-549.
Tritscher. AM: Goldstein. JA: Portier. CJ: McCoy. Z: Clark. GC: Lucier. GW. (1992). Dose-
       response relationships for chronic exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin in a rat
       tumor promotion model: Quantification and immunolocalization of CYP1A1 and
       CYP1A2 in the liver. Cancer Res 52: 3436-3442.
U.S. EPA (U.S. Environmental Protection Agency). (1986a). Guidelines for mutagenicity risk
       assessment [EPA Report]. (EPA/630/R-98/003). Washington, DC.
       http://www.epa.gov/iris/backgrd.html.
U.S. EPA (U.S. Environmental Protection Agency). (1986b). Guidelines for the health risk
       assessment of chemical mixtures (pp. 38). (EPA/630/R-98/002). Washington, DC: U.S.
       Environmental Protection Agency, Risk Assessment Forum.
       http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=22567.
U.S. EPA (U.S. Environmental Protection Agency). (1991). Guidelines for developmental
       toxicity risk assessment [EPA Report]. (EPA/600/FR-91/001). Washington, DC: U.S.
       Environmental Protection Agency, Risk Assessment Forum.
       http://www.epa.gov/iris/backgrd.html.
U.S. EPA (U.S. Environmental Protection Agency). (1994). Methods for derivation of inhalation
       reference concentrations and application of inhalation dosimetry. (EPA/600/8-90/066F).
       Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research
       and Development, Office of Health and Environmental Assessment, Environmental
       Criteria and Assessment Office.
       http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=71993.
U.S. EPA (U.S. Environmental Protection Agency). (1996). Guidelines for reproductive toxicity
       risk assessment [EPA Report]. (EPA/630/R-96/009). Washington, DC: U.S.
       Environmental Protection Agency, Risk Assessment Forum.
       http://www.epa.gov/raf/publications/pdfs/REPRO51 .PDF.
U.S. EPA (U.S. Environmental Protection Agency). (1998). Guidelines for neurotoxicity risk
       assessment. (EPA/630/R-95/001F). Washington, DC: U.S. Environmental Protection
       Agency, Risk Assessment Forum.
       http ://www. epa. gov/raf/publications/pdfs/NEUROTOX.PDF.
U.S. EPA (U.S. Environmental Protection Agency). (2000). Benchmark dose technical guidance
       document [external review draft]. (EPA/630/R-00/001). Washington, DC: U.S.
       Environmental Protection Agency, Risk Assessment Forum.
       http://www.epa.gov/raf/publications/benchmark-dose-doc-draft.htm.
U.S. EPA (U.S. Environmental Protection Agency). (2002). Guidelines for ensuring and
       maximizing the quality, objectivity, utility, and integrity, of information disseminated by
       the Environmental Protection Agency. (EPA/260/R-02/008). Washington, DC.
       http://www.epa.gov/quality/informationguidelines/documents/EP A_InfoQualityGuideline
       s.pdf.
U.S. EPA (U.S. Environmental Protection Agency). (2003). Exposure and human health
       reassessment of 2,3,7,8 tetrachlorodibenzo-p dioxin (TCDD) and related compounds
       [NAS review draft]. (EPA/600/P-00/001). Washington, DC: U.S. Environmental
       Protection Agency, National Center for Environmental Assessment.
       http://www.epa.gov/nceawwwl/pdfs/dioxin/nas-review/.

                                         5-20

-------
U.S. EPA (U.S. Environmental Protection Agency). (2005a). Guidelines for carcinogen risk
       assessment. (EPA/630/P-03/001F). Washington, DC.
       http://www.epa.gov/cancerguidelines/.
U.S. EPA (U.S. Environmental Protection Agency). (2005b). Supplemental guidance for
       assessing susceptibility from early-life exposure to carcinogens. (EPA/630/R-03/003F).
       Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
       http://www.epa.gov/cancerguidelines/guidelines-carcinogen-supplement.htm.
U.S. EPA (U.S. Environmental Protection Agency). (2006a). Approaches for the application of
       physiologically based pharmacokinetic (PBPK) models and supporting data in risk
       assessment (final report). (EPA/600/R-05/043F). Washington, DC: U.S. Environmental
       Protection  Agency, Office of Research and Development.
U.S. EPA (U.S. Environmental Protection Agency). (2006b). An inventory of sources and
       environmental releases of dioxin-like compounds in the United States for the years 1987,
       1995, and 2000 [EPA Report]. (EPA/600/P-03/002F). Washington, DC.
       http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=l 59286.
U.S. EPA (U.S. Environmental Protection Agency). (2008a). 2,3,7,8-Tetrachlorodibenzo-p-
       dioxin (TCDD) dose-response studies: Preliminary literature search results and request
       for additional studies. (EPA/600/R-08/119). Washington, DC: U.S. Environmental
       Protection  Agency, National Center for Environmental Assessment.
       http://cfpub.epa. gov/ncea/cfm/recordisplay. cfm?deid= 199923.
U.S. EPA (U.S. Environmental Protection Agency). (2008b). Framework for application of the
       toxicity equivalence methodology for polychlorinated dioxins, furans, and biphenyls in
       ecological  risk assessment. (EPA/100/R 08/004). Washington, DC: U.S. Environmental
       Protection  Agency, Office of Science Advisor.
       http://www.epa.gov/raf/tefframework/index.htm.
U.S. EPA (U.S. Environmental Protection Agency). (2009a). Summary of U.S. EPA dioxin
       workshop:  February 18-20, 2009, Cincinnati, Ohio. (EPA/600/R-09/027). Cincinnati,
       OH: U.S. Environmental Protection Agency, National Center for Environmental
       Assessment. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=205603.
U.S. EPA (U.S. Environmental Protection Agency). (2009b). Using probabilistic methods to
       enhance the role of risk analysis in decision-making with case study examples.
       (EPA/100/R-09/001). Washington, DC.
U.S. EPA (U.S. Environmental Protection Agency). (2010a). EPA's reanalysis of key issues
       related to dioxin toxicity and response to NAS comments. (EPA/600/R-10/038A).
       Washington, DC.
U.S. EPA (U.S. Environmental Protection Agency). (2010b). Recommended Toxicity
       Equivalence Factors (TEFs) for Human Health Risk Assessments of 2,3,7,8-
       Tetrachlorodibenzo-p- Dioxin and Dioxin-Like Compounds. (EPA/100/R-10/005).
       Washington, DC. http://www.epa.gov/raf/files/tefs-for-dioxin-epa-00-r-10-005-final.pdf.
U.S. EPA (U.S. Environmental Protection Agency). (2012).  Integrated risk information system
       (IRIS). Available online at http://www.epa.gov/iris/index.html.
Van den Berg, M;  Birnbaum, L; Bosveld, AT; Brunstrom, B; Cook, P; Feeley, M; Giesy,  JP:
       Hanberg, A; Hasegawa, R: Kennedy, SW: Kubiak, T; Larsen, JC: van Leeuwen, FX;
       Liem. AK: Nolt C: Peterson. RE: Poellinger. L: Safe. S:  Schrenk. D: Tillitt D: Tysklind.
       M: Younes, M: Waern, F; Zacharewski, T. (1998). Toxic equivalency factors (TEFs) for
       PCBs, PCDDs, PCDFs for humans and wildlife. Environ Health Perspect  106: 775-792.

                                         5-21

-------
Van Birgelen, AP: Van der Kolk, J: Fase, KM; Bol, I; Poiger, H; Brouwer, A; Van den Berg, M.
       (1995a). Subchronic dose-response study of 2,3,7,8-tetrachlorodibenzo-p-dioxin in
       female Sprague-Dawley rats. Toxicol Appl Pharmacol 132: 1-13.
       http://dx.doi.org/10.1006/taap.1995.1080.
Van Birgelen, AP: Smit, EA; Kampen, EVI: Groeneveld, CN: Fase, KM; Van der Kolk, J: Poiger,
       H: Van den Berg. M: Koeman, JH: Brouwer. A. (1995b). Subchronic effects of 2,3,7,8-
       TCDD or PCBs on thyroid hormone metabolism:  Use in risk assessment. Eur J
       Pharmacol 293: 77-85. http://dx.doi.org/10.1016/0926-6917(95)90021-7.
van Birgelen. AP: van den Berg. M. (2000). Toxicokinetics. Food  Addit Contam 17: 267-273.
       http://dx.doi.org/10.1080/026520300283342.
Van Den Hove, MF; Beckers, C: Devlieger, H: De Zegher, F; De Nayer, P. (1999). Hormone
       synthesis and storage in the thyroid of human preterm and term newborns: effect of
       thyroxine treatment. Biochimie 81: 563-570.
Van den Berg. M: de Vroom, E: Olie, K: Hutzinger, O. (1986). Bioavailability of PCDDs and
       PCDFs of fly ash after semi-chronic oral ingestion by guinea pig and Syrian golden
       hamster. Chemosphere 15: 519-533.
Van den Berg, M: Birnbaum, LS: Denison,  M: De Vito, M: Farland, W; Feeley, M: Fiedler, H:
       Hakansson, H: Hanberg, A; Haws, L; Rose, M: Safe, S; Schrenk, D; Tohyama, C:
       Tritscher. A: Tuomisto, J: Tysklind, M: Walker. N: Peterson. RE. (2006). The 2005
       World Health Organization reevaluation of human and mammalian toxic equivalency
       factors for dioxins and dioxin-like compounds. Toxicol Sci 93: 223-241.
       http://dx.doi.org/10.1093/toxsci/kfl055.
Van derMolen, G: Kooijman, A; Slob, W. (1996). A generic toxicokinetic model for persistent
       lipophilic compounds in humans: An application to TCDD. Fundam Appl Toxicol 31:
       83-94. http://dx.doi.0rg/10.1093/toxsci/31.l.83.
Van derMolen. GW: Kooijman. SAL. M: Michalek, JE:  Slob. W.  (1998). The estimation of
       elimination rates of persistent compounds: A re-analysis of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin levels in Vietnam veterans. Chemosphere 37: 1833-1844.
       http://dx.doi.org/10.1016/S0045-6535(98)00249-5.
Van derMolen. GW: Kooijman. BA: Wittsiepe, J: Schrev, P: Flesch-Janys, D: Slob. W. (2000).
       Estimation of dioxin and furan elimination rates with a pharmacokinetic model. J Expo
       Anal Environ Epidemiol 10: 579-585.
Vanden Heuvel JP: Clark. GC: Tritscher. A: Lucier, GW. (1994a). Accumulation of
       polychlorinated dibenzo-p-dioxins and dibenzofurans in liver of control laboratory rats.
       Fundam Appl Toxicol 23: 465-469.  http://dx.doi.Org/10.1093/toxsci/23.3.465.
Vanden Heuvel JP: Clark. GC: Kohn, MC: Tritscher. AM: Greenlee, WF: Lucier, GW: Bell
       DA. (1994b). Dioxin-responsive genes: examination of dose-response relationships using
       quantitative reverse transcriptase-polymerase chain reaction. Cancer Res 54: 62-68.
Vos, JG:  Moore, JA; Zinkl, JG. (1973). Effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin on the
       immune system of laboratory animals. Environ Health Perspect 5:  149-162.
Walker. NJ: Portier. CJ: Lax. SF: Crofts. FG: Li. Y: Lucier. GW: Sutter. TR. (1999).
       Characterization of the dose-response of CYP1B1, CYP1A1, and CYP1A2 in the liver of
       female Sprague-Dawley rats following chronic exposure to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin. Toxicol Appl Pharmacol 154: 279-286. http://dx.doi.org/10.1006/taap.1998.8595.
Wang. SL: Su. PH: Jong. SB:  Guo. YL: Chou. WL: Papke. O. (2005). In utero exposure to
       dioxins and polychlorinated biphenyls and its relations to thyroid function and growth

                                         5-22

-------
       hormone in newborns. Environ Health Perspect 113: 1645-1650.
       http://dx.doi.org/10.1289/ehp.7994.
Wang. X: Santostefano. MJ: Evans. MV: Richardson. VM: Diliberto. JJ: Birnbaum. LS. (1997).
       Determination of parameters responsible for pharmacokinetic behavior of TCDD in
       female Sprague-Dawley rats. Toxicol Appl Pharmacol 147: 151-168.
       http://dx.doi.org/10.1006/taap.1997.8242.
Wang. X: Santostefano. MJ: DeVito. MJ: Birnbaum. LS. (2000). Extrapolation of a PBPK model
       for dioxins across dosage regimen, gender, strain, and species. Toxicol Sci 56: 49-60.
Warner, M; Eskenazi, B; Mocarelli, P; Gerthoux, PM: Samuels, S: Needham, L: Patterson, D:
       Brambilla, P. (2002). Serum dioxin concentrations  and breast cancer risk in the seveso
       women's health study. Environ Health Perspect 110: 625-628.
Warner, M: Samuels, S: Mocarelli, P: Gerthoux, PM: Needham, L: Patterson, DG, Jr: Eskenazi,
       R (2004). Serum dioxin concentrations and age at menarche. Environ Health Perspect
       112: 1289-1292. http://dx.doi.org/10.1289/ehp.7004.
Warner, M: Eskenazi, B: Olive, PL; Samuels, S: Quick-Miles, S: Vercellini, P: Gerthoux, PM:
       Needham, L: Patterson, DG, Jr: Mocarelli, P. (2007). Serum dioxin concentrations and
       quality of ovarian function in women of seveso. Environ Health Perspect  115: 336-340.
       http://dx.doi.org/10.1289/ehp.9667.
Weber. LW: Lebofsky, M: Stahl. BU: Smith. S: Rozman, KK. (1995). Correlation between
       toxicity and effects on intermediary metabolism in  2,3,7,8-tetrachlorodibenzo-p-dioxin-
       treated male C57BL/6J and DBA/2J mice. Toxicol  Appl Pharmacol 131: 155-162.
       http://dx.doi.org/10.1006/taap.1995.1057.
Weber, R: Schmitz, HJ: Schrenk, D: Hagenmaier, H. (1997). Metabolic degradation, inducing
       potency, and metabolites of fluorinated  and chlorinated-fluorinated dibenzodioxins and
       dibenzofurans. Chemosphere 34: 29-40. http://dx.doi.org/10.1016/S0045-6535(96)00365-
       7.
Wendling, JM: Orth, RG: Poiger, H. (1990). Determination of [3H]-2,3,7,8-tetrachlorodibenzo-
       p-dioxin in human feces to ascertain its  relative metabolism in man. Anal Chem 62: 796-
       800. http://dx.doi.org/10.1021/ac00207a005.
White, KL, Jr: Lysy, HH: McCay, JA: Anderson, AC. (1986). Modulation of serum complement
       levels  following exposure to polychlorinated dibenzo-p-dioxins. Toxicol Appl Pharmacol
       84: 209-219. http://dx.doi.org/10.1016/0041-008X(86)90128-6.
WHO (World Health Organization). (1980). WHO laboratory manual for the examination of
       human semen and sperm-cervical mucus interaction. Singapore: Press Concern.
WHO (World Health Organization). (1998). Executive summary: Assessment of the health risk
       of dioxins: Re-evaluation of the tolerable daily intake (TDI). Geneva, Switzerland: WHO
       European Centre for Environmental Health and International Programme  on Chemical
       Safety.
WHO (World Health Organization). (2007). Assessment of iodine deficiency disorders and
       monitoring their elimination. Geneva, Switzerland: WHO Press.
       http://www.who.int/nutrition/publications/micronutrients/iodine deficiency/9789241595
       827/en/index.html.
WHO/UNICEF/ICCIDD (World Health Organization/ United Nations Children's Fund/
       International Council for the Control of Iodine Deficiency Disorders). (1994). Indicators
       for assessing iodine deficiency disorders and their control through salt iodization.
       (WHO/NUT/94.6).  Geneva: World Health Organization.

                                         5-23

-------
Wijchman, JG: de Wolf, BT; Graafe, R: Arts, EG. (2001). Variation in semen parameters derived
       from computer-aided semen analysis, within donors and between donors. J Androl 22:
       773-780.
Wittsiepe, J: Erlenkamper, B; Welge, P; Hack, A; Wilhelm, M. (2007). Bioavailability of
       PCDD/F from contaminated soil in young Goettingen minipigs. Chemosphere 67: S355-
       S364. http://dx.doi.0rg/10.1016/j.chemosphere.2006.05.129.
Wong. TK: Domin. BA: Bent PE: Blanton. TE: Anderson. MW: Philpot. RM. (1986).
       Correlation of placental microsomal activities with protein detected by antibodies to
       rabbit cytochrome P-450 isozyme 6 in preparations from humans exposed to
       polychlorinated biphenyls, quaterphenyls, and dibenzofurans. Cancer Res 46: 999-1004.
Yang, JZ; Agarwal SK; Foster, WG. (2000). Subchronic exposure to 2,3,7,8-tetrachlorodibenzo-
       p-dioxin modulates the pathophysiology of endometriosis in the cynomolgus monkey.
       Toxicol Sci 56: 374-381.
Zareba. G: Hoio. R: Zareba. KM: Watanabe. C: Markowskl VP: Baggs. RB: Weiss. B. (2002).
       Sexually dimorphic alterations of brain cortical dominance in rats prenatally exposed to
       TCDD. J Appl Toxicol 22: 129-137. http://dx.doi.org/10.1002/jat.839.
Zoeller, RT; Rovet J. (2004). Timing of thyroid hormone action in the developing brain: clinical
       observations and experimental findings. J Neuroendocrinol 16: 809-818.
       http://dx.doi.0rg/10.llll/i.1365-2826.2004.01243.x.
                                         5-24

-------
                                      EPA/600/R-10/038F
                                        www.epa.gov/iris
              APPENDIX A

Summary of External Peer Review and
   Public Comments and Disposition
                  January 2012
          National Center for Environmental Assessment
             Office of Research and Development
            U.S. Environmental Protection Agency
                  Cincinnati, OH

-------
         CONTENTS—APPENDIX A: Summary of External Peer Review and
                        Public Comments and Disposition
APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
     COMMENTS AND DISPOSITION	A-l
     A.I. GENERAL CHARGE QUESTIONS	A-2
         A. 1.1. SAB Comments and Recommendations and EPA Responses	A-2
               SAB Charge Question 1.1	A-2
               SAB Charge Question 1.2	A-3
     A.2. TRANSPARENCY AND CLARITY IN THE SELECTION OF KEY DATA
         SETS FOR DOSE-RESPONSE ANALYSIS	A-4
         A.2.1. SAB Comments and Recommendations and EPA Responses	A-4
               SAB Charge Question 2.1	A-4
               SAB Charge Question 2.2	A-5
               SAB Charge Question 2.3	A-5
         A.2.2. Summary of Public Comments and EPA Responses	A-9
     A.3. THE USE OF TOXICOKINETICS IN DOSE-RESPONSE MODELING FOR
         CANCER AND NONCANCER ENDPOINTS	A-ll
         A.3.1. SAB Comments and EPA Responses	A-ll
               SAB Charge Question 3.1	A-ll
                   SAB Charge Question 3.1.a	A-ll
                   SAB Charge Question 3.l.b	A-12
                   SAB Charge Question 3.l.c	A-13
                   SAB Charge Question 3.l.d	A-13
               SAB Charge Question 3.2	A-15
                   SAB Charge Question 3.2.a	A-15
                   SAB Charge Question 3.2.b	A-15
                   SAB Charge Question 3.2.c	A-16
               SAB Charge Question 3.3	A-17
               SAB Charge Question 3.4	A-17
               SAB Charge Question 3.5	A-17
         A.3.2. Summary of Public Comments and EPA Responses	A-18
     A.4. REFERENCE DOSE	A-21
         A.4.1. SAB Comments and EPA Responses	A-21
               SAB Charge Question 4.1	A-21
               SAB Charge Question 4.2	A-23
                   SAB Charge Question 4.2.a	A-23
                      SAB Charge Question 4.2.a.i	A-23
                      SAB Charge Question 4.2.a.ii	A-24
                   SAB Charge Question 4.2.b	A-25
                      SAB Charge Question 4.2.b.i	A-25
                      SAB Charge Question 4.2.b.ii	A-26
               SAB Charge Question 4.3	A-26
               SAB Charge Question 4.4	A-26
               SAB Charge Question 4.5	A-27

                                     A-i

-------
                          CONTENTS—(continued)


           SAB Charge Question 4.6	A-28
           SAB Charge Question 4.7	A-28
           SAB Charge Question 4.8	A-28
    A.4.2. Summary of Public Comments and EPA Responses	A-29
A.5. REFERENCES	A-38
                                 A-ii

-------
      APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
                            COMMENTS AND DISPOSITION
       EPA 's Reanalysis of Key Issues Related to Dioxin Toxicity and Response to NAS
Comments (Reanalysis) has undergone a formal, independent, expert panel review performed by
U.S. Environmental Protection Agency's (EPA's) Science Advisory Board (SAB) in accordance
with EPA guidance on peer review (2006c, 2000). The SAB Dioxin Review Panel held
two public face-to-face meetings to deliberate on the charge questions on July 13-15, 2010 and
October 27-29, 2010, as well as two public teleconferences on March 1 and 2, 2011. The SAB
Dioxin Review Panel was asked to consider the accuracy, objectivity, and transparency of EPA's
Reanalysis. Initially, the charge questions presented to the SAB Dioxin Review Panel were
divided into six sections:  General Charge Questions, Transparency and Clarity in the Selection
of Key Data Sets for Dose-Response Analysis, The Use of Toxicokinetics in the Dose-Response
Modeling for Cancer and Noncancer Endpoints, Chronic Oral Reference Dose, Cancer
Assessment, and Feasibility of'Quantitative Uncertainty Analysis From NAS Evaluation of the
2003 Reassessment.  Because of EPA's decision to release the cancer assessment and
quantitative uncertainty sections in a separate document, SAB and public comments related to
those topics are not addressed in this appendix but will be addressed in the Reanalysis Volume 2.
A summary of comments made by the SAB Dioxin Review Panel and EPA's responses to these
comments,  arranged by charge question, follow.  In many cases, the comments have been
synthesized and paraphrased in development of this appendix.  In response to a Federal Register
notice (75 FR 28610 [May 21, 2010]), EPA also received, comments from the public on the draft
document.  Each section provides EPA's charge question, followed by SAB comments and
specific recommendations related to the charge question, and then EPA's responses to the
recommendations. Major public comments that are relevant to specific sections, along with EPA
responses to the comment, are provided at the end of each respective section.  Section A.5 lists
the references cited in this Appendix.
                                         A-l

-------
A.l. GENERAL CHARGE QUESTIONS
A.1.1.  SAB Comments and Recommendations and EPA Responses
SAB Charge Question 1.1
Is the draft Response to Comments clear and logical? Has EPA objectively and clearly
presented the three key NRC recommendations?

Comment: In general, the Report was clear, logical, and responsive to many but not all of
National Academy of Sciences (NAS) recommendations; although there are opportunities for
improvement.  The Panel found that EPA was effective in developing a clear, transparent, and
logical response to NAS recommendations, and that EPA has objectively and clearly presented
the three key NAS recommendations. The Executive Summary was valuable in providing a
concise and accurate summary.  The Report was dense and repetitive in some places, and could
benefit from greater clarity in writing.  Although the Panel found that the Report was clear in its
presentation of the key NAS recommendations, it was not complete in consideration of
two critical elements: (1) nonlinear dose response for 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD) carcinogenicity and (2) uncertainty analysis.

       Response: EPA is moving forward to complete the draft Reanalysis and is planning to
       publish two reports (U. S. EPA's Reanalysis of Key Issues Related to Dioxin Toxicity and
       Response to NAS Comments Volumes 1  and 2 [Reanalysis Volumes 1 and 2]) that
       together will respond to the recommendations and comments  on TCDD dose-response
       assessment included in the NAS 2006 review.  The current report, Reanalysis Volume  1,
       includes the following information and corresponds to Sections 2 through 4 of the
       external review draft Reanalysis:
          1.   The study selection criteria used for the selection of studies for both noncancer
              and cancer TCDD dose-response analysis

          2.   The results of EPA's study selection process for both cancer and noncancer
              TCDD dose-response information

          3.   EPA's choice and use of a kinetic model to quantify appropriate dose metrics for
              both cancer and noncancer data sets

          4.   A noncancer  oral RfD for TCDD, including justification of approaches used for
              dose-response modeling of noncancer endpoints

          5.   A qualitative discussion of uncertainties in the RfD and a quantitative sensitivity
              analysis of the choices made in the development of points of departure (PODs) for
              RfD derivation
             Reanalysis Volume 2 will address the SAB comments related to the nonlinear
       dose response for TCDD carcinogenicity and quantitative uncertainty analysis. In
       Volume 2, EPA will complete the evaluation of cancer mode of action, cancer

                                          A-2

-------
       dose-response modeling, including justification of the approaches used for dose-response
       modeling of the cancer endpoints, and an associated quantitative uncertainty analysis.
       These issues correspond to Sections 5 and 6 of the external review draft Reanalysis.
              In addition to editing the document for greater clarity in writing, EPA has
       restructured Section 2 of the Reanalysis, moving large portions of summary text to
       appendices to reduce density and enhance readability of the document.

   Recommendation No. 1: Provide greater clarity and transparency in the discussion of
   studies that did not satisfy inclusion criteria. Given the enormity of this task, it can be done
   generally to indicate how the issue was considered.

       Response: In Sections 2.3.1 and 2.3.2, EPA has clarified further the study considerations
       and inclusion criteria for both the human and animal studies, respectively.  These
       clarifications included a statement that positive studies (i.e., studies reporting health
       outcomes) take precedence over null studies (i.e., studies not reporting health outcomes)
       for quantitative assessment. However, null studies are used by EPA when considering
       the biological significance of the critical endpoint(s) used as the basis for deriving an RfD
       and in qualitatively considering the overall database for hazard identification.
              EPA also has added a new Figure 4-2 that provides an overview of the
       disposition of all noncancer animal studies. For the noncancer animal studies,
       additional details are provided in Section 2 and Appendix D; a new Table D-2 shows
       the excluded animal studies and identifies the study inclusion criteria that were not met.
       For the epidemiologic studies that were evaluated, EPA reviewed and clarified the
       reasons for study exclusion; details are provided in Section 2 and Appendix C (see
       Tables C-2 through C-57).

   Recommendation No. 2: Carefully review the document using a qualified technical editor.

       Response: EPA has had the document reviewed by a qualified technical  editor.

   Recommendation No. 3. Include a glossary.

       Response: Section 1.5 now refers to the IRIS  online glossary available at
       http://epa.gov/iris/help_gloss.htm noting that this glossary provides definitions of terms
       typically used in IRIS documents, such as the Reanalysis.

   Recommendation No. 4: Find additional efficiencies (e.g., greater use of appendices and
   elimination of redundancies) to yield a more succinct and approachable document.

       Response: To improve readability, EPA has eliminated redundancies among sections of
       the document and moved the detailed epidemiologic and animal study summaries from
       the main text in Section 2 to Appendices C and D, respectively.

SAB Charge Question 1.2
Are there other critical studies that would make a significant impact on the conclusions of the
hazard characterization or dose-response assessment of the chronic noncancer and cancer
health effects ofTCDD?

                                          A-3

-------
Comment: The Panel did not identify any other critical studies that would impact the hazard
characterization or the dose-response assessment but feels that the Report should provide more
clarity on the exclusion of null epidemiologic studies.

   Recommendation No. 5: Provide more discussion and clarity on exclusion of null
   epidemiologic studies.

      Response: EPA has added as discussion of this issue in Section 2.3.1 with respect to
      epidemiologic study selection criteria.

A.2. TRANSPARENCY AND CLARITY IN THE SELECTION OF KEY DATA SETS
      FOR DOSE-RESPONSE ANALYSIS
      In general, the Panel favorably viewed EPA's efforts in developing the section of the

Report that presents how transparency and clarity was ensured (see Section 2) when selecting

key data sets. The comments and recommendations provided below will help EPA further
improve Section 2.


A.2.1. SAB Comments and Recommendations and EPA Responses
SAB Charge Question 2.1
Is this section responsive to the NAS concerns about transparency and clarity in data set
selection for dose-response analysis?

Comment: The Panel found that Section 2 was responsive to NAS concerns about transparency
and clarity. The Panel commended EPA's use of flow diagrams and Appendix B to increase
transparency and clarity.  The Panel noted, however, that clarity could be improved by providing
search words used for the MedLine searches.  The Panel also noted that the Report was overly
verbose, which was detrimental to its overall clarity.

      Response: EPA has further employed the use of flow diagrams and tables to show the
      disposition of studies and study/endpoint combinations in the process used to derive the
      TCDD RfD (e.g., see Figures 2-4, 4-2, and Tables D-l and D-2). EPA has added a new
      Appendix to the Reanalysis  (see Appendix I) that lists the search terms used to conduct
      the literature search. EPA has improved the readability of the document by moving
      summary text to appendices and eliminating redundancies in the text where feasible.

   Recommendation No. 6: Carefully and extensively edit to revise and consolidate Section 2
   and the Report as a whole. Restructure Section 2 to make it easier to follow a study from
   one section of the Report to another.  Then, use Section 2 as the foundation to improve
   overall document integration.

      Response: In response to these recommendations, EPA has conducted extensive editing
      and revisions to provide a clear, cohesive document.  To improve readability, the detailed
      epidemiologic and animal study summaries have been moved from the main text in
      Section 2 to Appendices C and D, respectively).  The rationale for study selection and
                                         A-4

-------
       tabular presentation of results remain the main focus of Section 2. Further, EPA has
       edited or added figures and tables to document the disposition of studies throughout the
       study selection process (see Figure 2-4 and Tables D-l and D-2) and for the development
       of candidate RfDs (see Figures 4-1, 4-2, and 4-3).

SAB Charge Question 2.2
Are the epidemiology and animal bioassay study criteria/considerations scientifically justified
and clearly described?

Comment: The Panel's discussion of Charge Question 2.2 is highly integrated with Charge
Question 2.3. Therefore, comments and specific recommendations that stem from these
two questions are presented together under Charge Question 2.3.

       Response: See recommendations and responses under Question 2.3 below.

SAB Charge Question 2.3
Has EPA applied the epidemiology and animal bioassay study criteria/considerations in a
scientifically sound manner? If not, please identify and provide a rationale for alternative
approaches.

Comment: The Panel found that study criteria and considerations were scientifically justified and
clearly described, and that they were presented in a scientifically sound manner, but
improvements could be made for clarity and on the rationale for decisions to include or exclude
particular studies or groups of studies  from the data sets. The panel also noted that the rationale
for distinct criteria for epidemiological and animal studies should be made stronger, and data set
selection for noncancer and cancer endpoints had room for further  clarification and justification.

   Recommendation No. 7: Better justify the rationale (including both scientific and practical
   reasons) for using studies where exposure is primarily to TCDD (or for animal studies only
   to TCDD) to calculate the reference dose.

       Response: EPA has added extensive text to Section 2.3 that discusses the rationale for
       focusing on TCDD studies, rather than studies on dioxin-like compounds (DLCs) or DLC
       mixtures. In identifying studies for quantitative TCDD dose-response analysis, EPA has
       focused on TCDD studies and  has not included studies on DLCs or DLC mixtures.
       Because the TCDD database is quite robust, inclusion of the DLC literature would likely
       increase the uncertainty in TCDD dose response unnecessarily. In addition, using studies
       evaluating information primarily or exclusively on TCDD,  as the index chemical,
       provides the most appropriate data for the risk assessment of dioxins and DLCs using the
       TEF  approach. EPA has included additional information to clarify that background DLC
       exposures are evaluated in the  context of the potential impact on TCDD-only
       quantification in certain cases as an uncertainty analysis (see new Section 4.5),
       particularly when TCDD exposures are relatively low.

   Recommendation No. 8: Incorporate studies with dioxin-like chemicals into a qualitative
   discussion of the weight-of-evidence for cancer and noncancer endpoints.

                                          A-5

-------
   Response: In the context of qualitative assessment of the critical effects, EPA has added
   a focused discussion of the Goodman et al. (2010) review of studies assessing DLC
   exposure and thyroid hormone levels in children (see response to Recommendation #34).
   The Goodman et al. (2010) review was evaluated with respect to elevated TSH levels in
   neonates, one of the co-critical endpoints forming the basis for the RfD.  EPA found no
   DLC exposure studies that evaluated the other co-critical endpoint, decreased sperm
   concentrations in men exposed to TCDD as boys.

Recommendation No. 9: Further clarify the justifications for study inclusion and exclusion
criteria/considerations more effectively and clearly. Specifically, remove criterion that
studies must explicitly state TCDD purity because it is highly unlikely that a study would
be conducted using impure TCDD.

   Response: EPA has removed the criterion for stating TCDD purity from the animal study
   selection criteria.

Recommendation No. 10: Revise the explanation of the in vivo mammalian bioassay
evaluation, indicating that the "study design is consistent with standard toxicological
practices" because it is too vague. If possible, provide a reference in which these practices
are described.

   Response: EPA has revised the explanation of this criterion to be clear that it excludes
   only those studies that use genetically-altered species.

Recommendation No. 11: Consider eliminating the use of the phrase "outside the range of
normal variability."

   Response: EPA has removed this phrase from the criteria.

Recommendation No. 12: Provide a definition when the term "common practice" is used,
and if possible, cite appropriate Agency documents.

   Response: EPA has removed the phrase "common practice" from the Reanalysis report
   and referenced the relevant Agency guidance documents where appropriate.  In addition,
   the Agency guidance used has been highlighted in a text box in Section 2.

Recommendation No. 13: Provide more discussion of data set limitations relevant to study
inclusion/exclusion criteria.

   Response: The epidemiology study summaries (Appendix C) have been edited with
   respect to study evaluation, meeting the study inclusion criteria and considerations, and
   suitability for dose-response modeling; Tables C-2 and C-3 summarize the cancer and
   noncancer studies, respectively,  identifying which criteria and considerations were met.

Recommendation No. 14: Better justify and explain considerations relating to selection of
epidemiology studies.
                                       A-6

-------
   Response: The descriptions for study quality considerations and study inclusion criteria
   have been edited for clarity.  Details of the implementation of these specific
   considerations and criteria in the study summaries and tables presented in Appendix C
   have also been edited.

Recommendation No. 15: Specifically, for Consideration #2 on Page 2-6 of the report, the
Panel recommends the following revisions: Define and clarify the term "susceptible to
important biases."  It is nonspecific, and the biases should be explained.

   Response: EPA has added clarifying language to Consideration #2 in Section 2 of the
   Reanalysis. The examination of biases included assessing the likelihood of selection
   bias, information bias, and confounding for the individual studies. EPA has also included
   text in the individual study summaries in Appendix C to specify possible sources of bias,
   and to determine the potential impact of these biases on individual study results.

Recommendation No. 16: Clarify what is meant by "control for potential confounding
exposures." Does this refer to only dioxin-like exposures?

   Response: EPA has added clarifying language to Consideration #2 to address this
   comment, which now reads "control for or account for confounding factors."  EPA has
   also provided explanations of specific confounding factors that were identified in the
   individual study summaries and tables in Appendix C. Assessment of the potential for
   confounding, therefore, was  not limited to dioxin-like chemicals and is specified for each
   study summary and summary tables as appropriate.

Recommendation No. 17: Clarify the phrase "bias arising from study design." Does  it
refer to selection bias, or is it used more broadly to describe how exposure and outcome are
measured and covariate data collected?

   Response: EPA has clarified Consideration #2 to address this comment; the current
   phrase "bias arising from limitations of study design" was referring to selection bias.
   EPA has also listed the main potential sources  of bias (e.g., selection bias, information
   bias, and confounding) earlier in Consideration #2 to help clarify this.

Recommendation No. 18: Define "bias arising from statistical analyses."  Might this  refer
to model misspecification?

   Response: EPA has added clarifying language to Consideration #2 to address this
   comment; the phrase "bias arising from statistical analyses" has been reworded to read
   "bias (e.g., selection or information bias) arising from limitations of the study design,
   data collection, or statistical  analysis." This would include model misspecification, such
   as adjustment for the incorrect functional form of certain confounders in multivariate
   regression modeling.

Recommendation No. 19: For Consideration #3 on Page 2-7 of the  report, the Panel
recommends the following revisions: Provide more discussion and clarity on the exclusion
of null epidemiologic studies.

                                       A-7

-------
   Response: EPA has added clarifying text under Consideration #3 to address this issue.
   This consideration addresses the use of null studies (i.e., studies reporting no association
   between TCDD and the health endpoint of interest) for the quantitative dose-response
   assessment used to derive an RfD; such studies are still used in qualitative assessments.
   Theoretically, a no-observed-adverse-effect level (NOAEL) can be identified from a null
   study and used to derive an RfD; that is, the highest available exposure dose from such a
   study could provide a NOAEL, which could serve as a basis for an RfD after appropriate
   uncertainty factors were applied.  However, a NOAEL from a study in which no adverse
   effects have been observed is not usually chosen for RfD derivation when other available
   studies demonstrate lowest-observed-adverse-effect levels (LOAELs). The large and
   comprehensive  database available to assess quantitative TCDD dose response provides
   many positive studies that  are considered stronger candidates for derivation of an RfD
   than the studies for which only a NOAEL can be identified. However, null studies are
   used by EPA to discuss the biological significance of the critical endpoint(s) used as the
   basis for deriving an RfD.

Recommendation No. 20: In Exclusion Criterion #3  on Page 2-7, define "reported dose."

   Response: EPA has deleted the sentence under Criterion #3 that contained this phrase as
   it did not enhance understanding of the criterion.

Recommendation No. 21: Clarify the discussion in Section 2 of the consideration of
confounding and other potential sources of bias.  Specifically, the Panel noted that the
differences between males and females with regard to TCDD half-life are discussed, but the
description of the number of males and females in each study population were often
missing or very difficult to determine.  Also, in the occupational cohort studies, the
possibility of men and women performing different job tasks also increased the possibility
that the men and women were exposed at different levels.  However, when the job
categories with assigned TCDD exposure levels were presented, there was  often no
discussion of the numbers by gender in the categories. For example, the Manz et al. study
(1991) of the Hamburg cohort (1,583 men and 399 women) does not describe the TCDD
categories by gender. In addition, the validity of the  TCDD exposure levels assigned to the
categories was examined "in a group of 48 workers who provided adipose tissue samples"
(page 2-41, lines 18-19). How were these workers selected?  How many were approached
but refused to provide a sample? Assessment of selection bias in this and other similar
circumstances was lacking in  some of the studies. This is particularly notable in the lack of
overall response rates reported for several of these studies. Inclusion of these factors in the
study review would be very helpful.

   Response: EPA has revised the summaries of the epidemiological studies in Appendix C
   to include clarifying text, response rates, and potential sources of bias where reported in
   the studies.

Recommendation No. 22: Clarify the discussion of the consideration that "statistical
precision, power, and study follow-up are sufficient." These metrics can be difficult to
determine with the smaller sample size populations, but there are studies that can be very
useful even given the small samples.

                                      A-8

-------
       Response: EPA has revised Consideration #5 and added clarifying text to address this
       issue. As stated in the consideration, EPA attempted to assess the possibility of not
       detecting an association that might be present due to limited statistical power of smaller
       studies.  In addition, EPA examined all reported effect estimates in each study
       irrespective of statistical significance.

A.2.2.  Summary of Public Comments and EPA Responses
Comment: Three commenters were concerned that the study inclusion criteria favored studies
showing positive associations between TCDD and health endpoints and that this would preclude
a weight-of-evidence analysis. The commenters were further concerned that the study inclusion
criteria in the draft Reanalysis were inconsistent with EPA's Information Quality Guidelines
(2002), Assessment Factors Handbook (2003), Risk Assessment Principles and Practices
documentation (2004), and the recommendations of the NAS committee that reviewed the 2003
Reassessment (NAS, 2006).

       Response: The study inclusion criteria apply only to the selection of data sets for dose-
       response modeling for the purpose of defining potential PODs and not to the elimination
       of studies from any further consideration. The focus of this process is on first identifying
       exposure levels associated with adverse effects, then determining an exposure level at
       which those effects do not occur.  The process does not eliminate "negative" studies for
       other purposes, such as supporting the cancer weight-of-evidence determination or
       assessing confidence in the endpoint(s) chosen for the POD for derivation of the RfD.
       EPA considered all studies, negative and positive, in the qualitative assessment of the
       RfD  in Section 4 of the Reanalysis. The study inclusion criteria are consistent with EPA
       RfD  and cancer assessment guidelines.  The  study selection process in this context is also
       consistent with the NAS committee recommendation that EPA justify the selection of
       studies for dose-response modeling.

Comment: One commenter asked EPA to consider recent publications addressing dioxin
toxicology in their selection of an overall data set. They provided the following list of
seven publications:

       Budinsky,  R.A., J.C. Rowlands, S. Casteel et al. (2008). A pilot study of oral
             bioavailability of dioxins and furans from contaminated  soils: Impact of
             differential hepatic enzyme activity and species differences. Chemosphere
             70:1774-86.
       Budinsky,  R.A., C.R. Kirman, LJ. Yost, B.F. Baker, L.L. Aylward, J.M. Zabik, J.C.
             Rowlands, T.F. Long, and T.  Simon.  (2009). Derivation of Soil Cleanup Levels
             for 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) Toxic Equivalence (TEQD/F) in
             Soil Through Deterministic and Probabilistic Risk Assessment of Exposure and
             Toxicity. Presentation at Society of Toxicology Annual Meeting. March.
       Charnley, G. and R.D. Kimbrough. (2006). Overview of exposure, toxicity and risks to
             children from current levels of 2,3,7,8-tetrachlorodibenzo-p-dioxin and related
             compounds in the USA. 2005. Food and  Chemical Toxicology 44:601-615.
                                          A-9

-------
       Garabrant D.H., A. Franzblau, J. Lepkowski, B.W. Gillespie, P. Adriaens, A. Demond, E.
             Hedgeman, K. Knutson, L. Zwica, K. Olson, T. Towey, Q. Chen, B. Hong, C-W.
             Chang, S-Y. Lee, B. Ward, K. LaDronka, W. Luksemburg, and M. Maier. (2009).
             The University of Michigan Dioxin Exposure Study: Predictors of human serum
             dioxin concentrations in Midland and Saginaw, Michigan.
       Hays, S.M. and L.L. Aylward. (2003). Dioxin risks in perspective: past, present, and
             future. Regulatory Toxicology and Pharmacology 37:202-217.
       Kimbrough R.D., C.A. Krouskas, M. Leigh Carson, T.F. Long, C. Bevan, and R.G.
             Tardiff.  (2009). Human uptake of persistent chemicals from contaminated soil:
             PCDD/Fs and PCBs. Regulatory Toxicology and Pharmacology 2009 Dec 24;
             [Epub ahead of print], Center for Health Risk Evaluation P.O. Box 15452
             Washington, DC 20003, United States.
       LaKind, J.S., S.M. Hays, L.L. Aylward, and D.Q. Naiman. (2009). Perspective on serum
             dioxin levels in the United States: an evaluation of the NHANES data. Journal of
             Exposure  Science and Environmental Epidemiology 19:435-441.
       Response: EPA has reviewed these studies and considered their applicability in
       informing the hazard identification dose response following TCDD exposure.  None of
       these studies provide in vivo mammalian dose-response study results that would be useful
       in quantitative dose-response analysis for derivation of an RfD or oral slope factor for
       TCDD, nor do they inform the hazard identification.  Therefore, none of these studies
       qualifies as an appropriate  study type in EPA's study selection process for quantitative
       TCDD dose-response assessment.

Comment: One commenter felt that the development of the proposed RfD was not transparent
because it did not rely on toxicological assessment work completed since the
2003 Reassessment.  Additionally, the commenter requested additional clarity and transparency
in the rationale for the Agency's selection of key data and more explanation of why EPA did not
pursue benchmark dose modeling for the two human data sets used to derive the RfD.

       Response: EPA collected and evaluated studies through October 2009, including studies
       from the 2003 Reassessment and newer studies found  via literature searches and through
       public submissions. EPA notes that the RfD is based on two studies published in 2008.
       In addition, EPA has included evaluations of several relevant studies published in 2010
       and 2011; EPA identified these studies as it continues  to monitor the dioxin health effects
       literature.
          Regarding the comment requesting additional transparency in the study selection
       process, EPA has provided additional clarity on the study inclusion criteria with
       revisions to the Reanalysis based on SAB and public  comments.
          EPA relied on the study authors' modeling of the epidemiologic study data, which
       included the important covariates affecting the relationship between health outcome and
       TCDD exposure. The current version of EPA's benchmark dose modeling software
       does not allow for modeling of covariates reported in epidemiologic studies.
                                         A-10

-------
A.3. THE USE OF TOXICOKINETICS IN DOSE-RESPONSE MODELING FOR
     CANCER AND NONCANCER ENDPOINTS
A.3.1.  SAB Comments and EPA Responses
SAB Charge Question 3.1
The 2003 Reassessment utilized first-order body burden as the dose metric. In the draft
Response to Comments document, EPA used a physiologically basedpharmacokinetic (PBPK)
model (Emondetal, 2006; 2005; 2004) with whole blood concentration as the dose metric
rather than first-order body burden. This PBPK model was chosen, in part, because it includes
a biological description of the dose-dependent elimination rate ofTCDD. EPA made specific
modifications to the published model based on more recent data. Although lipid-adjusted serum
concentrations (LASC) for TCDD are commonly used as a dose metric in the literature, EPA
chose whole blood TCDD concentrations as the relevant dose metric because serum and serum
lipidare not true compartments in the Emond PBPK models (LASC is a side calculation
proportional to blood concentration). Reviewers were asked to comment on Questions 3. l.a-d.

SAB Charge Question S.l.a
The justification of applying a PBPK model with whole blood TCDD concentration as a
surrogate for tissue TCDD exposure in lieu of using first-order body burden for the
dose-response assessment ofTCDD.

Comment: The use of whole blood concentration is a better choice than body burden, as was
used in the 2003 Reassessment, because it is more closely related to the biologically relevant
dose metric. However, the rationale for the use of blood concentration rather than lipid adjusted
serum concentration (LASC) should not be based on the Emond model structure. The question
that should be addressed is only whether blood concentrations or LASCs provide better
surrogates for cross-species and cross-study comparisons of free dioxin concentration in the
target tissues. LASC is the preferred measure for reporting dioxin biomonitoring data and is the
measurement reported in most of the human epidemiological  studies. A metric that considers
blood lipid content is also more likely to reflect free dioxin concentration in the plasma and,
hence, free concentration in the target tissue. The EPA pointed out that the LASC was related to
the blood concentration by a scalar; however, EPA incorrectly concluded that the metrics are
equivalent and later discussed the fact that the relationship between them was subject to
inter-individual and inter-species variation.  If the LASC were used to drive the distribution of
TCDD to tissues, the pharmacokinetic outcome would be different from using blood as the driver
because the tissue:blood ratio would differ.  If the blood fatblood and tissue:blood values were
accounted for in the model, the use of blood and LASC would be similar. It is not clear at this
point how this issue was addressed in the dose metric calculations.  Consideration of this issue is
unlikely to drastically affect the outcome of the risk calculations, but it would be important for a
quantitative uncertainty analysis.

   Recommendation No. 23: The use of the blood metric is acceptable for the PBPK model.
   Clarify how the model deals with studies that report the concentration of dioxin in plasma,
   serum, blood, or blood fatblood measurements.
                                         A-ll

-------
Response: The issue of whether LASC or whole-blood concentration is the more relevant metric
(for interspecies extrapolation) hinges on how the Emond rat PBPK model was calibrated.  The
rat model was calibrated to whole tissue concentrations (liver, fat, whole blood) and not LASC
or other tissue lipid concentrations. Relative whole-tissue concentrations reflect the relative
tissue fat content, so the difference in LASC:whole-blood ratios between rats and humans is
handled implicitly in the model.  The rat model intake predictions are a function of whole-blood
concentrations rather than LASC. The human model is structured the same way. Therefore,
human whole-blood concentrations should be equated with rat whole-blood concentrations for
obtaining the equivalent human intakes. EPA has clarified that the TCDD LASC values reported
in the epidemiology studies were used directly to estimate equivalent human intakes from the
Emond PBPK model.
       EPA also clarified that, for interspecies extrapolation, whole-blood concentrations were
used because distribution of TCDD to the liver and subsequent processing for dose-dependent
elimination in the liver in this model is dependent on whole-blood concentrations, not LASC.  In
both the Emond rodent and human models, LASC values are calculated post-processing by
application of scalars representing the proportion of plasma and fat in the whole-blood
compartment. That is, translating results from the rodent model to the human model requires an
estimate of the TCDD concentration in the whole-blood compartment whether starting from
whole-blood concentrations or LASC.  This approach assumes that differences in serum and
serum lipid fractions between rodents and humans do not result in large differences among the
species in the transfer of TCDD from blood to liver.

SAB Charge Question S.l.b
The scientific justification for using the Emond et al. model as opposed to other available TCDD
kinetic models.

Comment:  The Emond model provided the best available basis for the dose metric calculations
in the assessment;  however, additional discussion of other published models and quantitative
evaluation of the impact of model selection on dose metric predictions should also be provided.

   Recommendation No. 24: Discuss how the model was intended to be used in the
   assessment, which would then dictate why a particular model was selected.  That is, for the
   intended purposes, was the Emond model more robust and/or simpler than other models,
   and did it contain sufficient details for biological determinants deemed important by the
   Agency?

       Response:  EPA has clarified that the Emond PBPK model was used to (1) estimate oral
       intakes corresponding to measured LASC TCDD concentrations in human subjects and
       (2) estimate animal blood concentrations based on measured doses in bioassays as the
       appropriate dose metric for modeling equivalent human intakes. EPA has also clarified
       that the Emond model was selected  because of its technical sophistication for simulating
       physiological processes associated with TCDD and because the model covered all of the
       relevant life stages (particularly gestational and childhood exposures), which the
       alternative  model  (CADM) did not.  Other models were not presented because they did
       not account for dose-dependent elimination  processes, which EPA established as an a
       priori criterion for PBPK model selection, based on the current scientific understanding
       of TCDD kinetics.
                                         A-12

-------
SAB Charge Question S.l.c
The modifications implemented by EPA to the published Emond et al. model.

Comment: The model changes are minor, scientifically appropriate, and well supported.

       Response: No response necessary.

SAB Charge Question S.l.d
Whether EPA adequately characterized the uncertainty in the kinetic models.

Comment: The Report presents a reasonably thorough qualitative characterization of the
uncertainty in the kinetic models that is sufficient to support their use in the assessment;
however, a more quantitative uncertainty analysis is needed.  It is critical to demonstrate the
dependence of human equivalent dose (HED) and risk predictions on uncertainty and variability
in the model parameters. Dose metric uncertainty needs to be determined under the same
exposure conditions that dose metrics are calculated—both for the various studies that serve as
the basis for the dose-response assessments and for human exposures at the corresponding HEDs
and risk-specific doses.
       The Hill coefficients for CYPlal and CYPla2 induction used in the Emond model
were 1.0 and 0.6, respectively, based on fitting of kinetic data from single doses of dioxin
(Santostefano et al.. 1998: Wangetal.. 1997). However, Walker et al. (1999) subsequently
estimated a Hill coefficient of 0.94 for both CYPlal and CYPla2 induction using chronic
exposures, which were more relevant to the use of the Emond model in the dioxin risk
assessment.  The value of 0.6 used in the Emond model was well outside the confidence interval
of 0.78 to 1.14 reported by Walker et al. (1999).  The use of a Hill coefficient value well below
unity would lead to a nonlinear model behavior that is biologically implausible (hypersensitivity
to induction at doses near zero).  As a result, when the human model was used for extrapolation
to lower doses (as in the calculation of risk-specific doses), the  model would tend to estimate a
lower exposure level for a given blood concentration.  This effect could be seen in Table ES-1 of
the Report, where a 5 order-of-magnitude change in risk was associated with a
6 order-of-magnitude change in risk-specific dose. That is, the model-estimated risk-specific
doses in the vicinity of 10   risk were about a factor of 10 lower (more conservative) than linear
extrapolation.  The evidence for this parameter needs to be carefully reviewed and the reasonable
range of values determined.  At the least, the Emond human model calculations will need to be
repeated with multiple values to characterize the resulting uncertainty in the estimates.
       When this is done, the Agency should also consider increasing the fatblood partition in
the human model from 100 to 200 to be more consistent with the human data (Maruyama et al.,
2002: lidaetaL  1999: Patterson  et al.. 1989: Schecter and Ryan.  1989: Schecter et al.. 1989).
The Hill coefficient is not likely to have as significant an effect on calculations with the animal
models, because low-dose extrapolation was not performed in the animals, but this should also
be verified by sensitivity/uncertainty analysis of the animal models. Public comments were
submitted to the Panel, recommending consideration of a Hill coefficient value of 1.0 and
pointing out why lower values are inappropriate (comments from Drs. Thomas Starr, July 7,
2010 and October 26, 2010 and Melvin E. Andersen, November 4, 2010).
                                         A-13

-------
Recommendation No. 25: Undertake additional efforts to fully characterize the uncertainty
in the model, with special consideration of the Hill coefficient value.

   Response: In response to this comment, EPA has conducted a sensitivity analysis by
   varying each parameter in the PBPK models individually to determine the effect on the
   average whole-blood concentrations (as the dose metric used for species extrapolations
   and reference dose calculations).  In addition, the effect of varying the Hill parameter on
   the model fits to literature data was explored. In response to this comment, two sections
   were added to Section 3. Section 3.3.4.3.2.5 describes the results of the sensitivity
   analysis preformed on the PBPK models as suggested by the SAB reviewers, and
   Section 3.3.4.3.2.6 documents the impact of changing the Hill coefficient on PBPK
   model simulations of dioxin blood levels in humans. Included in this section is a
   sensitivity analysis using alternative CYP1A2 induction parameters determined from data
   presented in Budinsky et al. (2010). The Walker et al. (1999) CYP1 Al  and CYP1A2
   induction analysis, in which a value of 0.94 was found for the Hill coefficient, uses a
   different model structure formulation than the one in the Emond model, in which the
   parameters have different interpretations,  such that the Hill coefficient values represent
   different processes and are not strictly comparable.
       Further, in an additional sensitivity analysis reported in Section 4.5.1.1.1, EPA also
   evaluated the impact on the RfD of changing the Hill coefficient to a value of 1, noting
   that the Hill  coefficient was the most influential variable in the Emond PBPK model (see
   Section 3.3.4.3.2.5) and that the value of 0.6 results in a supralinear relationship between
   intake and blood concentrations at very low doses.  The value of 1 was chosen for the
   sensitivity analysis of the Hill coefficient because that is the lowest value where the
   model is no longer supralinear; otherwise the value of 1 has no biological or empirical
   basis.  When the Hill coefficient is set to a value of 1, and applying an uncertainty of 30
   (see Section 4.3.5), the resulting candidate RfD would be 2  x io~4 ng/kg-day
   (2 x 10'11 mg/kg-day).
       EPA's sensitivity analysis for the Emond PBPK model parameters also addresses
   the fatblood partition coefficient (PCps) issue (i.e., SAB's suggestion to increase the
   value to 200).  To clarify the nature of the parameter, the PCps  of 100 in the Emond
   model is a fitted value in the original rat model (WangetaL 1997), in which other
   parameters (including the value of 0.6 for the Hill coefficient, the most influential
   parameter in the model) were also fitted simultaneously  against animal and human data.
   EPA has evaluated the literature cited by the SAB and has concluded that a PCps of 160
   is more representative of the data presented in those papers. A value of 158 is estimated
   by Patterson et al. (1988) based on 50 individuals from Times Beach, MO.  lida et al.
   (1999) measured levels of 2,3,7,8-TCDD in blood and adipose  tissue from eight human
   subjects, who varied in age (19 to 82 years) and gender (four females and four males).
   Using the individual measurements presented in lida et al. (1999) and assuming relative
   lipid contents of 0.85 and 0.0057 in adipose tissue and blood, respectively, EPA
   estimated a mean and median PCpB of 166 and  161, respectively. A value of 247
   reported by Maruyama et al. (2002) was based on the data from lida et al. (1999),
   however, EPA was unable to reproduce the value of 247 reported by these authors.
   Schecter and Ryan (1989) present data on a single individual who was also exposed to
   high levels of DLCs and PCBs in an acute event (transformer explosion).  Several
   serum and fat measurements were taken over the next 5 years, during which time the
                                      A-14

-------
      patient lost 30 pounds and took medication to reduce serum lipids.  The combination of
      all of these factors suggest that the internal concentrations may not have equilibrated in
      this time frame and introduces too much uncertainty for use of these data in estimating a
      PCFB for TCDD.  Schecter et al. (1989) report fat TCDD concentrations but not blood
      or serum concentrations. In the sensitivity analysis that EPA conducted on the Emond
      PBPK model, the elasticity of a 50% increase in the fatblood partition coefficient at
      exposures equal to the RfD POD (0.02 ng/kg-day) was -0.064 (see Table 2-12), which
      means that increasing the parameter value from 100 to 150 would result in a 6.4%
      decrease in the TCDD blood concentration at this exposure level; a further increase to
      160 would result in about a 7% decrease. EPA estimates that, using the 160 value for
      the fatblood partition coefficient, the LOAEL corresponding to the Baccarelli et al.
      (2008) scenario would increase by 10% to 0.022 ng/kg-day, with no change in the RfD.
      The LOAEL corresponding to the Mocarelli et al. (2008) scenario would increase by
      40% to 0.028 ng/kg-day.

SAB Charge Question 3.2
Several of the critical studies for both noncancer and cancer dose-response assessment were
conducted in mice. A mouse PBPK model was developed from an existing rat model in order to
estimate TCDD concentrations in mouse tissues, including whole blood.  Reviewers were asked
to comment on Questions A.3.2.a-c.

SAB Charge Question 3.2.a
The scientific rationale for the development of EPA 's mouse model based on the published rat
model (Emondet al. 2006; 2005; 2004).

Comment: The Panel agrees that an appropriate approach was used to develop the mouse model
on the basis of the published rat model and the available mouse kinetic data.  It should be noted
that the NAS recommendation to use human  data for dose metric could be accomplished because
dose-dependent elimination of TCDD has been described in humans, albeit in just a few cases.
Dose-dependent elimination  has been reported repeatedly in animals, and the PBPK model
reflected this dose-dependence. Using CYP1A2 data from humans (caffeine metabolism) and
mice would offer an opportunity to validate and/or adjust the mouse model.

   Recommendation No. 26: Conduct an external peer review of the mouse model because it
   has not been published in the peer-reviewed literature.

      Response: EPA has recommended that the authors submit their work for publication in
      the peer-reviewed literature. Although EPA used revised estimates for some of the
      published parameters, no modifications were made to the structure of the Emond model.
      Using these revised parameters, EPA has described the evaluation of the PBPK model in
      Section 3. An important point is that the mouse data were not used directly in estimation
      of reference values.

SAB Charge Question 3.2.b
The performance of the mouse model in reference to the available data.
                                        A-15

-------
Comment: The Panel found that the mouse model performed reasonably well, apart from
under-prediction of urinary excretion data. The urinary excretion data can be improved by
taking into account the fact that urine contains metabolites only, which partition differently from
the parent compound. The model appeared to be adequate for use in estimating dose metrics for
the assessment, but with greater uncertainty than the rat and human models. This was considered
a reasonable approach to solve a deficiency in published PBPK models to meet the needs of this
assessment.
       The Panel noted, however, that the EPA's suggestion in the RfD chapter that the
clustering of mouse points of departure (PODs) at the lowest doses was due to mouse model
failure, was inappropriate, and should be rewritten.

   Recommendation No. 27: Use the mouse model and try to get the model published in the
   peer-reviewed literature to enhance scientific credibility.

       Response: EPA has revised the text describing the mouse PODs to eliminate the
       impression that the result was due to failure of the mouse PBPK model, which was not
       intended. See the  response above (Recommendation 26) regarding the comment on the
       publication of the mouse model.

SAB Charge Question 3.2.c
Whether EPA adequately  characterized the uncertainty in the mouse and rat kinetic models.
Please comment specifically on the scientific justification of the kinetic extrapolation factor from
rodents to humans.

Comment: EPA provided an adequate characterization of the qualitative uncertainty in the
mouse and rat kinetic models sufficient to justify their use, together with the human model, to
estimate rodent-to-human extrapolation factors. On the other hand, formal  recalibration of the
PBPK model parameters using a Hierarchical Bayesian approach such as Markov chain Monte
Carlo analysis was not considered necessary or particularly useful.  However, a more
quantitative uncertainty analysis is needed.

   Recommendation No. 28: Perform a more quantitative uncertainty analysis using methods
   suggested in response to Charge Question 6.2.l

       Response: In response to this recommendation and other comments, EPA has conducted
       a sensitivity analysis and added it to Section 3 (see Sections 3.3.4.3.2.5 and 3.3.4.3.2.6;
       also see response to Recommendation 25). EPA has undertaken additional quantitative
       sensitivity analyses for the kinetic modeling and some exposure assumptions relevant to
       the development of the RfD (see Section 4.5; see also responses to Recommendations 29
       and 32).
1 SAB comments on Sections 5 and 6 are not addressed in Volume 1 of the Reanalysis, but can be viewed at the
following URL: http://yosemite.epa.gov/sab/sabproduct.nsf/WebReportsLastMonthBOARD
/2A45B492EBAA8553852578F9003ECBC5/$File/EP A-SAB-ll-014-unsigned.pdf.
                                         A-16

-------
SAB Charge Question 3.3
Please comment on the use of the Emondet al. PBPK model to estimate human intakes based on
internal exposure measures.

Comment: The modified Emond model is the best available approach for estimating exposures
on the basis of internal exposure measurements.  Nevertheless, there is considerable uncertainty
associated with attempting to reconstruct prior exposures in a human population (e.g., Seveso).

   Recommendation No. 29: Describe the modeling of the Cheng et al. (2006), Mocarelli
   et al. (2008), and Baccarelli et al. (2008) studies in more detail, and quantitatively evaluate
   the impact of model parameter uncertainty and exposure uncertainty in these studies.

       Response: EPA has revised the document to describe the modeling of Mocarelli et al.
       (2008) and Baccarelli et al. (2008) in more detail. Sensitivity analyses pertaining to the
       choice of model inputs have been performed for Mocarelli et al. (2008) and  Baccarelli
       et al. (2008) and are described in Section 4.5 of the document.  Cheng et al.  (2006) is a
       cancer-modeling study and will be addressed in Volume 2 of this report.

SAB Charge Question 3.4
Please comment on the sensitivity analysis of the kinetic modeling (see Section 3.3.5).

Comment: The Report only presented the sensitivity analysis published by Emond  et al. (2006),
which was not entirely adequate for the purposes of this assessment. The analysis left out the
Hill coefficient, which was one of the most important parameters in the model for low-dose
extrapolation (Evans and Andersen, 2000). Moreover, model sensitivities were species, dose,
and dose-scenario dependent, so they need to be determined under the same exposure conditions
as those for which dose metrics were calculated:  both for the various studies that serve as the
basis for the dose-response assessments and for human exposures at the corresponding HEDs
and risk-specific doses. This represents the most pragmatic path forward for an evaluation of
model sensitivity as it relates to potential environmental regulation.

   Recommendation No. 30: Provide a sensitivity analysis of the model to authenticate the
   model for its intended purpose.

       Response: EPA has conducted a sensitivity analysis (see response to
       Recommendations 25 and 28).

SAB Charge Question 3.5
Both EPA 's noncancer and cancer dose-response assessments are based on a lifetime average
daily dose. Did EPA appropriately estimate lifetime average daily dose? If not, please suggest
alternative approaches that could be readily developed based on existing data.

Comment: The Panel agrees with the average daily dose calculation approaches,  but it was not
clear to some Panel members how the computational estimates of internal dose for newborns
were carried out because a lactation model was not used.  This is important because of the use of
TSH (thyroid stimulating hormone) in newborns as a critical effect.

                                         A-17

-------
   Recommendation No. 31: Explain how the early life-stage internal doses are calculated.

       Response: Internal TCDD doses for newborns were not estimated in the Reanalysis. The
       increased TSH levels at 72 hours after birth are modeled as a function of maternal
       exposure, with the assumption that the actual critical exposures occurred in utero and
       were not due to breast feeding. EPA has clarified that the Emond PBPK model accounts
       for physiological changes including body weight and tissue volumes over  different life
       stages, including during gestation.  The only life stage that is not accounted for in the
       Emond model is infants exposed to TCDD through breast milk.  The details of how the
       model estimates tissue and blood levels of TCDD during the other life stages following
       TCDD exposures are described in Section 3 and by Emond et al. (2006; 2005;  2004).

A.3.2.  Summary of Public Comments and EPA Responses
Comment: One commenter noted that CADM (i.e., Concentration- and Age-Dependent
Elimination Model) should be given more consideration as a credible alternative to the Emond
et al. model. When CADM and the Emond et al. model have been evaluated on the same human
data sets, CADM appears to provide substantially better results, and the Emond et al. model
appears to markedly overpredict the early serum concentration levels.  Another commenter noted
that CADM allows estimation of the relevant risk-specific doses using the PBPK  model but is
applied in the exposure range relevant to real-world exposures, reproduces the elimination
behavior of TCDD relevant to risk assessment and risk management, and takes into account
background body burdens of TCDD and non-TCDD  contributors  to TEQ and their impact on
TCDD elimination behavior.

       Response: EPA used the Emond model for human toxicokinetics because  the model
       covered all of the relevant life stages (particularly gestational and childhood exposures),
       which CADM does not, and also because of its technical sophistication for simulating
       physiological processes associated with TCDD toxicokinetics. The Emond model also is
       able to account for background TCDD and DLC body burdens and their impact on TCDD
       elimination behavior; pertinent simulations and discussions on these aspects have been
       added in the new Section 4.5.
           For animal bioassays, EPA undertook, and reported in the document,  modeling
       analyses that compared the predicted values from both the Emond PBPK model and
       CADM for all administered  doses. Throughout the document, separate simulations for
       both the PBPK model and CADM were conducted for comparison to experimental or
       literature data for animals. In Section 3, EPA presents extensive comparisons of the
       Emond model and CADM.  In Appendix E, EPA also presents whole blood, fat, and liver
       TCDD concentrations and body burdens that  were predicted by both the Emond model
       and CADM for each key animal bioassay.

Comment: One commenter noted that the Hill function dependence of CYP1A2 induction on
AhR-bound TCDD has a nonphysical, nonsensically infinite slope at zero dose, due to the fact
that its exponent parameter has a numerical value smaller than 1,  namely 0.6. This phenomenon
has no predictive value at low doses. According to the commenter, the values that are predicted
at low doses are simply artifactually constrained by the supralinear shape of the Hill function,
which is imposed by the data at far higher doses.  Because no data occur in the low-dose region
that is well below the EC50, no counterbalancing force exists that would keep the Hill exponent
                                        A-18

-------
value at or greater than 1.  This leads to artifactual and arbitrarily large increases in the oral slope
as the TCDD intake approaches zero.

       Response: EPA has conducted a sensitivity analysis for the Hill coefficient (see response
       to Recommendation 25) and has evaluated the impact of eliminating the supralinear
       behavior on relative human intakes.  Changing the Hill coefficient to 1, which results in
       linear low-dose behavior, and optimizing to a limited number of human data sets results
       in somewhat lower oral intake rate estimates associated with the TCDD serum
       concentrations in the range of interest (i.e., near the RfD and LOAEL POD). This result
       is well within  the range of other uncertainties evaluated by EPA (see Section 4.5).  EPA
       has concluded that, given the uncertainties in the value of this parameter and
       interdependent parameters in the model, and the lack of a  substantial impact on predicted
       intakes in the range of the POD for the RfD, there is no mechanistic or empirical basis on
       which to change the value of the Hill coefficient or related parameters.  In response to
       this comment, two sections were added to Section 3. Section 3.3.4.3.2.5 describes the
       results of the sensitivity analysis performed on the PBPK models as suggested by this
       reviewer and the SAB reviewers, and Section 3.3.4.3.2.6 illustrates the impact of
       changing the Hill coefficient on PBPK model simulations of dioxin blood levels using
       available human data.

Comment: Two commenters noted that EPA incorrectly assumed a partition factor of 100 for
TCDD in human fat compared to blood. The commenters state that available human data
demonstrate that the actual partition factor is between 150 and 200 (lida et al., 1999; Patterson et
al..  1989).

       Response: While EPA has not changed the value in the model,  a sensitivity analysis was
       conducted that indicated this is not a sensitive parameter in the model (see  response to
       Recommendation 25).

Comment: Some commenters felt that use of modeled concentrations is not acceptable for
deriving toxicity values when measured data are available. The commenters noted that EPA's
use of modeled whole-blood concentration results in underestimation of PODs, HEDs at the
BMDLs, and calculated reference dose.

       Response: EPA modeled the blood concentrations for the  rat exposures in NTP (2006),
       when actual liver and fat TCDD concentrations were reported in the study.  This was
       done primarily for consistency across all rat bioassays.  The whole liver concentrations
       are not likely to be relevant because they include TCDD bound to CYP1A2, which is not
       part of the biologically-active TCDD fraction.  However, in response to this comment,
       EPA has added a sensitivity analysis (See Section 4.5.1.2.) that evaluates the effect of
       using the measured fat TCDD concentrations on modeled  human intakes based on (NTP,
       2006).

Comment: Several commenters noted that the Emond et al. (2005) PBPK model did not account
for the enhanced elimination rate of TCDD observed in infants and children, which would
substantially underestimate the daily dose rates associated with identified target body burdens,
and, thus, underestimate the derived RfD estimated in modeling for the Mocarelli et al. (2008)

                                         A-19

-------
data set. Commenters provided references of Clewell et al. (2004), Ott et al. (1987), Hochstein
et al. (2001), Kerger et al. (2006), Leung et al. (2006), and Milbrath et al. (2009) and suggested
that EPA address the role of differential elimination rates in children in their quantitative analysis
of a reference dose.

       Response: The changes in elimination rate with age reported in Kerger et al. (2006) are
       thought to reflect growth processes as a child ages. The Emond PBPK model accounts
       for this phenomenon implicitly by modeling growth and age-related changes in fat
       content and physiology explicitly. Including an explicit variable-elimination term in the
       model would then "double count" for this effect. The TCDD half-life calculations in
       Kerger et al. (2006) are based on blood level rather than whole-body measurements.
       Blood levels  of the chemical are influenced by the dynamic processes of storage in fat
       deposits and elimination rates (including binding to proteins in the liver).  The inclusion
       of these physiological process and the dynamic interplay among them provide the
       biological basis for an observed increase in elimination rate in children. At early life
       stages, less fat volume in the body results in more TCDD available for deposit in liver.
       More TCDD in the liver results in a higher elimination rate.  Leung et al. (2006) indicated
       that the more rapid clearance in children was due to their lower fat content, which is
       accounted for in the model.

Comment: A commenter noted that non-TCDD TEQ contributes to the induction of CYP1A2,
which will influence the elimination rate for TCDD.  Given the current background body
concentrations of TCDD and other TEQ contributors, the commenter felt that the appropriate
application of the PBPK model would be to start from current background concentrations
(including some accounting for non-TCDD TEQ).

       Response: Induced levels of CYP1A2 due to dioxin are calculated using a Hill function.
       The relative difference between induced levels of CYP1A2 and basal levels of the
       enzyme are then used to describe the dose-dependent elimination rate for TCDD in the
       liver.  Application of the PBPK model to estimate the elimination of TCDD is based on
       an assumption that background effects of dioxin-like chemicals and any others that may
       influence CYP1A2 levels in the liver are implicitly included in the basal-level estimates.
       EPA also added a simulation of total TEQ background exposure as a sensitivity analysis
       in Section 4.5 to investigate this phenomenon. Issues pertaining tomodeling non-TCDD
       TEQ are discussed in Section 4.5 and, also in this Section, EPA has presented several
       alternative approaches for incorporating background DLC exposure into the derivation of
       the RfD.  In the sensitivity analysis, EPA estimates that average total-TEQ PODs based
       on background non-TCDD TEQ exposures could range from no change to the POD to
       2.5-fold higher than the TCDD-only POD of 0.02 ng/kg-day used in the derivation of the
       RfD.

Comment: Several commenters noted deficiencies and limitations with the PBPK model, and
some stated that EPA failed to adhere to its own guidance on selection and application of PBPK
models (i.e., U.S. EPA (2006a), Guidelines on PBPK Model Selection in Risk Assessments
report). Specifically, the PBPK model was not peer reviewed and was not validated.
Two commenters noted a need for an uncertainty analysis of key parameters in the model, such
as the Hill coefficient.

                                         A-20

-------
       Response: Although EPA used revised estimates for some of the published parameters,
       no modifications were made to the structure of the Emond model. Using these revised
       parameters, EPA describes the evaluation of the PBPK model in Section 3. Also, see
       the response to Recommendation 25 concerning the sensitivity analysis.

A.4. REFERENCE DOSE
A.4.1.  SAB Comments and EPA Responses
SAB Charge Question 4.1
The Mocarelli et al.  (2008) and Baccarelli et al. (2008) studies were selected as co-critical
studies for the derivation of the RfD. Is the rationale for this selection scientifically justified and
clearly described? Please identify and provide the rationale for any other studies that should be
selected, including the rationale for why the study would be considered a superior candidate for
the derivation of the RfD. In addition, male reproductive effects and changes in neonatal thyroid
hormone levels, respectively, were selected as the co-critical effects for the RfD. Please
comment on whether the selection of these critical effects is scientifically justified and clearly
described. Please identify and provide the rationale for any other endpoints that should be
selected as the critical effect.

Comment: The use of the Mocarelli et al. (2008) and Baccarelli et al. (2008) studies was
appropriate for identifying "cocritical" effects for the RfD calculation, and the rationale for
selecting these two studies over others was clearly described. However, the weaknesses of the
two studies were not always clearly delineated. For  example, in the Baccarelli (2008) study,
there was limited discussion of how the presence of polychlorinated dibenzo-p-dioxins (PCDDs),
polychlorinated dibenzofurans (PCDFs), and coplanar polychlorinated biphenyls (PCBs) that
were also found in the blood might confound the interpretation of TCDD association with
elevated TSH levels. In addition, there was no discussion of the potential impact of residential
histories (e.g., individuals who may have moved in and out of Zone A after the accident). The
Panel believes that more discussion of the strengths and weaknesses of these two  studies is
needed.
       The Panel found that in isolation from each other, and lacking a description of supportive
animal and epidemiological studies, the studies were less useful for setting the RfD, and
emphasizes the need to  consider supportive animal and epidemiological studies for dioxin and
dioxin-like compounds  in order to demonstrate a consistent and integrative signal of toxicity
across species and endpoints for TCDD. While Figures 4.3 and 4.4 show quantitative
comparisons across RfDs and benchmark dose lower bounds (BMDLs) from animal and
epidemiological studies, the figures do not indicate which endpoints are being measured, and
consistency in signal is  not readily apparent.
       The Panel noted that although it has been addressed in the Report, the discussion of the
known human age-specific variability in endpoints such as sperm counts should be expanded,
though the data from Mocarelli et al. (2008) do show ranges and variance (in Figure 3 and
Table 2), and neonatal TSH levels.

   Recommendation No. 32: Provide a discussion of the strengths and weaknesses  of the
   Mocarelli et al. (2008) and Baccarelli et al. (2008) studies with an indication of whether the
   weaknesses affect determination of the RfD.

                                         A-21

-------
   Response: In Appendix C, EPA presents an assessment of both the Baccarelli et al.
   (2008) and Mocarelli et al. (2008) studies, delineating their strengths and weaknesses.
   Section 4.4 identifies and describes qualitatively a number of uncertainties associated
   with the derivation of the RfD from the Baccarelli et al. (2008) and Mocarelli et al.
   (2008) studies. Additionally, in Section 4.5.1, EPA presents a quantitative sensitivity
   analysis that highlights the uncertainty associated with deriving an RfD from the
   Baccarelli et al. (2008) and Mocarelli et al. (2008) studies. In this analysis, EPA focused
   on several important assumptions that were made in defining variables for modeling the
   exposure history of the cohorts and in estimating a chronic intake leading to the observed
   effect; the analysis presents the quantitative impact of making alternative assumptions for
   those variables on the POD estimates. EPA also modeled the potential impact of
   background DLC exposure on the PODs derived from both of the principal  studies. EPA
   did not discuss the potential impact of residential histories because the PODs from both
   studies were based entirely on measured serum TCDD concentrations, irrespective of
   zone of residence. Zonal averages were not used in any way in the derivation of the RfD.
          With respect to age-specific variability in sperm concentrations as relates to the
   interpretation of Mocarelli et al. (2008), EPA notes that all the men evaluated in the study
   were between the ages of 22 and 31 at the time of semen collection and would not expect
   any substantial age-related differences.  EPA does present group sperm concentrations at
   one standard deviation below the mean as reported by Mocarelli et al. (2008).

Recommendation No. 33: Label the endpoints for studies included in Figures 4.3 and 4.4.

   Response: EPA agrees with the SAB Panel's recommendation and has modified
   Figure 4-4 by adding the last name of the first author of each study and the year of
   publication and Figure 4-5 by adding the health endpoint or health outcome as suggested.
   Table 4-5 lists the study endpoints described in Figure 4.3 along with other study
   information.

Recommendation No. 34: Discuss the comprehensive database of both animal studies and
human epidemiological studies, including studies with dioxin-like compounds (e.g., studies
cited in Goodman et al. (2010), together to demonstrate a consistent and integrative signal
of toxicity across species and endpoints for TCDD.

   Response: EPA methodology does not require that a consistent and integrative signal of
   toxicity across species and endpoints be demonstrated for derivation  of an RfD.
   However, concordance of effects, both qualitatively and quantitatively, across endpoints
   and species is considered, primarily in the assessment of confidence in the RfD. In
   response to this recommendation and consistent with EPA methodology, EPA has
   modified the Reanalysis as follows.
       Section 4.3.6 has been revised to provide additional supporting information for the
   critical effects noted in the two co-principal studies: neonatal thyroid effects from
   Baccarelli et al. (2008) and sperm effects from Mocarelli et  al. (2008).
       In Section 4.3.6.1, EPA has evaluated the Goodman et al. (2010) review and added
   a discussion of the findings. EPA concluded that, because  of relatively low DLC
   exposures in the studied populations and different timings of measurements in the cited
   studies, it would be unlikely that any consistent patterns would be detected.  EPA

                                      A-22

-------
       confirmed that there were no additional studies identified in this review that meet the
       selection criteria outlined in Section 2.
           EPA has added an analysis of the qualitative and quantitative concordance of key
       effects across species and studies in Appendix D and referenced in Section 4.4 as part of
       the discussion of qualitative uncertainty in the RfD.  The analysis includes effects from
       all of the animal and human studies listed in Table 4-5 in six categories: male
       reproductive effects, female reproductive effects, developmental effects,
       immunotoxicity, neurotoxicity, and thyroid toxicity.  Coverage of effects was expanded
       beyond those in Table 4-5 to include effects at doses higher than the LOAEL in each
       study.

SAB Charge Question 4.2
In the Seveso cohort, the pattern of exposure to TCDD is different from the average daily
exposure experienced by the general population. The explosion in Seveso created a high-dose
pulse of TCDD followed by low-level background dietary exposure in the exposed population. In
the population, this high-dose pulse of TCDD was slowly eliminated from body tissues over time.
There is uncertainty regarding the influence of the high-dose pulse exposure on the effects
observed later in life.

SAB Charge Question 4.2.a
Mocarelli et al. (2008) reported male reproductive effects observed later in life for boys exposed
to the high dose pulse of TCDD between the ages of 1 and 10. EPA identified a 10 year critical
exposure window. In the development of the  candidate RfD, EPA used an exposure averaging
approach that differs from the typical approach utilized for animal bioassays.  EPA determined
that the relevant exposure should be calculated as the mean of the pulse exposure and the
10-year critical exposure window average. Please comment on the following:

SAB Charge Question 4.2.a.i
EPA 's approach for identifying the exposure window and calculating average exposure for this
study.

Comment: The Panel discussed extensively extrapolation issues posed by the pattern  of exposure
from Seveso. Issues raised included the question of whether the same endpoints and/or dose
response would be expected from such exposure scenarios with high-dose acute exposures when
extrapolating to low-dose chronic exposures.

    Recommendation No. 35: Provide a discussion of published examples in which dioxin
    studies were conducted using both high-dose acute and low-dose chronic exposures in
    animals for the same endpoint and how the outcomes compare both qualitatively and
    quantitatively.  Determine whether similar results were observed for similar endpoints.
    Several chronic dioxin animal studies may be useful in this regard (Sandet al., 2010;
    Yoshizawa et al.. 2010: 2009).

       Response: EPA is aware of only one rodent toxicology study—Kim et al.
       (2003)—directly comparing health outcomes following the administration of either a high
       acute TCDD dose or a low longer-term continuous TCDD dose in animals where the

                                         A-23

-------
       long-term average tissue TCDD concentrations in both dose groups were comparable; the
       effects were more severe for the acute exposure regimen.
           Another animal study, Sand et al. (2010), used an initial-loading dose,
       weekly-maintenance-dose protocol in which the loading dose is 10 times higher than
       the weekly maintenance dose but did not evaluate the equivalent continuous exposure,
       and so does not inform the issue. Both of the Yoshizawa et al (2010; 2009) studies
       were analyses of the NTP (2006) study that is already presented in the Reanalysis, and
       has no acute vs. continuous component. One other study, Bell et al. (2007), mentioned
       in Recommendation 37 following, allows for acute/continuous comparison for in utero
       and lactational exposures, addressing a very different developmental period than the one
       in question for the Seveso cohort children (average age >6 years).  This study found that
       acute exposure had a significantly lower impact on preputial separation in male rat pups
       than did the equivalent continuous exposure (similar terminal TCDD body burdens), the
       opposite of the finding of Kim et al. (2003). EPA does not consider this finding very
       informative for the specific exposure scenario and critical exposure period relevant to
       theRfD.

   Recommendation No. 36: Discuss the life-stage-specific approach to hazard and
   dose-response characterization for children's health risk assessment found in EPA's
   Framework for Assessing Health Risks of Environmental Exposures to Children (U.S.
   EPA. 2006b).

       Response: The approach outlined in EPA's  Framework for Assessing Health Risks of
       Environmental Exposures to Children (U.S. EPA, 2006b) encourages evaluation of the
       potential for toxicity during all developmental lifestages, based on knowledge of external
       exposure, critical windows of development for different organ systems, MO As, anatomy,
       physiology, and behavior that can affect external  exposure and internal dose metrics.
       EPA has followed the framework in evaluating the available data for TCDD and in
       developing the Reanalysis.  The concepts explored in this framework are those that apply
       to all risk assessments—namely problem formulation, analysis, and risk characterization.
       The Reanalysis is not a risk assessment but rather a hazard identification and
       dose-response assessment for noncancer outcomes. It does not contain information on
       problem formulation or risk characterization; however, it does follow standard EPA
       procedures.

   Recommendation No. 37: Consider adding to the discussion, Bell et al. (2010), which
   summarized and presented data on some differences between chronic versus acute exposure
   in maternal transfer.

       Response: EPA considered this recommendation as discussed in the response to
       Recommendation 35.  An analysis of the data has led EPA to consider the findings of
       Bell et al. (2010) not to be informative in the context of the Seveso exposures on which
       the RfD is based.

SAB Charge Question 4.2.a.ii
Please comment on EPA 's designation of a 20% decrease in sperm count (and an 11% decrease
in sperm motility) as a LOAELfor Mocarelli et al. (2008).
                                         A-24

-------
Comment: The Panel found that changes from normal sperm counts and sperm motility are of
public health relevance and, therefore, of interest for determining an RfD.  There is general
support for EPA's approach of using the WHO reference value for determining relevant TSH
levels, but the Panel feels that further discussion of WHO reference values for male reproductive
parameters should be included in the Report.  Additionally, the Report should indicate that life
stage differences clearly exist in sperm counts in humans; cite and discuss the EPA life stage
document (U.S. EPA. 2006b).

    Recommendation No. 38: Include discussion of background information regarding WHO
    reference values for male reproductive parameters (e.g., Skakkebaek,  2010).

       Response: EPA has added additional discussion of WHO reference values for male
       reproductive parameters and a discussion of the Skakkebaek (2010) study in
       Section 4.3.4.2.

    Recommendation No. 39: Discuss standard deviations or range of changes from the
    Mocarelli (2008) study to provide a better understanding of the potential magnitude of
    effect.

       Response: In Section 4.3.4.2, EPA discusses the magnitudes and standard deviations of
       the effects reported in Mocarelli et al.  (2011).

SAB Charge Question 4.2.b
For Baccarelli et al. (2008), the critical exposure window occurs long after the high-dose pulse
exposure. Therefore, the variability in the exposure over the critical exposure window is likely
to be less than the variability in the Mocarelli et al. (2008) subjects.  EPA concluded that the
reported maternal exposures from the regression model developed by Baccarelli et al. (2008)
provide an appropriate estimate of the relevant effective dose as opposed to extrapolating from
the measured infant TCDD concentrations to  maternal exposure. Additionally, EPA selected a
LOAEL of 5 i^-units TSH per ml blood in neonates; as this was established by World Health
Organization (WHO) as a level above which there was concern about abnormal thyroid
development later in life.  Please comment on the following:

SAB Charge Question 4.2.b.i
EPA 's decision to use the reported maternal levels and the appropriateness of this exposure
estimate for the Baccarelli et al. (2008) study.

Comment, The Panel supports EPA's decision to use the Baccarelli et al. (2008) estimates of the
relevant effective doses. Because the bulk of the calculations were based on zonal averages,
clarify how these measurements relate to ranges and variations in exposure in utero.

       Response: The Baccarelli et al. (2008) calculations presented in the Reanalysis are
       derived from the individual exposure measures by the study authors and are not based on
       zonal averages.  EPA has clarified this for the RfD derivation in Section 4.3.
                                          A-25

-------
SAB Charge Question 4.2.b.ii
EPA 's designation of 5 ju-units TSHper ml blood as a LOAELfor Baccarelli et al. (2008).

Comment: The change in TSH levels reported by Baccarelli et al. (2008) was of public health
relevance and, therefore, of interest for determining an RfD.  Any follow-up data on thyroid
hormone levels in the population studied should be discussed in the Report, if available.

    Recommendation No. 40: Better describe the potential adverse health outcomes related to
    altered neonatal TSH levels (e.g., effects on both cognitive and motor deficits). For
    example, in addition to effects on growth, both cognitive and motor deficits have been
    found in young adults with congenital hypothyroidism (Oerbeck, 2007, 2003).  The Report
    could better describe the consequences of transient hypothyroidism on reproductive
    outcomes (e.g., Anbalagan et al., 2010).  Other references that relate to this question
    include Chevrier et al. (2007). Dimitropoulos et al. (2009). and Ye (2008).

       Response: EPA has added a discussion of the potential adverse health outcomes
       associated with altered neonatal  TSH levels in Section 4.3.4.1.  The discussion includes
       information about thyroid hormone disruption during pregnancy and the neonatal period,
       potentially leading to neurological deficiencies, particularly in the attention and memory
       domains(Oerbeck et al., 2005). It also addresses some of the uncertainties in the
       relationship between human neonatal TSH levels and measures of neurological function
       such as IQ. EPA also identified animal bioassays, reporting that perturbations in thyroid
       status can lead to altered brain development  (e.g., Sharlin et al., 2010;  Roviand et al.,
       2008; 2008; Auso et al., 2004: Lavado-Autric et al., 2003). Discussion of these findings
       has been added to Section 4.3.4.1.

SAB Charge Question 4.3
Please comment on the rationale for the selection of the uncertainty factors (UFs) for the RfD. If
changes to the selected UFs are proposed, please identify and provide a rationale.

Comment: The Panel agrees that the appropriate UFs were included. The exclusion or inclusion
of the UFs in the Report is obvious, clearly discussed, and adequately rationalized. The Report
would be more transparent if EPA included a short discussion for the basis of the decision not to
include a UF for data quality.

       Response: EPA has clarified its  choice of UFs for the candidate RfDs  in Section 4.3.5
       and Table 4-7.

SAB Charge Question 4.4
EPA did not consider biochemical endpoints (such as CYP induction, oxidative stress,  etc.) as
potential critical effects for derivation of the RfD for TCDD due to the uncertainties in the
qualitative determination of adversity associated with such endpoints and quantitative
determination of appropriate response levels for these types of endpoints in relation to TCDD
exposure.  Please comment on whether the decision not to consider biochemical endpoints is
scientifically justified and clearly described.
                                          A-26

-------
Comment: Biochemical endpoints such as P450 activation, increased oxidative stress, etc. may
be acceptable endpoints to establish PODs, particularly when the quantitative relationship
between the biochemical endpoint and an adverse health outcome is clearly evident. However,
with respect to TCDD, the Panel agrees that more traditional endpoints (e.g., immune, endocrine,
reproductive) are more appropriate because associations of these endpoints with health outcomes
are well studied and provide a stronger association to an adverse outcome than biochemical
endpoints.  However, because of the wealth of data on P450s and their importance in disease
development, normal development, and chemical response to exogenous agents, EPA should
discuss biochemical endpoints, particularly P450s, relevant to establishing and strengthening the
proposed reference dose.

       Response:  In general, there is a lack of information linking these particular endpoints to
       downstream adverse effects for the noncancer effects observed in the available studies.
       Some of these  endpoints, such as CYP (P450) induction  and oxidative stress are
       discussed in Section 5 of the 2010 External Review Draft of the Reanalysis in the context
       of the mode or action for carcinogenesis or are evaluated quantitatively as potential
       cancer precursor effects. EPA intends to consider these endpoints further in Volume 2 of
       the Reanalysis. In the context of noncancer effects, however, an expansive coverage of
       these endpoints will not necessarily provide a better understanding of the RfD, given the
       lack of information on the relevant modes of action. For these reasons, further analysis
       of these data with respect to their relevance to strengthening the reference dose was not
       conducted.

SAB Charge Question  4.5
In using the animal bioassays, EPA averaged internal blood TCDD concentrations over the
entire dosing period, including 24 hours following the last exposure. Please comment on EPA 's
approach for averaging exposures including intermittent and one day gestation exposure
protocols.

Comment: For animal studies, it has been shown that for some effects, acute exposure could give
different results than chronic exposure. For TCDD, however,  its persistence might suggest that
such differences would be partly negated. In Baccarelli et al. (2008), there was extensive
discussion regarding the use of the exposure average time for the TCDD concentrations. This is
of biological significance as several papers have indicated the unique aspects of high peak
exposure of TCDD as occurred in Seveso and in several of the animal studies.  The endpoints
affected as a result of these peaks do not always translate to impacts from lower chronic
exposures. It would be helpful to discuss any available animal studies comparing high-dose
acute versus low-dose chronic effects on similar endpoints for dioxin or dioxin-like compounds
(as stated earlier in this section).

       Response: See EPA's response to Recommendation 35.  For the Baccarelli et al. (2008)
       study, the exposures over the critical exposure window (gestation) were relatively
       constant compared to the exposures experienced by the subjects studied in Mocarelli
       et al. (2008) and other Seveso  cohort studies.
                                          A-27

-------
SAB Charge Question 4.6
Please comment on the benchmark dose (BMD) modeling conducted by EPA to analyze the
animal bioassay data and EPA 's choice of points of departure (PODs)from these studies.

Comment: The Panel agrees with the BMD modeling approaches used in this section.  However,
the justification for EPA's  conclusions that the animal data had sufficient limitations that
precluded their use to establish an RfD is quite diverse and poorly linked to specific studies.

   Recommendation No.  41: Discuss several of the best animal studies in some detail so that
   their limitations are more apparent.

       Response: Summaries of all of the studies are presented in Appendix D, with some
       discussion of their limitations.  Strengths and limitations of all of the animal bioassays at
       the lower end of the candidate  RfD range  are presented in Table 4-6. Two studies of note
       (Bell et al., 2007; NTP, 2006) are discussed in more detail in Section 4.4.  Table 4-4 and
       Appendix G, which summarizes the  BMD modeling, highlight some of the limitations of
       the BMD modeling for each modeled data set.

   Recommendation No.  42: Better  cite the endpoint guidance that is present within EPA
   documents for defending approaches used and application of BMD models for the critical
   effects: this is especially necessary given public comments that EPA was not following  its
   own guidelines.

       Response: In response to this comment, EPA has added Text Box 2-1.  In this text box,
       EPA identifies the risk assessment guidelines and guidance documents that it relied upon
       during development of the dose-response  assessment.

SAB Charge Question 4.7
For the animal bioassay modeling, EPA applied the kinetic extrapolation at the level of the POD
prior to applying the uncertainty factors because EPA has less confidence in the kinetic model
output at lower doses reflective of the RfD.  Please comment on whether the kinetic extrapolation
at the level of the POD prior to applying the uncertainty factors was scientifically justified and
clearly described.

Comment: The EPA approach of applying the kinetics on the actual data present at the POD is
preferred in this assessment (see additional discussion in the response to Charge Question 3).

       Response: No response necessary.

SAB Charge Question 4.8
Please comment as to whether EPA 's qualitative  discussion of uncertainty in the RfD is justified
and clearly described.

Comment: The Panel agreed that EPA provided a clear and justified discussion of the
uncertainties in deriving the RfD using the Seveso cohort. The Panel agrees with EPA that the
major limitation of the Seveso cohort is the uncertainty arising from how well the  effects

                                          A-28

-------
resulting from high-dose acute exposure translate to low-dose daily exposures. It may be useful
to re-review the animal studies to identify if there are any studies where dioxin or DLCs were
administered by acute as well as chronic (or even subchronic), and comparable endpoints were
examined. If so, the information can be used to help confirm or refute the accuracy of the
"average daily dose" adjustment. This is of particular concern in the Mocarelli study as "time
periods of susceptibility" appear in male reproductive development, and these periods (windows)
may be very short.  Animal studies, particularly those involving male reproduction, may be
helpful.

   Recommendation No. 43: It would be useful to include a discussion of potential
   uncertainty in the exposure estimates from the Baccarelli study. Serum dioxin levels were
   only established in a subset of the cohort (approximately 51) at the time of the study while
   dioxin levels from the main cohort  were estimated from data collected from zone of
   residence (A or B) at a much earlier time.

      Response: For derivation of the  POD, EPA used the regression modeling in Baccarelli
      et al. (2008), which was based only on the 51 infants with maternal TCDD measurements
      taken between 1992 and 1998 and did not depend on prior measurements in the main
      cohort. All  outcomes evaluated  for the derivation of the RfD are associated with
      individual serum concentrations rather than zonal averages. Baccarelli et al. (2008)
      extrapolated the measured values to the time of conception for each  of the
      51  pregnancies, which occurred  between 1994 and 2005. In Section 4.4, EPA has
      identified and clarified the qualitative uncertainties associated with deriving an RfD from
      both of the principal studies (Baccarelli et al., 2008; Mocarelli et al., 2008). EPA has
      also added Section 4.5.  In this section, EPA quantifies the impact of alternative
      assumptions about the exposures and pharmacokinetic for both the Baccarelli and
      Mocarelli studies. Also, see response to Recommendation 32.

   Recommendations No. 44: While the Panel agrees that the true dioxin-like-compound
   impact cannot be determined, it might be helpful to provide some general estimates of the
   variability that may occur at the proposed RfD.

      Response: In response to this comment, EPA has added Section 4.5  to the document. In
      this section, EPA quantifies the  impacts of alternative assumptions about the TCDD-only
      and DLC exposures on the PODs for both the Mocarelli (see Section 4.5.1.1.1) and
      Baccarelli (see Section 4.5.1.1.2) studies. In Section 4.5.2, EPA has estimated alternative
      PODs from  the NTP (2006) study based on different approaches to modeling TCDD only
      and the DLCs. In Section 4.5.2, EPA also has estimated potential PODs from  several
      different endpoints identified in  Seveso cohort studies (other than those used in
      developing the RfD) and has estimated the range of potential PODs based on
      uncertainties encountered in their analyses;  these uncertainties included the impacts of
      DLC background exposures.

A.4.2. Summary of Public Comments and EPA  Responses
Comment. Several  comments addressed the fact that when determining an RfD, EPA accounted
for only 2,3,7,8-TCDD exposures and did not account for exposures to dioxin-like chemicals.
The commenters noted that in human epidemiological studies, people are exposed to all
                                         A-29

-------
dioxin-like compounds regardless of the sources of their exposures. Specifically, the
commenters suggested that EPA did not account for these exposures in the Seveso population
when evaluating dose response and, thus, underestimated the reference doses derived from
Mocarelli et al. (2008) and Baccarelli et al. (2008).
       Response: EPA agrees that the human subjects studied in the epidemiological studies
       were subject to background DLC exposures from many sources. As a component of a
       sensitivity analysis, EPA has added an analysis of the impact of background DLC
       exposures on the RfD to the document in Section 4.5. In this analysis, EPA estimates
       background DLC exposures for several of the Seveso exposure scenarios, including those
       relevant to the Mocarelli et al. (2008) and Baccarelli et al. (2008) POD estimates. EPA
       summarizes the results of these sensitivity analyses in Figures 4-6 through 4-9.

Comment: One commenter noted that EPA's qualitative discussion of uncertainty in the
reference dose (pp. 4-28 to 4-32) is well written and clearly described.  Two commenters felt that
the rationale for the selection of the male reproductive effects (Mocarelli et al., 2008) and
changes in neonatal thyroid hormone levels (Baccarelli et al., 2008) as critical effects was clearly
described  and scientifically justified.  One commenter felt that the LOAEL selected from the
Mocarelli  et al. (2008) study was justified.  Commenters also felt that EPA's decision not to
consider biochemical endpoints (such as CYP induction, oxidative stress, etc.) as potential
critical effects for derivation of the RfD for TCDD is clearly described and scientifically
justified.

       Response: No response necessary.

Comment: Several commenters asked EPA to further address the uncertainties associated with
deriving an RfD from the Baccarelli  et al. (2008) and Mocarelli et al. (2008) studies.  Several
commenters noted that EPA does not include the use of the data from these studies for dose-
response modeling and reference dose derivation with a discussion of the clinical significance of
the effects, or the levels of change that represent an adverse effect for each endpoint.

       Response: In Section 4.4, EPA presents a discussion of the qualitative uncertainties
       associated with the development of an RfD from these two studies.  In response to this
       and other comments, EPA has expanded the discussion to include the potential clinical
       significance of the two effects encountered in these epidemiological studies: (1) elevated
       TSH levels in infants and (2) decreased semen quality in men that experienced elevated
       TCDD exposures as young boys. Further, in the sensitivity analysis added in Section 4.5,
       EPA evaluates some quantitative uncertainties in the derivation of PODs from the
       Baccarelli et al. (2008) and Mocarelli et al. (2008) studies.
Comment: Two commenters noted that the Agency substantially underestimated liver and
adipose tissue concentrations in the 2006 National Toxicology Program bioassay (NTP, 2006),
resulting in an approximate two-fold overestimate of TCDD potency.  EPA ignored reported
TCDD concentrations in adipose and liver tissue, which should have been used as the dosimetry
endpoints for extrapolation to human equivalent dosages.  The use of modeled data is not
acceptable for deriving toxicity values used in risk assessment when measured data are available;
unnecessary inaccuracies in the derivation of the RfDs are introduced.

                                          A-30

-------
       Response: In the sensitivity analysis presented in Section 4.5.2, EPA has estimated PODs
       based on the TCDD adipose concentrations reported in NTP (2006). EPA does not
       consider the whole liver concentrations to be relevant because they include TCDD bound
       to CYP1A2, which is not part of the biologically-active TCDD fraction. Because
       adequate human studies were available, animal studies including the above referenced
       NTP (2006) were not used to derive the RfD.

Comment: One commenter noted that several studies included in the Report examined the
effects of TCDD exposure on serum thyroid hormone concentrations (Crofton et al., 2005; Seo et
al., 1995; Sewall et al., 1995), which are lexicologically irrelevant and should be excluded from
the analysis.

       Response: EPA considers serum thyroid hormone levels to be lexicologically relevant, as
       indicators of hormonal imbalance and potential thyroid toxicity. EPA does not require
       the observation of overt clinical effects in this respect.

Comment: A commenter suggested that many of the animal studies, particularly developmental
studies, used dosing regimens that cannot be properly extrapolated to chronic exposures and,
thus, are inappropriate for derivation of a chronic RfD. The commenter noted that the weight of
evidence suggests that peak, rather than average, exposure level is most relevant to assessing the
effect of in utero and developmental exposure to TCDD on male rat reproductive system
parameters.

       Response: EPA defines the RfD as a lifetime protection value that includes all exposures
       and life  stages, not just long-term exposure. If shorter-term exposures over a particular
       critical window, such as in utero or early childhood, indicate greater susceptibility, the
       short-term exposures must be considered during the development of an RfD and can be
       the basis of an RfD. EPA has removed the word "Chronic" from the title of Section 4 in
       the Reanalysis to avoid confusion. EPA did not distinguish between peak and average
       exposure levels when evaluating male rat reproductive system effects because
       administered doses were fairly level, unlike the exposure scenario evaluated for the
       Seveso cohort.

Comment: A commenter noted that some of the health effects that are addressed in derivation of
an RfD are actually precancerous lesions (i.e., hypertrophy and hyperplasia), and as such, are
more appropriate for use in cancer risk assessment than for deriving a chronic RfD.

       Response: Hypertrophy and hyperplasia are not always considered to be precancerous.
       For the TCDD assessment, no POD is based solely on either of these effects.

Comment: One commenter noted that in developmental studies, the appropriate unit for
statistical analysis is the litter; many of the developmental studies considered by EPA, however,
incorrectly used the individual pup as the statistical unit for analysis (e.g., Shi et al., 2007; Hojo
et al.,  2002; Markowski et al., 2001: Ohsako etal., 2001). The commenter suggested that data
from developmental studies that have been incorrectly evaluated using the  individual pup should
not be used as the basis for derivation of an RfD. Alternatively, the original study  data could be
reanalyzed using the litter as the statistical unit of analysis.

                                          A-31

-------
       Response: EPA guidance calls for a litter-based approach for dichotomous outcomes
       when the data are reported on that basis.  All the endpoints in the studies identified by the
       commenter were continuous measures, to which the guidance does not apply. In
       addition, all the data were presented only by aggregated exposure groups, so that a
       litter-based analysis was not possible even if the responses could be dichotomized.

Comment: One commenter noted that some data are derived from guinea pigs, which are known
to be substantially more susceptible to the effects of TCDD treatment than humans.  Because of
the extreme sensitivity, an uncertainty factor of 3 for animal-to-human extrapolation is
unfounded for these studies.

       Response: There are few data to evaluate the relative sensitivities of guinea pigs and
       humans to TCDD. As shown in Table 4-5, guinea pigs are not necessarily more sensitive
       than other species. The use of a three-fold uncertainty factor for the toxicodynamic
       component of interspecies uncertainty (UFA) is standard EPA practice when using
       modeling the toxicokinetic extrapolation component (U.S. EPA, 1994).

Comment: One commenter suggested that several studies included in the analysis are limited by
the number of animals used (see Shi et al.,  2007; Franc et al., 2001; Sewall et al., 1995) and that
the determination of a NOAEL and LOAEL based on  the analyses as provided by the authors is
not appropriate for deriving a regulatory threshold value.

       Response: EPA has indicated such limitations  in the animal bioassay evaluations in
       Table 4-6. While EPA considered these  studies as possible POD candidates, the RfD is
       based on human epidemiological studies, not on data derived from animal bioassays.

Comment: One commenter felt that the LOAELs in the Van Birgelen et al. (1995a:  1995b) and
Fattore et al. (2000) studies were incorrectly  interpreted.  The commenter noted that, in the Van
Birgelen et al. (1995a: 1995b) study, the LOAEL should be based only on changes in thymus
weight because  other changes (i.e., liver retinoid levels) might only be adaptive responses and
cannot be considered toxic effects.  The commenter also noted that the LOAEL for the Fattore
et al. (2000) study should be interpreted as a  l-|ig/kg diet (2 |ig/day for 13-week  old female rats)
with a NOAEL of 0.2 |ig/kg (0.3 |ig/day for 13-week-old female rats) because of the
dose-dependent reduction in hepatic vitamin  A, with significant reductions at TCDD diet
concentrations of 1,  2, and 20 |ig/kg, but not  at 0.2 |ig/kg.

       Response: EPA acknowledges that there are uncertainties in the selection of specific
       effects in these studies but believes that it has appropriately interpreted these study
       endpoints in its development of candidate RfDs.  EPA does not consider depletion of
       liver retinoid levels to be adaptive in the  Van Birgelen et al. (1995a: 1995b) study.

Comment: Several commenters noted that EPA's evaluation of noncancer risk ignored the NAS
peer-review conclusions that the evidence for dioxin exposure as a cause of reproductive and
hormonal abnormalities is not strong and that there is no convincing evidence of adverse,
noncancer effects as a result of dioxin exposure.

       Response: In Sections 2 and 4 of the document, EPA identifies a number of additional
       epidemiology and toxicology studies that support associations between TCDD exposures
                                         A-32

-------
       and noncancer effects. Several important studies in this group (e.g., Baccarelli et al.,
       2008: Mocarelli et al.. 2008: Bell et al.. 2007: NTP. 2006) were published after the NAS
       report was published.

Comment: Some commenters suggested that there is a significant amount of uncertainty in the
Mocarelli et al. (2008) study, given that the reported demographics of the control population
were different from those of the exposure groups, and the study authors had no information on
TCDD levels in the control group.

       Response: The analysis in Mocarelli et al. (2008) was performed by grouped exposures
       across all subjects.  The lowest exposure group, being the reference group for the
       analysis, included individuals from all exposure zones, not just the "control" population
       (the non-ABR zone) mentioned by the commenter.  TCDD serum levels were measured
       in a subset of the non-ABR population as reported in Needham et al. (1997) and
       Mocarelli et al. (1991).  It is not clear how many, if any, of the individual exposures in
       the lowest exposure group were assigned a generic value rather than a measured one.
       Demographic differences among the individuals across all exposure groups were
       identified and considered as covariates in the analysis by Mocarelli et al. (2008).

Comment: One commenter noted that neither Mocarelli et al. (2008) nor EPA has explained the
biological mechanism by which dioxin demonstrated negative effects on sperm concentration in
1- to 9-year-old boys and positive effects on sperm concentration in 10- to 17-year-old boys.
Commenters questioned the study's assumption of 10 as a reasonable age for puberty in boys and
stated that 12-16 years is the average age at onset of puberty.

       Response: EPA agrees with the commenter that the mechanism of toxic action for this
       effect is not known.  For the establishment of an RfD, EPA does not require the
       establishment of a mechanism of toxic action. Neither the study authors nor EPA assume
       10 years to be the age of puberty onset; it is simply the age that the study authors used to
       divide their study population by magnitude of effect.

Comment: In the Baccarelli et al. (2008) and Mocarelli et al. (2008) studies, the populations of
interest were small, especially for the high-exposure group. This leads to questions about the
overall representativeness of the studies.

       Response: Both studies refer to specific age groups, specifically newborn infants and
       young children; therefore, the population is not a representative sample of the general
       population, but of a potentially sensitive population. In part, because of the small sample
       size, EPA used a factor of 3, rather than 1, for UFn to account for the possibility that all
       sensitive individuals might not be represented.

Comment: One commenter felt that the lack of data on maternal iodine status in the Baccarelli
et al. (2008) study could affect the neonatal TSH data.  The authors' explanation that potential
iodine-related effects would affect all  study groups evenly and would not impact the findings
was questionable.

       Response: Baccarelli et al. (2008) discount iodine status in the population as a
       confounder because exposed and referent populations all lived in a relatively small
                                         A-33

-------
       geographical area.  That an iodine deficiency was present in one and not the other is
       unlikely based on iodine levels in the soil.

Comment: One commenter stated that EPA used data that were not clinically significant and did
not demonstrate a dose-response relationship to derive an RfD. In determining the critical effect,
EPA had no information to verify that the persons with the potentially low values were
associated with higher exposures  to TCDD.

       Response: EPA does not require PODs used to derive RfDs to be based on effects that
       have demonstrable clinical significance. EPA has expanded the discussion of the
       potential significance of elevated neonatal TSH levels in the Reanalysis.

Comment: Several comments suggested that EPA did not acknowledge and address in an
appropriate weight-of-evidence evaluation several other credible studies for RfD development.
EPA excluded credible studies showing no adverse effect from dioxin, yet failed to address the
significant uncertainties associated with the studies used. The commenters felt that EPA should
use an approach that includes results from studies that report both positive and negative findings,
incorporates an appropriate dose range, and evaluates a biologically plausible endpoint.

       Response: In response to this comment and others,  EPA  has added an analysis of the
       qualitative and quantitative concordance of specific key effects across species in
       Appendix D.3 as a supplement to the existing discussion of the critical effects in Sections
       4.3 and 4.4.

Comment: Commenters noted that some of the animal studies used to support derivation of a
chronic RfD evaluate nonadverse endpoints, have not been specifically linked to adverse events,
were generally unsuitable, or were of questionable toxicological relevance. See Amin et al.
(2000). Cantoni et al. (19811 Fattore et al. (2000). Hojo et  al. (2002). Hutt et al. (2008).
Kattainen et al. (2001). Keller et al. (2008a: 2008b: 2007).  Li et al. (1997). Miettinen et al.
(2006). and Van Birgelen et al. (1995a: 1995b).
       Response: See response to Charge Question 4.4.
Comment: A commenter noted that some of the studies cited in support of EPA's derivation of
an RfD report findings that conflict with findings of other studies, thus indicating that the
associated responses to TCDD treatment have not been well-elucidated. The commenter also
added that the lack of agreement among studies regarding the evaluated responses following
TCDD treatment suggests that these endpoints likely are not sensitive indicators of
TCDD-mediated effects. Thus, they should not be used to support the derivation of an RfD.
(SeeAmin et al.. 2000: Gravetal.. 1995:  Bjerke and Peterson. 1994: Mably et al.. 1992.)

      Response: EPA's methods for developing RfDs do not require that all studies be positive
      for a given effect and take into account conflicting information when deciding on a
      critical effect. As mentioned previously in response to other comments, EPA has
      expanded the discussion of qualitative and quantitative concordance of effects across
      species and studies (Appendix D.3).
                                          A-34

-------
Comment: Several commenters stated that the sperm quality endpoints used for risk assessment
were of questionable clinical relevance. EPA failed to present a valid analysis of variability of
effects in the control.  The commenters felt that the critical effect should not be based on
"assumed" effects, but rather, on documented effects of clinical concern and that several
scientific and quantitative issues should be addressed regarding the underlying data used to
derive an RfD.

       Response: EPA does not require PODs to be based on effects that have demonstrable
       clinical significance (see response to SAB charge question 4.4).  EPA has framed the
       concern for the sperm quality endpoints in terms of shifts in the distributions of these
       measures in the general population. Such shifts could result in decreased fertility in men
       at the low end of these population distributions. In a new study, Mocarelli et al (2011)
       report that elevated TCDD exposures during and after pregnancy (via breast-feeding) led
       to similar sperm quality degradation. EPA has expanded the  discussion in Section 4.3.4.2
       regarding the significance of this endpoint.

Comment: Some commenters suggested that owing to limitations in  control for confounding
variables, difficulty in translating exposure scenario to the general population, and relevance of
the main outcome measure, the results of the Baccarelli et al. (2008)  study are suitable for
hypothesis generation but are not strong enough on their own for generation of an RfD. The
commenters additionally noted that neither Baccarelli et al. (2008) nor EPA presented any  data
that shows increasing TSH levels in the population during the years when dioxin exposures were
high and decreasing levels in more recent years, specifically the past 20 years.

       Response: Sections 4.4 and 4.5.1.2 describe and quantify the impacts of important
       sources of uncertainty in this analysis.  In response to the issue of historical infant TSH
       levels against changing background exposures, EPA has added a discussion of the
       Goodman et al. (2010) review of this issue in Section 4.3. EPA notes that the SAB
       agreed with the choice of principal studies, including Baccarelli  et al.  (2008).

Comment: Several commenters suggested that EPA did not sufficiently address the
appropriateness of using the Seveso cohort as a basis to derive an RfD, given that the exposure
levels of those nearest the explosion far exceeded what is observed in the general population.
Nevertheless, at least one reviewer felt that EPA was justified in using the exposure estimates
provided by the study authors to quantify exposure for the dose response.

       Response: In response to this comment and similar ones, EPA has, in addition to the
       existing discussion of the Seveso exposure scenarios in Section 4, added a sensitivity
       analysis in Section 4.5 that investigates in more detail the uncertainties in the exposure
       modeling.

Comment: Several commenters felt that the exposures in Seveso also included substantial
exposure to other confounding chemicals that contribute to the overall TEQ, which was not
accounted for in the analysis. They suggested that TCDD comprised only a small fraction  of the
total TEQ.
                                          A-35

-------
       Response: The released fluid mixture at Seveso reportedly contained TCDD, sodium
       trichlorophenate, ethylene glycol, and sodium hydroxide (Mocarelli et al., 2000), but the
       presence of other dioxin-like compounds was not reported.  However, as part of a
       sensitivity analysis, EPA has evaluated the impact of background DLC exposures for the
       Seveso population. In Section 4.5.1, EPA analyzes TEQ estimates based on background
       exposures to DLCs in the Baccarelli et al. (2008) and Mocarelli et al. (2008) studies. In
       Section 4.5.2, EPA analyzes TEQ estimates based on background DLC exposures for
       other studies of the Seveso cohort and has concluded that background DLC exposure is
       relatively small compared to TCDD at the LOAEL POD.

Comment: One commenter noted that, the study by Baccarelli et al. (2008) provided a clear basis
for estimating a NOAEL for impacts on neonatal TSH levels.  The identification of this robust
NOAEL, with substantial support from the weight of evidence from numerous other studies,
provides the basis for reduced uncertainty factors in the derivation of the RfD. The commenter
outlined an alternative method for deriving the RfD using the principal studies that EPA selected,
which included differences in calculating NOAEL/LOAEL values and applied UFs in Baccarelli
et al. (2008).

       Response: The SAB  has agreed with the approach that EPA has taken to derive the RfD
       from this  study. EPA could not define a NOAEL because it is not clear what maternal
       intake should be assigned to the group below a TSH level of 5 jiU/mL. In
       Section 4.5.1.2, EPA quantifies the impact of sources of uncertainty in a sensitivity
       analysis that examines the key elements encountered during the derivation of an RfD
       from Baccarelli et al. (2008), including a potential NOAEL.

Comment: One commenter noted that in the regression analysis plots from Baccarelli et al.
(2008) (Figure 2), which EPA cites as the basis of the RfD derivation,  if a benchmark of
10 |iU/mL had been used rather than 5 |iU/mL, the corresponding POD (in terms of a maternal
plasma TCDD concentration) would be > 1,200 ppt, as compared with 270 ppt.  The resulting
RfD would be about 5-fold higher.  If a 10  jiU/mL benchmark was applied to the Baccarelli et al.
(2008) regression analysis, there would be little basis for comparing exposures, because  no data
points exceeded 10 jiU/mL.

       Response: In Section 4.5.1.2, EPA addresses this issue in a sensitivity analysis of the
       Baccarelli et al. (2008)  study.  In this section, EPA estimates PODs based on alternative
       increases  in the neonatal TSH levels reported at different TCDD levels in Baccarelli et al.
       (2008). The highest TSH level considered for defining an alternate LOAEL was the
       highest one used by Baccarelli et al. (2008) in their regression model. The overall infant
       cohort included a number of TSH levels above 10 jiU/mL, but no maternal  TCDD
       concentrations were available for those infants. As it is impossible to determine what the
       regression slope would be had those data points been included, EPA did not evaluate the
       regression model beyond the highest TSH value in the modeled data set.

Comment: Several commenters suggested changing the uncertainty factors (UFs).  One
commenter suggested that EPA should reduce the intrahuman uncertainty factor (UFH) from 3 to
1 as the critical effects  observed in the co-principal studies were found in sensitive
subpopulations (children, neonates). Another commenter stated that EPA needs to address why

                                         A-36

-------
it did not include a UF to account for the unique susceptibility and vulnerability of children and
why it chose to use a UF of 3 (instead of 10) to account for human interindividual variability.

       Response: For human interindividual variability (UFn), EPA used a factor of 3 (10°'5)
       because the effects were elicited in sensitive populations. A further reduction to 1 was
       not made because the sample sizes were relatively small, which, combined with
       uncertainty in exposure estimation, may not fully capture the range of interindividual
       variability.  In addition, chronic effect-levels are not well defined for humans and could
       possibly be more sensitive.  EPA has added text to Table 4-7 and believes that the
       Report adequately describes the use of UFs.
           In the EPA's RfD methodology, there is not a separate UF to account for the unique
       susceptibility and vulnerability of children. Such differences are accounted for as part
       ofUFH.
                                         A-37

-------
A.5. REFERENCES
Amin, S: Moore, RW: Peterson, RE; Schantz, SL. (2000). Gestational and lactational exposure to
       TCDD or coplanar PCBs alters adult expression of saccharin preference behavior in
       female rats. Neurotoxicol Teratol 22: 675-682. http://dx.doi.org/10.1016/S0892-
       0362(00)00094-5.
Anbalagan, J: Sashi, AM; Vengatesh, G: Stanley, JA; Neelamohan, R; Aruldhas, MM. (2010).
       Mechanism underlying transient gestational-onset hypothyroidism-induced impairment of
       posttesticular sperm maturation in adult rats. Fertil Steril 93: 2491-2497.
Auso. E: Lavado-Autric. R: Cuevas. E: Del Rev. FE: Morreale De Escobar. G: Berbel P. (2004).
       A moderate and transient deficiency of maternal thyroid function at the beginning of fetal
       neocorticogenesis alters neuronal migration. Endocrinology 145: 4037-4047.
       http://dx.doi.org/10.1210/en.2004-0274.
Baccarelli, A; Giacomini, SM; Corbetta, C: Landi, MT; Bonzini, M; Consonni, D; Grillo, P;
       Patterson, DG: Pesatori,  AC: Bertazzi, PA. (2008). Neonatal thyroid function in Seveso
       25 years after maternal exposure to dioxin. PLoS Med 5: e!61.
Bell, PR: Clode, S: Fan, MQ: Fernandes, A; Foster, PM: Jiang, T; Loizou, G: MacNicoll, A;
       Miller. BG: Rose. M: Iran. L: White. S. (2007). Toxicity of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin in the developing male Wistar(Han) rat. II: Chronic dosing causes developmental
       delay. Toxicol  Sci 99: 224-233. http://dx.doi.org/10.1093/toxsci/kfml41.
Bell, PR: Clode, S: Fan, MQ: Fernandes, A; Foster, PM: Jiang, T; Loizou, G: MacNicoll, A;
       Miller, BG: Rose, M: Iran, L; White, S. (2010). Interpretation of studies on the
       developmental reproductive toxicology of 2,3,7,8-tetrachlorodibenzo-p-dioxin in male
       offspring. Food Chem Toxicol 48: 1439-1447.
       http://dx.doi.0rg/10.1016/i.fct.2010.04.005.
Bjerke, PL; Peterson, RE. (1994). Reproductive toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       in male rats: different effects of in utero versus lactational exposure. Toxicol Appl
       Pharmacol 127: 241-249. http://dx.doi.org/10.1006/taap.1994.1158.
Budinsky, RA: Rowlands. JC: Casteel S: Pent. G: Gushing. CA: Newsted, J: Giesv, JP: Ruby.
       MV: Aylward, LL. (2008). A pilot study of oral bioavailability of dioxins and furans
       from contaminated soils: Impact of differential hepatic enzyme activity and species
       differences. Chemosphere 70:  1774-1786.
       http://dx.doi.0rg/10.1016/i.chemosphere.2007.08.035.
Budinsky, RA: LeCluyse, EL; Ferguson, SS: Rowlands, JC: Simon, T. (2010). Human and rat
       primary hepatocyte CYP1A1 and 1A2 induction with 2,3,7,8-tetrachlorodibenzo-p-
       dioxin, 2,3,7,8-tetrachlorodibenzofuran, and 2,3,4,7,8-pentachlorodibenzofuran. Toxicol
       Sci 118: 224-235. http://dx.doi.org/10.1093/toxsci/kfq238.
Budinsky, RA, Kirman, C. R., Yost L. J., Baker, B. F., Aylward,  L. L., Zabik, J. M., Rowlands,
       J. C.. Long. T. F.. Simon. T... (2009). Derivation of Soil Cleanup Levels for 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin (TCDD) toxic equivalence (TEQD/F) in soil through
       deterministic and probabilistic risk assessment of exposure and toxicity. Paper presented
       at Society of Toxicology Annual Meeting, March 2009.
Cantoni, L: Salmona, M: Rizzardini, M. (1981). Porphyrogenic effect of chronic treatment with
       2,3,7,8-tetrachlorodibenzo-p-dioxin in female rats. Dose-effect relationship following
       urinary excretion of porphyrins. Toxicol Appl Pharmacol 57: 156-163.
                                          A-38

-------
Charnley, G, Kimbrough, R. D. (2006). Overview of exposure, toxicity and risks to children
       from current levels of 2,3,7,8-tetrachlorodibenzo-p-dioxin and related compounds in the
       USA. Food Chem Toxicol 44: 601-615.
Cheng. H: Aylward. L: Beall. C: Starr. TB: Brunei RC: Carrier. G: Delzell. E. (2006). TCDD
       exposure-response analysis and risk assessment. Risk Anal 26: 1059-1071.
       http://dx.doi.org/10.1111/j. 1539-6924.2006.00800.X.
Chevrier, J: Eskenazi, B; Bradman, A; Fenster, L; Barr, DB. (2007). Associations between
       prenatal exposure to polychlorinated biphenyls and neonatal thyroid-stimulating hormone
       levels in a Mexican-American population, Salinas Valley,  California. Environ Health
       Perspect 115: 1490-1496.
Clewell, HJ: Gentry, PR: Covington, TR; Sarangapani, R; Teeguarden,  JG. (2004). Evaluation of
       the potential impact of age- and gender-specific pharmacokinetic differences on tissue
       dosimetry. Toxicol Sci 79: 381-393. http://dx.doi.org/10.1093/toxsci/kfhl09.
Crofton, KM; Craft ES: Hedge, JM; Gennings, C: Simmons, JE: Carchman, RA; Carter, WH, Jr;
       DeVito, MJ. (2005). Thyroid-hormone-disrupting chemicals: Evidence for dose-
       dependent additivity or synergism. Environ Health Perspect 113: 1549-1554.
Dimitropoulos, A; Molinari, L: Etter, K; Lang-Muritano, M; Jenni, OG: Largo, RH; Latal, B.
       (2009). Children with congenital hypothyroidism: Long-term intellectual outcome after
       early high-dose treatment. Pediatr Res 65: 242-248.
Emond, C: Birnbaum, LS: DeVito, MJ. (2004). Physiologically based pharmacokinetic model for
       developmental exposures to TCDD in the rat. Toxicol Sci  80: 115-133.
       http://dx.doi.org/10.1093/toxsci/kfhll7.
Emond, C: Michalek, JE: Birnbaum, LS: DeVito, MJ.  (2005). Comparison of the use of a
       physiologically based pharmacokinetic model and a classical pharmacokinetic model for
       dioxin exposure assessments. Environ Health Perspect 113: 1666-1668.
Emond, C: Birnbaum, LS: DeVito, MJ. (2006). Use  of a physiologically based pharmacokinetic
       model for rats to study the influence of body fat mass and induction of CYP1A2 on the
       pharmacokinetics of TCDD. Environ Health  Perspect 114: 1394-1400.
Evans, MV: Andersen, ME. (2000). Sensitivity analysis of a physiological model for 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD): assessing the impact of specific model parameters
       on sequestration in liver and fat in the rat. Toxicol Sci 54:  71-80.
Fattore, E; Trossvik, C: Hakansson, H. (2000). Relative potency values derived from hepatic
       vitamin A reduction in male and female Sprague-Dawley rats following subchronic
       dietary exposure to individual polychlorinated dibenzo-p-dioxin and dibenzofuran
       congeners and a mixture therof. Toxicol Appl Pharmacol 165: 184-194.
       http://dx.doi.org/10.1006/taap.2000.8943.
Franc. MA: Pohjanvirta, R: Tuomisto, J: Okev, AB.  (2001). Persistent,  low-dose 2,3,7,8-
       tetrachlorodibenzo-p-dioxin exposure: effect on aryl hydrocarbon receptor expression in
       a dioxin-resistance model. Toxicol Appl Pharmacol 175: 43-53.
       http://dx.doi.org/10.1006/taap.2001.9222.
Garabrant, DH; Franzblau, A; Lepkowski, J: Gillespie, BW: Adriaens, P; Demond, A;
       Hedgeman, E: Knutson, K; Zwica, L: Olson, K; Towey, T; Chen, 0: Hong, B: Chang,
       CW: Lee. SY: Ward. B: Ladronka. K: Luksemburg. W: Maier. M. (2009). The University
       of Michigan Dioxin Exposure Study: Predictors of human serum dioxin concentrations in
       Midland and Saginaw, Michigan. Environ Health Perspect 117:  818-824.
       http://dx.doi.org/10.1289/ehp.11779.

                                        A-39

-------
Goodman, M; Squibb, K; Youngstrom, E; Anthony, LG: Kenworthy, L; Lipkin, PH; Mattison,
       PR; Lakind, JS. (2010). Using systematic reviews and meta-analyses to support
       regulatory decision making for neurotoxicants: Lessons learned from a case study of
       PCBs. Environ Health Perspect 118: 727-734. http://dx.doi.org/10.1289/ehp.0901835.
Gray. EL. Jr: Ostbv. J: Wolf. C: Miller. DB: Kelce. WR: Gordon. CJ: Birnbaum. L. (1995).
       Functional developmental toxicity of low doses of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
       and a dioxin-like pcb (169) in Long Evans rats and Syrian hamsters: reproductive,
       behavioral and thermoregulatory alterations. Organohalogen Compounds 25: 33-38.
Hays, SM; Aylward, LL. (2003). Dioxin risks in perspective: past, present, and future. Regul
       Toxicol Pharmacol 37: 202-217. http://dx.doi.org/10.1016/S0273-2300(02)00044-2.
Hochstein, MS, Jr: Render, JA; Bursian, SJ: Aulerich,  RJ. (2001). Chronic toxicity of dietary
       2,3,7,8-tetrachlorodibenzo-p-dioxin to mink. Vet Hum Toxicol 43: 134-139.
Hoio. R:  Stern. S: Zareba. G: Markowskl VP: Cox. C: Kost JT: Weiss. B. (2002). Sexually
       dimorphic behavioral responses to prenatal dioxin exposure. Environ Health Perspect
       110:247-254.
Hutt KJ: Shi. Z: Albertinl DF: Petroff. BK. (2008). The environmental toxicant 2,3,7,8-
       tetrachlorodibenzo-p-dioxin disrupts morphogenesis of the rat pre-implantation embryo.
       BMC Dev Biol 8: 1-12. http://dx.doi.org/10.1186/1471-213X-8-l.
lida, T; Hirakawa, H; Matsueda,  T; Magayama, J: Nagata, T. (1999). Polychlorinated dibenzo-p-
       dioxins and related compounds: Correlations of levels in human tissues and in blood.
       Chemosphere 38: 2767-2774.
Kattainen, H;  Tuukkanen, J: Simanainen, U: Tuomisto, JT: Kovero, O: Lukinmaa, PL;
       Alaluusua. S: Tuomisto. J: Viluksela. M. (2001).  In Utero/Lactational 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin Exposure Impairs Molar Tooth Development in Rats.
       Toxicol Appl Pharmacol  174: 216-224. http://dx.doi.org/10.1006/taap.2001.9216.
Keller, JM; Huet-Hudson, YM; Leamy, LJ. (2007). Qualitative effects of dioxin on molars vary
       among inbred mouse strains. Arch Oral Biol 52: 450-454.
       http://dx.doi.0rg/10.1016/i.archoralbio.2006.10.017.
Keller, JM: Huet-Hudson, Y; Leamy, LJ. (2008a). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       on molar development among non-resistant inbred strains of mice: A geometric
       morphometric analysis. Growth Development and Aging 71: 3-16.
Keller, JM: Zelditch, ML; Huet YM: Leamy, LJ. (2008b). Genetic differences in sensitivity to
       alterations of mandible structure caused by the teratogen 2,3,7,8-tetrachlorodibenzo-p-
       dioxin. Toxicol Pathol 36: 1006-1013. http://dx.doi.org/10.1177/0192623308327409.
Kerger. BD: Leung. HW: Scott. P: Paustenbach. DJ: Needham. LL: Patterson. DG. Jr: Gerthoux.
       PM; Mocarelli, P. (2006). Age- and concentration-dependent elimination half-life of
       2,3,7,8-tetrachlorodibenzo-p-dioxin in Seveso children.  Environ Health Perspect 114:
       1596-1602. http://dx.doi.org/10.1289/ehp.8884.
Kim, AH; Kohn, MC: Nyska, A; Walker, NJ. (2003). Area under the curve as a dose metric for
       promotional responses following 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure. Toxicol
       Appl Pharmacol 191: 12-21. http://dx.doi.org/10.1016/S0041-008X(03)00225-4.
Kimbrough. RD: Krouskas.  CA:  Carson. ML: Long. TF: Sevan. C: Tardiff RG. (2009). Human
       uptake of persistent chemicals from contaminated soil: PCDD/Fs and PCBs. Regul
       Toxicol Pharmacol 57: 43-54.
                                         A-40

-------
LaKind, JS, Hays, S. M., Aylward, L. L., Naitnan, D. 0. (2009). Perspective on serum dioxin
       levels in the United States: an evaluation of the NHANES data. J Expo Sci Environ
       Epidemiol 19:435-441.
Lavado-Autric, R: Auso, E; Garcia-Velasco, JV: Del Carmen Arufe, M; Escobar Del Rev, F;
       Berbel, P; Morreale De Escobar, G. (2003). Early maternal hypothyroxinemia alters
       histogenesis and cerebral cortex cytoarchitecture of the progeny. J Clin Invest 111: 1073-
       1082.  http://dx.doi.org/10.1172/JCI16262.
Leung, HW: Kerger, BD; Paustenbach, DJ. (2006). Elimination half-lives of selected
       polychlorinated dibenzodioxins and dibenzofurans in breast-fed human infants. J Toxicol
       Environ Health A 69: 437-443. http://dx.doi.org/10.1080/15287390500246886.
Li. X: Johnson. DC: Rozman. KK. (1997). 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
       increases release of luteinizing hormone and follicle-stimulating hormone from the
       pituitary of immature female rats in vivo and in vitro.  Toxicol Appl Pharmacol 142: 264-
       269. http://dx.doi.org/10.1006/taap.1996.8044.
Mably, TA; Bjerke, PL; Moore, RW: Gendron-Fitzpatrick, A; Peterson, RE. (1992). In utero and
       lactational exposure of male rats to 2,3,7,8-tetrachlorodibenzo-p-dioxin: 3. Effects on
       spermatogenesis and reproductive capability. Toxicol  Appl Pharmacol 114:  118-126.
       http://dx.doi.org/10.1016/0041-008X(92)90103-Y.
Manz. A: Bergen J: Dwyer. JH: Flesch-Janys. D: Nagel S: Waltsgott H. (1991).  Cancer
       mortality among workers in chemical plant contaminated with dioxin. Lancet 338: 959-
       964. http://dx.doi.org/10.1016/0140-6736(91)91835-I.
Markowski, VP:  Zareba, G: Stern, S: Cox, C: Weiss, B. (2001). Altered operant responding for
       motor reinforcement and the determination of benchmark doses following perinatal
       exposure to low-level 2,3,7,8-tetrachlorodibenzo-p-dioxin. Environ Health Perspect 109:
       621-627.
Maruyama, W; Yoshida, K; Tanaka,  T; Nakanishi, J. (2002). Determination of tissue-blood
       partition coefficients for a physiological model for humans, and estimation of dioxin
       concentration in tissues. Chemosphere 46:  975-985.
Miettinen, HM; Sorvari, R; Alaluusua, S: Murtomaa, M: Tuukkanen, J: Viluksela, M. (2006).
       The Effect of Perinatal TCDD exposure on caries susceptibility in rats. Toxicol Sci 91:
       568-575. http://dx.doi.org/10.1093/toxsci/kfil58.
Milbrath. MO: Wenger. Y: Chang. CW: Emond. C: Garabrant D: Gillespie. BW: Jolliet O.
       (2009). Apparent half-lives of dioxins, furans, and polychlorinated biphenyls as a
       function of age, body fat, smoking status, and breast-feeding. Environ Health Perspect
       117: 417-425. http://dx.doi.org/10.1289/ehp.11781.
Mocarelli, P: Needham, LL; Marocchi, A; Patterson, DG, Jr; Brambilla, P: Gerthoux, PM;
       Meazza, L: Carreri, V. (1991). Serum concentrations of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin and test results from selected residents of Seveso, Italy. J Toxicol Environ Health
       A 32:  357-366. http://dx.doi.org/10.1080/15287399109531490.
Mocarelli, P:  Gerthoux, PM: Ferrari,  E; Patterson Jr, DG: Kieszak, SM; Brambilla, P: Vincoli,
       N; Signorini, S: Tramacere, P: Carreri, V: Sampson, EJ: Turner, WE. (2000). Paternal
       concentrations  of dioxin and sex ratio of offspring. Lancet 355: 1858-1863.
       http://dx.doi.org/10.1016/S0140-6736(00)02290-X.
                                         A-41

-------
Mocarelli, P; Gerthoux, PM: Patterson, DG, Jr: Milani, S: Limonata, G: Bertona, M: Signorini,
       S; Tramacere, P; Colombo, L: Crespi, C: Brambilla, P; Sarto, C; Carreri, V: Sampson,
       EJ: Turner, WE; Needham, LL. (2008). Dioxin exposure, from infancy through puberty,
       produces endocrine disruption and affects human semen quality. Environ Health Perspect
       116: 70-77. http://dx.doi.org/10.1289/ehp.10399.
Mocarelli, P: Gerthoux, PM: Needham, LL: Patterson, DG, Jr: Limonta, G: Falbo, R: Signorini,
       S: Bertona, M: Crespi, C: Sarto, C: Scott, PK: Turner, WE: Brambilla, P. (2011).
       Perinatal exposure to low doses of dioxin can permanently impair human semen quality.
       Environ Health Perspect 119: 713-718.
NAS (National Academy of Sciences). (2006). Health risks from dioxin and related compounds:
       Evaluation of the EPA reassessment. Washington, DC: National Academy Press.
       http://www.nap.edu/catalog.php7record id=l 1688.
Needham, LL: Gerthoux, PM: Patterson, DG: Brambilla, P: Turner, WE: Beretta, C: Pirkle, JL:
       Colombo, L: Sampson, EJ: Tramacere, PL; Signorini, S: Meazza, L; Carreri, V: Jackson,
       RJ: Mocarelli, P. (1997). Serum dioxin levels in Seveso, Italy, population in 1976.
       Teratog Carcinog Mutagen 17: 225-240. http://dx.doi.org/10.1002/(SICI) 1520-
       6866(1997)17:4/5<225::AID-TCM5>3.0.CO:2-K.
NTP (National Toxicology Program). (2006). NTP technical report on the toxicology and
       carcinogenesis studies of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (CAS no.  1746-
       01-6) in female harlan Sprague-Dawley rats (gavage studies). (NTP TR 521; NIH
       Publication No. 06-4468).  Research Triangle Park, NC.
Oerbeck, B: Sundet, K: Kase, BF: Heyerdahl, S. (2005). Congenital hypothyroidism: No adverse
       effects of high dose thyroxine treatment on adult memory, attention, and behaviour. Arch
       Dis Child 90: 132-137. http://dx.doi.org/10.1136/adc.2003.043935.
Oerbeck, B, Reinvang, I,  Sundet, K., Heyerdahl, S. (2007). Young adults with  severe  congenital
       hypothroidism: Cognitive event related potentials (ERPs) and the significance of an early
       start of thyroxine treatment. Scand J Psychol 48: 61-67.
Oerbeck, B, Sundet, K., Kase, B. F., Heyerdahl, S. (2003). Congenital hypothyroidism: Influence
       of disease severity and L-thyroxine treatment on intellectual, motor, and school-
       associated outcomes in young adults [Review]. J Pediatr 112: 923-930.
Ohsako, S: Miyabara, Y; Nishimura, N; Kurosawa, S: Sakaue, M: Ishimura, R: Sato, M: Takeda,
       K: Aoki, Y; Sone,  H: Tohyama, C: Yonemoto, J. (2001). Maternal exposure to a low
       dose of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) suppressed the development of
       reproductive organs of male rats:  Dose-dependent increase of mRNA levels of 5a-
       reductase type 2 in contrast to descrease of androgen receptor in the pubertal ventral
       prostate. Toxicol Sci 60: 132-143.
Ott, MG: Olson, RA: Cook, RR: Bond, GG. (1987). Cohort mortality study of chemical workers
       with potential exposure to  the higher chlorinated dioxins. J Occup Environ Med 29: 422-
       429.
Patterson, DG: Needham,  LL: Pirkle, JL: Roberts, DW: Bagby, J: Garrett, WA: Andrews, JS:
       Falk, H: Bernert, JT: Sampson, EJ: Houk, VN. (1988).  Correlation between serum and
       adipose tissue levels of 2,3,7,8-tetrachlorodibenzo-p-dioxin in 50 persons from Missouri.
       Arch Environ Contam Toxicol 17: 139-143. http://dx.doi.org/10.1007/BF01056017.
                                         A-42

-------
Patterson, DG, Jr; Fiirst P; Henderson, LO: Isaacs, SG: Alexander, LR; Turner, WE; Needham,
       LL; Harmon, H. (1989). Partitioning of in vivo bound PCDDs/PCDFs among various
       compartments in whole blood. Chemosphere 19: 135-142.
       http://dx.doi.org/10.1016/0045-6535(89)90301-9.
Royland, JE; Parker, JS: Gilbert, ME. (2008). A genomic analysis of subclinical hypothyroidism
       in hippocampus and neocortex of the developing rat brain. J Neuroendocrinol 20: 1319-
       1338. http://dx.doi.0rg/10.llll/i.1365-2826.2008.01793.x.
Sand, S: Fletcher, N; von Rosen, D; Kalantari, F; Viluksela, M; Tuomisto, JT; Tuomisto, J: Falk-
       Filipsson, A; Hakansson, H. (2010). Quantitative and statistical analysis of differences in
       sensitivity between Long-Evans and Han/Wistar rats following long-term exposure to
       2,3,7,8-tetrachlorodibenzo-p-dioxin. Regul Toxicol Pharmacol 57: 136-145.
Santostefano, MJ: Wang. X: Richardson. VM: Ross. DG: DeVito, MJ: Birnbaum, LF. (1998). A
       pharmacodynamic analysis of TCDD-Induced Cytochrome 450 gene expression in
       multiple tissues: Dose and time-dependent effects. Toxicol Appl Pharmacol  151: 294-
       310.
Schecter, A: Mes, J: Davies, D. (1989). Polychlorinated biphenyl (PCB), DDT, DDE and
       hexachlorobenzene (HCB) and PCDD/F isomer levels in various organs in autopsy tissue
       from North American patients. Chemosphere 18: 811-818.
       http://dx.doi.org/10.1016/0045-6535(89)90201-4.
Schecter, AJ: Ryan. JJ. (1989). Blood and adipose tissue levels of PCDDs/PCDFs over three
       years in a patient after exposure to polychlorinated dioxins and dibenzofurans.
       Chemosphere 18: 635-642.
Seo, BW: Li. MH: Hansen, LG: Moore. RW: Peterson. RE: Schantz, SL. (1995). Effects of
       gestational and lactational exposure to coplanar polychlorinated biphenyl (PCB)
       congeners or 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on thyroid hormone
       concentrations in weanling rats. Toxicol Lett 78: 253-262.
Sewall CH: Flagler. N: Vanden Heuvel JP: Clark. GC: Tritscher. AM: Maronpot. RM: Lucier.
       GW. (1995). Alterations in thyroid function in female Sprague-Dawley rats following
       chronic treatment with  2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Appl Pharmacol
       132: 237-244.
Sharlin, DS: Tighe, D: Gilbert, ME; Zoeller, RT. (2008). The balance between oligodendrocyte
       and astrocyte production in major white matter tracts is linearly related to serum total
       thyroxine. Endocrinology 149: 2527-2536. http://dx.doi.org/10.1210/en.2007-1431.
Sharlin. DS: Gilbert. ME: Taylor. MA: Ferguson. DC: Zoeller. RT. (2010). The nature of the
       compensatory response to low thyroid hormone in the developing brain. J
       Neuroendocrinol 22: 153-165. http://dx.doi.Org/10.llll/i.1365-2826.2009.01947.x.
Shi. Z: Valdez, KE: Ting. AY: Franczak, A: Gum. SL: Petroff. BK. (2007). Ovarian endocrine
       disruption underlies premature reproductive senescence following environmentally
       relevant chronic exposure to the aryl hydrocarbon receptor agonist 2,3,7,8-
       tetrachlorodibenzo-p-dioxin. Biol Reprod 76: 198-202.
       http://dx.doi.org/10.1095/biolreprod.106.053991.
Skakkebaek, NE. (2010). reference ranges for semen quality and their relations to fecundity.
       Asian J Androl 12: 95-98.
                                         A-43

-------
U.S. EPA (U.S. Environmental Protection Agency). (1994). Methods for derivation of inhalation
       reference concentrations and application of inhalation dosimetry. (EPA/600/8-90/066F).
       Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research
       and Development, Office of Health and Environmental Assessment, Environmental
       Criteria and Assessment Office.
       http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=71993.
U.S. EPA (U.S. Environmental Protection Agency). (2000). Science Policy Council handbook:
       Peer review. (EPA 100-B-OO-OOl). Washington, DC: U.S. Environmental Protection
       Agency, Science Policy Council.
       http://www.ntis.gov/search/product.aspx? ABBR=PB2005109156.
U.S. EPA (U.S. Environmental Protection Agency). (2002). Guidelines for ensuring and
       maximizing the quality, objectivity, utility, and integrity, of information disseminated by
       the Environmental Protection Agency. (EPA/260/R-02/008). Washington, DC.
       http://www.epa.gov/quality/informationguidelines/documents/EPA InfoQualityGuideline
       s.pdf.
U.S. EPA (U.S. Environmental Protection Agency). (2003). Assessment factors for evaluating
       the quality of scientific and technical information, http://www.epa.gov/spc/assess.htm.
U.S. EPA (U.S. Environmental Protection Agency). (2004). An examination of EPA risk
       assessment principles and practices. (EPA/100/B-04/001). Washington, DC: U.S.
       Environmental Protection Agency, Office of the Science Advisor.
U.S. EPA (U.S. Environmental Protection Agency). (2006a). Approaches for the application of
       physiologically based pharmacokinetic (PBPK) models and supporting data in risk
       assessment (final report). (EPA/600/R-05/043F). Washington, DC: U.S.  Environmental
       Protection  Agency, Office of Research and Development.
U.S. EPA (U.S. Environmental Protection Agency). (2006b). A framework for assessing health
       risks of environmental exposures to children. (EPA/600/R-05/093A). Washington, DC:
       U.S. Environmental Protection Agency, National Center for Environmental Assessment.
U.S. EPA (U.S. Environmental Protection Agency). (2006c). Science policy council handbook:
       Peer review, 3rd edition. (EPA/100/B-06/002). Washington, DC: U.S. Environmental
       Protection  Agency, Science Policy Council. http://www.epa.gov/OSA/spc/2peerrev.htm.
Van Birgelen, AP: Van der Kolk, J: Fase, KM; Bol, I; Poiger, H; Brouwer, A; Van den Berg, M.
       (1995a). Subchronic dose-response study of 2,3,7,8-tetrachlorodibenzo-p-dioxin in
       female  Sprague-Dawley rats. Toxicol Appl Pharmacol 132: 1-13.
       http://dx.doi.org/10.1006/taap.1995.1080.
Van Birgelen, AP: Smit, EA; Kampen, IM; Groeneveld, CN: Fase, KM; Van der Kolk, J: Poiger,
       H: Van den Berg. M: Koeman, JH: Brouwer. A. (1995b). Subchronic effects of 2,3,7,8-
       TCDD  or PCBs on thyroid hormone metabolism: Use in risk assessment. Eur  J
       Pharmacol 293: 77-85. http://dx.doi.org/10.1016/0926-6917(95)90021-7.
Walker. NJ: Portier. CJ: Lax. SF: Crofts. FG: Li. Y: Lucier, GW: Sutter. TR. (1999).
       Characterization of the dose-response of CYP1B1, CYP1A1, and CYP1A2 in the liver of
       female  Sprague-Dawley rats following chronic exposure to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin.  Toxicol Appl Pharmacol 154: 279-286. http://dx.doi.org/10.1006/taap.1998.8595.
Wang. X: Santostefano, MJ: Evans. MV: Richardson. VM: Diliberto, JJ: Birnbaum, LS. (1997).
       Determination of parameters responsible for pharmacokinetic behavior of TCDD in
       female  Sprague-Dawley rats. Toxicol Appl Pharmacol 147: 151-168.
       http://dx.doi.org/10.1006/taap.1997.8242.

                                         A-44

-------
Ye, L; Leung, LK. (2008). Effect of dioxin exposure on aromatase expression in ovariectomized
       rats. Toxicol Appl Pharmacol 229: 102-108. http://dx.doi.Org/10.1016/i.taap.2008.01.003.
Yoshizawa. K: Brix. AE: Sells. DM: Jokinen. MP: Wvde. M: Orzech. DP: Kissling. GE: Walker.
       NJ: Nyska, A. (2009). Reproductive lesions in female Harlan Sprague-Dawley rats
       following two-year oral treatment with dioxin and dioxin-like compounds. Toxicol Pathol
       37: 921-937.
Yoshizawa, K, .; Walker, N, . J.: Nyska, A, .; Kissling, G, . E.: Jokinen, M, . P.: Brix, A, . E.:
       Sells, D, . M.: Wvde, M,  . E. (2010). Thyroid follicular lesions induced by oral treatment
       for 2 years with 2,3,7,8-tetrachlorodibenzo-p-dioxin and dioxin-like compounds in
       female Harlan Sprague-Dawley rats. Toxicol Pathol 38: 1037-1050.
       http://dx.doi.org/10.1177/0192623310382560.
                                         A-45

-------
                                   EPA/600/R-10/038F
                                     www.epa.gov/iris
        APPENDIX B
Dioxin Workshop Report
           January 2012
  National Center for Environmental Assessment
     Office of Research and Development
     U.S. Environmental Protection Agency
            Cincinnati, OH

-------
                               EPA/600/R-09/027
                                   May 2009
Summary of U.S. EPA
   Dioxin Workshop
    February 18-20,2009
         Cincinnati, Ohio
   National Center for Environmental Assessment
      Office of Research and Development
     U.S. Environmental Protection Agency
          Cincinnati, OH 45268

-------
                                     DISCLAIMER
       This document summarizes the discussions presented at the Dioxin Workshop in
February 2009, in Cincinnati, OH, as documented by the Session Co-Chairs. This document is
not all inclusive or binding. Conclusions and recommendations to the U.S. EPA may not
represent full consensus.  The views expressed in this document are those of the Dioxin
Workshop Panelists and do not necessarily reflect the views and policies of the U.S.
Environmental Protection Agency.  Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
Preferred Citation:
U.S. Environmental Protection Agency (U.S. EPA). (2009) Summary of U.S. EPA Dioxin Workshop:
February 18-20, 2009. U.S. Environmental Protection Agency, National Center for Environmental Assessment,
Cincinnati, OH. EPA/600/R-09/027.
                                           B-ii

-------
                        TABLE OF CONTENTS
DIOXIN WORKSHOP TEAM	B-iv
ACKNOWLEDGMENTS	B-iv

INTRODUCTION	B-l

     REFERENCES	B-2

SCIENTIFIC WORKSHOP TO INFORM THE TECHNICAL WORK PLAN FOR U.S.
   EPA'S RESPONSE TO NAS COMMENTS ON THE HEALTH EFFECTS OF
   DIOXIN PRESENTED IN U.S. EPA'S DIOXIN REASSESSMENT	B-3

     SESSION 1: QUANTITATIVE DOSE-RESPONSE MODELING ISSUES	B-3
     SESSION 2: IMMUNOTOXICITY	B-6
     SESSION 3 A: DOSE-RESPONSE FOR NEUROTOXICITY AND
         NONREPRODUCTIVE ENDOCRINE EFFECTS	B-8
     SESSION 3B: DOSE-RESPONSE FOR CARDIOVASCULAR TOXICITY
         AND HEPATOTOXICITY	B-ll
     SESSION 4A: DOSE-RESPONSE FOR CANCER	B-13
     SESSION 4B: DOSE-RESPONSE FOR
         REPRODUCTIVE/DEVELOPMENTAL TOXICITY	B-16
     SESSION 5: QUANTITATIVE UNCERTAINTY ANALYSIS OF DOSE-
         RESPONSE	B-20

APPENDIX A: 2009 U.S. EPA DIOXIN WORKSHOP AGENDA	B-24

APPENDIX B: 2009 U.S. EPA DIOXIN WORKSHOP QUESTIONS TO GUIDE
   PANEL DISCUS SIGNS	B-31

APPENDIX C: 2009 U.S. EPA DIOXIN WORKSHOP DRAFT SELECTION
   CRITERIA TO IDENTIFY KEY IN VIVO MAMMALIAN STUDIES THAT
   INFORM DOSE-RESPONSE MODELING FOR
   2,3,7,8-TETRACHLORODIBENZO-p-DIOXIN(TCDD)	B-34
                                B-iii

-------
                            DIOXIN WORKSHOP TEAM
The Dioxin Workshop Team, under the leadership of Peter W. Preuss, Director, NCEA,
comprised the following members:

National Center for Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency, Cincinnati, OH 45268
   Belinda S. Hawkins
   Janet Hess-Wilson
   Glenn Rice
   Jeff Swartout
   Linda K. Teuschler
   Bette Zwayer

Argonne National Laboratory, Argonne, IL 60439
   Maryka H. Bhattacharyya
   Andrew Davidson
   Mary E. Finster
   Margaret M. MacDonell
   David P. Peterson
ACKNOWLEDGMENTS

The Track Group, Alexandria, VA 22312
   Kara Hennigan
   Alan Minton
   Brandy Quinn

ECFlex, Inc., Fairborn, OH 45324
   Dan Heing
   Heidi Glick
   Amy Prues
   Lana Wood

IntelliTech Systems, Inc., Fairborn, OH 45324
   Cris Broyles
   Luella Kessler
   Stacey Lewis
   Linda Tackett
                                        B-iv

-------
                                  INTRODUCTION
       This document provides a summary of the Scientific Workshop to Inform EPA's
Response to National Academy of Science Comments on the Health Effects of Dioxin in EPA's
2003 Dioxin Reassessment.  The U.S. Environmental Protection Agency (U.S. EPA) and
Argonne National Laboratories (ANL), through an inter-Agency agreement with the U.S.
Department of Energy, convened this scientific workshop ("Dioxin Workshop") on February
18-20, 2009, in Cincinnati, Ohio.  The goals of the Dioxin Workshop were to identify and
address issues related to the dose-response assessment of 2,3,7,8-tetrachlorodibenzo-/>-dioxin
(TCDD). This report summarizes the discussions and conclusions from this workshop.
Previously, at the request of the U.S. EPA, the National Academy of Sciences (NAS) prepared a
report, Health Risks from Dioxin and Related Compounds: Evaluation of the EPA Reassessment
(NAS, 2006), which made a number of recommendations to improve the U.S. EPA's risk
assessment for TCDD (U.S. EPA, 2003).  The 3-day Dioxin Workshop was convened
specifically to ensure that the U.S. EPA's response to the NAS recommendations focuses  on the
key issues and reflects the most meaningful science.

   The Dioxin Workshop included seven scientific sessions:
   (1) Session 1:    Quantitative Dose-Response Modeling Issues
   (2) Session 2:    Immunotoxicity
   (3) Session 3 A:  Dose-Response for Neurotoxicity and Nonreproductive Endocrine Effects
   (4) Session 3B:  Dose-Response for Cardiovascular Toxicity and Hepatotoxicity
   (5) Session 4A:  Dose-Response for Cancer
   (6) Session 4B:  Dose-Response for Reproductive/Developmental Toxicity
   (7) SessionS:    Quantitative Uncertainty Analysis of Dose-Response

During each  session, the U.S. EPA asked a panel of expert scientists to:

   •   identify and discuss the technical challenges involved in addressing the key NAS
       comments on the TCDD dose-response assessment in the U.S. EPA Reassessment
       (U.S. EPA, 2003);

   •   discuss approaches for  addressing the key NAS comments; and

   •   identify important published, independently peer-reviewed literature, particularly studies
       describing epidemiologic and in vivo mammalian bioassays, which are expected to be
       most useful for informing the U.S. EPA's response.

       The sessions were followed by open comment periods during which members of the
audience were invited to address the Panels.  At the conclusion of the open comment periods, the
Panel Co-Chairs were asked to summarize and present the results of the panel discussions. The
summaries could include minority opinions stated by panelists.  The main points derived from
the session summaries were used to prepare this document. Additionally, this document includes
a list of the session panelists and their affiliations and three appendices. Appendix A presents
the Dioxin Workshop Agenda. Appendix B identifies the charge questions presented to the
Panel.  Appendix C describes draft study selection criteria proposed by the Dioxin Workshop
Team for consideration by the  workshop panelists.


                                          B-l

-------
REFERENCES
NAS (National Academy of Sciences).  2006. Health Risks from Dioxin and Related
Compounds: Evaluation of the EPA Reassessment. National Academies Press, Washington, DC
(July).  Available at http://www.nap.edu/catalog.php?record_id=l 1688.

U.S. EPA (U.S. Environmental Protection Agency).  2003. Exposure and Human Health
Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. NAS
review draft, Volumes 1-3 (EPA/600/P-00/00ICb, Volume 1). U.S. Environmental Protection
Agency, National Center for Environmental Assessment, Washington, DC (December).
Available at http://www.epa.gov/nceawwwl/pdfs/dioxin/nas-review/.
                                        B-2

-------
  SCIENTIFIC WORKSHOP TO INFORM THE TECHNICAL WORK PLAN FOR U.S.
  EPA'S RESPONSE TO NAS COMMENTS ON THE HEALTH EFFECTS OF DIOXIN
                PRESENTED IN U.S. EPA'S DIOXIN REASSESSMENT

Dioxin Workshop Co-Chairs: Peter W. Preuss and Glenn Rice

       The Dioxin Workshop session summaries were prepared by the session panel Co-Chairs
with input from the panelists, as requested by the U.S. EPA prior to the workshop.  The Co-
Chairs subsequently presented these summaries to all of the workshop participants during
designated periods at the workshop. In these summaries, the U.S. EPA asked that the Co-Chairs
summarize the key issues from the panel discussions. Because the sessions were not designed to
achieve consensus among the panelists, the summaries do not necessarily represent consensus
opinions; rather, they reflect the essence of the panel discussions. Some of the specific points
may represent the views of multiple panelists, while others only the views of a single panelist.
Prior to the summarizations, there were opportunities for public comments on the discussion
topics. Some Co-Chairs met with their sessions' panelists after their sessions ended to develop
these summaries, while others developed reports based on their personal notes. Because Session
5 was the last session of the workshop—with little time provided to develop the summary—the
Co-Chairs circulated a draft for comment by the Session 5 panelists after the workshop, prior to
finalizing the session summary. The U.S. EPA collected the session summaries and then
prepared this document. A draft of this document was distributed to all of the session Co-Chairs
to provide them with a final opportunity to comment and make revisions.  Finally, it should be
noted that U.S. EPA was not prescriptive to the session Co-Chairs with respect to the format of
the presentation materials and provided no specific instructions, resulting in unique formats
among the session summaries.
SESSION 1: QUANTITATIVE DOSE-RESPONSE MODELING ISSUES
       This session discussed the general dose-response modeling issues related to TCDD.
Many of these issues were highlighted by NAS (2006).  There was a general introductory
presentation on TCDD kinetics, including information and uncertainties pertaining to the
conversion of administered doses in animals to human body burden (BB) and additivity to
background issues. This presentation was followed by a Panel discussion on the state of the
science regarding dioxin dose-response modeling issues.

Session 1 Panelists (Session Co-Chairs are identified by  asterisk)
   •   Bruce Allen, Bruce Allen Consulting
   •   Lesa Aylward, Summit Toxicology
   •   Roger Cooke, Resources for the Future
   •   Kenny Crump, Louisiana Tech University
   •   MikeDeVito, U.S. EPA
   •   Dale Hattis, Clark University
   •   Rick Hertzberg, Biomath Consulting
   •   Rob McDowell, U.S. Department of Agriculture
   •   Jim Olson, State University of New York, University at Buffalo
                                         B-3

-------
   •   *Lorenz Rhomb erg, Gradient
   •   Woody Setzer, U.S. EPA
   •    * Jeff Swartout, U. S. EPA

Please note that the use of the term "concluded" or "recommended" in this summary does not mean that a consensus
was reached. Session Summaries were written from the material prepared by the non-EPA/ANL Co-Chair and
represent a synopsis of the panel discussions.
Key Study Selection Criteria
       The Panel discussed the advantages and disadvantages of using key study criteria
(Appendix C). They concluded that a priori criteria foster transparency and consistency, and
could deflect a posteriori criticism. However, the Panel also acknowledged that having a priori
criteria could introduce the potential for excluding useful data.  Although the key study criteria
provided by the U.S. EPA listed studies using TCDD only as a criterion, the Panel posed the
possibility of using closely related dioxin-like compounds (DLCs) as surrogates for TCDD. The
criterion for use of data from mammalian studies only was one criterion that received generalized
support due to the lack of extrapolation protocols for nonmammalian species. The Panel also
discussed the specific exposure-duration criterion and asked if there should be a preference for
longer-term rather than acute studies.  The Panel made three suggestions to modify U.S. EPA's
key study selection criteria:
   (1) Define more relevant exposure-level (i.e., dose) cut points using tissue concentrations.
   (2) Reword statistical criteria to include do-it-yourself analysis.
   (3) Reword the response criteria to clarify "outside of normal range."

Dose Metrics
       The Panel discussed the relative merits of various measures of dose for modeling TCDD
dose response. One general conclusion was that tissue concentration (TC) is the preferred
metric, especially lipid-adjusted TC, because this measure more closely approximates exposures
close to the target tissue when compared to administered doses.  However, the Panel
acknowledged that these data are often unavailable.  They further noted that BB, which is
defined as the concentration of TCDD in the body (ng/kg body weight) (U.S. EPA, 2003), might
be useful as a surrogate for TC provided the two measures were proportional.

       The Panel suggested that a linear approach to BB estimation, which was utilized by
U.S. EPA (2003), is too simplistic because this approach does not take into account toxicokinetic
issues related to TCDD—e.g., sequestration in the liver and fat, age-dependent elimination, and
changing elimination rates over time.  The Panel recommended the use of kinetic/mechanistic
modeling to the extent possible to quantify tissue-based metrics.

       The Panel raised the issue of whether the preferred dose metric would be different for
different endpoints and exposure durations. This led to the Panel's comment that the peak
exposure might be a more important metric than average BB for variable exposure scenarios.
Given this discussion about different exposure durations being relevant to a specific  endpoint,
the Panel suggested that the U.S. EPA also  consider peak measures in dose-response modeling.
                                          B-4

-------
       The last point raised in this part of the discussion centered on the possibility of dose
errors in experimental studies. The Panel highlighted the need for the U.S. EPA to consider dose
error (i.e., uncertainty in the x-axis of the dose-response curve) when using dose surrogates.

Dose-Response Modeling of Mammalian Bioassays
       The Panel considered several issues related to dose-response modeling of mammalian
bioassay data for TCDD: supralinearity and incomplete response data ("anchoring"), defining the
benchmark response (BMR) level with respect to establishing the point of departure (POD), and
the use of threshold modeling—as further explained below.

       The Panel discussed the specific issues of supralinearity and anchoring raised by the
U.S. EPA with respect to modeling noncancer endpoints.  The panel recognized that, for many of
the most sensitive endpoints,  the response at the lowest dose is high (e.g., quantal responses
above 25% and continuous endpoints differ substantially from the mean, often implying 100%
incidence in the treated animals).  This lack of response anchoring at the low end of the dose-
response curve (near the BMR) results in the higher responses determining the shape of the
curve.

       The Panel asked whether new tools might be needed or whether the current tools could be
applied differently. In the context of developing new tools, the Panel emphasized the need for
collaboration between biologists and mathematicians. When discussing application, the Panel
suggested that the problem with supralinearity might be overcome by simply dropping the
requirement for using the lower bound on the Benchmark Dose.  In addition, the Panel posed
several more approaches for further consideration in dose-response modeling by the U.S. EPA:
   (1) Combine similar data sets to fill in data gaps.
   (2) Use mechanistic approaches to model the data gaps.
   (3) Dichotomize continuous data.

Finally, the Panel acknowledged that, in certain  situations, there simply  may not be enough
information to provide meaningful answers.
       The Panel discussed the BMR level for establishing a POD in the context of deriving a
Reference Dose (RfD).  The Panel generally agreed that, while the effective  dose level (ED0i)
used in the 2003 Reassessment may be useful for comparative analysis across endpoints, the
EDoi estimates developed for all endpoints considered in the Reassessment were not appropriate
for deriving an RfD because they were not based on the effect's adversity. The panel noted that
EDoi also is much lower than typical EPA BMR levels. The Panel recommended that the U.S.
EPA work to define endpoint-specific BMRs based on the consideration of adversity. Given that
the same uncertainty factor framework is applied to all PODs, the Panel emphasized the need for
consistency in BMRs; numerical consistency is needed for quantal BMRs and consistency in the
choice of biological relevance should be applied for continuous BMRs.
       The Panel generally discouraged threshold modeling by stating that thresholds are very
difficult to pin down and suggested that the lower bound may always be zero.
                                          B-5

-------
Dose-Response Modeling of Epidemiological Studies
       The Panel noted that many studies have been published with measured concentrations of
TCDD that could be used for dose reconstruction. In this discussion, the Panel acknowledged
that use of these data would entail dealing with toxicity equivalence (TEQ) issues and
pharmacokinetic (PK) modeling. Pertaining to the use of these data for quantitative risk
assessment by the U.S. EPA, the Panel posed the question, "At what point does indirect or
confounded human data supersede controlled animal bioassay data?", or alternatively, "How
much human data uncertainty can we tolerate?" The Panel suggested, at the least, that the
epidemiologic data could be used to "ground-truth" the animal bioassay modeling results.

Supporting Information
       The Panel acknowledged that Ah receptor (AhR) binding affinities are not necessarily
tied to endpoint sensitivity, but they reiterated the need to consider mechanistic modeling to aid
in developing appropriate dose metrics or filling in data gaps in the existing dose-response data.

References

NAS (National Academy of Sciences). 2006.  Health Risks from Dioxin and Related
Compounds: Evaluation of the EPA Reassessment. National Academies Press, Washington, DC
(July). Available at http://www.nap.edu/catalog.php?record_id=l 1688.

U.S. EPA (U.S. Environmental Protection Agency).  2003. Exposure and Human Health
Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. NAS
Review Draft (EPA/600/P-00/001Cb). U.S. Environmental Protection Agency, National Center
for Environmental Assessment, Washington, DC. Available at
http://www.epa.gov/nceawwwl/pdfs/dioxin/nas-review/.
SESSION 2: IMMUNOTOXICITY
       The U.S. EPA plans to consider development of a quantitative dose-response assessment
for the immunologic effects associated with TCDD exposure.  Such an assessment would be
based on information in U.S. EPA (2003), NAS (2006) and key studies identified in this
workshop.  The purpose of this session was to identify and discuss key issues pertaining to dose-
response assessment for dioxin-induced immunologic effects.

Session 2 Panelists (Session Co-Chairs are identified by asterisk)
   •   Roger Cooke, Resources for the Future
   •   Rob Goble, Clark University
   •   *Belinda Hawkins, U.S. EPA
   •   Nancy Kerkvliet, Oregon State University
   •   Manolis Kogevinas, Centre for Research in Environmental Epidemiology
   •   Robert Luebke, U. S. EPA
   •   Paolo Mocarelli, University of Milan
   •   * Allen Silverstone, State University of New York, Upstate Medical University

-------
    •   Courtney Sulentic, Wright State University
    •   Nigel Walker, National Institute of Environmental Health Sciences

Please note that the use of the term "concluded" or "recommended" in this summary does not mean that a consensus
was reached. Session Summaries were written from the material prepared by the non-EPA/ANL Co-Chair and
represent a synopsis of the panel discussions.
Key Study Selection Criteria
       The Panel first addressed the Key Study Selection Criteria proposed by the U.S. EPA
(Appendix C). The Panel raised the issue that the key study criteria do not apply to most studies
designed to investigate immunotoxicity, including those used to calculate EDoiS (U.S. EPA,
2003). The Panel observed that most dioxin immunotoxicity studies are relatively high dose
(>200 ng/kg-d) acute studies and/or use parenteral rather than oral administration.

       The Panel discussed several studies often considered important for assessing the
immunotoxic effects of TCDD exposure. The Oughton et al. (1995) mouse bioassay was
discussed and, although the study does meet the proposed criteria, it could not be considered a
key study; specifically, the Panel contended that since there were no functional alterations
observed or measured in this bioassay, the changes in cellular phenotypes are only "suggestive"
of immune alterations and cannot be regarded as having immunopathologic significance.

       The Panel discussed two additional studies for further consideration by the U.S. EPA:

   •  Baccarelli et al. (2002).  The Panel discussed this as a potentially key human
       epidemiological study that should be reviewed and considered further by the U.S. EPA.
       It measured the level of IgG, demonstrating a significant decline relative to dioxin body
       burdens.

   •  Smialowicz et al. (2008). The Panel noted that this study identified the antibody response
       to sheep red blood cells (SRBCs) as the critical effect, labeling this  protocol as a
       functional assay. The Panel stated that if modeled, the U.S. EPA could calculate the
       BMR for this endpoint as 1 standard deviation from the control mean.

References

Baccarelli, A., P. Mocarelli, D.G. Patterson et al.  2002.  Immunologic effects of dioxin: New
results from Seveso and comparison with other studies.  Environ. Health Perspect.
NAS (National Academy of Sciences). 2006.  Health Risks from Dioxin and Related
Compounds: Evaluation of the EPA Reassessment.  National Academies Press, Washington, DC
(July). Available at http://www.nap. edu/catalog.php?record_id=l 1688.

Oughton, J.A., C.B. Pereira, O.K. Dekrey, J.M. Collier, A.A. Frank and N.I. Kerkvliet. 1995.
Phenotypic analysis of spleen, thymus, and peripheral blood cells in aged C57BI/6 mice
following long-term exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol. Sci. 25(l):60-69.
                                           B-7

-------
Smialowicz, R.J., MJ. DeVito, W.C. Williams and L.S. Birnbaum. 2008. Relative potency
based on hepatic enzyme induction predicts immunosuppressive effects of a mixture of
PCDDS/PCDFSandPCBS. Toxicol. Appl. Pharmacol. 227(3):477-484.

U.S. EPA (U.S. Environmental Protection Agency).  2003.  Exposure and Human Health
Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds.  NAS
Review Draft (EPA/600/P-00/001Cb).  U.S. Environmental Protection Agency, National Center
for Environmental Assessment, Washington, DC.  Available at
http://www.epa.gov/nceawwwl/pdfs/dioxin/nas-review/.
SESSION 3A: DOSE-RESPONSE FOR NEUROTOXICITY AND NONREPRODUCTIVE
ENDOCRINE EFFECTS

       The U.S. EPA plans to consider development of a quantitative dose-response assessment
for neurological and/or nonreproductive endocrine effects associated with TCDD exposure.
Such an assessment would be based on information in U.S. EPA (2003), NAS (2006) and key
studies identified in this workshop.  The purpose of this session was to identify and discuss key
issues pertaining to dose-response assessment for dioxin-induced neurological and/or
nonreproductive endocrine effects.

Session 3A Panelists (Session Co-Chairs are identified by asterisk)

   •  *Maryka Bhattacharyya, Argonne National Laboratory
   •  Mike DeVito, U.S. EPA
   •  Mary Gilbert, U.S. EPA
   •  Rob Goble, Clark University
   •  Nancy Kerkvliet, Oregon State University
   •  Fumio Matsumura, University of California-Davis
   •  Paolo Mocarelli, University of Milan
   •  Chris Portier, National Institute of Environmental Health Sciences
   •  Lorenz Rhomb erg, Gradient
   •  Allen Silverstone, State University of New York, Upstate Medical University
   •  Marie Sweeney, National Institute of Occupational Safety and Health
   •  *Bernie Weiss, University of Rochester

Please note that the use of the term "concluded" or "recommended" in this summary does not mean that a consensus
was reached. Session Summaries were written from the material prepared by the non-EPA/ANL Co-Chair and
represent a synopsis of the panel discussions.
What Are the Key Questions Regarding These Endpoints?
       The Panel used the following question to initiate discussion: "Are there identifiable
indices of neurotoxicity and nonreproductive endocrine effects in animal studies and human
populations? " Under this discussion topic, the Panel discussed three endpoints: neurotoxicity
(with focus on developmental exposures), thyroid dysfunction (e.g., thyroid hormone deficits),
and diabetes. The Panel also addressed the relevance of windows of vulnerability to each

-------
endpoint.  The Panel acknowledged that, in some cases, the window of exposure may precede the
window of expression of toxicity.

Epidemiological Study Selection
Developmental Neurotoxicity
       The Panel recognized that an unusual feature for this endpoint is that there are sufficient
human data for dose-response modeling (e.g., Dutch children [Huisman et al., 1995; Patandin et
al., 1999] and U.S. children [Jacobson and Jacobson, 1996]) and there is an internal dose metric
(serum concentrations). Additionally, the Panel discussed recent studies that address this
endpoint in humans (from Japan [reference not provided] and Holland [e.g., Koopman-Esseboom
et al., 1996; Vreugdenhil et al., 2002]). For continued investigation into this endpoint, the Panel
raised two issues to the U.S. EPA:

    •   Conduct an evaluation of whether a modeled effect can be attributed to TCDD and not
       some other persistent organic pollutant (POP), although the Panel recognized that it is
       unlikely U.S. EPA will be able to distinguish among these exposures because other POPs
       are intrinsic confounders in the Dutch study.

    •   Allow animal data to inform the dose-response modeling of epidemiological data.

Thyroid Dysfunction
       The Panel identified the availability of human data for this endpoint (e.g., Calvert et al.,
1999; Koopman-Esseboom et al., 1994). Much of the thyroid dysfunction literature has been
published since the 2003 Reassessment (e.g., Wang et al., 2005; Baccarelli et al., 2008).  The
Panel also noted the availability of an internal dose metric (serum concentrations).  Additionally,
the Panel discussed the mechanistic studies in animals that link TCDD to thyroid dysfunction.
For continued investigation into this endpoint, the Panel raised three issues for the U.S. EPA to
consider:

    •   Consider the newly available human data since the Reassessment.

    •   Investigate and clarify of the role of TCDD-induced thyroid dysfunction in
       developmental neurotoxicity.

    •   Evaluate and determine whether an effect can be attributed to TCDD or other
       contaminants.

Diabetes
       The Panel discussed that data suggest that diabetes incidence in those under 55 years old
may be associated with exposure to PCBs. They acknowledged that whether this is a dioxin-like
compound (DLC) mediated effect or whether other POPs are responsible is still undetermined.
The Panel also acknowledged that no animal  model exists for the investigation of xenobiotic-
induced diabetes, and that separating the injury dose level from the current body burdens would
depend on good pharmacokinetics in humans. For continued investigation into this endpoint, the
Panel listed two issues for the U.S. EPA to consider:

    •   Results from the Anniston study and the Great Lakes Fishermen study (references not
       provided) should be examined for dose metrics (both studies examine human PCB
       exposures).
                                          B-9

-------
   •   Changes of adipose tissue status need to be considered, given that dieting can cause
       release of lipid-soluble contaminants.

References

Baccarelli, A., S.M. Giacomini, C. Corbetta et al.  2008. Neonatal thyroid function in Seveso 25
years after maternal exposure to dDioxin. PLoS Med.  5(7):el61.
doi:10.1371/journal.pmed.0050161.

Calvert, G.M., M.H. Sweeney, J. Deddens and O.K. Wall.  1999.  Evaluation of diabetes
mellitus, serum glucose, and thyroid function among United States workers exposed to
2,3,7,8-tetrachlorodibenzo-p-dioxin. Occ. Env. Med.  56:270-276.

Huisman, M., C. Koopman-Esseboom, V. Fidler et al.  1995. Perinatal exposure to
polychlorinated biphenyls and dioxins and its effect on neonatal neurological development.
Early Hum. Devel. 41(2):111-127.

Jacobson, J.L. and S.W. Jacobson.  1996.  Intellectual impairment in children exposed to
polychlorinated biphenyls in utero. N. Engl. J. Med. 335:783-789.

Koopman-Esseboom, C., N. Weisglas-Kuperus, M.A.J. deRidder, C.G. Van derPaauw,
L.G.M.Th. Tuinstra and PJJ. Sauer.  1996.  Effects of polychlorinated biphenyl/dioxin exposure
and feeding type on infants' mental and psychomotor development.  J. Pediatr.  97(5):700-706.

Koopman-Esseboom, C., D.-C. Morse, N. Weisglas-Kuperus et al. 1994. Effects of dioxins and
polychlorinated biphenyls on thyroid hormone status of pregnant women and their infants.
Pediatr. Res.  36:468-473.

Patandin, S., C.I. Lanting, P.G.H. Mulder, E.R. Boersma, PJJ. Sauer and N. Weisglas-Kuperus.
1999. Effects of environmental exposure to polychlorinated biphenyls and dioxins on cognitive
abilities in Dutch children at 42 months of age.  J. Pediatr.  134:33-41.

NAS (National Academy of Sciences).  2006.  Health Risks from Dioxin and Related
Compounds: Evaluation of the EPA Reassessment.  National Academies Press, Washington, DC
(July).  Available at http://www.nap.edu/catalog.php?record_id=l 1688.

U.S. EPA (U.S. Environmental Protection Agency). 2003. Exposure and Human Health
Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds.  NAS
Review Draft (EPA/600/P-00/001Cb). U.S. Environmental Protection Agency, National Center
for Environmental Assessment, Washington, DC.  Available at
http://www.epa.gov/nceawwwl/pdfs/dioxin/nas-review/.

Vreugdenhil,  H.J., C.I. Lanting, P.G. Mulder, E.R. Boersma andN. Weisglas-Kuperus. 2002.
Effects of prenatal PCB and dioxin background exposure on cognitive and motor abilities in
Dutch children at school age.  J. Pediatr. 140:48-56.
                                         B-10

-------
Wang S.L., P.H. Su, S.B. Jong, Y.L. Guo, W.L. Chou and O. Papke. 2005. In utero exposure to
dioxins and polychlorinated biphenyls and its relations to thyroid function and growth hormone
in newborns.  Environ. Health Perspect.  113:1645-1650.
SESSION 3B: DOSE-RESPONSE FOR CARDIOVASCULAR TOXICITY AND
HEPATOTOXICITY

       The U.S. EPA plans to consider development of a quantitative dose-response assessment
for cardiovascular and/or hepatic effects associated with TCDD exposure. Such an assessment
would be based on information in U.S. EPA (2003), NAS (2006) and key studies identified in
this workshop. The purpose of this session was to identify and discuss key issues pertaining to
dose-response assessment for dioxin-induced cardiovascular and/or hepatic effects.

Session 3B Panelists (Session Co-Chairs are identified by asterisk)

   •   Bob Budinksy, Dow Chemical
   •   Manolis Kogevinas, Centre for Research in Environmental Epidemiology
   •   Rob McDowell, U.S. Department of Agriculture
   •   Jim Olson, State University of New York, University at Buffalo
   •   Marian Pavuk, Agency for Toxic Substances and Disease Registry
   •   * Jeff Swartout, U. S. EPA
   •   *Mary Walker, University of New Mexico
   •   Nigel Walker, National Institute of Environmental Health Sciences

Please note that the use of the term "concluded" or "recommended" in this summary does not mean that a consensus
was reached.  Session Summaries were written from the material prepared by the non-EPA/ANL Co-chair and
represents a synopsis of the panel discussions.
Key Study Selection Criteria
       The Panel initially focused on the draft key study selection criteria offered by the
U.S. EPA (Appendix C). The panel recommended that for cardiovascular effects, which are not
usually observed in rodents, the use of knockout mouse models (ApoE KO and LDLR KO) be
moved to the "primary" column because only these studies establish the cardiovascular toxicity
model in mice.

       The panel also was concerned that the gavage procedure can increase mouse blood
pressure. Consequently, the panel recommended that gavage studies not be used for the blood
pressure endpoint (i.e., only dietary dosing studies should be considered).

Human Health Endpoints
       In relation to the hepatic endpoint, the Panel acknowledged the large body of dose
response information on hepatic effects in rodents and that enzyme (mostly CYP1A1) induction
was a sensitive effect. However, the Panel cited the lack of linkage of CYP1A1 to downstream
events, which complicates the toxicological interpretation of this endpoint, and concluded that
                                         B-ll

-------
the more important liver effects in rodents are probably on the "road to cancer." The Panel noted
that hepatic effects were not seen in the epidemiological studies, but acknowledged that these
studies were not designed to detect them.

       In relation to the cardiovascular endpoint, the Panel identified hypertension and ischemic
heart disease (THD) as two key endpoints from the epidemiological studies.  The Panel
recommended that the U.S. EPA perform a meta-analysis of these data. The Panel also
commented that recent animal studies support the observations linking TCDD exposure to IHD
and hypertension. In particular, the National Toxicology Program (NTP) study shows
inflammatory and structural effects on resistant vascular arterioles (NTP, 2006). Additional
evidence from the study suggests that the vascular effects may be CYP1A1-dependent. The
Panel suggested that the NTP study data might be used as a surrogate for dose-response
modeling of hypertension and that such an approach would be supported by data on the role of
AhR in vascular function and remodeling.

POD Issues
       The Panel was not supportive of 1% of maximal response (EDoi), which was utilized in
the 2003 Reassessment. The Panel concluded that the POD should depend on the specific
endpoint and recommended the following to the U.S. EPA:

   •   For continuous measures, base the BMR on difference from control. Consider the
       adversity level—at what point does the endpoint become adverse?

   •   For incidence data, set the BMR to a fixed-risk level.

Supporting Information
       The Panel posed several suggestions to the U.S. EPA for reducing uncertainty and
improving the knowledge base for TCDD toxicity.

   •   Use in vitro data to define uncertainties, such as the relative sensitivity between rodents
       and humans and around the definition of a POD.

   •   Consider studies on dioxin-like compounds (DLCs).

   •   Use PK modeling to define the dose metric for hepatic effects.

   •   Use body burden or serum concentrations for cardiovascular endpoints.
Finally, the Panel recommended that U.S. EPA finish the reassessment quickly and establish a
definitive plan to review and incorporate new data as they become available.

References

NAS (National Academy of Sciences).  2006. Health Risks from Dioxin and Related
Compounds: Evaluation of the EPA Reassessment. National Academies Press, Washington, DC
(July). Available at http://www.nap.edu/catalog.php?record_id=l 1688.
                                         B-12

-------
NTP (National Toxicology Program). 2006. Toxicology and Carcinogenesis Studies of
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) (CAS No. 1746-01-6) in Female Harlan Sprague-
Dawley Rats (Gavage Studies).  U.S. Department of Health and Human Services. NTP TR 521.
Research Triangle Park, NC (April).

U.S. EPA (U.S. Environmental Protection Agency). 2003. Exposure and Human Health
Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds.  NAS
Review Draft (EPA/600/P-00/001Cb). U.S. Environmental Protection Agency, National Center
for Environmental Assessment, Washington, DC. Available at
http://www.epa.gov/nceawwwl/pdfs/dioxin/nas-review/.
SESSION 4A: DOSE-RESPONSE FOR CANCER
       The U.S. EPA plans to consider development of a quantitative dose-response assessment
for cancer associated with TCDD exposure.  Such an assessment would be based on information
in U.S. EPA (2003), NAS (2006) and key studies identified in this workshop. The purpose of
this session was to identify and discuss key issues pertaining to dose-response assessment for
dioxin-induced cancer.

Session 4A Panelists (Session Co-Chairs are identified by asterisk)
   •   Lesa Aylward, Summit Toxicology
   •   Kenny Crump, Louisiana Tech University
   •   Dale Hattis, Clark University
   •   * Janet Hess-Wilson, U. S. EPA
   •   Karen Hogan, U.S. EPA
   •   Manolis Kogevinas, Centre for Research in Environmental Epidemiology
   •   Marian Pavuk, Agency for Toxic Substances and Disease Registry
   •   Chris Portier, National Institute of Environmental Health Sciences
   •   Lorenz Rhomb erg, Gradient
   •   Jay Silkworm, General Electric
   •   *Nigel Walker, National Institute of Environmental Health Sciences

Please note that the use of the term "concluded" or "recommended" in this summary does not mean that a consensus
was reached.  Session Summaries were written from the material prepared by the non-EPA/ANL Co-chair and
represent a synopsis of the panel discussions.

Key Study Selection
       The Panel discussed both human and rodent studies.  In reviewing the epidemiological
data, the Panel agreed the EPA should focus on four cohort studies (Dutch cohort, NIOSH
cohort, BASF accident cohort, and Hamburg cohort) and pointed out that there are numerous
updates and reevaluations of data now in the literature and others will be published soon.  The
Panel stated that it is appropriate for the U.S. EPA to consider the increase in total cancers for
modeling human cancer data, however, Non-Hodgkin's lymphoma, and lung tumors are the main
TCDD-related cancer types seen in humans exposed to TCDD.  The Panel suggested the U.S.
EPA focus the quantitative dose-response modeling on the human data.
                                         B-13

-------
       In reviewing the rat data, the Panel identified four new NTP rodent cancer bioassays with
liver and lungs as the main target organs. However, they suggested that dose-response modeling
efforts should model "all cancers" from these NTP data sets as well and use tumor incidence—
not individual rats as measures.

Key Study Selection Criteria
       The Panel discussed whether data for TCDD only should be used or if PCB126 could be
used to develop a dose-response curve. From this discussion, the Panel reached a general
agreement that limiting the dose-response modeling and cancer assessment to TCDD only would
be the best approach.

       Regarding the oral dosing regimens, the Panel discussed the differences in results from
different bioassays.  They concluded that there were insufficient data to pick between oral  feed
(Kociba et al., 1978) and oral gavage (NTP, 2006) studies, but stated "If all aspects of studies
were equal, an oral feed study is preferred." However, given that current data sets are not  equal,
they agreed that U.S. EPA should consider both feed and gavage studies.

       The Panel put forth the recommendation that studies that include initiation-promotion
model  data and TgAC transgenic model data from oral exposure studies should be excluded from
the primary category in the key study selection criteria (Appendix C lists the draft study selection
criteria distributed prior to the meeting).  Studies from both classifications  should be moved to
the second tier.

       The Panel was also unsupportive of the "response magnitude outside the range of normal
variability" criterion, as they did not believe it was applicable to a cancer endpoint.

Critical Endpoints to Consider
       The Panel recognized that the MOA for TCDD includes cell growth/differentiation
dysregulation, that different endpoints (tumor types) across species may be expected, and that
there are differences in tumor sites across species. The Panel further acknowledged that there is
insufficient information to determine if rodent tumor types observed are relevant to humans.
Thus, the Panel suggests the following:

   •   U.S. EPA should consider all the observed cancer endpoints in its evaluation.

Nonlinear (aka threshold) Versus Linear Dose-Response Modeling
       The Panel agreed that NTP bioassays appear to demonstrate nonlinear dose response, but
they expressed concern about using animal data to infer slope and dose response for humans.
The Panel pointed out that there are differences in slopes across different bioassays, and
specifically, that some appear linear while others appear nonlinear. Given  the observation of
both nonlinear vs. linear, the Panel concluded that neither could be ruled out for extrapolation
below the POD simply based on the available data. One panelist noted that U.S. EPA Cancer
Guidelines (U.S. EPA, 2005) state that only if one can demonstrate that the MOA has a threshold
dose-response shape, and can exclude all other potential linear MO As, can one use a nonlinear
model. Lastly, the Panel noted that there are data and rationales to support use of both linear and
                                          B-14

-------
nonlinear response below POD. From this discussion, the Panel raised one possibility to the U.S.
EPA:

   •   Both linear and nonlinear model functions should be considered in the dose-response
       analysis.

Dose Metrics
       In considering human data, the Panel expressed a preference for lipid-adjusted serum
levels over body burden (BB), and they expressed concerns over the assumptions used in the
back calculation of the BB in the epidemiologic cohorts.  In considering the rat data, the Panel
supported the use of BB—especially lipid-adjusted BB. The Panel, however, did express
concern over the sequestering of TCDD in liver and then the use of liver levels in BB
calculations.

Supporting Information—Biologically-Based Dose-Response (BBDR) Models and MOA
       The Panel discussed BBDR.  Though once considered an attractive proposition, BBDR
models may mask uncertainty within the models, necessitating them to be used with greater
caution. The Panel suggested two issues for the U.S. EPA to consider:

   •   If there is a published model, use it if it is valid—do not generate a new model.
   •   Focus on the actual experimental data to drive the analysis.

References

Kociba, R.J., D.G. Keyes, I.E. Beyer et al.  1978.  Results of a two-year chronic toxicity and
oncogenicity study of 2,3,7,8-tetrachlorodibenzo-^-dioxin in rats.  Toxicol. Appl. Pharmacol.
46:279-303.

NAS (National Academy of Sciences).  2006. Health Risks from Dioxin and Related
Compounds: Evaluation of the EPA Reassessment. National Academies Press, Washington, DC
(July). Available at http://www.nap.edu/catalog.php?record_id=l 1688.

NTP (National Toxicology Program). 2006. Toxicology and Carcinogenesis Studies of
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) (CAS No. 1746-01-6) in Female Harlan Sprague-
Dawley Rats (Gavage Studies).  U.S. Department of Health and Human Services. NTP TR 521.
Research Triangle Park, NC (April).

U.S. EPA (U.S. Environmental Protection Agency). 2003. Exposure and Human Health
Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. NAS
Review Draft (EPA/600/P-00/001Cb). U.S. Environmental Protection Agency, National Center
for Environmental Assessment, Washington, DC. Available at
http://www.epa.gov/nceawwwl/pdfs/dioxin/nas-review/.

U.S. EPA (U.S. Environmental Protection Agency). 2005. Guidelines for Carcinogen Risk
Assessment. U.S. Environmental Protection Agency Risk Assessment Forum.
EPA/630/P-03/001F.
                                         B-15

-------
SESSION 4B: DOSE-RESPONSE FOR REPRODUCTIVE/DEVELOPMENTAL
TOXICITY
       The U.S. EPA plans to consider development of a quantitative dose-response assessment
for reproductive and developmental effects associated with TCDD exposure. Such an
assessment would be based on information in U.S. EPA (2003), NAS (2006) and key studies
identified in this workshop.  The purpose of this session was to identify and discuss key issues
pertaining to dose-response assessment for dioxin-induced reproductive and developmental
effects.

Session 4B Panelists (Session Co-Chairs are identified by asterisk)

   •   Barbara Abbott, U. S. EPA
   •   Bruce Allen, Bruce Allen Consulting
   •   Roger Cooke, Resources for the Future
   •   George Daston, Procter & Gamble
   •   MikeDeVito, U.S. EPA
   •   Rob Goble,  Clark University
   •   *Fumio Matsumura, University of California-Davis
   •   Paolo Mocarelli, University of Milan
   •   Brian Petroff, University of Kansas
   •   *GlennRice, U.S. EPA
   •   Marie Sweeney, National Institute of Occupational Safety and Health
   •   Mary Walker, University of New Mexico
   •   Bernie Weiss, University of Rochester

Please note that the use of the term "concluded" or "recommended" in this summary does not mean that a consensus
was reached.  Session Summaries were written from the material prepared by the non-EPA/ANL Co-Chair and
represent a synopsis of the panel discussions.
A Major Question Posed During this Workshop Session was "Are Human Embryos and
Infants Less Sensitive to Dioxin Exposures Than Some Experimental Animals?"
       The Panel recognized that animal data show a wide range of species sensitivity to dioxin
for a given developmental or reproductive endpoint. Presently, there are data for some endpoints
that show that human sensitivity is comparable to experimental animals (e.g., semen quality),
and for other endpoints the data demonstrate that humans are insensitive compared to  other
species (e.g., cleft palate). Lastly, the Panel recognized that there are some endpoints  for which
relative human sensitivity remains uncertain.

Key Study Selection
       The Panel reviewed the charge questions (Appendix B), discussed them, and listed two
issues for the U.S. EPA to consider:

   •   Concerning key study determination, use a stepwise approach that is dependent upon the
       information available and needed to address the question.
                                         B-16

-------
   •   Concerning the key studies informing the POD and the POD endpoint choice, use the
       POD to depart from what is certain and use a high-confidence study that has found
       effects at a low enough level at which other effects are protected.

The Panel also developed Table 1, based on the information presented in this session.  Table 1
identifies specific reproductive  and developmental effects of concern, listing whether an effect
has been observed in test animals and epidemiologic cohorts.  It also identifies the EDio
estimated by the U.S. EPA (2003) for health effects observed in rodent bioassays. If the U.S.
EPA did not report an EDio  for an effect, the table identifies a study where the effect was
reported and the lowest study dose where the effect was observed. Table 1 also identifies the
epidemiologic cohort where the specific reproductive and developmental effects were  observed.

Epidemiological Study Utility
       The Panel reviewed the  charge questions (Appendix B), discussed them, and made two
suggestions to the U.S. EPA:

   •   Concerning the ability of epidemiological studies to inform critical  effects, start with
       concordance across species (including humans) for the spectrum of effects.

   •   Concerning the ability of epidemiological studies to inform dose-response modeling, start
       with the epidemiology and then go to animal  data if the dose response has not been well
       characterized for an endpoint of interest and compare to animal data as a reality check.

Animal Model  Utility
       The Panel reviewed and discussed the charge questions (Appendix B).  Table 1, which
identifies the effects that occur in animals and also have relevance to humans, summarizes much
of this discussion. Regarding the influence of mode of action (MO A) on animal model choice,
the Panel concluded that by  evaluating concordance among health effects reported in
epidemiologic and animal bioassay data, the U.S. EPA could identify a set  of plausible
reproductive and developmental effects to consider.  Actual animal and human MOA
information is helpful in that it creates comfort with the animal models and in defining the
boundaries of possible effects.
                                         B-17

-------
TABLE 1
Reproductive/Developmental Effects of Concern for Human Health
Endpoint
Sperm Count/Motility
Sex Ratio
Delayed Puberty Males
Delayed Puberty in Females
Cleft Palate
Premature Senescence
Hormones E2
Low Birth Weight
Reproductive Cycling
(prolongation)
Rodent
(EDiong/kg-d)
Yes (6.2-28;
66-200)
No
Yes (94)
Yes
Yes (6300-6400)
Yes
Yes
Yes (190)
Yes
Human
Yes
Yes, Seveso
Yu-cheng
No in Seveso
No
No, Seveso
Yes, Males —
Seveso
Suggestive
effect in Seveso
in first 8 years
after exposure
Yes, Seveso
Prepubertal
exposure
Notes
ED10 bases Mabley et al. (1992a,b) caudal
sperm count and daily sperm production
range from 6.2-28; Gray et al. (1997)
epididymal sperm count and total testis sperm
counts range from 66-200.

ED10 basis rat male puberty delay Gray et al.
(1997). Need to qualify epidemiology data
because of cohort PCDD/PCDFs exposures.
Gray and Ostby (2002) report delayed
puberty in female offspring of pregnant rats
receiving a single dose of 1 ug TCDD/kg on
GD 15.
ED10 basis Birnbaum et al. (1989).
Franczak et al. (2006) report that rats
prematurely entered reproductive senescence,
after receiving cumulative TCDD doses as
low as 1.7 ug TCDD/kg. They considered
first occurrence of prolonged interestrous
interval (>6 d) as evidence of onset of
reproductive senescence.
Li et al. (1995) report serum estradiol-17p
(E2) concentrations induced by equine
Chorionic Gonadotropin injection were
significantly elevated in female rats orally
administered 10 ug/kg TCDD onPND 22.
While E2 decreased dramatically in control
animals during the preovulatory LH surge, it
did not in TCDD-treated rats.
ED10 basis Gray et al. (1997).
Franczak et al. (2006) report loss of normal
cyclicity in female rats at 8 months of age
following a cumulative dose of 1.7 ug
TCDD/kg.
B-18

-------
Supporting Information
       The Panel reviewed the charge questions (Appendix B), discussed them, and made two
suggestions to the U.S. EPA:

   •   Concerning deviation from default approaches for noncancer endpoints, there needs to be
       a careful assessment of the POD and the application of uncertainty factors in light of
       PK/pharmacodynamics (PD), population characteristics  and variability, and MOA
       information.

   •   Concerning the MOA's ability to clarify endpoint and the incorporation of a cascade of
       cellular event into dose-response for noncancer endpoint, any study that helps inform the
       dose response should be considered—including studies not specific to dioxins.
       Complicated mechanistic models need not be developed. Standard dose-response models
       can be applied. One can look at the  cascade of events in a stepwise,  simple way.

References

Birnbaum, L.S., M.W. Harris, L.M. Stocking et al.  1989. Retinoic acid and 2,3,7,8-
tetrachlorodibenzo-p-dioxin selectively enhance teratogenesis in C57BL/6N mice.  Toxicol.
Appl. Pharmacol. 98:487-500.

Franczak, A., A. Nynca, K.E. Valdez, K.M. Mizinga and B.K. Petroff.  2006.  Effects of acute
and chronic exposure to the aryl hydrocarbon receptor agonist 2,3,7,8-tetrachlorodibenzo-
p-dioxin on the transition to reproductive senescence in female  Sprague-Dawley rats. Biol.
Reprod. 74:125-130.

Gray, L.E. and J.S. Ostby. 2002.  In utero 2,3,7,8-tetrachlorodibenzo-^-dioxin (TCDD) alters
reproductive morphology and function in female rat offspring.  Toxicol. Appl. Pharmacol.
133(2):285-294.

Gray, L.E., J.S. Ostby and W.R. Kelce.  1997.  A dose-response analysis of the reproductive
effects of a single gestational dose of 2,3,7,8-tetrachlorodibenzo-p-dioxin in male Long Evans
Hooded rat offspring. Toxicol. Appl. Pharmacol.  146:11-20.

Li, X., D.C. Johnson and K.K. Rozman. 1995. Reproductive effects of 2,3,7,8-
tetrachlorodibenzo-p-dioxin (TCDD) in female rats: ovulation, hormonal regulation, and possible
mechanism(s). Toxicol. Appl. Pharmacol.  133:321-327.

Mably, T.A., D.L. Bjerke, R.W. Moore et al. 1992a.  In utero and lactational exposure of male
rats to 2,3,7,8-tetrachlorodibenzo-p-dioxin.  3. Effects on spermatogenesis and reproductive
capability.  Toxicol. Appl. Pharmacol.  114:118-126.

Mably, T.A., R.W.  Moore, R.W. Goy et al.  1992b. In utero and lactational exposure of male
rats to 2,3,7,8-tetrachlorodibenzo-p-dioxin.  2. Effects on sexual behavior and the regulation of
luteinizing hormone secretion in adulthood. Toxicol. Appl. Pharmacol. 114:108-117.
                                          B-19

-------
NAS (National Academy of Sciences). 2006. Health Risks from Dioxin and Related
Compounds: Evaluation of the EPA Reassessment. National Academies Press, Washington, DC
(July).  Available at http://www.nap.edu/catalog.php?record_id=l 1688.

U.S. EPA (U.S. Environmental Protection Agency).  2003. Exposure and Human Health
Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. NAS
Review Draft (EPA/600/P-00/001Cb).  U.S. Environmental Protection Agency, National Center
for Environmental Assessment, Washington, DC.  Available at
http://www.epa.gov/nceawwwl/pdfs/dioxin/nas-review/.
SESSION 5: QUANTITATIVE UNCERTAINTY ANALYSIS OF DOSE-RESPONSE
       This session addressed the uncertainty analysis to be considered for the dose-response
assessments.  The session opened with a presentation on current estimates of dioxin exposure
levels. Then it focused on the factors to include in the scope of an uncertainty analysis including
dioxin kinetics.

Session 5 Panelists (Session Co-Chairs are identified by asterisk)
   •   Bruce Allen, Bruce Allen Consulting
   •   Lesa Aylward, Summit Toxicology
   •   Roger Cooke, Resources for the Future
   •   Kenny Crump, Louisiana Tech University
   •   MikeDeVito, U.S. EPA
   •   Dale Hattis, Clark University
   •   *Rick Hertzberg, Biomath Consulting
   •   Nancy Kerkvliet, Oregon State University
   •   Leonid Kopylev,  U. S. EPA
   •   Rob McDowell, U.S. Department of Agriculture
   •   Lorenz Rhomb erg, Gradient
   •   Woody Setzer, U.S. EPA
   •   Marie Sweeney, National Institute of Occupational Safety and Health
   •   *Linda Teuschler, U.S. EPA

Please note that the use of the term "concluded" or "recommended" in this summary does not mean that a consensus
was reached.  Session Summaries were written from the material prepared by the non-EPA/ANL Co-Chair and
represent a synopsis of the panel discussions.

The Panel summarized the NAS comments regarding uncertainty. Areas for improvement
include:

   •   Ensure "transparency, thoroughness, and clarity in quantitative uncertainty analysis."

   •   Describe  and define (quantitatively to the extent possible) the variability and uncertainty
       for key assumptions used for each key endpoint-specific risk assessment, including
       choices of data set, point of departure, dose-response model, and dose metric.

   •   Incorporate probabilistic models to represent the range of plausible values.
                                         B-20

-------
   •   Assess goodness-of-fit of dose-response models.

   •   Provide upper and lower bounds on central tendency estimates for all statistical estimates.

   •   When quantification is not possible, clearly state it, and explain what would be required
       to achieve quantification.

Identification of Important Uncertainties
       The Panel reviewed the charge questions (Appendix B), discussed them, and listed eight
issues for consideration by the U.S. EPA:

   •   Concerning species and strain differences in the U.S. EPA's Response to NAS, current
       U.S. EPA procedures do not take this into account when selecting one data set for risk
       assessment.  Issues include "Where are humans in the distribution of potencies that can
       be generated? How likely is it that human response is similar to the selected data? Can
       we infer inter-individual variability from these differences?"

   •   Concerning the use of animal data for cross species extrapolation to humans (PK and PD
       uncertainties), issues to consider include differences in distribution and responses
       following bolus doses from those of subchronic and chronic protocols; uncertainty in
       liver doses due to sequestration;  differences in receptor binding affinity among
       congeners; and age factors (e.g.,  assumption of a lifetime constant daily dose for a cancer
       extrapolation).

   •   Concerning the description of AhR response, biochemical changes occur at lower doses
       than toxicological changes.  There should be an effort to identify the biochemical changes
       that would mark Ah receptor binding to inform the BMR,  and, thus, prevent toxicity.

   •   Concerning model uncertainty, the mathematical model choice depends on endpoint.
       There should be an effort towards determining what is the most sensitive endpoint(s) for
       humans and conducting animal studies  to model that endpoint(s).

   •   Concerning exposure and dose response in human studies, ensure  enough similarity to
       current human exposure profiles (mixture composition) so that a dose-response
       assessment can be done. Incorporate new epidemiological studies. Evaluate
       concordance with animal data and consistency across studies. Panel-acknowledged
       uncertainties include exposure estimates from person to person, shape of human dose-
       response  curve, healthy worker effect, and age dependence.

   •   Concerning POD determination, uncertainty factors are inherently mathematically
       inconsistent and that should be conveyed in the discussion of uncertainties when
       interpreting the POD.

   •   Concerning dose metric, tissue concentration is preferred.  It should be evaluated against
       a background of variability in  AhR-binding expression. There is uncertainty in what
       level of binding should be considered, in different cell types, tissues, life stage
       (development).  The relationship between dose metric and causation of adverse effects
       should be examined.
                                          B-21

-------
Low-Dose Extrapolation
       The Panel reviewed the charge questions and discussed them (Appendix B). The Panel
concluded that curve-fitting uncertainty (for a given dataset, dose metric, and model) can be
characterized and is useful, but, by itself, it is an incomplete characterization of uncertainty. The
Panel acknowledged the difficulty of fully characterizing uncertainty, especially quantitatively.
Some panelists argued that the problem is insurmountable and that no meaningful uncertainty
analysis is likely to be performable.  Other panelists contended that, the difficulties
notwithstanding, "good-faith" efforts to do something practical and forthright to characterize
uncertainty in low-dose extrapolation would be useful and important. The Panel clarified "good
faith" as meaning a characterization that is useful and not misleading to decision makers and is
inclusive of approaches that have meaningful support in the scientific community as a whole.
Being in "good faith" is more important than being complete (i.e., addressing every uncertain
element), especially since completeness is not a realistic goal. From this discussion, the Panel
listed four issues for consideration by the U.S. EPA:

   •   Review alternative data sets, dose metrics, and models to see where consequential
       uncertainties and impacts on low-dose implications arise.

   •   Consider the impacts of choices among plausible alternative data sets, dose metrics,
       models, and other more qualitative choices—issues include how much difference the
       choices make and also how much relative credence should be put to each alternative as a
       way of gauging and describing the landscape of imperfect knowledge
       regarding possibilities for the true dose-response.
       •   Hard to do quantitatively, since the factors are not readily expressed as statistical
          distributions, but can describe the rationale for believing/doubting each alternative in
          terms of available supporting evidence, contrary evidence, and needed assumptions.
       •   Expert judgment methods may be helpful in characterizing the relative weights of
          scientific credibility among alternatives. The expert judgment process, when
          conducted systematically, can be thought of as adding data to the assessment of
          credibility of alternatives, rather than as just an opinion poll.
       •   Information on plausibility of alternative low-dose extrapolation approaches can
          come from external considerations of mode of action, and not just from statistical
          success at fitting particular (high-dose) data sets.

   •   Characterizing uncertainty through a variety of approaches could be tried, and their
       relative merits and shortcomings discussed, as a way forward.

   •   Consider the sources of potential error, particularly in epidemiological data (e.g., TEF
       uncertainty and variation in congener mixtures) and if possible quantify their impact on
       the dose-response assessment.

Considerations for Conducting Uncertainty Analysis
       Overall, the Panel was split on whether U.S. EPA should do quantitative uncertainty
analyses.  The Panel noted that if done on only  some of the uncertainties, then results would be
misleading and could be misused. Ultimately, the Panel listed seven issues for consideration by
the U.S. EPA:
                                          B-22

-------
   •   The Panel recapped what some consider as being the first integrated risk assessment, with
       structured expert judgment and uncertainty analysis, i.e., the Rasmussen Report
       (WASH-1400; U.S. Nuclear Regulatory Commission, 1975).  In their discussion of the
       report, the Panel noted that in addition to standard event tree/fault tree modeling, this
       report also tackled difficult model uncertainty issues involved in accident progression,
       dispersion of released pollutants in the atmosphere,  environmental transport, exposure,
       health, and economic impacts. And though the Panel also recognized that this method
       was no longer state-of-the-art, the Panel contended that it represents a good example of a
       structured approach and methodology that could be  built upon.

   •   The Panel also discussed TEQs used in epidemiological studies, based on intake, and
       recognized that the key uncertainty in what was measured was not just intake but also
       involved PK/PD issues. The Panel acknowledged that the TEQ system is regularly used
       on a concentration basis, but they expressed concern that the qualification becomes lost.
       TEQs ignore pharmacokinetics and the common practice of rounding to orders of
       magnitude introduces more error.

   •   Structure the risk assessment along MOA steps—identify key biochemical measures
       (-5-10) common across toxic endpoints and identify the degree of meaningful change in
       effect or effect variance.  Make a table with all  options for data set, model, etc.; make
       best estimates/choices and determine which of these choices matter the most to the
       answer.

   •   Use expert panels—expert judgment can be collected scientifically (procedures are
       published).  But there are known biases; central tendency estimates work much better
       than extremes.

   •   Use supporting studies to fill in critical data gaps—Info filling methods do exist (e.g., PK
       modeling).  Put short-term studies into the "supporting info" category (unless, of course,
       the risk assessment is for acute exposures, such as chemical spills).

   •   Be creative in the analysis of uncertainty.  Intermediate steps between AhR binding and
       the end processes can be hypothesized based on data, experiences, and analogies related
       to other chemicals.

   •   The 2003 Reassessment presented potency estimates on wide variety of
       endpoints/models; needed to be more transparent in that discussion. Statistical graphics
       can be used to convey uncertainties.

Reference

U.S. Nuclear Regulatory Commission. 1975. Reactor Safety Study: An Assessment of Accident
Risks in U.S. Commercial Nuclear Power Plants.  WASH-1400 (NUREG-75-014). Washington,
DC.
                                         B-23

-------
            APPENDIX A: 2009 U.S. EPA DIOXIN WORKSHOP AGENDA


                              SCIENTIFIC WORKSHOP
    TO INFORM THE TECHNICAL WORK PLAN FOR U.S. EPA'S RESPONSE TO
             NAS COMMENTS ON THE HEALTH EFFECTS OF DIOXIN
                PRESENTED IN U.S. EPA'S DIOXIN REASSESSMENT

                                    Cincinnati, OH

                              Date: February 18-20, 2009


BACKGROUND/WORKSHOP OBJECTIVE
       At the request of the U.S. Environmental Protection Agency (U.S. EPA), the National
Academy of Sciences (NAS) prepared a report, Health Risks from Dioxin and Related
Compounds: Evaluation of the EPA Reassessment (NAS, 2006), that made a number of
recommendations to improve the U.S. EPA's risk assessment for 2,3,7,8-tetrachlorodibenzo-
/>-dioxin (TCDD). In response, the U.S. EPA will prepare a technical report that addresses key
comments on the dose-response assessment for TCDD. The U.S. EPA intends to develop its
response through a transparent process that provides multiple opportunities for input.

       To assist in this effort, a Workshop will be held to inform the U.S. EPA's evaluation of
the NAS recommendations. The Workshop will be open to the public. At the Workshop, the
U.S. EPA will solicit input from expert scientists and the public.

       The goal of the Workshop is to ensure that the U.S. EPA's response to the NAS
comments focuses on the key issues and reflects the most meaningful science. The three main
objectives of the Workshop are to (1) identify and discuss the technical challenges involved in
addressing the NAS key comments on the TCDD dose-response assessment in the U.S. EPA
Reassessment (U.S. EPA, 2003), (2) discuss approaches for addressing these comments, and
(3) identify key published, independently peer-reviewed literature, particularly studies describing
epidemiologic and in vivo mammalian bioassays, which are expected to be most useful for
informing the U.S. EPA response.

       Workshop participants will be encouraged to think broadly about the body of scientific
information that can be used to inform the U.S. EPA's response and to participate in open
dialogue regarding ways in which the science can best be used to address the key dose-response
issues. This Workshop  is similar to scientific workshops being conducted under the new review
process for the National Ambient Air Quality Standards (NAAQS)1 that assess health-related
information for criteria pollutants.
1 Please see http://www.epa.gov/ttn/naaqs/ for more information on the new NAAQS review process.


                                         B-24

-------
       The Workshop discussions are expected to build upon two prior publications:
       1.  Exposure and Human Health Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin
          (TCDD) and Related Compounds (U.S. EPA, 2003). This external review draft
          provides a comprehensive reassessment of dioxin exposure and human health effects.
          This "dioxin reassessment" was submitted in October 2004 to the National Academy
          of Sciences (NAS) for review.
       2.  Health Risks from Dioxin and Related Compounds: Evaluation of the EPA
          Reassessment (NAS, 2006).

       Workshop participants are encouraged to review both of these documents and other
relevant materials (e.g., the National Toxicology Program report on TCDD [NTP, 2006]) before
the meeting because they provide important insights into the key questions and challenges.
There are a number of open comment periods that are intended to facilitate a broad discussion of
the issues.

       Scientists with significant expertise and experience relevant to the health effects of
TCDD or dioxin-like compounds and associated topics will be asked to serve on "expert panels"
for discussions throughout the Workshop. Workshop panelists will include a wide range of
experts representing many scientific areas needed to assess TCDD dose-response (e.g.,
epidemiology, human and animal toxicology, nuclear receptor biology, dose-response modeling,
risk assessment, and uncertainty analysis).  The Workshop panelists will be asked to highlight
significant and emerging research and to make recommendations to the U.S. EPA regarding the
design and  scope of the technical response to NAS comments on the dose-response analysis for
TCDD—including, but not limited to, recommendations for evaluating associated uncertainty.
Open comment periods will follow each panel discussion session.  Public participation will be
encouraged by way of these designated open comment periods and, also,  by participation in the
scientific poster session planned for the second evening (February 19).

       U.S. EPA will use the input received during this Workshop as the foundation for its
development of a technical work plan for responding to the NAS comments on the TCDD dose-
response analysis. The work plan will outline the schedule, process, and  approaches for
evaluating the relevant scientific information and addressing the key issues.  The work plan also
will identify the key literature to be utilized in U.S. EPA's response.

       As a follow-on activity to this Workshop, a panel is being established under the Federal
Advisory Committee Act (FACA) to guide and review the U.S. EPA's response to NAS
comments.  The FACA panel will be asked to conduct a consultation with the Agency on the
draft technical work plan. At the same time, the public will also have the opportunity to provide
comments to the FACA panel on the work plan. The final technical work plan will guide the
development of the technical  report that will constitute the U.S. EPA's response to NAS
comments.  During the development of this response, the U.S. EPA will seek advice from the
FACA panel and the public several times. Finally, the FACA panel will be asked to review the
technical report in a public forum.

       The preliminary Agenda presented on the following pages may be revised prior to the
Workshop following review by the session Co-Chairs; the dates and general timing of the
                                         B-25

-------
sessions, however, will not change. A final Agenda and a set of charge questions, intended to
provide general direction for the Workshop discussions, will be posted on the Workshop Internet
site (http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=199923) prior to the meeting.

       A poster session will be held on the evening of the second day (February 19). The
purpose of this poster session is to provide a forum for scientists to present recent studies
relevant to TCDD dose-response assessment and to encourage open discussion about these
presentations.

REFERENCES
NAS (National Academy of Sciences). 2006. Health Risks from Dioxin and Related
Compounds: Evaluation of the EPA Reassessment. National Academies Press, Washington, DC
(July).  Available at http://www.nap.edu/catalog.php?record_id=l 1688.

NTP (National Toxicology Program). 2006. Toxicology and Carcinogenesis Studies of
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) (CAS No. 1746-01-6) in Female Harlan Sprague-
Dawley Rats (Gavage Studies). U.S. Department of Health and Human Services. NTP TR 521.
Research Triangle Park, NC (April).

U.S. EPA (U.S. Environmental Protection Agency). 2003. Exposure and Human Health
Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds., NAS
review draft, Volumes  1-3 (EPA/600/P-00/001Cb, Volume 1). U.S. Environmental Protection
Agency, National Center for Environmental Assessment, Washington, DC (December).
Available at http://www.epa.gov/nceawwwl/pdfs/dioxin/nas-review/.
                                         B-26

-------
WORKSHOP AGENDA


                                   Davl


8:00-9:00          Registration

9:00-9:30          Welcome/Purpose of Meeting/Document Development Process

9:30-9:45          Panel Comments/Questions on Charge


9:45-2:45        Session 1: Quantitative Dose-Response Modeling Issues
                  (Hall of Mirrors)

     9:45-10:10     Background/Introductory Remarks

     10:10-10:35    TCDD Kinetics: Converting Administered Doses in Animals to
                  Human Body Burdens
                  Presenter: Michael Devito

     10:35-11:30    Panel Discussion

     11:30-1:00     Lunch

     1:00-2:00      Panel Discussion cont.

     2:00-2:45      Open Comment Period


2:45-3:05        Break


3:05-5:15        Session 2: Immunotoxicitv (Hall of Mirrors)

     3:05-3:15      Background/Introductory Remarks

     3:15-4:45      Panel Discussion

     4:45-5:15      Open Comment Period
                                     B-27

-------
8:00-8:30

    8:00-8:15

    8:15-8:30
                 Day 2

Report-Outs for Sessions 1 and 2 (Hall of Mirrors)

Report-Out for 1: Quantitative Dose-Response Modeling Issues

Report-Out for 2: Immunotoxicity
8:30-11:30

8:30-11:30


    8:30-8:45

    8:45-11:00
8:30-11:30


    8:30-8:45

    8:45-11:00
Sessions 3A and 3B (concurrent sessions)

Session 3A: Dose-Response for Neurotoxicitv and
Nonreproductive Endocrine Effects (Hall of Mirrors)

Background/Introductory Remarks

Panel Discussion
     11:00-11:30   Open Comment Period
Session 3B: Dose-Response for Cardiovascular Toxicitv and
Hepatotoxicitv (Rookwood Room)

Background/Introductory Remarks

Panel Discussion
     11:00-11:30   Open Comment Period

11:30-1:00       Lunch
1:00-2:00
Report-Outs for Sessions 3A and 3B (Hall of Mirrors)
The structure of the session report-outs will include the following:

      •   Summary of session presentation including minority opinion
      •   Public comments
      •   Discussion

     1:00-1:15     Report-Out for 3A: Dose-Response for Neurotoxicity and
                 Nonreproductive Endocrine Effects

     1:15-1:30     Open Comment Period
                                    B-28

-------
    1:30-1:45     Report-Out for 3B: Dose-Response for Cardiovascular Toxicity and
                 Hepatotoxicity

    1:45-2:00     Open Comment Period
2:00-5:15        Sessions 4A and 4B (concurrent sessions)

2:00-5:15        Session 4A: Dose-Response for Cancer (Hall of Mirrors)

    2:00-2:15     Background/Introductory Remarks

    2:15-4:45     Panel Discussion

    4:45-5:15     Open Comment Period


2:00-5:15        Session 4B: Dose-Response for
                 Reproductive/Developmental Toxicitv (Rookwood Room)

    2:00-2:15     Background/Introductory Remarks

    2:15-4:45     Panel Discussion

    4:45-5:15     Open Comment Period


6:45-8:15        Poster Session (Rosewood Room)



                                  Day 3


8:30-9:30        Report-Outs for Sessions 4A and 4B (Hall of Mirrors)

    8:30-8:45     Report-Out for 4A: Dose-Response for Cancer

    8:45-9:00     Open Comment Period

    9:00-9:15     Report-Out for 4B: Dose-Response for Reproductive/Developmental
                 Toxicity

    9:15-9:30     Open Comment Period
                                   B-29

-------
9:30-3:30        Session 5: Quantitative Uncertainty Analysis of Dose-
                  Response (Hall of Mirrors)

     9:3 0-9:40      Background/Introductory Remarks

     9:40-10:10     Evidence of a Decline in Background Dioxin Exposures in Americans
                  Between the 1990s and 2000s
                  Presenter: Matt Lorber

10:10-10:30      Break

     10:30-11:30    Panel Discussion

     11:30-1:00     Lunch

     1:00-2:15      Panel Discussion cont.

     2:15-2:30      Break

     2:30-3:00      Open Comment Period

     3:00-3:15      Report-Out for 5: Quantitative Uncertainty Analysis of Dose-
                  Response

     3:15-3:30      Closing Remarks

3:30              Adjourn
                                     B-30

-------
                 APPENDIX B: 2009 U.S. EPA DIOXIN WORKSHOP
                   QUESTIONS TO GUIDE PANEL DISCUSSIONS
SESSION 1

Dose Metric
Considering all of the endpoints or target tissues, and species that U.S. Environmental Protection
Agency (U.S. EPA)'s dose-response modeling might evaluate, what are the best measures of
dose (e.g., ingested, tissue concentrations, body burden, receptor occupancy, other surrogate) and
why?
Developing Dose-Response Models from Mammalian Bioassays
How best can the point of departure (POD) be determined when the response range is
incompletely characterized (i.e., high response at the lowest dose or low response at the highest
dose; observed in several key 2,3,7,8-Tetrachlorodibenzo-p-Dioxin [TCDD] studies)?

If considered to be biologically plausible, how can a threshold be incorporated into a dose-
response function (e.g., for TCDD cancer data)?

How can nonmonotonic responses be incorporated into the dose-response function?
Developing Dose-Response Models from Epidemiological Studies
How can the epidemiological data be utilized best to inform the TCDD exposure-response
modeling? Which epidemiological studies are most relevant?
Supporting Information
For those toxicological endpoints that are Ah receptor-mediated, how would the receptor kinetics
influence the shape of the dose-response curve? How would downstream cellular events affect
the shape of the dose-response curve? How can this cascade of cellular events be incorporated
into a quantitative model of dose-response?
                                        B-31

-------
SESSIONS 2, 3A, 3B, 4A, AND 4B

Key Study Selection
For this endpoint, what refinements should be made to the draft criteria for selection of key
studies?

What are the specific effects of concern for human health for this endpoint?

Based on the draft criteria for the selection of key studies, what are the key studies informing the
shape of the dose-response curve above the POD and the choice of the POD for this endpoint?
Epidemiological Study Utility
How and to what extent do the epidemiological data inform the choice of critical effect?

How can the epidemiological data inform the quantitative dose-response modeling?
Animal Model Utility
Are there types of effects observed in animal models that are more relevant to humans than
others?  To what extent does information on mode of action (MOA) influence the choice of
animal model (species, strain, sex)?


Supporting Information

Are there studies that establish a sufficient justification for departure from the default procedures
that address the shape of the dose-response curve below the POD under the cancer guidelines?

Are there studies that establish a sufficient justification for departing from U.S. EPA's default
approaches for noncancer endpoints?

To what extent can MOA information clarify the identification of endpoints of concern and dose-
response metric for this endpoint?  How can the cascade of cellular events for this endpoint be
incorporated into a quantitative model of dose response?
                                          B-32

-------
SESSION 5

For cancer and noncancer TCDD dose-response assessments, U.S. EPA is interested in
developing a quantitative uncertainty analysis addressing both parameter and model uncertainty,
if feasible. Uncertainties will include, among others, choice of endpoint; underlying study
uncertainties; choice of dose metric; interspecies extrapolations such as kinetic uncertainties; and
choice of dose-response model, including threshold models. The U.S. EPA is currently
examining techniques and tools for uncertainty analysis—including Bayesian and frequentist
approaches.
Identification of Important Uncertainties
What are the major uncertainties pertaining to modeling the animal data?
       Consider the dose metric (species or tissue specificity), vehicle of administration,
       exposure frequency, exposure duration, and POD determination (e.g., benchmark
       response selection or no-observed-adverse-effect level/lowest-observed-adverse-effect
       level identification).

What are the major uncertainties pertaining to dose-response modeling below the POD?
       Consider how receptor kinetics and downstream cellular event information might be used
       to bound the uncertainties associated with dose-response modeling below the POD.

What are the major uncertainties in cross-species extrapolation (e.g., half-lives, tissue
distribution, and toxicodynamics)?
       Consider the primary species dosed with TCDD: mice, hamsters,  rats, guinea pigs, and
       monkeys.

What are the major uncertainties pertaining to intrahuman variability?
       Consider what data sets would be useful to represent sensitive subpopulations.

What are other significant sources of uncertainty for the cancer and noncancer assessments?
Considerations for Conducting Uncertainty Analysis
What data sets could be used to quantify uncertainties in cancer and noncancer TCDD dose-
response assessments?
       Consider dioxin-like compound dose-response data.
       Consider MOA information.

What are the appropriate techniques for the TCDD dose-response uncertainty analysis, and what
are their respective strengths and weaknesses of these approaches as applied to TCDD?
                                          B-33

-------
        APPENDIX C: 2009 U.S. EPA DIOXIN WORKSHOP DRAFT SELECTION CRITERIA TO IDENTIFY KEY IN
                       VIVO MAMMALIAN STUDIES THAT INFORM DOSE-RESPONSE MODELING FOR
                                        2,3,7,8-TETRACHLORODIBENZO-/?-DIOXIN(TCDD)a
Study Feature
Chemical, purity,
matrix/medium
Peer review
Study design,
execution, and
reporting
Study subject:
species, strain, and
sensitivity for given
endpoint; litter; life
stage; gender
Exposure route
Dose level
Exposure frequency,
duration, and timing
Controls
Response
Statistical evaluation
Selection Rationale
Pr/ma/y"
TCDD-only doses included, purity specified,
matrix in which TCDD is administered is identified
Independently peer-reviewed, publicly available
Clearly documented and consistent with standard
toxicological principles, testing protocols,
and practice (i.e., endpoint-appropriate,
particularly for negative findings)
Mammalian species
Strain and gender identified
Animal age at beginning of treatment identified
Litter confounders (within/between) accounted for
Oral
Lowest dose <200 ng/kg-d for noncancer
endpoints and <1 ug/kg-d for cancer
Dosing regimen characterized and explained
Appropriate and well characterized
Effect relevant to human health
Magnitude outside range of normal variability
Clearly described and appropriate to the endpoint
and study design (e.g., per error variance,
magnitude of effect)
Secondary*
TCDD purity or matrix not clearly identified
Supplementary materials accompanying
peer-reviewed publication
Testing protocol provides incomplete
coverage of relevant endpoint-specific
measures, particularly for negative findings
Mammalian species, in vivo, but only
studying an artificially sensitive subject
(e.g., knockout mouse)
Parenteral (e.g., intravenous, intramuscular,
intraperitoneal, subcutaneous)
Lowest dose >200 ng/kg-d for noncancer
endpoints, or >1 .0 ug/kg-d for cancer

Effect reported, but with no negative control
Precursor effects, or adaptive responses
potentially relevant to human health
Limited statistical context
Currently Excluded
Studies of dioxin-like compounds
(DLCs) or mixtures
Not formally peer-reviewed; literature
not publicly available
Studies not meeting standard
principles and practices
Non-mammalian or not in vivo
Inhalation, dermal, ocular

Characterization/explanation
missing or cannot be determined

Lethality

td
      ' NAS (2006) commented that the selection of data sets for quantitative dose-response modeling needed to be more transparent. These draft criteria are
        offered for consideration at the kickoff workshop. These criteria would be used to identify candidate studies of non-human mammals that would be used to
        define the point-of-departure (POD). These criteria are not designed for hazard identification or weight-of-evidence determinations. Studies addressing data
        other than direct TCDD dose-response in mammals (including toxicokinetic data on absorption, distribution, metabolism, or elimination; information on
        physiologically-based pharmacokinetic [PBPK] modeling, and mode of action data) will be evaluated separately.
      1 Presents preliminary draft criteria for evaluating a study being considered for estimating a POD in a TCDD dose-response model.
       Presents preliminary draft criteria that could qualify a study as primary with support from other lines of evidence (e.g., PBPK modeling), when no study for an
        endpoint meets the "primary" criteria.

-------
                                     EPA/600/R-10/038F
                                      www.epa.gov/iris
               APPENDIX C

Summaries and Evaluations of Cancer and
   Noncancer Epidemiologic Studies for
    Inclusion in TCDD Dose-Response
                 Assessment
                  January 2012
           National Center for Environmental Assessment
              Office of Research and Development
             U.S. Environmental Protection Agency
                   Cincinnati, OH

-------
    CONTENTS—Appendix C: Summaries and Evaluations of Cancer and Noncancer
        Epidemiologic Studies for Inclusion in TCDD Dose-Response Assessment
LIST OF TABLES	C-vi

APPENDIX C.   SUMMARIES AND EVALUATIONS OF CANCER AND
      NONCANCER EPIDEMIOLOGIC STUDIES FOR INCLUSION IN TCDD
      DOSE-RESPONSE ASSESSMENT	C-8
      C.I. EVALUATION OF EPIDEMIOLOGIC STUDIES FOR DOSE-
           RESPONSE ASSESSMENT	C-8
           C.I.I.  Cancer	C-8
                  C.I. 1.1.  Cancer Cohorts	C-9
                          C.I.1.1.1. The NIOSH cohort	C-9
                                   C.I.1.1.1.1.   Fingerhut et al. (1991a)	C-9
                                   C.I.1.1.1.2.   Steenland et al. (1999)	C-15
                                   C.I.1.1.1.3.   Steenland etal. (2001b)	C-19
                                   C.I.1.1.1.4.   Cheng et al. (2006)	C-25
                                   C.I.1.1.1.5.   Collins et al. (2009)	C-28
                          C.I.1.1.2. The BASF cohort	C-31
                                   C. 1.1.1.2.1.   Thiess and Frentzel-Beyme
                                                (1977) and Thiess etal. (1982)	C-31
                                   C.I.1.1.2.2.   Zoberetal. (1990)	C-33
                                   C.I.1.1.2.3.   OttandZober(1996a)	C-35
                          C.I. 1.1.3. The Hamburg cohort	C-38
                                   C.I.1.1.3.1.   Manz etal. (1991)	C-39
                                   C.I.1.1.3.2.   Flesch-Janys etal. (1995)	C-43
                                   C.I.1.1.3.3.   Flesch-Janys etal. (1998)	C-46
                                   C.I.1.1.3.4.   Becheretal. (1998)	C-47
                          C.I. 1.1.4. The Seveso cohort	C-50
                                   C.I.1.1.4.1.   Bertazzi etal. (2001)	C-51
                                   C.I.1.1.4.2.   Warneretal. (2002)	C-53
                                   C.I.1.1.4.3.   Pesatori et al. (2003)	C-56
                                   C.I.1.1.4.4.   Baccarelli et al. (2006)	C-57
                                   C.I.1.1.4.5.   Consonni etal. (2008)	C-58
                          C.I.1.1.5. Chapaevsk study	C-59
                                   C.I.1.1.5.1.   Revich et al. (2001)	C-59
                          C. 1.1.1.6. The Air Force Health ("Ranch Hands" cohort)
                                   study	C-61
                                   C.I.1.1.6.1.   Akhtar et al. (2004)	C-62
                                   C.I.1.1.6.2.   Michalek and Pavuk (2008)	C-65
                          C. 1.1.1.7. Other studies of potential relevance to dose-
                                   response modeling	C-67
                                   C.I. 1.1.7.1.   Hooiveld et al. (1998)—
                                                Netherlands workers	C-67
                                        C-ii

-------
                    CONTENTS (continued)
                         C.I. 1.1.7.2.   t' Mannetje et al. (2005)—New
                                      Zealand herbicide sprayers	C-70
                         C.I. 1.1.7.3.   McBride et al. (2009b)—New
                                      Zealand herbicide sprayers	C-73
                         C.I. 1.1.7.4.   McBride et al. (2009a)—New
                                      Zealand herbicide sprayers	C-76
       C.I.1.2. Key Characteristics of Epidemiologic Cancer Studies	C-77
       C.I. 1.3. Feasibility of TCDD Cancer Dose-Response Modeling—
               Summary Discussion by Cohort	C-78
               C. 1.1.3.1.  Using the NIO SH cohort in dose-response
                         modeling	C-78
               C.I. 1.3.2.  Using the BASF cohort in dose-response
                         modeling	C-80
               C. 1.1.3.3.  Using the Hamburg cohort in dose-response
                         modeling	C-81
               C.I. 1.3.4.  Using the Seveso cohort in dose-response
                         modeling	C-81
               C.I.1.3.5.  Using the Chapaevsk related data in dose-
                         response modeling	C-83
               C. 1.1.3.6.  Using the Ranch Hands cohort in dose-response
                         modeling	C-83
       C. 1.1.4. Discussion of General Issues Related to Dose-Response
               Modeling	C-83
               C.I. 1.4.1.  Ascertainment of exposures	C-83
               C.I. 1.4.2.  Latency intervals	C-84
               C.I.1.4.3.  Use of the SMR metric	C-84
               C.I.1.4.4.  All cancers versus site-specific	C-87
               C.I. 1.4.5.  Summary of epidemiologic cancer study
                         evaluations for dose-response modeling	C-87
C.1.2.  Noncancer	C-87
       C.I.2.1. Noncancer Cohorts	C-88
               C.1.2.1.1.  The NIOSH cohort	C-88
                         C.1.2.1.1.1.   Steenland et al. (1999)	C-88
                         C.1.2.1.1.2.   Collins et al. (2009)	C-89
               C.1.2.1.2.  The BASF cohort	C-90
                         C.1.2.1.2.1.   OttandZober	C-90
               C.1.2.1.3.  The Hamburg cohort	C-92
                         C.1.2.1.3.1.   Flesch-Janys et al. (1995)	C-92
               C.1.2.1.4.  The Seveso Cohort—SWHS	C-94
                         C. 1.2.1.4.1.   Eskenazi et al.  (2002b)—
                                      menstrual cycle characteristics	C-95
                         C.I.2.1.4.2.   Eskenazi et al.  (2002a)—
                                      endometriosis	C-97

-------
     CONTENTS (continued)
          C.1.2.1.4.3.    Eskenazi et al. (2003)—birth
                        outcomes	C-99
          C.I.2.1.4.4.    Warner et al. (2004)—age at
                        menarche	C-101
          C.I.2.1.4.5.    Eskenazi et al. (2005)—age at
                        menopause	C-102
          C.I.2.1.4.6.    Warner et al. (2007)—ovarian
                        function	C-105
          C.I.2.1 A.I.    Eskenazi et al. (2007)—uterine
                        leiomyoma	C-106
C.I.2.1.5. Other Seveso noncancer studies	C-108
          C.I.2.1.5.1.    Bertazzi etal. (1989); Consonni
                        et al. (2008)—mortality
                        outcomes	C-108
          C.I.2.1.5.2.    Mocarelli et al. (2000; 1996)—
                        sex ratio	C-110
          C.1.2.1.5.3.    Baccarelli et al.  (2004; 2002)—
                        immunologic effects	C-113
          C.I.2.1.5.4.    Landi et al. (2003)—gene
                        expression	C-115
          C.I.2.1.5.5.    Alaluusua et al.  (2004)—
                        developmental dental defects	C-117
          C.I.2.1.5.6.    Baccarelli et al.  (2005)—
                        chloracne	C-119
          C.I.2.1.5.7.    Baccarelli et al.  (2008)—
                        neonatal thyroid hormone levels....C-120
          C.I.2.1.5.8.    Mocarelli et al. (2008)—sperm
                        effects	C-123
C.1.2.1.6. The Chapaevsk study	C-125
          C.I.2.1.6.1.    Revich et al. (2001)—mortality
                        and reproductive health	C-125
C. 1.2.1.7. The Air Force Health ("Ranch Hands"  cohort)
          study	C-126
          C.1.2.1.7.1.    Henriksen et al., (1997)	C-127
          C.1.2.1.7.2.    Longnecker and Michalek (2000)..C-130
          C.1.2.1.7.3.    Michalek et al. (2001a)	C-133
          C.I.2.1.7.4.    Michalek et al. (200lb)—hepatic
                        health outcomes	C-136
          C.1.2.1.7.5.    Michalek et al. (2001c)—
                        peripheral neuropathy	C-139
          C.I.2.1.7.6.    Pavuk et al. (2003)—thyroid
                        health endpoints	C-142
               C-iv

-------
                        CONTENTS (continued)
                             C.I.2.1.7.7.   Michalek and Pavuk (2008)—
                                         diabetes	C-145
                   C.I.2.1.8.  Other noncancer studies of TCDD	C-146
                             C.I.2.1.8.1.   Ryanetal. (2002)—sex ratio	C-146
                             C.I.2.1.8.2.   Kang et al.(2001)—long-term
                                         health effects	C-148
                             C.1.2.1.8.3.   McBride et al. (2009a) —
                                         noncancer mortality	C-152
                             C.I.2.1.8.4.   McBride et al. (2009b)—
                                         noncancer mortality	C-153
           C.I.2.2. Feasibility of Dose-Response Modeling for Noncancer	C-154
           C.I.2.3. Summary of Epidemiologic Noncancer Study Evaluations
                   for Dose-Response Modeling	C-155
C.2.  EVALUATION TABLES FOR CANCER STUDIES	C-167
     C.2.1. NIOSH Cohort Studies	C-167
     C.2.2. BASF Cohort Studies	C-172
     C.2.3. The Hamburg Cohort	C-173
     C.2.4. The Seveso Cohort Studies	C-177
     C.2.5. The Chapaevsk Study	C-182
     C.2.6. The Air Force Health ("Ranch Hands") Study	C-182
     C.2.7. Other Studies of Potential Relevance to Dose-Response Modeling	C-184
C.3.  EVALUATION TABLES FOR NONCANCER STUDIES	C-187
     C.3.1. NIOSH Cohort	C-187
     C.3.2. BASF Cohort	C-189
     C.3.3. Hamburg Cohort	C-190
     C.3.4. The Seveso Women's Health Study	C-191
     C.3.5. Other Seveso Noncancer Studies	C-198
     C.3.6. Chapaevsk Study	C-204
     C.3.7. Air Force Health ("Ranch Hands") Study	C-205
     C.3.8. Other Noncancer Studies ofDioxin	C-211
C.4.  REFERENCES	C-215
                                 C-v

-------
                                   LIST OF TABLES
C-l.          Summary of epidemiologic cancer studies (key characteristics)	C-157
C-2.          Epidemiologic cancer study selection considerations and criteria	C-159
C-3.          Epidemiologic noncancer study selection considerations and criteria	C-163
C-4.          Fingerhut et al. (1991a)—All cancer sites, site-specific analysis	C-167
C-5.          Steenland et al. (1999)—All cancer sites combined, site-specific analysis	C-168
C-6.          Steenland et al. (2001b)—All cancer sites combined	C-169
C-7.          Cheng et al. (2006)—All cancer sites combined	C-170
C-8.          Collins et al. (2009)—All cancer sites combined, site-specific analysis	C-171
C-9.          Zober et al. (1990)—All cancer sites combined, site-specific analysis	C-172
C-10.         Ott and Zober (1996a)—All cancer sites combined	C-172
C-ll.         Manz et al. (1991)—All cancer sites combined, site-specific analyses	C-173
C-12.         Flesch-Janys et al. (1995); Flesch-Janys et al. (1996) erratum—All cancer
              sites combined	C-l74
C-l3.         Flesch-Janys et al. (1998)—All cancer sites combined, site-specific
              analysis	C-l 75
C-14.         Becher et al. (1998)—All cancer sites combined	C-176
C-15.         Bertazzi et al.  (2001)—All cancer sites combined, site-specific analyses	C-177
C-16.         Pesatori et al. (2003)—All cancer sites combined, site-specific analyses	C-178
C-17.         Consonni et al. (2008)—All cancer sites combined, site-specific analyses ....C-179
C-18.         Baccarelli et al. (2006)—Site-specific analysis	C-180
C-19.         Warner et al. (2002)—Breast cancer incidence	C-181
C-20.         Revich et al. (2001)—All cancer sites combined, and site-specific analyses..C-182
C-21.         Akhtar et al. (2004)—All cancer sites combined and site-specific analyses...C-182
C-22.         Michalek and Pavuk (2008)—All cancer sites combined	C-183
C-23.         't Mannetje et al.  (2005)—All cancer sites combined, site specific analyses..C-l84
C-24.         McBride et al. (2009a)—All cancer sites combined, site-specific analysis ....C-185
C-25.         McBride et al. (2009b)—All cancer sites combined, site-specific analysis ....C-185
C-26.         Hooiveld et al. (1998)—All cancer sites combined, site-specific analysis	C-186
C-27.         Steenland et al. (1999)—Mortality (noncancer)	C-187
C-28.         Collins et al. (2009)—Mortality (noncancer)	C-188
C-29.         Ott and Zober (1996a)—Mortality (noncancer)	C-189
C-30.         Flesch-Janys et al. (1995); Flesch-Janys et al. (1996) erratum—Mortality
              (noncancer)	C-l 90
C-31.         Eskenazi et al. (2002b)—Menstrual cycle  characteristics	C-191
C-32.         Eskenazi et al. (2002a)—Endometriosis	C-192
C-33.         Eskenazi et al. (2003)—Birth outcomes	C-193
C-34.         Warner et al. (2004)—Age at menarche	C-194
C-35.         Eskenazi et al. (2005)—Age at menopause	C-195
C-36.         Warner et al. (2007)—Ovarian function	C-196
C-37.         Eskenazi et al. (2007)—Uterine leiomyoma	C-197
C-38.         Mocarelli et al. (2008)—Semen quality	C-198
C-39.         Mocarelli et al. (2000)—Sex ratio	C-198
C-40.         Baccarelli et al. (2008)—Neonatal thyroid function	C-199
C-41.         Alaluusua et al. (2004)—Developmental dental defects	C-200

                                          C-vi

-------
                         LIST OF TABLES (continued)
C-42.        Bertazzi et al. (2001)—Mortality (noncancer)	C-201
C-43         Consonni et al. (2008)—Mortality (noncancer)	C-202
C-44.        Baccarelli et al. (2005)—Chloracne	C-203
C-45.        Baccarelli et al. (2004; 2002)—Immunological effects	C-203
C-46.        Revich et al. (2001)—Mortality (noncancer) and reproductive health	C-204
C-47.        Henriksen et al. (1997)—Diabetes	C-205
C-48.        Longnecker and Michalek (2000)—Diabetes	C-206
C-49.        Michalek et al. (2001a)—Hematological effects	C-206
C-50.        Michalek et al. (2001b)—Hepatic abnormalities	C-207
C-51.        Michalek et al. (200 lc)—Peripheral Neuropathy	C-208
C-52.        Pavuk et al. (2003) —Thyroid function and disorders	C-209
C-53.        Michalek and Pavuk (2008)—Diabetes	C-210
C-54.        McBrideetal. (2009b)—Mortality  (noncancer)	C-211
C-55.        McBride et al. (2009a)—Mortality  (noncancer)	C-212
C-56.        Ryan et al. (2002)—Sex ratio	C-213
C-57.        Kangetal. (2001)—Long term health consequences	C-214
                                         C-vii

-------
APPENDIX C.    SUMMARIES AND EVALUATIONS OF CANCER AND NONCANCER
                  EPIDEMIOLOGIC STUDIES FOR INCLUSION IN TCDD
                             DOSE-RESPONSE ASSESSMENT
   C.I.   EVALUATION OF EPIDEMIOLOGIC STUDIES FOR DOSE-RESPONSE
          ASSESSMENT
          This appendix summarizes and evaluates studies for potential use in tetrachlorodibenzo-
   p-dioxin (TCDD) dose-response assessment using the study evaluation considerations and
   inclusion criteria for epidemiologic data (see Section 2.3.1).  Those studies that meet the study
   inclusion criteria are listed in Section 2 of this document in Tables 2-1 and 2-2, for cancer and
   noncancer, respectively. The following sections, C. 1.1 and C. 1.2, for cancer and noncancer
   studies, respectively, are organized by epidemiologic study population. In Section C.I.I,
   following a brief summary of each cohort, its associated cancer studies are then summarized
   chronologically, assessed for methodological considerations relative to epidemiologic cohorts
   and studies and evaluated for suitability for TCDD dose-response assessment.  In Section C.I.2,
   summaries of the cohorts are not repeated, but are still used as an organizing element for this
   section. The reader is referred back to the cancer section for the cohort summaries. Following
   the heading for the cohort, its associated noncancer studies are then summarized chronologically,
   assessed for methodological considerations relative to epidemiologic cohorts and studies and
   evaluated for suitability for TCDD dose-response assessment.
          Sections C.2 and C.3 of this appendix provide specific details of the study selection
   criteria results for the cancer and noncancer epidemiologic studies, respectively.  This includes a
   table for each study with information on how each of the five considerations and three criteria
   were evaluated, and why each study was or was not selected by U.S. Environmental Protection
   Agency (EPA) for TCDD quantitative dose-response assessment.

   C.I.I.  Cancer
          In the 2003 Reassessment, EPA selected three cohort studies from which to conduct a
   quantitative dose-response analysis: the National Institute for Occupational Safety and Health
   (NIOSH) cohort (Steenland et al.. 2001b). the BASF  cohort (Ott and Zober. 1996b). and the
   Hamburg cohort (Becher et al., 1998). Although these studies were deemed suitable for a
   quantitative dose-response analysis, the criteria EPA  used to reach this conclusion were unclear.

-------
In this section, the study selection criteria and methodological considerations presented in
Section 2.3.1 are systematically applied to evaluate a number of studies to determine their
suitability for inclusion in dose-response modeling. In addition to the three cohorts used in
previous TCDD quantitative risk assessment, considerations are applied to other relevant TCDD
epidemiologic data sets that were identified through a literature review for epidemiologic studies
of TCDD and cancer up through 2009.  Study summaries and suitability for quantitative
dose-response analysis evaluations are discussed below.

C.I. 1.1.   Cancer Cohorts
C.l.1.1.1.  The NIOSHcohort
       In  1978, the NIOSH undertook research that identified workers employed by U.S.
chemical companies that made products contaminated with TCDD between 1942 and 1982.
TCDD was generated in the production of 2,4,5-trichlorophenol and subsequent processes. This
chemical was used to make 2,4,5-trichlorophenoxyacetic acid (2,4,5-T), which was a major
component of the widely-used defoliant, Agent Orange. The NIOSH cohort is the  largest cohort
of occupational workers studied to date, and has been the subject of a series of investigations
spanning more than two decades. It is important to note that this cohort consists mostly of male
workers that were chronically exposed to TCDD via daily occupational exposure, as compared to
an acute accidental exposure  scenario seen with  other cohorts.  The investigations have
progressed from a comparison of the mortality patterns of the cohort to the U.S. general
population to dose-response modeling using serum-derived estimates of TCDD that have been
back-extrapolated several decades. Analyses of cancer data from the NIOSH cohort that are
addressed in this section include studies published by Fingerhut et al. (1991a), Steenland et al.
(2001b; 1999), Cheng et al. (2006). and Collins et al. (2009).

C.l.1.1.1.1.  Finserhut et al (1991a)
C.I.1.1.1.1.1.  Study summary
       The investigation of Fingerhut and her colleagues published nearly two decades ago
attracted widespread attention (Fingerhut et al.. 199la).  This retrospective study examined
patterns of cancer mortality for 5,172 male workers who comprised the NIOSH cohort, which
combined workers from the company-specific cohorts of Dow Chemical (Ott et al., 1987; Cook,

                                          C-9

-------
1981) and the Monsanto Company (Zack and Gaffey, 1983; Zack and Suskind, 1980). These
workers were employed at 12 plants producing chemicals contaminated with TCDD.  The
production processes were assumed to be the same in all 12 plants.  Almost all workers in the
cohort (97%) had production or maintenance jobs with processes involving TCDD
contamination.  On average, workers were employed for 2.7 years in specific processes that
involved TCDD contamination, and overall, were employed for 12.6 years. Serum TCDD
samples were obtained from 253 workers (gender not specified) from two plants (selection
criteria and response rates not specified in the study).  Due to the high correlation between the
logarithm  of serum TCDD levels and the logarithm of years of exposure (Pearson  correlation
coefficient = 0.72), the study used duration of exposure as a surrogate for TCDD exposure. The
mortality follow-up began in 1940  and extended until the end of 1987.  Vital status was
determined using records from the  Social Security Administration, the Internal Revenue Service,
or the National Death Index. The ascertainment of vital status in the cohort was nearly complete,
with less than 1% of the cohort not followed up until death or the end of the study period.
Two-hundred two workers were  excluded because plant records did not show duration of
exposure,  and 67 women were excluded. No additional data were presented on study
participants to determine how representative they were of the overall study cohort. Comparisons
of mortality were made relative to the U.S. male general population and expressed using the
standardized mortality ratios (SMRs) and 95% confidence intervals (CIs). Life-table methods
were used to generate person-years of risk accrued by cohort members at each plant.
Person-years and corresponding  deaths were tabulated across age, race, and year of death strata,
which permitted the SMRs to be adjusted for the potential confounding influence from these
three characteristics. No unadjusted SMRs were presented in the paper.  The cross-classification
of person-years and deaths was also done across several exposure-related groupings, including
duration of employment, years since first exposure, years since last exposure, and  duration of
exposure.  Employment duration was categorized as <5, 5- <10, 10- <15, 15- <20, 20- <25,
25- <30, and >30 years. The variable "years since first exposure" (<10,  10- <20, and >20 years)
was used to evaluate associations for different latency periods.  The analysis was jointly
stratified by duration of employment and for varying latency intervals to  evaluate whether cohort
members with higher cumulative TCDD levels had higher cancer mortality rates than those
cohort members with lower cumulative levels.
                                         C-10

-------
       Overall, the cohort of workers had slightly elevated cancer mortality than the general
population (SMR = 1.15, 95% CI = 1.02-1.30). Comparisons to the general population,
however, yielded no statistically significant excess for any site-specific cancer. Cancer mortality
was examined for the subset of workers that worked for at least one year and had a latency
interval of at least 20 years (n = 1,520).  The 1-year cut-point was selected based on analyses of
serum levels in a subset of 253 workers which revealed that every worker employed for at least
one year had a lipid-adjusted serum level that exceeded the mean (7 ppt).  Relative to the
U.S. general population, statistically significant excesses in cancer mortality were  observed for
all cancers (SMR = 1.46, 95% CI = 1.21-1.76), cancers of the respiratory system (SMR = 1.42,
95% CI = 1.03-1.92), and for soft tissue sarcoma (SMR = 9.22, 95% CI = 1.90-26.95) among
this subset of 1,520 male workers. The elevated SMR for soft tissue sarcoma, however, was
based on only three cases  in this subset.
       SMRs also were generated across joint categories of duration of exposure and period of
latency for deaths from all cancer sites (combined), and cancer of the trachea, bronchus, and
lung.  Increased SMRs were observed in strata defined by longer duration of exposure and
latency, but no statistically significant linear trends were found.

C.I.1.1.1.1.2.  Study evaluation
       This cohort was the largest of four the International Agency  for Research on Cancer
(IARC) considered in its 1997 classification of TCDD as a Group 1  human carcinogen (IARC,
1997).  Duration of employment in processes that involved TCDD contamination was used as a
surrogate measure of cumulative exposure.  This was based on a high correlation detected
between serum TCDD levels and duration of exposure.  These 253 workers selected from
two plants each had their last exposure 15-37 years prior to evaluation. In using this exposure
metric, Fingerhut et al. (1991a) made the implicit assumption that concentrations of TCDD
exposures were equivalent at all production plants. Doses for individual cohort members were
not reconstructed for these analyses, although they were in subsequent analyses of this cohort.
       Workers in this cohort were also exposed to other  chemicals, which could have
introduced bias if these chemicals were associated with both TCDD exposure and  the health
outcomes being examined. At one plant, workers were exposed to 4-aminobiphenyl. Previous
investigators also reported that workers at another plant were exposed to 2,4,5-T and

                                          C-ll

-------
2,4-dichlorophenoxyacetic acid (2,4-D) (Bond et al.. 1989: 1988: Ottetal.. 1987). Although this
study did not examine the impact of confounding by other occupational coexposures, subsequent
analyses of this cohort showed that associations between cumulative TCDD and all cancer
mortality persisted after excluding workers exposed to pentachlorophenols from the analyses
(Steenland et al., 1999). Further, the removal of workers who died from bladder cancer did not
substantially change the dose-response relationship between TCDD and cancer mortality from all
other sites combined. This finding suggests that exposures to 4-aminobiphenyl distort the
association between cancer mortality and TCDD exposure. Overall, there is little evidence of
confounding by these coexposures among this cohort; however, exposure to other possible
confounders, such as dioxin-like compounds (DLCs), was not examined.
       The study collected no information on the smoking behaviors of the workers, and
therefore, the SMRs do not account for possible differences in the prevalence of smoking that
existed between the workers and the general population. For several reasons, however, the
inability to take into account smoking is unlikely to have been an important source of bias.  First,
mortality from other smoking-related causes of death such as nonmalignant respiratory disease
were not more common in the cohort than in the general population (SMR = 0.96,
95% CI = 0.54-1.58). Second,  stratified analyses of workers with at least a 20-year latency
(assuming this subset shared similar smoking habits) revealed that excesses were apparent only
among those who were exposed for at least 1 year.  Specifically, when compared to the general
population, the SMR among workers exposed for at least  1 year with a latency of 20 years was
1.46 (95% CI = 1.21-1.76), while those exposed for less than 1 year had an SMR of 1.02
(95% CI = 0.76-1.36). Third, for comparisons of cancer mortality between blue-collar workers
and the general population, smoking is unlikely to  explain cancer excesses of greater than
10-20% (Siemiatycki et al.,  1988). Finally, the investigators found no substantial changes in the
results  for lung cancer when risks were adjusted for smoking histories obtained in 1987  from
223 workers employed at two plants. These data were used to adjust for the expected number of
lung cancer deaths expected  in the entire cohort (Fingerhut et al., 199la). Following this
adjustment, a small change was observed in the SMR for lung cancer in the overall  cohort from
1.11 (95% CI = 0.89-1.37) to 1.05  (95% CI = 0.85-1.30). Similarly, only a slight change in the
SMR for lung cancer in the higher exposure subcohort was noted from an SMR of 1.39
(95% CI = 0.99-1.89) to 1.37 (95% CI = 0.98-1.87).
                                         C-12

-------
       The use of death certificate information from the National Death Index is appropriate for
identifying cancer outcomes. For site-specific cancers such as soft tissue sarcoma, however, the
coding of the underlying cause of death is more prone to misclassification (Percy et al., 1981).
Indeed, a review of tissues from four men concluded to have died from soft-tissue sarcoma
determined that two deaths had been misclassified (Fingerhut et al., 199la).  A review of hospital
data revealed that two other individuals had soft tissue sarcomas that were not identified by death
certificate information. The use of death certificate information to derive SMRs for cancer as a
whole is likely not subject to significant bias; the same might not hold true, however, for some
site-specific cancers such as soft tissue sarcoma.
       Using the SMR metric to compare an occupational cohort with the general population is
subject to what is commonly referred to as the "healthy worker effect" (Li and Sung,  1999; Choi,
1992).  The healthy worker effect is a bias that arises because those healthy enough to be
employed have lower morbidity and mortality rates than the general population.  The healthy
worker effect is likely to be larger for occupations that are more physically demanding
(Aittomaki et al., 2005; Checkoway et al., 1989), and the healthy worker effect is considered to
be of little consequence in the interpretation of cancer mortality (Monson, 1986; McMichael,
1976).  Few cancers are associated with a prolonged period of poor health that would affect
employability long before death. Also recognized is that, as the employed population ages, the
magnitude of the healthy worker effect decreases as the absolute reduction in mortality becomes
relatively smaller (McMichael, 1976).  The mortality follow-up of occupational cohorts
generally spans several decades, which should minimize the associated healthy worker effect in
such studies. Bias could also be introduced in that workers who are healthier might be more
likely to stay employed and therefore accrue higher levels of exposure. In the NIOSH cohort,
however, mortality was ascertained for those who could have left the workforce or retired by
linking subjects to the National Death Index.  Although internal cohort comparisons can
minimize the potential for the healthy worker effect for the reasons  presented above, for cancer
outcomes, the SMR statistic is a valuable tool for characterizing whether occupational cohort are
more likely to die of cancer than the general population. Moreover, stratified analyses across
categories of duration of exposure, or latency periods within a cohort can yield important
insights about which workers are at greatest risk. Perhaps most important, subsequent analyses
of the NIOSH cohort that presented risk estimates derived from external comparisons using the
                                          C-13

-------
SMR were remarkably consistent with rate ratios derived using an internal referent (Steenland et
al.. 1999).

C.I.1.1.1.1.3.  Suitability of data for TCDD dose-response modeling
       This cohort meets most of the identified considerations for conducting a quantitative
dose-response analysis for mortality from all cancer sites combined.  The NIOSH cohort is the
largest cohort of TCDD-exposed workers, exposure characterization at an individual level is
possible but not available in this particular study, and the follow-up period is long enough to
evaluate latent effects. Although there is no direct evidence of any important source of bias,
confounding may be present due to a lack of consideration of DLCs.  For the purpose of
quantitative dose-response modeling, it is important to note that subsequent studies of this cohort
adopted methods that greatly improved the characterization of TCDD exposure in the NIOSH
cohort and increased the follow-up interval (Cheng et al., 2006; Steenland et al., 2001b). As
such, for all practical purposes, due consideration for dose-response modeling should focus on
the more recently developed data sets.
       For quantitative dose-response modeling for individual cancer sites, the data are much
more limited. A statistically significant positive association with TCDD was noted only for
soft-tissue sarcoma among those with more than 1 year of exposure and 20 years of latency
(SMR = 9.22, 95% CI = 1.90-26.95).  However there were only three deaths from soft tissue
sarcoma among this exposed component of the cohort, and four deaths in total in the overall
cohort. Also, misclassification of outcome for soft-tissue sarcoma through death registries is
well recognized and supported with additional review of tissue from two of the men.
Specifically, tissues from the four men who died of soft-tissue sarcoma revealed that only two of
these cases were coded correctly.
       Although subsequent analyses of the NIOSH cohort did not show evidence of
confounding by other occupational exposures, the design of this initial publication of the NIOSH
cohort did not allow for examination of exposures to other possible confounders, such as DLCs.
Duration of exposure was used as a surrogate for cumulative TCDD exposure; therefore,
effective doses could not be estimated. Therefore, dose-response modeling was not conducted
for this study.
                                          C-14

-------
C.l.1.1.1.2.  Steenland et al (1999)
C.I.1.1.1.2.1.  Study summary
       A subsequent analysis of the NIOSH cohort extended the follow-up interval of Fingerhut
et al. (199la) by 6 years (i.e., from 1940-1993) and improved the characterization of TCDD
exposure (Steenland et al., 1999).  A key distinction from the work of Fingerhut et al. (199 la)
was the exclusion of several workers that had been included in the previous mortality analyses.
The authors excluded 40 workers who were either female, had never worked in TCDD-exposed
departments, or had missing date of birth information.  An additional 238 workers were excluded
as occupational data for characterizing duration of exposure were lacking, preventing their use in
a subcohort dose-response analysis. This subcohort was further reduced by excluding workers
from four plants (n = 591) because the information on the degree of TCDD contamination in
work histories was  limited, preventing the characterization of TCDD levels by job type.
Thirty-eight additional workers were excluded from the eight remaining plants because TCDD
contamination could not be estimated. Finally, 727 workers were excluded because they had
been exposed to pentachlorophenol. Exposures were assigned to 3,538  (69%) male members of
the  overall cohort, a population substantially reduced from the 5,172 on which Fingerhut et al.
(1991a) reported. Steenland et al. (1999) also evaluated the mortality experience of a subcohort
of 608 workers with chloracne who had no exposure to pentachlorophenol.
       For each worker, a quantitative exposure score for each day of work was calculated based
on the concentration of TCDD (ug/g) present in process materials, the fraction of the day
worked, and a qualitative contact level based on estimates of the amount of TCDD exposure via
dermal absorption or inhalation. The authors derived a cumulative measure of TCDD exposure
by summing the exposure scores across the working lifetime history for each worker. The
authors validated this cumulative exposure metric indirectly by comparing values obtained for
workers with and without chloracne. Such a validation is appropriate, given that chloracne is
considered a clinical sign of exposure to high doses of dioxin (Ott et al., 1993). The median
exposure score among those with chloracne was 11,546 compared with  77 among those without
(Steenland and Deddens. 2003).
       Cancer mortality was compared using two approaches.  As in Fingerhut et al. (1991a),
external comparisons were made to the U.S. general population using the SMR statistic.  The
authors adjusted the SMR statistics for race, age,  and calendar time.  They also applied life-table

                                         C-15

-------
methods to characterize risks across the subcohort of 3,538 workers with exposure data by
categorizing the workers into seven cumulative exposure groups.  The cut-points for these
categories were selected so that the number of deaths in each category was nearly equal to
optimize study power. Life-table analyses were extended further to consider a 15-year lag
interval, which in a practical sense means that person-years at risk would not begin to accrue
until 15 years after the first exposure occurred. The person-years and deaths that occurred in the
first 15 years were included in the lowest exposure grouping. The Cox proportional  hazards
model was used to characterize risk within the cohort.  Cox regression was used to provide an
estimate of the hazard ratios and the 95% CIs for ischemic heart disease, all cancers combined,
lung cancer, smoking related cancers, and all other cancers.  The authors also performed Cox
regression analyses using the seven categories of exposure, adjusting the regression coefficients
for both year of birth and age. The regression models were run for both unlagged and lagged
(15 years) cumulative exposure scores.
       Overall, when compared with the U.S. general population, a slight excess of cancer
mortality (from all sites) was noted in the 5,132 cohort study population (SMR = 1.13,
95% CI = 1.02-1.25).  This result did not substantially differ from the earlier finding that
Fingerhut et al. (199la) published (SMR =1.15, 95% CI = 1.03-1.30). Site-specific analyses
revealed statistically significant excesses relative to the U.S. general population for bladder
cancer (SMR = 1.99, 95% CI = 1.13-3.23) and for cancer of the larynx (SMR = 2.22,
95% CI = 1.06-4.08).  In the chloracne subcohort (n = 608), SMRs of 1.25
(95% CI = 0.98-1.57) and 1.45 (95% CI = 0.98-2.07) were found for all cancer sites and for
lung cancer, respectively, relative to the general population. The authors also found statistically
significant excesses for connective and soft tissue sarcomas (SMR = 11.32,
95% CI = 2.33-33.10) and for lymphatic and hematopoietic malignancies (SMR = 3.01,
95% CI= 1.43-8.52).
       External comparisons made by grouping workers into septiles of cumulative TCDD
exposure and generating an  SMR for each septile using the U.S. population as the referent group
suggested a dose-response relationship. For all cancer sites combined, workers in the highest
exposure score category had an SMR of 1.60 (95% CI = 1.15-1.82); increases also were
observed in the sixth (SMR = 1.34) and fifth (SMR = 1.15) septiles.  The two-sided/>-value
associated with the test for trend for cumulative TCDD exposure was statistically significant
                                          C-16

-------
(p = 0.02).  A similar approach for lung cancer revealed virtually the same pattern. The
incorporation of a 15-year latency for the analyses of all cancer deaths, in general, produced
slightly higher SMRs across the septiles, although a slight attenuation of effect was noted in the
highest septile (SMRuniagged = 1.60 vs. SMRiagged = 1.54). For a 15-year lag, the lung cancer
SMRs were mixed compared to the unlagged results with some septile exposure categories
increasing and others decreasing relative to the lowest exposure group.
       For the internal cohort comparisons using Cox regression analyses, higher hazard ratios
were found among workers in the higher exposure categories than those in the lowest.  The linear
test for trend, however, was not statistically significant (p = 0.10). The associations across the
septiles for the unlagged exposure for the internal cohort comparisons were not as strong as for
the  external cohort comparisons.  The opposite was true, however, for cumulative exposures
lagged 15 years.
       Relative to the lowest septile, stratified analyses revealed  increased hazard ratios in the
upper septiles of the internal cohort comparisons for both smoking- and nonsmoking-related
forms of cancer. The test for linear trend was statistically significant for all other cancers (after
smoking-related cancers were excluded). These analyses suggest that the overall cancer findings
were not limited to an interaction between TCDD and smoking. Additional sensitivity analyses
by the authors indicated the findings for smoking-related cancers were largely unaffected by the
exclusion of bladder cancer cases. This observation suggests that exposure to 4-aminobiphenyl,
which occurred at one plant and might have contributed to an increased number of bladder
cancers, did not substantially bias the relationship between TCDD and all cancers combined.
       The investigators also evaluated the dose-response relationship with a Cox regression
model separately for each plant using internal cohort comparisons and found some heterogeneity.
This finding is not unexpected particularly given the relatively small number of cancer deaths at
each plant,  and given that exposures were quite low for one plant at which no positive
association was found. The variability among plants was taken into account by modeling plant
as a random effect measure in the Cox model, which produced little change in the slope
coefficient (P = 0.0422 vs. P = 0.0453, respectively).
                                          C-17

-------
C. 1.1.1.1.2.2.  Study evaluation
       This study represents a valuable extension from that published by Fingerhut et al.
(1991a).  Internal comparisons were performed to help minimize potential biases associated with
using an external comparison group (e.g., healthy worker effect, and differences in other risk
factors between the cohort and the general population). That similar dose-response relationships
were found for internal and external comparison populations suggests that the bias due to the
healthy worker effect in the cohort is minimal for cancer mortality. More importantly, the
construction of the cumulative exposure scores provides an improved opportunity to evaluate
dose-response relationships compared with the length of exposure and duration of employment
metrics that Fingerhut et al. (199la) used.
       A potential limitation of the NIOSH study was the inability to account for cigarette
smoking.  If cigarette smoking did contribute to the increased cancer mortality rates in this and
other cohorts, increased cancer mortality from exposure to TCDD would be expected only for
smoking-attributable cancers. This study found associations with TCDD for both smoking- and
nonsmoking-related cancers, including a stronger association for nonsmoking-related cancers.
Therefore, the data provide evidence that associations between TCDD and cancer mortality are
not likely due to cigarette smoking.
       The findings regarding latency should be interpreted cautiously as the statistical power in
the study to compare differences across latency intervals was limited. Caution also should be
heeded, given that latency intervals can vary on an individual basis as they are often
dose-dependent (Guess and Hoel 1977). The evaluation of whether TCDD acts as either an
initiating or promoting agent (or both) is severely constrained by the reliance on cancer mortality
data rather than incidence data. This constraint is due to the fact that survival time can be quite
lengthy and can vary substantially across individuals and by cancer subtype. For example, the
5-year survival among U.S. males for all cancer sites combined ranged between 45 and 60%
(Clegg et al., 2002). When only mortality data are available, evaluating the time between when
individuals are first exposed and when they are first diagnosed with cancer is nearly impossible.
       Starr (2003) suggested that Steenland et al. (1999) focused too heavily on the exposures
that incorporated a 15-year period of latency and that those who experienced high exposures
would inappropriately contribute person-years to the lowest exposure group "irrespective of how
great the workers' actual cumulative exposure scores may have been."  Most cancer deaths

                                          C-18

-------
would, however, typically occur many years postemployment. Given that the follow-up interval
of the cohort was lengthy and the average exposure duration was 2.7 years, at the time of death,
person-years for those with high cumulative exposures would be captured appropriately.  The
median 5-year survival for all cancers is approximately 50% (Clegg et al., 2002), so applying a
minimum latency of 5 years when using cancer mortality rather than cancer incidence data is
needed to assure that the exposure metric captures exposures before diagnosis.  Increasing this
latency period, for example to 10 or 15 years, would eliminate consideration of exposures that
occur in the period between tumor occurrence and tumor detection (diagnosis), and allows for an
appropriate focus on exposures that act either early or late in the pathogenic process. If the
association of TCDD with cancer is causal, effects might become apparent only at high
exposures and with adequate latency.  As such, IARC has concluded that a latency interval of
15 years could be too short (IARC. 1997).  EPA considers the Steenland et al. (1999)
presentation to be balanced in that they provided the range in lifetime excess risk estimated
across the various models used.  The authors' finding that the models with a 15-year lag
provided a statistically significant improvement in fit based on the chi-square test statistic should
not be readily dismissed.

C. 1.1.1.1.2.3.  Suitability of data for TCDD dose-response modeling
       This study meets most of the epidemiologic considerations for conducting a quantitative
dose-response analysis for mortality from all cancer sites combined.  This study excludes a large
number of workers who were exposed to pentachlorophenol, thus eliminating the potential for
bias from this exposure.  Relative to the earlier study by Fingerhut et al. (199la), improvements
were made to the methodology applied to assign TCDD exposures to the workers. This study,
however, is superseded by Steenland et al.  (200Ib), who provide a more detailed presentation
and modeling of the NIOSH cohort data. Therefore, dose-response modeling was not pursued
for this study, but was for the subsequent NIOSH study by Steenland et al. (2001b).

C.l.1.1.1.3. Steenland et al (2001b)
C.I.1.1.1.3.1.  Study summary
       In 2001, Steenland et al. (200Ib) published a risk analysis using the NIOSH cohort that,
for the first time, incorporated serum measures in the derivation of TCDD exposures for

                                         C-19

-------
individual workers.  The authors applied the same exclusion criteria to the entire cohort of
workers across the 12 plants in the Steenland et al. (1999) study, leaving 3,538 male workers for
which risk estimates could be calculated.  Unlike previous analyses of the NIOSH cohort that
considered several different mortality outcomes, the analyses presented in Steenland et al.
(200 Ib) focused exclusively on mortality from all cancers sites combined. The authors observed
256 cancer deaths in the cohort between 1942 and the end of 1993.  All risks estimated in the
Steenland et al. (200 Ib) study were based on internal cohort comparisons.
       Characterization of TCDD exposure levels among the workers was based on serum
measures obtained in 1988 from 199 workers who were employed in one of the eight plants.
Only those workers with both TCDD serum measures and previously developed exposure scores
(Steenland et al., 1999) were used to estimate the relation between these different exposure
metrics. Based on these findings, cumulative TCDD serum levels were estimated on an
individual basis for all 3,538 workers following restriction to a subset of 170 workers whose
1988 serum measures were greater than the upper range of background levels (10 ppt) (Steenland
etal.. 2001b).
       The authors developed a regression model estimated the level of TCDD at the time of last
exposure for the 170 workers.  The model was based on the estimated half-life of TCDD, the
known work history of each worker, a pharmacokinetic model for the storage and excretion of
TCDD, and exposure scores for each job held by each worker over time.  The resulting  equation
follows:
sure = ^1988 CXp(kAt)                            (Eq. C-l)
                            iast exposu
       The first-order elimination rate constant (X) was based on a half-life of 8.7 years
previously reported for the Ranch Hands cohort (Michalek et al., 1996). The background rate of
TCDD exposure was assumed to be 6.1 ppt, which was based on the median level in a sample of
79 unexposed workers in the NIOSH cohort (Piacitelli etal., 1992). This value was subtracted
when TCDD values were back-extrapolated, and then added again after the back-extrapolation
was completed. A background level of 5 ppt also was used in some of the analyses with minimal
demonstrable effects on the results.  Sensitivity analyses also were incorporated to consider a
                                         C-20

-------
7.1-year half-life estimate that had been developed for the earlier Ranch Hands study (Pirkle et
al.. 1989).
       After back-extrapolating to obtain TCDD serums levels at the time of last exposure, the
investigators estimated cumulative (or "area under the curve") TCDD serum levels for every
cohort member. This estimation procedure was the same method Flesch-Janys et al. (1998)
applied to the Hamburg cohort to derive a coefficient for relating serum levels to exposure
scores. The "area under the curve" approach integrates time-specific serum levels over the
employment histories of the individual workers.  The slope coefficient was estimated using a
no-intercept linear regression model. This model is based on the assumption that a cumulative
score of zero is associated with no serum levels above background.
       Cox regression was also used to model the continuous measures of TCDD.  A variety of
exposure metrics were considered that took into  account different lags, nonlinear relationships
(e.g., log-transform and cubic spline), as well as  threshold and nonthreshold exposure metrics.
Categorical analyses were used to evaluate risks  across TCDD exposure groups, while different
shapes of dose-response curves were evaluated through the use of lagged and unlagged
continuous TCDD measures. Categorical analyses of TCDD exposure were conducted using the
Cox regression model to derive estimates of relative risk (RR) as described by hazard ratios and
95% CIs. The reference group in this analysis was those workers in the lowest septile
cumulative exposure grouping (<335 ppt-years). The septiles were chosen based on cumulative
serum levels that considered no lag and also a 15-year lag.
       The investigators also conducted dose-response analyses using the toxicity equivalence
(TEQ) approach. The TEQ is calculated as the sum of all exposures to dioxins and furans
weighted by the potency of each specific compound.  In this study, TCDD was assumed to
account for all dioxin exposures in the workplace.  For background TEQ levels, the investigators
used a value of 50 ppt in the dose-response modeling. This is based on the assumption that
TCDD accounted for 10% of the toxicity of all dioxins and furans  (WHO, 1998), and is
equivalent to using a background level of 5 ppt/yr that was used in the derivation of cumulative
serum TCDD levels. A statistically  significant dose-response pattern was observed for all cancer
mortality and TCDD exposure based on log of cumulative TEQs with a 15-year lag. A
comparison of the overall model chi-square values indicated that the fit of this model was not as
good as that for TCDD.
                                         C-21

-------
       The hazard ratios among workers grouped by categories of cumulative TCDD exposure
(lagged 15 years) suggested a positive dose-response relationship.  Steenland et al. (200Ib)
found statistically significant excesses in the higher exposure categories compared to the lowest
septile.  The RR was 1.82, (95% CI = 1.18-2.82) for the sixth septile (7,568-20,455 ppt-years)
and 1.62, (95% CI = 1.03-2.56) for the seventh septile (>20,455 ppt-years).  Cox regression
indicated that log TCDD serum concentrations (lagged 15 years) was positively associated with
cancer mortality (P = 0.097, standard error [P] = 0.032, p < 0.003). A statistically significant
improvement in fit was observed when a 15-year lag interval was incorporated into the model
                                            r\
compared to a model with no such lag (Model % with 4 degrees of freedom = 7.5).  Results were
similar when using a half-life of 7.1 years rather than 8.7 years. The excess lifetime risk of death
from cancer at age 75 for TCDD intake (per 1.0-picogram per kilogram [pg/kg] of body weight
[BW] per day) was about 0.05-0.9% above a background lifetime risk of cancer death of 12.4%.
The results from the best-fitting models provide lifetime risk estimates within the ranges derived
using data from the Hamburg cohort (Becher et al., 1998).
       In both categorical and continuous analyses of TCDD based on a linear model, the
dose-response pattern tailed off at high exposures suggesting nonlinear effects.  This
phenomenon could be due to saturation effects (Stayner et al.,  2003) or, alternatively, could have
resulted from increased exposure misclassification of higher exposures (Steenland et al.,  2001b).
Specifically, some of the highest exposures might have been poorly estimated as they occurred in
workers exposed to  short-term high exposures during the clean-up of a spill. The choice of a
linear model to develop data from a single time point can also result in exposure
misclassification in those individuals that have differences in the length of exposure (Emond  et
al., 2005). Misclassification would be less likely at low concentrations where dose-dependent
elimination is minimal.

C. 1.1.1.1.3.2.   Study evaluation
       An important consideration in the Steenland et al. (200Ib) study was the use of a small
subset of workers (n = 170) to infer exposures for the remainder of the cohort. Although there is
limited information  in the study to determine how representative the 199 workers were of the
overall workers in that plant, the authors  report that exposures from the plant in which these
170 subjects worked were in the middle of the exposure distribution of the eight U.S. chemical

                                          C-22

-------
plants the authors had previously studied.fSteenland et al., 1999)  This subset did comprise
surviving members of the cohort (in 1988), and therefore, the frequency distribution of their year
of birth would have differed from the rest of the cohort. Furthermore, these workers were
employed at a single plant that had less detailed work histories than the other plants; thus, the
development of the exposure scores differed between this plant and the others.  Also, many of
the workers at this plant had the same job title and were employed during the same calendar
period. The use of serum data from this subset adds a level of uncertainty that is not readily
characterized.  The study report only states that the serum levels were available for these
individuals, but it does not provide any indication of how or why the individuals were selected
for serum evaluation or if there were a number of individuals that declined to give samples.
Thus, it is hard to gauge how representative this population is of the plant cohort.  Despite these
limitations, the use of these sera data to derive cumulative measures for all cohort workers seems
warranted given the strong correlation observed between the exposure scores, and TCDD serum
levels estimates at the time of last exposure (Spearman r = 0.90).
       The authors performed an extensive series of sensitivity analyses and considered several
alternative exposure metrics to the simple linear model. The lifetime excess risk above
background was nearly twice as high for the log cumulative serum measures with  a 15-year lag
when compared to the piecewise linear models with no lag.  An important observation was that
the exposure metric based on cumulative serum (lagged 15 years) did not fit the data as well as
the cumulative exposure score used in earlier  analyses (Steenland et al., 1999).  A priori, one
would expect that a better fit would be obtained with serum-based measures because serum
provides  a better measure of relevant biological dose. As the authors noted, inaccuracies
introduced in estimating the external-based exposure scores could have contributed to a poorer
fit of the  data.  Alternatively, exposure misclassification error could be introduced if serum
samples based on the 170 workers were not representative of the entire cohort.  Although the
serum-based measures did not fit the data as well as the exposures scores, the authors regarded
them as providing a reasonable fit based on an improvement in log likelihood of 3.99 (between
the log cumulative serum  model and the log cumulative exposure score model).  Moreover, the
serum-based measures enabled better characterization of risk in units (pg/kg-day) that can be
used in regulating exposures.
                                          C-23

-------
C. 1.1.1.1.3.3.  Suitability of data for TCDD dose-response modeling
       This study meets all of the epidemiologic considerations for conducting a quantitative
dose-response analysis for mortality from all cancer sites combined. As mentioned previously,
the NIOSH cohort is the largest assembled to date for which TCDD-related risks of cancer
mortality can be estimated.  The use of serum-based measures provides an objective measure of
TCDD exposure. Repeated measures in other study populations have provided reasonable
estimates of the half-life of TCDD, which permitted exposures to be back extrapolated in this
cohort.
       The authors have made extensive efforts to evaluate a wide variety of nonlinear and
linear models with varying lengths of latency and log transformations. The model chi-square test
statistics were fairly similar for the log cumulative serum (15-year lag) (Model X,2(4df) = 11.3)
model and the piecewise linear model (no lag) (Model X,2(5df> = 12.5).  These models, however,
produced results with twofold differences in lifetime excess risks.  These differences underscore
the importance of characterizing uncertainty in modeling approaches when conducting
dose-response analysis.
       The Steenland et al.  (200Ib) study characterizes risk in terms of pg/kg of BW per day.
Given that tolerable daily intake dioxin levels are typically expressed in pg/kg of BW (WHO,
1998), the presentation of risks using these units is an important advance from the earlier
analyses that used exposure scores (Steenland et al., 1999).  Many of the Steenland et al. (200 Ib)
findings are consistent with earlier work from this cohort, which is not surprising given that
exposures scores were used to derive serum-based levels for the cohort. The findings of excess
lifetime risks obtained for the best- fitting model are also consistent with those derived from the
Hamburg cohort (Becher et al., 1998). This study meets the epidemiologic considerations noted
previously as there is no evidence that the study is subject to bias from confounding due to
cigarette smoking or other occupational exposures. Given the considerable efforts to measure
effective dose to TCDD among the study participants, this study also meets the requisite
dose-response modeling criteria and will be used in quantitative dose-response analyses of
cancer mortality.
                                          C-24

-------
C.l.1.1.1.4.  Chens et al (2006)
C.I.1.1.1.4.1.  Study summary
       Cheng et al. (2006) undertook a subsequent quantitative risk assessment of 3,538 workers
in the NIOSH cohort using serum-derived estimates of TCDD.  This dose-response analysis was
published after the 2003 Reassessment document was released. The goal of this study was to
examine the relationship between TCDD and cancer mortality (all sites combined) using a new
estimate of dose that estimated TCDD as a function of both exposure intensity and age using a
kinetic model.  This physiologically-based pharmacokinetic model has been termed the
"concentration- and age-dependent elimination model" (CADM) and was developed by Aylward
et al. (2005b).  This model describes the kinetics of TCDD following oral exposure to humans by
accounting for key processes affecting kinetics by simulating the total concentration of TCDD
based on empirical  consideration of hepatic processes (see Section 3.3).  An important feature of
this kinetic model is that it incorporates concentration- and age-dependent elimination of TCDD
from the body; consequently, the effective half-life of TCDD elimination varies based on
exposure history, body burden, and age of the exposed individuals.  The study was motivated by
the reasoning that back-calculations of TCDD using a first-order elimination model and a
constant half-life of 7-9 years  underestimated exposure to TCDD among workers. This
underestimate, in turn, would result in overestimates of the carcinogenic potency of TCDD.
       As with the earlier Steenland et al. (200Ib) analyses, the cohort follow-up period was
extended from 1942 until the end of 1993 and work histories were linked to a job exposure
matrix  to obtain cumulative TCDD scores. Two cumulative serum lipid exposure metrics (in
ppt-years) were constructed using the data obtained from the sample of 170 workers. The first
replicated the metric used in a  previous  analysis of the cohort (Steenland et al., 200 Ib) and was
based on a first-order elimination model with an 8.7-year half-life (Michalek et al., 1996).  The
second metric was based on CADM and had two first-order elimination processes (Aylward et
al., 2005a). This metric assumes that the elimination of TCDD in humans occurs at a faster rate
when body concentrations are high and  at slower rates in older individuals (Avlward et al.,
2005a:  2005b).  The model was optimized using individuals for which serial measures of serum
TCDD were available. These measures were obtained from 39 adults with initial serum levels
between 130 and 144,000 ppt (Aylward et al., 2005b). This group included 36 individuals who
had been exposed in the Seveso accident and 3 exposed in Vienna, Austria. In practice, for

                                         C-25

-------
serum levels greater than 1,000 ppt, the effective half-life would be less than 3 years, and for
serum TCDD levels less than 50 ppt, the effective half-life would be more than 10 years
(Aylward et al., 2005b). Results from the model indicate that men eliminate TCDD faster than
women do as demonstrated previously by Needham et al. (1994). These age- and
concentration-dependent processes were assumed to operate independently on TCDD in hepatic
and adipose tissues, and TCDD levels in liver and adipose tissue were assumed to be a nonlinear
function of body concentration.  Cheng et al. (2006) calibrated CADM using a dose of 156 ng
per unit of exposure score and assumed a background exposure rate of 0.01 ng/kg-month. The
average TCDD ppt-years derived from CADM with a 15-year lag was 4.5-5.2 times higher than
with the first-order elimination model.  The two metrics, however, were highly correlated based
on a Pearson correlation coefficient of 0.98 (p < 0.001). Comparisons of fit between the CADM
and first-order elimination model were made using R2 values and presented in Aylward et al.
(2005b).
       Cheng et al. (2006) compared the mortality experience of NIOSH workers to the U.S.
general population using the SMR statistic. SMR statistics  also were generated separately for
each of the  8 plants and for all plants combined. Cox regression models were used to analyze
internal cohort dose response.  These models used age as the time variable, and penalized
smoothing spline functions of the CADM metric also were considered.  The possible
confounding effects of other occupational exposures and other regional population differences
were assessed by repeating analyses after excluding one plant at a time. Lagged and unlagged
TCDD exposures were analyzed separately, and stratified analyses allowed risk estimates to be
compared between smoking- and nonsmoking-related cancers. Cheng et al. (2006) adjusted the
slope estimates derived from the Cox model for the potential confounding effects  of race and
year of birth.
       Overall, a statistically significant excess in all  cancer mortality in the cohort occurred
relative to the general population (SMR =1.17, 95% CI = 1.03-1.32). The plant-specific SMRs
ranged from 0.62-1.87, with a statistically significant excess evident only for plant 10
(SMR = 1.87, 95% CI = 1.35-2.52). For lung cancer mortality, the overall SMR was not
statistically significant (SMR = 1.11, 95% CI = 0.89-1.37). A statistically significant excess of
lung cancer also was found for plant 10 (SMR = 2.35, 95%  CI = 1.44-3.64). The SMRs between
                                         C-26

-------
smoking- (SMR = 1.22, 95% CI = 1.01-1.45) and nonsmoking-related cancers (SMR =1.12,
95% CI = 0.94-1.33) were similar.
       For the internal cohort analyses of serum-derived measures, the authors were able to
replicate the one-compartmental model used previously (Steenland et al., 2001b). As had been
noted by Steenland et al. (200 Ib), an inverse-dose-response pattern was seen for individuals with
high exposures (above 95th percentile); this type of pattern is frequently observed in occupational
studies (Stayner et al., 2003). Excluding these data produced a stronger association between
TCDD and all-cancer mortality. In fact, only when the upper 2.5% or 5% of observations was
removed did a statistically significant positive association become evident with the
untransformed, unlagged data. Similarly, when the model incorporated a lag of 15 years, a
statistically significant association was noted only for the untransformed TCDD ppt-years with
the upper 5% of observations removed. Stratified analyses revealed little difference in the
association between TCDD and smoking- and nonsmoking-related cancers, and the removal of
one plant at a time from the analyses of TCDD ppt-years changes did not substantially change
the slope.

C.I.1.1.1.4.2.  Study evaluation
       The authors reported that CADM provided an improved fit over the one-compartmental
model, but presented no evidence  regarding any formal test of statistical significance. A
comparison ofR2 values presented in Aylward et al. (2005b), however, does reveal that the R2
value increased from 0.27 (first-order compartmental model with an 8.7-year half-life) to 0.40
for CADM.  TCDD exposures estimated using CADM were approximately fivefold higher than
the one-compartmental model estimates among cohort members with higher levels of exposure.
Differences in exposure estimates between the two metrics were less striking among individuals
with lower TCDD exposures. The net effect was that CADM produced a 6- to 10-fold decrease
in the estimated risks compared to those previously reported (Steenland et al., 2001b).
Nonetheless, the estimates produced by CADM span more than two orders of magnitude under
various assumptions. Further uncertainties arise from between-worker variability of TCDD
elimination rates, possible residual confounding, and the variability associated with the use of
data obtained from other cohorts.  Nevertheless, the use of the CADM to estimate TCDD
                                         C-27

-------
exposure is considered a significant advantage over the previous first-order body burden
calculations.

C.I.1.1.1.4.3.  Suitability of data for TCDD dose-response modeling
       The value of including the NIOSH cohort data has already been established based on
investigations by Steenland et al. (2001b: 1999). The decision to include data from the
quantitative dose-response analysis by Cheng et al. (2006) relates to the added value that the
CADM exposure estimates would provide.  The earlier modeling work of Aylward et al. (2005b)
provided some support for a modest improvement of the fit of CADM over the first-order
compartmental model, and they also confirmed previous studies that found that TCDD
elimination rates varied by age and sex.  Recent work by Kerger et al. (2006) also demonstrates
that the half-life for TCDD is shorter among Seveso children than in adults, and that body
burdens influence the elimination of TCDD in humans. That estimates of half-lives among men
have been remarkably consistent, with mean estimates ranging between 6.9 and 8.7 years
(Needham et al.. 2005: Michalek et al.. 2002: Flesch-Janys et  al..  1996: PirkleetaL 1989).
however, is noteworthy. Based on the underlying strengths of the NIOSH cohort data and efforts
by Cheng et al. (2006) to improve estimates of effective dose, these data support further
dose-response modeling.

C.l.1.1.1.5. Collins et al (2009)
C.I.1.1.1.5.1.  Study summary
       In a recent study, Collins et al. (2009) investigated the relationship between serum TCDD
levels and mortality rates in a cohort of trichlorophenol workers (gender not specified) exposed
to TCDD. These workers were part of the NIOSH cohort having accounted for approximately
45% of the person-years in an earlier analysis (Bodner et al., 2003).  The investigators completed
an extensive dioxin serum evaluation of workers employed by the Dow Chemical plant in
Midland, Michigan, that made 2,4,5-trichlorophenol (TCP) from 1942 to 1979 and 2,4,5-T from
1948 to 1982. Collins et al. (2007) and Aylward et al. (2007)  developed historical TCDD
exposure estimates for all TCP and 2,4,5-T workers.  This study represents the largest group of
workers from a single plant ever studied for the health effects  of TCDD.  Little information on
how vital status was ascertained, was provided in this paper or in the Bodner et al. (2003) report

                                         C-28

-------
of mortality in this cohort.  Although the authors indicate that death certificates were obtained
from the states in which the employees died, it is unclear whether vital status was ascertained
from company records or through record linkage to the National Death Index is unclear.
       The follow-up interval for these workers spanned the period between 1942 and 2003.
Thus, the study included 10 more years of follow-up than earlier investigations of the entire
NIOSH cohort.  Serum samples were obtained from 280 former workers (selection criteria
including data on gender were not specified) in 2004-2005.  A simple one-compartment first-
order pharmacokinetic model and elimination rates as estimated from the BASF cohort were
used (Flesch-Janys etal., 1996).  The "area under the curve" approach was used to characterize
workers' exposures over the course of their working careers and provided a cumulative measure
of exposure. Analyses were performed with and without 165 of the 1,615 workers exposed to
pentachlorophenol to evaluate the impact of these exposures.
       External comparisons of cancer mortality rates to the general U.S. population were made
using SMRs.  Internal cohort comparisons of exposure-response relationships were made using
the Cox regression model.  This model used age as the time variable, and was adjusted for year
of hire and birth year. Only those causes of death for which an excess was found based on the
external comparisons or for which previous  studies had identified a positive association were
selected for dose-response analyses.
       A total of 177 cancer deaths were observed in the cohort. For the external comparison
with the U.S. general population, overall,  no statistically significant difference was observed in
all cancer mortality among all workers (SMR = 1.0, 95% CI = 0.8-1.1). Results obtained after
excluding workers exposed to pentachlorophenol were similar (SMR = 0.9, 95% CI = 0.8-1.1).
Excess mortality in the cohort was found for leukemia (SMR = 1.9, 95% CI = 1.0-3.2) and soft
tissue sarcoma (SMR = 4.1, 95% CI = 1.1-10.5).  Although not statistically significant SMRs for
other lymphohemopoietic cancers included non-Hodgkin lymphoma (SMR = 1.3, 95% CI = 0.6,
2.5) and Hodgkin disease (SMR = 2.2, 95% CI = 0.2, 6.4).
       Internal cohort comparisons using the Cox regression model were performed for all
cancers combined, lung cancer, prostate cancer, leukemia, non-Hodgkin lymphoma, and
soft-tissue sarcoma.  Whether the internal comparisons excluded those workers exposed to
pentachlorophenol is not entirely clear from the text or accompanying table, but presumably they
do not.  The RR was 1.002 (95% CI = 0.991-1.013) for all cancer mortality per  1 ppb-year

                                         C-29

-------
increase in cumulative TCDD exposure was not statistically significant. Except for soft tissue
sarcomas, no statistically significant exposure-response trends were observed for any cancer site.
For soft tissue sarcoma, analyses were based on only four deaths.

C.I.1.1.1.5.2. Study evaluation
       A key limitation of this study is that SMRs were not derived for different periods of
latency for the external comparison group analysis.  The original publication on the NIOSH
cohort found that SMRs increased when a 20-year latency period was incorporated (Fingerhut et
al., 1991a), and similar patterns have been observed in other occupational cohorts (Ott and
Zober, 1996a: Manz et al., 1991) and among Seveso residents (Consonni et al., 2008).
Additionally, dose-response analyses showed marked increases in slopes with a 15-year latency
period (Cheng et al., 2006; Steenland and Deddens,  2003). In this context, the absence of an
elevated SMR for cancer mortality is consistent with previous findings of the NIOSH cohort.
Additional analyses published subsequently (Collins et al., 2010) found no excess cancer
mortality in the cohort relative to the general population when a latency period of 20 years was
applied (SMR = 1.0, 95% CI = 0.8-1.1).
       Unfortunately, the Collins  et al. (2009) study did not include a categorical analysis of
TCDD exposure and cancer mortality.  This categorical analysis would have enabled an
evaluation of whether a nonlinear  association exists between TCDD exposure and cancer risk.
The analyses of both Cheng et al. (2006) and Steenland et al. (200 Ib) suggest an attenuation of
effects at higher doses, and several investigations have considered log-transformed associations
as a means to address nonlinearity. Also, the earlier plant-specific dose-response analyses of
Steenland et al. (200Ib) are not consistent with the findings for the Midland plant that Collins
et al. (2009) presented. In response to the letter by Villeneuve and Steenland (2010) that
highlighted the value of characterizing risk across categories of TCDD exposure, Collins et al
(2010) reported SMRs across three cumulative exposure levels of  0.1-374.9, 375.0-1,999.9, and
2,000-112,253 ppt-month categories. No excess cancer mortality, as captured by the SMR, was
observed in any of the three exposure categories for analyses conducted with no latency and a
20-year latency.  Given that excesses were not noted in the NIOSH cohort until approximately
14,000 ppt-months, the upper exposure grouping (2,000-112,253 ppt-months) used by Collins
et al. (2010) may not be able to differentiate possible associations  at higher exposure levels.

                                          C-30

-------
C.I.1.1.1.5.3.  Suitability of data for dose-response modeling
       The Collins et al. (2009) study used serum levels to derive TCDD exposure estimates and
does not appear to be subject to important biases. The reliance on data from one plant offers
some advantages over the multiplant analyses, as heterogeneity in exposure to other occupational
agents would be lower.  The number of individuals who provided serum samples (n = 280) is
greater than the 170 individuals used to derive TCDD estimates for the NIOSH cohort, but there
was no information presented in either study to assess how representative subjects who provided
samples were of the larger cohort.  The authors found a statistically significant dose-response
trend for soft tissue sarcoma mortality and TCDD exposures. Therefore, this study is considered
suitable for quantitative dose-response analysis.

C.l.1.1.2.  The BASF cohort
       In 1953, dioxin contamination occurred as a result of an autoclave accident during the
production of trichlorophenol at the BASF plant in Ludwigshafen, Germany. A second dioxin
incident occurred in 1988 that was attributed to the blending of thermoplastic polyesters with
brominated flame retardants.  Of the two events, the one on November 13,  1953, was associated
with more severe acute health effects,  including chloracne that resulted in immediate
hospitalizations for seven workers.  These adverse events were not linked to TCDD until  1957
when TCDD was identified as a byproduct of the production of trichlorophenol and was shown
to induce chloracne (Zober et al., 1994). Zober and colleagues (1998) noted that with the 1988
accident, affected individuals did not exhibit clinical symptoms or chloracne, but rather were
identified through "analytical measures."  In both instances, efforts were made to limit the
potential for exposure to employees.

C.l.1.1.2.1. Thiess and Frentzel-Beyme (1977) and Thiess et al (1982)
C.I.1.1.2.1.1.  Study summary
       A study of the mortality of workers employed at the BASF plant was first presented in
1977 (Thiess and Frentzel-Beyme, 1977)  with subsequent updates in both 1982 (Thiess et al.,
1982). and in 1990 (Zober etal.. 1990). In the first published paper (Thiess et al.. 1982).
74 employees involved in the 1953 accident were traced and their death certificate information
extracted.  Of these, 66 suffered from chloracne or severe dermatitis.  Observed deaths were

                                         C-31

-------
compared to the expected number using three external reference groups: the town of
Ludwigshafen (n = 180,000), the district of Rhine-Hessia-Palatinate (n = 1.8 million), and the
Federal Republic of Germany (n = 60.5 million).  Another comparison group was assembled by
selecting age-matched employees taken from other cohorts under study. This additional
comparison was aimed at avoiding potential biases associated with healthy worker effect when
using an external referent.
       During a follow-up interval of up to 26 years (1953-1979), 21 individuals died.  Of
these, seven deaths were from cancer. The expected number of cancer deaths derived for the
three external comparison groups  ranged between 4.1 and 4.2, producing an SMR of 1.7
(p-values ranged between 0.12 and 0.14). Excess mortality was found for stomach cancer based
on the external comparisons (p < 0.05); however, this was  based on only three cases. No other
statistically significant excesses were found with the external comparisons made to the other
cohorts of workers.

C. 1.1.1.2.1.2.  Study evaluation
       In the Thiess et al. (1982) study, no TCDD exposures were derived for the workers, thus
no dose-reconstruction was performed. The findings from this study are severely limited by the
small size of the cohort.  The 74 workers followed in this cohort represent the smallest number of
workers across the occupational cohorts (McBride et al., 2009a: 2009b: Michalek and Pavuk,
2008: Steenland et al.. 2001b: BecheretaL 1998: Hooiveld et al.. 1998: Fingerhut et aL 1991b)
that have investigated TCDD exposures and cancer mortality. Mechanisms of follow-up were
excellent as all individuals were traced, and death certificates were obtained from all deceased
workers.
       Although the study does compare the mortality experience to other occupational cohorts,
the paper provides insufficient information to adequately interpret these findings. For example, a
description of these occupations is lacking making it impossible to determine whether these
cohorts were exposed to other occupational carcinogens that might have confounded the
associations between TCDD exposure and cancer mortality.
                                         C-32

-------
C. 1.1.1.2.1.3.  Suitability of data for TCDD dose-response modeling
       Subsequent data assembled for the BASF cohort provide more detailed exposure
characterization, and also include information for 243 male workers employed at the plant. As
such, this study did not meet the considerations for further dose-response analysis.

C.l.1.1.2.2.  Zober et al (1990)
C.I.1.1.2.2.1.  Study summary
       Zober et al. (1990) also examined the mortality patterns of those involved in the 1953
accident at the BASF plant.  As detailed in their paper, the size of the original cohort was
expanded to 247 workers through efforts to locate  all who were exposed in the accident or during
the clean-up.  Three approaches were followed in assembling the cohort. Sixty-nine cohort
members were identified from the company physician's list of employees exposed as a result of
the accident (Subcohort Cl). Sixty-six of these workers were included in the original study
population of workers Thiess et al. (1982) examined.  Eighty-four other workers who were
potentially exposed to TCDD due to their involvement in demolitions or operations  were added
to the cohort.  This group included 43 firemen, 18  plant workers, 7 bricklayers, 5 whitewashers,
4 mechanics, 2 roofers, and  5 individuals in other occupations (Subcohort C2).  The cohort was
further augmented through the Dioxin Investigation Program, which sought to locate those who
were involved in the 1953 accident and were still alive in 1986. Current and former workers
enrolled in the study were asked to identify other current or former coworkers (including
deceased or retired) who might have been exposed from the accident. This third component of
94 workers (Subcohort C3) included 27 plant workers, 16 plumbers, 10 scaffolders,
10 professionals, 7 mechanics, 6 transportation workers, 5 bricklayers, 5 laboratory  assistant,
3 insulators, and 5 individuals in other occupations. A medical examination was performed for
those identified through the  Dioxin Investigation Program, and blood measures were obtained for
28 of these workers.
       External comparisons of the workers' mortality experience to the general population of
the Federal Republic of West Germany were made using SMRs.  Person-years were tabulated
across strata defined by calendar period, sex, and age-group. Sixty-nine deaths including 23
from cancer were  detected among the workers during the 34-year follow-up period (November
17,  1953  through December 31, 1987). Cause-specific death rates for these same strata were

                                         C-33

-------
available for the Federal Republic of West Germany.  Stratified analyses were conducted to
examine variations in the SMRs according to years since first exposure (0-9, 10-19, and
>20 years) for each of the three subcohorts, as well as 114 workers with chloracne.
       Although it was consistent in magnitude with findings from the NIOSH cohort, a
statistically significant SMR for all cancer mortality was not observed (SMR =1.17,
90% CI = 0.80-1.66).  The SMRs for each of the three subcohorts varied substantially.  For
Subcohorts Cl, C2, and C3, the SMRs were 1.30 (90% CI = 0.68-2.26), 1.71
(90% CI = 0.96-2.83), and 0.48 (90% CI = 0.13-1.23), respectively. The SMRs increased
dramatically when analyses were restricted to those with 20 or more years since first exposure in
Subcohort Cl (SMR = 1.67, 90% CI = 0.78-3.13) and Subcohort C2 (SMR = 2.38,
90% CI = 1.18-4.29).  Meanwhile, in a subgroup analysis of those with chloracne, for the period
of 20 or more years after first exposure, a statistically significant excess in cancer mortality was
noted (SMR = 2.01; 90% CI = 1.22-3.15).

C. 1.1.1.2.2.2.  Study evaluation
       An important limitation of the study is the manner in which the cohort was constructed.
Subcohort C3 was constructed by identifying individuals who were alive in 1986.  This resulted
in 97 active and retired employees who participated in the program, with 94 included in the
analysis. Although these individuals did identify other workers who might have also retired or
died, inevitably, some individuals who had died were not included in the cohort. This would
serve to underestimate the SMRs that were generated with external comparisons to the German
population. Indeed, cancer mortality rates in this subcohort were about half of what would have
been expected based on general population rates (SMR = 0.48, 90% CI = 0.13-1.23).
Additionally, more than half of Subcohort C2 were firemen (43 of 84), who were likely exposed
to other occupational carcinogens. Quantitative analyses of epidemiologic data for firefighters
have demonstrated increased cancer risk for several different forms of cancer (Youakim, 2006).
Therefore, potential confounding from other occupational exposures of the firefighters could
have contributed to the higher SMR in Subcohort C2 cohort and is a concern. Data on cigarette
smoking were not available either. No excess for nonmalignant respiratory disease was found,
however, suggesting this might not be an important source  of bias.
                                          C-34

-------
C.I. 1.1.2.2.3.  Suitability of data for TCDD dose-response modeling
       As with the Thiess et al. (1982) publication, individual-level estimates of workers'
exposures were not made. Lack of exposure estimates precludes a quantitative dose-response
analysis using these data. Also, the study design is not well suited to characterization of risk
using the SMR statistic.  Mortality is likely under-ascertained in the large component of the
cohort that was constructed through the identification of surviving members of the cohort.

C.l.1.1.2.3.  Ott andZober (1996a)
C.I.1.1.2.3.1.  Study summary
       Ott and Zober (1996a) extended the analyses of the BASF cohort to include estimates of
individual-level measures of TCDD. The researchers also investigated associations with cancer
mortality and incidence.  The cohort follow-up period of 39 years extended until December 31,
1992, adding 5 years to the previously published study (Zober et al., 1990). Ott and Zober
(1996a) identified incident cases of cancer using occupational medical records, death certificates,
doctor's letters, necropsy reports, and information from self-reported surveys sent to all
surviving cohort members.  Self-reported cancer diagnoses were confirmed by contacting the
attending physician.
       This  study characterized exposure by two methods: (1) determining chloracne status of
the cohort members, and (2) estimating cumulative TCDD  (ug/kg) levels.  In 1989, serum
measures were sought for all surviving members of the 1953 accident, and serum TCDD levels
were quantified for 138 individuals.  These serum levels were used to estimate cumulative
TCDD concentrations for all 254 members  of the accident cohort.  Ott et al. (1993) published a
description of the exposure estimation procedure, which was a regression model that accounted
for the circumstances and duration of individual exposure.  The average internal half-life of
TCDD was estimated to be 5.8 years based on repeated serum sampling of 29 individuals.  The
regression model allowed for this half-life to vary according to the percentage of body fat, and
yielded half-lives of 5.1 and 8.9 years among those with 20% and 30% body fat, respectively.
Previous analyses of this cohort had used a half-life of 7.0 years (Ott etal., 1993).
       TCDD half-life has been reported to increase with percentage of body fat in both
laboratory mammals (Geyer et al..  1990) and humans (Zober and Papke. 1993).  Ott and Zober
(1996a) contend  that observed correlations  with chloracne severity and cumulative estimates of

                                         C-35

-------
TCDD exposure indirectly validated this exposure metric. Specifically, the mean TCDD
concentration for those without chloracne was 38.4 ppt; for those with moderate and severe
forms of chloracne, the mean was 420.8 ppt and 1,008 ppt, respectively.
       Unlike the NIOSH cohort, individual-level data were collected for other cancer risk
factors. These factors included body mass index at time of first exposure, history of
occupational exposure to p-naphthylamine and asbestos, and history of smoking. Smoking data
were available for 86% of the cohort.  SMRs were based on the external referent population of
West Germany.  For cancer incidence, Ott and Zober (1996a) generated standardized incidence
ratios (SIRs) using incidence rates for the state of Saarland (1970-1991) as the external referent.
They calculated SMRs (and SIRs) for three or four categories of cumulative TCDD levels:
<0.1 ug/kg, 0.1-0.99 ug/kg and >1 ug/kg.  The Cox regression model was used to characterize
risk within the cohort using a continuous measure of TCDD. These analyses considered the
potential confounding influence of age, smoking, and body mass index using a stepwise
regression modeling approach.  The Cox modeling employed a stratified approach using the date
of first exposure to minimize possible confounding between calendar period and exposure. The
three first exposure groups were: exposure within the first year of the accident, exposure between
1 year after the accident and before 1960, and exposure after 1959.  The Cox regression
estimates were presented in terms of conditional risk ratios (i.e., hazard ratios adjusted for body
mass index, smoking and  age).
       Although no statistically significant excess relative to the general population was
detected for all cancer mortality, there was some suggestion of an exposure-response
relationship. In the 0.1-0.99 ug/kg, 1-1.99 ug/kg, and >2.00 ug/kg exposure groups, the  all
cancer SMRs were 1.2 (95% CI = 0.5-2.3), 1.4 (95% CI = 0.6-2.7) and 2.0 (95% CI = 0.8-4.0),
respectively.  Higher SMRs for cancer (all sites combined) were also found with an increased
interval since exposure first occurred. Specifically, when observed versus expected counts of
cancer were compared in the time interval 20 years after first exposure, the SMR in the highest
combined exposure group (>1 ug/kg) was 1.97 (95% CI = 1.05-5.36).  An excess in lung  cancer
also was noted with the same lag in this exposure group (SMR = 3.06, 95% CI = 1.12-6.66).
For cancer incidence, a statistically significant increased SIR for lung or bronchus cancer  was
observed in the highest combined exposure (>1 ug/kg)  category (SIR = 2.2, 95% CI = 1.0-4.3),
                                         C-36

-------
but no other statistically significant associations were detected for any other cancer site. No
cases of soft-tissue sarcoma were found among the cohort members in this analysis.
       Cox regression models also were used to conduct internal cohort comparisons by
generating hazard ratios as measures of relative risk for TCDD exposures with adjustment for
smoking, age and body mass index.  A statistically significant association between TCDD dose
(per ng/kg) and cancer mortality was detected (RR = 1.22, 95% CI = 1.00-1.50), but not for
cancer incidence (RR = 1.11, 95% CI = 0.91-1.35).  Statistically significant findings were
observed for stomach cancer mortality (RR = 1.46, 95% CI = 1.13-1.89) and incidence
(RR= 1.39, 95% CI= 1.07-1.69).
       The Ott and Zober (1996a) study also compared the relationship between TCDD
exposure categories and cancer mortality from all sites combined according to smoking status.
Associations were noted between increased exposure to TCDD and mortality from cancer among
current smokers, but not among never or former smokers.

C. 1.1.1.2.3.2.  Study evaluation
       The Ott and Zober (1996a) study characterizes exposure to TCDD at an individual level.
Therefore, unlike past studies of this cohort, these data can provide an opportunity for
conducting quantitative dose-response modeling.  As with the more recent studies involving the
NIOSH cohort, serum samples were obtained from surviving cohort members and then used to
back-extrapolate TCDD values for all cohort members. In the BASF cohort, however, serum
data were available for a much higher percentage of cohort members (54%) than in the NIOSH
cohort (5%).  An additional study strength was the collection of questionnaire data, which
allowed for the potential confounding influence of cigarette smoking and body mass index to be
taken into account.
       The Ott and Zober (1996a) study also evaluates the relationship between TCDD and
cancer incidence.  Most cohort studies of TCDD-exposed workers have relied solely on mortality
outcomes. The availability of incidence  data better allows for period of latency to be described,
and moreover, to characterize risks associated with cancers that typically have long survival
periods.  The  authors provide few details on the expected completeness of ascertainment for
incident cancer cases, which makes determining any  associated bias difficult.  They do, however,
suggest that nonfatal cancers are more likely to have  been missed in the earlier part of the

                                         C-37

-------
follow-up.  The net result of differential case ascertainment over time makes evaluating
differences in risk estimates across different periods of latency impossible.
       The small sample size of the cohort (n = 243 men) limited the statistical power to detect
small associations for some of the exposure measures. This also effectively limited the ability to
analyze dose-response relationships quantitatively, particularly across strata such as time since
exposure. For site-specific analyses, the cancer site with the most cancer deaths was the
respiratory system (n = 11). Given the evidence of an exposure-response relationship noted for
all cancer sites combined, quantitative dose-response analysis using these cohort data would be
limited to the evaluation of this endpoint.
       The most important limitation of this study is related to the construction of the
third component of the cohort.  As mentioned earlier, this cohort was assembled by actively
seeking out surviving members of the cohort in the mid-1980s. The mortality experience of this
cohort is much lower than that of the general population over the entire follow-up, a result that is
expected given that the large component of the cohort was made up of individuals known to be
alive as of 1986. The net result is likely an underestimate of the SMR.

C.I. 1.1.2.3.3.  Suitability of data for TCDD dose-response modeling
       This study was included in the quantitative dose-response modeling for the
2003 Reassessment (U.S. EPA, 2003). The characterization of exposure data and availability of
other risk factor data at an individual level are appropriate for use in quantitative dose-response
analyses.

C. 1.1.1.3.  The Hamburg cohort
       The Hamburg cohort has been the subject of several cancer risk assessments.  As  with the
NIOSH and BASF cohorts, analyses have progressed from basic comparisons of mortality rates
to those in the general population to more sophisticated internal cohort analyses involving the
reconstruction of TCDD exposures using serum measures.  This cohort consists of approximately
1,600 workers who were employed in the production of herbicides at a plant in Hamburg,
Germany during 1950-1984 (Becher et al.. 1998: Flesch-Janys et al.. 1995).  The herbicides
produced included 2,4,5-T, p-hexachlorocyclohexane and lindane. The production of TCP and
2,4,5-T was halted in 1954 following a chloracne outbreak. The plant ceased operations  in 1984.

                                          C-38

-------
Approximately 20 different working areas were identified, which, in turn, were grouped into
five main areas based on putative TCDD exposure levels.  One working area was deemed to be
extremely contaminated, having TCDD exposures at least 20-fold higher than in other areas. In
this section, the studies undertaken in this cohort that have examined cancer mortality are
summarized.

C.l.1.1.3.1.  Manz et al (1991)
C.I.1.1.3.1.1.  Study summary
       Manz et al. (1991) investigated patterns of mortality in the Hamburg cohort.  The study
population consisted of 1,583 workers (1,184 men, 399 women) who were employed for at least
three months between  1952 and 1989. Casual workers were excluded as they lack sufficient
personal identifying information thereby not allowing for associations with mortality outcomes
to be examined. Vital  status was determined using community-based registries of inhabitants
throughout West Germany. Cause of death until the end of 1989 was determined from medical
records for all cancer deaths and classified based on the ninth revision of the International
Classification of Diseases (WHO, 1978). Although Manz et al. (1991) present some data on
cancer incidence for the cohort, the data are incomplete as  information was available on only
12 cases; 103 (93 men and 20 women) cancer deaths were  observed in the cohort.
       In this study, the authors used information on production processes to group workers into
categories of low, medium, or high exposure to TCDD. This information was based on TCDD
concentrations in precursor materials, products, waste, and soil from the plant grounds, measured
after the plant closed in 1984.  The distribution of workers into the low, medium, and high
exposure groups was 186 (79 men and 107 women), 901 (636 men and 265 women), and
496 (469 men and 27 women), respectively. The authors examined the validity of the
three exposure categories using a separate group of 48 workers not selected for the cohort who
volunteered to provide adipose tissue samples. Selection criteria and response rate information
for the 48 volunteers were not provided, nor was there any indication that comparisons were
made between the 48 volunteers and the individuals included in the study cohort. The median
exposure of the 37 volunteers in the high group was 137 ng/kg and 60 ng/kg in the remaining 11.
Although the results indicate higher TCDD levels in the high-exposure group, combining the
lower two groups precludes separate validation of the two exposure groups.  In addition, the

                                         C-39

-------
authors reported that some exposure misclassification was likely given that 5 of the 37 workers
classified in the high exposure group had adipose levels lower than background (20 ng/kg).
Information about chloracne in the cohort was incomplete, and, therefore, was not used as a
marker of TCDD exposure.  Other surrogate measures of exposure were considered in this study,
including duration of exposure and year of first employment. For the latter measure,
employment that began after 1954 was assumed to result in much lower exposures given that
production of 2,4,5-T and TCP stopped in 1954.
      External comparisons of cancer mortality were made by calculating SMRs using the
general population of West Germany as a referent.  Comparisons of mortality in the cohort also
were made to a separate cohort of 3,417 gas supply workers to avoid bias from the healthy
worker effect.  Vital status and cause of death in the gas supply workers were determined using
the same methods as in the Hamburg cohort.  SMRs were calculated relative to both referent
populations (West Germany and gas supply workers) across low, medium, and high TCDD
exposure groups. The  comparison of mortality to the gas supply workers, however, extended
only until the end of 1985, whereas,  comparisons to the general population extended until 1989.
Stratified analyses were undertaken to calculate SMRs for each of the three exposure groups for
categories of duration of employment (<20 versus >20 years) and date of entry into the cohort
(<1954vs. >1954).
      When compared to the  general population, overall cancer mortality was elevated in male
cohort members (SMR = 1.24, 95% CI = 1.00-1.52) but not in females (SMR = 0.80,
95% CI = 0.60-1.05).  A twofold  increase in female breast cancer mortality was noted although
it did not achieve statistical significance at the alpha level of 0.05 (SMR = 2.15,
95% CI = 0.98-4.09).  The SMR among men was further increased when analyses were
restricted to workers who were employed for at least 20 years (SMR = 1.87,
95% CI = 1.11-2.95).  Analyses restricted to those in the highest exposure group produced an
even higher SMR for those with at least 20 years of employment (SMR = 2.54,
95% CI = 1.10-5.00).  Statistically significant excesses in risk were detected among those who
first worked before 1954, but not afterward. Furthermore, a dose-response trend was observed
across increasing exposure categories in the subset of workers employed before 1954. The
SMRs using the cohort of gas supply workers as the referent group for the low, medium, and
high groups in  this subset were 1.41  (95% CI = 0.46-3.28), 1.61 (95% CI =  1.10-2.44), and 2.77

                                         C-40

-------
(95% CI = 1.59-4.53), respectively. This finding is consistent with what was known about
TCDD exposures levels at the plant, namely, that TCDD concentrations were much higher
between 1951 and 1954, with subsequent declining levels after 1954.
       Generally speaking, patterns of excess mortality were similar when the cohort of gas
workers was used as a reference group. The overall SMR for men was 1.39
(95%CI= 1.10-1.75); and was 1.82 (95% CI = 0.97-3.11) when analyses were restricted to
workers with 20 or more years of employment. A  dose-response trend also was observed across
exposure categories when analyses were restricted to those employed for at least 20 years. In
particular, with these analyses, no cancer deaths were observed among those in the lowest
exposure group, while the SMRs in the middle and high exposure groups were 1.36
(95% CI = 0.50-2.96) and 3.07 (95% CI= 1.24-6.33).
       SMRs also were generated for several site-specific cancers relative to the West German
general population and the gas worker cohort. No  statistically significant excesses were
observed using the general population reference. In contrast, statistically significant excesses
were observed for lung cancer (SMR = 1.67, 95% CI = 1.09-2.44) and hematopoietic system
cancer (SMR = 2.65, 95% CI = 1.21-5.03) relative to the gas workers cohort.

C. 1.1.1.3.1.2.  Study evaluation
       The Manz et al. (1991) findings indicate an excess of all cancer mortality among the
workers with the highest exposures, particularly those who worked for at least 20 years and were
employed before 1954.  The findings across categories of exposure within the subsets of workers
employed for at least 20 years and before 1954, particularly using the cohort of gas supply
workers, are consistent with a dose-response relationship. These elevated cancer mortality rates
found among those employed before 1954 occurred at a time where TCDD exposures were
highest. Other carcinogenic coexposures, such as benzene, asbestos, and dimethyl sulfate, could
have occurred among this population.  Given that no substantial changes in the production
processes at the Hamburg plant occurred after 1954, comparable levels of these coexposures
would be expected before and after 1954. Exposures to these other chemicals varied across
different departments/groups; therefore, confounding was unlikely since a strong association
between concentrations of these chemicals and TCDD exposures was not evident.  No
                                         C-41

-------
information, however, was presented on potential exposure to other DLCs which may confound
the associations that were detected.
       Detailed information on workers' smoking behaviors was not collected.  Limited
evidence indicated, however, that smoking prevalence between the Hamburg cohort and the gas
supply workers cohort was quite similar. A nonrepresentative sample of 361 workers in the
Hamburg cohort and the sample of 2,860 workers in the gas supply cohort found that the
self-reported smoking prevalence was 73 and 76% in these two cohorts, respectively. This
suggests that the two cohorts are comprised predominantly of smokers.  The similarity in overall
smoking prevalence suggests that comparisons of cancer mortality between the two groups are
not unduly influenced by an inability to adjust for smoking.

C. 1.1.1.3.1.3.  Suitability of data for TCDD dose-response modeling
       The data compiled for the Manz et al. (1991) study do satisfy many of the considerations
for conducting quantitative dose-response analysis; health outcomes appear to be ascertained in
an unbiased manner, and exposure was characterized on an individual-level basis. However, as
demonstrated in later studies, there was a large DLC component that was not quantified or
assessed in this study. Dose-response associations between TCDD  and cancer mortality were
detected, with stronger associations observed with increased  periods of latency and for those who
first worked when TCDD was at higher levels.
       The size of the cohort, although not as large as the NIOSH cohort, does offer sufficient
statistical power to evaluate TCDD-related risk for all cancers combined. The data are limited,
however, for characterizing cancer risks among women; only 20 cancer deaths occurred in the
399 women included in the cohort. It is unlikely that the excess cancer risks using the external
reference population are due to uncontrolled effects from  smoking since dose-response patterns
were strengthened when comparisons were made to the cohort of gas supply workers rather the
general population referent where smoking rates were likely lower.  The inability to account for
other occupational exposure when TCDD exposures were much higher (pre-1955) could result in
confounding if these other exposures were related to TCDD and the health outcomes under
consideration. This data set would be suitable for quantitative dose-response modeling if the
exposure characterization of the cohort could be improved using biological measures of dose.
                                         C-42

-------
C.l.1.1.3.2.  Flesch-Janys et al (1995)
C.I.1.1.3.2.1.  Study summary
       In 1995, Flesch-Janys et al. (1995) published an analysis of the male employees from the
Hamburg cohort that extended the follow-up to 40 years (1952-1992). Inclusion of these three
additional years of follow-up resulted in a sample size of 1,189 male workers.
       The authors estimated a quantitative exposure variable for concentrations of TCDD in
blood at the end of exposure (i.e., when employment in a department ended) and above German
median background TCDD levels.  The TCDD exposure assessment defined 14 production
departments according to TCDD levels in various products in the plant, in waste products, and in
various buildings. The time (in years) each worker spent in each department then was
calculated.  Concentrations of TCDD were determined in 190 male workers using serum
(n = 142) and adipose tissue samples (n = 48). Selection criteria and response rate information
was not provided for this subsample. The authors used a first-order kinetic model to calculate
TCDD levels  at the end of exposure for the 190 workers with available polychlorinated
dibenzo-p-dioxin (PCDD) and -furan (PCDF) at various time points. Half-lives were calculated
from an elimination  study of 48 workers from this cohort, and the median TCDD background
level was estimated at 3.4 ng/kg blood fat from the German population (Flesch-Janys et al.,
1994; Papke et al., 1994).  Using the one-compartment, first-order kinetic model, the half-life of
TCDD was estimated to be 6.9 years (Flesch-Janys, 1997).  Increased age and higher body fat
percentage were associated with increased TCDD half-life, while smoking was associated with a
higher decay rate for most of the congeners examined  (Flesch-Janys et al., 1996). Cumulative
TCDD exposures for all 1,189 workers were estimated by summing exposures over the time
spent in all production departments (expressed in terms of ng/kg of blood fat) in combination
with quantitative estimates based on the blood and adipose samples from the 190 workers. The
contribution of each working department on overall PCDD exposure was estimated using
ordinary least squares regression.  The authors also applied a metric of total toxicity equivalence
(TOTTEQ) as the weighted sum of all congeners where weights were TEQs that denoted the
toxicity of each congener relative to TCDD.
       Similar to previous analyses on this cohort, comparisons were made using an external
referent group of workers from a gas supply company  (Manz et al., 1991). In contrast to
previous analyses where SMR statistics were generated using this "external" reference, however,

                                         C-43

-------
Flesch-Janys et al. (1995) used Cox regression. The Cox regression models treated the gas
worker cohort as the referent group, and six exposure groups were defined from serum-derived
cumulative TCDD estimates. The groups were determined by using the first four quintiles with
the upper two exposure categories corresponding to the ninth and tenth deciles of the cumulative
TCDD.  Internal cohort comparisons used those workers in the lowest quintile as the referent
group, as opposed to the cohort of gas workers. A similar approach was used to model TEQs.
No known TCDD exposures occurred in the gas workers, so they were assigned exposures based
on the median background levels in the general population.  RRs were calculated based on
exposure above background levels; in other words, background levels were assumed to be
equivalent across all workers and also for those employed by the gas supply company.  The RRs
derived using the Cox model were adjusted for total duration of employment, age, and year when
employment began.
      The Cox regression with the cohort of gas workers as the referent exposure group yielded
a linear dose-response relationship between cumulative TCDD exposure and cancer mortality for
all sites combined (p < 0.01). The RRs for all-cancer mortality were 1.59, 1.29, 1.66, 1.60, 1.70,
and 3.30. For four of the six categories (excluding the referent group), the RRs were statistically
significant (p < 0.05); in the highest TCDD exposure category (344.7-3,890.2 ng/kg) the RR
was 3.30 (95% CI = 2.05-5.31).  Similar findings were evident with TOTTEQ.  A dose-response
pattern for all cancer mortality (p < 0.01) based on the internal cohort comparisons was also
detected.
      The authors performed an additional analysis to evaluate the potential confounding role
of dimethylsulfate.  Although no direct measures of dimethylsulfate were available, the
investigators repeated analyses by excluding  149 workers who were employed in the department
where dimethylsulfate was present. A dose-response pattern persisted for TCDD and cancer
mortality (p < 0.01), and those in the highest exposure group (344.7-3,890.2 ng/kg of blood fat)
hadaRRof2.28(95%CI= 1.14-4.59).

C. 1.1.1.3.2.2.  Study evaluation
      The Flesch-Janys et al. (1995) study used serum-based measures to determine cumulative
exposure to TCDD at the end of employment for all cohort members.  They used the standard
one-compartment, first-order kinetic model and samples obtained from 190 male workers. This

                                         C-44

-------
quantitative measure of exposure permits an examination of a dose-response relationship.
However, there is not enough information provided on the selection of these 190 workers to
determine how representative they were of the larger cohort. Confounding for other
occupational exposures is unlikely to have biased the results.  A dose-response relationship
persisted after excluding workers exposed to dimethylsulfate. Other potential exposures of
interest included benzene and isomers of hexachlorocyclohexane. Exposure to these agents,
however, was highest in the hexachlorocyclohexane and lindane department, where TCDD
exposures were lower. Confounding was unlikely due to exposure to these chemicals, since a
strong association between concentrations of these chemicals and TCDD exposures was not
evident (due to considerable variability in concentrations across different departments/groups).
As outlined earlier, the study findings are unlikely to be biased for cigarette smoking as the
prevalence  of smoking in the cohort was similar to that in the comparison population. Moreover,
more recent analyses of serum-based TCDD exposure measures found no correlation with
smoking status in this cohort (Flesch-Janys  et al., 1995)—a necessary condition for confounding
to occur.
       The authors used an exposure metric that quantified the cumulative TCDD exposure of
workers at the time they were last exposed.  As a result, the authors were unable to characterize
risks associated with this metric for different periods of latency despite a lengthy follow-up
period.  Subsequent analyses constructed time-dependent measures of cumulative TCDD  and
accounted for excretion of TCDD during follow-up.
       In contrast to most risk assessments of TCDD exposure, this study modeled the
relationship between other DLCs and the risk of cancer mortality using the TOTTEQ metric.

C.I. 1.1.3.2.3.  Suitability of data for TCDD dose-response modeling
       The data used in this study  satisfy most of the considerations developed for performing a
quantitative dose-response analysis. However, latency period was not examined in this study.
Dose-response analyses were, therefore, limited to a subsequent study of this cohort (Becher et
al..  1998). which did examine latency.
                                          C-45

-------
C.l.1.1.3.3.  Flesch-Janys et al (1998)
C.I.1.1.3.3.1.  Study summary
       Flesch-Janys et al. (1998) undertook another analysis on this cohort that incorporated
additional sera data collected from 275 workers (39 females and 236 males). The follow-up
period was the same as that used in the  1995 publication, with mortality follow-up extending
until December 31, 1992. Analyses were based on 1,189 males who were employed for at least
3 months from January 1, 1952 onward. The authors continued this dose-response analysis to
address limitations in their previous work.  One limitation was that the previous method did not
account for the elimination of TCDD while exposures were being accrued during follow-up. A
second limitation was that the amount of time workers spent in different departments was not
considered. In the 1998 study, the "area under the curve" approach was used because it accounts
for variations in concentrations over time and reflects cumulative exposure to TCDD. The
authors used a first-order kinetic model to link blood levels and working histories to derive
department-specific dose rates for TCDD.  The TCDD background level of 3.4 ng/kg blood fat
for the German population was used (Papke et al., 1994). The dose rates were applied to
estimate the concentration of TCDD at every point in time for all  cohort members.  A cumulative
measure expressed as ng/kg blood fat multiplied by years was calculated and used in the SMR
analysis.  SMRs were calculated using general population mortality rates for the German
population between 1952 and 1992. No lag period was incorporated into the derivation of the
SMRs. The SMRs were  estimated for the entire  cohort and for exposure groups based on
quartiles obtained from the area under the curve. Linear trend tests were also performed. The
overall SMR for cancer mortality in the cohort was 1.41 (95% CI = 1.17-1.68). This SMR value
was higher than the SMR of 1.21 reported for this same cohort with 3 fewer years of follow-up
(Manz et al., 1991).  In terms of site-specific cancer mortality, excesses were found for
respiratory cancer (SMR = 1.71, 95% CI = 1.24-2.29) and rectal cancer (SMR =  2.30,
95% CI = 1.05-2.47). Increased risk for lymphatic and hematopoietic cancer (SMR = 2.16,
95% CI= 1.11-3.17) were also noted largely attributable (SMR = 3.73, 95%CI= 1.20-8.71)to
lymphosarcoma (i.e., non-Hodgkin lymphoma).  A dose-response relationship was observed
across quartiles of cumulative TCDD for all-cancer mortality (p < 0.01). The SMRs for these
quartiles were 1.24, 1.34, 1.34, and 1.73. Dose-response relationships were not observed for
lung cancer or hematopoietic cancers using this same metric. Dose-response relationships were

                                         C-46

-------
not observed with cumulative TEQ for any of the cancer sites examined (i.e., all cancers, lung
cancer, hematopoietic cancer).

C. 1.1.1.3.3.2.  Study evaluation
       The approach used in the Flesch-Janys et al. (1998) study offers a distinct advantage over
earlier analyses of the same cohort.  The authors used sera data on 275 male and female subjects
to estimate department-specific dose rates, although it is unclear whether data on females were
used to estimate TCDD levels among the males examined in the cancer mortality analysis.
Three more years of follow-up were available, and the characterization of exposure using the
"area under the curve" better captures changes in cumulative exposure using a person-years
approach when compared to estimates of cumulative TCDD at the time of last exposure. As
noted previously, other occupational exposures or cigarette smoking are unlikely to have biased
the study findings. A sufficient length of follow-up had accrued, and dose-response relationships
were evident. DLCs were evaluated in this study. For TCDD, the mean  concentration was
101.3 ng/kg at the time of measurement. For other higher chlorinated congeners, the
corresponding mean (without TCDD) was 89.3 ng/kg.

C.I. 1.1.3.3.3.  Suitability of data for TCDD dose-response modeling
       The data used in this study satisfy most of the considerations developed for performing a
quantitative dose-response analysis. However, latency was not examined in this study.
Dose-response analyses were, therefore, limited to a subsequent study of this cohort (Becher et
al., 1998) which did examine latency and supersedes the Flesch-Janys et  al. (1998) study.

C.l.1.1.3.4.  Becher et al (1998)
C.I.1.1.3.4.1.  Study summary
       The Becher et al. (1998) quantitative cancer risk assessment for the Hamburg cohort was
highlighted in the 2003 Reassessment as being appropriate for conducting dose-response
analysis.  The integrated TCDD concentration over time, as estimated in  the Flesch-Janys et al.
(1998) study, was used as the exposure variable.  Estimates of the half-life of TCDD based on
the sample of 48  individuals with repeated measures were incorporated into the model that
back-calculated TCDD exposures to the end of the employment (Flesch-Janys et al., 1996). This
                                          C-47

-------
method took into account the age and body fat percentage of the workers.  In Becher et al.
(1998), the analysis used the estimate of cumulative dose (integrated dose or area under the
curve) as a time-dependent variable.
       Poisson and Cox regression models were used to characterize dose-response
relationships. Both models were used to conduct internal comparisons where a person-years
offset was used, and to an external comparison where an offset of expected number of deaths
was used. The person-years offset was used to account for varying person-time accrued by
workers across exposure categories. The use of the expected number of deaths as an offset
allows risks to be described in relation to that expected in the general population. Within each
classification cell of deaths and person-years, a continuous value TCDD and TEQ levels based
on the geometric mean were entered into the Poisson model. For the Cox model, accumulated
dose was estimated based on area under the curve for TCDD, TEQ, TEQ without TCDD, and
p-hexachlorocyclohexane.  These other coexposure metrics were adjusted for in the Cox
regression analyses. Other covariates considered included in the models were year of entry, year
of birth, and age at entry into the cohort.  A background level of 3.4 ng/kg blood fat for the
German population was used (Papke et al., 1994). A variety of latencies was evaluated (0, 5, 10,
15, and 20 years),  and attributable and absolute risks were estimated.  The unexposed cohort of
gas workers was used for most internal analyses.
       Internal and external comparisons using the Poisson model found positive associations
with TCDD exposure and mortality from all cancers combined. The slope associated with the
continuous measure of TCDD (ng/kg blood fat *  years) for the internal comparison was 0.027
(p < 0.001), which decreased to 0.0156 (p = 0.07) after adjusting for age and calendar period.
The slope for the external comparison was 0.0163 (p = 0.055);  this estimate was not adjusted for
other covariates. For TEQ, the slopes based on the internal comparisons were 0.0274 (p < 0.001)
in the univariate model and 0.0107 (p = 0.175) in the multivariate model after adjusting for age
and calendar period. The external estimate of slope for TEQ was 0.0109 (p = 0.164). Cox
regression of TCDD across six exposure categories, with a lag  of 0 years,  found a statistically
significant linear trend (p = 0.03) and those in the upper exposure group had a RR of 2.19
(95% CI = 0.76-6.29).  These estimates were adjusted for year of entry, age at entry, and
duration of employment. A similar pattern was observed with the Cox regression analysis of
                                         C-48

-------
TEQ; the linear test for trend, however, was not statistically significant at the alpha level of 0.05
(p = 0.06).
       Cox regression models that included both TCDD and TEQ (excluding TCDD) were
applied. In this model, the slope (P) for TCDD was 0.0089 (p = 0.058), while the coefficient for
TEQ (excluding TCDD) was -0.024 (p = 0.70). This suggests that confounding by other DLCs
was unlikely and the increased risk of cancer was due to TCDD  exposure.  For all TEQs
combined, the slope was 0.0078 (p = 0.066).
       The authors used multiple Cox models to evaluate the effect of latency.  The slope
estimates for both TCDD and TEQ increased dramatically with increasing latency.  The slope
estimates for TCDD increased from 0.0096 to 0.0160 (p < 0.05) when latency was increased
from 0 to 20 years. Similar changes in the TEQ slopes were noted (0.0093 to 0.0157).
Evaluations of dose-response curves found that the best-fitting curve was concave in shape,
thereby yielding higher risk at low exposure. Differences between the fit of the class of models
considered [i.e., RR(x,P) = exp (P \og(kx = 1))], however, were small.
       Attributable risks were generated only for TCDD, as the data suggested no effects with
other TEQs.  The additional lifetime risk of cancer assuming a daily intake of 1  pg TCDD/kg
body weight/day was estimated to range between 0.001 and 0.01.

C.I.1.1.3.4.2.  Study evaluation
       The Becher et al. (1998) study represents perhaps the most detailed analyses performed
on any cohort to date.  The findings were robust, as similar patterns were found with and without
using the gas supply worker cohort as the referent group. Exposures to other potential
confounding coexposures, such as DLCs, were taken into account, and workers with exposure to
other carcinogens (e.g., lindane) were excluded. Furthermore, latency was examined in this
study, unlike earlier studies of this cohort. Although the TCDD exposure estimates were derived
from a sample of 275 workers with repeated serum measures, the authors indicate that the
production department-specific estimates were in agreement with a priori expectations based on
an understanding of the chemistry and available industrial  hygiene data. The authors also
reported no differences in dose rate estimates related to gender or short durations of employment.
Similar to other studies, the potential for exposure misclassification based on limited number of
                                         C-49

-------
biomarker samples is hard to determine without more information on the representativeness of
the participants who provided samples.

C.I.1.1.3.4.3.  Suitability of data for TCDD dose-response modeling
       This study was included in the quantitative dose-response modeling for the
2003 Reassessment (U.S. EPA. 2003).  The data in the Becher et al. (1998) study are suitable for
conducting quantitative dose-response modeling.  The exposure data capture cumulative
exposure to TCDD as well as exposures to other DLCs.  The length of the follow-up is sufficient,
and the study does not appear to be subject to confounding or other types of biases. Therefore,
this study is utilized in quantitative dose-response analysis.

C. 1.1.1.4.  The Seveso cohort
       Several studies have evaluated the morbidity and mortality effects of residents exposed to
TCDD following a July 10,  1976, accidental release through an exhaust pipe at a chemical plant
in the town of Meda near Seveso, Italy. The released fluid mixture contained 2,4,5-T, sodium
trichlorophenate, ethylene glycol, and sodium hydroxide. Vegetation in the area showed
immediate signs of damage, and in the days following the accident, residents developed nausea,
headaches,  eye irritation, and dermal lesions, particularly children.
       This accident transported TCDD up to 6 km from the plant. Soil samples taken near the
plant revealed average levels of TCDD that ranged from 15.5 ug/m2 to 580.4 ug/m2 in the most
contaminated area near the plant (referred to as Zone A) (Bertazzi et al., 2001).  Zone A covered
87 hectares and extended 2,200 m south from the plant.  Another, more distant contaminated
zone (Zone B)  covering 270 hectares also had contaminated soil levels, but the TCDD
concentration range was much lower (1.7-4.3 ug/m3). A reference zone (Zone R), which
surrounded the two contaminated areas, had lower TCDD soil levels (range: 0.9-1.4 ug/m3) and
included approximately 30,000 residents.  Following the accident, most residents in Zone A left
the area.  Although residents in Zone B remained, they were under strict regulations to avoid
consuming homegrown products. In total, 736, 4,737, and 31,800 individuals lived in Zones A,
B, and R, respectively.  Within days of the accident, 3,300 animals (mostly poultry and rabbits)
were found dead.  Emergency slaughtering was undertaken to prevent TCDD from entering the
food chain, and within 2 years more than 80,000 animals had been slaughtered.  Mechanisms

                                         C-50

-------
were put into place for long-term follow-up of these residents.  Unlike the other occupational
cohort studies, the follow-up of this population allows for risks to be characterized for females.
       The mortality studies from Seveso published to date have not incorporated serum TCDD
levels that were measured in individuals. Needham et al. (1997) describe the collection of serum
samples from a sample of the exposed population and control subjects in 1976.  In 1988, human
exposure to TCDD was assessed by measuring small volumes of serum remaining from medical
examinations done in 1976.  An examination of these data revealed some of the highest serum
TCDD levels ever reported,  that the half-life of TCDD  in this population was between 7 and
8 years, and that half-life varied between women and men.  The half-life of TCDD in serum was
longer in women (~9 years)  than in men (~7 years) (Needham et al., 1994). In this report, the
findings of studies that characterized cancer risks in relation to exposure to TCDD from the 1976
accident are highlighted. These studies include comparisons of cancer mortality rates to the
general population based on zone of residence at the time of accident  (Consonni et al., 2008;
Bertazzi etal., 2001). More recent work done by Warner et al. (2002) investigated the
relationship between serum-based measures of TCDD and breast cancer among participants in
the Seveso Women's Health Study (SWHS).

C.l.1.1.4.1.  Bertazzi et al (2001)
C.I.1.1.4.1.1.  Study summary
       Several studies have reported on the mortality experience of Seveso residents. The more
recent publications having a longer follow-up of the cohort are evaluated here. In 2001, the
findings from a 20-year mortality study of Seveso residents was published (Bertazzi et al., 2001).
The Bertazzi et al. (2001) study was an extension of the 10- and 15-year follow-ups for mortality
(Pesatori et al.,  1998; Bertazzi etal., 1997; 1989) and the 10-year follow-up for cancer incidence
(Bertazzi et al.. 1993).
       In this cohort, TCDD exposures were assigned to the population using a three-level
categorical variable representative of the individual's place  of residence (Zones A, B, or R) at the
time of the accident or when the person first became a resident of the  zone, if that was after
1976. An external comparison to the province of Lombardy was  made by generating rate ratios
(RR) using Poisson regression techniques.  Person-years of follow-up were tabulated across
strata defined by age, zone of residence, duration of residence, gender, calendar time, and

                                         C-51

-------
number of years that had elapsed since the time of exposure.  Mortality rates during the
preaccident period also were compared to evaluate potential changes in rates due to the accident
and to evaluate whether patterns were consistent before and after the accident.
       No overall excess in mortality rates from all cancer sites combined was observed in
Zones A or B (combined) when compared to the reference population of Lombardy
(n = 9 million residents) (RR = 1.0, 95% CI = 0.9-1.2). Analyses of site-specific cancer
mortality revealed statistically significant excesses among residents in Zones A or B (combined)
for cancer of the rectum (RR = 1.8, 95% CI = 1.0-3.3) and lymphatic and hematopoietic
malignancies (RR = 1.7, 95% CI = 1.2-2.5). Lymphatic and hematopoietic malignancies were
elevated in women (RR = 1.8, 95% CI = 1.1-3.2) and in men (RR = 1.7, 95%  CI = 1.0-2.8).
       Analyses stratified by the number of years since first exposure  (i.e., 1976) revealed
higher risk among men with an increased number of years elapsed.  Similar to other studies, the
RR for all cancers (combined) was 1.3 (95% CI = 1.0-1.7) among men 15-20 years after first
exposure. No such increase after 15 years postexposure, however, was noted in women
(RR = 0.8, 95% CI = 0.6-1.2).

C.I.1.1.4.1.2.  Study evaluation
       Ascertainment of mortality appears to be excellent.  Vital status was established using
similar methods for both the exposed and reference populations.  No individual data were
collected and, therefore, the possibility that confounding by individual characteristics such as
cigarette smoking cannot be entirely dismissed. Bertazzi et al. (2001)  do note that the
sociodemographic characteristics of residents in the three zones were similar based on
independently conducted surveys, and no differences in chronic respiratory disease were found
across the different zones.  If excess mortality was attributable to cigarette smoking, such
excesses would be expected to be evident during the entire  study period. Latency analyses
revealed elevated risks 15-20 years postaccident. Finally, no excesses were observed for other
smoking-related cancers of the larynx, esophagus, pancreas, and bladder.  The observed excesses
in all  cancer mortality do not appear to be attributed to differential smoking rates between the
two populations.
       To examine potential for bias due to noncomparability in the two study populations, a
comparison of cancer mortality rates between the Seveso regions and the reference population of

                                         C-52

-------
Lombardy was conducted. Elevated rates for brain cancer mortality were noted in Seveso
relative to Lombardy, but the higher rates of leukemia mortality were found in Lombardy
relative to Seveso.  That no excess was reported for all cancer sites combined lends credence to
the hypothesis that the exposure to TCDD from the accident increased rates of cancer after a
sufficient period of latency.
       Stratified analyses were performed across several categorical variables including gender
and time since exposure.  The numbers of cancer site-specific deaths are quite small in many of
the 5-year increments since first exposure.  The study, therefore, has limited statistical power to
detect differences in mortality rates among the comparison groups for many cancer sites.
       Bertazzi et al. (2001) assigned exposures based on zone of residence.  Soil sampling
within each zone revealed considerable variability in TCDD soil levels within each zone.
Moreover, some individuals would have left the area shortly after the accident, and determining
the extent to which individuals in Zone B who were subject to the  recommendations near the
time of the accident adhered to them is difficult. As a result, exposure misclassification is
possible, and the use  of individual measures of TCDD level in serum is  preferred over zone of
residence for determining exposure.  As noted by the authors, the study  is better suited to "hazard
identification" than to quantitative dose-response analysis.

C.I.1.1.4.1.3.  Suitability of data for TCDD dose-response modeling
       Given the variability in soil TCDD levels within each zone and the lack of individual
level, no effective dose can be estimated for quantitative dose-response  analyses.  Uncertainty in
identifying the critical exposure window for the Seveso cohort is a key limitation.  The
evaluation of this study indicates that this study is not suitable for  quantitative dose-response
analysis.

C.l.1.1.4.2.  Warner et al (2002)
C.I.1.1.4.2.1.  Study summary
       To date, Warner et al. (2002) is the only published investigation  of the relationship
between  serum-based measures of TCDD and cancer in  Seveso. Eligible participants from the
SWHS (see Section C.I.2.1.4 for details) were women who, at the time  of the accident in 1976,
were 40 years of age  or younger, had lived in one of the most highly contaminated zones (A or

                                          C-53

-------
B), and had adequate sera collected soon after the explosion.  Enrollment in SWHS was begun in
March 1996 and lasted until July 1998.  Of the total 1,271 eligible women, 981 agreed to
participate in the study. Cancer cases were identified during interview and confirmed through
review of medical records. Information on other risk factors including reproductive history and
cigarette smoking was obtained through interview.
       Serum volumes greater than 0.5 mL collected between 1976 and 1981 were analyzed.
Most sera were collected in 1976/77 (n = 899); samples were collected in 1978-1981 for
54 women, and in 1996/97 for 28 women.  For samples collected after 1977, serum TCDD levels
were back-extrapolated using a first-order  kinetic model with a 9-year half-life (Pirkle et al.,
1989).  For 96 women with undetectable values, a serum level that was equal to one-half the
detection level was used.
       Analyses were based only on women who provided serum samples; no extrapolation of
values to a larger population was done.  Risks were therefore generated using data collected at an
individual level. Serum TCDD was analyzed as both a continuous variable and a categorical
variable. The distribution of serum TCDD levels of the 15 cases of breast cancer was examined
in relation to the distribution of all  women in the SWHS.  The median exposure was  slightly
higher among with the 15 cases of breast cancer (71.8 ppt) compared to those without (55.1 ppt),
and the exposure distribution among breast cancer cases appeared to be shifted to the right (i.e.,
the exposures were higher but followed the same distribution); however, no formal test of
significance was conducted.
       Warner et al. (2002) used Cox proportional hazards models to evaluate the risk of breast
cancer in relation to TCDD serum  levels while controlling for a number of potential risk factors.
In all, 21 women had been diagnosed with cancer, and of these, 15 cases were cancer of the
breast.  The analysis revealed that for every 10-fold increase in TCDD log-serum levels (e.g.,
from 10 to 100 ppt) the risk of breast cancer increased by a factor of 2.1 (95% CI = 1.0-4.6).
Risk estimates also were generated across  four categories (<20, 20.1-44, 44.1-100, >100 ppt),
with the lowest category used  as the reference. The RRs estimated in the third  and fourth highest
exposure categories were 4.5 (95% CI = 0.6-36.8) and 3.3 (95% CI = 0.4-28.0).  Although
statistical significance was not achieved for either category, likely because of the small number
of cases, the greater than threefold  risk evident in both categories is worth noting.  Given that the
reference category had only one incident case underscores the limited inferences that can be
                                         C-54

-------
drawn from these analyses.  The authors adjusted for numerous potential confounders, but
observed no differences between the crude and adjusted results; the authors, therefore, presented
unadjusted risks.

C. 1.1.1.4.2.2.  Study evaluation
       The findings from the Warner et al. (2002) study differ from reports in earlier studies in
which mortality outcomes noted the absence of an SMR association. The design of this study is
much stronger than earlier ones, given the improved characterization of exposure, the ability to
compare incidence rates within the cohort, the ability to control for potential confounding
variables at an individual  level, and the availability of incident outcomes.  The use of incident
cases (versus mortality data) should also help minimize potential bias due to disease survival.
Another important advantage was the ability to measure TCDD near the time of the accident,
thereby reducing the potential for exposure measurement error.
       A potentially important limitation of the Warner et al. (2002) study was that information
was collected only from those who were alive as of March 1996.  Therefore, TCDD and other
relevant risk factor data could not be collected for those who had previously died of breast
cancer.  Thirty-three women could not participate because they were either too ill or had died.
Of these, three died of breast cancer.  Given that there were only 15 breast cancer cases, the
exclusion of these 3 cases could have dramatically impacted the findings in either direction.
       Another limitation was that, at the  time of the follow-up, most women were still
premenopausal and therefore, most of the  cohort (average age = 40.8 years) had not yet attained
the age of greater risk of breast cancer (average age at diagnosis among the cases in this cohort
was 45.2 years).  Although comparable data from Italy were not found, the median age of
diagnosis for breast cancer among U.S. women from 2003-2007 was 61 years (Altekruse et al.,
2010). An ongoing follow-up of the cohort should be completed by 2010, which should allow
for increased number of incident breast cancers to be identified.  Given that the current analyses
were based only on 15 incident cases, this will substantially improve the statistical power of the
study. A secondary benefit is that the increased follow-up will allow for an investigation of
possible differential effects  according to the age the women were at the  time of exposure.
                                          C-55

-------
C.I. 1.1.4.2.3.  Suitability of data for TCDD dose-response modeling
       Several aspects of the Warner et al. (2002) study are weaknesses in the consideration of
this study for further dose-response modeling. Only 15 cases of breast cancer were available,
and no increases in risk were found with serum TCDD exposures between 20.1 and 44 ppt
(n = 2) when compared to those with <20 ppt (n = 1). The average age at the time of enrollment
was 40.8 years while the average age at diagnosis among the cases was 45.2 years.  As most
women had not yet reached the age when breast cancer cases are typically diagnosed, additional
follow-up of the cohort would improve the quantitative dose-response analysis and strengthen
this study. A key strength of this study, however, is that Warner et al. (2002) includes an
investigation of the relationship between individual serum-based measures of TCDD and cancer
in Seveso. Despite the weaknesses, this study meets the evaluation considerations and criteria
for inclusion and will be analyzed for quantitative dose-response modeling.

C.l.1.1.4.3.  Pesatori et al (2003)
C.I.1.1.4.3.1.  Study summary
       Pesatori et al. (2003) published a review of the short- and long-term studies of morbidity
and mortality outcomes in the Seveso cohort in 2003. This paper presented cancer incidence
data from 1977 to 1991 for Seveso males and females residing in Zones A, B and R relative to an
external population (i.e., uncontaminated areas). Mortality data are also presented for a 20-year
follow-up (1976-1996) relative to the reference population. As in the original Bertazzi et al.
(2001) study, RRs were estimated using Poisson regression. No associations were noted for zone
of residence and all cancer mortality for either males or females. Although no cases were
reported in Zones A and  B, soft tissue sarcoma incidence rates were higher among males from
Zone R (RR = 2.6, 95% CI = 1.1-6.3). Among males, residence in Zones A and B was
associated with lymphatic and hematopoietic cancer (RR =1.9, 95% CI = 1.1-3.1).  This
increased risk was due primarily to non-Hodgkin lymphoma, which accounted for 8  of the
15 incident cases (RR = 2.6, 95% CI = 1.3-5.3). Among females, increased incidence of
multiple myeloma (RR = 4.9, 95% CI = 1.5-16.1), cancer of the vagina (RR = 5.5,
95% CI = 1.3-23.8), and cancer of the biliary tract (RR = 3.0, 95% CI = 1.1-8.2) was associated
with residence in Zones A and B.
                                         C-56

-------
C. 1.1.1.4.3.2.  Study evaluation
       Limitations of the Pesatori et al. (2003) study included exposure misclassification from
the use of an ecological measure of exposure (i.e., region of residency at time of accident) and
low statistical power for some health endpoints. For example, all of the RRs presented above for
specific cancer mortality among females in the Pesatori et al. (2003) study were based on fewer
than five incident cases.

C.I. 1.1.4.3.3.  Suitability of data for TCDD dose-response modeling
       As with the studies of mortality among Seveso residents, the Pesatori et al. (2003) study
does not capture  TCDD exposure on an individual basis, and soil TCDD levels considerably vary
within each zone. Therefore, the  quality of the exposure data is inadequate for estimating the
effective dose needed for quantitative dose-response analysis.

C.l.1.1.4.4.  Baccarelli et al (2006)
C.I.1.1.4.4.1.  Study summary
       Given previous findings from Seveso, Baccarelli et al. (2006) examined t(14;18)
translocations in  the DNA of circulating lymphocytes of 144  healthy dioxin-exposed individuals.
These translocations are associated with the development of cancer, namely follicular
lymphomas. The study included  144 individuals selected from a previous population of
211 healthy subjects representative of the Seveso area, and 101 who had developed chloracne.
The investigators analyzed data from 72 (52 females and 20 males) high-TCDD plasma level
individuals (>10  ppt) and 72 (41 females and 31 males) low-TCDD plasma levels (<10 ppt),
matched for history of chloracne and smoking. A three-level categorical exposure variable was
used to evaluate dose response. This variable was developed by dividing those with exposures
>10 ppt into two groups: 10- <50  ppt, and 50-475.0 ppt. Trained interviewers administered a
questionnaire that collected data on demographic characteristics, diet, and residential and
occupational history.
       The prevalence of t(14;18) was estimated as those individuals having a t(14;18) positive
blood sample divided by the t(14;18) frequency (number of copies per million lymphocytes).
Baccarelli et al. (2006) found that the frequency of t( 14; 18) was  associated with plasma TCDD
levels, but no association between TCDD and the prevalence of t(14;18) was detected.

                                          C-57

-------
C.I.1.1.4.4.2.  Study evaluation
       Whether the frequency of t(14;18) associated with plasma TCDD levels translates into an
increased risk of lymphoma is uncertain as prospective data of TCDD on those who developed
non-Hodgkin lymphoma are lacking. Moreover, the t(14;18) translocation could be an important
event in the pre-B stage cell that contributes to tumorigenicity, however subsequent exposure to
carcinogenic agents might be necessary for t(14;18) cells to develop into a malignancy (Hoglund
et al.. 2004).

C.I.1.1.4.4.3.  Suitability of data for TCDD dose-response modeling
       Given that current TCDD plasma levels were measured for this study, it is unclear if the
effects of lymphocyte translocations may be due to an initial high exposure or are a function of
the cumulative exposure accrued over a longer time window.  Additionally, whether the
frequency of t(14;18) associated with plasma TCDD levels translates into an increased risk of
lymphoma is unknown. Dose-response analysis for this outcome, therefore, was not conducted.

C.l.1.1.4.5.  Comonni et al (2008)
C.I.1.1.4.5.1.  Study summary
       Consonni et al. (2008) analyzed cancer mortality in the Seveso cohort with the addition
of a 25-year follow up period. Similar analytic methods as Pesatori et al. (2003) were applied
with 25 years of follow-up added to the analysis (Consonni  et al., 2008).  An important addition
in this paper was the presentation of RRs for Zone R, which had the lowest TCDD  levels.
Poisson regression models were used to calculate RRs of mortality using Seregno as the
reference population. Cancer deaths observed in Zones A and B were 42 and 244, respectively.
       No statistically significant differences in all cancer mortality relative to the  reference
population were noted in any of the zones (Zone A: RR = 1.03, 95% CI = 0.76-1.39;  Zone B:
RR = 0.92, 95% CI = 0.81-1.05; Zone R: RR = 0.97, 95% CI = 0.92-1.02). Statistically
significant excesses in mortality from non-Hodgkin lymphoma (RR = 3.35,
95% CI =  1.07-10.46) and multiple myeloma  (RR = 4.34, 95% CI = 1.07-17.52) were observed
in the area with the highest TCDD levels (Zone A). No other statistically significant increases in
cancer mortality relative to the reference population were apparent. The absence of elevated
breast cancer mortality among women in this study was noteworthy, as this finding differs from

                                         C-58

-------
the results of a study of Seveso women for which TCDD exposures were estimated using serum
samples (Warner et al.. 2002).

C.I.1.1.4.5.2.  Study evaluation
       Although no individual-level data on smoking were available, the potential for
confounding is likely minimal. Independent smoking surveys found that smoking prevalence
rates in Desio, one of cities affected by the accident, were similar to those in districts just outside
the study area (Cesana et al., 1995). As mentioned earlier, one would expect elevated RRs over
the entire study period if smoking had biased the study results, and not just after 15-20 years
since exposure to TCDD.

C.I.1.1.4.5.3.  Suitability of data for TCDD dose-response modeling
       The lack of individual-level exposure data precludes quantitative dose-response modeling
using these data.

C. 1.1.1.5.  Chapaevsk study
       Industrial contamination of dioxin in the Chapaevsk region of Russia has been the focus
of research on environmentally-induced cancers and other adverse health effects.  The
Chapaevsk region is located in the Samara region of Russia and has a population of 83,000. The
region is home to a chemical plant that produced lindane and  its derivatives between 1967 and
1987, which are believed to be responsible for local dioxin contamination.  Soil sampling has
demonstrated a strong gradient of increased TCDD concentrations with decreased proximity to
the chemical plant (Revich et al., 2001).

C.l.1.1.5.1. Revich et al (2001)
C.I.1.1.5.1.1.  Study summary
       Revich et al. (2001) used a cross-sectional study to compare mortality rates of Chapaevsk
residents to two external populations of Russia and the region of Samara.  Mortality rates for all
cancers combined among males in Chapaevsk were found to be 1.2 times higher when compared
to the Samara region as a whole and 1.3 times higher than Russia. Similar to other studies, a
statistically significant excess was noted in men (SMR = 1.8,  95% CI = 1.6-1.9) but not in

                                         C-59

-------
women (SMR = 0.9, 95% CI = 0.8-1.1). Among men, the excess was highest for the
smoking-related cancers of the lung (SMR = 3.1, 95% CI = 2.6-3.5) and larynx (SMR = 2.3,
95% CI = 1.2-3.8) and urinary organs (SMR = 2.6, 95% CI = 1.7-3.6). Among females, there
was no increased SMR for all cancer sites combined, but excesses for breast cancer (SMR = 2.1,
95% CI = 1.6-2.7) and cancer of the cervix (SMR = 1.5, 95% CI = 1.0-3.1) were statistically
significant.
      Revich et al. (2001) also compared age-standardized cancer incidence rates in Chapaevsk
to those in Samara. Although statistical tests examining these differences were not reported,
higher incidence rates were observed for all cancers combined, cancer of the lip, cancer of the
oral cavity,  and lung and bladder cancer among males in Chapaevsk. Considerably lower cancer
incidence rates also were observed for prostate cancer, cancer of the esophagus, and
leukemia/lymphoma among males from Chapaevsk.  Among females, incidence rates were
higher in 1998 for all cancers in Chapaevsk when compared to Russia and the Samara region, an
observation that appears somewhat counter to the presented SMR of 0.9 for all cancer mortality
from 1995-1998.  Similar to the mortality findings, rates of breast and cervical cancer incidence
among women in Chapaevsk were higher than in Russia. Leukemia/lymphoma rates were higher
among women in Chapaevsk than the reference populations of Samara and Russia.  This finding
is contrary to the results for males where lower rates of leukemia/lymphoma were observed in
Chapaevsk.

C.I.1.1.5.1.2.  Study evaluation
      Although the Revich et al. (2001) findings suggest TCDD exposures in Chapaevsk are
quite high relative to other parts of the world (Akhmedkhanov et al., 2002), the evaluation of
health outcomes to date is based on ecological data.  One limitation is that insufficient details are
provided by the authors to gauge the completeness and coverage of the cancer registry and
mortality data. Given the ecological nature of the data, the authors did not adjust for the
influence of other risk factors (e.g., smoking, reproductive characteristics) that could contribute
to increased cancer rates for lung cancer in men and breast cancer in women. In addition,
occupational exposures  may have also contribute to these SMR and SIR differences for cancer
outcomes that varied considerably between men and women.
                                         C-60

-------
       Future research in Chapaevsk includes plans to conduct a breast cancer case-control
study.  Women who were born from 1940 onward and who have been diagnosed with breast
cancer before the age of 55 were included in the study, although the plan to characterize TCDD
using serum is uncertain (Revich et al., 2005).

C.I.1.1.5.1.3.  Suitability of data for TCDD dose-response modeling
       This study did not meet most of the study considerations and criteria for inclusion in a
quantitative dose-response assessment. Given the lack of exposure data on an individual basis,
no effective dose can be estimated for this study population. Therefore, no dose-response
modeling was conducted for this study.

C.l.1.1.6.  The Air Force Health ("Ranch Hands" cohort) study
       Between 1962 and 1971, the U.S. military sprayed herbicides over Vietnam to destroy
crops that opposition forces depended upon, to clear vegetation from the perimeter of U.S. bases,
and to reduce the ability of opposition forces to hide.  These herbicides were predominantly a
mixture of 2,4-D; 2,4,5-T; picloram; and cacodylic acid (Committee to Review the Health
Effects in Vietnam Veterans of Exposure to Herbicides, 2006). A main chemical sprayed was
Agent Orange, which was a 50% mixture of 2,4-D and 2,4,5-T.  TCDD was produced as a
contaminant of 2,4,5-T and had levels ranging from 0.05 to 50 ppm (Committee to Review the
Health Effects in Vietnam Veterans of Exposure to Herbicides, 1994).  A series of studies have
investigated cancer outcomes among Vietnam veterans.  A review of military records to
characterize exposure to Agent Orange led Stellman and Stellman (1986) to conclude that
assignment of herbicide levels should  not be based solely on self-reports or a crude measure such
as military branch or area of service within Vietnam.  Investigations have been performed on the
Ranch Hands cohort, which consisted  of those who were involved in the aerial spraying of
Agent Orange between 1962 and 1971. More elaborate methods were used to characterize
exposures among these individuals, and these studies are summarized below.
                                         C-61

-------
C.l.1.1.6.1.  Akhtar et al (2004)
C.I.1.1.6.1.1.  Study summary
       Akhtar et al. (2004) investigated the incidence of cancer in the Ranch Hand cohort. The
Ranch Hand Unit was responsible for aerial spraying of herbicides, including Agent Orange, in
Vietnam from 1962 to 1971.  Cancer incidence in the Ranch Hand cohort was compared to a
cohort that included other Air Force personnel who served in Southeast Asia during the same
period but were not involved in the spraying of pesticides. Study participation was voluntary,
but there was no indication of the participation rate for either the Ranch Hand cohort or the
comparison group.  Health outcomes were identified during the postservice period that extended
from the time each  veteran left Southeast Asia until December 31, 1999.  The Akhtar et al.
(2004) study took into account concerns that both the comparison and spraying cohorts had
increased risks of cancer, and addressed the possibility that workers with service in Vietnam or
Southeast Asia might have increased cancer risk.  The authors addressed the latter concern by
adjusting risk estimates for the time spent in Southeast Asia and for the proportion of service
time spent in Vietnam.
       The Ranch Hand cohort comprised 1,196 men, and the comparison cohort had
1,785 men. The comparison  cohort was selected by matching date of birth, race, and occupation
(i.e., officer pilot, officer navigator, nonflying officer, enlisted flyer, or enlisted ground
personnel). TCDD levels were determined using serum levels collected from veterans who
completed a  medical examination in 1987. Blood measures also were taken in 1992, 1997, and
2002 for subjects with no quantifiable TCDD levels in 1987, those who refused in 1987, and
those new to the study; however, the 2002 data were not available for the Akhtar et al (2004)
analyses. For those who did not have a serum measure taken in 1987, but provided one in
subsequent years, TCDD levels were back-extrapolated to 1987 using a first-order kinetic model
that assumed a half-life of 7.6 years.  Those with nonquantifiable levels were assigned a value of
the limit of detection divided by the square root of 2. A total of 1,009 and 1,429 individuals in
the Ranch Hand and comparison cohorts, respectively, provided serum measures that were used
in the risk assessment.  Veterans also were categorized according to the time their tours ended.
This date corresponded to changes in herbicide use. These categories were before 1962 or after
1972 (no herbicides were used), 1962-1965 (before Agent Orange was used), 1966-1970 (when
Agent Orange use was greatest), and 1971-1972 (after Agent Orange was used).  Information on

                                         C-62

-------
incident cases of cancer in the cohort was determined from physical examinations and medical
records. Some malignancies were discovered at death and coded by using the underlying cause
of death as detailed on the death certificate. A total of 134  and 163 incident cases of cancer were
identified in the Ranch Hand and comparison cohorts, respectively.  Akhtar et al. (2004) describe
case ascertainment verified by record review as being complete.
       External comparisons were made based on the expected cancer experience derived from
U.S. national rates by using SIRs and their corresponding 95% confidence intervals. Incident
events and person-year contributions per group were tabulated by 5-year calendar and age
intervals.
       When compared to the general population, no statistically  significant excesses  in all
cancer incidence were observed for either the Ranch Hand (SIR = 1.09, 95% CI = 0.91-1.28) or
the comparison cohort (SIR = 0.94, 95% CI = 0.81-1.10).  Statistically significant differences
were found for three site-specific cancers in the Ranch Hands cohort relative to the general
population. Excesses were noted for malignant melanoma (SIR = 2.33, 95% CI = 1.40-3.65)
and prostate cancer (SIR = 1.46, 95% CI = 1.04-2.00).  In contrast, a reduced SIR was found for
cancers of the digestive system (SIR = 0.61, 95% CI = 0.36-0.96). The excess in prostate cancer
was also noted in the comparison cohort (SIR =  1.62, 95% CI = 1.23-2.10) relative to the
general population. External comparisons were repeated by restricting the cohorts to the period
when  Agent Orange was used (1966-1970).  Again, no statistically significant excesses in all
cancer incidence were noted in the Ranch Hand veterans (SIR = 1.14, 95% CI = 0.95-1.37) or in
the comparison cohort (SIR = 0.94, 95% CI = 0.80-1.11).  Statistically significant excesses
persisted for malignant melanoma (SIR = 2.57, 95% CI = 1.52-4.09) and prostate cancer
(SIR = 1.68, 95% CI = 1.19-2.33) in the Ranch Hand veterans.  No other statistically significant
differences were found among Ranch Hands personnel.
       For internal cohort analyses, veterans were assigned to one of four exposure categories.
Those in the comparison cohort were assigned to the "comparison category." Ranch Hand
veterans that had TCDD serum levels <10 ppt were assigned to the "background" category.
Those with a TCDD levels >10 ppt had their TCDD level estimated  at the end of their Vietnam
service with a first-order kinetic model that used a half-life of 7.6  years. These
back-extrapolated values that were less than  118.5 ppt were assigned to a "low" exposure group,
while those with values above 118.5 ppt were classified as "high" exposure. Akhtar et al. (2004)
                                          C-63

-------
used Cox regression models to describe risks across the exposure groups using the comparison
category as the reference. Risks were adjusted for age at tour, military occupation, smoking
history, skin reaction to sun exposure, and eye color.  Internal cohort analyses were restricted to
those who spent no more than 2 years in Southeast Asia and Ranch Hand workers who served
exclusively in Vietnam, and the comparison cohort who served exclusively outside of Vietnam.
       Statistically significant excesses of cancer incidence (all sites combined) were observed
in the highest two exposure groups. A statistically significant trend (p = 0.04) was detected
based on the RRs for the background, low, and high exposure groups: 1.44
(95% CI = 0.82-2.53); 2.23 (95% CI = 1.24-4.00), and 2.02 (95% CI= 1.03-3.95).  For
malignant melanoma, a statistically significant trend (p = 0.004) was detected, and the RRs
across the three increasing exposure categories were 2.99, 7.42, and 7.51, with statistically
significant results for the low and high exposure groups. The corresponding risk estimates for
prostate cancer were 1.50, 2.17, and 6.04 with statistically significant results only detected for
the high exposure group.

C.I.1.1.6.1.2.  Study evaluation
       An important strength of this study is the manner in which TCDD exposure was
estimated.  Serum data were available for most veterans, and therefore, generalizing exposure
from a  small sample of cohort members is not a concern as was the case with the NIOSH and
Hamburg cohorts. Back-extrapolating to derive past exposures was based on a methodology that
has been applied in many of the cohorts, thereby facilitating risk comparisons. An additional
strength of the study is the examination of cancer incidence as a measure of disease occurrence
rather than mortality. There is limited potential for gauge how representative the study
participants were given the lack of information provided on participation rates for either the
Ranch Hands or the comparison group. The analysis by Akhtar et al. (2004) was restricted to
individuals who spent no more than 2 years in Southeast Asia. Previous research had
demonstrated that increased time spent in Southeast Asia was associated with an increased risk
of cancer.  Confounding might have been introduced given that the comparison cohort spent
much more time  in Southeast Asia than the Ranch Hands. To illustrate, the median number of
days spent in Southeast Asia was 790 for comparison  cohort members, and the median days for
the Ranch Hand cohort in the background, low, and high exposure groups were 426, 457, and

                                          C-64

-------
397, respectively. After restricting to those who spent at most 2 years, statistically significant
associations were observed for all cancer sites combined, prostate cancer, and malignant
melanoma using the internal cohort comparisons.
       Given that 2,4,5-T and 2,4-D were used in equal concentrations in Agent Orange, there is
some concern regarding the ability to distinguish independent health effects for TCDD from
coexposures to these two herbicides.  However, in a large cohort study, called the Agricultural
Health Study, these herbicides were 2 of 50 pesticides and herbicides evaluated in a cohort of
more than 55,000 (mostly male) pesticide applicators in the United States and more than
33,000 spouses.  Although statistically significant associations were shown between prostate
cancer and several individual pesticides in this cohort (Alavanja et al., 2005), neither 2,4,5-T nor
2,4-D was associated with prostate cancer in that study (Alavanja et al., 2003): no associations
were found for these 2 herbicides and lung cancer either (Alavanja et al., 2004).  Therefore,
based on these Agricultural Health Study results, the dose-response relationship detected for
prostate cancer in the Akhtar et al. (2004) Ranch Hands study seems unlikely to be due to 2,4-D
or 2,4,5-T exposures.

C.I.1.1.6.1.3.  Suitability of data for TCDD dose-response modeling
       The ascertainment of incident cases and characterization of exposure to TCDD based on
serum measures are strong features of the cohort. Based on findings from another study
(Alavaniaetal., 2005: 2004: 2003), confounding by 2,4-D and 2,4-T does not appear likely to be
responsible for the exposure-response relationships found for prostate cancer and TCDD
exposures.  Therefore, this study was found suitable for quantitative TCDD  dose-response
analysis.

C.l.1.1.6.2.  Michalek andPavuk (2008)
C.I.1.1.6.2.1.  Study summary
       Michalek and Pavuk  (2008) published an updated analysis of the  incidence of cancer and
diabetes in the cohort of Ranch Hand veterans.  As with the Akhtar et al. (2004) analysis, the
study included a comparison cohort of other Air Force veterans who served  in Southeast Asia at
the  same time but were not involved with the spraying of herbicides.  This study extended
previous analyses (Akhtar et al.. 2004: Henriksen et al.. 1997) by stratifying the results by the

                                          C-65

-------
number of days of herbicide spraying, calendar period of service, and the time spent in Southeast
Asia.  Veterans who attended at least one of five examinations were eligible for inclusion.
Incident cancer cases also were identified from medical records.
       The methods used to determine TCDD exposures were as described above in the review
of the Akhtar et al. (2004) study. Blood measures taken in 1992, 1997, and 2002 were all
included in this new analysis. The study report did not provide the number of men with
measurements  at the different time points or the number who refused to partake at any time
point. TCDD dose at the end of service in Vietnam was assigned to Ranch Hands that had
TCDD levels above background using a first-order kinetic model and constant half-life of
7.6 years.  Each veteran was then assigned to one of four dose categories: comparison veteran,
background (i.e., Ranch Hands with 1987 levels of TCDD <10 ppt), low (Ranch Hands with
1987 levels of TCDD >10-91 ppt), and high (Ranch Hands with 1987 levels of TCDD >91 ppt).
Serum TCDD estimates were available for 1,597 veterans (men) in the comparison cohort, and
986 veterans (men) in the Ranch Hand cohort. The comparison cohort was selected by matching
on date of birth, race, and military occupation of the Ranch Hands.
       Michalek and Pavuk (2008) used Cox regression to characterize risks of cancer incidence
across the three upper exposure categories using the comparison cohort as the referent group.
Risk estimates were adjusted for year of birth, race, smoking, body mass index at the qualifying
tour, military occupation,  eye color, and skin  reaction to sun exposure. Tests for trend for
increased risk of cancer were conducted by testing the continuous covariate logioTCDD.
       Without stratification, no association between the TCDD exposure categories and RR of
all-site cancer incidence was observed.  Those in the highest exposure group had an RR of 0.9
(95% CI = 0.6-1.4).  Stratified analyses by calendar period of service  showed a more
pronounced risk for those who served before  1986 (when higher amounts of Agent Orange were
used). A statistically significant dose-response trend (p < 0.01) was observed for cancer risk and
loglOTCDD exposure. The RRs for the background, low, and high groups used in these
comparisons were 0.7 (95% CI = 0.4-1.3) with/? = 0.26, 1.7 (95% CI  = 1.0-2.9) with/? = 0.03,
and 1.5 (95% CI = 0.9-2.6) with/? = 0.14.  The strongest statistically significant increase,
however, was noted when analyses were restricted to those who had served before  1968, had
sprayed for at least 30 days before 1967, and had spent less than 2 years in Southeast Asia.  A
RR of 1.4 (95% CI = 1.1-1.7) per log(TCDD) exposure was detected (trend test/? = 0.005)
                                         C-66

-------
among this subgroup, while categorical exposures also suggested associations in the Low
(RR=1.7, 95% CI = 0.8-3.5) and High (RR=2.2, 95% CI = 1.1-4.4) groups relative to the
comparison group.

C. 1.1.1.6.2.2.  Study evaluation
       Michalek and Pavuk (2008) used the same study population as Akhtar et al. (2004), and
so it shares the same basic strengths and limitations as noted above.  The follow-up, however,
extends an additional 5 years (until the end of 2004), resulting in additional cancer data for
analysis and the inclusion of the serum data from 2002. Also, in this study, all analyses were
further adjusted for the number of days of spraying, which had not been done before. The
findings for the dose-response analyses were not as compelling as the earlier Akhtar et al. (2004)
findings, which was due in part to increased cancer risks in 2005 in the comparison cohort with
years spent in SEA.

C.I. 1.1.6.2.3.  Suitability of data for TCDD dose-response modeling
       As stated above for the Akhtar et al. (2004) study, the ascertainment of incident cases and
characterization of exposure to TCDD based on serum measures are strengths of the cohort.  In
addition, newer data and additional statistical adjustments improved the strength of the analysis.
This study, Michalek and  Pavuk (2008), was suitable for quantitative dose-response analysis of
TCDD.

C. 1.1.1.7.  Other studies of potential relevance to dose-response modeling
C.I.1.1.7.1. Hooiveld et  al (1998)—Netherlands workers
C.I.1.1.7.1.1.  Study summary
       Hooiveld et al. (1998) reanalyzed the mortality experience of a cohort of workers
employed in two chemical plants in the Netherlands using 6 additional years of follow-up from
an earlier study (Bueno de Mesquita et al., 1993). The cohort consisted of those employed
between 1955 and June 30, 1985, and vital status was ascertained until December 31, 1991 (i.e.,
36 years of follow-up). These cohort members were involved in the synthesis and formulation of
phenoxy herbicides, of which the main product was 2,4,5-trichlorophenoxyacetic acid and
monochloroacetic acid. This cohort, with a shorter follow-up interval than the original study
                                          C-67

-------
ft' Mannetje et al., 2005), was included in the IARC international cohort.  The cohort consisted
of 1,167 workers, of which 906 were alive at the end of the follow-up.  The average length of
follow-up was 22.3 years, and only 10 individuals were lost to follow-up.
       The authors used detailed occupational histories to assign exposures. Workers were
classified as exposed to phenoxy herbicides or chlorophenols and contaminants if they worked in
selected departments (i.e., synthesis, finishing, formulation, packing, maintenance/repair,
laboratory, chemical effluent waste, cleaning, shipping-transport, or plant supervision); were
exposed to the accident in 1963; or were exposed by proximity (i.e., if they entered an exposed
department at least once a week).  The 1963 accident was the result of an uncontrolled reaction
in the autoclave in which 2,4,5-trichlorophenol was synthesized; an explosion resulted, with
subsequent release of PCDDs that included TCDD. Based on these methods of exposure
assignment, 562 workers were deemed to be exposed to phenoxy herbicides or chlorophenols,
and 567 were unexposed.  Due to limited information, exposure could not be determined for
27 workers.
       TCDD exposures also were assigned using  serum measured on a sample of workers who
were employed for at least 1 year and started working before 1975. DLCs including PCDDs
were also measured in the serum samples but were not analyzed for this study. Of the
144 subjects who were invited to provide samples,  94 agreed.  TCDD levels were
back-extrapolated to the time of maximum exposure using a one-compartment, first-order kinetic
model that used a half-life estimate of 7.1 years.  The mathematical model used was
In(TCDDmax) = In(TCDD) + lag  x ln(2)/7.1.  The  lag was defined as the number of years since
last exposure for those exposed by virtue of their normal job duties. For those exposed as a
result of the accident in 1963, the lag was defined as the number of years since the accident
occurred.
       The authors made external comparisons of cohort mortality to the Netherlands population
using SMRs.  Poisson regression was used to perform internal cohort comparisons using
unexposed workers as the referent. RRs (measured using rate ratios) generated from the Poisson
model also were used to compare mortality based on low, medium, and high TCDD
serum-derived categories. The Poisson model included the following covariates as adjustment
factors: age, calendar period at end of follow-up, and time since first exposure.
                                         C-68

-------
       When compared to the general population, workers had an excess mortality from cancer
(SMR = 1.5, 95% CI = 1.1-1.9), based on 51 cancer deaths.  Generally, no excesses were
observed for site-specific cancers.  The exception included eight deaths from cancers of the
urinary organs (SMR = 3.9, 95% CI = 1.7-7.6). Although not statistically significant, SMRs
comparable in magnitude to other studies were detected for non-Hodgkin lymphoma
(SMR = 3.8, 95% CI = 0.8-11.0) and Hodgkin disease (SMR = 3.2, 95% CI = 0.1-17.6). A
statistically significant excess of cancer mortality (n = 20 deaths among workers) also was
observed relative to the general population when analyses were restricted to those exposed from
the 1963 accident (SMR = 1.7, 95% CI = 1.1-2.7). Three deaths from prostate cancer were also
noted among these workers (SMR = 5.2, 95% CI = 1.1-15.3), but no excess was observed with
any other cancer site.
       Internal cohort comparison also demonstrated an increased risk of all  cancer mortality
among those exposed to phenoxy herbicides, chlorophenols, and contaminants relative to those
unexposed (RR = 4.1, 95% CI = 1.8-9.0). A statistically significant increased risk was also
noted for respiratory cancer mortality (RR = 7.5, 95% CI = 1.0-56.1).  Analyses across
categories of TCDD exposure revealed excesses in cancer mortality for all cancer sites
combined; however, no dose-response trend was apparent.

C.I.1.1.7.1.2.  Study evaluation
       Several other studies that have characterized cohorts  by TCDD levels have used the area
under the curve approach and thus have derived an exposure metric that is time dependent.
Hooiveld et al. (1998) instead created an exposure metric to  capture the maximum exposure
attained during the worker's employment. Characterizing risks using this metric assumes that
other TCDD exposures accrued during a workers'  lifetime are not relevant predictors of cancer
risk.

C.I.1.1.7.1.3.  Suitability of data for TCDD dose-response modeling
       One study limitation is that, although DLCs were measured in the serum samples,
mortality associations were reported for TCDD only. There  is some utility in examining
dose-response analyses using the alternative exposure metrics that were constructed for this
cohort. However, the small number of identified cancer deaths, exposure assessment limitations

                                          C-69

-------
(based on a nonrepresentative sample, and maximum exposure level) and concern over potential
confounding by coexposures preclude using these data for a dose-response analysis.

C.I.1.1.7.2.  t' Mannetje et al.  (2005)—New Zealand herbicide sprayers
C.I.1.1.7.2.1.  Study summary
      t' Mannetje et al. (2005) described the mortality experience of a cohort of New Zealand
workers who were employed in a plant located in New Plymouth.  The plant produced phenoxy
herbicides and pentachlorophenol between 1950 and the mid-1980s.  This study population also
was included in the international cohort of producers and sprayers of herbicides that was
analyzed by IARC (Kogevinas et al., 1997; Saracci et al., 1991). In this 2005 study, analyses
were restricted to those who had worked at least 1 month; clerical, kitchen, and field research
staff were excluded. The authors followed up 1,025 herbicide producers and 703 sprayers from
1969 and 1973, respectively, until the end of 2000.
      The cohort consisted of two components: those involved with the production of
herbicides and those who were sprayers.  For the herbicide producers, exposures were
determined by consulting occupational history records; no direct measures of exposure were
available. Each department of employment was assigned to one of 21 codes as in the IARC
international cohort (Saracci etal., 1991). Industrial hygienists and factory personnel with
knowledge of potential exposures in this workforce classified each job according to potential to
be exposed to TCDD, other chlorinated dioxins, and phenoxy herbicides.  Exposure was defined
as a dichotomous variable (i.e.,  exposed and unexposed). Among  producers, 813 (713 men and
100 women) were classified as exposed, with the remaining 212 (gender not specified)
considered unexposed.
      The "sprayer" component of the cohort includes those who were registered in the national
registry  of applicators at any time from January 1973 until the end of 1984. For the sprayers,
detailed occupational information was lacking.  Exposure was, therefore, based on an exposure
history questionnaire completed in a previous study of congenital malformations (Smith et al.,
1982). This questionnaire, administered to 548 applicators in 1980 and 232 applicators in 1982,
achieved a high response rate (89%). Participants were asked to provide information about
2,4,5-T-containing product use on an annual basis from 1969 up to the year the survey was
completed.  As the use of 2,4,5-T ceased in the mid-1980s, data on occupational  exposure to

                                         C-70

-------
TCDD among these workers are fairly complete.  Virtually all sprayers (699 [697 men and
2 women] of 703) were deemed to have been exposed to TCDD, higher chlorinated dioxins, or
phenoxy herbicides.
       Deaths among workers were identified through record linkage to death registrations in the
New Zealand Health Information Service.  Electoral rolls, drivers' licenses, and social security
records also were consulted to confirm  identified deaths. External comparisons of mortality
were made to the New Zealand population using the SMR statistic. The mortality follow-up for
the producers began on January 1, 1969 and extended until December 31, 2000.  For the
sprayers, the follow-up period extended from January 1, 1973 until December 31, 2000.  A total
of 43 cancer deaths occurred in the producer group and 35 cancer deaths occurred in the sprayer
group in the cohort.  Stratified analyses by duration of employment and department were
conducted. The departments examined for producers included synthesis, formulation and lab,
maintenance and waste, packing and transport, other, and unexposed. SMRs were generated
using the New Zealand population as an external  referent. A linear test for trend was applied to
evaluate dose-response trends according to categories of duration of employment. Stratified
analyses also were also done for sprayers who started working before 1973, as TCDD levels in
2,4,5-T produced at the New Zealand plant dropped dramatically after 1973.  Although an SMR
was presented for female producers, given that only one cancer death was observed, this study
can provide no insight on differential risks between the sexes.
       Among TCDD-exposed producers, for all cancers combined, no statistically significant
excess in mortality was found when compared to the general population (SMR = 1.24,
95% CI =  0.90-1.67). No dose-response trend in the SMRs for all cancers was observed with
duration of employment (p = 0.44).  No statistically significant elevated SMR was observed in
any of the duration of employment categories for any of the six specific departments examined.
A statistically significant positive linear trend, however, was noted among synthesis workers
(p = 0.04). There was some suggestion of reduced mortality in the upper exposure levels for
workers in the formulation and lab departments. For sprayers, the SMR for all cancer sites
combined was not elevated relative to the New Zealand  general  population (SMR = 0.82,
95% CI =  0.57-1.14), nor was a dose-response pattern observed with increasing  duration of
employment (p = 0.86). Additionally, no statistically significant excess in cancer mortality for
all sites combined was evident in workers who were first employed either before 1973
                                         C-71

-------
(SMR = 0.75, 95% CI = 0.50-1.07) or from 1973 onwards (SMR =1.81, 95% CI = 0.59-4.22).
For site-specific cancer mortality, an excess of multiple myeloma was observed among
production workers relative to the general population (SMR =5.51, 95% CI = 1.14-16.1). This
SMR was based on three deaths. No statistically significant excess (or deficit) of mortality was
found for any other cancer site examined in either the sprayers or the producers.

C. 1.1.1.7.2.2.  Study evaluation
       The physical activity demands of spraying contribute to a healthy worker effect that
manifests itself in a lower SMR based for both external comparisons to the general population as
a referent, and that generated relative to the producers in the cohort.  The lack of individual-level
TCDD data resulted in the analyses being based upon job title and duration of employment.
Thus, intra-cohort comparisons were precluded due to a lack of an unexposed group (e.g., the
sprayers), limited exposure contrasts and the small number of cancer deaths.
       The dose-response pattern with duration of employment coupled with the observation
that higher levels of exposure to TCDD occurred among workers in the synthesis department is
an important finding. These workers were, however, also  exposed to several other contaminants
that include processing  chemicals, technical products, intermediates, and byproducts (Kauppinen
et al., 1993). These included phenoxy herbicides and DLCs such as chlorinated dioxins.  Since
the dichotomous exposure measure was based on exposure to TCDD, chlorinated dioxins and
phenoxy herbicides, the associated dose-response analyses presented in this study should be
interpreted cautiously in light of the inability to either characterize or control for these potential
confounders.  As such, these coexposures might have contributed to the dose-response pattern
observed with increased duration of employment in the synthesis workers.

C.I. 1.1.7.2.3.  Suitability of data for TCDD dose-response modeling
       Although the study authors completed a subsequent analysis of this cohort using
serum-derived TCDD (McBride et al., 2009b), the lack of individual-level TCDD exposures
precludes dose-response modeling.
                                          C-72

-------
C.I.1.1.7.3.  McBride et al. (2009b)—New Zealand herbicide sprayers
C.I.1.1.7.3.1.  Study summary
       McBride et al. (2009b) recently published the mortality experience of the New Zealand
cohort in relation to serum estimates of TCDD levels.  This study included  1,599 workers who
were employed between 1969 and November 1, 1989,  which was the date that 2,4,5-T was last
used.  The study report does not specify how many of the individuals were men or women, but
using the percentage that were men lost to follow-up (73% of 1,261 were men) and not lost to
follow-up (76% of 338 were men) would indicate 1,001 men and 598 women were included in
the original cohort.  As in their study published earlier in the same year (McBride et al., 2009a),
the follow-up period extended from the first day of employment until December 31, 2004. Vital
status was ascertained through record linkage to the New Zealand Health Information Service
Mortality Collection and the Registrar General's Index to Deaths for deaths up to 1990.
       All current and former workers who lived within 75 km of the plant were invited to
provide serum samples.  A total of 346 of the eligible workers (68%, gender not specified)
provided samples, which represented 22% of the overall study population (346/1,599). Based on
the serum measures, 70% (241/346) had been exposed to TCDD. This percentage is similar to
the estimated 71% of workers who were deemed to have been exposed based on a review of
occupational records.  The mean serum TCDD value was 9.9 ppt. The highest exposures were
observed for those employed in the trichlorophenol operation (23.4 ppt). Values among
unexposed workers averaged 4.9 ppt, which is close to the background level of 3.9 ppt among
individuals of similar age in the New Zealand general population (Bates et al., 2004). Details on
smoking histories of individuals were also collected for the 346 individuals who provided serum,
allowing for  an examination of the potential confounding influence that smoking might have on
derived risk estimates for TCDD.
       Cumulative exposure to TCDD, as a time-dependent metric, was estimated for each
worker. A detailed description of the methods used to derive TCDD exposure was described in
Aylward et al. (2009). The qualitative TCDD scores available for those with serum measures
were used to estimate the cumulative exposures based  on a half-life of 7 years.  A
time-dependent estimate of TCDD exposure was derived and the area under the curve was used
to estimate cumulative workplace TCDD exposures above background levels.  Model
performance appeared modest as the model explained only 30% of the variance (adjusted R )

                                        C-73

-------
when these TCDD exposure estimates were compared with actual serum levels (Avlward et al.,
2009).
       As with previous analyses of the cohort (McBride et al., 2009a: f Mannetje et al., 2005),
external comparisons to the New Zealand general population were made using the SMR. The
SMR also was used to compare mortality across four exposure groups relative to the general
population, as defined by the serum TCDD estimates: 0-68.3, 68.4-475.0, 475.1-2085.7, and
>2085.8 ppt-month.  The proportional hazards model also was used to conduct internal cohort
comparisons across these same four exposure groups. In these analyses, age was used as the
time variable, and the covariates of date of hire, sex, and birth year were included in the
proportional hazards model. The cut-points for these four exposure categories were chosen so
that approximately equal numbers of deaths were included in each category.
       Consistent with earlier SMR analyses of the same cohort, no increased cancer mortality
was observed among "ever" exposed workers when compared to the general population
(SMR =1.1, 95% CI = 0.9-1.4). No statistically significant excess  was noted for any of the
site-specific cancers, although there was some suggestion of increased risk of soft tissue sarcoma
(SMR = 3.4, 95% CI = 0.1-19.5), multiple myeloma  (SMR = 2.2, 95% CI = 0.2-8.1),
non-Hodgkin lymphoma (SMR = 1.6, 95% CI = 0.3-4.7), and cancer of the rectum (SMR = 2.0,
95% CI = 0.7-4.4). No statistically significant increase in cancer mortality (all sites combined)
was found in any of the four exposure categories as measured by the SMR statistic, nor was a
dose-response trend noted with increasing exposure categories. No  dose-response trends (based
on SMR analyses) were noted  for five site-specific cancers examined (i.e., digestive organs,
bronchus, trachea and lung, soft tissue sarcomas, lymphatic and hematopoietic tissue, and
non-Hodgkin lymphoma), although SMRs for three of the four exposure categories exceeded 2.0
for non-Hodgkin lymphoma.
       In contrast to the external cohort comparisons, the RRs generated with the proportional
hazards model supported a dose-response trend, as rate ratios increased across increasing TCDD
exposure categories. The RRs and 95% confidence intervals for all  cancer mortality relative to
the lowest of the four groups were 1.05  (95% CI = 0.48-2.26), 1.38 (95% CI = 0.64-2.97) and
1.58 (95% CI = 0.71-3.52).  Neither the linear (p = 0.29) or quadratic (p = 0.82) test for trend,
however, was statistically significant. An increased risk of lung cancer mortality was observed
in the highest TCDD exposure category relative to the lowest although the precision of this risk
                                         C-74

-------
estimates was poor and was not statistically significant (RR = 5.75, 95% CI = 0.76-42.24). The
test for trend for lung cancer also was not statistically significant.
       A smoking survey was administered to a sample of surviving workers of this cohort, and
smoking prevalence was found to be slightly higher among those with higher cumulative
exposure (61%) compared to lower exposures (51-56%). These minor differences in smoking
prevalence were unlikely to explain the five-fold increase in risk of lung cancer found in the
highest exposure category. Although the smoking data assessment was a strength of the study, it
was limited to only sample of workers and was not available for those who died of lung cancer,
or other causes of death.

C. 1.1.1.7.3.2.  Study evaluation
       Given high rates of emigration, loss to follow-up (21%) was a potential concern in this
study. If comparable emigration rates did occur among the general population then the SMRs
would be underestimated. It is unclear to what extent emigration occurred among the general
population and whether emigration in both the worker and general populations was dependent on
health status.  If emigration rates were comparable among these two populations, the associated
bias from the under-ascertainment of mortality in the lost to follow-up group would likely
attenuate a positive association between TCDD and cancer mortality.  Among the worker
population, there was not much evidence of differential loss to follow-up with respect to
exposure as average exposures were lower (3.2 ppt) among those loss to follow up compared to
those with complete follow-up (5.7 ppt). Previous studies among this population also found
slightly higher loss to follow-up rates among the unexposed (23%) compared to the exposed
(17%) workers ft' Mannetie et  al.. 2005).
      McBride et al. (2009b) did not present results using a continuous measure of TCDD
exposure (lagged or unlagged) as was done in most other occupational cohorts. Additionally, the
modeling did not consider the use of different periods of latency.

C.I. 1.1.7.3.3.  Suitability of data for TCDD dose-response modeling
       There was limited evidence of dose-response relationships  between TCDD exposure and
the cancer outcomes that were examined. There is also no evidence that the authors considered
exposure metrics that are consistent with environmental cancer-causing agents such as exposure

                                         C-75

-------
modeling that takes latency into account.  Given that past occupational cohort studies of
TCDD-exposed workers have consistently demonstrated stronger association with lag interval of
15 years, such an approach should be applied to this cohort. This precludes this study from
consideration for quantitative dose-response modeling.

C.I.1.1.7.4.  McBride et al. (2009a)—New Zealand herbicide sprayers
C.I.1.1.7.4.1.  Study summary
       McBride et al. (2009a) published an updated analysis of the mortality of the New Zealand
cohort. The follow-up period was from January 1, 1969 to December 31, 2004 extending the
previous study by an additional 4 years.  In contrast to the previous study where the cohort
comprised individuals employed for at least 1 month prior to 1982 (or 1984) (f Mannetje et al.,
2005), the cohort in this study consisted of all those who worked at least one day between
January 1, 1969 and October 1, 2003.  This resulted in a cohort of 1,754 workers, of which
247 died in the follow-up interval.  Twenty-two percent of the cohort members were lost to
follow-up, which could be a source of selection bias if loss to follow-up was related to both the
exposure metrics and the health outcome of interest. Previous data from this cohort (f Mannetje
et al., 2005), however,  showed fairly comparable loss to follow-up among the unexposed (23%)
and the exposed populations (17%).
       Comparisons to the New Zealand general population were made using the SMR statistic.
Stratified analyses were conducted by duration of employment (<3 months, >3 months), sex,
latency (<15 years, >15 years), and period of hire (<1976, >1976). The authors defined latency
as the period between the day last worked and the earliest of date of death, date of emigration or
loss to follow-up,  or December 31, 2004.
       The overall SMR for mortality from all cancer sites combined relative  to the New
Zealand population was 1.01 (95% CI = 0.85-1.10). Although not statistically significant, there
was suggestion of an increased risk of rectal cancer (SMR = 2.03, 95% CI = 0.88-4.01). SMRs
for lymphatic and hematopoietic cancers (overall SMR =  1.21, 95% CI = 0.52-2.39) included
3.12 (95% CI = 0.08-17.37) for Hodgkin disease, 1.59 (95% CI = 0.43-4.07)  for non-Hodgkin
lymphoma, and 1.66 (95% CI = 0.20-5.99) for multiple myeloma.  No statistically significant
excess of cancer mortality was noted among workers employed for <3 months (SMR = 1.19,
95% CI = 0.65-2.00), or for >3 months (SMR = 0.98, 95% CI = 0.75-1.26). A statistically

                                         C-76

-------
significant excess of digestive cancers was found for those who worked fewer than 3 months
relative to the New Zealand population (SMR = 2.52, 95% CI = 1.15-4.78).  No excesses were
observed for any site-specific cancers when analyses were restricted to those who worked for 3
or more months. No statistically significant elevated SMRs were found for all cancers
(combined) either for a latency period of fewer than 15 years (SMR =1.14, 95% CI = 0.72-1.71)
or a latency period of >15 years (SMR = 0.96, 95% CI = 0.72-1.26).  Similarly, no statistically
significant excess in cancer mortality was observed for all cancer sites combined, or any
site-specific cancer when analyses were stratified by date of hire (<1976,>1976) or by sex. The
SMR among women who were employed at the site was 0.68 (95% CI = 0.45-1.00).

C.I.1.1.7.4.2.  Study evaluation
       High rates of emigration in New Zealand (9% among workers in the cohort) contributed
to a fairly high loss to follow-up (22% among workers)  during the study period. The loss to
follow-up would reduce the overall mortality estimates among the workers, which could
underestimate the SMRs if loss to follow-up (and health status) was not comparable in the
general population.  For example, it is unclear if workers and the general population who
emigrated were less healthy than those who did not. Previous data from the cohort suggests that
loss to follow-up rates were slightly higher among those with lower exposures (McBride  et al.,
2009b: t' Mannetie et al.. 2005).

C.I.1.1.7.4.3.  Suitability of data for TCDD dose-response modeling
       This study extended the mortality follow-up of an earlier study and included stratified
analyses to investigate effect modification by period of latency, sex, and date of hire.  A key
limitation was the lack of direct measures of exposure for study participants which precluded
estimating effective dose needed for dose-response modeling.  As such, this study did not meet
the considerations and criteria for inclusion in quantitative dose-response analysis.

C.I. 1.2.   Key Characteristics of Epidemiologic Cancer Studies
       Table C-l summarizes the key characteristics of the available epidemiologic studies of
TCDD exposure and cancer.  It compares the length of follow-up, latency period used, half-life
                                          C-77

-------
for TCDD used, and the fraction of TEQs accounted for by TCDD (when applicable) for each
study.

C. 1.1.3.  Feasibility of TCDD Cancer Dose-Response Modeling—Summary Discussion by
         Cohort
C. 1.1.3.1.  Using the NIOSH cohort in dose-response modeling
       It is important to evaluate the NIOSH cohort with respect to its suitability to conduct
dose-response modeling of TCDD and cancer. This cohort is the largest assembled to date,
direct measures of TCDD based on serum sampling are available, and the lengthy follow-up
interval allows for latent effects to be taken into account. Further, although this cohort consists
mostly of male workers, these workers were occupationally exposed to TCDD daily, as
compared to the acute accidental exposures of other occupational cohorts.  Although the most
recent analyses of a subset of the NIOSH cohort showed no association between serum TCDD
levels and cancer mortality, the exposure category cutpoints did not allow for examination of
health effects above levels for which associations had been observed in the larger NIOSH cohort
(Collins etal.. 2010: 2009).
       Most published studies of the NIOSH cohort did not evaluate exposures to DLCs. An
exception is the analysis by Steenland et al. (2001b). Although Steenland et al. (200Ib) did not
incorporate individual-level data on  DLCs, based on their previous work (Piacitelli etal., 1992)
they assumed that TEQ occupational exposures occurred as a result of TCDD alone in this
population. TCDD exposures provided a better fit to the data than the TEQ-based metric, and
15-year latencies improved the fit for both metrics (relative to unlagged exposures).  The lifetime
risk estimates for an increase in 10 TEQs (pg/kg of body weight/day/sex) ranged from
0.05-0.18%.  The value added for this measure is the incorporation of the contribution of other
DLCs to the background rates.
       Blue collar workers, such as  those in the NIOSH cohort, typically have higher rates of
smoking than the general population (Lee et al., 2007; Bang and Kim, 2001). This potential
source of confounding would be expected to produce a higher SMR for lung cancer mortality,
and could contribute to the excess noted in the cohort with longer lag intervals.  This bias,
however, likely is not large as no statistically significant excess of nonmalignant respiratory
mortality was found in these workers.  Any associated bias from smoking would be expected to
be smaller for comparisons conducted within the cohort, as fellow workers would be expected to
                                         C-78

-------
be more homogeneous with respect to their risk factor profile than with an external general
population referent group.  Stratified analyses using both internal and external comparison
groups also did not identify important differences in associations with TCDD exposure between
smoking and nonsmoking cancers. Thus, fatal cancer risk estimates reported for workers in the
NIOSH cohort appear to provide a reasonable estimate of the carcinogenic potency of TCDD.
       Although the Steenland et al. (200Ib) study did not directly account for the possible
confounding effects of other occupational exposure, the authors did address this source of
potential bias.  No known occupational exposures to carcinogens occurred, with the exception of
4-aminobiphenyl, which occurred at only one plant. Two deaths from mesothelioma also
occurred in the cohort, so some exposure to asbestos was possible (Fingerhut et al., 1991a). The
statistical analyses suggested that the inability to control for other occupational exposures would
not have unduly affected risk estimates generated from internal cohort comparisons.  For
instance, the removal of one plant at a time from the analysis did not materially change
dose-response estimates generated from the Cox model  (Cheng et al., 2006).  Moreover, adding a
variable to represent each plant in the Cox regression had little impact on the risk estimates.
Given that other occupational exposures varied by plant, a change in risk estimates would be
expected if such exposures were strong confounders.
       The Cheng et al. (2006) analysis provides important information about the impact of
applying kinetic models to the data. The CADM TCDD kinetic model resulted in dramatic
decreases in the TCDD cancer mortality risk estimates when compared to the one-stage
compartmental model that had been applied. Although Cheng et al. (2006) suggested that the
CADM provides a better fit to the data than the typically used simple one-compartmental model,
statistical comparisons of model fit were not reported.  Therefore, there is value in presenting the
range in risk estimates across different models when characterizing dose-response relationships.
       Finally, the half-life of TCDD is generally recognized to vary according to body fat
percentage, and this information was not available for the NIOSH workers. The inability to
account for between-worker variability in body fat would introduce exposure measurement error.
That body fat percentage would not be expected to correlate with cumulative exposure to TCDD
exposure, however, would limit the potential for misclassification bias.  The effect of any
nondifferential exposure measurement error likely would serve to attenuate the risk estimates of
the study.
                                          C-79

-------
C. 1.1.3.2.  Using the BASF cohort in dose-response modeling
       The availability of blood lipid data for TCDD allows for characterization of cumulative
TCDD exposures in the BASF cohort. TCDD blood lipid data were collected for 90% of the
surviving members of the cohort (138 of 154) and these serum measures were used to generate
TCDD exposure estimates for all 254 cohort members. Therefore, the potential for
misclassification error from extrapolating these exposures to the entire cohort is less likely than
for the NIOSH cohort where sera data were available for only a small fraction of workers.  These
BASF serum data were, however, collected long after the accident (36 years) and had to be
back-extrapolated to derive the initial exposures.
       The data on this cohort included several risk factors such as cigarette smoking and body
mass index.  One advantage is that cumulative TCDD levels by body mass index can be
estimated on an individual-level basis. As expected, the derived cumulative measures appear to
correlate well with severity scores of chloracne.  The finding that more pronounced risks were
found 15-20 years after first exposure are also consistent with findings from several other
cohorts (Bertazzi etal.. 2001: Fingerhut et al.. 1991b: Manzetal.. 1991).
       A key limitation of the BASF cohort is its relatively small sample size (n = 243), which
limits the ability to evaluate dose-response relationships for site-specific cancers.  Also, the
quality of the ascertainment of cancer incidence cannot be readily evaluated as the geographic
area of the cohort is not covered by a tumor registry. Ott and Zober (1996a) state that nonfatal
cancers could have been more likely to be missed in early years, which could partially contribute
to the higher standardized incidence ratio found for cancer with longer latencies. Commenting
on risk differences derived from incident and decedent cancer outcomes is difficult. Among
those comprising the cohort, the ascertainment of incident outcomes was recognized to be less
complete in early years. Although the ascertainment of mortality outcomes was generally
regarded to be good among the 243 workers, some workers who died or moved likely were
missed when the cohort was constructed.  These deaths would have been more likely to have
occurred several years before the second component of the cohort was assembled.
       The use of the SMR statistic for this study population is associated with important
sources of uncertainties.  Deaths were surely missed, particularly for the third component of the
cohort that accounts for approximately 38% (94/247) of the entire cohort; this factor would serve
to underestimate the overall SMR.  As mentioned before, this component of the cohort was

                                         C-80

-------
assembled through the recruitment of workers known to be alive in 1986. Despite this limitation,
the characterization of exposure data and availability of other risk factor data at an individual
level allow the development of quantitative dose-response analyses.

C.I. 1.3.3.  Using the Hamburg cohort in dose-response modeling
       The Hamburg cohort lacked data on cigarette smoking, and, therefore, effect estimates
could not be adjusted for this covariate.  Additional analyses that excluded lung cancers resulted
in an even stronger dose-response relationship between all cancer mortality and TCDD.  Serum
levels of TCDD also were also not associated with smoking status in a subgroup of these workers
(Flesch-Janys et al., 1995) suggesting that smoking unlikely confounds the association between
all cancer mortality and TCDD.
       An important limitation of the cohort is the reliance on blood and tissue measurements of
190 workers that likely represent a highly selective component of the cohort. This subset of
workers was identified at the end of the observation period, and therefore, excludes  workers who
died or could not be traced.  There are uncertainties in deriving department- and period-specific
estimates for a period that extends over three decades using this number of workers.
Additionally, the criteria applied to the reference population could have introduced some bias.
Workers were included only in the reference group if they had been employed for at least
10 years in a gas supply industry.  The criteria were much different for the workers who were
exposed to TCDD (only 3 months of employment). As a result, the reference group likely would
be more susceptible to the healthy worker effect. Internal cohort comparisons, which should be
void of such bias, however, generally produced results similar to those based on the external
comparison population. In summary, the Becher et al. (1998) study meets the criteria and
additional epidemiologic considerations which allowed for development of quantitative
dose-response analyses.

C.I. 1.3.4.  Using the Seveso cohort in dose-response modeling
       Unlike many of the occupational cohorts that were examined, data from the Seveso
cohort are representative of a residential population whose primary exposure was from a single
TCDD release. A notable exception is the BASF cohort where workers were exposed principally
                                          C-81

-------
through two accidents that occurred in the plant.  The Seveso data, therefore, might permit
cancer dose-response investigations in women and children.
       Uncertainty in identifying the critical exposure window for most of the outcomes related
to the Seveso cohort is a key limitation.  An important feature of the Seveso cohort, however, is
that TCDD levels were much lower among those in the highest exposure zones in Seveso
(medians range from 56-136 ng/kg) (Eskenazi et al., 2004) than those in the occupational
cohorts who had TCDD exposures that were sometimes more than 1,000 ng/kg.  Given these
dramatic exposure differences in exposures, the standardized mortality ratios (after incorporating
a 15-20 year latency period) for all cancer sites combined are remarkably similar between the
Seveso and the occupational cohort analyses. Perhaps more importantly, the data from Seveso
might be more relevant for extrapolating to lower levels, given that exposures to TCDD are
two orders of magnitude higher than background levels (Smith and Lopipero, 2001), and lower
than many of the exposures  observed in the other occupationally exposed cohorts.
       The Warner et al. (2002) study found a positive association between serum levels of
TCDD and breast cancer.  As noted previously, ascertainment of incident cases for all cancers
would allow for a dose-response relationship to be evaluated. Moreover, future breast cancer
analyses in this cohort that would increase sample size should strengthen the quantitative
dose-response analyses of this specific cancer site.  The strengths of the Warner et al. (2002)
study outlined earlier suggest that this study should be considered for cancer dose-response
modeling.
       Earlier Seveso studies likely are unsuitable for conducting quantitative risk assessment.
These previous studies used an indirect measure of TCDD exposure, namely, zone of residence.
Soil concentrations of TCDD varied widely in these three zones (Zone A: 15.5-580.4 ppt;
Zone B: 1.7-4.3  ppt; and Zone R: 0.9-1.4 ppt), which could have resulted in considerable
exposure misclassification.  The Warner et al. (2002)  study greatly improved the characterization
of TCDD exposure using serum measures, and also allowed for control  of salient risk factors that
may have resulted in bias  due to confounding.
       At this time it is unclear whether any study has examined the relationship between cancer
and serum estimates of TCDD among Seveso males exposed from the 1976 accident.
                                          C-82

-------
C.I. 1.3.5.  Using the Chapaevsk related data in dose-response modeling
       Currently, individual-level exposure data are lacking for residents of this area and there is
no established cohort for which cancer outcomes can be ascertained. These limitations,
therefore, preclude the inclusion of Chapaevsk data in a quantitative dose-response analysis.

C.I. 1.3.6.  Using the Ranch Hands cohort in dose-response modeling
       Study strengths of the Ranch Hand cohort includes a relatively large cohort with
individual-level serum measurements taken overtime in 1987, 1992, 1997, and 2002.  In
addition, TCDD levels for later years were back-extrapolated to 1987 using a first-order  kinetic
model that assumed a half-life of 7.6 years. Although the isolation of TCDD effects from those
of other agents found in Agent Orange raised some concerns about confounding, results  from a
large agricultural cohort found no association between 2,4-D or 2,5-T and prostate cancer or lung
cancer (Alavanja et al., 2005; 2004; 2003). It was determined that dose-response analyses would
be conducted on this population using both the (Michalek and Pavuk, 2008)) and Akhtar et al.
(2004) studies.

C. 1.1.4.  Discussion of General Issues Related to Dose-Response Modeling
C.I.1.4.1.  Ascertainment of exposures
       Several series of epidemiologic data have used serum measures to estimate TCDD
exposures.  Serum data offer a distinct advantage in that they provide an objective means to
characterize TCDD exposure at the individual level. The serum measures in the occupational
cohorts, however, are limited in two important ways. First, these samples are generally collected
from small  subsets of the larger cohorts; therefore, using these measures to extrapolate to the
remainder of the cohort could introduce  bias due to exposure misclassification. The
second limitation is related to estimating the half-life of TCDD. As noted previously, exposures
to TCDD were back-extrapolated several decades from the date that serum samples were
collected among surviving members of several cohorts.  This approach was used in the NIOSH,
Ranch Hands, BASF, New Zealand, and Hamburg cohorts.  The reported half-life of TCDD
among these populations was reported between 7.1 to 9.0 years and the half-life has been shown
to vary with several individual characteristics including age, body fat composition, and smoking.
The derivation of half-lives from a sample of workers, and application of these estimates to

                                         C-83

-------
retrospectively characterize exposure can introduce uncertainty into the lifetime exposure
estimates. It is important to note, however, that sensitivity analyses results in several studies
have been fairly consistent when evaluating the impact of half-life of TCDD (Steenland et al.,
2001b: Flesch-Janys et al., 1995). In addition, the reliance on surviving cohort members for
serum samples can introduce bias as it assumes their distribution of TCDD exposures was the
same among those who died.
       A unique advantage of the Seveso study is that serum measures were taken shortly after
the accident, and therefore characterization of TCDD exposure in this population does not
depend on assumptions needed to back-extrapolate exposures several decades.

C.I.1.4.2.  Latency intervals
       Many of the epidemiologic studies indicate stronger associations between TCDD and
cancer outcomes once a latency period has been considered.  Generally, risks are higher when a
latency period of 15-20 years is included.  As noted previously, this observation is consistent
with many other environmental carcinogens such as radon, radiation, and cigarette smoking.
That recent exposures do not contribute to increased cancer risk provides  some support that the
initiation and promotion phases might occur many years before death making recent exposures
irrelevant for these analyses.  The ability to discriminate between models  of varying latency,
however, was not possible in many  studies. The application of biologically-based modeling
could provide additional important insights on which phase(s) of carcinogenesis TCDD exerts an
influence. Such modeling, however, would necessitate having data  on an individual-level basis.
Ideally, this modeling would use cancer incident rather than mortality outcomes given that the
median survival time exceeds 5 years for many cancer sites.

C.l.1.4.3.  Use of the SMR metric
       The occupational cohorts and the studies in Seveso and Chapaevsk have relied on the
SMR to make inferences regarding the effects of TCDD on mortality.  When compared to the
general population, the healthy worker effect may result in a downward bias in the SMR. This
often can manifest as SMRs less than 1 for several causes of mortality.  The effect of this bias is,
however, generally smaller for cancer outcomes. Cancer outcomes, whether incidence or death,
                                          C-84

-------
typically occur later in life and do not generally affect an individual's ability to work at earlier
ages.
       There are several approaches that can be taken to minimize potential biases introduced by
the healthy worker effect, which would account for workers being healthier than the general
population. Comparisons of mortality (or cancer incidence) can be made to other cohorts of
similar workers.  If done properly, this can allow for some control of characteristics such as
sociodemographic characteristics and smoking as the two populations can be matched by these
factors.  However, it may be the case that other working populations are exposed to other
harmful exposures, thereby making it difficult to estimate risk associated with a specific agent
(such as TCDD) in the cohort of interest. A second and preferred approach to control for the
healthy worker effect, should it prove feasible, is to conduct comparisons of health outcomes in
relation to exposure within the cohort. These comparisons are less likely to be influenced by
other potential confounding variables such as smoking,  socioeconomic  status, and other
occupational exposures that are generally more homogeneous within the cohort relative to
external populations. Moreover, the mechanisms used to identify health outcomes and follow
individuals over time are generally applied in the same manner to all cohort members.  Taken
together, where different comparisons have been made to generate risk  estimates, those that have
been conducted using internal cohort comparisons are preferable.
       In addition to potential bias from the healthy worker effect, the comparison of SMRs
between studies is not always straightforward and is not recommended by some (Myers and
Thompson, 1998; Rothman, 1986).  The SMR is the ratio of the observed number  of deaths to
the expected number of deaths and is often referred to as the method of indirect standardization.
The expected number of deaths is estimated by multiplying the number of person-years tabulated
across individuals in the cohort,  stratified by age, by rates from a reference population that  are
available for the same strata. Therefore, each population cohort will have an estimated number
of cases derived using a different underlying age structure.  As outlined by Rothman (1986), the
mortality rates might not be directly comparable to each other, although the impact of such bias
will be much less if the age-distribution  of the cohorts is similar.  While it might be reasoned that
the TCDD exposed workers would have similar age distributions this is in fact not the case
(Becher et al.. 1998: Ott etal., 1993: Thiess et al.. 1982). This may be due to exposure occurring
both chronically, as well as from acute exposures due to accidental releases that happened at
                                          C-85

-------
various times at different plants. This is evident with the Hamburg and the BASF cohorts, as
most individuals comprising the BASF cohort were employed at the time of the accident
(1953/1954), while most of the Hamburg cohort (852/1048) was employed after 1954; the
follow-up of these cohorts ended at approximately the same time.
       The method of direct standardization allows for a more meaningful comparison of
mortality rates to be made between cohorts. With this approach, weights (usually based on age
and sex) are drawn from a standard population and are, in turn, applied to disease rates for the
same strata observed in the cohort of interest.  A comparison of weighted rates between different
cohorts would then be based on the same population standard.
       Despite these limitations in comparing SMRs between studies, Armstrong (1995) argues
that the comparisons are valid if the underlying stratum specific rates in each exposure grouping
are in constant proportion to external rates.  Comparisons of the SMRs between studies will be
biased only if there is an interaction between age and TCDD (i.e., the RR of disease due to
exposure differs by age). For cancer outcomes, the finding that associations become stronger
after a period of latency is incorporated into the analyses suggests that this assumption does not
hold true.  That is, risk estimates would be lower among young workers.  Similarly, for
noncancer outcomes,  some of the data from the Seveso cohort suggests differential effects
according to the age at exposure.
       The use of the SMR might also be biased in that workers exposed to TCDD could be
subject to more intensive follow-up than the general  population, and as a result, differential
coding biases with cause of death might occur. Moreover, some cohorts (e.g., the BASF cohort)
have been assembled, in part, by actively seeking out survivors exposed to accidental releases of
dioxins. As such, they would not include persons who have died or who were lost to follow-up.
This would result in underascertainment of deaths and SMRs developed from these data.  The
use of an internal cohort comparison offers distinct advantages to overcome potential sources of
selection bias. Given these uncertainties about the comparability across the different studies,
conducting a meta-analysis of cancer outcomes for TCDD using the SMR statistic is not
warranted for this analysis.
                                         C-86

-------
C.I.1.4.4.  All cancers versus site-specific
       An important consideration for quantitative dose-response modeling is the application of
models for all cancers combined, or for site-specific cancers.  Consistency is often lacking for
site-specific cancers, which might be due in large part to the relatively small number of cases
identified for site-specific cancers in the cohorts.  Although the risk estimates produced for all
cancer sites have important limitations and uncertainties, the data are far more consistent in
terms of the magnitude of an association and latency intervals. The IARC evaluation has put
forth the possibility of a pleuripotential mode of action between TCDD and the occurrence of
cancer. Despite the criticism of this assertion by some (Cole et al., 2003), the general
consistency of an increased risk for all-cancer mortality across the occupational cohorts when
latency intervals have been incorporated, provides adequate justification for dose-response
quantification of all cancer sites combined.

C.I.1.4.5.  Summary of epidemiologic cancer study evaluations for dose-response modeling
       All epidemiologic cancer studies summarized above were evaluated for suitability of
quantitative dose-response assessment using the TCDD-specific considerations and study
inclusion criteria. The results of this evaluation are summarized in a  matrix style array (see
Table C-2).  Table 2-1 in  Section 2 of this document summarizes the  key epidemiologic cancer
studies suitable for further TCDD dose-response analyses.

C.1.2. Noncancer
       In this section, the available epidemiologic data that could be  used in a dose-response
analysis for noncancer endpoints are evaluated. Because many of the key studies also evaluated
cancer outcomes, the noncancer studies are presented in the same order as in Section C.I.I.
Generally, the strengths and limitations of the cancer studies also apply to the noncancer
outcomes. In this section, key features of these studies that have direct relevance to modeling of
noncancer outcomes in particular are highlighted. To reduce redundancy, a detailed overview of
many of these cohorts and studies are not provided here. Instead, the reader should refer to
Section C.I. 1.1.
                                          C-87

-------
C.I.2.1.  Noncancer Cohorts
C.l.2.1.1.  The NIOSHcohort
       See general summary of the NIOSH cohort in Section C. 1.1.1.1.

C.l.2.1.1.1. Steenland et al (1999)
C.I.2.1.1.1.1.  Study summary
       The 1999 published report of NIOSH workers exposed to TCDD also conducted external
cohort comparisons to the U.S. general population using SMRs for mortality outcomes other than
cancer (Steenland et al., 1999). Analyses are based on 3,538 male workers employed at 8 plants
from 1942 to 1984. Four of the 12 plants originally analyzed were excluded due to lack of
records on the degree of TCDD contamination in the work processes or information was lacking
for work histories needed to estimate TCDD exposure.  Workers were excluded if they were
female (n = 40) or were lacking data to evaluate exposure (n = 238).  SMRs were based on a
mortality follow-up that was extended until the end of 1993. Cox regression analyses were used
to compare mortality risk in relation to TCDD  exposure within the cohort.

C. 1.2.1.1.1.2.  Study evaluation
       Overall, no statistically significant differences in all-cause mortality (SMR = 1.03,
95% CI = 0.97-1.08) were observed. Mortality from ischemic heart disease (SMR = 1.09,
95% CI = 1.00-1.20) and accidents (SMR = 1.25, 95% CI = 1.03-1.50) was slightly elevated.
Based on the external comparison population, the dose-response relationship for ischemic heart
disease observed with the SMRs calculated across TCDD exposure septiles was not statistically
significant (p = 0.14). Overall, no excess risk was observed for diabetes,  cerebrovascular
disease, or nonmalignant respiratory disease using the external population comparisons. Internal
cohort comparisons using the Cox regression model were performed using 0 and 15-year lag
intervals. A dose-response trend was observed for the derived ratios across the unlagged
cumulative TCDD exposure septiles for ischemic heart disease (p = 0.05) and diabetes
(p = 0.02).  For ischemic heart disease mortality, those in the upper two septiles had rate ratios of
1.57 (95% CI = 0.96-2.56) and 1.75 (95% CI = 1.07-2.87), respectively, relative to those in the
lowest septile.  In contrast, an inverse dose-response relationship was observed for diabetes
mortality.  The inverse  association found for diabetes is inconsistent with the positive association

                                          C-88

-------
reported in the Ranch Hands study (Michalek and Pavuk, 2008).  However, previous reports
have questioned the use of death certificates as the means to ascertain diabetes as these deaths
may be under-reported especially among those with diabetes who die from cancer (McEwen et
al.. 2006).

C. 1.2.1.1.1.3.  Suitability of data for TCDD dose-response modeling
       There was no evidence of a dose-response relationship between TCDD exposure and
ischemic heart disease mortality in this study or other cohorts. The inverse association with
diabetes also precludes dose-response analysis  for this outcome.  As all outcomes were based on
mortality,  dose-response modeling was not conducted for this study.

C.l.2.1.1.2.  Collins et al (2009)
C.I.2.1.1.2.1.  Study summary
       Collins et al. (2009)  described the mortality experience of Dow employees who worked
in Midland, Michigan. This plant produced 2,4,5-trichlorophenol between 1942 and 1979, and
2,4,5-T between 1948 and 1982. The cohort consisted of 1,615 workers (number of each gender
not specified) exposed to TCDD from as early as 1942; the follow-up of the cohort extended
until 2003.
       TCDD exposures were derived using serum samples obtained from 280 surviving
individuals (gender and selection criteria not reported). A simple one-compartment, first-order
pharmacokinetic model was used to estimate time-dependent TCDD measures. The area under
the curve approach was then applied to estimate cumulative TCDD exposure above background.
A half-life of 7.2 years for TCDD based on earlier work was incorporated into the exposure
estimation (Flesch-Janys etal.,  1996).
       Collins et al. (2009)  made an external comparison of the mortality rates of the cohort to
the U.S. general population using the SMR. Noncancer causes of death included all causes,
diabetes, cerebrovascular disease, nonmalignant respiratory disease, cirrhosis of the liver, and
accidents.  Overall, no statistically significant difference in all-cause mortality of these workers
was detected when compared to the general population (SMR = 0.9, 95% CI = 0.9-1.0).  Except
for cirrhosis  of the liver (SMR = 0.4, 95% CI = 0.1~0.8), no differences were found for any of
the noncancer causes of death relative to the general population.

                                         C-89

-------
       Internal cohort analyses based on cumulative measures of TCDD were conducted for
mortality from diabetes, ischemic heart disease, and nonmalignant respiratory disease using the
Cox regression model. These models adjusted for possible confounders such as year of hire and
birth year. No statistically significant associations were found between the continuous measure
of TCDD exposure and these causes of death.

C. 1.2.1.1.2.2.  Study evaluation
       Given that the external comparisons may result in bias from the healthy worker effect,
results from the internal cohort comparisons using the Cox regression model are preferred.
These analyses were performed for diabetes, ischemic heart disease, and nonmalignant
respiratory disease. TCDD levels for these workers were estimated using a simple
one-compartment pharmacokinetic model (Avlward et al., 2007). Because participation rates
and selection criteria for the 280 individuals providing samples were not reported, it is not
possible to determine how representative these individuals are  of the larger cohort. The hazard
ratios generated from the Cox regression model were not statistically significant for any of the
three noncancer outcomes modeled.

C.I.2.1.1.2.3.  Suitability of data for TCDD dose-response modeling
       No increased risks were observed for any of the noncancer outcomes reported in Collins
et al. (2009). As all outcomes were based on mortality, dose-response modeling was not
conducted for this study.

C.l.2.1.2.  The BASF cohort
       See general summary of the BASF  cohort in Section C.I. 1.1.2.

C.l.2.1.2.1. OttandZober
C. 1.2.1.2.1.1.  Study summary
       In 1996, Ott and Zober (1996a)  published a report on the mortality experience of the
cohort of 243 BASF male workers who were accidentally exposed to 2,3,7,8-TCDD in  1954 or
in the clean up that followed. The mortality follow-up of this cohort extended until the end of
1992. External comparisons of mortality were made with the German  population.  Internal
                                         C-90

-------
cohort comparisons were also made by estimating cumulative TCDD for the cohort using serum
measures that were obtained from 138 workers.  Ott et al. (1993) provided a detailed account of
the methodology to estimate TCDD. The 138 workers were selected based on a set of criteria of
duration of exposure (relative to the timing of the accident). There was no indication of the
participation rate among these workers, although some employee subgroups were over- and
under-represented. Briefly, a cumulative measure of TCDD expressed in ng/kg was derived, by
first estimating the half-life of TCDD using individuals who had repeated serum measures; the
half-life was estimated to be 5.8 years.  Individual-level  data on body fat were used to account
for the influence of body fat on decay rates. Half-life estimates of TCDD varied (range:
5.1-8.9 years) and were dependent on body fat composition (20% and  30%, respectively). This
approach differed from  previous analysis of this cohort that used a constant 7-year half-life (Ott
et al., 1993). TCDD levels at the time of serum sampling were then  estimated as the product of
TCDD concentration in blood lipid and the total lipid weight for each worker.  Nonlinear models
then were applied to estimate the contribution of duration of exposure to TCDD dose
extrapolated to the time of exposure.
      External comparisons to the German population using the SMR statistic also were
examined across dose categories.  The noncancer causes of death examined by Ott and Zober
(1996a) included all-cause mortality, diseases of the circulatory system, ischemic heart disease,
diseases of the digestive system, external causes, suicide, and  residual causes of death.  Overall,
no statistically significant differences in the SMR with the general population for all-causes of
death (SMR = 0.9, 95% CI = 0.7-1.1), nor any other causes of death examined were found.
      Ott and Zober (1996a) performed internal cohort comparisons using Cox regression.
These analyses found no dose-response patterns when cause-specific mortality was examined
across increasing cumulative TCDD exposure categories. Although  an inverse association for
diseases of the respiratory system (SMR = 0.1, 95% CI = 0.0-0.8) was detected, it was based
only on 1 reported death.  Many comparisons were limited by small  sample sizes as only
92 deaths occurred in the cohort, and of these, 31 were from cancer.  Also, the third component
of the cohort was identified primarily from former employees who were alive in 1986. As a
result, the SMR based on the general population was likely underestimated by the exclusion of
deceased workers.
                                         C-91

-------
C. 1.2.1.2.1.2.  Study evaluation
      As noted previously, caution should be exercised in the interpretation of SMR for
noncancer outcomes as they could be influenced by the healthy worker effect.  Although the
mechanism of identifying vital status appears to be excellent and unbiased, SMRs might be
underestimated due to the manner in which the cohort was constructed.  Specifically, a large
component of the cohort was assembled by actively seeking out former workers known to be
alive in 1986.

C.I.2.1.2.1.3.  Suitability of data for TCDD dose-response modeling
      No dose-response patterns were observed between TCDD and the noncancer outcomes in
the Ott and Zober (1996a) study.  Therefore, dose-response modeling was not conducted.

C.l.2.1.3.  The Hamburg cohort
      See general summary of the Hamburg cohort in Section C.I. 1.1.3.

C.l.2.1.3.1. Flesch-Janys et al (1995)
C. 1.2.1.3.1.1.  Study summary
      Flesch-Janys et al. (1995) reported on the mortality experience of a cohort of individuals
employed by an herbicide-producing plant in Hamburg, Germany, covering the period 1952 to
1992. As described in more detail in Section C. 1.1.1.3, the authors developed a cumulative
measure of TCDD using serum measures from 190 workers.  Selection criteria and response
rates for this subsample were not specified. This study also examined the  relationship between
total TEQ and mortality. In the study population, the mean TEQ without TCDD was 155 ng/kg,
and for the mean TEQ including TCDD was 296.5 ng/kg.
      Risks relative to the unexposed referent group of gas workers were estimated using Cox
regression across six exposed TCDD groups (i.e., the first four quintiles, and the ninth and
tenth deciles). A linear dose-response relationship was found with all causes of mortality and
cardiovascular mortality (p < 0.01).  The RR for all cardiovascular deaths  in the upper exposure
category was  1.96 (95% CI = 1.15-3.34), although there was no evidence  of a linear
dose-response trend (p = 0.27). The dose-response relationship was  strongest for ischemic heart
disease, with a RR of 2.48 (95% CI = 1.32-4.66) in the highest exposure group. A
dose-response relationship was also observed across TEQ groupings for all cause mortality,
                                         C-92

-------
cardiovascular disease mortality, and ischemic heart disease mortality. The authors did not
perform joint modeling of TEQ (without TCDD) and TCDD, so determining the extent that
DLCs contributed to an increased risk of mortality is not possible.

C. 1.2.1.3.1.2.  Study evaluation
       The Flesch-Janys et al. (1995) study lacks information on other potential risk factors for
cardiovascular disease, which could result in confounding if those risk factors are also related to
TCDD exposure.  Dose-response patterns were strong, however, and persisted across numerous
TCDD (and TEQ) exposure categories based on the use of an external reference group (i.e., gas
workers) or based on the internal comparison. The findings based on the internal comparison are
noteworthy in that these groups should be more homogenous with respect to confounding
factors.  As noted previously, the poor correlation between TCDD and smoking among workers
and similar smoking prevalence estimates between the workers and the external gas company
workers suggest that smoking was not likely a confounder of the TCDD and cardiovascular
disease relationship. No other evaluation of noncancer mortality outcomes has been undertaken
in this cohort since 1995.
       A strength of the Flesch-Janys et al. (1995) study was that it included the collection of
blood serum which provided an objective measure of TCDD exposure. Blood serum data,
however, were obtained only for 16% of the cohort.  However, the selection criteria and
participation rate for individuals providing blood serum is not provided to evaluate how
representative these individuals are of the larger cohort. The assumption of the first-order kinetic
elimination model is critical, given that measures were taken at the end of follow-up. The model
also assumed the half-life of TCDD was 6.9 years. If the kinetics are not first-order, or if the
half-life estimate is inaccurate, estimates of TCDD levels during exposure would be biased,
particularly for workers having longer periods between exposure and PCDD and PCDF assays.
Sensitivity analyses completed by the authors suggest that such bias is not likely to present
because the results were unaffected when different model assumptions regarding kinetic and
half-lives were examined.  The lack of an impact on RR estimates with varying half-life
estimates was similar to findings by Steenland et al.  (2001b).
                                          C-93

-------
c
  1.2.1.3.1.3.  Suitability of data for TCDD dose-response modeling
       Despite the aforementioned study strengths, the study focused on fatal outcomes such as
all cause mortality, cardiovascular disease mortality, and ischemic heart disease mortality. As all
outcomes were based on mortality, dose-response modeling was not conducted for this study.

C.l.2.1.4.   The Seveso Cohort—SWHS
       Eskenazi et al. (2000) presented an overview of the SWHS. The SWHS is the first
comprehensive epidemiologic study of the reproductive health of a female population exposed to
TCDD. The primary objective of the SWHS is to investigate the relationship of TCDD and
several reproductive endpoints, including endometriosis, menstrual cycle characteristics, birth
outcomes, infertility, and age at menopause. A second phase of follow-up that focuses on
osteoporosis, thyroid hormone, breast cancer, diabetes, and metabolic syndrome is not yet
completed.
       Women were eligible for participation in the SWHS if they resided in Zones A and B (the
most contaminated areas) at the time of the explosion, were 40 years  of age or younger at the
time of the explosion in 1976, and samples of their blood were collected and stored between
1976 and 1980.  The enrollment of women in the SWHS began in March 1996 and continued
until July 1998.  Of the  1,271 eligible women,  17 could not be found, 21 had died, and 12 were
too ill to participate. Of the 96% remaining women, 80% (n = 981) participated in the study.
Participation in the SWHS included a blood draw and an interview by a trained nurse who was
blind to subjects' TCDD level and zones of residence at the time of the accident.  The interview
included detailed information on potential confounders including occupational, medical, and
reproductive, and pregnancy history. Women who were premenopausal were also asked to
undergo a vaginal ultrasound and pelvic exam  and to complete a daily diary on menstruation.
       Depending on the health outcome under study, TCDD exposures were characterized for
the women at different times. For example, TCDD exposure levels were estimated at the time of
the accident for  some studies and at the time of conception for others. The SWHS study
population has been  used to investigate associations between maternal TCDD levels and the
following health outcomes: menstrual cycle characteristics (Eskenazi et al., 2002b):
endometriosis (Eskenazi et al., 2002a): birth outcomes (Eskenazi et al., 2003): age at menarche
(Warner et al.. 2004): age at menopause (Eskenazi et al.. 2005): uterine leiomyomas (Eskenazi et
                                         C-94

-------
al., 2007): and ovarian function (Warner et al., 2007). An evaluation of the studies in
chronological order is presented in this section.

C.l.2.1.4.1.  Eskenazi et al. (2002b)—menstrual cycle characteristics
C.I.2.1.4.1.1.  Study summary
       Eskenazi et al. (2002b) evaluated serum TCDD exposures in relation to several menstrual
cycle characteristics in the SWHS. A total of 981 women who were 40 years of age or younger
at the time of the accident comprised the SWHS.  The following exclusion criteria was applied
44 years of age or older, women with surgical or natural menopause, those with Turner's
syndrome, and those who in the past year had been pregnant, breastfed, or used an intrauterine
device or oral contraceptives.
       A trained interviewer collected data on menstrual cycle characteristics using a
questionnaire.  Women were asked to indicate how long their menstrual cycles were, whether the
cycles were regular (e.g., irregular cycle defined as length varied by more than 4 days), how
many days the menstrual flow lasted, and whether this flow was "scanty, moderate, or heavy."
Information was also collected on obstetric and gynecological conditions.  TCDD exposures
were derived from serum samples collected in 1976-1985. The authors selected the earliest
available serum sample, and back-extrapolated to 1976 values using either the Filser model
(Kreuzer et al., 1997) for women aged 16 years or younger in 1976 (n = 20) or the first-order
kinetic model (n = 6) (Pirkle et al., 1989).
       Serum TCDD levels were transformed using the logic scale, and the relationships
between these levels and length of menstrual cycle and days of menstrual flow were examined
using linear regression. The authors applied logistic regression to characterize the risk between
logioTCDD and heaviness of flow or regularity of cycle. In these analyses, moderate or heavy
flow and regular cycle were used as the reference categories. Stratified analysis was performed
by menarcheal status at the time of the accident.
       Overall, the association with TCDD exposure (per 10-fold increase) and length of
menstrual cycle was not statistically significant for premenarcheal (P = 0.93, 95% CI = -0.01,
1.86) women or postmenarcheal women (P = -0.03, 95% CI = -0.61, 0.54). The corresponding
estimates found for days of menstrual flow were P = 0.18 (95% CI = -0.15, 0.51) and P = 0.16
(95% CI = -0.18,  0.50), respectively. Reduced flow was not associated with TCDD when

                                          C-95

-------
compared to moderate or heavy flow (odds ratio [OR] = 0.84, 95% CI = 0.44, 1.61); effect
modification by menarcheal status, however, was evident (p = 0.03). Specifically, women
exposed to TCDD who were premenarcheal had lower odds of reduced flow, while those
exposed to TCDD who were postmenarcheal did not.  Finally, statistically significant ORs were
found between serum TCDD levels (per 10-fold increase) and having an irregular cycle
(OR = 0.46,  95% CI = 0.23, 0.95). This inverse association was evident in both premenarcheal
(OR = 0.50,  95% CI = 0.18, 1.38) and postmenarcheal women (OR = 0.41, 95% CI = 0.15, 1.16).

C. 1.2.1.4.1.2.  Study evaluation
      Overall, the Eskenazi et al. (2002b) study reported some associations between TCDD and
menstrual cycle characteristics among women exposed before menarche. Exposures to TCDD
were well characterized using serum samples available on an individual-level basis,  and the
design allowed for the influence of other risk factors to be controlled. Analysis of TCDD levels
and the length of menstrual cycle in premenarcheal women produced associations that were
largely not statistically significant at the alpha level of 0.05, but may have some biological
relevance. However, it is unclear whether the endpoints that were measured constitute adverse
health outcomes as they are not definitive markers of ovarian dysfunction. Another source of
uncertainty is measurement error due to the subjective nature of menstrual flow reporting. Any
resulting misclassification of the outcome would be expected to be nondifferential, as the
measurement error is unlikely to be dependent on TCDD exposure.

C.I.2.1.4.1.3.  Suitability of data for TCDD dose-response modeling
      Rigon et al. (2010) reported the median age at  menarche to be 12.4 in Italian females,
which would establish a critical window of susceptibility between birth and about 13 years of
age.  The determination of a lowest-observed-adverse-effect level (LOAEL) is difficult, as there
is no independent measure of an adversity threshold to establish the toxicological significance of
a given increase in menstrual cycle length. The study authors did not present data for unexposed
premenarcheal girls (in 1976),  so an appropriate reference population is not available.  However,
an approximate LOAEL can be estimated from Figure 1  in Eskenazi et al. (2002b), noting that
both the length of the menstrual cycle and its variance increases above TCDD concentrations of
                                         C-96

-------
about 1,000 ppt.  This study is suitable for further consideration for quantitative dose-response
modeling.

C.l.2.1.4.2.  Eskenazi et al (2002a)—endometriosis
C.l.2.1.4.2.1.  Study summary
       The SWHS provided the opportunity to investigate the association between serum TCDD
levels and endometriosis (Eskenazi et al., 2002a). The rationale the authors provided for
undertaking this study was the experimental animal studies that suggested an association, the
high prevalence of endometriosis among infertile women where breast milk concentrations of
dioxin are high, and the unknown etiology of endometriosis.  The study consisted of 601 women
who were younger than 30 years at the time of the Seveso accident.  Stored sera that had been
collected between 1976 and 1980 were available for these women.
       The researchers classified women as having endometriosis based on laparoscopy,
symptom report, gynecologic examinations, and vaginal ultrasound. Endometriosis cases were
identified by a positive ultrasound or if a woman had endometriosis noted on a laparoscopy or
laparotomy.  A woman was classified as nondiseased if she had surgery without a finding of
endometriosis or if she had a negative ultrasound, exam, and symptom history. Given that
laparoscopy could not be performed on women unless clinically indicated, there was less
certainty regarding endometriosis diagnoses among those without an ultrasound or prior
laparoscopy. These remaining women without clinical confirmation were classified as
"uncertain" based solely on positive symptom history.
       TCDD was measured in sera in 1976 for 93% of the women. Values for women whose
serum TCDD levels were collected after 1977 and had values exceeding 10 ppt were
back-extrapolated to 1976 using either the Filser model (<16 years of age) (Kreuzer et al., 1997)
or a first-order kinetic model (>16 years) (Pirkle et al., 1989).  These estimates of TCDD were
then modeled as both continuous (on a log scale) and categorical (<20, 20.1-100, and >100 ppt)
exposures.
       Polytomous logistic regression was applied to generate RRs for internal cohort
comparisons.  In relation to women in the lowest exposure category, the RR for endometriosis
among women in the middle and upper categories was 1.2 (90% CI = 0.3-4.5) and 2.1
                                         C-97

-------
(90% CI = 0.5-8.0), respectively.  The trend tests were not statistically significant for either the
categorical (p = 0.25) or the continuous measures of TCDD (p = 0.84).

C. 1.2.1.4.2.2.  Study evaluation
       Based on the results of a validation study they conducted in a clinical population, the
study authors found that symptom history was not predictive of disease, but that ultrasound had
excellent specificity and sensitivity for ovarian endometriosis.  Thus, there was some potential
for disease misclassification among the uncertain group who were classified solely on symptom
history. Although this disease misclassificationis could have resulted in missed cases of
endometriosis, it is unlikely to have biased the study findings. Bias is unlikely to result from
differential (by exposure status) symptom reporting for the following reasons: the study
interviewers and respondents were unaware of study hypotheses, the interviewers, respondents
and investigators who made the diagnoses did not know the TCDD levels, and the Centers for
Disease Control and Prevention laboratory had no information about disease. Younger women
were likely to be under-represented as those who had never been sexually active could not be
examined due to cultural reasons; thus residual confounding by age is a possibility despite
statistical  adjustment in the regression models. Other DLCs (PCDD, PCDFs, or polychlorinated
biphenyls [PCBs]) were not considered because of small serum volumes, but any potential TEQ
exposures occurring in the population were thought to be mostly attributable to TCDD in the
exposed women.  Although individual-level serum samples were available, a biologically-
relevant critical exposure window for this effect cannot be established.

C.l.2.1.4.2.3.  Suitability of data for TCDD dose-response modeling
       There were no statistically significant dose-response patterns observed with either
log-transformed TCDD exposures or across TCDD exposure categories, and the elevated risks
among those with higher exposures had very wide confidence intervals (that included unity). In
addition, because of the lack of definitive measures of endometriosis and the inability to define a
critical exposure window, quantitative dose-response analysis was not conducted for this
outcome.
                                          C-98

-------
C.l.2.1.4.3.  Eskenazi et al (2003)—birth outcomes
C.l.2.1.4.3.1.  Study summary
       Eskenazi et al. (2003) examined the relationship between serum TCDD levels and birth
outcomes. Analyses were based on 745 of the 981 women from the SWHS who agreed to
participate (80% of the cohort) and reported having been pregnant (n = 1,822). Many of these
pregnancies (888 pregnancies among 510 women) occurred after the accident in 1976.  Analysis
of spontaneous abortions was restricted to 769 pregnancies among 476 women that did not end
in abortion or in ectopic or molar pregnancy. Congenital anomalies were evaluated for the
672 pregnancies that did not end in spontaneous abortion. For the birth outcomes of fetal growth
and gestational age, analysis was performed using 608 singleton births from women without
hypertensive pregnancy disorders or diabetes.
       TCDD exposures were based on serum measures, most of which were taken shortly after
the accident.  Serum was collected in 1976-1977 for 413 women, between 1978 and  1981 for
12 women, and in 1996 for 19 women whose samples were not viable. For samples collected
between 1976 and 1981, the first serum sample collected was used.  TCDD exposures based on
serum samples collected after 1977 onward were back-extrapolated to 1976 using the Filser
toxicokinetic model (Kreuzer et al., 1997).
       Statistical analyses were performed on all pregnancies that ended between 1976 and the
time of interview. The authors also restricted the analysis to those pregnancies occurring within
the first 8 years (1976-1984) or roughly the first TCDD half-life after the explosion (Pirkle et al..
1989),  since the expectation was that exposure body burden would be greatest during this period.
A continuous measure of logic TCDD (base  10 scale) was used to investigate associations with
adverse birth outcomes. Logistic regression was used to characterize the relationship between
TCDD exposure spontaneous abortions, small for gestational age, and preterm birth (<37 weeks
gestation). Linear regression was used to describe the relationship between TCDD and birth
weight (in grams) and gestational age (in weeks) estimates.
       The risk estimates were adjusted for various characteristics that included sex  of infant,
history of low birth weight child, maternal height, maternal body mass index, maternal
education, maternal smoking during pregnancy, and parity.  No associations were detected
between TCDD serum levels and spontaneous abortion for pregnancies between 1976 and 1998
(OR = 0.8, 95% CI = 0.6-1.2), or those between 1976 and 1984 (OR = 1.0, 95% CI = 0.6-1.6).

                                         C-99

-------
No statistically significant associations (ORs ranged from 1.2-1.8) were found between
logic TCDD levels and preterm delivery or small for gestational age. The authors also saw no
association between TCDD exposure and mean birth weight among the entire population.
Although it was not statistically significant, the mean birth weight for pregnancies restricted to
between  1976 and 1984 decreased by 92 grams (0 = -92, 95% CI = -204 to 19) for every
10-fold increase in TCDD serum level.

C. 1.2.1.4.3.2.  Study evaluation
       This study was well-designed with individual-level exposure data, although there is some
uncertainty in extrapolating limited serum data to such narrow critical windows of exposure
especially among women who were pregnant many years after the explosion in 1976. While the
study lacked exposure data for the fathers, the authors indicated that only a small proportion
were believed to have high exposures to TCDD. A key limitation of the study was a reliance on
self-reported measures of pregnancy history subject to maternal recall error. For example, birth
weight was often reported only to the nearest 100 grams. This measurement error could lead to
some misclassification of the birth outcomes. The observation  that a large proportion of Seveso
women had a voluntary abortion because of fears of possible birth defects due to exposures from
the accident suggest that awareness bias is also possible as a result of differential reporting of
birth outcomes according to exposure status. Statistically significant associations were not
evident, although the mean birth-weight findings among those assumed to have the highest
TCDD body burden (exposed during first 8 years (1976-1984)) may have some toxicological
significance.  As the study authors point out, those who were potentially the most vulnerable at
the time of the accident (the youngest) had not yet completed their childbearing years.  Thus,
further follow-up of this cohort should help elucidate whether subjects with higher TCDD
exposures had an increased risk of adverse birth outcomes.

C.I.2.1.4.3.3.  Suitability of data for TCDD dose-response modeling
       No statistically significant associations were found in the study; in addition, possible
awareness bias could have influenced the self-reported measures of birth outcomes.  The authors
did not report TCDD levels at the time of pregnancy and EPA cannot extrapolate serum
concentrations measured in 1976 to the times of the pregnancies in these women based on the

                                         C-100

-------
information reported in the study.  Therefore, quantitative dose-response modeling was not
conducted for this study.

C.l.2.1.4.4.  Warner et al. (2004)—age at menarche
C.l.2.1.4.4.1.  Study summary
       Warner et al. (2004) examined the relationship between TCDD and age at menarche in
the SWHS cohort. As described earlier in this report, the SWHS comprised 981 participants.
This study was restricted only to those who were premenarcheal at the time of the accident
(n = 282). The proportional hazards model was used to examine the relationship between TCDD
exposures and age at menarche.  Age at menarche was determined by questionnaire administered
by a trained interviewer. Covariates examined as potential confounders included height, weight,
body mass index, athletic training at the time of interview, smoking, and alcohol consumption.
       TCDD exposures were determined using serum samples collected from 257 (91%) of
these women between 1976 and 1977.  For the remaining women, TCDD levels were quantified
from measures collected between 1978 and 1981 (n = 23, 8%) and in 1996 (n = 2, 1% collected
due to inadequate volume of older samples).  TCDD levels determined after 1977 were
back-extrapolated to the time of the explosion in 1976.  TCDD was modeled as both a
continuous variable (logioTCDD) and a categorical variable based on quartile values (<55.9,
56-140.2, 140.3-300, >300 ppt). The lowest group was further subdivided into those with levels
<20, and >20 ppt; this cut-point represented background levels found in a sample of women
living in an unexposed area.
       No association (hazard ratio [HR] = 0.95, 95% CI = 0.83-1.09) was detected between
age at menarche and a 10-fold increase in serum TCDD concentrations (from 10 ppt to 100 ppt).
Analyses restricted to those who were younger than 8 in 1976 produced similar results
(HR =  1.08, 95% CI = 0.89-1.30). No dose-response trend was observed with categorical
measures of TCDD among all women,  as well as those under the age of 8. A 10-fold increase in
serum TCDD concentrations were later reported to be associated with an earlier age of menarche
(HR =  1.20, 95% CI = 0.98-1.60, p for trend = 0.07) when analyses were restricted to 84 women
under the age of 5 at the time of the accident (Warner and Eskenazi, 2005).
                                        C-101

-------
C.l.2.1.4.4.2.  Study evaluation
       An important strength of the Warner et al. (2004) study is the ability to characterize
TCDD exposures using serum samples that were collected shortly after the accident occurred.
The outcome of interest, age at menarche, was determined by asking women "At what age did
you get your first menstrual period?" Previous work suggests that self-reported measures of age
at menarche decades later have modest agreement with responses provided during adolescence
with recall varying by education and by history of an adverse birth outcome (Cooper et al.,
2005). Although it seems unlikely, information  bias could be introduced in the Seveso study if
recall of age of menarche varied according to exposure levels.  The results from the analysis in
the original paper (Warner et al., 2004) were largely null there was some suggestion of an
association between elevated TCDD levels and earlier age of menarche in the follow-on
communication (Warner and Eskenazi, 2005). These more recent findings  lend some support to
the suggestion of Wolff et al. (2005) that the first 5 years of life may be the most relevant
exposure period for determination of an effect on age at menarche. However, the actual change
in the age at menarche relative to TCDD serum concentrations was not reported and cannot be
established from the information presented by the study authors.

C.l.2.1.4.4.3.  Suitability of data for TCDD dose-response modeling
       No major biases were evident, but some  sources of uncertainty remain which complicate
interpretation of the  study results and potential application to dose-response modeling.  The
study also showed limited evidence of an association between age at menarche and TCDD
exposure and little evidence of a dose-response relationship.  It remains unclear to what extent
age at menarche represents an adverse health effect. Thus, EPA cannot assess the biological
significance of this finding and cannot establish  a LOAEL for this effect. Therefore, quantitative
dose-response assessment was not conducted for this study, but it was included in the reference
dose (RfD) uncertainty analysis presented in Section 4.5.2.

C.l.2.1.4.5. Eskenazi et al. (2005)—age at menopause
C.l.2.1.4.5.1.  Study summary
       Eskenazi et al. (2005) evaluated the relationship between the age at onset of menopause
and serum levels of TCDD among women in the SWHS. Of the 981 (80% of women contacted)

                                         C-102

-------
women who agreed to participate in SWHS, this analysis was restricted to those who had not
reached natural menopause before the time of the accident and who were at least 35 years of age
at the time of the interview.  The recruitment and interview of women occurred approximately 20
to 22 years after the accident (March 1996-July 1998).
       The population was divided into quintiles of serum TCDD levels for the categorical
analysis.  For most women (n = 564), TCDD levels were estimated from samples provided in
1976-1977.  For the remaining women included in these analyses, TCDD levels were estimated
from samples collected between  1978 and 1982 (n = 28) and between 1996 and 1997 (n = 24;
collected due to insufficient volume of earlier sample). As noted previously, exposure levels for
women with post-1977 detectable levels of TCDD were back-extrapolated to 1976 using either
the first-order kinetic model (Pirkle etal., 1989) (>16 years at time of accident) or the Filser
model (<16 years at time of accident) (Kreuzer et al., 1997).  Women were classified as
premenopausal if they were  still menstruating or if they had amenorrhea as a result of pregnancy
or lactation (at the time of interview) with an indication of subsequent menstruation based on
maintained diaries or further examination.  Subjects for which amenorrhea had persisted for at
least 1 year with no apparent medical explanation were classified into a natural menopause
category. The  category, surgical menopause, pertained to women with a medically confirmed
hysterectomy or an oophorectomy.  Finally, impending menopause was defined for subjects in
which menstruation had been absent for 2 months, but who provided evidence of subsequent
menstruation, or had a secretory endometrial lining, or indicated less predictable cycles in the
previous 2-5 years.  If participants' menopausal status could not be determined, they were
grouped into the "other" category. This category included those for whom status could not be
determined due to current use of oral contraceptives, hormone replacement therapy, or previous
cancer chemotherapy.
       Statistical analysis was based on both a continuous measure of log-transformed TCDD
exposures and  categories based on quintiles (<20.4 ppt; 20.4-34.2 ppt; 34.3-54.1 ppt;
54.2-118.0 ppt; >118.0 ppt). The Cox model was used to generate hazard ratios as estimates of
relative risks and their 95% confidence intervals examining natural menopause as the outcome.
Several covariates previously identified as associated with menopausal status in the literature
were considered  as potential confounders. These covariates included body mass index, physical
                                         C-103

-------
activity, premenopausal smoking, education, marital status, history of heart disease and other
medical conditions, and other reproductive characteristics.
       A statistically significant association with onset of menopause was not detected
(RR = 1.02, 95% CI = 0.8-1.3) based on the logTCDD continuous measure.  The RRs were
found to increase across the second through fourth quintiles (RRs = 1.1, 1.4, and 1.6,
respectively) of serum TCDD categories in relation to those in the lowest  category, but not in the
upper quintile (RR = 1.0,  95% CI = 0.6-1.8). A statistically significant trend was detected
across the first four quartiles (p = 0.04) but not across all five quintiles (p  = 0.44).  However,
when the 24 women who  had back-extrapolated TCDD levels from 1996 were excluded, the
hazard ratios were slightly larger in magnitude.  Compared with women in the lowest quintile,
HRs for risk of earlier menopause were 1.2 (p = 0.5) for quintile 2, 1.6 (p  = 0.08) for quintile 3,
1.7 (p = 0.05) for quintile 4, and 1.2 (p = 0.5) for quintile 5, with a statistically significant trend
(p = 0.02) across the first  four quintiles.  Eskenazi et al. (2005) suggested  that the stronger results
following exclusion of 1996 measures may have been due to reduced exposure measurement
error and less exposure misclassification.

C.l.2.1.4.5.2.  Study evaluation
       The categorical exposure results from this study support a nonmonotonic
dose-related-association for earlier menopause with increased serum TCDD levels up to
approximately 118-ppt TCDD serum. Eskenazi et al. (2005) speculated that the inverse "U"
shape of the dose-response relationship is explained by the mimicking of hormones at lower
doses of a chemical, while at higher levels the toxic effect of a chemical does not have the
capacity to either inhibit or stimulate hormonal effects. Similar dose-response relationships have
been observed for TCDD for other endpoints in other studies for both humans and rodents (e.g.,
Mocarelli et al.. 2008: NTP. 2006: Steenland et aL 2001aX  although none with such  a
pronounced drop in response at higher exposures.  Overall, the findings suggest the possibility  of
a nonlinear dose-response relationship for age of onset of menopause with TCDD, with increased
risks in the 4*  quintile and perhaps the 3r quintile. However, the actual change in the age at
menopause relative to TCDD serum concentrations was not reported and cannot be established
from the information presented by the study authors.  The biological significance of these
                                          C-104

-------
findings is unclear. A biologically-relevant critical exposure window for this effect cannot be
established.
       A study limitation is the potential for residual confounding due to adjustment based on
current smoking status and not at the time of onset of menopause. It is unclear to what extent
smoking status may differ between these two time periods and whether smoking is related to
TCDD exposures in this cohort.

C.l.2.1.4.5.3.  Suitability of data for TCDD dose-response modeling
       Because the critical window of exposure that would cause an effect on age at menopause
is not apparent and EPA could not determine with confidence the biological significance of this
result for the establishment of a LOAEL, a quantitative dose-response assessment was not
conducted for this study in the context of the RfD derivation. However, this study is included in
the RfD uncertainty analysis presented in Section 4.5.2.

C.l.2.1.4.6.  Warner et al. (2007)—ovarian function
C.l.2.1.4.6.1.  Study summary
       Warner et al. (2007) investigated the association between serum TCDD levels and
ovarian function in subjects in the SWHS who were younger than 40 in 1976 and for whom sera
collected after the accident had been stored.  These women were recruited from March 1996 until
July 1998. Ovarian function analysis was limited to 363 women between 20 and 40 years  of age
and who were not using oral contraceptives.  Of these, 310 underwent transvaginal ultrasound
and were included in the functional ovarian cyst analysis.  Ninety-six women were in the
preovulatory stage of their menstrual cycles and were included in the follicle analysis.  For the
hormone analysis, 126 women who were in the last 2 weeks of their cycle were included.
       The authors used logistic regression to examine the relationship between TCDD and the
prevalence of ovarian follicles greater than 10 mm. Linear regression models were used to
examine the continuous outcomes: number of ovarian follicles >10 mm and diameter of
dominant ovarian follicle. Covariates considered for inclusion in the model were age at
ultrasound, age at accident, age at menarche, marital status, parity, gravidity, lactation history,
current body mass index, age at last birth, and smoking history. For the serum hormone
analyses, estradiol and progesterone were measured in blood at the time of interview. Ovulation

                                         C-105

-------
status was defined as a dichotomous variable (yes/no) based on a serum progesterone cut-point
value of 3 ng/mL.
       The adjusted ORs across categories of TCDD exhibited no dose-response trend for the
presence of follicles in relation to TCDD in the follicular phase; also, no statistically significant
differences were noted in any of the upper exposure categories relative to those in the lowest.
The adjusted OR for the continuous measure of logioTCDD was 0.99 (95% CI = 0.4-2.2). A
similar nonstatistically significant finding was found for logioTCDD in relation to ovulation in
both the luteal (OR = 0.99, 95% CI = 0.5-1.9) and mid-luteal phases (OR = 1.03,
95% CI = 0.4-2.7).  Progesterone and estradiol also were not related to serum TCDD levels for
either the luteal or mid-luteal phases (p = 0.51 andp = 0.47).

C.l.2.1.4.6.2.  Study evaluation
       The investigators found no relationship between serum TCDD levels and serum
progesterone and estradiol levels among women who were in the luteal phase at the time of
blood draw.  No association with number of ovarian follicles detected from ultrasound.
Although no association was found, the authors suggested that the lack of significant results
could be  because the women in SWHS were all exposed postnatally and the relevant and critical
time period for an effect might be in utero.

C.l.2.1.4.6.3.  Suitability of data for TCDD dose-response modeling
       Because of the lack of a defined critical exposure window and absence of associations
between  TCDD and adverse health effects in this study, quantitative dose-response assessment
was not conducted for this study; however, this study is included in the RfD uncertainty analysis
presented in Section 4.5.2.

C.l.2.1.4.7.  Eskenazi et al. (2007)—uterine leiomyoma
C.l.2.1.4.7.1.  Study summary
       Associations between TCDD exposures and uterine leiomyomata (i.e., fibroids), which
are benign estrogen-dependent tumors, were examined among 956 women in the SWHS
(Eskenazi et al., 2007).  The sample population was based on the original 981  SWHS participants
excluding 25 women diagnosed with fibroids before the date of the accident (July 10, 1976).

                                         C-106

-------
Women who previously had fibroids were identified both through the administered questionnaire
and the review of medical records. Transvaginal ultrasounds were performed for 634 women to
determine if they had fibroids at the time of follow-up. Women who had a fibroid diagnosis in
their medical records dated after the accident did not need to have an ultrasound. Similar to
other SWHS studies, exposure to TCDD was estimated using serum collected from women
shortly after the time of the accident, between 1978 and 1981 and in 1996. TCDD levels were
back-extrapolated to 1976 levels.
       The study authors performed statistical analyses using two definitions of fibroids as
outcome measures. The first was fibroids detected before the study, and the second was fibroids
detected via ultrasound. A proportional odds method Dunson and Baird (2001) developed was
used to model the cumulative odds of onset of fibroids. This method combines historical and
current information of diagnoses of fibroids.  Continuous and categorical measures of TCDD
were modeled. Regression models were adjusted for known or suspected risk factors of fibroids
including: parity, family history of fibroids, age at menarche, body mass index, smoking, alcohol
use, and education.
       Categorical measures of TCDD showed an inverse dose-response relationship with the
onset of fibroids. Relative to those with TCDD levels less than 20 ppt, those having TCDD
exposures between 20.1 and 75.0 ppt and greater than  75.0 ppt (at time of measurement) had
hazard ratios of 0.58 (95% CI = 0.41-0.81), and 0.62 (95%  CI = 0.44-0.89), respectively.  The
hazard ratio was 0.83 (95% CI = 0.65-1.07) for a continuous measure of logioTCDD. The study
authors concluded that TCDD may have antiestrogenic effects in  the uterine myometrium, in
contrast to the suggestion of estrogenic effects previously found in the breast (Warner et al.,
2002).

C.l.2.1.4.7.2.  Study evaluation
       The strengths of the Eskenazi et al. (2007) study included the longitudinal design,
individual-level serum measures (most taken within 2  years of the accident), and the ability to
include outcomes among those who did not take an ultrasound by using an adapted statistical
approach.  An important limitation was that the differences in risk by the stage of development
could not be assessed as all women were exposed postnatally, and only 4 cases were observed
among those who were premenarcheal at the time of exposure.  The authors found a

                                         C-107

-------
statistically-significant reduction in risk for uterine fibroids in SWHS women having TCDD
exposures between 20.1 and 75.0 ppt and greater than 75.0 ppt.  A biologically-relevant critical
exposure window for this effect cannot be established.

C.l.2.1.4.7.3.  Suitability of data for TCDD dose-response modeling
       Although this association is suggestive of anti-estrogenic activity, EPA was unable to
establish the biological significance of the findings at any particular exposure level for
establishing a LOAEL. Because a LOAEL could not be established for anti-estrogenic activity
(Eskenazi et al., 2007), quantitative dose-response modeling was not conducted.

C. 1.2.1.5.  Other Seveso noncancer studies
       See general summary of the Seveso cohort in Section C.I. 1.1.4.

C.l.2.1.5.1. Bertazzi et al (1989); Consonni et al (2008)—mortality outcomes
C.I.2.1.5.1.1.  Study summary
       Several studies have evaluated the mortality of Seveso residents exposed to TCDD
following the 1976 accident. The earlier section of this report described the designs of these
studies and discussed their findings as they relate to cancer mortality.  In this section, some of
the findings for other causes of death are described. A key feature of these studies is that
patterns of mortality among Seveso residents were investigated according to their zone of
residence at the time of explosion relative to general population rates.
       A 10-year mortality follow-up of residents of Seveso was published in 1989 (Bertazzi et
al., 1989). Poisson regression was used to derive RRs for those who had lived in Zone A at the
time of explosion using a referent group consisting of inhabitants who had lived in the
uncontaminated study area. Between 1976 and 1986, no statistically significant difference was
observed in all-cause mortality relative to the general population among those who lived in the
most highly exposed area (Zone A) at the time of the accident.  This finding was evident in both
males (RR = 0.86, 95% CI = 0.5-1.4) and females  (RR =1.14, 95% CI = 0.6-2.1). A
statistically significant excess in circulatory disease mortality was found among males relative to
those in the referent population (RR = 1.75, 95% CI = 1.0-3.2); this increased risk was more
pronounced when the follow-up period was restricted to the first 5 years after the accident
(1976-1981) (RR = 2.04, 95% CI =  1.04-4.2).  Between 1982 and  1986, the RR decreased
                                         C-108

-------
substantially and was not statistically significant (RR = 1.19, 95% CI = 0.4-3.5). Among
females, a risk similar in magnitude was detected for circulatory disease mortality although it
was not statistically significant (RR = 1.89, 95% CI = 0.8-4.2).  Contrary to the calendar
period-specific findings for males, the excess of circulatory mortality among females occurred
between 1982 and 1986 (RR = 2.91, 95% CI = 1.1-7.8) and not between 1976 and 1981
(RR =1.12, 95% CI = 0.3-4.5).  The number of deaths in this cohort with the 10 years of
follow-up was relatively small; in Zone A, 16 deaths were observed among males and 11 among
females.
       The most recently published account of the mortality experience of Seveso residents
provides further information on follow-up of these residents until the end of 2001 (25 years after
the accident) (Consonni et  al., 2008).  Three exposure groups were considered: Zone A (very
high contamination), Zone B (high contamination), and Zone R (low contamination). The
reference population consisted of those residents who lived in unaffected surrounding areas, as
well as residents of five nearby towns.  The authors used Poisson regression to compare
mortality rates for each zone relative to the reference population.
       For all causes of death, no excess was found in Zone A, B, or R relative to the reference
population. Statistically significant excesses were noted for those who lived in Zone A relative
to the reference population for chronic rheumatic heart disease (RR = 5.74,
95% CI = 1.83-17.99) and chronic obstructive pulmonary disease (RR = 2.53,
95% CI = 1.20-5.32).  These risks, however, were based on only 3 and 7 deaths, respectively.
For those in Zone A, no statistically significant excesses in mortality were noted for diabetes,
accidents, digestive diseases, ischemic heart disease, or stroke.  Among Zone A residents,
stratified analysis by time since accident showed increased rates of circulatory disease 5-9 years
since the accident (RR = 1.84, 95% CI = 1.09-3.12). Increased mortality from diabetes relative
to the reference population was noted  among females who lived in Zone B (RR = 1.78,
95% CI= 1.14-2.77).

C. 1.2.1.5.1.2.  Study evaluation
       The ascertainment of mortality in this cohort appears to be nearly complete.
Misclassification of some health outcomes, such as diabetes, may occur due to the use of death
certificate data.

                                         C-109

-------
       The characterization of exposure is based on zone of residence. Soil sampling indicated
considerable variability in TCDD soil levels, and therefore, the generation of risks based on zone
of residence likely does not accurately reflect individual exposure. Exposure misclassification
might also occur because residency in the areas does not necessarily reflect whether the
individual would have been present in the area at the time the accident occurred.  Any exposure
misclassification would likely be nondifferential which would tend to bias the risk estimates
towards the null.
       Although some excess of circulatory disease mortality was found, the finding was not
consistent between men and women. Moreover, excess circulatory disease mortality was more
pronounced among men within the first 5 years of exposure, while, for women, the excess was
more pronounced in years 5-10. Numerous other risk factors for circulatory disease were not
controlled for in these analyses and may be confounders if related to TCDD exposure.  Taken
together, the possibility that TCDD increased circulatory disease mortality based on these data is
tenuous at best.

C.I.2.1.5.1.3.  Suitability of data for TCDD dose-response modeling
       There is considerable uncertainty in these data due to the potential for outcome and
exposure misclassification. The lack of the individual-level TCDD levels and the examination
only of fatal outcomes reported in this study are not a suitable basis for development of an RfD.
For these reasons, dose-response analysis for this outcome is not conducted.

C.l.2.1.5.2.  Mocarelli et al (2000;  1996)—sex ratio
C.l.2.1.5.2.1.  Study summary
       A letter to the editor was the first report of a possible change in the sex ratio from dioxin
among Seveso residents following the July  10, 1976 accident (Mocarelli etal., 1996). The
authors reported that 65% (n = 48) of the 74 total births that had occurred from April 1977 to
December 1984 were females.  This male to female ratio of 26:48 (35%) is significantly different
from the worldwide birth ratio of 106 males: 100 females (51%) (James, 1995). Between 1985
and 1994, the Seveso male to female ratio leveled out at 60:64 (48%). The authors suggested
that the finding supported the hypothesis that dioxin might alter the sex ratio through several
possible mechanistic pathways.

                                         C-110

-------
       Mocarelli et al. (2000) later reported on an investigation of serum-based TCDD measures
in parents and the sex ratio of offspring. In this study, serum samples were collected from
mothers and fathers who lived in nearby areas at the time of the explosion, were between the
ages of 3 and 45 at the time of the explosion, and produced offspring between April 1, 1977 and
December 31, 1996. The study population included 452 families and 674 offspring, and serum
measures were available for 296 mothers and 239 fathers.  An estimate of TCDD at the time of
conception was also examined in relation to male to female birth ratios. TCDD exposure
estimates between the years of 1976 and 1996 were estimated using Filser's model (Kreuzer et
al..  1997).
       Mocarelli et al. (2000) used chi-square test statistics to compare observed sex ratio to an
expected value of 0.51 in this Seveso population. Concentrations of TCDD were modeled as
categorical variables in several ways.  First, a dichotomous variable was used whereby
unexposed parents were defined as those who lived outside Zones A, B, and R or had a serum
TCDD concentration of less than 15 ppt; parents with exposures of 15 ppt or higher were
considered exposed. Second, a trichotomous exposure variable was created that consisted of
parents who (1) lived outside Zones A, B, and R or had serum concentrations of less than 15 ppt,
(2) had serum concentrations of 15-80 ppt, and (3) had  serum concentrations that exceeded
80 ppt. These cut-points were chosen as they represented tertiles based on the distribution of
TCDD among parents.  Analyses were conducted separately for paternal and maternal TCDD
levels.
       The overall proportion of 0.49 male births (based on male to female ratio of 328:346) was
not significantly different from the expected proportion of 0.51 (p > 0.05).  Statistically
significant differences  were found, however, if both parents  had TCDD levels >15 ppt (sex
ratio = 0.44) or just the father had serum TCDD levels >15 ppt (sex ratio = 0.44). No
statistically significant differences were found when the fathers had TCDD levels less than
15 ppt, irrespective of the maternal levels.  A dose-response pattern in the sex ratio was found
across the paternal exposure categories. That is, the sex ratio decreased with increased paternal
TCDD levels (linear test for trend, p = 0.008).  In the unexposed group, the sex ratio (male to
female) was 0.56 (95% CI = 0.49-0.61), while in the highest exposure group
(281.0-26,400.0 ppt) the corresponding sex ratio was 0.38 (95% CI = 0.28-0.49).
                                         C-lll

-------
       Stratified analyses by age at paternal exposure revealed that the sex ratio was altered to a
greater degree among fathers who were younger than 19 at the time of the explosion.  The male
to female ratio among the unexposed fathers was 0.56 (95% CI = 0.50-0.62), while it was 0.38
(95% CI = 0.30-0.47) for those younger than 19 when exposed and 0.47 (95% CI = 0.41-0.53)
for those exposed after 19. Regardless of the  age at the time of exposure, however, fathers who
were exposed had a statistically significantly different birth ratio (they were more likely to father
girls) than those who were unexposed (p < 0.05).
       Separate analysis of birth ratios based on paternal TCDD exposure estimated at the time
of conception did not show the same dose-response pattern but did show strong evidence of
consistently decreased male births relative to females.  More specifically, the male to female
birth ratios  among the four successive quartiles (first through fourth) were 0.41,  0.33, 0.33,
and 0.46.

C. 1.2.1.5.2.2.  Study evaluation
       Mocarelli et al. (2000) based the characterization of TCDD exposure on serum samples,
which is an objective method for characterizing dose. Unlike for the occupational cohorts, serum
measures for this study were taken close to the time of the accident, and therefore,
back-extrapolation of TCDD exposures is unnecessary. Maternal TCDD levels at the time of
conception  did not demonstrate a dose-response relationship, but paternal exposures resulted in
consistently reduced male to female birth ratios (range: 0.33-0.46). Paternal exposures received
before the age of 19 at the time of the explosion were more strongly associated with a reduced
male to female ratio than those received after  the age of 19.
       The methods used to identify births appear to be appropriate. Even if some births were
missed, there is no reason to believe that ascertainment would be related to TCDD exposure and
the sex of the baby. Therefore, no bias is suspected due to incomplete birth ascertainment. The
authors  report that the findings did not differ when age at conception was dichotomized (< or
>35 years). They also state that age at conception was, on average, similar across calendar years.
However, some uncertainty remains as to what degree this influenced  the sex ratio given that the
lowest mean age of conception periods (1973-1976 and 1977-1984) also corresponded with the
lowest reported male:female ratios.
                                         C-112

-------
C.I.2.1.5.2.3.  Suitability of data for TCDD dose-response modeling
       TCDD exposures were well-characterized, and internal cohort analyses demonstrate an
association between paternal TCDD levels and birth ratio, particularly when exposure occurred
before 19 years of age.  Although the data are suggestive of an effect earlier in life, perhaps even
pre-pubertal, the biologically-relevant critical  exposure window of susceptibility cannot be
defined with any confidence for this endpoint. Quantitative dose-response assessment was not
conducted for Mocarelli et al. (2000) in the context of the RfD derivation. However, this study is
included in the RfD uncertainty analysis presented in  Section 4.5.2.

C.l.2.1.5.3.  Baccaretti et al (2004; 2002)—immunolo8ic effects
C.l.2.1.5.3.1.  Study summary
       The relationship between TCDD and immunological effects was evaluated in a sample of
Seveso residents (Baccarelli et al., 2004; Baccarelli et al., 2002). Both studies were based on
findings from 62 individuals who were randomly selected during December 1992 and March
1994 from Zones A and B. An additional randomly selected 59 subjects were chosen from the
surrounding noncontaminated areas during the same time period. Residency was based on where
subjects lived at the time of the accident (July 10, 1976) (Landi etal., 1998).  Frequency
matching ensured that the two groups of subjects were similar with respect to age, sex, and
cigarette smoking status.
       TCDD levels were determined by mass spectrometric analysis of plasma samples.
TCDD levels at the time of sampling were obtained, and estimates of levels at the time of the
accident also were estimated by assuming an 8.2-year half-life (Landi et al.,  1998).  Exposure to
other DLCs for both the TCDD contaminated and noncontaminated areas were reported to be at
background levels.  The plasma was also used to characterize levels of the immunoglobulins (Ig)
IgG and IgM and the complement components C3 and C4.  One subject was excluded due to lack
of an immunological evaluation. Analyses are, therefore, based on 58 subjects in the
noncontaminated areas and 62 individuals from the contaminated areas.
       Nonparametric tests were applied to test for differences between the two groups.
Multiple regression also was used to describe the relationship between the variables. Adjustment
was made for several potentially confounding variables that were collected via questionnaire.
                                         C-113

-------
       An inverse association was noted with TCDD levels and plasma IgG levels; this result
remained statistically significant after adjusting for other potential confounding variables in the
regression models. Specifically, the regression coefficient and/?-value for the unadjusted
(P = -035;p = 0.0002) and adjusted model were noted to be similar. In the 2004 analysis, the
authors present IgG, IgM, IgA, C3, and C4 median and interquartile values across TCDD
exposure quintiles. Decreased levels of IgG were observed in the highest exposure groups.
Specifically, the median values across the five quintiles (for lowest to highest) were 1,526;
1,422; 1,363; 1,302; and 1,163. The Kruskal-Wallis test for differences across the TCDD
categories was statistically significant (p = 0.002), which is consistent with the findings for the
continuous measures of TCDD.  This finding persisted after excluding those subjects with
inflammatory diseases and those who used antibiotics or nonsteroidal anti-inflammatory drugs.
For the other plasma measures, no dose-response relationship was apparent based on median
values for IgM, IgA, C3, or C4 across TCDD quintiles. The authors highlight the need for
additional research, particularly given the excess of lymphatic tumors noted in the area.

C. 1.2.1.5.3.2.  Study evaluation
       Both TCDD exposure and health outcome measures are relatively well characterized.
TCDD exposures, however, are based on concurrent serum measures and are far-removed from
the initial peak-exposure event. Therefore, back-extrapolation to  earlier time periods of exposure
would be highly uncertain.  EPA cannot determine with confidence whether the health outcome
is a result of current exposure or longer-term continuous exposure to elevated TCDD levels.
Furthermore, EPA cannot determine what effect the much higher  initial peak exposure might
have had on the outcome observed 17 years later.  A dose-response relationship between TCDD
and IgG was evident in the unadjusted model, but no  details are provided on any changes that
may be present when other covariates were added to the model.
       Interpreting the inverse association between TCDD exposure and IgG in terms  of clinical
significance is not possible. The 24% reduction in IgG at the highest exposures cannot be linked
to any adverse health outcome without more specific  testing. The IgG values reported are much
higher than those associated with antibody immunodeficiency disorders, as discussed by
Baccarelli et al. (2002). The biologically-relevant critical window of TCDD exposure  associated
                                         C-114

-------
with possible IgG impacts is uncertain, because it is unclear whether the current serum TCDD
levels or the higher prior TCDD serum levels are associated with these impacts.

C.I.2.1.5.3.3.  Suitability of data for TCDD dose-response modeling
       Although the data support an inverse dose-response relationship between IgG and TCDD,
the biological significance of the findings are too uncertain to define a LOAEL or a NOAEL.
Further the critical window of exposure that would cause an effect on IgG levels is not known
and thus does not allow for estimation of the effective TCDD exposure. For these reasons, these
data were not suitable for quantitative dose-response modeling.

C.l.2.1.5.4. Landi et al. (2003)—gene expression
C.l.2.1.5.4.1.  Study summary
       The impact  of TCDD on the aryl hydrocarbon receptor (AhR) was evaluated by Landi
et al. (2003) in a population-based study of Seveso residents.  AhR,  a mechanistically based
biomarker of dioxin response, must be present for manifestation of most of the toxic effects of
TCDD, including tumor promotion and  immunological and reproductive system effects (Puga et
al., 2000; Safe, 1986).  AhR activates the transcription of several metabolizing enzymes in
addition to certain genes (Whitlock, 1999). The primary objective of the study was to determine
whether plasma levels of TCDD and TEQ are associated with the AhR-dependent pathway in
lymphocytes among Seveso residents. The genes involved in the pathway that were examined
included: AhR, aryl hydrocarbon receptor nuclear translocator, CYPA1 Al and CYP1B1
transcripts, and CYP1A1-associated 7-ethoxyresorufm O-deethylase (EROD).
       Study recruitment occurred from December 1992 to March 1994. A total of 62 subjects
were randomly chosen from the highest exposed zones in Seveso (Zones A and B), while
59 were chosen from the noncontaminated area (non-ABR). Those  chosen from the
noncontaminated zone were matched by age,  sex, and smoking. Assignment of zones was based
on place of residence where subjects lived at the time of the accident in 1976. Subjects provided
data via questionnaire on a variety of sociodemographic and behavioral risk factors, including
cigarette smoking.  Multivariate models were adjusted for a variety of confounders including:
age, gender, date of assay, actin expression, postculture viability, experimental group,  and cell
growth.

                                         C-115

-------
       TCDD levels were determined using high-resolution gas chromatography, and 21 other
dioxins, or DLCs, were measured to examine TEQ.  Eleven measurements taken on the
121 subjects were deemed inadequate and excluded, but no further information was provided on
these exclusions. Nine subjects from Zone B and fourteen subjects from Zone ABR had TCDD
levels below detection, and were assigned a value equal to the lipid-adjusted detection limit
divided by the square root of 2.  The toxic equivalent for the mixture of DLCs (i.e., TEQ) was
calculated by summing the products of the concentration of each congener by its specific toxic
equivalency factor.
       The subjects provided between 5 and 50 mL of whole blood, which was centrifuged to
separate mononuclear cells. The cells were frozen and later thawed. Cells were cultured,
removed from the culture medium, and resuspended in a stimulation medium,  14 mL of which
was used for RNA analysis. Reverse transcription-PCR was conducted and EROD was assayed.
Differences in gene expression and EROD activity observed for various cell culture conditions
were compared using paired t-tests.  The unpaired Student's t-test was applied to test for
differences between groups, while a Bonferroni factor was used to account for multiple
comparisons. Data for continuous variables were log-transformed.
       TCDD accounted for 26% of the TEQ among the study subjects, but varied by zone (35%
in zone A and 18% in zone non-ABR).  After adjusting for confounding, AhR was inversely
related to plasma TCDD levels in uncultured cells (p < 0.03) and in mitogen-stimulated cells
(p < 0.05). EROD was lower in cells cultured from subjects with higher plasma TCDD and TEQ
levels, and the corresponding continuous measure of EROD was statistically significant
(p < 0.05). No statistically significant associations with TCDD or TEQ were found with ARNT
or CYP1B1 in uncultured cell medium,  nor with CYP1 Al or CYP1B1 in mitogen-stimulated
cells.  In general, females had lower AhR transcripts and higher levels of dioxin.
       Collectively, the findings suggest that TCDD exposure might reduce AhR expression in
unstimulated cells.  Therefore, TCDD could exert an influence on the AhR pathway regulation.

C.l.2.1.5.4.2.  Study evaluation
       The study used biologically-based measures of both TCDD exposures and biomarkers or
AhR.  Subjects were randomly selected from the larger cohort; some individuals with severe
medical illnesses were excluded (Landi et al., 1998).  Although few details are provided on the

                                        C-116

-------
number of subjects excluded for these reasons, given the objective nature of the biomarker
outcomes that were evaluated, such exclusions are unlikely to be an important source of bias.
The exclusion rates were also reported to be low and comparable across the zones (five subjects
from the noncontaminated zone non-ABR and four subjects from zone B).
       A strength of the study was the examination of other DLCs via the TEQ analysis.  A
limitation of the study included the relatively small number of subjects which resulted in the
grouping of several covariates, including TCDD exposures, into a small number of categories.
As such, slope coefficients derived from modeling continuous measures were emphasized in the
data presentation. Another key limitation of the study is the uncertainty of how effects on AhR
translate into subsequent development of cancer and other chronic health effects.

C.l.2.1.5.4.3.  Suitability of data for TCDD dose-response modeling
       It is unclear how associations between AhR biomarkers and TCDD levels translate into
an increased risk of adverse health effects. Dose-response analysis for this outcome, therefore,
was not conducted.

C.l.2.1.5.5. Alaluusua et al. (2004)—developmental dental defects
C.l.2.1.5.5.1.  Study summary
       Alaluusua et al.  (2004) examined the relationship between TCDD and dental defects,
dental caries, and periodontal disease among Seveso residents who were children at the time of
the accident. Subjects were randomly selected from those individuals who had previously
provided serum samples in 1976, which was shortly after the accident. A total of 65 subjects
who were less than 9.5 years of age at the time of the accident, and who lived in Zones A, B, or
R were invited to participate. Recruitment was initiated 25 years after the time of the Seveso
accident. An additional 130 subjects from the surrounding area (outside Zones A, B, or R or
"non-ABR zone") having the same age restriction were recruited.  Subjects were frequency
matched by age,  sex, and education.  Questionnaires were  administered to these individuals to
collect detailed information on dental and medical histories, education, and smoking behaviors.
Ten subjects who had completed at least high school were randomly excluded from the non-ABR
zone to create groups with similar educational profiles.  Participation rates  for the ABR and
non-ABR zones were 74% and 58%, respectively.

                                         C-117

-------
       One dentist who was blind to the patients' TCDD exposure levels assessed dental
aberrations.  Dental caries were assessed using recommendations of the World Health
Organization (WHO).  Periodontal status was described following a detailed evaluation of the
surfaces of the teeth.  A radiographic examination was done to identify missing teeth, alveolar
bone loss, deformities in the roots, and jaw cysts.
       Comparisons of the presence of dental enamel defects according to exposure status were
made using logistic regression. Chi-square test statistics were applied to compare the
distributions in the prevalence of dental defects across several categorical covariates (i.e.,
education, age, and serum TCDD level). For those who were younger than 5 at the time of the
accident, dental defects were more prevalent among patients in zone ABR (42%) than those in
the non-ABR zone (26%) (p = 0.14). Zone ABR is characterized by higher levels of soil TCDD
levels relative to non-ABR.  Serum levels permitted an improved characterization of risk as they
were available at an individual level, rather than using a zone of residence.  The continuous
measure of serum TCDD was associated with developmental dental defects (p = 0.007) and
hypodontia (p = 0.05). The authors classified less-exposed individuals in the non-ABR zones as
the reference population  and also examined exposure tertiles for the ABR residents.  The
prevalence of dental effects for the reference group was 26% (10/39).  The prevalence of dental
effects in the 1st, 2nd and  3rd tertile exposure groups was 10% (1/10), 45% (5/11) and 60% (9/15),
respectively. A total of 12.5% of the zone ABR subjects had missing permanent teeth (lateral
incisors and  second premolars) compared with 4.6% of the zone non-ABR residents. For zone
ABR subjects,  missing teeth were more frequent with higher serum TCDD levels.

C.l.2.1.5.5.2.  Study evaluation
       TCDD  exposures were characterized using serum measures for those who resided in
zone ABR in 1976 (within a year of the accident).  Alaluusua et al. (2004), however, provide few
details about the sampling frame used to identify these participants. Despite this, it is important
to note that a dose-response pattern was observed between TCDD exposure and presence of
developmental dental defects in the ABR population alone (p = 0.016). This finding is based on
27 subjects with developmental dental defects. This positive association provides support for a
quantitative dose-response modeling of developmental dental defects.  The numbers of such
subjects are small, however, with one, five, and nine subjects having defects in the exposure

                                         C-118

-------
tertiles ; the concentration ranges in the 1st, 2nd and 3rd tertiles were 31-226, 238-592, and
700-26,000 ng/kg TCDD, respectively.
C.l.2.1.5.5.3.  Suitability of data for TCDD dose-response modeling
       The considerations for conducting a dose-response analysis have been satisfied with the
study population.  A critical window of exposure can be defined for the subjects with
individual-level serum samples. The enamel defects combined with the prevalence of missing
permanent teeth in the higher-exposed subjects allows for a LOAEL to be established for the 2n
tertile exposure range. A NOAEL is evident for the 1st tertile and a NOAEL and LOAEL could
be established. Dose-response analyses were conducted for this outcome.

C.l.2.1.5.6.  Baccarelli et al  (2005)—chloracne
C.l.2.1.5.6.1.  Study summary
       Baccarelli et al. (2005) published findings from a case-control study of 110 chloracne
cases and 211 controls. The authors collected information on pigment characteristics and an
extensive list of diseases.  This study was performed to yield information about the health status
of chloracne cases, TCDD-chloracne exposure response, and factors that could modify TCDD
toxicity.  TCDD was measured from plasma from subjects recruited during 1993 to 1998.
Following adjustment for confounding, TCDD was associated with chloracne (OR = 3.7,
95% CI = 1.5-8.8), and the  risk of chloracne was considerably higher in subjects younger than 8
at the time of the accidents (OR = 7.4, 95% CI = 1.8-30.3). Among individuals with lighter hair,
the association between TCDD and chloracne was stronger than among those with darker hair.

C.l.2.1.5.6.2.  Study evaluation
       Statistical  power was limited in this study especially to assess potential interactions.
Study strengths included unique distribution of age and sex of chloracne cases, characterization
of individual-level TCDD exposures using sera samples, and the availability of both clinical and
epidemiologic data. Although a dose-response relationship was observed, chloracne is a rare
health outcome likely only to occur among those highly exposed.
                                         C-119

-------
C.l.2.1.5.6.3.  Suitability of data for TCDD dose-response modeling
       Given the very high TCDD levels needed to cause chloracne (Ott etal., 1993), this health
endpoint would not be considered as the basis for the RfD.  Therefore, dose-response analyses
for the Baccarelli et al. (2005) study were not conducted.

C.l.2.1.5.7.  Baccarelli et al. (2008)—neonatal thyroid hormone levels
C.l.2.1.5.7.1.  Study summary
       Baccarelli et al. (2008) investigated the relationship between thyroid function and TCDD
among offspring of women who were of reproductive age at the time of the 1976 accident.  This
health endpoint is relevant because thyroid function is important for energy metabolism and
nutrients and for stimulating growth and development of tissues. Neonatal thyroid function at
birth is evaluated through blood thyroid-stimulating hormone (b-TSH). Two related analyses
were conducted as part of this investigation: (1) the Residence-Based Population Study and
(2) the Plasma Dioxin Population Study.
       For the Residence-Based analysis, the study population of 1,772 women was selected
based on the following criteria: having lived in the highly contaminated areas (Zones A or B) at
the time of the accident or between July  10, 1976 and December 31, 1947; were of fertile age
(born after 1947); and were alive as of January 1, 1994.  A random sample of 1,772 unexposed
women who lived in the reference area was selected from the 55,576 eligible female participants
using frequency matching by year of birth to the exposed women and residency in the reference
area at the time  of the accident. The reference area represents the noncontaminated areas that
surround the three zones of decreasing exposure (Zones  A, B and R).  Population registry offices
(n = 472) were contacted to detect children born to these women. Records could be traced  for
virtually all subjects (1761/1772 exposed; 1762/1772 unexposed). Children born outside the
Lombardy area  (n = 156) were excluded as b-TSH could not be obtained for them. The analyses
were based on the remaining 56, 425, and 533 singletons born between January 1, 1994 and June
30, 2005 in Zone A, B, and from the reference area, respectively.
       Thyroid function is tested in all newborns by b-TSH measures in the region of Lombardy
where Seveso is located. These measures were obtained from blood samples taken 72 hours after
birth using a standardized protocol.  The b-TSH  levels were log transformed to approximate a
normal distribution.  Linear regression analysis was used to conduct test for trends in mean

                                         C-120

-------
b-TSH levels across different covariates. Logistic regression was used to assess associations
between elevated b-TSH levels defined by the cutpoint of 5 uU/mL and residence in particular
zones of contamination. The 5 |iU/mL cutpoint for thyroid stimulating hormone (TSH)
measurements in neonates was recommended by WHO (1994) for use in neonatal population
surveillance programs. Although WHO established the standard for increased neonatal TSH in
the context of iodine deficiency disease, the toxicological implications are the same for TCDD
exposure and include increased metabolism and clearance of thyroxine (T4). Fisher's exact tests,
Wilcoxon nonparametric tests, and generalized estimating equations were used to adjust the
standard errors of the regression coefficients due to correlation between siblings.
       Results from the Residence-Based analysis indicate that mean levels of b-TSH were
positively associated with average soil TCDD concentrations in the three areas (Zone A: 1.66
uU/mL; Zone B: 1.35 uU/mL; and Zone R: 0.98 uU/mL) (p < 0.001).  Plasma TCDD levels also
were shown to be much higher in a group of 51 newborns that had b-TSH levels >5 uU/mL.
Compared to the reference population, adjusted ORs were elevated for Zone B (OR =  1.90,
95% CI = 0.94-3.86) and Zone A (OR = 6.63, 95% CI = 2.36-18.6).  These ORs were adjusted
for gender, birth weight, birth order, maternal age at delivery, hospital, and type of delivery. The
adjusted ORs however differed only slightly from those that were unadjusted (Zone B
OR = 1.79, 95% CI = 0.92-3.50; Zone A OR = 6.60, 95% CI = 2.45-17.8). Of the risk factors
considered, only gender and birth weight were identified as independent predictors of neonatal
b-TSH levels.
       The Plasma Dioxin Population analysis included  children born to 109 women who were
part of the Seveso Chloracne Study (Baccarelli et al., 2005). A total of 51 children were born  to
38 of these women, of these 12 lived in Zone A,  10 in  Zone B, 20 in Zone R, and 9 from the
reference population.  All children in this analysis from zones A and B were also part of the
Residence-Based population study (which included all zone A and B women), while none of the
children from zone R and the reference area were sampled in the Residence-Based population
study.  Several congeners including TCDD were measured in maternal plasma collected from
December 1992 to September 1998.  TCDD levels were extrapolated to the date of delivery
using a first-order pharmacokinetic model (Michalek et al.. 1996). The elimination rate used was
9.8 years based  on the mean half-life estimate from a previous study of women in the Seveso
region (Michalek et al., 2002). TEQs were calculated  for a mixture of DLCs by multiplying the
                                         C-121

-------
concentration of each congener by its toxicity equivalence factor.  The maternal average TEQ
was 44.8 ppt (range: 1 1.6-330.4) among 51 mothers. The measurement of noncoplanar PCBs
occurred only later in the study (1996) and, therefore, total mean TEQs (i.e., including the sum
of PCDDs, PCDFs, coplanar PCBs, and noncoplanar PCBs) are available only on a subset
(n = 37) of the population. DLCs were examined as earlier studies suggested associations
between the sum of PCBs, or individual congeners having decreased thyroxine (Sandau et al.,
2002: Longnecker et al.. 2000). and increased TSH (Alvarez-Pedrerol et al.. 2008: Chevrier et
al., 2007).  The following confounders were examined by the authors in the plasma dioxin
models: maternal body mass index, smoking habits, alcohol consumption, and neonatal  age in
hours at the time of the b-TSH measurement.
       For the Plasma Dioxin analysis, the authors used a linear regression model to examine the
association between maternal TCDD levels and b-TSH.  The standardized regression coefficient
obtained from this model was 0.47 (p < 0.001). For the evaluation of TEQs, a similar
association was  noted for PCDDs, PCDFs, and coplanar PCBs (n = 51, p = OA5,p = 0.005) but
not with noncoplanar PCBs (n = 37, P = 0. 16, p = 0.45).  Statistically significant associations
between b-TSH with plasma TCDD, PCDDs, PCDFs, and coplanar PCBs, but not with
noncoplanar PCBs, were found based on multivariate regression models adjusted for gender,
birth weight, birth order,  maternal age at delivery, hospital, and type of delivery. No association
was detected for the sum of all total TEQs from the measured compounds (n = 37, p = 0.31,
C.l.2.1.5.7.2.  Study evaluation
       The Baccarelli et al. (2008) study satisfies the epidemiologic considerations and criteria
for determining whether dose-response modeling should be pursued. The outcome is well
defined, and a dose-response pattern was observed. The study also contained a sub study that
characterized TCDD and exposures to other DLCs and used serum measures for a sample of
mothers. Results were consistent among the zone of residence analysis and the substudy based
on plasma measures. Although they examined potential confounding factors, a study limitation
was that this assessment was based on statistical significance alone.
                                        C-122

-------
C.l.2.1.5.7.3.  Suitability of data for TCDD dose-response modeling
       Given the potential for exposure misclassification due to variability in TCDD soil levels
within each zone, modeling should rely on individual-level TCDD exposures derived from the
plasma sampling substudy. Data from this study population provide an opportunity for
quantitative dose-response analyses as the critical exposure window of 9 months can be used for
exposure assessment purposes.

C.l.2.1.5.8.  Mocarelli et al (2008)—syerm effects
C.l.2.1.5.8.1.  Study summary
       Mocarelli et al. (2008) examined the relationship between TCDD and endocrine
disruption and semen quality in a cohort of Seveso men. Study participants included 397 of the
eligible 417 males (<26 years old in 1976) from Zone A and nearby contaminated areas who had
serum TCDD levels measured in 1976. Frozen serum samples collected from 1976 to 1977 were
used to derive TCCD exposures. In addition, 372 healthy blood donors not living in the
TCCD-contaminated area were invited to participate. The researchers collected a health
questionnaire and semen samples from participants. Analyses were based on 257 individuals in
the exposed group and 372 in the comparison group.  Of the 257 exposed men, 135 (53%)
without disease agreed to participate, while 184 of the 372  (49%) recruited men in the
comparison group participated.  Semen samples were collected postmasturbatory at home.
Ejaculate volume, sperm motility, and sperm concentration were measured on these samples.
Fasting blood samples also were collected from the subjects for reproductive hormone analyses,
including 17p-estradiol (£2), follicle stimulating hormone (FSH), inhibin B, LH, and
testosterone.
       The researchers estimated serum concentrations of TCDD from samples provided in
1976-1977, and also in 1997-1998 for individuals whose earlier samples had TCDD values that
exceeded 15 ppt. Serum concentrations for the comparison group were assumed to be less than
15 ppt in 1976 and 1977  and <6 ppt in 1998/2002 on the basis of serum results for residents in
uncontaminated areas.  The exposed and comparison groups were divided into three groups
based on their age in 1976: 1-9, 10-17, and 18-26 years. Mocarelli et al. (2008) applied a
general linear model to the sperm and hormone data and included exposure status, age, smoking
                                         C-123

-------
status, body mass index, and occupational exposures as covariates. The study authors addressed
the potential for confounding factors.
       Men exposed between the ages of 1 and 9 had reduced semen quality 22 years later.
Reduced sperm quality included decreases in sperm count (p = 0.025), progressive sperm
motility (p = 0.001), and total number of motile sperm (p = 0.01) relative to the comparison
group. The opposite pattern was observed for several indices of semen quality among those aged
10-17 at the time of the accident; this included a statistically significant increase in sperm count
(p = 0.042).  The clinical significance of this increase is unknown.  For the hormone analyses,
those in the exposed group had lower serum £2 levels, and higher follicle stimulating hormone
concentrations. Neither testosterone levels nor inhibin B concentrations were associated with
TCDD exposure.

C.l.2.1.5.8.2.  Study evaluation
       The findings of the Mocarelli et al. (2008) study support the hypothesis that exposure to
TCDD in infancy/prepuberty reduces sperm quality. The changes in serum £2 and FSH
concentrations are of unknown clinical significance, and it is unclear whether they represent
adverse health endpoints. Further, it may be noted that the collection of a single semen sample is
not suitable for accurate evaluation of semen effects in an individual, but is less of a concern for
evaluation of the population average.  Although most semen analysis studies have low
compliance rates in general population samples (20-40%) (Muller et al., 2004; J0rgensen et al.,
2001), the compliance rate in this study was much higher (60%). Given that the compliance
rates were similar between the exposed and comparison groups and the strong differences
detected across the two age groups, selection bias appears unlikely in this study.

C.l.2.1.5.8.3.  Suitability of data for TCDD dose-response modeling
       The health outcomes are well defined in the Mocarelli  et al. (2008) study, and exposures
are well characterized using serum data.  Because the men exposed to elevated TCDD levels
between the ages of 1 and 9 had reduced semen quality 22 years later, it is difficult to identify the
relevant time interval over which TCDD dose should be considered.  Specifically, it is difficult
to discern whether this effect is a consequence of the initial high exposure between  1 and 9 years
of age or a function of the cumulative exposure for this entire exposure window beginning at the

                                         C-124

-------
early age. However, the differences between these two dose estimates (the initial high exposure
versus the cumulative exposure for the 9 year window) are minimal (i.e., within an order of
magnitude). Despite the uncertainty in estimating the critical window of exposure,
dose-response analysis for this outcome was conducted.

C.I.2.1.6.  The Chapaevsk study
       See general summary of the Chapaevsk study in Section C.I. 1.1.5.

C.l.2.1.6.1. Revich et al. (2001)—mortality and reproductive health
C.I.2.1.6.1.1.  Study summary
       Revich et al. (2001) describe a series of investigations that have evaluated adverse health
outcomes among residents of Chapaevsk where ecological measures of TCDD have been noted
to be higher than expected.  In the earlier cancer section of this report, the cross-sectional
comparisons of mortality that the authors carried out between Chapaevsk residents and a general
population reference were described. Although the general focus of this paper is on cancer, the
authors examined other adverse health outcomes.
       For all-cause mortality, rates were found to be higher in Chapaevsk relative to the Samara
region and other nearby towns.  The magnitude of this increase, however, was not quantified in
the review by Revich et al. (2001)  Cardiovascular mortality accounted for nearly two-thirds of
women's deaths and almost  half of those among men.  The rates  of cardiovascular mortality
among Chapaevsk men have been reported to be 1.14 times higher than those in Russia.
       Revich et al. (2001) also reported on the occurrence of adverse reproductive events.
Although the authors indicated that official medical information was used to make comparisons
between regions, no details were provided about data quality, completeness, or surveillance
differences across areas. The presented rates for reproductive health outcomes should be
interpreted cautiously.  A higher rate of spontaneous abortions (24.4 per 100 pregnancies
finished by delivery) was found in Chapaevsk women relative to rates that ranged between 10.6
and 15.2 found in five other areas.  The frequency of preeclampsia also was found to be  higher in
Chapaevsk women (44.1/100) relative to other towns, as was the proportion of low birth-weight
babies and preterm births. The percentage of newborns with low birth weight was slightly larger
in Chapaevsk (7.1%) when compared to other towns in Samara (5.1-6.2%); observed
differences, however, were not statistically significant. The authors also reported on the sex ratio
                                         C-125

-------
of newborns born between 1983 and 1997.  These ratios (boys:girls) were highly variable and
ranged between 0.79 and 1.29. Given the annual variability of this ratio on a year-to-year basis,
it is unclear if this is largely due to natural fluctuations and to what extent this may result from
prior TCDD (or other contaminants) exposure TCDD and other contaminants.

C. 1.2.1.6.1.2.  Study evaluation
       The review by Revich et al. (2001) highlights analyses that have been undertaken using
largely cross-sectional data. Although soil sampling measures appear to demonstrate decreasing
levels of TCDD in the soil with increasing distance from the plant, at this time, no
individual-level TCDD exposure data are available.  Increased rates of mortality relative to the
Samara region in Russia were observed among Chapaevsk men for all cancer sites combined;
this excess risk however, was not observed among women.  Although the authors provide
compelling evidence of increased adverse events among residents of Chapaevsk, the study lacks
a discussion about the validity of comparing health data across regions, and suffers from inherent
limitations from ecological studies such as exposure misclassification and potential for
confounding.

C.I.2.1.6.1.3.  Suitability of data for TCDD dose-response modeling
       Insufficient details  are provided by the authors to gauge the completeness and coverage
of the cancer registry and mortality data.  Health outcomes were studied on the basis of
information in the official medical statistics. As with the cancer outcomes presented in this
study, the data for noncancer outcomes are limited by the absence of TCDD levels on an
individual-level basis and information on other potential confounding variables that could have
biased the results.  The cross-sectional nature of the data that were presented does not provide
the necessary level of detail needed to estimate effective dose given the lack of individual-level
exposure data. Therefore,  a quantitative  dose-response analysis was not conducted.

C.l.2.1.7.  The Air Force Health ("Ranch Hands" cohort) study
       See general summary of the Ranch Hands cohort in Section C. 1.1.1.6.
                                         C-126

-------
C.l.2.1.7.1. Henriksen et al, (1997)
C.I.2.1.7.1.1.  Study summary
       Henriksen et al. (1997) investigated the relationship between TCDD exposure and
diabetes among participants of the Air Force Health Study (AFHS). This study included
veterans of Operation Ranch Hand who served in Southeast Asia between 1962 and 1971 and
were exposed to high levels of dioxin from the spraying of Agent Orange during flight
operations and the maintenance of aircraft and herbicide spray equipment. In addition, it
included a comparison group of other Air Force veterans who also  served in Southeast Asia
during the same period, but were not actively involved in the spraying of herbicides. This
comparison group was selected by matching to the Ranch Hands on the basis of age, race, and
military occupation. Data from physical examinations in 1982, 1985,  1987, and 1992 were used
for the study. The cohort initially consisted of 1,108 Ranch Hands and 1,494 veterans in the
control cohort.
       Incident diabetes from the end of the tour of duty through June 1995 was identified based
responses provided from questionnaires administered from at least  one of the four examinations,
followed by verification of medical records and laboratory results.  Study subjects were
classified as diabetics if they had a verified history of diabetes mellitus by medical diagnosis or if
they exhibited a 2-hour postprandial glucose laboratory value of >200 mg/dL. A total of
315 incident cases of diabetes were identified; of these, 169 occurred in the comparison cohort.
The authors also examined associations between TCDD and the following health outcomes:
severity of diabetes, time to onset of diabetes, and glucose abnormalities. Diabetes severity was
determined based on a review of the medical records, and questionnaire responses and classified
as insulin therapy, oral medication, diet only, or no control.  Fasting glucose and 2-hour
postprandial glucose were used to identify glucose abnormalities.  The 100-gm glucose load for
the postprandial assay was not given to known diabetics. The outcome time-to-onset of diabetes
was defined as the number of years between the end of the last tour of duty in Southeast Asia,
and initial diagnosis of diabetes. For those without diabetes, the time to onset of diabetes was
the number of years since the end of tour of duty and the last physical examination; this time-to
onset value was right-censored.
       Serum dioxin levels were first estimated using high resolution gas chromatography/high
resolution mass spectrometry using samples collected in the 1987 interview.  Those whose

                                         C-127

-------
dioxin levels were not quantifiable in 1987 and those who refused or were new to the study were
asked to provide serum in 1992 to measure dioxin.  Dioxin levels were then estimated for the
Ranch Hands at the end of the tour of duty by assuming a constant half-life of 8.7 years. The
Ranch Hands were classified on the basis of this TCDD exposure estimate into one of
three groups (Background, Low, or High). The study excluded those with a history of diabetes
before service in Southeast Asia, those with no measure of dioxin, and those in the comparison
group with a dioxin level that exceeded 10 ppt which was regarded as the threshold level for
background exposure. The analyses of diabetes mellitus and TCDD exposure were based on
2,265 veterans (989 Ranch Hands,  1276 Comparison veterans).
       The relative risk (and confidence intervals) of diabetes was estimated using the ratio of
the prevalence of diabetes in Ranch Hands veterans relative to the comparison group using the
method of Rothman (1986). The risk of diabetes was associated with TCDD exposure, and
Ranch Hands in the highest exposure group had a relative risk of 1.5 (95% CI = 1.2, 2.0) relative
to those in the comparison cohort.  A subsequent analysis of this cohort further adjusted for the
effects of triglycerides, which slightly attenuated this risk estimate (RR = 1.4, 95% CI = 1.1-1.8)
(Michalek et al., 1998). The severity of diabetes was associated with dioxin exposure. For
example, among those who required insulin therapy for the management of their diabetes, the
relative risk was among those in the High dioxin exposure group relative to  those in the lowest
2.4 (95% CI=0.9 - 6.4). Time to onset of diabetes was found to be inversely related to exposure
to dioxin, and this association persisted across veterans stratified by body fat percentage. Serum
insulin abnormalities, as determined by the 2-hour postprandial glucose measure, were positively
associated with dioxin exposure in nondiabetics.  Specifically, among Ranch Hands in the High
dioxin exposure category, the prevalence  of those with abnormal insulin values was 8.4%
compared to 2.5% among those in the comparison cohort (RR=3.4, 95% CI=1.9 - 6.1).

C. 1.2.1.7.1.2.  Study evaluation
       A strength of this study is its relatively large sample size  of 2,265 veterans, and identified
cases  of diabetes (n = 315).  Moreover, there is a large range in exposure to  TCDD across the
study population (i.e., the comparison cohort as well as veterans  of the Operation Ranch Hands).
The study was able to achieve a high level of participation, and lengthy follow-up interval with
                                         C-128

-------
data from four physical examinations.  As documented by Michalek et al (200 Ic), few veterans
were lost to attrition over the four physical examinations.
       The methods used to identify newly diagnosed cases of diabetes following the tour of
duty were valid, and the study evaluated several different measures associated with diabetes.
The associations observed between these different health measures (i.e., diabetic status, time to
onset of diabetes, severity of diabetes, and insulin abnormalities) were consistent, and therefore,
strengthen the argument that exposure to TCDD may contribute to the development of insulin
resistance and diabetes.
       The use of serum measures to estimate TCDD exposure was also a strength of the study.
The authors estimated dioxin levels in veterans at the end of their tour of duty using a constant
half-life of 8.7 years, and conducted additional sensitivity analyses across strata of subjects
grouped by body fat percentages. These results produced similar associations.
       Unlike the subsequently published study by Longnecker and Michalek (2000) which is an
essentially cross-sectional analysis of the comparison cohort, the analysis presented in this study
is longitudinal. The dramatically higher exposure to TCDD among the Ranch Hand component
of the cohort during their tour of duty allows for diabetes prevalence,  severity, time to onset, as
well as glucose abnormalities among nondiabetics to be compared across groups that differed  by
TCDD exposure before these health outcomes were determined.
       An important limitation of the study was raised by Slade (1998) who noted that
interactions between plasma lipid fractions, dioxin, and diabetes could produce a spurious
association between dioxin and diabetes. In her letter, she noted that hyperinsulinemia, insulin
resistance, impaired glucose tolerance and diabetes are all associated with lipid abnormalities,
and the corresponding change in lipid fractions may elevate dioxin levels. As exposure to TCDD
was estimated in 1987, and in some cases 1992, it is possible that these lipid abnormalities may
have distorted the back-extrapolation of TCDD exposure estimates at the end of the tour of duty
in Vietnam.  The authors were not able to directly evaluate the magnitude of this source of
measurement error because  no lipid samples were stored for this cohort that would allow for
dioxin to be measured. Subsequent analysis to respond to these comments found little change in
the  risk estimates for diabetes after adjusting for triglycerides (Michalek et al.. 1998). However,
dioxins have also been shown to affect triglyceride levels in both animals and in humans, and
                                         C-129

-------
therefore the influence of triglycerides may be responsible for a noncausal association between
dioxin and the health outcomes in this study.

C.I.2.1.7.1.3.  Suitability of data for TCDD dose-response modeling
       The use of the individual-level TCDD serum measures and the identification of diabetes
through medical records and objectively-based serum tests are strengths.  TCDD levels were
estimated based on samples collected in 1987, and in some cases 1992; the study authors note
that these samples were collected 20 to 30 years after the TCDD exposures.  If there are
diabetogenic effects of TCDD, it is unclear whether TCDD-mediated diabetes onset might be a
consequence of an elevated TCDD exposure event over a relatively short period of exposure
(during service) or chronic TCDD exposure over a longer window of time. Estimation of peak
exposures 20 years earlier is highly uncertain.  Also, the longer potential exposure window
occurred during a time period of decreasing exposure to TCDD and DLCs (Lorber and Phillips,
2002) further impeding the ability to estimate effective exposures. The uncertainty in identifying
a critical period of exposure precluded the estimation of an effective TCDD exposure.
Therefore, a quantitative dose-response analysis was not conducted for this study.

C.l.2.1.7.2. Lonsnecker andMichalek (2000)
C.l.2.1.7.2.1.  Study summary
       Longnecker and Michalek (2000) evaluated the relationship between serum levels of
TCDD and the incidence of diabetes and levels of serum glucose and insulin among veterans in
the AFHS.  However, unlike the earlier work on diabetes by Henriksen et al. (1997), and
Michalek et al. (2003), this study did not include those in operation Ranch Hand that were more
highly exposed to TCDD from the spraying of Agent Orange. Instead, this study was restricted
to the comparison group of male veterans in the AFHS who were never in contact with
dioxin-contaminated herbicides,  and whose serum TCDD levels were thought to fall within the
same range as the background levels found in the United States. These veterans included air and
ground personnel who participated in aircraft missions in Southeast Asia between August 1961
and May 1972. The manner in which this cohort of nonsprayers was assembled was originally
described by Wolfe et al. (1990).  A total of 1,667 comparison group veterans (i.e., non-Ranch
hands) were invited to participate in AFHS examinations in 1982. Subsequent examinations

                                        C-130

-------
were also conducted in 1985, 1987, and 1992. Participation rates were high (>70%) among this
comparison group of veterans, with 1,197 subjects available for analyses.
       Incident diabetes following each veteran's tour of duty was the primary health outcome
under study. This outcome was defined by either (i) self-reported physician diagnosis of
diabetes at any of the examinations (1982, 1987, and 1992) with subsequent verification of
medical records through June 1995, or (ii) by a postchallenge glucose test using 100 g of glucose
(positive status >200 mg/dL) in 1992. All incident cases of diabetes were type II. Levels of
serum and insulin were also measured using fasting, and 2-hour postchallenge tests in
nondiabetics.
       Serum dioxin levels were estimated using high resolution gas chromatography/high
resolution mass spectrometry using samples collected in the 1987 interview. For a small number
of veterans (n = 21) dioxin levels were estimated using serum collected in 1997.  For the
108 subjects with TCDD levels below the level of detection (1.25 ng/kg lipid), they were
assigned a TCDD level of 0.625 mg/kg. Those with serum TCDD levels above 10 ng/kg were
excluded as were those who lacked complete data for the covariates of interest. The covariates
that were examined as potential confounders included age, dioxin, body mass index, waist size,
and family history of diabetes, postchallenge glucose, and triglycerides. Analyses were based on
the remaining 1,197 veterans, and among these 169 incident cases of diabetes were identified.
       Logistic regression was used to  estimate the odds ratios and 95% confidence intervals of
diabetes across quartiles of serum TCDD levels, as well as in relation to a linear increase in
4.0 ng/kg of TCDD. The natural logarithm of serum-insulin levels was modeled again TCDD
levels using linear regression.  Results were adjusted for year of birth, race, military occupation,
body mass index at 1992, body mass  index at time of TCDD measurement and waist size in
1992. Ordinary least squares regression was used to evaluate associations between serum
glucose or insulin measures and quartiles of TCDD exposure. Adjustment was made for the
same covariates used in the logistic regression analysis.
       The adjusted odds ratio for diabetes increased with higher serum TCDD levels.
Specifically, an increase of 4.0 ng/kg of serum TCDD yielded an adjusted odds ratio of 1.55
(95% CI = 1.09-2.20).  After further adjustment for serum triglyceride levels,  the corresponding
odds ratio remained positive but was  attenuated (OR = 1.37, 95% CI = 0.96-1. 97). Associations
were also observed between serum TCCD and serum glucose (and insulin) levels, although some
                                         C-131

-------
of these were not statistically significant following adjustment for confounding.  This implies
that TCDD may contribute to increased insulin resistance and increased glucose levels among
those not satisfying the formal criteria for the diagnosis of diabetes.  The addition of serum
triglycerides to this model weakened these associations. The findings for both the outcomes of
diabetes and serum glucose were essentially unchanged after excluding subjects whose serum
TCDD was measured after 1987.

C. 1.2.1.7.2.2.  Study evaluation
       A strength of this study is the relatively large sample size (n = 1197) and corresponding
number of incident cases of diabetes (n = 169). However, while exposure levels are well
characterized using serum-based measure of TCDD, the primary limitation of this study is that
the analysis is essentially cross-sectional. The measurement of serum levels of TCDD occurred
following onset of diabetes for many of the veterans. On the other hand, associations between
dioxin exposure and diabetes during the most recent follow-up interval were dependent on serum
based TCDD exposures taken much earlier in 1987. In short, the findings did not account for the
timing of the exposure in relation to when diabetes was diagnosed. Therefore, the associations
may be noncausal. As noted by the authors, the onset of diabetes may have affected  dioxin
levels via the increased solubility of dioxides within increased serum triglycerides. Diabetes is
recognized to increase triglyceride levels, and adjustment for triglycerides attenuated the findings
in this  study.  Unlike the earlier study by Henriksen et al.  (1997), this study excluded the Ranch
Hand workers that had considerably higher exposures.  The much smaller range in exposures
along with the potential for  serum triglycerides to affect dioxin levels implies that there is a
greater potential for exposure misclassification across the groups used in this study than those
used by Henriksen et al (1997).
       The ascertainment of incident diabetes relied on either a self-reported measure with
confirmation through medical records, or a postglucose challenge serum test. These  are valid
methods to identify cases of diabetes mellitus. The possibility existed that those with lower
dioxin levels may have been less likely to participate in the follow-up examination, thereby,
leading to an under-ascertainment of diabetes among those with lower dioxin level.  However,
given a positive association was noted based on 1992 examination alone, and that participation
                                          C-132

-------
rates among those with 1987 dioxin less than the median was 91%, this potential source of bias
would likely be modest.

C.I.2.1.7.2.3.  Suitability of data for TCDD dose-response modeling
       The use of the individual-level TCDD serum measures and the identification of diabetes
through medical records and objective serum tests are strengths of this study, however, the
potential noncausal role of serum triglycerides cannot be dismissed. Additionally, there is
uncertainty in determining the critical window of exposure.  This was essentially a
cross-sectional analysis of diabetes in relation to a single point-in-time measure of TCDD
background exposure level that may have occurred over an approximate 20-year interval.
Considering the uncertainty in estimating the biologically relevant exposure window and the
uncertainty in estimating peak exposures 20 years prior to measurement, a quantitative
dose-response analysis was not conducted.

C.l.2.1.7.3. Michalek et al (2001a)
C.l.2.1.7.3.1.  Study summary
       Michalek et al. (200la) examined the relationship between TCDD exposure and
hematopoietic effects among veterans in the Air Force Health Study. A description of the overall
study design has been  described earlier, and can be found in the paper by Wolfe et al (1990).
This study included both veterans in the Ranch Hand unit, as well as those in a comparison
cohort who were not involved in the spraying of herbicides.
       The study used data collected from medical examinations and self-reported
questionnaires completed in 1982, 1985, 1987, and 1992. TCDD  levels were estimated using
serum collected in 1987, with some additional samples taken in 1992 for those who lacked
TCDD measurements.  In total, TCDD was assayed for 2,198 veterans. TCDD levels below the
limit of detection were assigned a value of 0 ppt. The study  excluded veterans with no TCDD
measure, those with TCDD levels above the level of detection but below the level of
quantification, and comparison subjects whose TCDD levels exceeded 10 ppt serum lipid
(threshold for background exposure). A first order kinetics model with a constant half-life of
8.7 years was used to estimate the initial TCDD dose at the end of the veterans' tours of duty in
Southeast Asia. Veterans were classified into four dioxin exposure groups: comparison cohort,

                                         C-133

-------
Ranch Hand—Background (<10 ppt), Ranch Hand—Low (10- <94 ppt), and Ranch
Hand—High (>94 ppt).
       At each of the four physical examinations, the following hematological characteristics
were measured: red blood cell count, hemoglobin, hematocrit, mean corpuscular volume, white
blood cell count, platelet count and erythrocyte sedimentation rate. Veterans who participated in
at least one examination, and who had a TCDD measurement were included unless they had a
fever (body temperature greater than 100°F) or they tested positive for human immunodeficiency
virus.
       Michalek et al. (200la) applied a linear regression model (adjusted for other covariates)
to calculate estimated mean differences in the various hematological measures among the
comparison group and the three other exposure groups. An adjusted test for trend was also
applied to the restricted group of Ranch Hand veterans. Logistic regression was used to estimate
the adjusted odds ratio for abnormally high or low hematological characteristics across TCDD
exposure categories. The measures of association were adjusted for the percentage of body fat,
year of birth, race, military occupation, and life-time smoking patterns. A secondary analysis of
mean corpuscular volume adjusted for current alcohol consumption was undertaken.
       There were no statistically significant differences in the mean values for red blood cell
counts, hematocrit, and white blood cell counts across the TCDD exposure categories in any of
the four examination periods. For three of the four examination periods, there was no
association observed between TCDD and hemoglobin.  Relative to the comparison group, the
mean corpuscular volumes were elevated among those in the highest exposure category in all
examination periods, while platelet counts were higher in three of the four periods.  Overall,
corpuscular volumes were about 1% higher among the most highly exposed Ranch Hands
compared to the comparison cohort, while the corresponding increase  was 4% with platelet
counts.
       Logistic regression analysis of abnormal red blood cell counts  across TCDD exposure
categories was hampered by small sample sizes. Typically, there were fewer than
four abnormalities in each of the four examination periods. In contrast, there was some evidence
for abnormally high platelet counts,  abnormally high mean corpuscular volume, and abnormally
high hematocrit in the highest Ranch Hand exposure group in some, but not all examination
periods.
                                        C-134

-------
       Michalek et al. (200la) suggested that the increased corpuscular volumes may be
explained by the noncausal effects of TCDD on serum triglycerides.  Other possible explanations
are also available for these associations, such as increased gamma-glutamyl transferase.

C. 1.2.1.7.3.2.  Study evaluation
       Strengths of the study included an assessment of dioxin at an individual-level using
serum based measures, a lengthy follow-up period that extended 30 years postservice, multiple
physical examination, and the use of valid methods of hematological function.  There are some
uncertainties in the estimation of TCDD exposure given serum was drawn decades after the
exposure period. Exposure misclassification may have been introduced from measurement error
in exposure estimates due to variations in metabolism, use of an assumed half-life of TCDD, and
calculations based on first-order decay.  The authors note considerable uncertainty in the
classification of the Background Ranch Hand veteran group as it comprised a mixture of exposed
and unexposed individuals. However, it is hard to gauge whether any exposure misclassification
would be differential by the health endpoints that were examined.
       For the most part, there were no  associations between hematological measures and
TCDD exposure. As noted by the authors, the associations between TCDD and mean
corpuscular volume may not be causally related. It may be a spurious association due to the
influence of TCDD on triglycerides levels which in turn affect corpuscular volume, or be due to
an increased prevalence of liver impairment previously noted in the cohort (Grubbs etal., 1995).
The positive association between TCDD and platelet count cannot be attributed directly to
TCDD given that many health conditions, which were not controlled for in the  analysis, may
have influenced platelet levels. Furthermore, the relationships identified are not supported by
other animal or epidemiologic literature, making interpretation of the associations difficult.

C.I.2.1.7.3.3.  Suitability of data for TCDD dose-response modeling
       There was no consistent association between TCDD serum levels and the hematological
measures of red and white blood cell counts, hemoglobin,  hematocrit, and erythroctyes. While
corpuscular volume and platelet counts were both positively associated with TCDD levels at
multiple examinations, evaluations of the data did not determine whether increases in these
measures were due to TCDD exposure during the Vietnam War.  These increases may be due to

                                         C-135

-------
noncausal associations from increased levels of triglycerides, or increased prevalence of mild
liver abnormalities among those with higher exposures (Grubbs et al., 1995), or the presence of
other comorbid health conditions that were not controlled for in the analysis. The findings of
associations that were small in magnitude between hematological function and TCDD likely
have little clinical relevance, but could provide some insight on biological mechanism of disease
from exposure to dioxin.
       This study analyzes the potential for associations between point-in-time measures of
TCDD serum levels and changes in hematological measures that may have occurred at any time
over approximately a 30-year interval, which precludes estimation of an effective TCDD
exposure over time.  EPA is uncertain whether TCDD-mediated changes in hematological
measures are the consequence of an elevated TCDD exposure event over a relatively short period
of exposure (during service) or chronic TCDD exposure over a longer window of time due to
slow TCDD elimination rates. Also, the long potential exposure window occurred during a time
period of decreasing background exposure to TCDD and DLCs (Lorber and Phillips, 2002) likely
decreasing the accuracy of the estimated exposure levels. Given the uncertainty in defining the
critical window of exposure and the inability to estimate an effective TCDD exposure over time,
quantitative dose-response analysis was not conducted for this study.

C.l.2.1.7.4. Michalek et al (2001b)—hepatic health outcomes
C.l.2.1.7.4.1.  Study summary
       Michalek et al. (200Ib) investigated the association between TCDD and the prevalence
of liver disease, and other indices of hepatic function in the Air Force Health Study.  The study
population included both Ranch Hands, as well as a comparison group  of veterans.  A detailed
description of the study design and methods is provided in earlier sections, as well as the paper
by Wolfe etal. (1990).
       This study relied on data collected at physical examinations conducted in 1982, 1985,
1987, and 1992. TCDD levels were estimated using serum collected in 1987, with some
additional samples taken in 1992 for those who lacked TCDD measurements. In total, TCDD
was assayed for 2,198 veterans. TCDD levels below the limit of detection were assigned a value
of 0 ppt.  The study excluded veterans with no TCDD measure, those with TCDD levels above
the level of detection but below the level of quantification, and comparison subjects whose

                                         C-136

-------
TCDD levels exceeded 10 ppt serum lipid (threshold for background exposure). A first order
kinetics model with a constant half-life of 8.7 years was used to estimate the initial TCDD dose
at the end of the veterans' tours of duty in Southeast Asia.  Veterans were classified into four
dioxin exposure groups: (i) Comparison cohort, (ii) Ranch Hand—Background (<10 ppt),
(iii) Ranch Hand—Low (10- <94 ppt), and (iv) Ranch Hand—High (>94 ppt).
       At each examination, participants were asked whether (1) a physician had informed them
that they had an enlarged liver, cirrhosis, or other liver condition (2) a physician had determined
presence or absence of hepatomegaly by palpitation, or (3) the presence or absence of liver
function test abnormalities through laboratory examination. All self-reported cases of liver
disease were confirmed through verification of medical records through 1993. In 1992, several
indices of liver function were measured using serum. These include: alanine aminotransferase,
aspartate aminotransferase, y-glutamyltransferase, lactic dehydrogenase, alkaline phosphatase,
and total bilirubin
       Michalek et al. (2001b)conducted statistical analysis for the  measures of liver function
collected during the 1992 examination, since they state that "the liver function test results for
1992 were not consistently different from those of previous examination." Mean values of liver
function were compared across the four categories of exposure using a linear model with a
log-transformation of liver function measures to enhance normality.  An adjusted test for trend
was also applied to the restricted cohort of Ranch Hands veterans.  All analysis was adjusted for
the history of liver disease, percentage of body fat, year of birth, race, military occupation,
lifetime industrial chemical exposure, lifetime degreasing chemical  exposure, as well as life-time
smoking and alcohol consumption. Enlisted Ranch Hands who had served in the ground crew
were analyzed separately because this subgroup was found to have the highest TCDD exposure.
The numbers of veterans included in the analysis of liver function tests across Comparison,
Background, Low and High TCDD exposure groups were 1195, 398, 262, and 264, respectively.
Logistic regression was used to evaluate the association between TCDD exposure and the
prevalence of liver diseases. These analyses were done among those who volunteered for at least
one examination, with valid dioxin measures, and excluded those with a history of liver disease
before their service in  Southeast Asia.  The numbers of veterans included in the analysis of liver
disease prevalence across Comparison, Background, Low and High TCDD exposure groups was
1,266; 420; 284; and 283, respectively.

                                         C-137

-------
       There was no association between TCDD exposure and hepatomegaly, or nonalcoholic
chronic liver disease (p-value linear test for trend=0.6). TCDD exposure was found to be
associated with other liver disorders. Compared to non-Ranch Hand veterans, the adjusted odds
ratio in the "high" exposure group was 1.6 (95% CI = 1.2-2.1). Laboratory measures associated
with these disorders were also found to be increased.  An increased level(s) of transaminase or
lactate dehydrogenase was found in veterans in the "high" exposure group (OR = 2.7,
95% CI = 1.4-5.1), and a dose-response trend was noted across exposure categories (p = 0.03).
Additionally, an increased odds ratio for nonspecific liver abnormalities was found in the same
"high" exposure group (OR = 1.4, 95% CI = 1.0-2.0), while no association was noted for
hepatomegaly. There were no statistically significant dose-response trends between TCDD and
any of the mean hepatic measures (AST, ALT, GOT, LDH, Alkaline phosphatase, or total
bilirubin) based on the 1992 serum data, although p-values for tests of trends for alkaline
phosphatase and y-glutamyltransferase (GOT) were 0.06. Statistically significant increases
(p < 0.05) in mean GGT levels were noted among those in the highest TCDD exposure group
relative to the comparison cohort. No consistent patterns were detected when results were
stratified by drinking history or current alcohol use, but GGT levels tended to increase across
current drinking levels,

C.l.2.1.7.4.2.  Study evaluation
       Strengths of this study include the high rate of participation, low attrition rate,
appropriately matched comparison group, and the decade long follow-up period. Within some of
the exposure categories, relatively few cohort members were diagnosed with several of the liver
conditions following their tours of duty. For example, there were only 10 veterans in the high
exposure group diagnosed with hepatomegaly, and only 5 diagnosed with nonalcoholic liver
disease and cirrhosis.  As such, the statistical power to detect some associations that may be
present was limited.

C.l.2.1.7.4.3.  Suitability of data for TCDD dose-response modeling
       The results do not unequivocally support a relationship between liver damage  and TCDD
exposure. Confounding and reverse causality cannot be eliminated as possible explanations of
the study results, and the clinical significance of the results (which were small in magnitude) is

                                         C-138

-------
unclear. Additionally, there is uncertainty in determining the critical window of exposure. This
study analyzes the potential for associations between point-in-time measures of TCDD serum
levels and possible changes in hepatic measures that may have occurred at any time over
approximately a 30-year interval. Thus, it is unclear whether the differences in serum enzyme
levels and liver function measures potentially affected by TCDD exposures are the consequence
of an elevated TCDD exposure event over a relatively short period of exposure (during service)
or chronic TCDD exposure over a longer window of time due to slow TCDD elimination rates.
Also, the long potential exposure window occurred during a time period of decreasing
background exposure to TCDD and DLCs (Lorber and Phillips, 2002) further impeding the
ability to estimate dose accurate. Considering the uncertainty in estimating the biologically
relevant exposure window and the uncertainty in estimating peak exposures 20 years prior to
measurement, a quantitative dose-response analysis was not conducted.

C.l.2.1.7.5.  Michalek et al.  (2001c)—peripheral neuropathy
C.l.2.1.7.5.1.  Study summary
       Michalek et al. (200Ic) studied the relationship between TCDD exposure and peripheral
neuropathy among veterans in the Air Force Health Study. The study included the Ranch Hands
who were involved in the  spraying of herbicides in Southeast Asia, as well as a comparison
cohort of veterans.  The study population and design has been described earlier in this section,
and is detailed in the publication by Wolfe et al. (1990).
       This study relied on data collected at physical examinations conducted in 1982, 1985,
1987, 1992 and 1997.  TCDD levels were estimated using  serum collected in 1987, with some
additional samples taken in 1992 for those who lacked measures. In total, TCDD was assayed
for 2,198 veterans.  TCDD levels below the limit of detection were assigned a value of 0 ppt.
The study excluded veterans with no  TCDD measure, those with TCDD levels above the level of
detection but below the level of quantification, and comparison  subjects whose TCDD levels
exceeded 10 ppt serum lipid (i.e., the threshold for background exposure). A first-order kinetics
model with a  constant half-life of 8.7 years was used to estimate the TCDD levels at the end of
the veterans' tours of duty in Southeast Asia. Veterans were classified into four dioxin exposure
groups: (i) Comparison cohort, (ii) Ranch Hand—Background (<1 ppt), (iii) Ranch Hand—Low
(10- <94 ppt), and (iv) Ranch Hand—High  (>94 ppt).

                                         C-139

-------
       Blinded neurological examinations were conducted on volunteers at each of the
five examinations by staff who were blinded to the veterans' exposure levels. These
neurological examination included evaluations of cranial nerves, muscle strength in both lower
and upper limbs, sensory perception of pain, light touch, vibration, proprioception, activity of
deep tendon reflexes, stance, gait, hand and foot coordination, and tremor. Velocities of nerve
conduction were conducted in 1982, while vibrotactile thresholds of the left and right toes were
measured in 1992 and 1997. The study excluded veterans with a history of neurological
disorders prior to their service in Southeast Asia. The analysis also excluded veterans with
disorders that could interfere with peripheral nerve assessments. These conditions included:
quadriplegia, injuries or amputations, and alcohol-related disorders. Diabetes status was also
determined as described by Longnecker and Michalek (2000). Michalek et al. (200 Ic) analyzed
data using main effects logistic and linear regression models.  An adjusted test for trend was also
applied. All measures of association were adjusted for body mass index, year of birth, height,
and alcohol consumption.  As in the Michalek et al. (200Ib) study, enlisted Ranch Hands who
had served in the ground crew were analyzed separately. Diabetics and nondiabetics were also
analyzed separately.  Furthermore, the data was analyzed in two rounds, with the second round
excluding veterans with neurologic conditions with known causes unrelated to dioxin exposure,
which could impact the neurological findings.
       No association was observed between TCDD and nerve conduction velocities in 1982,
and there were no statistically significant associations found for 'any symmetrical peripheral
abnormalities' in 4 of the 5 examinations. However, based on the 1997 examination, those in the
highest exposure category had an increased risk of any symmetrical peripheral abnormality
(OR =  1.8, 95% CI = 1.2-2.7).  These associations were stronger for 'probable' symmetrical
peripheral neuropathy than they were for those designated as possible. There was no evidence of
effect measure modification by diabetes status for TCDD associations with probable  peripheral
neuropathy in the 1997.  An interaction was found  between diabetes status and current dioxin
exposure for diagnosed neuropathy in 1997.  Additional restrictions excluding veterans with
diseases, disorders or other exposures that may have produced neuropathic symptoms resulted in
groups that were too small to further analyze.
                                          C-140

-------
C.l.2.1.7.5.2.  Study evaluation
       The strengths of this study are the same as described for the Michalek et al. (200 la:
2001b) studies. Uncertainty in the critical window of exposure, as well as uncertainty in
exposure classification present in the Michalek et al. (200 Ib), are also weaknesses of this study.
The Michalek et al. (200 Ic) study attempts to characterize risks of neuropathy while accounting
for the possible modifying influence of diabetes. While the associations are strong, they are
limited by the relatively small number of cases in the "high" exposure group.  Moreover,
associations were for the most part, confined to only one of the five examination intervals.  A
large number of comparisons were conducted in this study using multiple measures of
neuropathy that were assessed at up to 5 examination periods. As a result, the multiple
comparisons performed increase the chance of detecting a false-positive association due to the
number of statistical hypothesis tests performed.

C.l.2.1.7.5.3.  Suitability of data for  TCDD dose-response modeling
       The dose-response relationship between TCDD exposure and peripheral neuropathy is
strong, and supported by several important strengths. However, associations were not consistent
across the different examinations, and further work is needed to evaluate the relationship
between diabetes and peripheral neuropathy in this cohort.  Some comparisons are limited by a
small number of outcomes particularly in the highest exposure group. Additionally, there is
uncertainty in the critical window of exposure. This study analyzes the potential for associations
between peripheral neuropathy and point-in-time measures of TCDD serum levels that may have
occurred at any time over approximately a 30-year interval, making it difficult to calculate a
TCDD effective dose over time.  Thus, it is unclear whether the peripheral neuropathies are the
consequence of an elevated TCDD exposure event over a relatively short period of exposure
(during service) or chronic TCDD exposure over a longer window of time due to slow TCDD
elimination rates.  Also, the long potential exposure window occurred during a time period of
decreasing background exposure to TCDD and DLCs (Lorber and Phillips, 2002) further
impeding the ability to estimate dose accurately.  For these reasons, a quantitative dose-response
analysis was not conducted for this study.
                                         C-141

-------
C.l.2.1.7.6.  Pavuk et al (2003)—thyroid health endyoints
C.l.2.1.7.6.1.  Study summary
       Pavuk et al. (2003) published an analysis that examined the effects of TCDD exposure on
thyroid function among veterans enrolled in the AFHS.  A summary of the design of the AFHS
study and methods have been already described in this section, and are provided in greater detail
in the paper by Wolfe et al. (1990).  This current study included both those involved with
Operation Ranch Hand, as well as a comparison cohort of other veterans who served in Southeast
Asia but who were not involved with spraying of herbicides. The objective of this study was to
examine associations between TCDD levels estimated in 1987 and  several measures of thyroid
function, as well the incidence of six different thyroid diseases following the completion of the
veterans' tours of duty.
       The study used data collected from medical examinations and self-reported
questionnaires completed in 1982,  1985, 1987,  1992, and 1997. TCDD levels were estimated
using serum collected in 1987, with some additional samples taken in 1992 and 1997 for those
who lacked measures. For those with serum measures taken in 1992 or 1997, a first order
kinetics model with a constant half-life of 8.7 years was used to extrapolate values to 1987.
Veterans were classified into four dioxin exposure groups: comparison cohort, Ranch Hand—
Background (<10 ppt), Ranch Hand—Low (10- <94 ppt), and Ranch Hand—High (>94 ppt).
       Thyroid diseases that occurred following the veterans' tours of duty were identified
through self-report of physician diagnosis at any of the five physical examinations and verified
from medical records. The following conditions were considered: unspecified goiter, nontoxic
nodular goiter, thyrotoxicosis, acquired hypothyroidism, thyroiditis, and other disorders of the
thyroid. Congenital hypothyroidism was not examined as this  condition would have prevented
individuals from entering the military.  Serum samples were used to obtain measures of thyroid
function. T4 and TSH were estimated at each of the five examinations, while triiodothyronine
percent (T3%) was determined in 1982, 1985, and 1987. The free thyroxine index (FTI) was
only estimated in 1982. Veterans who participated in at least one examination,  and who had a
TCDD measurement were included unless they were being treated with thyroid medication, had
a previous thryroidectomy or irradiation, or were  diagnosed with a thyroid disease before their
service had ended.
                                         C-142

-------
       For each physical examination, cross-sectional analysis was performed to compare the
mean levels of TSH, T4, T3%, and FTI across the four TCDD exposure categories. A repeated
measures linear model was used to compare mean TSH, T4, and T3% values across exposure
categories using data from all five examinations combined. This model took into account the
repeated nature of the data by using an autoregressive order one covariance structure.  Logistic
regression was used to estimate the OR of thyroid diseases across TCDD exposure categories, as
well as abnormally high TSH levels across the five examinations. These models were adjusted
for confounding by age, race, and military occupation.
       No association was found between TCDD and any of the  six thyroid diseases that were
examined. In four of the five examinations, higher TSH values were observed in the higher
TCDD exposure categories. A dose-response relationship was observed in the longitudinal
analyses of these data (p = 0.002).  The ORs of an abnormally high TSH among the high
exposure Ranch Hand  group ranged from 1.4 to 1.9 relative to the comparison group, but was not
statistically significant in any of the five examinations (p > 0.05). No significant associations
were reported with  either the cross-sectional or longitudinal analyses of the total T4 levels
(mean), T3% uptake, or FTI.

C.l.2.1.7.6.2.  Study evaluation
       The overall  size of the cohort was relatively large as analyses were based on 1,009 Ranch
Hands, and 1,429 comparison veterans. However, there were relatively few thyroid disorders
identified among these veterans following their tour of duty. Specifically, there were only
188 such veterans, and therefore, analyses of the relationship between these six different
disorders and the four categories of TCDD exposure was limited by statistical power.
       Strengths of this study include the estimation of TCDD levels using serum, and the
consideration of several different outcome measures of thyroid disorders from questionnaire
data, as well as serum TSH, T3% uptake, T4, and FTI measurements. Thyroid function was
assessed multiple times using serum-based measures that are valid and widely used. While the
authors did not take into account the timing of disease onset for the thyroid conditions examined,
the serum-based measures of TCDD in 1987 allowed for veterans to be classified according to
exposure status prior to onset of disease.  In particular, these exposure levels among the Ranch
                                         C-143

-------
Hands could be attributed to exposure received during their tours in Southeast Asia, and only
thyroid conditions that occurred following the tour of duty were considered.
       There was no association found between serum-based measures of TCDD and any of the
six thyroid conditions examined (unspecified goiter, nodular goiter, hyperthyroidism, thyroiditis,
or other thyroid disease).  The only thyroid measure that was associated with TCDD levels was
TSH. Higher levels of TSH were observed among those in the higher exposure categories, and a
dose-response relationship was observed when data across all examinations were modeled.
However, those in the highest exposure group did not have a statistically significant increased
risk of abnormal TSH levels irrespective of when the examination date. Taken together, the
findings suggest that TCDD may increase TSH levels which are a marker for an underactive
thyroid. Lower TSH levels over the long term may increase the risk of hypothyroidism, or
indicate thyroid hormone resistance.  However, the clinical implications are unclear in light of
the absence of an association between TCDD and any of the six thyroid conditions that were
examined. As noted by the authors, this cohort may not yet be old enough to determine whether
TCDD exposure increases the risk of developing thyroid disease.

C.l.2.1.7.6.3.  Suitability of data for TCDD dose-response modeling
       There was no association between TCDD exposure and any of the six thyroid diseases
that were examined. Further, there was no association between cross sectional or longitudinal
analyses of TCDD and T4, T3% uptake, or FTI. While a dose-response trend was observed with
TCDD and TSH levels, evidence of a statistically significant increase in abnormally high TSH
levels was not observed among veterans in the highest exposure group.  Additionally, there is
uncertainty in the critical window of exposure. This study examined associations between
thyroid conditions and measures of thyroid disorders with point-in-time measures of TCDD
serum levels that may have occurred at any time over approximately a 30-year interval. As a
whole, these analyses do not support an association between TCDD exposure and comprised
thyroid function, and therefore, a quantitative dose-response analysis was not conducted for this
study.
                                         C-144

-------
C.l.2.1.7.7. Michalek andPavuk (2008)—diabetes
C.l.2.1.7.7.1.  Study summary
       Michalek and Pavuk (2008) examined both the incidence of cancer and the prevalence of
diabetes in the cohort of Ranch Hand workers exposed to TCDD. As noted previously, these
veterans were responsible for aerial spraying of Agent Orange in Vietnam between 1962 and
1971. Exposure to TCDD was estimated using serum collected from (1) participants in 1987 or
(2) participants in 1992, 1997, and 2002 for those who had no quantifiable TCDD result in 1987,
those who refused in 1987, and those subjects who were new to the study.  Exposure to TCDD
was estimated using a first-order pharmacokinetic model with a half-life of 7.6 years and
provided an estimate of TCDD at the end of the tour of duty in Vietnam. Veterans were grouped
into four categories: comparison, background, low, and high. Diabetes was identified from
diagnoses during the post-Vietnam era from medical records.  Overall, no differences were
shown in the RR of diabetes between the Ranch Hand unit and the reference group (RR = 1.21,
p = 0.16).  Stratified analyses by days of spraying (<90 days, >90 days),  however, revealed a
significant increase in risk of diabetes (RR = 1.32, p = 0.04) among those who sprayed for at
least 90 days. A dose-response relationship was also evident when logioTCDD was modeled in
the combined cohort. Also, stratification by calendar period showed a dose-response relationship
for those whose last year of service was during or before 1969.

C.l.2.1.7.7.2.  Study evaluation
       The Michalek and Pavuk (2008) study provides an opportunity to characterize risks of
diabetes as the study is  not subject to some of the potential bias of case ascertainment based on
death certificates (D'Amico et al., 1999). The quality of the TCDD exposure estimates is good,
given that serum data were available at an individual-level basis for all Ranch Hand and
comparison veterans used in the cohort. However, there is significant uncertainty in the
biologically-relevant critical window of exposure. Also, the long lag between initial  exposure
and sera measurements limits the estimation of peak exposures 20 years  earlier.

C.l.2.1.7.7.3.  Suitability of data for TCDD dose-response modeling
       The reported dose-response relationship between TCDD and diabetes in the Michalek
and Pavuk (2008) study is supported by study strengths, including the use of the individual-level

                                         C-145

-------
TCDD serum measures and the identification of diabetes through medical records.  However, it
is unclear whether the diabetes cases are the consequence of an elevated TCDD exposure event
over a relatively short period of exposure (during service) or chronic TCDD exposure over a
longer window of time due to slow TCDD elimination rates. In addition, the long potential
exposure window occurred during a time period of decreasing background exposure to TCDD
and DLCs (Lorber and Phillips, 2002) further impedes the ability to estimate dose accurately.
For these reasons, a quantitative dose-response analysis was not conducted for this study.

C. 1.2.1.8.  Other noncancer studies of TCDD
       See general summaries of the Netherlands and New Zealand cohorts in Section C. 1.1.1.7.

C.l.2.1.8.1. Ryan et al (2002)—sex ratio
C.I.2.1.8.1.1.  Study summary
       Ryan et al. (2002) conducted an investigation on the sex ratio in offspring of pesticide
workers who were involved with the production of trichlorophenol and the herbicide 2,4,5-T in
Ufa, Bashkortostan, Russia. Ufa was the site of a state agrochemical plant that has been in
operation since the 1940s. Between 1961 and 1988, the plant employed more than 600 workers,
most in their early 20s. Females, however, accounted for about 15% of the workforce that
produced 2,4,5-T and 30% for 2,4,5-trichlorophenol.
       Serum samples previously taken in 1992 among 60 men, women, and children from the
factory and city of Ufa showed TCDD  exposures that were approximately 30 times higher than
background levels (Ryan and  Schecter, 2000). Blood data were subsequently measured on a
sample of 20 workers between 1997 and 2000, and on 23 2,4,5-trichlorophenol workers between
1997 and 2001.  In all, 84 individuals (67 men and 19 women) who provided blood samples
formed the basis of the analysis in this  study. Of these, 55 (43  men and 12 women) were
exposed to 2,4,5-T and 29 (22 men and 7 women) were exposed to 2,4,5-trichlorophenol. There
is no indication on how the individuals that were asked to provide and those who did provide
serum samples were selected.  Ryan et al. (2002) reviewed company records for these workers to
determine the number, sex, and date of birth of any children; birth data were available for
198 workers (150 men and 48 women). Awareness of the study led other workers who had not
                                        C-146

-------
provided serum to provide information on births that occurred 9 months after the time of first
employment in the factory.
       The authors calculated descriptive statistics for the 198 workers and compared them to
values for the city of Ufa between 1959 and 1996.  Tests of statistical significance were made
using the z-test, and the chi-square test. The observed proportion of male births (0.40) among
the factory workers was much lower than that for the city of Ufa (0.51) (p < 0.001).  Stratified
analyses revealed that this lower ratio was observed only among those paternally exposed to
TCDD.  Specifically, the proportion of male births among exposed fathers was 0.38  and among
exposed mothers was 0.51.  This pattern was observed in both the workers exposed to 2,4,5-T
(proportion of male births = 0.40) and 2,4,5-trichlorophenol (proportion of male births = 0.35).

C. 1.2.1.8.1.2.  Study evaluation
       The Ryan et al. (2002) findings are consistent with earlier work completed for Seveso
residents (Mocarelli et al., 2000). Although individual-level serum measures were available for
84 individuals, exposure-response relationships with birth ratios were not performed on these
data.  This approach would have been preferred and consistent with that which Mocarelli et al.
(2000) used. All comparisons were made using an external comparison group, namely the sex
ratio observed in Ufa between 1959 and 1996.
       Although serum measures were used to describe TCDD exposure for a sample of the
workers (selection  criteria for these workers was not provided), individual-level dose estimates
were not calculated for the study population. Specifically, exposures were characterized many
years after exposure, and no attempt was made to back-extrapolate to the time of conception.
The two groups of workers in the study also reportedly had high exposure levels of
1,2,3,7,8-pentachlorodibenzo-p-dioxin. So, the group level exposure classification (by plant) did
not allow consideration of potential confounding due to other DLCs. Another limitation of the
study is that the study population is likely nonrepresentative of all workers employed at the plant.
Participants included only those willing to provide serum samples and those who volunteered to
participate in the study after learning about it in a public forum.  If participation was dependent
on TCDD exposures and the reproductive health of these subjects, then bias may have occurred.
                                         C-147

-------
C.I.2.1.8.1.3.  Suitability of data for TCDD dose-response modeling
       The findings are notable in their consistency with those found in Seveso residents by
Mocarelli et al. (2000). For the Ryan et al. (2002) study, serum data were quantified at an
individual-level basis. Risk estimates, however, were not derived in relation to these exposures
but instead in two separate subgroups (2,4,5-T and 2,4,5-trichlorophenol workers).  Because of
this important limitation and the uncertainty in the biologically-relevant critical window of
exposure, a quantitative dose-response analysis was not conducted for this study.

C.l.2.1.8.2.  Kans et al(2001)—Ions-term health effects
C.I.2.1.8.2.1.  Study summary
       Kang et al. (2001) investigated the relationship between self-reported health measures
and serum-based measures of TCDD in a group of 1,499 Vietnam veterans and a control group
of 1,428 non-Vietnam veterans. The study subjects were identified from (1) reports of Army
Chemical Corps detachments in Vietnam between 1966 and 1971,  (2) personnel records of
individuals involved in chemical operations who were on active duty between 1971 and 1974,
and (3) class rosters of personnel who were trained at Fort McClellan in Alabama between 1965
and 1973. The comparison group was selected so that branch of service, time period, and
military occupation were similar to those of the subjects with the exception that they did not
serve in Vietnam. Although 2,872 Vietnam veterans and 2,732 non-Vietnam veterans were
identified as potential subjects, those who were deceased as of December 1998 and those who
had previously participated in a pilot study were excluded. The study targeted 2,247 Vietnam
and 2,242 non-Vietnam veterans.
       Exposure to TCDD was characterized for subsets of the study population that provided
blood samples, specifically 795 of 1,085 (73%) Vietnam veterans and 102 of 157 (65%)
non-Vietnam veterans. Details on these individuals selected for participation in the serum dioxin
study were not presented. The  authors did state, however, that due to economic constraints, only
897 serum samples could be analyzed.  Blood specimens were collected in 1999-2000 at
individuals'  homes. TCDD concentrations were analyzed by laboratory staff blind to the group
status (i.e., Vietnam or non-Vietnam) of the study subjects.
       Prevalent health outcomes were ascertained by self-reported information on selected
conditions diagnosed by a medical doctor. The following conditions were included: diabetes,

                                         C-148

-------
hepatitis (all types combined), heart disease, all cancer, nonmalignant chronic respiratory
diseases, and hypertension.  Health-related quality of life was evaluated using the SF-36 survey
instrument (Ware etal., 1993).
       Eligible veterans whose current residences (4,119 total) could be identified were
contacted for study participation. Survey participation rates were 73% for Vietnam veterans,
yielding data for 1,499 individuals, and 69% for non-Vietnam veterans, yielding data for
1,428 non-Vietnam veterans.  The survey data showed that, relative to non-Vietnam veterans,
Vietnam veterans were more likely to be regular smokers and to be obese. They also were more
likely to be enlisted personnel, and a much higher proportion was 51 years of age or older
(83% vs. 58%). After adjusting for age, race, smoking status, rank, and body mass index, the
prevalence of self-reported health conditions was found to be statistically significantly higher in
the Vietnam group.  The adjusted ORs were as follows: diabetes, OR =1.16
(95% CI = 0.91, 1.49); hepatitis, OR = 1.85 (95% CI = 1.30, 2.64); heart condition, OR =  1.09
(95% CI = 0.87, 1.38); all cancer, OR = 1.46 (95% CI = 1.02, 2.10); nonmalignant respiratory
condition, OR = 1.41 (95% CI =1.13, 1.76); and hypertension,  OR = 1.06 (95% CI = 0.89, 1.27).
       For those with Vietnam service, the mean serum TCDD concentrations were higher
among those who reported spraying herbicides (4.3 ppt) than those who did not (2.7 ppt)
(p < 0.001).  The investigators did not back-extrapolate serum levels to the time when
individuals last sprayed.  The adjusted ORs (adjusted for age, cigarette smoking, body  mass
index, rank,  and race) for most chronic health conditions examined revealed increased
prevalence among Vietnam sprayers relative to non-Vietnam sprayers. These ORs included:
diabetes, OR = 1.49 (95% CI =1.10, 2.02); hepatitis, OR = 1.40 (95% CI = 0.92, 2.12); heart
condition, OR = 1.41 (95% CI = 1.06, 1.89); all cancer, OR = 1.36 (95% CI = 0.91, 2.04);
nonmalignant respiratory condition, OR = 1.57 (95% CI = 1.20, 2.07); and hypertension,
OR= 1.26(95%CI=1.00,  1.58).
       The investigators also examined the possibility of over-reporting of chronic health
conditions by comparing the prevalence of self-reported conditions among 357 Vietnam sprayers
who mean serum TCDD levels of 2.5 ppt compared to those who had levels less than 2.5 ppt.
Prevalence of diabetes, heart condition, and hypertension, was higher among those with mean
serum TCDD levels of 2.5 ppt, although no levels of statistical significance were reported. Data
for cancer were not presented.
                                         C-149

-------
C. 1.2.1.8.2.2.  Study evaluation
       Data were collected from only half of the individuals in the study target population, so
there is some potential for selection bias in this study. First, the study excluded those who had
died before 1999, excluding potentially important TCDD-related adverse health effects that
could result in death more than two decades after veterans had been actively spraying. Survey
participation rates were 73% for Vietnam veterans and 69% for non-Vietnam veterans. If those
in poorer health were less inclined to participate, the prevalence of the selected chronic health
conditions would be understated. Selection bias due to study participation could also be possible
if, for example, those in poorer health also had higher (or lower) exposures than those not
participating in the study.  The lack of direct evidence of differential participation and reports of
comparable prevalence rates of hypertension and diabetes to other general  populations suggests
that selection bias may be minimal.
       Because the data collected are cross-sectional, they are not well suited for evaluating the
relationship between the timing of exposure and the onset of disease. Whether any of the data
could help identify when the chronic health conditions were diagnosed is unclear.  Given the
long period covered by the study, many of the self-reported health conditions likely were
diagnosed some time ago, perhaps closer to the time of potential TCDD exposure. Such detail is
needed to characterize health risks associated with specific TCDD levels, particularly given that
TCDD levels have been demonstrated to decrease from time of last exposure.
       An important strength of the study is the availability of blood sera for a subset of the
study population, which allows for individual-level estimates of TCDD exposure.  Serum TCDD
levels were available for only 897 subjects, however, which limits the ability to examine the
relationship between measures of TCDD and prevalence of health outcomes without restricting
the sample  size or extrapolating exposure levels to the whole study population. For example,
among sprayers with available TCDD exposure data only 60 cases  of diabetes and 69 cases of
heart disease were examined relative to exposure.  Also, the small number of cancers precluded a
site-specific cancer analysis. Moreover, whether these TCDD  levels are representative of the
larger eligible population is difficult to gauge, given that deceased veterans and those whose
current residences could not be determined were excluded.
       The study relied on self-reported measures of disease prevalence.  The ascertainment of
chronic health conditions using self-reported data can be fraught with difficulties.  For example,

                                          C-150

-------
the sensitivity of self-reported data when compared to medical diagnosis has been shown to be
poor for conditions such as diabetes and hypertension (Okura et al., 2004). As Kang et al. (2006)
state, prevalence studies are not be well suited to examine rare diseases with short survival times
such as cancer.  In addition, self-report of physician-diagnosed cancers by study subjects often
lacks the sensitivity needed in most epidemiologic studies as they can be influenced by a variety
of factors including age and education (Navarro et al., 2006).
       The potential for biases in the reporting of health outcomes between the sprayers and the
non-Vietnam veterans (i.e., differential by TCDD exposure status) is plausible, given the public
attention that spraying of Agent Orange has received. Although the authors examined whether
over-reporting was related to outcome prevalence among herbicide sprayers (prior to collection
and determination of actual TCDD serum levels), the possibility exists that these subjects
reporting could be influenced by their perceived level of exposure from herbicide spraying. The
authors also examined the  potential for misreported diabetes by conducting a medical records
review of 362 veterans.  Seventy-nine percent of the self-reported diabetes cases were confirmed
with medical records. The documentation rate was also comparable between the Vietnam
veterans and the non-Vietnam veterans suggesting that differential reporting was not an issue for
this health outcome.
       Because the Vietnam veterans group comprised professional sprayers, it is not
unreasonable to assume that they would have been exposed to other potentially harmful agents
either during their service in Vietnam, or from the end of their service to when they provided
data in 1999-2000.  This study did not control for other, potentially relevant occupational
exposures.

C.I.2.1.8.2.3.  Suitability  of data for TCDD dose-response modeling
       Although the study demonstrates increased prevalence of several chronic health
conditions, these findings should be interpreted with caution due to the potential for selection
and recall biases.  Because of the lack of demonstrated dose-response relationships with cancer
or other outcomes and uncertainty in the biologically-relevant critical exposure window, a
quantitative dose-response analysis was not conducted for this study.
                                          C-151

-------
C.l.2.1.8.3. McBride et al (2009a) —noncancer mortality
C.l.2.1.8.3.1.  Study summary
       The McBride et al. (2009a) mortality study of New Zealand workers employed as
producer or sprayers with potential exposure to TCDD was described earlier in this report.
These individuals were employed at a plant that manufactured 2,4,-dichlorophenoxyacetic acid,
and later 2,4,5-T and 4-chloro-2-methyphenoxyacetic acid. In 1987, the plant closed and 2,4,5-T
production ceased in 1988.
       The cohort consisted of 1,754 individuals who were employed for at least one day at the
New Plymouth site between January 1, 1969, and October 1, 2003. Vital  status was determined
until the end of 2004, and 247 deaths occurred during this time period.  Comparisons of mortality
were made to the New Zealand general population.  Exposure was characterized by duration of
employment. Person-years of follow-up were tabulated across strata defined by  age, calendar
period, duration of employment, sex, latency, and period of hire. Analyses were stratified to
compare risks by duration of employment (<3 or>3 months), latency (<15 or >15 years),  and
period of hire (<1976 or >1976).
       Overall, no statistically significant differences in all-cause mortality relative to the
general population were found among those who worked for at least 3 months (SMR = 0.92,
95% CI = 0.80-1.06) or for less than 3 months (SMR = 1.23, 95% CI = 0.91-1.62). No
statistically significant excesses were found for mortality from diabetes, cerebrovascular disease,
heart disease, or accidents.  The incorporation of a latency period of 15  years revealed no
statistically significant excesses for these same causes of death. Similarly, no excesses for any
cause of death were noted among those who were hired either before or after 1976.
       In subsequent analyses of the same cohort that used estimated TCDD levels from  serum
samples, McBride et al. (2009b) found no excesses for all-cause mortality or mortality from
diabetes or heart disease.

C. 1.2.1.8.3.2.  Study evaluation
       For the McBride et al. (2009a) study, the size of the cohort is large enough to characterize
mortality risks relative to the general population for most common causes of deaths.  An
important limitation of this study is the loss to follow-up of a substantial percentage of workers
(22%).  This would have impacted statistical power by reducing the number of deaths among the

                                         C-152

-------
workers.  If this incomplete ascertainment of mortality outcomes did not occur in a similar
fashion with the general population then the results may also be biased.
      For noncancer causes of death, the use of the SMR statistic is more likely to be
influenced by the healthy-worker effect. Therefore, the findings obtained for these outcomes
should be interpreted with caution. Subsequent analyses published by the same authors
(McBride et al., 2009a) provide improved characterization of TCDD exposure using serum
samples.

C. 1.2.1.8.3.3.  Suitability of data for dose-response analysis
      Overall, no associations were evident between surrogate measures of TCDD (duration of
employment, year of hire) and noncancer mortality outcomes.  As all outcomes were based on
mortality, dose-response modeling was not conducted for this study.

C.l.2.1.8.4.  McBride et al (2009b)—noncancer mortality
C.l.2.1.8.4.1.  Study summary
      McBride et al. (2009b) further analyzed the cohort of New Zealand workers to include
estimates of TCDD exposure based on serum samples.  Current and former employees who were
still alive and living within 75  km of the site were asked to provide serum samples. Samples
were collected from 346 workers representing 22% (346/1599) of the entire study population.
These serum measures were used to estimate cumulative TCDD levels for all workers. The
exposure assessment approach by Flesch-Janys et al. (1996) was used to estimate time-dependent
exposures based on area under the curve models.  This was based on a one-compartment
first-order kinetic model with a half-life of 7.2 years.
      Comparisons of mortality were made to the general population using the SMR. The Cox
proportional hazards model was used to conduct an internal cohort analysis across
four categories of cumulative TCDD levels for diabetes and ischemic heart disease mortality.
The RRs generated from these models were adjusted for sex, hire year, and birth year. No
diabetes deaths were observed among women, and therefore, analysis of this outcome was
limited to men.
      Relative to the general  population, no difference in the all-cause mortality experience was
observed in exposed cohort members (SMR = 1.0, 95% CI = 0.9-1.2).  Similarly, no excess in

                                         C-153

-------
these workers was observed for heart disease (SMR =1.1, 95% CI = 0.9-1.5); cerebrovascular
disease (SMR =1.1, 95% CI = 0.6-1.9); diabetes (SMR = 0.7, 95% CI = 0.2-2.2); or
nonmalignant respiratory disease (SMR = 0.8, 95% CI = 0.4-1.4). For the internal cohort
analysis, the RR associated with cumulative categorical TCDD measure was 1.0 for both
diabetes and ischemic heart disease.

C.l.2.1.8.4.2.  Study evaluation
       The McBride et al. (2009b) study extends their earlier work in two ways. First, serum
measures were used to estimate cumulative TCDD with  methodology that has been applied to
several other cohorts of workers exposed to TCDD. Second, they used regression analyses that
examined individual-level TCDD exposures in relation to various outcomes as part of the
internal cohort comparisons.  For noncancer outcomes, no dose-response associations with
TCDD were observed with the internal comparisons.  Also, as found with earlier analyses of this
same cohort, no excess noncancer mortality relative to the New Zealand general population was
observed.
       Associations between TCDD and diabetes have been found previously in TCDD-exposed
populations, most notably in the Ranch Hands cohort (Michalek and Pavuk, 2008). In this
cohort, only five deaths from diabetes were identified, and of these, only three occurred among
those who were exposed to TCDD.  The study, therefore, has limited statistical power to
characterize associations between TCDD and mortality from diabetes.  Further, the identification
of diabetes deaths is subject to misclassification errors due to under-reporting (McEwen et al.,
2006).

C.l.2.1.8.4.3.  Suitability of data for TCDD dose-response modeling
       McBride et  al. (2009b) found no statistically significant associations in any of the
noncancer causes of death. As all outcomes were based on mortality, dose-response modeling
was not conducted for this study.

C. 1.2.2.  Feasibility of Dose-Response Modeling for Noncancer
       Relatively few study populations permit  quantitative dose-response modeling to be
performed for noncancer outcomes. The serum collected among Seveso men and women

                                         C-154

-------
provide an opportunity to characterize risks for several health conditions in relation to TCDD
exposure. The collection of these serum samples, shortly after the accident does not require the
back-extrapolation of TCDD levels as in the occupational cohorts, which should reduce the
exposure assessment uncertainty and minimize the potential for exposure misclassification.
       An added feature of the SWHS is the detailed collection of other risk factor data from
trained interviewers. These data allow for risk estimates to be adjusted for potential confounding
variables. For the evaluations of reproductive health outcomes, this adjustment is critical given
there are various documented risk factors for the different outcomes that were examined.  For
some health outcomes, continued follow-up of the cohort is needed, given that several of the
Seveso studies suggest that those exposed at a very young age might be more susceptible to
subsequent adverse health effects.
       The findings of positive associations and dose-response relationships with serum-based
measures of TCDD suggest several noncancer health outcomes could be associated with TCDD
exposure. These health outcomes include neonatal thyroid function, sex ratio, diabetes, and
semen quality. Although findings have suggested an association between TCDD and age at
menopause, they were not statistically significant and no dose-response trend was observed.
Weak or nonstatistically significant associations have been noted for endometriosis and
menstrual cycle characteristics and do not support quantitative dose-response analyses.
       Associations between TCDD exposure and cardiovascular  disease have been noted in
some, but not all, of the occupational  cohorts, and also shortly after the accident among Seveso
residents. Findings from the cohort studies based on external comparisons using the SMR
statistic should be interpreted cautiously due to potential bias from the healthy worker effect.
Because the magnitude of the healthy worker bias is recognized to be larger for cardiovascular
diseases than for cancer outcomes, risk estimates in some occupational cohorts might be
underestimated for cardiovascular outcomes.  Information on cardiovascular risk factors
generally was not captured in these studies, and sensitivity analyses were generally designed to
examine risk estimates generated for cancer outcomes.

C.l.2.3.   Summary of Epidemiologic Noncancer Study Evaluations for Dose-Response
          Modeling
       All epidemiologic noncancer studies  summarized above were evaluated for suitability of
quantitative dose-response assessment using the TCDD-specific considerations and study
                                          C-155

-------
inclusion criteria. The results of this evaluation are summarized in a matrix style array (see
Table C-3).  The key epidemiologic noncancer studies suitable for further TCDD dose-response
assessment are presented in Table 2-2 in Section 2 of this document.
                                         C-156

-------
Table C-l. Summary of epidemiologic cancer studies (key characteristics)
Publication
Length of
follow-up
Latency period
Half-life for TCDD
Fraction of TEQs
accounted for by
TCDD
NIOSH Cohort
Fingerhutetal. (1991a)
Steenlandetal. (1999)
Steenland et al. (200 Ib)
Cheng et al. (2006)
Collins et al. (2009)
1942-1987
1942-1993
1942-1993
1942-1993
1942-2003
0, 20 years
0, 15 years
0, 15 years
0, 10, 15 years
None
N/A
N/A
8.7 years (Michalek et
al.. 1996)

8.7 years (Michalek et
al.. 1996). and CADM
(Aylward et al.. 2005a)
7.2 years (Flesch-Janvs
etal.. 1996)
N/A
N/A
TCDD accounted for all
occupational TEQ; 10%
of background
N/A
N/A
BASF Cohort
Thiessetal. (1982)
Zober et al. (1990)
Ott and Zober (1996a)
1953-1980
1953-1987
1953-1991
None
Years since first
exposure: 0-9,
10-19, and 20+
None
N/A
N/A
5.8 years
N/A
N/A
N/A
Hamburg Cohort
Manz et al. (1991)
Flesch-Janys et al.
(1995)

Flesch-Janys et al.
(1998)
Becheretal. (1998)
1952-1989
1952-1992
1952-1992
1952-1992
None, used
duration of
employment
(<20, >20 years)
None
None
0, 5, 10, 15 and
20 years
N/A
7.2 years Flesch-Janys
et al. (1994)
7.2 years Flesch-Janys
et al. (1996). also used
decay rates that were
function of age and fat
composition
7.2 years Flesch-Janys
et al. (1996) took into
account age and fat
composition
N/A
Mean TEQ without
TCDD was 155 ng/kg;
mean TEQ with TCDD
was 296.5 ng/kg
Mean concentration of
TCDD was 10 1.3 ng/kg;
for TEQ (without
TCDD) mean exposure
was 89.3 ng/kg
Not described
                                 C-157

-------
Table C-l. Summary of epidemiologic cancer studies (key characteristics)
(continued)
Publication
Length of
follow-up
Latency period
Half-life for TCDD
Fraction of TEQs
accounted for by
TCDD
Seveso Cohort
Bertazzi et al. (2001)
Warner et al. (2002)
Pesatori et al. (2003)
Baccarelli et al. (2006)
Consonni et al. (2008)
1976-1996 Periods
postexposure: 0,
[o-4, 5-9, 10-14,
15-19 years
1976-1998 Hone
1976-1996
1976-1998
1976-2001
Period
)ostexposure: 20
years
Period
)ostexposure: 22
years
Periods
)ostexposure: 0,
0-4, 5-9, 10-14,
15-19, 20-24
years
N/A
8 years (Pirkle et al..
1989)
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Chapaevsk Cohort
Revich et al. (2001)
Cross-
sectional
study
(1995-1998)
N/A
N/A
N/A
Ranch Hand Cohort
Akhtar et al. (2004)
Michalek and Pavuk
(2008)
1962-1999
1962-2004
^one
^one, but
stratified by
period of service
N/A
7.6 years
N/A
N/A
New Zealand Cohort
t'Mannetje et al. (2005)
McBride (2009b)
1969-2000
(herbicide
producers);
1973-2000
(herbicide
sprayers)
1969-2004
N/A
None
N/A
N/A
N/A
N/A
New Zealand Cohort (continued)
McBride et al. (2009b)
1969-2004
None
7 years
N/A
Dutch Cohort
Hooiveld et al. (1998)
1955-1991
3ostexposure
)eriods: 0-19
years, >19 years
7.1 years
N/A
                                C-158

-------
           Table C-2.  Epidemiologic cancer study selection considerations and criteria

Cancer
NIOSH Cohort
Fingerhut et al. (1991a)
all cancer sites, site-specific analyses
Steenland et al. (1999)
all cancer sites combined, site-specific
analyses
Steenland et al. (200 Ib)
all cancer sites combined
Cheng et al. (2006)
all cancer sites combined
Collins et al. (2009)
all cancer sites combined, site-specific
analyses
BASF Cohort
Thiess et al. (1982)
all cancer sites combined, site-specific
analyses
Methods
clear and
unbiased
Risk
estimates
not
susceptible
to biases
Association
between
TCDD and
adverse
health effect
Individual-
level
exposures
Study size
and follow-
up
adequate
Considerations

A/
A/
A/
A/
A/

A/

X
A/
A/
A/
A/

X

X
A/
A/
A/
A/

X

X
A/
A/
A/
A/

X

A/
A/
A/
A/
A/

X










Published
in peer-
reviewed
literature.
Exposure
primarily to
TCDD
Effective
exposure
estimable
Criteria

A/
A/
A/
A/
A/

A/

X
A/
A/
A/
A/

X

A/
A/
A/
A/
A/

X
Pass for
dose-
response
analyses?
Y/N

N
Na
Y
Y
Y

N
n
i
^O

-------
           Table C-2.  Epidemiologic cancer study selection considerations and criteria (continued)

Cancer
BASF Cohort (continued)
Zober et al. (1990)
all cancer sites combined, site-specific
analyses
Ott and Zober (1996a)
all cancer sites combined
Hamburg Cohort
Manz et al. (1991)
all cancer sites combines, site-specific
analyses
Flesch-Janys et al. (1995)
all cancer sites combined
Flesch-Janys et al. (1998)
all cancer sites combined, site-specific
analyses
Becher et al. (1998)
all cancer sites combined
Seveso Cohort
Bertazzi et al. (2001)
all cancer sites combined, site-specific
analyses
Warner et al. (2002) - SWHS
breast cancer incidence
Pesatori et al. (2003)
all cancer sites combined, site-specific
analyses
Baccarelli et al. (2006) - SWHS
site specific analysis
Methods
clear and
unbiased
Risk
estimates
not
susceptible
to biases
Association
between
TCDD and
adverse
health
effect,
Individual-
level
exposures
Study size
and follow-
up
adequate
Considerations

A/
A/

A/
A/
A/
A/

A/
A/
A/
A/

A/
A/

A/
A/
A/
A/

A/
A/
A/
A/

X
A/

A/
A/
A/
A/

A/
A/
X
X

X
A/

A/
A/
A/
A/

X
A/
X
A/

X
A/

A/
A/
A/
A/

A/
A/
A/
A/















Published
in peer-
reviewed
literature.
Exposure
primarily to
TCDD
Effective
exposure
estimable
Criteria

A/
A/

A/
A/
A/
A/

A/
A/
A/
A/

X
A/

X
A/
A/
A/

A/
A/
X
A/

X
A/

A/
X
A/
A/

X
A/
X
A/
Pass for
dose-
response
analyses?
Y/N

N
Y

N
N
Nb
Y

N
Y
N
Nc
n
i
O

-------
           Table C-2.  Epidemiologic cancer study selection considerations and criteria (continued)

Cancer
Consonni et al. (2008)
all cancer sites combined, site-specific
analyses
Chapaevsk Cohort
Revich et al. (2001)
all cancer sites combined, site-specific
analyses
Ranch Hands Cohort
Akhtar et al. (2004)
all cancer sites combined, site-specific
analyses
Michalek and Pavuk (2008)
all cancer sites combined
Dutch Cohort
Hooiveld et al. (1998)
all cancer sites combined, site-specific
analyses
New Zealand Cohort
t' Mannetje et al. (2005)
all cancer sites combined, site-specific
analyses
McBride et al. (2009b)
all cancer sites combined, site-specific
analyses
McBride et al. (2009a)
all cancer sites combined, site-specific
analyses
Methods
clear and
unbiased
Risk
estimates
not
susceptible
to biases
Association
between
TCDD and
adverse
health
effect,
Individual-
level
exposures
Study size
and follow-
up
adequate
Considerations
A/

X

A/
A/

A/

A/
A/
A/
A/

X

A/
A/

X

X
A/
X
A/

X

A/
A/

A/

A/
X
X
X

X

A/
A/

A/

A/
A/
A/
A/

A/

A/
A/

X

A/
A/
X














Published
in peer-
reviewed
literature.
Exposure
primarily to
TCDD
Effective
exposure
estimable
Criteria
A/

A/

A/
A/

A/

A/
A/
A/
A/

X

A/
A/

A/

X
A/
X
X

X

A/
A/

X

X
X
X
Pass for
dose-
response
analyses?
Y/N
N

N

Y
Y

N

N
N
N
n

-------
             Table C-2.  Epidemiologic cancer study selection considerations and criteria (continued)

     aThis study has been superseded and updated by Steenland et al. (2001b).
     bBecher et al. (1998) assessed this same cohort taking cancer latency into account, thereby superseding this study.
     °It is unknown whether the frequency of t(14;18)translocations in lymphocytes relates specifically to an increased risk of non-Hodgkin lymphoma. Given this
     lack of obvious adverse effect, dose-response analyses for this outcome were not conducted.

     A/ = Consideration/criterion satisfied; X = Consideration/criterion not satisfied.
o
 I

K>

-------
           Table C-3. Epidemiologic noncancer study selection considerations and criteria

Noncancer
NIOSH Cohort
Steenland et al. (1999)
mortality (noncancer) -ischemic heart
disease
Collins et al. (2009)
mortality (noncancer)
BASF Cohort
Ott and Zober (1996a)
mortality (noncancer)
Hamburg Cohort
Flesch-Janys et al. (1995)
mortality (noncancer)
Seveso Cohort-SWHS
Eskenazi et al. (2002b)
menstrual cycle characteristics
Eskenazi et al. (2002a)
endometriosis
Methods
clear and
unbiased
Risk
estimates not
susceptible to
biases
Association
between
TCDD and
adverse
health effect
Individual-
level
exposures
Study size
and follow-
up adequate
Considerations

A/
A/

A/

A/

A/
A/

X
A/

A/

A/

A/
A/

A/
X

X

A/

A/
X

A/
A/

A/

A/

A/
A/

A/
A/

A/

A/

A/
X












Published
in peer-
reviewed
literature
Exposure
primarily to
TCDD
Effective
exposure
estimable
Criteria

A/
A/

A/

A/

A/
A/

A/
A/

A/

A/

A/
A/

X
X

X

X

A/
X
Pass for
dose-
response
analyses?
Y/N

N
N

N

N

Y
N
n

-------
           Table C-3. Epidemiologic noncancer study selection considerations and criteria (continued)

Noncancer
Seveso Cohort-SWHS (continued)
Eskenazi et al. (2003)
birth outcomes
Warner et al. (2004)
age at menarche
Eskenazi et al. (2005)
age at menopause
Warner et al. (2007)
ovarian function
Eskenazi et al. (2007)
uterine leiomyoma
Seveso Cohort-Other Studies
Bertazzi et al. (2001)
mortality (noncancer)
Consonni et al. (2008)
mortality (noncancer)
Seveso Cohort-Other Studies
(continued)
Mocarelli et al. (2000)
sex ratio
Baccarelli et al. (2004: 2002)
immunological effects
Landi et al. (2003)
gene expression
Alaluusua et al. (2004)
developmental dental defects
Baccarelli et al. (2005)
chloracne
Baccarelli et al. (2008)
neonatal thyroid function
Methods
clear and
unbiased
Risk
estimates not
susceptible to
biases
Association
between
TCDD and
adverse
health effect
Individual-
level
exposures
Study size
and follow-
up
adequate
Considerations

X
A/
A/
A/
A/

A/
A/

A/
A/
A/
A/
A/
A/

X
A/
A/
A/
A/

A/
A/

A/
A/
A/
A/
A/
A/

X
X
X
X
A/

X
X

A/
X
X
A/
A/
A/

A/
A/
A/
A/
A/

X
X

A/
A/
A/
A/
A/
A/

A/
A/
A/
A/
A/

A/
A/

A/
A/
X
A/
A/
A/


















Published
in peer-
reviewed
literature
Exposure
primarily to
TCDD
Effective
exposure
estimable
Criteria

A/
A/
A/
A/
A/

A/
A/

A/
A/
A/
A/
A/
A/

A/
A/
A/
A/
A/

A/
A/

A/
A/
A/
A/
A/
A/

X
A/
X
X
X

X
X

X
X
X
A/
A/
A/
Pass for
dose-
response
analyses?
Y/N

N
Na
N
N
N

N
N

N
N
N
Y
Nb
Y
n

-------
           Table C-3. Epidemiologic noncancer study selection considerations and criteria (continued)

Noncancer
Mocarelli et al. (2008)
semen quality
Chapaevsk Study
Revich et al. (2001)
mortality (noncancer) and
reproductive health
Ranch Hands Cohort
Henriksen et al. (1997)
diabetes
Longnecker and Michalek (2000)
diabetes
Michalek et al. (200 la)
hematological effects
Michalek et al. (200 Ib)
hepatic abnormalities
Michalek et al. (200 Ic)
peripheral neuropathy
Pavuketal. (2003)
thyroid function and disorders
Michalek and Pavuk (2008)
diabetes
Ufa Cohort
Ryan et al. (2002)
sex ratio
Vietnam Veterans Cohort
Kang et al. (2001)
long-term health consequences
New Zealand Cohort
McBride et al. (2009a)
mortality (noncancer)
Methods
clear and
unbiased
Risk
estimates not
susceptible to
biases
Association
between
TCDD and
adverse
health effect
Individual-
level
exposures
Study size
and follow-
up
adequate
Considerations

A/

X

A/
A/
A/
A/
A/
A/
A/

X

X

A/

A/

X

X
X
X
X
X
A/
A/

X

X

X

A/

X

A/
A/
X
A/
A/
X
A/

X

X

X

A/

X

A/
X
A/
A/
A/
A/
A/

X

A/

A/

A/

A/

A/
A/
A/
A/
X
X
A/

A/

A/

A/




















Published
in peer-
reviewed
literature
Exposure
primarily to
TCDD
Effective
exposure
estimable
Criteria

A/

A/

A/
A/
A/
A/
A/
A/
A/

X

A/

A/

A/

X

A/
A/
A/
A/
A/
A/
A/

X

X

X

A/

X

X
X
X
X
X
X
X

X

X

X
Pass for
dose-
response
analyses?
Y/N

Y

N

N
N
N
N
N
N
N

N

N

N
n

-------
             Table C-3.  Epidemiologic noncancer study selection considerations and criteria (continued)


Noncancer
McBride et al. (2009b)
mortality (noncancer)

Methods
clear and
unbiased

Risk
estimates not
susceptible to
biases
Association
between
TCDD and
adverse
health effect

Individual-
level
exposures

Study size
and follow-
up
adequate
Considerations

A/

A/

X

A/

X






Published
in peer-
reviewed
literature

Exposure
primarily to
TCDD

Effective
exposure
estimable
Criteria

A/

A/

X

Pass for
dose-
response
analyses?
Y/N

N
o
     aEPA cannot assess the biological significance of this finding and cannot establish a LOAEL for this effect.
     bChloracne is considered to be an outcome associated with high TCDD exposures; thus this study was not considered further in RfD derivation.
     A/ = Consideration/criterion satisfied. X = Consideration/criterion not satisfied.
Oi
Oi

-------
C.2.   EVALUATION TABLES FOR CANCER STUDIES
C.2.1. NIOSH Cohort Studies

      Table C-4. Fingerhut et al. (1991a)—All cancer sites, site-specific analysis
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. The data sources to ascertain vital status and cause of death information
were the Social Security death files, the National Death Index, and the Internal Revenue Service.
Vital status could be determined for 98% of the cohort.
Risk estimates are not susceptible to important biases.
Consideration not satisfied. While the authors provide compelling arguments that suggest risks are
not unduly biased by lack of cigarette smoking data, they acknowledge potential biases that could
exist for other occupational exposure (e.g., asbestos) for which data were lacking.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. There was not a statistically significant linear trend of increasing
mortality with increased duration of exposure.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration not satisfied. This study used duration of exposure, at an individual level, as a
surrogate measure of TCDD. Duration of exposure determined by number of years workers were
involved in processes involving TCDD contamination. Exposure was determined by reviewing, at
each plant, operating conditions, job duties, records of TCDD levels in industrial hygiene samples,
intermediate reactants, products, and wastes. Exposure assessment was limited and the uncertainty
related to exposure measures not fully addressed.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. This is the largest of the occupational cohorts that has been exposed to
TCDD. The cohort consisted of 5, 172 workers and a total of 265 cancer deaths. Site-specific
mortality analyses, including soft tissue sarcoma (n = 4), was limited by small numbers.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. New England Journal of Medicine, 1991; 324:212-218. Authors address the
possibility of bias from lack of control for potential confounders such as smoking and other
occupational exposures. They address limitations of using death certificates for identifying certain
causes of deaths, and limitations of using duration of employment as an exposure metric.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria not satisfied. Since this study used duration of exposure as the exposure metric,
dose-response relationships cannot be quantified.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. Models incorporated period of latency, and a surrogate measure of cumulative
TCDD exposure was modeled. The follow-up interval was sufficiently long (1942-1987).

Overall, quantitative exposure data are lacking on an individual-level basis. Further dose-response
analysis should consider updated data for this cohort that includes serum-based measures of
TCDD, in addition to an extension of the follow-up period. Given these limitations, this study is
not further evaluated for TCDD dose-response assessment.
                                      C-167

-------
Table C-5. Steenland et al. (1999)—All cancer sites combined, site-specific
analysis
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
1 Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. The study evaluated mortality from all cancer sites (combined). As
described in the paper, the sources of vital status and cause of death information were received
from the Social Security death files, the National Death Index, and the Internal Revenue Service.
Vital status was known for 99.4% of the cohort members, cause of death information is available
for 98% of the decedents.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Occupational exposure to asbestos and 4-aminobiphenyl contributed to
some excess cancer, but no evidence of confounding for the relationship between TCDD and all
cancer mortality was detected following removal of workers who died of bladder cancer. No
information is available for cigarette smoking, although dose-response patterns were stronger for
nonsmoking related cancers. This finding suggests that smoking is not responsible for excess
cancer risk that was observed in the cohort.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. When a 15-year lag interval was incorporated into the exposure metric a
statistically significant dose-response pattern was observed for all cancer sites combined with both
a continuous measure of TCDD (p = 0.05) as well as one that was log-transformed (p < 0.001).
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. The study conducted detailed sensitivity analyses and evaluated different
assumptions regarding latency, log-transformed TCDD exposures, and half-life values for TCDD.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. This is the largest of the occupational cohorts with exposures to TCDD.
The cohort consisted of 5,132 male workers and a total of 377 cancer deaths. This permits
characterization of risk for all cancer sites (combined).

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Journal of the National Cancer Institute, 1999; 91(9):779-786. The authors
discussed the potential for bias from smoking, and other occupational exposures for which data foi
both were lacking at an individual basis.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Exposure scores assigned on an individual level using a job-exposure matrix
(JEM). The job-exposure matrix was based on estimated factor of contact with TCDD in each job,
level of TCCD contamination of materials at each plant over time, and proportion of day worker
could be in contact with materials. These factors were multiplied together to derive a daily
exposure score, which was accumulated over the working history of each worker to obtain a
cumulative measure of TCDD.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. The follow-up of the cohort extended from 1942 until the end of 1993. Greater
than 25 years of follow-up have accrued in cohort allowing for latency to be examined. Different
assumptions on the half -life of TCDD were evaluated and produced similar results. Latency
intervals were incorporated, with strongest associations noted with an interval of 15 years.

This study meets the criteria and considerations noted above but has been superseded and updated
by Steenland et al.QOOlb). Therefore, this study was not considered for further dose-response
analyses.
                                  C-168

-------
Table C-6.  Steenland et al. (2001b)—All cancer sites combined
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. The study evaluated mortality from all cancer sites (combined). As
described by Steenland et al. (1999) the sources of vital status and cause of death information
were received from the Social Security death files, the National Death Index, and the Internal
Revenue Service. Vital status was known for 99.4% of the cohort members, cause of death
nformation is available for 98% of the decedents.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Occupational exposure to asbestos and 4-aminobiphenyl contributed to
some excess cancer, but no evidence of confounding for the relationship between TCDD and all
cancer mortality was detected following removal of workers who died of bladder cancer. No
nformation is available for cigarette smoking, although dose-response patterns were similar
)etween smoking and nonsmoking related cancers. There is no available information in the study
to determine how representative the 199 workers were of the overall workers in that plant, or the
potential for this to result in exposure misclassification.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. Increased risk estimates were observed in the higher cumulative exposure
categories. The dose-response curve was not linear at higher doses.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied.
Exposure metrics considered included cumulative TCDD, loglOTCDD, average exposure, and a
cubic spline model was also evaluated. Exposure response relationships were also evaluated using
TEQs. Exposure scores were assigned on an individual level using a job-exposure matrix. The
ob-exposure matrix was based on estimated factor of contact with TCDD in each job, level of
TCCD contamination of materials at each plant over time, and proportion of day worker could be
n contact with materials. Serum levels were measured in 199 workers at one of 8 plants in 1998.
Different estimate of the half -life of TCDD were used, and similar results were produced. The
)aper presented a range in risk estimates thereby conveying the range of uncertainties in risk
estimates derived using different measures of exposure.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. This is the largest of the occupational cohorts with exposures to TCDD.
The cohort consisted of 3,538 male workers and a total of 256 cancer deaths.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Am J Epidemiol, 2001, 154(5):451-458. However, additional details to assess
uncertainties associated with characterizing serum data in a subset of workers to remainder of
cohort are lacking.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. The metrics considered included cumulative TCDD, loglOTCDD, average
exposure, and a cubic spline model was also evaluated. Exposure response relationships were also
evaluated using TEQs. Serum lipid TCDD measurements from 170 workers whose TCDD levels
were greater than 10 ppt (the upper ranges of a background level) were used along with JEM
nformation, work histories, and a pharmacokinetic elimination model to estimate dose rates per
unit exposure score. In this regression model, the estimated TCDD level at the time of last
exposure was modeled as a function of exposure scores. The coefficient relating serum levels and
exposure scores was then used to estimate serum TCDD levels over time from occupational
exposure (minus the background level) for all 3,538 workers. Time-specific serum levels were
then integrated over time to derive a cumulative serum lipid concentration due to occupational
exposure for each worker.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. Greater than 25 years of follow-up have accrued in cohort allowing for latency
to be examined. Different assumptions on the half-life of TCDD were evaluated producing
similar results.
                                C-169

-------
Conclusion
Overall, criteria have been satisfied.  This study was modeled in the 2003 Reassessment and is
considered for further dose-response evaluations herein.	
       Table C-7.  Cheng et al. (2006)—All cancer sites combined
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
1 Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. The study evaluated cancer mortality. The vital status and the
information regarding the cause of death were extracted from the Social Security death files, the
National Death Index, and the Internal Revenue Service (Steenland et al.. 1999). Vital status was
known for 99.4% of the cohort members, while cause of death information is available for 98% of
the decedents.
Risk estimates are not susceptible to important biases.
Consideration satisfied. This is the same data set used in the Steenland et al. (200 Ib) paper.
Occupational exposure to asbestos and 4-aminobiphenyl contributed to some excess cancer, but no
evidence of confounding for the relationship between TCDD and all cancer mortality was detected
following removal of workers who died of bladder cancer. No information is available for
cigarette smoking, although dose-response patterns were similar between smoking and
nonsmoking related cancers.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. Slope coefficients are available for all cancers combined under a varying
set of assumptions. Little evidence of an association was found when lag interval was not taken
into account. Associations strengthened with incorporation of a 10 to 15 year lag interval. Dose
response was nonlinear at higher exposures, suggesting a nonlinear relationship or increased
exposure misclassification at higher levels.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Compared to the 1st order models, the CADM provided a better fit for the
serum sampling data. CADM exposure estimates are higher than those based on an age only,
constant 8.7-year half-life model. As discussed by Aylward et al. (2005b), model exposure
estimates are influenced not only by choice of elimination model, but also by choices in regression
procedure (e.g., log transformation, use of intercept, and incorporation of background dose term).
Other limitations or uncertainties in exposure assessment include the following
Job-exposure matrix based on limited sampling data, and subjective judgment on contact times
and factors
Inability to take into account interindividual variability in TCDD elimination kinetics
Dose-rate regressions are based on a small sample of the cohort with serum measures; therefore,
regression results may not be representative of remainder of the cohort.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. Largest cohort of TCDD exposed workers. The risk estimates are based
on a total of 256 cancer deaths.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Risk Analysis, 2006; 4:1,059-1,071. Additional details to assess uncertainties
associated with characterizing serum data can be found in Aylward et al. (2005b); Risk Anal.
25(4):945-956.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Cumulative serum lipid concentrations were estimated for each worker. No
other DLCs were assessed in this analysis.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. Concentration and age-dependence of TCDD elimination and
two compartments (hepatic and adipose tissue) were taken into account when estimating TCDD
exposures. Nearly 50 years of follow-up were available permitting an evaluation of latency.

This study met the main criteria and considerations. The study is considered for further
dose-response analyses.
                                             C-170

-------
Table C-8. Collins et al. (2009)—All cancer sites combined, site-specific
analysis
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Vital status complete for all but two workers.
Risk estimates are not susceptible to important biases.
Consideration satisfied. No information collected on smoking status, but no excess in lung cancer
or nonmalignant respiratory diseases noted. Analyses took into account potential for exposure to
pentachlorophenol.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. No dose-response pattern was observed with all cancer sites combined,
lowever, a dose-response pattern was observed with soft tissue sarcoma. The study found no
association between TCDD and death from most types of cancer.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. The authors used serum from 280 former TCP workers to estimate
listorical exposure levels of TCDD, furans, and polychlorinated biphenyls (PCBs) for all 1,615
workers. Exposure assessment included detailed work history, industrial hygiene monitoring, and
the presence of chloracne cases among groups of workers. This data was integrated into a 1-
compartment, first-order pharmacokinetic to determine the average TCDD dose associated with
obs in each group, after accounting for the presence of background exposures estimated from the
residual serum TCDD concentration in the sampled individuals. The authors did not evaluate
departures from linearity, or examine skewness at higher exposures. Exposure levels were not
provided.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. Largest study of workers employed in one center, and a total of
177 deaths from cancer were observed. Limited precision in the relative risk estimate was noted
for soft tissue sarcoma and TCDD exposures.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Published in Am J Epidemiol, 2009, 170(4):501-506. The authors discuss
imitations of using death certificates for identifying deaths from soft tissue sarcoma for which a
)ositive association was noted, assumptions in exposure characterization, and effects of cigarette
smoking.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. This study has the largest number of serum samples obtained from a specific
plant.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. Although specific analyses of latency were not reported, this cohort had a
sufficient length of follow-up for cancer mortality outcomes.

The authors found a statistically significant dose-response trend for soft tissue sarcoma mortality
and TCDD exposures. The all-tumor results are not amenable to dose-response analysis because
they found no effect. Therefore, this study is considered for quantitative dose-response analysis
for the soft tissue sarcoma mortality results, only.
                                  C-171

-------
C.2.2. BASF Cohort Studies
      Table C-9. Zober et al. (1990)—All cancer sites combined, site-specific
      analysis
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
1 Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. A large component of the cohort (94 out of 247 workers) was assembled
sy actively seeking out workers who were alive in 1986 through the "Dioxin Investigation
Programme." As a result, it is likely a number of deaths were missed due to the recruitment of
survivors. This underascertainment is supported by much lower all cancer SMR one component of
the cohort (SMR = 0.48, 95% CI = 0. 13-1.23) relative to the general population.
Risk estimates are not susceptible to important biases.
Consideration satisfied. See above discussion of underascertainment in mortality for some of the
cohort members. Although it is likely that other coexposures occurred (e.g., among firefighters),
confounding could only occur if these coexposures were associated with both the endpoint and
exposure (TCDD) being considered.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. Workers were not categorized on the basis of their exposure, but
rather their mortality experience compared to control cohort and the general population. The
design of the study does not allow for dose response to be examined.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration not satisfied. Although years since first exposure was examined, exposure
assessment was based on working in various occupational cohorts. Since there was no quantitative
assignment of TCDD exposures, the associated uncertainties could not be evaluated.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration not satisfied. There were only 23 cancer deaths in the entire cohort. As such, this
study lacked adequate statistical power to detect cancer mortality differences that were moderate in
magnitude.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Int Arch Occup Environ Health, 1990,62:139-157. The authors address issues
related to the healthy worker effect, multiple comparisons, smoking, and small size of the cohort.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria not satisfied. Risks were derived by comparing mortality rates of the three cohort subsets
relative to a control cohort and the general population by time since first exposure categories.
Workers were not assigned exposures. There were no quantitative estimates of TCDD exposure.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. While the study was able to indirectly look at variations in risk estimates
related to latency by using time since exposure, there were no quantitative estimates of TCDD
exposure.

This study is not suitable for dose-response analysis, as it failed the inclusion criteria. Most
notably, the lack of exposure data does not permit the use of these data for a dose-response
analysis.
      Table C-10. Ott and Zober (1996a)—All cancer sites combined
1. Consideration
Response
2. Consideration
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Mortality ascertainment appeared to be fairly complete. The
ascertainment of cancer incidence is more difficult to judge as geographical area not covered by a
cancer registry.
Risk estimates are not susceptible to important biases.
                                      C-172

-------
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Consideration satisfied. Information was collected on smoking status, body mass index (BMI),
and other occupational exposures, however a large portion of the cohort was firefighters who may
have been exposed to other occupational carcinogens. However, the recruitment of survivors may
results in under-ascertainment of mortality.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. Increased cancer incidence was observed in the highest TCDD cumulative
exposure category. Risks were most pronounced when a period of 20 years since first exposure
was incorporated into the model.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Cumulative measure of TCDD expressed was derived from serum
measures. Exposure was also estimated by chloracne status of the cohort members. The authors
have not addressed the potential implication of deriving TCDD exposure estimates for the whole
cohort using sera data that were available for only about half of the cohort.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. For all cancer sites combined, there were 3 1 deaths. It is the smallest of
the occupational cohorts, but the deaths can be grouped into quartiles to allow for evaluation of
dose-response relationships.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Occupational and Environmental Medicine, 1996,53:606-612. A large
component of the cohort (94 out of 247 workers) was assembled by actively seeking out workers
who were alive in 1986 through the "Dioxin Investigation Programme." As a result, it is likely a
number of deaths were missed due to the recruitment of survivors. This underascertainment is
supported by much lower all cancer SMR one component of the cohort (SMR = 0.48, 95% CI =
0.13-1.23) relative to the general population (Zober et al, 1990).
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Serum samples, taken in 1989, were available for 138 surviving workers out of
254 and allowed for cumulative TCDD levels to be estimated using regression techniques in the
remainder of the cohort.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. Exposure assignment took into the affect that body mass index had on TCDD
half-lives. TCDD levels estimates through back-extrapolation of serum levels based on half-life
estimates obtained from previous studies. Latency was considered with stronger association
observed in external comparisons incorporating a latency of 20 years. The follow-up of the cohort
was lengthy (>50 years).

Given a part of the cohort was based solely on survivors in the in the mid-1980s, the SMR statistic
derived from this study underestimates excess mortality relative to the general population. The
cohort also includes some firefighters who are recognized to be exposed to other carcinogenic
agents — these exposures may be confounding the associations that were reported. However,
exposure to TCDD was quantified and the effective dose and oral exposure estimable. Overall,
criteria have been satisfied. This study was modeled in the 2003 Reassessment and is considered
for further dose-response evaluations herein.
C.2.3. The Hamburg Cohort
      Table C-ll. Manz et al. (1991)—All cancer sites combined, site-specific
      analyses
1. Consideration
Response
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Deaths were identified through medical records of the cohort members. A
review of death certificates of the identified cancer deaths found a high degree of concordance
(51/54). One of the 136 noncancer death certificates examined indicated an "occult" neoplasm.
                                       C-173

-------
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Risk estimates are not susceptible to important biases.
Consideration satisfied. Smoking data were similar between exposed and nonexposed cohort
)ased on independent samples. Occupational exposures for which individual data are lacking are
unlikely to explain dose response with TCDD. The potential impacts of any exposure
misclassification is hard to gauge, but the authors reported that some misclassification was likely
given that 5 of the 37 workers classified in the high exposure group had adipose levels lower than
)ackground (20 ng/kg).
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. Dose-response patterns across three levels of exposure observed among
those who started work before 1954, and among those who worked for 20 years or longer. Dose-
response patterns not evident across whole cohort, among those with less than 20 years of
smployment, or among those who started after 1954.
Exposure assessment methodology is clear and characterizes individual-level exposures..
Consideration satisfied. Categorical exposures were based on TCDD concentrations in precursor
materials, products, waste, and soil from the plant grounds, measured after the plant closed in
1984. Exposure uncertainty examined using a separate group of 48 workers who provided adipose
tissue samples. Other surrogate measures of exposure were considered in this study, including
duration of exposure and year of first employment.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. For all cancer sites combined, there were 65 cancer deaths for the
comparison to the comparison cohort of gas workers. The study is underpowered to look at site-
specific cancers.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Lancet 1991, 338:959-964. The authors discussed the potential for
misclassification from the use of death certificates, the healthy worker effect and the related use of
a comparison cohort of gas supply workers, other occupational exposures present at the plant,
potential impact and the lack of smoking data.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria not satisfied. Exposure consisted of a large DLC component that was not quantified.
Given crude TCDD exposure categorization data, no quantitative exposure metric was derived.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. Exposure metrics were constructed that took into account duration of exposure,
and periods when exposure was highest. However, exposure estimates did not consider lagged
sxposure.

This study is not amenable to further TCDD dose-response analysis and is not considered further
lere because it consisted of a large DLC component that was quantified and no quantitative
sxposure metric was derived. The dose-response patterns of risks observed across the three
sxposure groups provide compelling support for an association between TCDD and cancer
mortality, particularly, given the associations observed when analyses restricted to those who were
lired when TCDD exposures were known to be much higher, and among those who worked for at
east 20 years. Subsequent studies improved the exposure assessment through the use of serum
measures.
Table C-12. Flesch-Janys et al. (1995); Flesch-Janys et al. (1996) erratum-
All cancer sites combined
1. Consideration
Response
2. Consideration
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Medical records used to identify deaths over the period 1952-1992.
Risk estimates are not susceptible to important biases.
                                 C-174

-------
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Consideration satisfied. Similarity in smoking rates between control cohort and the exposed
workers was similar based on independent surveys. Occupational exposures to benzene, and
dimethyl sulfate were unlikely to bias dose-response pattern observed as these exposures occurred
in production departments with low-medium levels of exposure.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. Dose-response relationship observed across 6 exposure categories, with
the cohort of gas supply workers used as the referent.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Exposure assessment methodology is clear and adequately characterizes
individual-level exposures. The limitations and uncertainties in the exposure assessment are
considered.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. For all cancer sites combined, there were 124 deaths in the exposed
cohort, and 283 in the cohort of gas supply workers. No site-specific cancers were examined in
this paper.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Am J Epidemiol, 1995, 1442:1165-1175. The authors discuss the potential role
of other occupational exposures (i.e., dimethyl sulfate, solvents, and benzene), smoking, and
suitability of the comparison cohort of gas supply workers.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Serum and adipose tissues were used to estimate TCDD exposure in
190 workers. A one-compartment first-order kinetic model was used to estimate exposure at end
of exposure for these workers. Regression methods were then used to estimates TCDD exposures
for all workers.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. Exposure was based on half-life estimates from individuals with repeated
serum measures. Other dioxin-like compounds were considered with the TOTTEQ of
polychlorinated dibenzo-p-dioxins and furans exposure metric. No consideration, however, was
given to latency or lagged exposures.

The exposure data used within this study are well-suited to a dose-response analysis given the
associations observed, the characterization of exposure using serum, and quality of ascertainment
of cancer outcomes. However, subsequent methods have been applied to the cohort to derive
different exposures to TCDD using area under the curve approaches, which updates the analysis
herein. Therefore, subsequent studies (i.e., Becher et al., 1998) will supersede this evaluation.
Table C-13. Flesch-Janys et al. (1998)
specific analysis
-All cancer sites combined, site-
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Mortality follow-up was extended until the end of 1992, an increase in
3 years from previous analyses of the cohort.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Exposure was well characterized using sera data. While serum samples
provided only from a subsample of surviving workers, these levels were consistent with expected
levels in different production departments. The authors examined other potential occupational
coexposures (e.g., p-hexachlorocyclohexane) and indirectly examined the potential effect of
smoking on the associations that were detected.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. A dose-response relationship across quartiles of TCDD was observed
with cancer mortality based on the SMR statistic (SMRs = 1.24, 1.34, 1.34, 1.73), and a linear test
for trend was statistically significant (p = 0.01).
                                  C-175

-------
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. The exposure measure was an integrated TCDD concentration over time
estimate that back-calculated TCDD exposures to the end of the employment. Categorical and
continuous TCDD exposures were examined in relation to the health outcome. These efforts
improve the exposure assessment of earlier studies.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. For all cancer sites combined, there were 124 cancer deaths.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Environ Health Perspect, 1998, 106(2):655-662. The authors address
uncertainties in the estimation of exposure, describe the potential for confounding from (3-2,4,5-T,
hexachlorocyclohexane, and cigarette smoking. In fact, they showed that blood levels of TCDD
were not associated with smoking in a subsample suggesting little bias from lack of smoking data.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Serum samples, taken from 190 workers were used to derive TCDD levels for
the entire cohort. Methods used to estimate exposure took into account elimination of TCDD
during employment periods when exposure took place, and the methods of the area under the curve
was used as it takes into account variations in concentration over time, and reflects cumulative
exposure.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. Exposure estimated based on half-lives observed in individuals with repeated
samples. Area under the curve approach was used which is an improvement from past
characterizations of exposure in this cohort.

The study provides data suitable for dose-response modeling. Derivation of exposure was done
using current understanding of elimination of TCDD. Estimates of risks were derived from
external comparisons to the general population that are unlikely to be biased by healthy worker
effect, but risks generated using internal cohort comparisons would be preferable. Becher et al.,
(1998) assessed this same data taking cancer latency into account, therefore Flesch-Janys et al.,
( 1998) will not be further considered for dose-response modeling.
Table C-14.  Becher et al. (1998)—All cancer sites combined
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Medical records used to identify deaths over the period 1952-1992. The
follow-up interval was lengthy.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Risks adjusted for exposures to TEQ, p-hexachlorobenzene, and
employment characteristics. Smoking was shown to be similar to the comparison cohort of gas
workers.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. A variety of exposure measures for both TCDD and TEQs found positive
associations with cancer mortality.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. The exposure measure was an integrated TCDD concentration over time
estimate that back-calculated TCDD exposures to the end of the employment. Categorical and
continuous TCDD exposures were examined in relation to the health outcome. Different models
explored the shape of the dose-response curve. These efforts improve the exposure assessment of
earlier studies.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. For all cancer sites combined, there were 124 cancer deaths.

Study is published in the peer-reviewed scientific literature.
                                C-176

-------
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Criteria satisfied. Environ Health Perspect, 1998, 106(2):663-670. The authors discuss
uncertainties associated with their use of exposure metrics, inability to evaluate effects for PCDD/
PCDF other than dioxin due to high correlations with (3-HCH, and inability to characterize risks
associated with exposures in children.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. The authors derived a measure of cumulative dose as a time-dependent variable
("area under curve") using serum measures available in a sample of 275 workers.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. TCDD levels estimates through back-extrapolation of serum levels based on
half-life estimates obtained from previous studies. Latency was considered, and a variety of
exposure metrics including nonlinear relationships were evaluated.

In this paper, a variety of exposure metrics were found to be positively associated with cancer
mortality. The additional lifetime risk of cancer corresponded to a daily intake of Ipg ranged
setween .01 and 0.001. This study was modeled in the 2003 Reassessment and is considered for
further dose-response evaluations herein.
C.2.4.  The Seveso Cohort Studies
       Table C-15. Bertazzi et al. (2001)—All cancer sites combined, site-specific
       analyses
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Mortality appears to be well captured from the vital statistics registries in
the region (99% complete). Vital status was ascertained using similar methods for both the
exposed and reference populations. Both cancer and noncancer mortality outcomes were
evaluated. Ideally, would have evaluated incident rather than decedent outcomes for cancer.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Individual-level data on potential confounders (i.e., age, calendar period,
and gender) were adjusted for. Information from other independent surveys suggests similarity
between smoking behaviors across the regions. Comparison of cancer mortality rates before the
time of the accident between the regions also revealed no differences.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied (for all cancers combined). No statistically significant excesses noted in
Zone A, or Zone B relative to reference area. Evidence of an exposure-response relationship was
detected for lymphatic and hematopoietic tissues by number of years since first exposure.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration not satisfied. Subjects were assigned to one of the zones (A, B, R, or reference)
based on official residence on the day of the accident or at entry into the area. Exposure
misclassification is likely and lack of individual-level data precludes an examination of this source
of error.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. In total, 27, and 222, cancer deaths were found among residents of Zones
A, and B, respectively. This allowed examined of gender-specific effects.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Am JEpidemiol, 2001 Jun 1; 153(11):1031-1044. Authors discuss
completeness of mortality ascertainment, diagnostic accuracy of death certificates particularly with
respect to diabetes, limited available of blood dioxin measures that did not permit estimation of
TCDD dose on an individual-level basis.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
                                        C-177

-------
Response
3. Criteria
Response

Conclusion
Criteria not satisfied. Individual-level exposure data are unavailable. Exposure based on place of
residence at time of the explosion. Soil sampling performed indicated considerable variability in
TCDD levels within each region. In addition, place of residency at time of explosion does not
ensure individuals were at their home around the time of the accident.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. An ecological measure of exposure (region of residency at time of accident)
was used to categorize individuals according to their possible exposure. Latencies were
considered. While such an approach has value for identifying wherever excesses occurred among
highly exposed populations, it is not precise enough to conduct a quantitative dose-response
analysis.

The lack of individual -level exposure data precludes quantitative dose-response modeling using
these data.
Table C-16. Pesatori et al. (2003)
analyses
-All cancer sites combined, site-specific
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Mortality was ascertained from 1977-1996, and, as reported in other
related manuscripts, appears to be well captured from the vital statistics registries in the region
(99% complete). Cancer incidence data was available from 1977-1991.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Individual-level data on potential confounders (i.e., age, calendar period,
and gender) were adjusted for.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. Although risk of all cancer mortality was not associated with zone of
residence, increased risk of cancer incidence was observed in Zone A. Among men, excess
lymphatic and hematopoietic cancer incidence was observed in Zone A (primarily to non-Hodgkin
lymphoma). Soft tissues sarcoma cancer incidence was also associated with residence in Zone R
among males, but not the more highly exposed zones (A and B). Among females living in Zones
A and B, higher rates were observed for multiple myeloma (RR = 4.9, 95% CI = 1.5-16.1), cancer
of the vagina (RR = 5.5, 95% CI = 1.3-23.8), and cancer of the biliary tract (RR = 3.0,
95%CI= 1.1-8.2).
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration not satisfied. Subjects were assigned to one of the zones (A, B, R, or reference)
based on official residence on the day of the accident or at entry into the area. Exposure
misclassification is likely and lack of individual-level data precludes an examination of this source
of error.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied for some endpoints, although several of the cancer specific mortality results
among women were based on very small number of deaths (i.e., <5).

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Occup Environ Med, 1998; 55:126-131. Authors discuss limitations such as
residency-based exposure assignment, absence of smoking, differential and death certification in
exposed versus nonexposed areas.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria not satisfied. Individual-level exposure data are unavailable. Exposure based on place of
residence at time of the explosion. Soil sampling performed indicated considerable variability in
TCDD levels within each region. In addition, place of residency at time of explosion does not
ensure individuals were at their home around the time of the accident.
Effective exposure is estimable latency and window(s) of exposure are examined.
                                  C-178

-------
Response

Conclusion
Criteria not satisfied. An ecological measure of exposure (region of residency at time of accident)
was used to categorize individuals according to their possible exposure. Latencies were
considered. While such an approach has value for identifying wherever excesses occurred among
highly exposed populations, it is not precise enough to conduct a quantitative dose-response
analysis.

No dose-response patterns evident in the study, and the study lacked quantifiable measures of
TCDD at an individual-level basis. The data are not well suited for dose-response analysis.
Table C-17. Consonni et al. (2008)
analyses
-All cancer sites combined, site-specific
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Mortality appears to be well captured from the vital statistics registries in
the region (99% complete). Both cancer and noncancer mortality evaluated, although diagnostic
accuracy of death certificates is likely low. Ideally, would have evaluated incident rather than
decedent outcomes for cancer.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Individual-level data on potential confounders (i.e., age, calendar period,
and gender) were adjusted for. Comparison of cancer mortality rates before the time of the
accident between the regions also revealed no differences. Information from other independent
surveys suggests similarity between smoking behaviors across the regions.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied for some outcomes. For all cancer sites combined, no evidence of dose
response was observed relative to general population across Zones A, B and R. Only statistically
significant excess found in Zone A was for chronic rheumatic disease but based on only three
deaths. Higher cancer excesses were found in Zone A after a latency period was incorporated;
however, no dose-response relationship observed with this latency period. Evidence of an
exposure-response relationship was detected for lymphatic and hematopoietic tissues by zone of
residence.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration not satisfied. Subjects were assigned to one of the zones (A, B, R, or reference)
based on official residence on the day of the accident or at entry into the area. Exposure
misclassification is likely and lack of individual-level data precludes an examination of this source
of error.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. In total, 42, 244, and 1,848 cancer deaths were found among residents of
Zones A, B, and R respectively.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Am J Epidemiol, 2008, 167:847-858. Authors discuss potential for selection
bias, limitation of residential based measure of exposure, similarities of mortality ascertainment in
exposed and referent populations, and multiple testing.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria not satisfied. Individual-level exposure data are unavailable. Exposure based on place of
residence at time of the explosion. Soil sampling performed indicated considerable variability in
TCDD levels within each region. In addition, place of residency at time of explosion does not
ensure individuals were at their home around the time of the accident.
Effective exposure is estimable latency and window(s) of exposure are examined.
                                 C-179

-------
Response

Conclusion
Criteria not satisfied. An ecological measure of exposure (region of residency at time of accident)
was used to categorize individuals according to their possible exposure. Latencies were
considered. While such an approach has value for identifying wherever excesses occurred among
highly exposed populations, it is not precise enough to conduct a quantitative dose-response
analysis.

The lack of individual -level exposure data precludes quantitative dose-response modeling using
these data.
Table C-18. Baccarelli et al. (2006)—Site-specific analysis
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
1 Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Polymerase chain reaction methods were used to describe outcome
measures. The prevalence of t(14; 18) was estimated as those individuals having a t(14; 18)
positive blood sample divided by the t(14; 18) frequency (number of copies per million
lymphocytes).
Risk estimates are not susceptible to important biases.
Consideration satisfied. Questionnaire data were used to collect information on cigarette smoking.
Other potential confounders (age, smoking status, and duration of smoking). In addition, both
exposure and outcome were objectively and accurately measured.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. Associations were detected between the frequency of t(14; 18) and
plasma TCDD levels as well as zone of residence at the time of the explosion. No association was
detected for these exposure measures and prevalence of t(14; 18). A dose-response trend was
detected for TCDD and the mean number of t(14;18) translocations/106 lymphocytes, however the
relevance of t(14; 18) in lymphocytes to non-Hodgkin lymphoma is uncertain.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. The authors highlight that exposure metrics represent both past and
current body burdens. They employ several different exposure metrics of TCDD: place of
residence (Zone A, B, R or reference), categorical serum measures, a linear term, log (base 10)
transformed TCDD, and individuals with chloracne diagnosed after the accident.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. Analyses are made using 72 highly exposed, and 72 low exposed
individuals.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Carcinogenesis, 2006, 27(10):2001-2007. The authors discuss the limitation of
using t(14; 18) translocations as an outcome measure, and the uncertain role it plays in the
development of non-Hodgkin lymphoma.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. A total of 144 subjects were included in the study. This included 72 subjects
who had low exposures, and 72 who had high exposures based on serum concentrations.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. A variety of measures were employed including current TCDD levels, as well as
surrogates of exposure at the time of the accident.

While an association was observed with the frequency of t(14; 18) translocation, it is uncertain
whether this translates into an increased risk of non-Hodgkin lymphoma. Given the speculative
nature of this endpoint and lack of demonstrated adverse effect, dose-response analyses for this
outcome were not conducted.
                                   C-180

-------
Table C-19. Warner et al. (2002)—Breast cancer incidence
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Diagnoses of incident breast cancer were based on interview and
information from medical records appears thorough. Of the 15 cases of breast cancer, 13 were
confirmed by pathology and the remaining 2 by surgery report only. Three cases of breast cancer
were excluded which represents a large proportion of the total cases identified. This would reduce
sample size and could result in bias if the exclusion was association with TCDD exposure.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Information was collected on an extensive series of risk factors by using
an interviewer administered questionnaire. Participation rates for the survey were fairly good
(80%).
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. Limited evidence (not statistically significant) of a dose response when
TCDD was analyzed as a categorical variable; only one breast cancer case was in the referent
exposure category. In the analysis of TCDD as a continuous measure (logi0TCDD), the hazard
ratio associated with a 10-fold increase in TCDD serum levels was 2. 1 (95% CI = 1.0-4.6).
Exposure assessment methodology is clear and characterizes individual-level exposures..
Consideration satisfied. Different exposure metrics were considered in these analyses (categorical,
continuous, measures on a log-scale). Exposure data are of high quality as they are based on
serum samples taken among women near the time of the accident. As such, exposure assignment
is not dependent on as many assumption as used in occupational cohorts were back-extrapolation
for many years had to be performed.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration somewhat satisfied. Inadequate follow-up for cancer limited the number of cases
available. Sample size also limited the conclusions draw from the categorical analysis based on
very few cases for some exposure categories.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Paper published in Environ Health Perspect, 2002 Jul, 1 10(7):625-628. A majoi
limitation of the study is the small number of incident cases of breast cancer (n = 15), important
strengths of the study include characterization of TCDD using serum collected near the time of the
accident.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Serum was used to estimate TCDD levels in 981 of 1,271 eligible women who
had lived in either of the two contaminated sites in 1976. Data represent an objective measure of
TCDD near the time of the exposure. Data obtained near the time of exposure which minimized
the potential for exposure misclassification.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. Exposure characterized using serum measures obtained close to the time of the
accident.

While characterization of exposure and availability of other risk factor data at an individual-level
basis are important strengths of this study, small sample size (n = 15 cases) based on inadequate
follow-up is a key limitation. Quantitative dose-response analyses were conducted using this
study, but continued follow-up of the study population or consideration of all cancer outcomes
would be valuable.
                                 C-181

-------
C.2.5. The Chapaevsk Study
      Table C-20. Revich et al. (2001)—All cancer sites combined, and site-specific
      analyses
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration cannot be evaluated. Insufficient details are provided in the paper to gauge the
completeness and coverage of the cancer registry and mortality data. Health outcomes were
examined on the basis of information in the official medical statistics.
Risk estimates are not susceptible to important biases.
Consideration not satisfied. Given the aforementioned limitations of this ecological study, it is
unclear to what extent the results may be subject to bias
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. Dose response was not evaluated as exposure was based on residency
in the region vs. no residency.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration not satisfied. No individual-level exposure estimates were used.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. A total of 476 cancer deaths were observed among males, and 376 cancer
deaths observed among females. The precision of the SMRs is demonstrated with fairly narrow
confidence intervals for many causes of death.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Published in Chemosphere, 2001, 43(4-7):95 1-966. Authors do not address the
completeness of the mortality follow-up, and whether there are differences in mortality surveillance
between regions. The authors do acknowledge, however, that new investigations being undertaken
would characterize exposure using serum-based measures.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria not satisfied. It is a cross-sectional study that compares mortality rates between regions.
No individual-level exposure data available.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. No individual-level exposure estimates were used in the study.

These cancer data are cross-sectional in nature; therefore, dose-response analyses were not
conducted for this study.
C.2.6. The Air Force Health ("Ranch Hands") Study
      Table C-21. Akhtar et al.
      analyses
-All cancer sites combined and site-specific
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Cancer incidence and mortality based on information from repeated
medical examinations, medical records and death certificate.
Risk estimates are not susceptible to important biases.
Consideration satisfied. The risk estimates were adjusted for a number of factors measured on an
individual level, including smoking.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. There is evidence of a dose response for all cancers and for some site-
specific cancers (i.e., malignant melanoma, and prostate cancer).
Exposure assessment methodology is clear and characterizes individual-level exposures.
                                       C-182

-------
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Consideration satisfied. High quality exposure data for most veterans was collected, so
extrapolation to other members of the cohort was not required. The serum dioxin measurements
also correlated well with reported skin exposure to herbicide in Vietnam, but collection of the
samples 25 years later required back-extrapolation.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. In total, 117 incidence cancers identified in the Ranch Hands cohort. For
those sites with a dose-response association, malignant melanoma and prostate cancer, there were
16 and 34 incident cases, respectively.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Published in J Occup Environ Med, 2004, 46(2): 123-136. Authors highlight
that this is only cancer incidence study in US veterans, and the lengthy interval of follow-up
(35-40 years) — both important strengths of the study. They addressed potential bias from healthy-
worker effect, and uncertainties surrounding the estimation of TCDD exposure (extrapolation 30
years after exposure), as well as exposure to other chemical exposures. Study uses incident
outcomes for cancer.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Individual exposure estimates are based on measurements of dioxin serum lipid
concentrations. They were available for 1,009 Ranch Hands and 1,429 in the comparison cohort.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. TCDD exposures at the end of duty were estimated by back-extrapolating 1987
serum values.

This study is suitable for TCDD dose-response modeling of cancer outcomes data.
Table C-22. Michalek and Pavuk (2008)—All cancer sites combined
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Cancer incidence was ascertained through the use of medical records.
Death certificate were used to identify some malignancies. Little data is provided on the number
of individuals lost to follow-up, however the same mechanisms of case ascertainment were
applied to both the comparison and Ranch Hand cohorts.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Information collected from repeated physical examinations allowed for
the adjustment of risk factors such as smoking and exposure related factors such military
occupation and number of years served.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied for some comparisons. Statistically significant associations were noted
with cancer incidence and TCDD when analyses were restricted to workers who served at most
two years in Southeast Asia and those who sprayed more than 30 days before 1967.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Initial TCDD dose were estimated at the end of the tour of duty for the
Ranch Hands. Individual-level serum dioxin measurements correlated well with correlated with
days of spraying and calendar period of service, but collection of the samples roughly 20 years
later required back-extrapolation.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. A total of 347 incident cases of cancer were used in the analyses. For
stratified analyses, statistical power is more limited. For example, only 67 incident cancer in the
subset of workers who spent less than 2 years in Southeast Asia, and sprayed for at least 30 days
before 1967.

Study is published in the peer-reviewed scientific literature.
                               C-183

-------
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Criteria satisfied. J Occup Environ Med 2008; 50:330-340. The authors discuss issues related to
exposure misclassification error, and suggest approaches for improving characterization of days of
spraying. Congener specific data were unavailable, thereby not allowing for congener specific
risks or adjustments to be made.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. TCDD data was available for 986 veterans in the Ranch Hand cohort, and 1,597
members of the comparison cohort.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. TCDD exposures at the end of duty were estimated by back-extrapolating 1987
serum values.

This study is suitable for TCDD dose-response modeling of cancer outcomes.
C.2.7. Other Studies of Potential Relevance to Dose-Response Modeling
      Table C-23. 't Mannetje et al. (2005)
      analyses
-All cancer sites combined, site specific
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. National records for death registrations through the New
Zealand Health Information Service. Subjects not registered as having died during the study
period were confirmed to be actually alive and resident in New Zealand using the New Zealand
Electoral Roll, drivers' license, and social security records.
Risk estimates are not susceptible to important biases.
Consideration not satisfied. Seventeen percent of workers were lost to follow up but it is unclear
if bias resulted. The dichotomous exposure measure was based on exposure to TCDD, chlorinated
dioxins and phenoxy herbicides, so confounding is a possibility by these coexposures.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. Dose-response evidence for duration of employment and elevated
mortality noted only in synthesis workers.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Exposure measures were limited to duration of employment and
exposed/unexposed.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. For all cancer sites combined, there were 43 cancer deaths among the
production workers, and 35 such deaths among the sprayers. Site-specific cancer analyses are
limited by small sample sizes.

Study is published in the peer-reviewed scientific literature.
Criteria not satisfied. Occup Environ Med, 2005; 62:34-40. A high percentage of the cohort was
lost to follow-up (17%). The authors fail to mention this important limitation in this paper.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria not satisfied. This study used duration of exposure, at an individual level, as a surrogate
measure of TCDD.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. Exposure was defined according to duration, and not concentrations of
TCDD. Latency intervals were not evaluated.

Overall, quantitative exposure data are lacking for TCDD and limited dose-response relationships
were observed across duration of exposure categories. Furthermore, confounding by coexposures
is a possibility. Taken together, these data are not suitable for inclusion in a dose-response
analysis
                                       C-184

-------
        Table C-24.  McBride et al. (2009a)—All cancer sites combined, site-specific
        analysis
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
1 Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. The New Zealand Health Information Service Mortality Collection and
the Registrar-General's Index to Deaths. Additional searches were based on the last known
address from the work record; the electoral roll and the habitation index; the telephone book; the
internet; and Terranet property information database. An additional search was carried out through
the Births, Deaths, and Marriages office of the New Zealand Department of Internal Affairs.
Lastly, automated personnel and pension records were also used to locate past New Plymouth
workers and identify some deaths.
Risk estimates are not susceptible to important biases.
Consideration not satisfied. Considerable amount of workers were lost to follow up (22%), but it
is unclear if bias resulted. The dichotomous exposure measure was based on exposure to TCDD,
chlorinated dioxins and phenoxy herbicides, so confounding is a possibility by these coexposures.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. Some SMRs for site-specific cancers were elevated but not
statistically significant. There was no examination of dose-response effects.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Dichotomous exposure (exposed/unexposed) and duration of employment
were examined from job exposure classification assessed via occupational history records
industrial hygienists/factory personnel knowledge and questionnaires. Authors discuss limitations
in the assignment of exposure among cohort members.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration not satisfied. A low number of deaths (n = 76) may have limited ability to detect
effects small in magnitude and exposure-response relationships.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Published in Occup Medicine, 2009; 59(4):255-263. The authors highlight
cohort lost to follow-up (22%), the limited size of the cohort, differences in cohort definitions
)etween sprayers and producers, and the potential for other exposures during employment at the
plant.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria not satisfied. TCDD exposures were not quantified.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. Effective dose could not be estimated given the lack of individual-level
TCDD exposure data.

The study lacks the quantification of exposures at an individual level, precluding dose-response
analysis. This study is not considered further in the dose-response modeling analysis.
        Table C-25.  McBride et al. (2009b)
        analysis
                               -All cancer sites combined, site-specific
1.  Consideration
Methods used to ascertain health outcomes are clearly identified and unbiased.
Response
Consideration satisfied. The New Zealand Health Information Service Mortality Collection and
the Registrar-General's Index to Deaths were used to identify deaths. Additional searches were
based on the last known address from the work record; the electoral roll and the habitation index;
the telephone book; the internet; and several other public databases in New Zealand. An additional
search was carried out through the Births, Deaths, and Marriages office of the New Zealand
Department of Internal Affairs. Lastly, automated personnel and pension records were also used to
locate past New Plymouth workers and identify some deaths.	
                                               C-185

-------
2. Consideration
Response
3. Consideration
Response
1 Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Risk estimates are not susceptible to important biases.
Consideration satisfied. Workers lost to follow-up (21%) were an unlikely source of bias since
there was no evidence that this loss was differential in the internal analyses of workers.
Confounding by sex, hire year, and birth year was addressed by adjustment in regression models.
Potential confounding by other coexposures (e.g., 2,4,6-TCP) unlikely to have resulted in bias, due
to presumed poor correlation with TCDD.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. Although not statically significant, elevated SMRs (>1.6) were noted
for soft tissue sarcoma, non-Hodgkin Lymphoma, multiple myeloma and rectal cancer. The linear
test for trend for TCDD exposure was not statistically significant for all cancer sites (combined), as
well as lung cancer mortality. Dose-response relationships were not apparent across quartiles of
TCDD exposure for all cancer sites combined, digestive cancers, lung cancer, soft tissue sarcomas
or non-Hodgkin Lymphoma.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Cumulative exposure to TCDD as a time-dependent metric was estimated
for each worker from serum samples, but the authors did not examine a continuous measure of
TCDD exposure (lagged or unlagged).
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. The adequate statistical power to detect associations that were present was
a strength of the study owing to the large sample size (» = 1,599 workers), extensive follow-up
period (35 years) and considerable exposure gradient.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Published in J Occup Environ Med 5 1 : 1049-1056. This paper discussed the
strengths and limitation of the study
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Serum measures available for 346 workers were used to derive TCDD exposures
for the entire cohort using the area under the curve approach.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. Although, effective dose could be estimated from serum-derived cumulative
exposure estimates, the exposure models did not consider different latency periods.

Given that no dose-response relationships were found, the data are not suited to dose-response
analysis.
Table C-26. Hooiveld et al.
analysis
-All cancer sites combined, site-specific
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Outcomes were mortality. Few deaths expected to be missed since only
5% of the cohort was lost to follow-up or had emigrated.
Risk estimates are not susceptible to important biases.
Consideration not satisfied. Although dioxin-like compounds (PCDDs, PCDFs, and PCBs) were
measured in the serum samples, these were not incorporated into the analysis. Therefore,
confounding cannot be ruled out as an explanation of the reported association.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. A dose-response pattern was observed for internal cohort comparison for
all cancer mortality, with RRs of 5.0 and 5.6 for the medium and high exposure, respectively.
Dose-response patterns evident for lung cancer as well.
Exposure assessment methodology is clear and characterizes individual-level exposures..
                                  C-186

-------
Response
5. Consideration
Response

1. Criteria
Response
Consideration satisfied. Detailed occupational histories to assign dichotomous exposures
(exposed/unexposed) based on maximum exposure levels. Although serum data also collected for
TCDD and other coexposures (PCDDs, PCDFs, and PCBs), study only presents data for TCDD
exposure. TCDD exposures at time of maximum exposure were extrapolated from measured
serum.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration not satisfied for internal cohort comparisons in either men or women. Among men,
only 7 cancer deaths were observed among those in the unexposed part of the cohort, and 5 1
among exposed workers. For external cohort comparisons, a total of 20 deaths were observed.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Am J Epidemiol, 1998, 147:891-901. The authors address potential limitations
of estimating TCDD exposure from a subsample of surviving workers, lack of smoking data, the
healthy worker effect, and relevance of other occupational exposures.
2. Criteria (Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Response [Criteria satisfied. Serum samples were obtained from 94 of 144 subjects who were asked to
participate in serum measurement study. Of these, a further 44 excluded due to absence due to
poliday or work (n = 22), and nonexposed workers excluded because matching exposed worker not
participating (n = 20). TCDD levels were extrapolated to the time of maximum exposure.
3. Criteria
Response

Conclusion
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. Exposures assigned based on levels at maximum exposure. Assignment of
exposure based on nonrepresentative sample of 50 survivors among the occupational cohort.

The small number of identified cancer deaths, limitations in terms of the exposure assignment
(based on nonrepresentative sample, and maximum exposure level) and concern over potential
confounding by coexposures preclude using these data for a dose-response analysis.
C.3.   EVALUATION TABLES FOR NONCANCER STUDIES
C.3.1. NIOSH Cohort

      Table C-27.  Steenland et al. (1999)—Mortality (noncancer)
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
1 Consideration
Response
Methods used to ascertain health outcomes are clearly identified and unbiased..
Consideration satisfied. The study evaluated mortality from all cancer sites (combined). As
described in the paper, the sources of vital status and cause of death information were received
from the Social Security death files, the National Death Index, and the Internal Revenue Service.
Vital status was known for 99.4% of the cohort members, cause of death information is available
for 98% of the decedents.
Risk estimates are not susceptible to important biases.
Consideration not satisfied. External comparisons for all-cause and cardiovascular mortality do
not appear to be affected by the "healthy worker effect" as similar patterns were observed with
internal cohort comparisons. Nonetheless, internal cohort comparisons are unable to adjust for
many of the individual -level risk factors for cardiovascular disease.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. A dose-response relationship was observed with ischemic heart disease
(linear test for trend p = 0.05), and with TCDD on a log-transformed scale the />-value was <0.00 1 .
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. The study conducted detailed sensitivity analyses and evaluated different
assumptions regarding latency, log-transformed TCDD exposures, and half-life values for TCDD.
Associations were stronger for log-transformed values, and latency intervals of 15 years.
                                    C-187

-------
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. This is the largest of the occupational cohorts with exposures to TCDD.
The cohort consisted of 5,132 male workers and a total of 456 deaths from ischemic heart disease.
This permits characterization of risk for all cancer sites (combined).

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Journal of the National Cancer Institute, 1999, 91(9):779-786. The authors
discussed the potential for bias from smoking, and other occupational exposures for which data for
both were lacking at an individual basis.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Exposure scores assigned at an individual level based on JEM. The JEM was
based on estimated factor of contact with TCDD in each job, level of TCCD contamination of
materials at each plant over time, and proportion of day worker could be in contact with materials.
These factors were multiplied together to derive a daily exposure score, which was accumulated
over the working history of each worker to obtain a cumulative measure of TCDD.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. The follow-up of the cohort extended from 1942 until the end of 1993.
Greater than 25 years of follow-up have accrued in cohort allowing for latency to be examined.
Different assumptions on the half-life of TCDD were evaluated and produced similar results.
Latency intervals were incorporated, with strongest associations noted no lag. Suggests
mechanisms occur at the same time as exposure. However, noncancer mortality is not a viable
endpoint to consider for further dose-response analysis.

TCDD exposures were quantified in this study, and a dose-response relationship was observed
with ischemic heart disease mortality. The sample size was sufficient, and the follow-up interval
was lengthy. However, no individual-level data were available for cardiovascular conditions, and
the inability to adjust for these exposures introduces considerable uncertainty into the risk
estimates. Furthermore, noncancer mortality is not considered a viable endpoint for dose-response
analysis.
Table C-28. Collins et al. (2009)—Mortality (noncancer)
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Vital status complete for all but two workers.
Risk estimates are not susceptible to important biases.
Consideration satisfied. No information collected on smoking status, but no excess in lung cancer
or nonmalignant respiratory diseases noted. Analyses took into account potential for exposure to
pentachlorophenol. External cohort comparisons should be interpreted cautiously due to healthy
worker effect, but internal cohort comparisons should not be influence by this bias.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. No statistically significant mortality excess for any noncancer
mortality outcome evaluated. This included ischemic heart disease, stroke, nonmalignant
respiratory disease, ulcers, cirrhosis, and external causes of death (accidents). Modeling of
continuous measure of TCDD was not related to diabetes, ischemic heart disease, or nonmalignant
respiratory mortality.
Exposure assessment methodology is clear and characterizes individual-level exposures.
                                  C-188

-------
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusions
Consideration satisfied. The authors used serum samples from 280 former TCP workers to
estimate historical exposure levels of TCDD, furans, and polychlorinated biphenyls for all
1,615 workers. Exposure assessment included detailed work history, industrial hygiene
monitoring, and the presence of chloracne cases among groups of workers. This data was
ntegrated into a 1 -compartment, first-order pharmacokinetic to determine the average TCDD dose
associated with jobs in each group, after accounting for the presence of background exposures
estimated from the residual serum TCDD concentration in the sampled individuals. The authors
did not evaluate departures from linearity, or examine skewness at higher exposures. No
presentation of exposure levels was provided.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. A total of 662 deaths were observed. Of these, 218 were from ischemic
leart disease, and 16 from diabetes (two outcomes for which associations have been noted
elsewhere).

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Published in Am J Epidemiol, 2009, 170(4):501-506. The authors discuss
)otential for exposure misclassification, large size of the cohort, lengthy follow-up interval, and
arge number of workers who provided serum from which TCDD exposures were estimated.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. This study has the greatest number of serum samples obtained from a specific
plant.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. Noncancer mortality is not a viable endpoint to consider for further dose-
response analysis.

No dose-response associations were noted for noncancer mortality outcomes. The data are,
therefore, not suited for dose-response modeling.
C.3.2. BASF Cohort
      Table C-29. Ott and Zober (1996a)—Mortality (noncancer)
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Mortality ascertainment appeared to be fairly complete.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Information was collected on smoking status, body mass index, and other
occupational exposures, however a large portion of the cohort was firefighters who may have been
exposed to other occupational carcinogens. However, the recruitment of survivors may results in
under-ascertainment of mortality.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. For external cohort comparisons across the three TCDD exposure
categories, there was no dose-response pattern observed for any of the noncancer causes of death.
Cox regression risk estimates for all cause or circulatory disease mortality when TCDD was
modeled as a continuous variable were not statistically significant.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Cumulative measure of TCDD expressed was derived from serum
measures. Exposure was also estimated by chloracne status of the cohort members. The authors
have not addressed the potential implication of deriving TCDD exposure estimates for the whole
cohort using sera data that were available for only about half of the cohort.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. For all causes of death, there were 92 deaths, while 37 circulatory deaths.
Many of the cause-specific death had less than 5 deaths in the upper exposure category.

                                     C-189

-------
1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Occup Environ Med, 1996, 53:606-612. A large component of the cohort was
assembled by actively seeking out workers who were alive in the mid 1980s. As a result, it is
likely a number of deaths were missed. This is supported by much lower SMRs in this component
of the cohort published in earlier studies of the cohort. This underascertainment of mortality
results in biased SMR statistics (underestimated). The authors do highlight the value of the serum
sased measures to estimate TCDD exposure
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Serum samples, taken in 1989, were available for 138 surviving workers out of
254 and allowed for cumulative TCDD levels to be estimated using regression techniques in the
remainder of the cohort.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. Exposure assignment took into the affect that body mass index had on
TCDD half-lives. TCDD levels estimates through back-extrapolation of serum levels based on
half-life estimates obtained from previous studies. Latency was considered with stronger
association observed in external comparisons incorporating a latency of 20 years. The follow-up
of the cohort was lengthy (>50 years). However, noncancer mortality is not a viable endpoint to
consider for further dose-response analysis.

No associations noted with any noncancer deaths. External comparisons should be treated
cautiously especially for cardiovascular mortality which is recognized to often be biased by the
healthy -worker effect. In the absence of any outcome with an association with TCDD exposure,
dose-response analyses of these data were not undertaken.
C.3.3. Hamburg Cohort
      Table C-30. Flesch-Janys et al. (1995); Flesch-Janys et al. (1996) erratum-
      Mortality (noncancer)
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
1 Consideration
Response
5. Consideration
Response

1. Criteria
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Medical records used to identify deaths over the period 1952-1992.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Similarity in smoking rates between control cohort and the exposed
workers was similar based on independent surveys. Occupational exposures to benzene, and
dimethyl sulfate were unlikely to bias dose-response pattern observed as these exposures occurred
in production departments with low to medium levels of TCDD exposure.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. Dose-response relationship observed for all-cause mortality,
cardiovascular mortality, and ischemic heart disease mortality across 6 exposure categories, with
the cohort of gas supply workers used as the referent. The linear tests for trend for these three
outcomes were all statistically significant (p < 0.05).
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. The exposure measures was an integrated TCDD concentration over time
estimate that back-calculated TCDD exposures to the end of the employment. Categorical and
continuous TCDD exposures were examined in relation to the health outcome. These efforts
improve the exposure assessment of earlier studies.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. For all causes of death combined, there were 414 deaths in the exposed
cohort, and 943 in the cohort of gas supply workers. A total of 157 and 76 deaths from
cardiovascular disease, and ischemic heart disease were noted. The corresponding number in the
cohort of gas supply workers was 459, and 205, respectively.

Study is published in the peer-reviewed scientific literature.
                                       C-190

-------
Response
Criteria satisfied. Am J Epidemiol, 1995, 1442:1165-1175. The authors discuss the potential role
of other occupational exposures (i.e., dimethyl sulfate, solvents, benzene), smoking, and suitability
of the comparison cohort of gas supply workers.	
2.  Criteria
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Response
Criteria satisfied. Serum and adipose tissues were used to estimate TCDD exposure in
190 workers. A one-compartment first-order kinetic model was used to estimate exposure at end
of exposure for these workers. Regression methods were then used to estimates TCDD exposures
for all workers.
3.  Criteria
Effective exposure is estimable latency and window(s) of exposure are examined.
Response
Criteria not satisfied. Exposure based on half-life estimates from individuals with repeated serum
measures.  Other DLCs were considered with the TOTTEQ exposure metric. Noncancer mortality.
tiowever, is not a viable endpoint to consider for further dose-response analysis.	
Conclusion
Although, the exposure data used within this study are well-suited to a dose-response analysis for
all-cause and cardiovascular mortality given the associations observed, use of noncancer mortality
endpoint is not amenable for further dose-response analysis.	
C.3.4.  The Seveso Women's Health Study
        Table C-31. Eskenazi et al. (2002b)—Menstrual cycle characteristics
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
1 Consideration
Response
5. Consideration
Response

1. Criteria
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Information was also obtained from medical records for all obstetric and
gynecologic conditions. Information on menstrual cycles was obtained from questionnaires.
Women were asked about length of cycles, regularity, how many days flow lasted, and heaviness
of menstrual flow (scanty, moderate, or heavy). Measurement error is likely for the subjective
nature of self-reported menstrual parameters but specificity and sensitivity is difficult to ascertain
due to lack of validation data for these measures.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Detailed risk factor information was collected from questionnaire,
allowing for the potential confounding influence of many risk factors to be controlled for. The
length of cycle study findings may have been affected by the presence of a few outliers.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. A positive dose-response relationship was found with TCDD among
women who were premenarcheal at time of the explosion and longer menstrual cycle. Increased
TCDD exposure was associated with a lower relative risk of scanty menstrual flow. No
association was noted with these two outcomes among postmenarcheal women. A decreased risk
of irregular cycles was also observed with higher TCDD levels.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Serum concentrations of TCDD offer improved exposure assessment,
although delineating the critical exposure window is challenging given the nature of the very
high initial exposure.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. Cohort was large enough as analyses were conducted on 301 women.

Study is published in the peer-reviewed scientific literature.
                                                C-191

-------
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Criteria satisfied. Am J Epidemiol, 2002; 156(4)383-392. Limitations included an inability to
assess affects on menstrual cycle at time body burdens were the highest (at time of the accident).
Also, TCDD was estimated for 1976, not concurrent with their cycles in the previous year, and a
arge number of women were excluded due to intrauterine device or oral contraceptive use.
Strengths included population-based nature of study, with characterization of exposure using
serum, and levels of other polychlorinated dibenzo-p-dioxins and dibenzofurans were at
)ackground levels. Findings for length of menstrual cycle may be unduly influenced by the
presence of some outliers.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. The study population was based on 301 women as those who were over the
age of 44 were excluded, as well as women with surgical of natural menopause, women with
Turner's syndrome, those who had been pregnant or breastfed in the past year, and those who
lad used an intrauterine device or oral contraceptives. For 272 women, TCDD levels were based
on serum data provided in 1976; TCDD levels were back-extrapolated to 1976 levels for the
other 29 women.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. Ideally, TCDD exposures would be concurrent with reporting of cycle
characteristics. Herein, TCDD exposures were based on levels in 1976; however, given the long
lalf-life of TCDD and the same follow-up interval for all women, TCDD exposures in 1976
should correlate well with levels near the time of interview. Further, the critical window of
exposure can be estimated for the women that were premenarcheal at the time of the accident (12
years).

This study meets all of the criteria and considerations for further dose-response analysis.
Although it is difficult to define the biologically relevant critical window of exposure for
quantitative exposure calculations, the critical window of susceptibility is assumed to occur
)etween birth and 13 years of age.
Table C-32. Eskenazi et al. (2002a)—Endometriosis
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Results of a pilot study showed that ultrasounds had excellent specificity
and sensitivity for ovarian endometriosis. Those with uncertain case status were analyzed
separately from cases.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Although more than half of the women were classified as 'uncertain' with
respect to endometriosis disease status, these subjects were analyzed separately from those with
endometriosis detected by laparoscopy or ultrasound. Bias is unlikely since disease
misclassification is not likely to be differential with respect to TCDD exposure status.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. While an increased risk of endometriosis was observed across the
3 TCDD categories, these risks were not statistically significant relative to the lowest exposure
category. The test for trend based on a continuous measure (logi0TCDD) was also not statistically
significant.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Serum concentrations of TCDD offer improved exposure assessment,
although delineating the critical exposure window is challenging given the nature of the very high
initial exposure.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration not satisfied. Only a total of 19 cases of endometriosis were identified, and more
than half of the subjects were listed as uncertain regarding endometriosis incidence.

Study is published in the peer-reviewed scientific literature.
                                C-192

-------
Response
Criteria satisfied. Environ Health Perspect 2002; 110(7)629-634. Author's highlight that this is
the first study to examine the relationship between TCDD and endometriosis, and the availability
of sera data to estimate TCDD levels.  Limitations included the small number of women with
endometriosis, and inability to confirm disease status for those without ultrasound or laparoscopy.
Finally, young women may  have been underrepresented due to cultural difficulties in examining
women who had never been sexually active.	
2.  Criteria
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Response
Criteria satisfied. Eligible study subjects were women between 1 month and 40 years of age at
time of accident.  These analyses excluded virgins, those with Turner's syndrome, and women who
refused the examination of ultrasound.  Serum data were available for the 601 participants on
which the analyses are based.  Of these, 559 had serum measures taken in 1976/77, 25 between
1978 and 1981, and 17 women in 1996.
3.  Criteria
Effective exposure is estimable latency and window(s) of exposure are examined.
Response
Criteria not satisfied. TCDD exposure was estimated at the time of "conception attempt" using
serum measures, with extrapolation from 1976 levels using half-life assumptions. It is difficult to
identify the relevant time interval over which TCDD dose should be considered for dose-response
analysis.  The critical window of exposure is unknown.	
Conclusion
Various reasons preclude the use of these data to conduct dose-response analysis. This includes the
lack of a statistically significant association, the large number of women for which endometriosis
disease status was "uncertain", and uncertainty in estimating the critical period of exposure.	
        Table C-33. Eskenazi et al. (2003)—Birth outcomes
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration not satisfied. Outcomes were identified through self-reported questionnaires and
subject to measurement error. Although there is no direct evidence of bias from differential
reporting, women tended to over-report birth weight, and underreport birth defects in children. As
a large number of women in Seveso underwent voluntary abortion in the first year after the
explosion, an awareness bias may have contributed to differential reporting of pregnancy histories.
Risk estimates are not susceptible to important biases.
Consideration not satisfied. See above.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. There was no association between spontaneous abortions and
log10TCDD, or with small for gestational age. There was some suggestion of decreased mean birth
weight and increased ORs for small for gestational age with TCDD exposure among pregnancies
occurring in the first eight years following the accident; however, none of these achieved statistical
significance atp < 0.05.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Serum concentrations of TCDD offer improved exposure assessment,
although delineating the critical exposure window is challenging given the nature of the very high
initial exposure.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. For spontaneous abortions there were 769 pregnancies. Fetal growth and
gestational age analysis was carried out on 608 singleton births that occurred postexplosion.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Environ Health Perspect, 2003, 1 1 1(7):947-953. The authors highlight potential
limitation of reliance on self-reported data to ascertain pregnancy outcomes. They also address the
relevance of paternal exposures to TCDD on the developing fetus — such exposure data were not
considered in this study.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
                                                C-193

-------
Response
3. Criteria
Response

Conclusion
Criteria satisfied. A total of 745 women in the SWHS had reported getting pregnant, of these 510
women were pregnant after the explosion (888 pregnancies). Analyses of spontaneous abortions
based on 476 women (excludes those with voluntary abortion, ectopic pregnancy, or molar
pregnancy). TCDD measured for 413 women in 1976/77, 12 women between 1978 and 1981, and
1996 for 19 women.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. TCDD exposures were extrapolated to 1976 values. However, there is
considerable uncertainty in estimating exposure levels for narrow critical windows of exposure
(e.g., trimesters during pregnancy) especially for pregnancies that occurred many years after the
explosion in 1976.

The findings of the study are somewhat limited due to the reliance on self -reported information for
pregnancy outcomes and possible awareness bias. The findings were not statistically significant.
Considered together with the uncertainty in estimating exposure levels for narrow critical windows
of exposure, dose-response analyses for this study were not conducted.
Table C-34. Warner et al. (2004)—Age at menarche
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. In this study age at menarche was based on retrospective recall 5 to
19 years before the interview. Previous work suggests moderate to high correlations between
actual and recalled menarche, misclassification of outcome would bias risk estimates towards the
null (assuming nondifferential misclassification).
Risk estimates are not susceptible to important biases.
Consideration satisfied. Data collected from self-reported questionnaires allow for the potential
confounding influence of many risk factors to be taken into account. Some misclassification of
outcome may bias risk estimates towards the null.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. There was no association between TCDD levels and the age at
menarche with either the continuous or categorical measures of TCDD in the primary publication,
However, suggestive evidence of an association between serum TCDD concentrations and earlier
age of menarche (HR = 1 .20, 95% CI = 0.98-1 .60, p for trend = 0.07) among 84 women under the
age of 5 at the time of the accident was noted in a follow-up communication from Warner &
Eskenazi (2005)to be when analyses were restricted. The consideration is not satisfied because, in
the context of the RfD derivation, considerable uncertainty remains as to whether associations with
age at menarche represent an adverse health effect.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Serum concentrations of TCDD offer improved exposure assessment,
although delineating the critical exposure window is challenging given the nature of the very high
initial exposure.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. Cohort was large enough as analyses were performed using 282 women
who were premenarcheal at the time of the explosion.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Environ Health Perspect, 2004, 112:1289-1292. Authors discuss use of pooled
serum from residents of the unexposed zone, and that those in lowest exposure group had high
exposures relative with contemporary levels for the area. Strengths of study include use of serum
to estimate TCDD exposure.
                                C-194

-------
2.  Criteria
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Response
Criteria satisfied. The SWHS included women between 1 month and 40 years of age at time of
accident who attempted to get pregnant after the explosion (n = 463). This study is restricted to
those who were premenarcheal at the time of the explosion (n = 282). Serum was collected for
these women, primarily in 1976-1977 (n = 257), between 1978 and 1981 for 23, and in 1996-1997
for the 2 remaining women.	
3.  Criteria
Effective exposure is estimable latency and window(s) of exposure are examined.
Response
Criteria satisfied. TCDD exposures in 1976 were estimated by extrapolation serum levels obtained
after this date using the Filser model. Both categorical and continuous measures of exposure were
modeled.  In utero measures of exposure are likely most relevant exposure based on findings from
animal studies.
Conclusion
No association between TCDD levels and age at menarche was reported in the primary
publication; however, a follow-up communication from Warner & Eskenazi (2005) reported a
10-fold increase in serum TCDD concentrations to be associated with an earlier age of menarche
(HR = 1.20, 95% CI = 0.98-1.60, p for trend = 0.07) when analyses were restricted to 84 women
under the age of 5 at the time of the accident. The TCDD exposure characterization of study
subjects was based on serum data, and no major biases were introduced from the study design or
analytical methods that were used.  In the context of the RfD derivation, considerable uncertainty
remains as to whether associations  with age at menarche represents an adverse health effect,
Therefore, dose-response analyses were not conducted for this study.	
        Table C-35. Eskenazi et al. (2005)—Age at menopause
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Outcome measures were obtained based on self-reported data collected
from questionnaires. Studies have shown that self -reports of age at menopause are reported with
accuracy and reliability, and among women with surgical menopause, the self-reported age
correlated well with that on the medical records.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Data obtained from the questionnaire allow for the potential confounding
influence of several potential confounders to be examined.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. Risks of earlier menopause increased in the first four quintiles, with a
statistically significant trend. No increased risk was noted in the highest exposure category (HR =
1.0 relative to lowest exposure group). The study authors suggest this is due to the "inverted U"
dose response often seen with hormonally active compounds. Additionally, no statistically
significant association was noted with logi0TCDD for the individual quintiles. More importantly,
the biological significance of this result for the establishment of a LO AEL (that is needed in the
context of the RfD derivation) could not be determined with confidence.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Serum concentrations of TCDD offer improved exposure assessment,
although the critical exposure window is uncertain.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. The study included 6 16 women. Of these, 260 were premenopausal, 169
classified as natural menopause, 83 as surgical menopause, 24 as impending menopause, 33 as
premenopausal, and 58 in an "other" category.

1. Criteria IStudy is published in the peer-reviewed scientific literature.
Response
Criteria satisfied. Environ Health Perspect, 1 13 :858-862 (2005). The authors highlight that this is
first study to look at relationship between dioxin and age at menopause. Limitations of the study
were that the lowest exposure group (<20.4 ppt) included exposure levels that are far higher than
)ackground, and age at menopause was based on retrospective recall. A strength of study is ability
to characterize TCDD using serum measures.
                                               C-195

-------
2. Criteria
Response
3. Criteria
Response

Conclusion
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. The Seveso Women's Health Study collected serum sample which allowed
TCDD exposures to be characterized. Those women (n = 616) who had not reached natural
menopause at the time of the accident were included in the study. Serum measures collected in
1976/77 were available for 564 women, for 28 women, sera was collected between 1978 and 1981,
while for 24 women, sera was collected in 1996/97.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. TCDD levels were estimated at the time of the explosion using available
information on TCDD half-life. However, it is difficult to identify the relevant time interval over
which TCDD dose should be considered for dose-response analysis. The critical window of
exposure can be estimated but is large and highly uncertain.

The biological significance of this result for the establishment of a LO AEL (that is needed in the
context of the RfD derivation) could not be determined with confidence. Therefore, dose-response
analyses were not conducted for this study.
Table C-36. Warner et al. (2007)—Ovarian function
1. Consideration Methods used to ascertain health outcomes are clearly identified and unbiased.
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response
Consideration satisfied. Ovarian cyst analysis based on women who underwent ultrasound
(n = 3 10). Ovarian follicle analysis based on self-report on menstrual cycle and done in women in
preovulatory cycle (n = 96) at time of ultrasound. Hormonal analysis based on women in last 14
days of cycle (n = 129).
Risk estimates are not susceptible to important biases.
Consideration satisfied. Data collected from self-reported questionnaires allow for the potential
confounding influence of many risk factors to be taken into account. Some misclassification of
outcome based on self-reports of menstrual cycle may bias risk estimates towards the null.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. There was no association between serum TCDD levels and the
number or size of ovarian follicles. TCDD was also not associated with the odds of ovulation.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Serum concentrations of TCDD offer improved exposure assessment,
although delineating the critical exposure window is challenging given the nature of the very high
initial exposure.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. Cohort was large enough as analyses were performed using 129 women
for ovulation outcome, and hormone analyses based on 87 women in luteal, and 55 in midluteal
phases.

1. Criteria IStudy is published in the peer-reviewed scientific literature.
Response
2. Criteria
Response
3. Criteria
Response

Criteria satisfied. Environ Health Perspect, 2007,115:336-340. An important limitation cited by
the authors was that women may not have been exposed at critical period (prenatally). Phases of
the cycle may also have been misclassified as this was based on self-reported data. Strength, first
study to have examined ovarian function and TCDD exposures.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. The SWHS included women between 1 month and 40 years of age at time of
accident who were between 20-40 years of age and not using oral contraceptives at follow-up (n =
363). Of these, serum was collected for 330 women between 1976 and 1977, between 1978 and
1982 for 25 women, and between 1996 and 1997 for 8 women.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. There is a lack of a defined critical window of exposure in this study.

                                C-196

-------
Conclusion
Because of the lack of a defined critical exposure window and absence of associations between
TCDD and adverse health effects in this study, quantitative dose-response assessment was not
conducted for this study. For these reasons, dose-response analyses were not conducted for this
study.
        Table C-37.  Eskenazi et al.
                             -Uterine leiomyoma
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
1 Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Outcomes were determined using two definitions: current fibroids, or past
diagnosis of fibroids. For past diagnosis of fibroids, self-reported data and medical records were
used to determine whether women were previously diagnosed with fibroids, these were confirmed
with medical records. A total of 25 women indicated they had never been diagnosed with fibroids.
Medical records indicate a past diagnosis for these women, and they were classified as such. For
current fibroids, this was determined at the time of the interview for 634 women using transvaginal
ultrasound examinations.
Risk estimates are not susceptible to important biases.
Consideration satisfied. In the SWHS questionnaires were administered to the participants and
detailed data for reproductive characteristics, smoking, body mass index, and alcohol use were
collected so risks could readily be adjusted for these covariates.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied, but inverse associations reported. An inverse dose-response pattern with
the percentage of women diagnosed (current and past history — combined) with fibroids across 3
categories of exposure. Namely, the percentages of wo men with fibroids inthe<20, 20.1-75.0,
and >75.0 ppt categories were 41.1%, 26.8%, and 20.0%, respectively.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. A variety of different exposure metrics were considered including linear,
categorical, splines, and logioTCDD.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. A total of 25 1 women were found to have fibroids, and there were 62,
110, and 79 women with fibroids diagnosed in the 3 TCDD exposure categories.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Am J Epidemiol, 2007, 166:79-87. In this study, the authors found an inverse
association between TCDD and uterine leiomyoma risk. The authors highlighted strengths of the
study that included the longitudinal design, serum measures taken at an individual-level basis and
most taken within 2 years of the accident, ability to include outcomes among those who did not
take an ultrasound by using an adapted statistical approach. An important limitation that was the
differences in risk by the stage of development could not be assessed as all women were exposed
postnatally, and only 4 cases were observed among those who were premenarcheal at the time of
exposure.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Final sample consisted of 956 women in the Seveso Women's Health Study
without a history of fibroids. For 872 of these women, serum was collected in 1976 and 1977. For
56 women, TCDD was measured in women between 1978 and 1981, and for 28 women the serum
was collected in 1996.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. TCDD exposures were back extrapolated to expected levels in 1976 (at the
time of the accident). However, it is difficult to identify the relevant time interval over which
TCDD dose should be considered for dose-response analysis. The critical window of exposure is
uncertain.

Because the critical window of exposure is uncertain, dose-response analyses were not conducted
for this study.
                                               C-197

-------
C.3.5. Other Seveso Noncancer Studies
      Table C-38. Mocarelli et al. (2008)—Semen quality
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Serum levels of TCDD were measured on an individual basis for men in
exposed areas; pooled samples from men in uncontaminated areas were measured to assess
background TCDD exposure levels.
Risk estimates are not susceptible to important biases.
Consideration satisfied. While compliance rates may have introduced some possible bias, this
does not seem likely as different effects noted between the 22-3 1 and 32-39 year old age groups.
Information collected for other risks factors, which have been used as adjustment factors in the
models.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. Figure 3 (Mocarelli et al.. 2008) suggests dose-response relationship
among those aged 1-9 at the time of the accident for sperm concentration and motility.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Serum concentrations of TCDD offer improved exposure assessment,
although delineating the critical exposure window is challenging.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. Analyses are based on 135 males exposed to TCDD.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Environmental Health Perspective s, 2008, 116(l):70-77. The authors describe
strengths associated with characterization of exposure (using serum samples), and
representativeness of study population. Limitation of study includes low compliance (but high for
semen sample studies), namely, 60% among a group of healthy men. The compliance rate was
higher among exposed group (69%).
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Involved males, <16 years old at time of accident.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. TCDD exposures were based on serum samples. Serum samples were drawn (in
1997/1998) from participants whose 1976 samples were above 15 ppt. Pooled samples obtained in
1997/98 were used to describe background TCDD levels in uncontaminated areas. The associated
between TCDD exposure and semen quality was found statistically significant for the boys with 1
and 9 years of age at the time of the accident. This provides a critical window of exposure to
estimate TCDD concentration.

Health outcomes are exposures are well characterized using serum data. However, the men
exposed between the ages of 1 and 9 to elevated TCDD levels had reduced semen quality 22 years
later. It is difficult to discern whether this effect is a consequence of the initial high exposure
between 1 and 9 years of age or a function of the cumulative exposure for this entire exposure
window beginning at the early age. Nonetheless, dose-response analyses for this outcome were
conducted.
      Table C-39. Mocarelli et al.
-Sex ratio
1. Consideration
Response
2. Consideration
Response
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Birth records examined for those who lived in parents who lived in the
area and who provided serum samples.
Risk estimates are not susceptible to important biases.
Consideration satisfied.
                                       C-198

-------
3. Consideration
Response
1 Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. Paternal TCDD exposures were associated with an increased probability
of female births (p = 0.008).
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Serum samples were used to estimate maternal and paternal TCDD levels.
No discussion of exposure levels in reference population.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. Statistically significant findings achieved.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. The Lancet, 2000, 355:1858-1863.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Serum levels of TCDD were obtained from parents using samples provided in
1976/77 '. Serum measures available for 296 mothers and 239 fathers.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. Serum based measures of TCDD were obtained shortly after the accident.
TCDD levels were also extrapolated to the time of conception. Although paternal pubertal
exposures may be a key critical window for sex differentiation, it is difficult to identify the
relevant time interval over which TCDD dose should be considered for dose-response analysis.

The data from this study demonstrate a positive dose-response relationship with pubertal and pre-
pubertal paternal TCDD levels at the time of the accident and increased likelihood for female
births. However, it is difficult to identify the relevant time interval over which TCDD dose should
be considered; specifically, it is difficult to discern whether this effect is a consequence of the
initial high exposure during childhood or a function of the cumulative exposure for this entire
exposure window beginning at the early age. Dose-response analysis for this outcome was not
conducted, because EPA could not define the critical exposure window.
Table C-40.  Baccarelli et al. (2008)—Neonatal thyroid function
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
1 Consideration
Response
5. Consideration
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Measures of b-TSH are taken using a standardized protocol 72 hours after
sirth. These b-TSH measures are taken on all newborns born in the region of Lombardy which
includes Seveso.
Risk estimates are not susceptible to important biases.
Consideration satisfied. For the comparisons involving place of residence at the time of the
accident, exposure misclassification is likely given variability in soil TCDD exposure levels within
these areas. For the individual TCDD measures (n = 5 1) reported in the study figures, exposure
misclassification is unlikely.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. Mean neonatal b-TSH was 0.98uU/ml [0.90-1.08] in the reference area,
1.35uU/ml [1.22-1.49] in zone B, and 1.66uU/ml [1.19-2.31] in zone A (p< 0.001). The plotted
frequency distributions have similar shapes, but have shifted to the right for areas of higher
exposures. Neonatal b-TSH was correlated with current maternal plasma TCDD ((3-0.47,
v < 0.001) in the 5 1 newborns for which individual maternal serum TCDD values were available.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. TEQs were measured among the 38 women for which serum samples were
available and were defined for a mixture of dioxin-like compounds. Maternal mean total TEQs
(PCDDs, PCDFs, coplanar PCBs, and noncoplanar PCBs) was 41.8 ppt. Two measures of
exposure included place of residence at time of accident and plasma samples obtained from
mothers at the time of delivery. Similarities in positive dose-response relationships give stronger
weight to the findings.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
                                 C-199

-------
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Consideration satisfied. For plasma based estimate of maternal TCDD there were 5 1 mother-child
pairs. Only seven children in total were found to have b-TSH levels in excess of 5 uU/mL.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. PLOS Medicine 2008; 5(7)1133-1142. The authors discuss the strength of the
study related to characterization of exposure using serum sampling, and ability to adjust for factors
related to b-TSH or TCDD levels (gender, birth weight, birth order, maternal age, hospital and type
of delivery). They also highlight that a limitation of study was that the influence of mother-child
dioxin transfer through colostrum could not be assessed because no information on breast-feeding
)efore b-TSH measurement was available.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. In the population-based study, eligible women who resided in zones A and B at
the time of the accident (n = 1,772) were matched to nonexposed women. In the study based on
)lasma dioxin measurements, participants were the 5 1 children born to 38 women from zones A, B,
I, or a reference zone for which plasma dioxin measurements were available.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. Maternal TCDD levels were estimated at the time of delivery based on plasma
samples, and the critical window of exposure was assumed to be the 9-month gestational period.

The data provide an opportunity for conducting dose-response analyses.
Table C-41.  Alaluusua et al. (2004)—Developmental dental defects
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Ascertainment of dental health was done blind to place of residence, used
standard protocol for caries developed by the WHO, and the clinical examination supplemented by
radiographic examination.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Additional risk factor information was collected on questionnaires. These
factors were considered as adjustment factors. The potential for participation bias is not possible
to ascertain given the available information. The potential impact of exposure misclassification is
also unknown, but the there is some suggestion that some individuals in the non-ABR zone may
higher TCDD levels than expected based on background exposure concentrations.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. Increased prevalence of developmental enamel effects found with
increased TCDD serum measures. Namely, prevalence in unexposed region was 26%, whereas in
the low, middle, and high TCCD groups the prevalence was 10%, 40%, and 60%, respectively.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. TCDD exposure level based on serum lipids. No discussion of exposure
levels in reference population.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. Despite small numbers, statistically significant findings were achieved.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Environmental Health Perspectives, 2004, 112(13): 13 13-13 18. Authors
mention two important strengths of the study: characterization of TCDD exposure using serum
collected shortly after the time of the accident, and the fact that developmental defects are
permanent in nature. Therefore, they represent a health outcome can evaluated years later. Little
discussion was made of the impact of differential compliance rates between the exposed (74%) and
nonexposed (58%) groups. Authors mention two important strength of the study: characterization
of TCDD exposure using serum collected shortly after the time of the accident, and the fact that
developmental defects are permanent in nature. Therefore, they represent a health outcome can
evaluated years later. Little discussion was made of the impact of differential compliance rates
between the exposed (74%) and nonexposed (58%) groups.
                                C-200

-------
2. Criteria
Response
3. Criteria
Response

Conclusion
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Serum levels of TCDD could be estimated for children in exposed areas. No
serum levels were available for reference group of children, and assumption of zero exposure was
made. This seems reasonable.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. It is difficult to discern whether this effect is a consequence of the initial high
exposure during childhood or a function of the cumulative exposure of the entire exposure window
beginning at early age. However, assumptions can be made regarding the critical window of
exposure and the relevant dose can be calculated.

The considerations for conducting a dose-response analysis have been satisfied with the study
population of only those subjects who lived in the ABR zone at the time of the accident; exposure
data are unavailable for those in the referent area. While is difficult to identify the relevant time
interval over which TCDD dose should be considered, dose-response analyses were conducted for
this outcome.
Table C-42. Bertazzi et al. (2001)—Mortality (noncancer)
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
1 Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. For some causes of death methods highly specific mortality appears to be
well captured from the vital statistics registries in the region (99% complete). Some health
outcomes (e.g., diabetes) are subject to misclassification using death certificate data.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Although individual-level data for individual risk factors are not
available, the potential for confounding is likely minimal. For e.g., independent surveys suggests
similarity between smoking behaviors across the regions. Exposure misclassification based on
place of residency likely to bias risk estimates towards the null.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. While a dose-response relationship was observed for chronic
obstructive pulmonary disease across Zones A, and B, this relationship was not.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration not satisfied. Exposure classification was based on the address of the
residence on the date of the accident or when the person first entered the area. Although TCDD
blood levels were also measured, these were not examined with respect to health outcomes. The
lack of individual -level data also precluded an examination of these uncertainties.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. A total of 494 noncancer deaths were found among residents of Zones A,
and B, respectively. This allowed examined of gender-specific effects.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Am JEpidemiol, 2001, 153:1031-1044. Authors discuss lack of
individual-level exposure data and other risk factors (e.g., smoking), difficulties in extrapolating to
)ackground levels, diagnostic accuracy of using death certificates. Strengths included similarities
)etween exposed and comparison population for several risk factors, completeness of follow-up,
and consistent methods to identify mortality outcomes in the exposed and comparison populations.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria not satisfied. Individual-level exposure data are unavailable. Exposure based on place of
residence at time of the explosion. Soil sampling performed indicated considerable variability in
TCDD levels within each region. In addition, place of residency at time of explosion does not
ensure individuals were at their home around the time of the accident.
Effective exposure is estimable latency and window(s) of exposure are examined.
                                  C-201

-------
Response

Conclusion
Criteria not satisfied. An ecological measure of exposure (region of residency at time of accident)
was used to categorize individuals according to their possible exposure. Latencies were
considered. While such an approach has value for identifying whether excesses occurred among
highly exposed populations, it is not precise enough to conduct dose-response analyses.
Furthermore, noncancer mortality is not a viable endpoint to consider for further dose-response
analysis.

Study is not suitable for dose-response analysis due to mortality as endpoint and lack of
individual-level exposure data.
Table C-43 Consonni et al.
-Mortality (noncancer)
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
1 Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. For some causes of death detection methods were highly specific;
mortality appears to be well captured from the vital statistics registries in the region (99%
complete). Some health outcomes (e.g., diabetes) are subject to misclassification using death
certificate data.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Although individual-level data for individual risk factors are not
available, the potential for confounding is likely minimal. For e.g., information from other
independent surveys suggests similarity between smoking behaviors across the regions. Exposure
misclassification based on place of residency is likely to bias risk estimates towards the null.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. Statistically significant association noted in most highly exposed area
for chronic rheumatic disease and chronic obstructive pulmonary disease. Dose-response pattern
noted across Zones A, B and R for circulatory disease mortality 5-9 years after the accident.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration not satisfied. Lack of individual-level data precludes an examination of these
uncertainties.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. However, only three deaths from diabetes occurred among residents of
Zone A. The limitation related to statistical power is exacerbated for stratified analyses carried out
by number of years since the accident.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Am J Epidemiol, 2008, 167:847-858. Authors discuss potential for selection
bias, limitation of residential based measure of exposure, similarities of mortality ascertainment in
exposed and referent populations, and multiple testing.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria not satisfied. Individual-level exposure data are unavailable. Exposure based on place of
residence at time of the explosion. Soil sampling performed indicated considerable variability in
TCDD levels within each region. In addition, place of residency at time of explosion does not
ensure individuals were at their home around the time of the accident.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. An ecological measure of exposure (region of residency at time of accident)
was used to categorize individuals according to their possible exposure. Latencies were
considered. While such an approach has value for identifying whether excesses occurred among
highly exposed populations, it is not precise enough to conduct dose-response analyses.
Furthermore, noncancer mortality is not a viable endpoint to consider for further dose-response
analysis.

Study is not suitable further dose-response evaluation due to noncancer morality endpoint.
                                C-202

-------
Table C-44. Baccarelli et al. (2005)—Chloracne
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Chloracne cases identified using standardized criteria.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Important potential confounders were included in the quantitative
analyses conducted by the study authors.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. Plasma TCDD was associated with an increased risk of chloracne. The
odds ratios increased in a dose-response pattern across zone of residence.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Authors discussed implications of differential elimination rates by age
and body growth.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. A total of 101 chloracne cases were identified, and 211 controls were
selected. Statistically significant findings were observed in several comparisons, although
statistical power was limited to assess potential interactions.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. British Journal of Dermatology, 2005, 152,459-465. The authors detail the
limited statistical power they had available in the study. They also highlight study strengths that
included uniqueness of age and sex distribution of chloracne cases, characterization of TCDD that
could be done using sera samples, and availability of both clinical and epidemiologic data.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. TCDD was estimated in both chloracne cases and control using serum measures.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria satisfied. Serum based measures of TCDD were obtained shortly after the accident.
Chloracne is thought to be caused by the initial high exposure.

Exposure to TCDD at sufficiently high levels is recognized to cause chloracne. This study
provides limited relevance to dose-response modeling of TCDD as exposure levels typically
observed in the general population are much lower.
Table C-45. Baccarelli et al.
-Immunological effects
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Common methods were used to describe blood levels of plasma
immunoglobulins (IgA, IgG, and IgM) and complement components (C3 and C4).
Risk estimates are not susceptible to important biases.
Consideration satisfied. Both exposure and outcome were objectively and accurately measured.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. While plasma IgG levels were inversely related with TCDD, it is
uncertain whether this outcome is adverse.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Both categorical (quintiles) and continuous measures of TCDD were
examined in the dose-response analysis.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. Analyses are made using 72 highly exposed, and 72 low exposed
individuals.

Study is published in the peer-reviewed scientific literature.
                                 C-203

-------
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Criteria satisfied. Toxicology letters, 2004, 149:287-293 and Environ Health Perspect, 2002,
110(12): 1169-1 173. The authors highlight that few studies have looked at immunological effects
of TCDD in humans, that the current study was able to exclude those with concurrent medical
conditions, and the ability to characterize exposure using serum measures. Limitations addressed
were the uncertainty about the clinical relevance of the dose-response pattern found, and the
relatively small size of the study population.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. A total of 120 subjects were included in the study. This included 62 randomly
selected from the high exposed zone, and 58 selected from the reference area.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. Dose-response relationships were examined using current TCDD levels.
However, it is difficult to identify the relevant time interval over which TCDD dose should be
considered for dose-response analysis.

An inverse dose-response relationship between IgG and TCDD was observed. However, the
biological significance of a decrease in IgG for the establishment of a LOAEL (needed in the
context of the RfD derivation) could not be determined with confidence. Further the critical
window of exposure that would cause an effect on IgG levels is not known and thus does not allow
for estimation of the effective TCDD exposure. Therefore, dose-response analyses were not
conducted for this outcome.
C.3.6. Chapaevsk Study
      Table C-46. Revich et al.
      health
-Mortality (noncancer) and reproductive
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration not satisfied. Insufficient details are provided in the paper to gauge the completeness
and coverage of the cancer registry and the mortality data. Health outcomes were examined on the
basis of information in the official medical statistics.
Risk estimates are not susceptible to important biases.
Consideration not satisfied. Given the aforementioned limitations of this ecological study, it is
unclear to what extent the results may be subject to bias.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. Dose response was not evaluated as exposure was based on residency
in the region vs. no residency.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration not satisfied. No individual-level exposure estimates were used.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. Population-based data over several years were used to make comparisons
at the ecological level.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Published in Chemosphere, 2001, 43(4-7):95 1-966.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria not satisfied. It is a cross-sectional study that compares mortality rates between regions.
No individual-level exposure data available.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. No individual-level exposure estimates were used in the study.

These cancer data are cross-sectional in nature; therefore, dose-response analyses were not
conducted for this study.
                                       C-204

-------
C.3.7.  Air Force Health ("Ranch Hands") Study
      Table C-47. Henriksen et al. (1997)—Diabetes
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Newly diagnosed cases of diabetes following the completion of the
veterans' tours of duty were identified from self-reported questionnaire data with verification from
medical records, or by using a postchallenge glucose serum test. Disease severity was determined
sased on questionnaire, and review of medical records. Fasting glucose and 2-hour postprandrial
glucose tests were used to identify glucose abnormalities among nondiabetics.
Risk estimates are not susceptible to important biases.
Consideration not satisfied. Adjustment was made for a number of risk factors related to diabetes
(e.g., BMI, family history, smoking). However, variations in the solubility of dioxin due to
setween-subject differences in lipid fractions may account for the positive association observed.
Many of the health outcomes under study (i.e., diabetes, impaired glucose tolerance, insulin
resistance) are associated with lipid abnormalities.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. There were statistically significant positive associations noted between
TCDD and diabetes, as well as changes in serum glucose levels, reduced time to onset of diabetes,
severity of diabetes, and glucose abnormalities among nondiabetics. While many of the
comparisons are based on small numbers, overall, the associations are consistent across the
outcomes that were examined.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. The methods used to estimates TCDD levels are clearly described, and
capture exposure at an individual-level many years before the health outcome was determined.
The authors describe the limitations of the exposure assessment within the paper. Sensitivity
analyses were undertaken for several of the key associations. The key limitation is that the
associations may be caused by differences in lipid fractions between individuals.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. There were a total of 2,265 veterans and 315 cases of diabetes. There was
very little attrition across the four physical examinations performed in 1982, 1985, 1987 and 1992.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. The paper was published in Epidemiology 1997;8:252-258. The discussion
contains an appropriate discussion of the strengths and weaknesses of the study.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Serum was used to characterize TCDD exposure. While the quantification of
TCDD levels at the time the tour of duty ended may be misspecified due to between-subject
differences in lipid fractions, the methods used were able to reasonably discriminate between those
veterans with high and low exposures.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. The nature of the data preclude identification of the critical window of
exposure to be examined and a effective dose to be calculated for this endpoint.

While the health outcomes and TCDD exposures were characterized using valid methods, the
nature of the data preclude identification of the critical window of exposure to be examined. Thus,
dose-response modeling was not conducted for this study.
                                      C-205

-------
Table C-48.  Longnecker and Michalek (2000)—Diabetes
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Newly diagnosed cases of diabetes following the completion of the
veterans' tours of duty were identified from self-reported questionnaire data with verification from
medical records, or by using a postchallenge glucose serum test. Glucose and insulin measures
were obtained among nondiabetics using fasting and 2-yr post challenge serum test.
Risk estimates are not susceptible to important biases.
Consideration not satisfied. Adjustment was made for a number of risk factors related to diabetes
(e.g., BMI, family history, smoking). However, the analysis was cross-sectional in nature, and
therefore was unable to take into account the timing of exposure in relation to diagnosis of
diabetes. The increased solubility of dioxin in triglycerides, whose levels are higher in diabetics,
may account for the positive association observed.
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. There were statistically significant positive associations noted between
TCDD and diabetes, as well between TCDD and serum glucose and insulin levels.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration not satisfied. The methods used to estimate TCDD levels are clearly described and
are able to determine exposures at an individual level. However, the range of exposures is small
given the exclusion of the more highly exposed Ranch Hand veterans. It is possible that between-
subject difference in lipids and triglycerides may introduce an important source of exposure
measurement error. The authors describe the limitations of the exposure assessment within the
>aper. The key limitations include the cross-sectional nature of the data, and the noncausal
associations that may be caused by triglycerides.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. There were a total of 1,197 veterans and 169 cases of diabetes. Levels or
participation across the multiple physical examinations were high.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. The paper was published in Epidemiology 2000;! l(l):44-48. The discussion
contains an appropriate discussion of the strengths and weaknesses of the study.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Serum-based measures are an objective and valid method to determine TCDD
sxposure levels.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. The diabetes cases were identified over a nearly 25-year interval. The
nature of the data and analysis preclude identification of the critical window of exposure and
sstimation of an effective dose for this study.

While the health outcomes and TCDD exposures were characterized using valid methods, the data
are essentially cross-sectional and thus are unable to evaluate associations between TCDD and
diabetes that can take into account the timing of the exposure. Given the narrow range in TCDD
sxposures in this study, particularly given the Ranch Hand workers were excluded, these between-
subject differences may introduce an important source of bias. Further, the nature of the analysis
precludes identification of the critical window of exposure. Thus, dose-response modeling was not
conducted for this study.
Table C-49.  Michalek et al. (200la)—Hematological effects
1. Consideration
Response
2. Consideration
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Hematological measures were determined from serum samples obtained
across four physical examinations.
Risk estimates are not susceptible to important biases.
                                C-206

-------
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Consideration not satisfied. Associations between TCDD and platelet counts may be influenced
sy other health conditions not accounted for by the study design. The positive association noted
setween TCDD and mean corpuscular volume may be noncausal. Specifically, this association
may be due to raised triglycerides levels or increased prevalence of liver impairment among those
more highly exposed to TCDD.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. Most hematological measures were not consistently associated with
TCDD across the different physical examination periods. While positive associations between
TCDD and platelet counts and mean corpuscular volumes were observed, they were not consistent
with a dose-response relationship as statistically significant differences, relative to those in the
lowest exposure group, were observed only among those in the highest exposure group.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. The methods used to estimate TCDD exposure are clearly described, and
capture exposure at an individual level prior to the diagnosis of the health outcome under study.
The authors describe the limitations of the exposure assessment within the paper.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. Continuous measures of hematological function approximately 2,200
veterans at four physical examinations. The study lacked adequate statistical power to perform the
secondary analysis of the relationship between TCDD and abnormally high red blood cell counts.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. The paper was published in Archives of Environmental Health, 200 1 ;
56(7):396-405.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Serum was used to characterize TCDD exposure at end of tour of duty. Given
exposures dropped dramatically for the Ranch Hands following their tours of duty, exposure to
TCDD prior to disease onset is reasonably characterized, though some misclassification between
those in the comparison group and those in the lowest Ranch Hand exposure grouping is
inevitable. Serum-based measures of hematological function were obtained at multiple
examinations which permitted dose-response relationships to be evaluated at four time intervals.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. There is uncertainty in the critical window of exposure. This study analyzes
the potential for associations between point-in-time measures of TCDD serum levels and changes
in hematological measures that may have occurred at any time over approximately a 30-year
interval. The clinical relevance of reported outcomes also is uncertain.

While the health outcomes and TCDD exposures were characterized using valid methods, most
hematological measures were not associated with TCDD. For corpuscular volume and blood
platelet levels an association with TCDD was detected. However, this association may be
noncausal and the influence of other confounders cannot be entirely ruled out. The clinical
relevance of these outcomes is also uncertain. Further, no dose-.response trend was observed with
either of these two hematological measures. Additionally, there is uncertainty in the critical
window of exposure. For these reasons, dose-response modeling was not conducted for this study.
Table C-50. Michalek et al. (200 lb)—Hepatic abnormalities
1. Consideration
Response
2. Consideration
Response
3. Consideration
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Hepatic function measures were determined from serum samples obtained
across four physical examinations, and the prevalence of liver disorders was determined using self-
reported data verified by medical records.
Risk estimates are not susceptible to important biases.
Consideration not satisfied. Associations between TCDD and liver function may be influenced by
other health conditions not accounted for by the study design.
Study demonstrates an association between TCDD and adverse health effect.
                                C-207

-------
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Consideration satisfied. No dose-response trend was observed with most measures of liver
[unction. There was no association between TCDD and hepatomegaly or nonalcoholic chronis
iver disease and cirrhosis. However, an association between TCDD was observed with y-
glutamyltransferase, and increased odds ratios of several hepatic disorders were observed among
those in the highest TCDD exposure group relative to the comparison cohort.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. The methods used to estimate TCDD exposure are clearly described, and
capture exposure at an individual level prior to the diagnosis of the health outcome under study.
The authors describe the limitations of the exposure assessment within the paper.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. Continuous measures of liver function approximately 2,200 veterans
during the 1992 physical examination. For some liver conditions, there were few prevalent cases
across the exposure categories, however, statistically significant differences were observed for
many conditions when comparisons where made between those in the highest exposure group
relative to the lowest.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. The paper was published in Annals of Epidemiology 200 1 ; 1 1 :304-3 1 1 .
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Serum was used to characterize TCDD exposure at end of tour of duty. Given
sxposures dropped dramatically for the Ranch Hands following their tours of duty, exposure to
TCDD prior to disease onset is reasonably characterized, though some misclassification between
those in the comparison group and those in the lowest Ranch Hand exposure grouping is
nevitable. Serum-based measures of liver function were obtained at the 1992 examination which
permitted dose-response relationships to be examined.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. There is uncertainty in the critical window of exposure. This study analyzes
the potential for associations between point-in-time measures of TCDD serum levels and liver
disease that may have occurred at any time over approximately a 25-year interval the clinical
relevance of the health endpoints that were examined is uncertain.

The results do not unequivocally support a relationship between liver damage and TCDD
sxposure. Confounding and reverse causality cannot be eliminated. Additionally, there is
uncertainty in the critical window of exposure. This study analyzes the potential for associations
)etween point-in-time measures of TCDD serum levels and liver disease that may have occurred at
any time over approximately a 25-year interval, making it difficult to calculate a cumulative
TCDD effective dose over time. For these reasons, dose-response modeling was not conducted for
this study.
Table C-51.  Michalek et al.
-Peripheral Neuropathy
1. Consideration
Response
2. Consideration
Response
3. Consideration
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. The outcomes were determined using a standardized neurological exam
conducted by a board certified neurologist blinded to exposure status. A number of difference
measures of peripheral neuropathy were obtained over multiple physical examinations.
Risk estimates are not susceptible to important biases.
Consideration not satisfied. Some of the observed associations may be due to residual
confounding by diabetes.
Study demonstrates an association between TCDD and adverse health effect.
                                C-208

-------
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Consideration satisfied. For some measures of peripheral neuropathy, the data were suggestive of
a dose-response relationship, particularly for probable symmetrical peripheral neuropathy.
However, only data from the 1997 examination yielded statistically significant increased odds ratio
in the highest exposure category relative to the comparison cohort. Associations between TCDD
and diagnosed peripheral neuropathy were evident in both 1992 and 1997, however, there were
very few veterans diagnosed with this condition.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. The methods used to estimate TCDD exposure are clearly described, and
capture exposure at an individual level prior to the diagnosis of the health outcome under study.
The authors describe the limitations of the exposure assessment within the paper.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration not satisfied. There were very few cases of peripheral neuropathy, particularly in
the most highly exposed groups. Statistical significance was only achieved in a few instances, and
in some cases, the odds ratios could not be estimated.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Neurotoxicology 2001: 22:479-490.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Serum was used to characterize TCDD exposure at end of tour of duty. Given
exposures dropped dramatically for the Ranch Hands following their tours of duty, exposure to
TCDD prior to disease onset is reasonably characterized, though some misclassification between
those in the comparison group and those in the lowest Ranch Hand exposure grouping is
inevitable.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. There is uncertainty in the critical window of exposure which impacts the
ability to calculate an effective TCDD over time. This study analyzes the potential for associations
setween point-in-time measures of TCDD serum levels and peripheral neuropathy that may have
occurred at any time over approximately a 30-year interval.

While an association was noted between peripheral neuropathy and TCDD levels, these
comparisons were limited by a small number of outcomes particularly within the highest exposure
group. Statistical significance was only achieved for some measures of peripheral neuropathy
using data from the 1997 examination, but not in the other 4 examination periods. Residual
confounding by undiagnosed diabetes may have distorted the measures of association, and this bias
cannot be fully dismissed. Additionally, there is uncertainty in the critical window of exposure
which precludes calculation of a cumulative TCDD effective dose over time. Multiple
comparisons arising from conducting statistical tests of significant over multiple time periods, and
measure of neuropathy raise the possibility of detecting a false-positive (spurious) association. For
these reasons, dose-response modeling was not conducted for this study.
Table C-52.  Pavuk et al. (2003) —Thyroid function and disorders
1. Consideration
Response
2. Consideration
Response
3. Consideration
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Thyroid diseases among veterans in the Air Force Health Study were
identified using questionnaire data collected in up to five examinations that were verified by a
review of medical records. Measures of thyroid function were also determined using serum
samples.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Exposure to TCDD was assessed using serum, and reasonably classified
veterans based on their exposure prior to disease onset. Appropriate methods were used to analyze
the data both longitudinally and cross-sectionally.
Study demonstrates an association between TCDD and adverse health effect.
                                C-209

-------
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Consideration not satisfied. There were no statistically significant associations between TCDD
and thyroid diseases. No associations were noted between serum-based measures of thyroid
function (T4, T3%, or FTI) and TCDD levels. While the data suggest a dose-response relationship
setween TCDD and TSH levels, the clinical implications are unclear. There were no statistically
significant increased risks of abnormal TSH levels among those in the highest exposure group
relative to the lowest for any of the five examination periods.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. The methods used to estimate TCDD exposure are clearly described, and
capture exposure at an individual level prior to the diagnosis of the health outcome under study.
The authors describe the limitations of the exposure assessment within the paper.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration not satisfied. There were 188 veterans who were diagnosed with a thyroid condition
following their tour of duty, and comparisons between 6 different thyroid diseases and four TCDD
exposure categories had poor statistical power. While there was a suggestion of increased TSH
abnormalities among Ranch Hand in the highest exposure group, these findings did not achieve
statistical significance for any of the 5 examination periods. Further follow-up of this cohort is
needed as the age distribution of the cohort may be too young to detect associations between
TCDD and thyroid function.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. The paper was published in Annals of Epidemiology 2003; 13:335-343. The
authors have discussed the strengths and limitations of the study.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Serum was used to characterize TCDD exposure as of 1987. Given exposures
dropped dramatically for the Ranch Hands following their tours of duty, exposure to TCDD prior
to disease onset is reasonably characterized. Serum-based measures of thyroid function were
obtained at multiple examinations which permitted dose-response relationships to be evaluated
30th cross sectionally and longitudinally.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. There is uncertainty in the critical window of exposure which impacts the
ability to calculate an effective TCDD over time. This study analyzes the potential for associations
setween point-in-time measures of TCDD serum levels and thyroid conditions and measures of
thyroid disorders that may have occurred at any time over approximately a 30-year interval.

While the health outcomes and TCDD exposures were characterized using valid methods, no
associations were observed between TCDD and any of the six thyroid conditions studied.
Additionally, no associations were noted with T4, FTI, or T3% in either cross-sectional or
longitudinal analyses. There is some support for a dose-response relationship between TCDD and
TSH, however, no statistically significant increase in abnormal TSH levels were observed among
those in the highest exposure group at any of the 5 examinations. Therefore, the clinical
implications of this dose-response relationship are unclear, particularly in light of the lack of
associations between TCDD and any of the thyroid disorders examined. Additionally, there is
uncertainty in the critical window of exposure, which precludes calculation of a cumulative TCDD
effective dose over time. For these reasons, dose-response modeling was not conducted for this
study.
Table C-53. Michalek and Pavuk (2008)—Diabetes
1. Consideration
Response
2. Consideration
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. Prevalent diabetes identified from medical records from repeated medical
check-ups. Preferred method of ascertaining outcome relative to use of death certificates.
Risk estimates are not susceptible to important biases.
                               C-210

-------
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Consideration satisfied. Adjustment was made for a number of risk factors related to diabetes
(e.g., BMI, family history, smoking) and other factors likely strongly associated with TCDD
exposure (e.g., last calendar year of service, occupation, etc.).
Study demonstrates an association between TCDD and adverse health effect.
Consideration satisfied. The RR for an increase in 10 units was 1.29 (p< 0.001), and the risks
across the background, low and high exposure categories, relative to the unexposed were 0.86,
1.45, and 1.68.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Initial TCDD dose were estimated at the end of the tour of duty for the
Ranch Hands. Individual-level serum dioxin measurements correlated well with correlated with
days of spraying and calendar period of service, but collection of the samples roughly 20 years
later required back-extrapolation.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. There were a total of 439 cases of diabetes identified.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. J Occup Environ Medicine, 2008, 50:330-340. The authors address strengths
and limitations related to the accuracy of the one-compartment pharmacokinetic model, impact of
the covariate time spent in Southeast Asia, and potential exposure misclassification on days
sprayed.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. TCDD estimates were derived using serum samples.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. The nature of the data did not allow for latency or critical windows of
exposure to be determined.

Because the nature of the data did not allow for the critical windows of exposure to be identified,
dose-response modeling was not conducted for this study.
C.3.8. Other Noncancer Studies of Dioxin
      Table C-54. McBride et al.
-Mortality (noncancer)
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. The New Zealand Health Information Service Mortality Collection and
the Registrar-General's Index to Deaths were used to identify deaths. Additional searches were
based on the last known address from the work record; the electoral roll and the habitation index;
the telephone book; the internet; and Terranet property information database. An additional search
was carried out through the Births, Deaths, and Marriages office of the New Zealand Department
of Internal Affairs. Lastly, automated personnel and pension records were also used to locate past
New Plymouth workers and identify some deaths.
Risk estimates are not susceptible to important biases.
Consideration satisfied. Workers lost to follow-up (21%) were an unlikely source of bias since
there was no evidence that this loss was differential in the internal analyses of workers.
Confounding by sex, hire year, and birth year was addressed by adjustment in regression models.
Potential confounding by other coexposures (e.g., 2,4,6-TCP) unlikely to have resulted in bias, due
to presumed poor correlation with TCDD.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. No associations were detected for mortality and the TCDD exposure
surrogates. No dose-response trend was observed across the exposure categories of TCDD.
Exposure assessment methodology is clear and characterizes individual-level exposures.
                                       C-211

-------
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Consideration satisfied. Cumulative exposure to TCDD as a time -dependent metric was estimated
for each worker from serum samples, but the authors did not examine a continuous measure of
TCDD exposure (lagged or unlagged).
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration not satisfied. Although the study had a large sample size (n=l,599 workers),
extensive follow-up period (35 years) and considerable exposure gradient, a limited number
noncancer deaths occurred. As such, mortality for some outcomes such as diabetes (based on 5
deaths) did not have adequate statistical power to examine potential associations. T he loss to
follow-up of 21% of workers was also substantial. This would have impacted statistical power by
reducing the number of deaths among the workers.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Published in J Occup Environ Med, 2009, 5 1 : 1049-1056. The other studies in
the cohort highlight the 21% of the cohort lost to follow-up and the potential for other exposures
during employment at the plant.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria satisfied. Serum measures available for 346 workers were used to derive TCDD exposures
for the entire cohort using the area under the curve approach.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. Effective dose could be estimated from serum-derived cumulative exposure
estimates. Also, noncancer mortality is not a viable endpoint to consider for further dose-response
analysis.

A considerable portion of the cohort was lost to follow-up, and no dose-response associations were
reported. In addition, since all outcomes were based on mortality, dose-response modeling was not
conducted for this study
Table C-55. McBride et al. (2009a)—Mortality (noncancer)
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
1 Consideration
Response
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration satisfied. The New Zealand Health Information Service Mortality Collection and
the Registrar-General's Index to Deaths were used to identify deaths. Additional searches were
based on the last known address from the work record; the electoral roll and the habitation index;
the telephone book; the internet; and Terranet property information database. An additional search
was carried out through the Births, Deaths, and Marriages office of the New Zealand Department
of Internal Affairs. Lastly, automated personnel and pension records were also used to locate past
New Plymouth workers and identify some deaths.
Risk estimates are not susceptible to important biases.
Consideration not satisfied. Considerable amount of workers were lost to follow up (22%), but it
is unclear if bias resulted. The dichotomous exposure measure was based on exposure to TCDD,
chlorinated dioxins and phenoxy herbicides, so confounding by these coexposures is possible.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. There was no associations detected for mortality and the TCDD
exposure surrogates. Because no individual exposure estimates were available for these analyses,
dose response could also not be evaluated.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Dichotomous exposure (exposed/unexposed) and duration of employment
were examined from job exposure classification assessed via occupational history records
industrial hygienists/factory personnel knowledge and questionnaires. Authors discuss limitations
in the assignment of exposure among cohort members.
                                C-212

-------
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. The size of the cohort is large enough to characterize mortality risks
relative to the general population for most common causes of deaths. A limitation of this study is
the loss to follow-up of a substantial percentage of workers (22%). This would have impacted
statistical power by reducing the number of deaths among the workers.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Published in Occup Medicine, 2009, 59(4):255-263. The authors highlight
cohort lost to follow-up, the limited size of the cohort, differences in cohort definitions between
sprayers and producers, and the potential for other exposures during employment at the plant.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria not satisfied. TCDD exposures were not quantified. The dichotomous exposure measure
was based on exposure surrogates of TCDD, chlorinated dioxins and phenoxy herbicides.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. Effective dose could not be estimated given the lack of individual-level
exposure data. Noncancer mortality is not a viable endpoint to consider for further dose-response
analysis.

The study lacks the quantification of exposures at an individual level, and a considerable portion of
the cohort was lost to follow-up. In addition, since all outcomes were based on mortality, dose-
response modeling was not conducted for this study.
Table C-56.  Ryan et al. (2002)—Sex ratio
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration not satisfied. Company records were used to identify births, the date of birth, and
the sex of the child. No information was provided on the expected completeness of identifying
sirths in this manner. Moreover, the study was expanded to include workers who heard about the
study in a public forum. Therefore, the study could be influenced by participation bias.
Risk estimates are not susceptible to important biases.
Consideration not satisfied. See above.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. The study compared birth ratios among men and women employed at
the plant to the general population. No categories of exposure were examined.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration not satisfied. This is not relevant as no analyses were done in relation to exposure
levels.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. For the categories of exposure used (yes/no), and the stratified analyses
sy sex and subcohort, the study allows for the birth ratios to be estimated with sufficient precision.

Study is published in the peer-reviewed scientific literature.
Criteria not satisfied. Published in Environ Health Perspect, 2002, 1 10(1 1):A699-A701. The
authors discussed the limitations of using serum collected many years after they stopped working
to estimate TCDD exposures when the preferred metric would be TCDD levels at the time of
conception. They did not address issues about the representativeness of the study participants to
the entire cohort of workers, nor did they address the limitation of not being able to conduct dose-
response analyses using individual-level TCDD data.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria not satisfied. While serum measures were available for 84 of the 198 participants of the
study, birth ratios were compared between the cohort of 2,4,5-T and 2,4,5-trichlorphgenol workers
relative to the city of Ufa. There was no attempt to derive birth ratios in relation to exposure
levels. The serum data were only used to demonstrate that these workers, on average, had TCDD
levels 30 times higher than Ufa residents.
                                C-213

-------
3.  Criteria
Effective exposure is estimable latency and window(s) of exposure are examined.
Response
Criteria not satisfied. TCDD exposures were based on serum measures taken in some cases many
years after children were born; no attempt was made to back-extrapolate to the time of conception.
Conclusion
Risk estimates have not been derived in relation to TCDD exposure levels. Uncertainties exist
about the representativeness of the participants in relation to the cohort as a whole, and
insufficient details are provided to evaluate the extent in which all births were identified. While
these data could not be used for quantitative dose-response modeling, the much lower male:female
sirth ratio among exposed fathers is consistent with the finding by Mocarelli et al, and lends
support to those findings. Dose-response modeling was not conducted for this study.	
       Table C-57.  Kang et al.
                          -Long term health consequences
1. Consideration
Response
2. Consideration
Response
3. Consideration
Response
4. Consideration
Response
5. Consideration
Response

1. Criteria
Response
2. Criteria
Response
3. Criteria
Response

Conclusion
Methods used to ascertain health outcomes are clearly identified and unbiased.
Consideration not satisfied. Data collected from only half of the individuals in the study target
>opulation, thus, there is some potential for selection bias in this study. The study excluded those
who had died before 1999, excluding potentially important TCDD-related adverse health effects
that could result in death more than two decades after veterans had been actively spraying. Survey
>articipation rates were modest: 72.9% for Vietnam veterans and 69.2% for non- Vietnam
veterans. If those in poorer health were less inclined to participate, the prevalence of the selected
chronic health conditions would be understated. The study relied on serf -reported measures of
disease prevalence increasing the possibility of recall bias.
Risk estimates are not susceptible to important biases.
Consideration not satisfied. See above.
Study demonstrates an association between TCDD and adverse health effect.
Consideration not satisfied. The data collected are cross-sectional, they are ill-suited for
svaluating the relationship between the timing of exposure and the onset of disease.
Exposure assessment methodology is clear and characterizes individual-level exposures.
Consideration satisfied. Serum TCDD levels were available for 897 subjects, although the entire
study population consisted of a group of 1,499 Vietnam veterans and a control group of 1,428
non- Vietnam veterans.
Study size and follow-up adequate to estimate risk and ensure sufficient statistical power.
Consideration satisfied. Size of study population likely provided sufficient study power to
observe effects.

Study is published in the peer-reviewed scientific literature.
Criteria satisfied. Published in Chemo sphere in 2001. The authors discussed the limitations of
using collected sera.
Exposure is primarily to TCDD and can be quantified to assess dose-response relationships.
Criteria not satisfied. While serum TCDD measures were available for some of the study
participants, there was no analysis of other contaminant exposures in the study population.
Effective exposure is estimable latency and window(s) of exposure are examined.
Criteria not satisfied. The critical exposure window could not be identified for the study.

A number of potential biases are present in this study. There is also potential confounding of
results from exposures to other contaminants that have not been evaluated in the population. The
critical exposure window cannot be determined. Dose-response modeling was not conducted for
this study.
                                              C-214

-------
C.4.   REFERENCES
Aittomaki, A; Lahelma, E: Roos, E: Leino-Arias, P; Martikainen, P. (2005). Gender differences
       in the association of age with physical workload and functioning. Br Med J 62: 95-100.
Akhmedkhanov, A; Revich, B: Adibi, JJ: Zeilert, V: Masten, SA: Patterson, DG, Jr: Needham,
       LL; Toniolo, P. (2002). Characterization of dioxin exposure in residents of Chapaevsk,
       Russia. J Expo Anal Environ Epidemiol 12: 409-417.
Akhtar, FZ: Garabrant DH: Ketchum, NS: Michalek, JE. (2004). Cancer in US Air Force
       veterans of the Vietnam war. J Occup Environ Med 46: 123.
Alaluusua, S; Calderara, P: Gerthoux, PM: Lukinmaa, PL; Kovero, O: Needham, L: Patterson, J,
       r, D. G.: Tuomisto, J: Mocarelli, P. (2004). Developmental dental aberrations after the
       dioxin accident in Seveso. Environ Health Perspect 112: 1313-1318.
Alavanja, MC: Samanic, C: Dosemeci, M: Lubin, J: Tarone, R: Lynch, CF: Knott C; Thomas,
       K: Hoppin, JA: Barken J: Coble. J: Sandier. DP: Blair. A. (2003). Use of agricultural
       pesticides and prostate cancer risk in the Agricultural Health Study cohort. Am J
       Epidemiol 157: 800-814. http://dx.doi.org/10.1093/aje/kwg040.
Alavanja, MC: Dosemeci, M: Samanic, C: Lubin, J: Lynch, CF: Knott C: Barker, J: Hoppin, JA:
       Sandier, DP; Coble, J: Thomas, K: Blair, A. (2004). Pesticides and lung cancer risk in the
       agricultural health study cohort. Am J Epidemiol 160: 876-885.
       http://dx.doi.org/10.1093/aje/kwh290.
Alavanja, MC: Sandier, DP; Lynch, CF: Knott C: Lubin, JH: Tarone, R: Thomas, K: Dosemeci,
       M: Barker, J: Hoppin, JA: Blair, A. (2005). Cancer incidence in the agricultural health
       study. Scand J Work Environ Health 31: 39-45.
Altekruse, SF: Kosary, CL: Krapcho, M: Neyman, N; Aminou, R: Waldron, W; Ruhl, J:
       Howlader, N; Tatalovich, Z: Cho, H: Mariotto, A; Eisner, MP: Lewis, PR: Cronin, K:
       Chen, HS: Feuer, EJ: Stinchcomb, DG: Edwards, BK. (2010). SEER cancer statistics
       review, 1975-2007. Bethesda, MD: National Cancer Institute.
       http://seer.cancer.gOv/csr/l 975_2007/.
Alvarez-Pedrerol, M: Ribas-Fito, N: Torrent, M: Carrizo, D; Garcia-Esteban, R: Grimalt, JO:
       Sunyer, J. (2008). Thyroid disruption at birth due to prenatal exposure to beta-
       hexachlorocyclohexane. Environ Int 34: 737-740.
Armstrong, BG. (1995). Comparing standardized mortality ratios. Ann Epidemiol 5: 60-64.
Aylward, LL: Brunei RC: Carrier, G: Hays, SM: Gushing, CA: Needham, LL: Patterson, DG:
       Gerthoux, PM: Brambilla, P: Mocarelli, P. (2005a). Concentration-dependent TCDD
       elimination kinetics in humans: Toxicokinetic modeling for moderately to highly exposed
       adults from Seveso, Italy, and Vienna, Austria, and impact on dose estimates for the
       NIOSH cohort. J Expo Anal Environ Epidemiol  15: 51-65.
       http://dx.doi.org/10.1038/sj.jea.7500370.
Aylward, LL: Brunei RC: Starr, TB: Carrier, G: Delzell, E: Cheng, H: Beall, C. (2005b).
       Exposure reconstruction for the TCDD-exposed NIOSH cohort using a concentration-
       and age-dependent model of elimination. Risk Anal 25: 945-956.
       http://dx.doi.0rg/10.llll/i.1539-6924.2005.00645.x.
Avlward, LL: Bodner, KM: Collins, JJ: Hays, SM. (2007). Exposure reconstruction for a dioxin-
       exposed cohort: Integration of serum sampling data and work histories. Organohalogen
       Compounds 69: 2063-2066.
                                        C-215

-------
Avlward. LL: Bodner. KM: Collins. JJ: Wilken. M: McBride. D: Burns. CJ: Hays. SM:
       Humphry, N. (2009). TCDD exposure estimation for workers at a New Zealand 2,4,5-T
       manufacturing facility based on serum sampling data. J Expo Sci Environ Epidemiol
       TEA: 1-10. http://dx.doi.org/10.1038/ies.2009.31.
Baccarelli, A; Mocarelli, P; Patterson, DG, Jr: Bonzini, M: Pesatori, AC: Caporaso, N: Landi,
       MT. (2002). Immunologic effects of dioxin: new results from Seveso and comparison
       with other studies. Environ Health Perspect 110: 1169-1173.
Baccarelli, A; Pesatori, AC: Masten, SA: Patterson, DG, Jr: Needham, LL: Mocarelli, P:
       Caporaso, NE: Consonni, D: Grassman, JA: Bertazzi, PA; MT, L. (2004). Aryl-
       hydrocarbon receptor-dependent pathway and toxic effects of TCDD in humans: a
       population-based study in Seveso, Italy. Toxicol Lett 149: 287-293.
       http://dx.doi.0rg/10.1016/i.toxlet.2003.12.062.
Baccarelli, A; Pesatori, AC: Consonni, D: Mocarelli, P: Patterson, DG, Jr: Caporaso, NE:
       Bertazzi, PA; Landi, MT. (2005). Health status and plasma dioxin levels in chloracne
       cases 20 years after the Seveso, Italy accident. Br J Dermatol 152: 459-465.
       http://dx.doi.org/10.1111/J.1365-2133.2005.06444.X.
Baccarelli, A; Hirt, C: Pesatori, AC: Consonni, D:  Patterson, DG, Jr: Bertazzi, PA; Dolken, G:
       Landi, MT. (2006). t(14;18) translocations  in lymphocytes of healthy dioxin-exposed
       individuals from Seveso, Italy. Carcinogenesis 27: 2001-2007.
       http://dx.doi.org/10.1093/carcin/bgl011.
Baccarelli, A; Giacomini, SM: Corbetta, C: Landi, MT: Bonzini, M: Consonni, D: Grillo, P:
       Patterson, DG: Pesatori, AC: Bertazzi, PA. (2008). Neonatal thyroid function in Seveso
       25 years after maternal exposure to dioxin.  PLoS Med 5: e!61.
Bang, KM; Kim, JH. (2001). Prevalence of cigarette smoking by occupation and industry in the
       United States. Am J Ind Med 40: 233-239.
Bates. MN: Buckland, SJ: Garrett N: Ellis. H: Needham. LL: Patterson. DG. Jr: Turner. WE:
       Russell,  DG. (2004). Persistent organochlorines in the serum of the non-occupationally
       exposed New Zealand population. Chemosphere 54: 1431-1443.
       http://dx.doi.0rg/10.1016/i.chemosphere.2003.09.040.
Becher, H: Steindorf, K: Flesch-Janys, D. (1998). Quantitative cancer risk assessment for dioxins
       using an occupational cohort. Environ Health Perspect 106: 663-670.
Bertazzi, A: Pesatori, AC: Consonni, D: Tironi, A: Landi, MT: Zocchetti, C. (1993). Cancer
       incidence in a population accidentally exposed to 2,3,7,8-tetrachlorodibenzo-para-dioxin.
       Epidemiology 4: 398-406.
Bertazzi, PA: Zocchetti, C: Pesatori, AC: Guercilena, S: Sanarico, M: Radice, L. (1989). Ten-
       year mortality study of the population involved in the Seveso incident in 1976. Am J
       Epidemiol 129: 1187-1200.
Bertazzi, PA: Zocchetti, C: Guercilena, S: Consonni, D: Tironi, A: Landi, MT: Pesatori, AC.
       (1997). Dioxin exposure and cancer risk: A 15-year mortality study after the "Seveso
       accident". Epidemiology 8: 646-652.
Bertazzi, PA: Consonni, D: Bachetti, S: Rubagotti, M: Andrea Baccarelli, A: Zocchetti, C: AC,
       R (2001). Health effects of dioxin exposure: a 20-year mortality study. Am J Epidemiol
       153:  1031-1044.
Bodner, K: Collins, J: Bloemen, L: Carson, M. (2003). Cancer risk for chemical workers exposed
       to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Occup Environ Med 60: 672-675.
       http://dx.doi.0rg/10.1136/oem.60.9.672.

                                         C-216

-------
Bond. GG: Wetterstroem. NH: Roush. GJ: McLaren. EA: Lipps. TE: Cook. RR. (1988). Cause
       specific mortality among employees engaged in the manufacture, formulation, or
       packaging of 2,4-dichlorophenoxyacetic acid and related salts. Occup Environ Med 45:
       98-105. http://dx.doi.0rg/10.1136/oem.45.2.98.
Bond, GG: McLaren, EA: Brenner, FE; Cook, RR. (1989).  Incidence of chloracne among
       chemical workers potentially exposed to chlorinated dioxins. J Occup Med 31: 771-774.
       http://dx.doi.org/10.1097/00043764-198909000-00017.
Bueno de Mesquita, HB; Doornbos, G: Van der Kuip, DA;  Kogevinas, M; Winkelmann, R.
       (1993). Occupational exposure to phenoxy herbicides and chlorophenols and cancer
       mortality in The Netherlands. Am J Ind Med 23: 289-300.
       http://dx.doi.org/10.1002/ajim.4700230206.
Cesana, GC: de Vito, G: Ferrario, M; Sega, R: Mocarelli, P. (1995). Trends of smoking habits in
       northern Italy (1986-1990): The WHO MONICA Project in Area Brianza, Italy. Eur J
       Epidemiol 11: 251-258.
Checkoway, H; Pearce, N; Crawford-Brown,  DJ. (1989). Research methods in occupational
       epidemiology. New York, NY: Oxford University Press.
Cheng. H: Aylward. L: Beall. C: Starr. TB: Brunei RC: Carrier. G: Delzell. E. (2006). TCDD
       exposure-response analysis and risk assessment. Risk Anal 26:  1059-1071.
       http://dx.doi.0rg/10.llll/i.1539-6924.2006.00800.x.
Chevrier, J: Eskenazi, B; Bradman, A; Fenster, L; Barr, DB. (2007). Associations between
       prenatal exposure to polychlorinated biphenyls and  neonatal thyroid-stimulating hormone
       levels in a Mexican-American population, Salinas Valley, California. Environ Health
       Perspect 115: 1490-1496.
Choi, BC. (1992). Definition, sources, magnitude, effect modifiers, and strategies of reduction of
       the healthy worker effect. J Occup Med 34: 979-988.
Clegg. LX: Li. FP: Hankev. BF: Chu. K: Edwards. BK. (2002). Cancer survival among US
       whites and minorities: a SEER (Surveillance, Epidemiology, and End Results) Program
       population-based study. Arch Intern Med 162: 1985-1993.
Cole, P; Trichopoulos, D: Pastides, H: Starr, T; Mandel, JS. (2003). Dioxin and cancer: A critical
       review [Review]. Regul Toxicol Pharmacol 38: 378-388.
Collins. JJ: Bodner. KM: Wilken. M: Haidar. S: Burns. CJ: Budinsky. RA: Martin. GD: Carson.
       ML; Rowlands, JC.  (2007). Serum concentrations of chlorinated dibenzo-p-dioxins and
       dibenzofurans among former Michigan trichlorophenol and pentachlorophenol workers. J
       Expo Sci Environ Epidemiol 17: 541-548. http://dx.doi.org/10.1038/si.ies.7500558.
Collins, JJ: Bodner, K; Aylward, LL: Wilken, M: Bodnar, CM. (2009). Mortality rates among
       trichlorophenol workers with exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Am J
       Epidemiol 170: 501-506. http://dx.doi.org/10.1093/aje/kwpl53.
Collins. JJ: Bodner. K: Aylward. LL. (2010).  Three Authors Reply. Am J Epidemiol  171: 130-
       131. http://dx.doi.org/10.1093/aie/kwp381.
Committee to Review the Health Effects in Vietnam Veterans of Exposure to Herbicides
       (Institute of Medicine Committee to Review the Health Effects in Vietnam Veterans of
       Exposure to Herbicides). (1994). Veterans and agent orange: Health effects of herbicides
       used in Vietnam. Washington, DC: The National Academies Press.
       http://www.nap.edu/catalog.php7record  id=2141.
                                        C-217

-------
Committee to Review the Health Effects in Vietnam Veterans of Exposure to Herbicides
       (Institute of Medicine Committee to Review the Health Effects in Vietnam Veterans of
       Exposure to Herbicides). (2006). Veterans and Agent Orange: update 2000 (7 ed.).
       Washington, DC: National Academies Press.
Consonni, D; Pesatori, AC: Zocchetti, C: Sindaco, R: D'Oro, LC: Rubagotti, M; Bertazzi, PA.
       (2008). Mortality in a population exposed to dioxin after the Seveso, Italy, accident in
       1976: 25 years of follow-up. Am J Epidemiol 167: 847-858.
       http://dx.doi.org/10.1093/aie/kwm371.
Cook, RR. (1981). Dioxin, chloracne, and soft tissue sarcoma. Lancet 317: 618-619.
       http://dx.doi.org/10.1016/S0140-6736(81)92070-5.
Cooper. GS: Klebanoff, MA: Promislow. J: Brock. JW: Longnecker. MP. (2005).
       Polychlorinated biphenyls and menstrual cycle characteristics. Epidemiology 16: 191-
       200. http://dx.doi.org/10.1097/01.ede.0000152913.12393.86.
D'Amico, M: Agozzino, E; Biagino, A; Simonetti, A; Marinelli, P. (1999). Ill-defined and
       multiple causes on death certificates-a study of misclassification in mortality statistics.
       Eur J Epidemiol 15: 141-148.
Dunson, DB; Baird,  DP. (2001). A flexible parametric model for combining current status and
       age at first diagnosis data. Biometrics 57: 396-403.
Emond, C: Michalek, JE; Birnbaum, LS: DeVito, MJ.  (2005). Comparison of the use of a
       physiologically based pharmacokinetic model and a classical pharmacokinetic model for
       dioxin exposure assessments. Environ Health Perspect 113: 1666-1668.
Eskenazi, B; Mocarelli, P; Warner, M: Samuels, S: Vercellini, P; Olive, D: Needham, L;
       Patterson. D: Brambilla, P. (2000). Seveso Women's Health Study: A study of the effects
       of 2,3,7,8-tetrachlorodibenzo-p-dioxin on reproductive health. Chemosphere 40: 1247-
       1253.
Eskenazi, B; Mocarelli, P: Warner, M: Samuels, S: Vercellini, P: Olive, D: Needham, LL:
       Patterson DG, J, r; Brambilla, P: Gavoni, N; Casalini, S: Panazza, S: Turner, W;
       Gerthoux, PM. (2002a). Serum dioxin concentrations and endometriosis: A cohort study
       in Seveso, Italy. Environ Health Perspect 110:  629-634.
Eskenazi, B: Warner, M: Mocarelli, P: Samuels, S: Needham, LL: Patterson, DG, Jr: Lippman,
       S: Vercellini. P: Gerthoux. PM: Brambilla. P: Olive. D. (2002b). Serum dioxin
       concentrations and menstrual cycle characteristics. Am J Epidemiol 156: 383-392.
Eskenazi, B: Mocarelli, P: Warner, M: Chee, WY: Gerthoux, PM: Samuels, S: Needham, LL:
       Patterson, DG, Jr. (2003). Maternal serum dioxin levels and birth outcomes in women of
       Seveso, Italy. Environ Health Perspect 111: 947-953.
Eskenazi, B: Mocarelli, P: Warner, M: Needham, L: Patterson, DG, Jr: Samuels, S: Turner, W:
       Gerthoux, PM: Brambilla, P. (2004). Relationship of serum TCDD concentrations and
       age at exposure of female residents of Seveso,  Italy. Environ Health Perspect 112: 22-27.
       http://dx.doi.org/10.1289/ehp.6573.
Eskenazi, B: Warner, M: Marks, AR: Samuels, S: Gerthoux, PM: Vercellini, P: Olive, PL;
       Needham, L: Patterson, D, Jr: Mocarelli, P. (2005). Serum dioxin concentrations and age
       at menopause. Environ Health Perspect 113: 858-862.
Eskenazi, B: Warner, M: Samuels,  S: Young, J: Gerthoux, PM: Needham, L: Patterson, D: Olive,
       D: Gavoni, N: Vercellini, P: Mocarelli, P. (2007). Serum dioxin concentrations and  risk
       of uterine leiomyoma in the Seveso Women's Health Study. Am J Epidemiol 166: 79-87.
       http://dx.doi.org/10.1093/aie/kwm048.

                                         C-218

-------
Fingerhut MA; Halperin, WE; Marlow, DA; Piacitelli, LA; Honchar, PA; Sweeney, MH: Greife,
       AL; Dill, PA; Steenland, K; Suruda, AJ. (199la). Cancer mortality in workers exposed to
       2,3,7,8-tetrachlorodibenzo-p-dioxin. N Engl J Med 324: 212-218.
Fingerhut MA; Halperin, WE; Marlow, DA; Piacitelli, LA; Honchar, PA; Sweeney, MH: Greife,
       AL: Dill PA: Steenland. K: Suruda. AJ. (1991b). Mortality of U.S. workers employed in
       the production of chemicale contaminated with 2,3,7,8-tetrachlorodibenzo-p-dioxin
       (TCDD). Cincinnati, OH: U.S. Department of Health and Human Services.
Flesch-Janys, D: Gurn, P: Jung, D; Konietzko, J: Manz, A; Papke, O. (1994). First results of an
       investigation of the elimination of polychlorinated dibenzo-p-dioxins and dibenzofurans
       (PCDD/F) in occupationally exposed persons. Organohalogen Compounds 21: 93-99.
Flesch-Janvs, D: Berger, J: Gurn. P: Manz. A: Nagel, S: Waltsgott, H: Dwver, JH. (1995).
       Exposure to polychlorinated dioxins and furans (PCDD/F) and mortality in a cohort of
       workers from a herbicide-producing plant in Hamburg, Federal Republic of Germany.
       Am JEpidemiol 142: 1165-1175.
Flesch-Janvs, D: Becher, H:  Gurn. P:  Jung. D: Konietzko. J: Manz. A: Papke. O. (1996).
       Elimination of polychlorinated dibenzo-p-dioxins and dibenzofurans in occupationally
       exposed persons. J Toxicol Environ Health 47: 363-378.
Flesch-Janvs, D. (1997). Analyses of  exposure to polychlorinated dibenzo-p-dioxins, furans, and
       hexachlorocyclohexane and different health outcomes in a cohort of former herbicide-
       producing workers in Hamburg, Germany. Teratog Carcinog Mutagen 17: 257-264.
       http://dx.doi.org/10.1002/(SICni520-6866(1997)17:4/5<257::AID-TCM8>3.0.CO:2-H.
Flesch-Janvs, D: Steindorf, K: Gurn, P: Becher, H. (1998). Estimation of the cumulated exposure
       to polychlorinated dibenzo-p-dioxins/furans and standardized mortality ratio analysis of
       cancer mortality by dose in an occupationally exposed cohort. Environ Health Perspect
       106: 655-662.
Gever, HJ:  Scheunteit I: Rapp, K: Kettrup, A: Korte, F: Greim, H: Rozman, K. (1990).
       Correlation between  acute toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and
       total body fat content in mammals. Toxicology 65: 97-107.
Grubbs, WD: Lustik, MB: Brockman, AS: Henderson. SC: Burnett FR: Land. RG: Osborne, DJ:
       Rocconi, VK: Schrieber, ME:  Williams. DE: Wolfe. WH: Michalek, JE: Miner. JC:
       Henriksen, GL: Swaby, JA. (1995). The Air Force Health Study: An epidemiologic
       investigation of health effects  in Air Force personnel following exposure to herbicides.
       1992 Follow-up examination results. Washington, DC: United States Air Force.
Guess, HA; Hoel, DG.  (1977). The effect of dose on cancer latency period. J Environ Pathol
       Toxicol 1:279-286.
Henriksen,  GL: Ketchum, NS: Michalek, J: Swaby, JA. (1997). Serum dioxin and diabetes
       mellitus in veterans of Operation Ranch Hand. Epidemiology 8: 252-258.
Hoglund, M; Sehn, L; Connors, JM; Gascoyne, RD; Siebert, R; Sail, T; Mitelman, F; Horsman,
       DE. (2004). Identification of cytogenetic subgroups and karyotypic pathways of clonal
       evolution in follicular lymphomas. Genes Chromosomes Cancer 39:  195-204.
       http://dx.doi.org/10.1002/gcc.10314.
Hooiveld, M; Heederik, DJ:  Kogevinas, M; Boffetta, P: Needham, LL; Patterson,  DG, Jr; Bueno-
       de-Mesquita, HB. (1998). Second follow-up of a Dutch cohort occupationally exposed to
       phenoxy herbicides, chlorophenols, and contaminants. Am J Epidemiol  147: 891-901.
IARC (International Agency for Research on Cancer). (1997). Polychlorinated dibenzo-para-
       dioxins and polychlorinated dibenzofurans. Lyon, France.

                                        C-219

-------
James. WH. (1995). What stabilizes the sex ratio? Ann Hum Genet 59: 243-249.
      http://dx.doi.0rg/10.llll/i.1469-1809.1995.tb00744.x.
J0rgensen, N; Andersen, AG: Eustache, F; Irvine, DS: Suominen, J: Petersen, JH; Andersen, AN:
      Auger, J: Cawood, EH; Horte, A; Jensen, TK; Jouannet P; Keiding, N; Vierula, M;
      Toppari, J: Skakkebaek, NE. (2001). Regional differences in semen quality in Europe.
      HumReprod 16: 1012-1019.
Kang, HK; Dalager, NA: Needham, LL: Patterson Jr, DG: Matanoski, GM: Kanchanaraksa, S;
      Lees, PSJ. (2001). US army chemical corps Vietnam veterans health study: Preliminary
      results. Chemosphere 43: 943-949. http://dx.doi.org/10.1016/S0045-6535(00)00455-0.
Kang. HK: Dalager. NA: Needham. LL: Patterson. DG. Jr: Lees. PS: Yates, K: Matanoski. GM.
      (2006). Health status of Army Chemical Corps Vietnam veterans who sprayed defoliant
      in Vietnam. Am J Ind Med 49: 875-884. http://dx.doi.org/10.1002/aiim.20385.
Kauppinen, T; Kogevinas, M; Johnson, E: Becher, H: Bertazzi, PA; Bueno de Mesquita, HB:
      Coggon, D: Green, L; Littorin, M: Lynge E Mathews, J: Neuberger, M: Osman, J:
      Pannett B: Pearce, N: Winkelmann, R: Saracci, R. (1993). Chemical exposure in
      manufacture of phenoxy herbicides and chlorophenols and in spraying of phenoxy
      herbicides. Am J Ind Med 23:  903-920.
Kerger. BD: Leung. HW: Scott. P: Paustenbach, DJ: Needham. LL: Patterson. DG. Jr: Gerthoux,
      PM; Mocarelli, P. (2006). Age- and concentration-dependent elimination half-life of
      2,3,7,8-tetrachlorodibenzo-p-dioxin in Seveso children. Environ Health Perspect 114:
      1596-1602. http://dx.doi.org/10.1289/ehp.8884.
Kogevinas, M: Becher, H; Benn, T; Bertazzi, PA; Boffetta, P: Bueno-de-Mesquita, HB: Coggon,
      D: Colin, D; Flesch-Janys, D: Fingerhut M: Green, L; Kauppinen, T; LJttorin, M: Lynge,
      E: Mathews, JD: Neuberger, M: Pearce, N: R, S. (1997). Cancer mortality in workers
      exposed to phenoxy herbicides, chlorophenols, and dioxins an expanded and updated
      international cohort study. Am J Epidemiol 145: 1061-1075.
Kreuzer, PE: Csanadv, GA: Baur, C: Kessler, W: Papke, O: Greim, H: Filser, JG. (1997).
      2,3,7,8-Tetrachlorodibenzo-p -dioxin (TCDD) and congeners in infants. A toxicokinetic
      model of human lifetime body burden by TCDD with special emphasis on its uptake by
      nutrition. Arch Toxicol 71: 383-400.
Landi, MT; Consonni, D: Patterson, DG, Jr: Needham, LL: Lucier, G: Brambilla, P: Cazzaniga,
      MA: Mocarelli, P: Pesatori, AC: Bertazzi, PA: NE, C. (1998).  2,3,7,8-
      Tetrachlorodibenzo-p-dioxin plasma levels in Seveso 20 years after the accident. Environ
      Health Perspect 106: 273-277.
Landi, MT: Bertazzi, PA; Baccarelli, A; Consonni, D: Masten, S: Lucier, G: Mocarelli, P:
      Needham, L: Caporaso, N: Grassman, J. (2003). TCDD-mediated alterations in the AhR-
      dependent pathway in Seveso, Italy, 20 years after the accident. Carcinogenesis 24: 673-
      680. http://dx.doi.org/10.1093/carcin/bgg002.
Lee, DJ: Fleming, LE: Arheart, KL: LeBlanc, WG: Caban, AJ: Chung-Bridges, K: Christ,  SL:
      McCollister, KE; Pitman, T. (2007). Smoking rate trends in U.S. occupational groups:
      The 1987 to 2004 National Health Interview Survey.  J Occup Environ Med 49: 75-81.
      http://dx.doi.org/10.1097/JOM.Ob013e31802ec68c.
Li, CY; Sung, FC. (1999). A review of the healthy worker effect in occupational epidemiology.
      Occup Med (Lond) 49:  225-229.
                                        C-220

-------
Longnecker. MP: Gladen. BC: Patterson. DG. Jr: Rogan. WJ. (2000). Polychlorinated biphenyl
       (PCB) exposure in relation to thyroid hormone levels in neonates. Epidemiology 11: 249-
       254.
Longnecker, MP: Michalek, JE. (2000). Serum Dioxin level in relation to Diabetes Mellitus
       among Air Force veterans with background levels of exposure. Epidemiology 11: 44-48.
Lorber, M; Phillips, L. (2002). Infant exposure to dioxin-like compounds in breast milk. Environ
       Health Perspect 110: A325-A332.
Manz. A: Bergen J: Dwver. JH: Flesch-Janys. D: Nagel.  S: Waltsgott H. (1991). Cancer
       mortality among workers in chemical plant contaminated with dioxin. Lancet 338: 959-
       964. http://dx.doi.org/10.1016/0140-6736(91)91835-I.
McBride. PI: Burns. CJ: Herbison. GP: Humphry. NF: Bodner. K: Collins. JJ. (2009a). Mortality
       in employees at a New Zealand agrochemical manufacturing site. Occup Med (Lond) 59:
       255-263. http://dx.doi.org/10.1093/occmed/kqp030.
McBride. PI: Collins. JJ: Humphry. NF: Herbison. P: Bodner. KM: Aylward. LL: Burns. CJ:
       Wilken, M. (2009b). Mortality in workers exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin
       at a trichlorophenol plant in New Zealand. J Occup Med 51: 1049-1056.
       http://dx.doi.org/10.1097/JOM.Ob013e3181b571ae.
McEwen, LN: Kim, C; Haan, M: Ghosh, D; Lantz, PM; Mangione, CM; Safford, MM; Marrero,
       D: Thompson, TJ: Herman, WH. (2006). Diabetes reporting as a cause of death: Results
       from the Translating Research Into Action for Diabetes (TRIAD) study. Diabetes Care
       29: 247-253.
McMichael  AJ. (1976). Standardized mortality ratios and the "healthy worker effect":
       Scratching beneath the surface. J Occup Med 18:  165-168.
Michalek. JE: Pirkle. JL: Caudill. SP: Tripathl RC: Patterson. DG. Jr: Needham. LL. (1996).
       Pharmacokinetics of TCDD in veterans of Operation Ranch Hand: 10-year follow-up. J
       Toxicol Environ Health 47: 209-220.
Michalek, JE: Ketchum, NS: Akhtar, FZ. (1998). Postservice mortality of US Air Force veterans
       occupationally exposed to herbicides in Vietnam: 15-year follow-up. Am J Epidemiol
       148:  786-792.
Michalek. JE: Akhtar. FZ: Longnecker. MP: Burton. JE.  (200la). Relation of serum 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) level to hematological examination results in
       veterans of Operation Ranch Hand. Arch Environ Health 56: 396-405.
       http://dx.doi.org/10.1080/00039890109604474.
Michalek, JE: Ketchum, NS: Longnecker, MP. (200Ib). Serum dioxin and hepatic abnormalities
       in veterans of Operation Ranch Hand. Ann Epidemiol 11: 304-311.
       http://dx.doi.org/10.1016/S1047-2797(00)00218-0.
Michalek. JE: Akhtar. FZ: Arezzo, JC: Garabrant DH: Albers, JW. (200Ic). Serum dioxin and
       peripheral neuropathy in veterans of Operation Ranch Hand. Neurotoxicology 22: 479-
       490.
Michalek. JE: Pirkle, JL: Needham, LL: Patterson. DG, Jr: Caudill SP:  Tripathl RC: Mocarelli,
       F. (2002). Pharmacokinetics of 2,3,7,8-tetrachlorodibenzo-p-dioxin in Seveso adults and
       veterans of operation Ranch Hand. J Expo Anal Environ Epidemiol 12: 44-53.
Michalek. JE: Ketchum. NS: Tripathl RC. (2003). Diabetes mellitus and 2,3,7,8-
       tetrachlorodibenzo-p-dioxin elimination in veterans of Operation Ranch Hand. J Toxicol
       Environ Health A 66: 211-221. http://dx.doi.org/10.1080/15287390306373.
                                        C-221

-------
Michalek, JE: Pavuk, M. (2008). Diabetes and cancer in veterans of Operation Ranch Hand after
       adjustment for calendar period, days of spraying, and time spent in Southeast Asia. J
       Occup Environ Med 50: 330-340. http://dx.doi.org/10.1097/JOM.Ob013e31815f889b.
Mocarelli, P: Brambilla, P: Gerthoux, PM: Patterson Jr. DG: Needham, LL. (1996). Change in
       sex ratio with exposure to dioxin. Lancet 348: 409.
Mocarelli, P: Gerthoux, PM: Ferrari, E: Patterson Jr, DG: Kieszak, SM: Brambilla, P: Vincoli,
       N; Signorini, S; Tramacere, P: Carreri, V: Sampson, EJ: Turner, WE. (2000). Paternal
       concentrations of dioxin and sex ratio of offspring. Lancet 355: 1858-1863.
       http://dx.doi.org/10.1016/S0140-6736(00)02290-X.
Mocarelli, P: Gerthoux, PM: Patterson, DG, Jr: Milani, S: Limonata, G: Bertona, M: Signorini,
       S: Tramacere, P: Colombo, L: Crespi, C: Brambilla, P: Sarto, C: Carreri, V: Sampson,
       EJ: Turner, WE; Needham, LL.  (2008). Dioxin exposure, from infancy through puberty,
       produces endocrine disruption and affects human semen quality. Environ Health Perspect
       116: 70-77. http://dx.doi.org/10.1289/ehp.10399.
Monson, RR. (1986). Observations on the healthy worker effect. J Occup Med 28: 425-433.
Muller, A; De La Rochebrochard, E: Labbe-Decleves, C: Jouannet, P: Bujan, L: Mieusset, R: Le
       Lannou, D: Guerin, JF: Benchaib, M: Slama, R: Spira, A. (2004).  Selection bias in semen
       studies due to self-selection of volunteers. Hum Reprod 19: 2838-2844.
       http://dx.doi.org/10.1093/humrep/deh521.
Myers, JE: Thompson, ML. (1998). Meta-analysis and occupational epidemiology. Occup Med
       (Lond) 48: 99-101.
Navarro, C: Chirlaque, MD: Tormo, MJ: Perez-Flores, D: Rodriguez-Barranco, M: Sanchez-
       Villegas, A; Agudo, A; Pera, G: Amiano, P: Dorronsoro, M: Larrafiaga, N: Quiros, JR:
       Ardanaz, E: Barricarte, A; Martinez, C: Sanchez, MJ: Berenguer, A; Gonzalez, CA.
       (2006). Validity of self reported diagnoses of cancer in a major Spanish prospective
       cohort study. J Epidemiol Community Health 60: 593-599.
       http://dx.doi.org/10.1136/iech.2005.039131.
Needham, LL: Gerthoux, PM: Patterson, J, r, D. G.: Brambilla, P: Prikle,  JL: Tramacere, PL;
       Turner, WE: c, B: Sampson, EJ: P, M. (1994). Half-life of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin in serum of Seveso adults: interim report. Organohalogen Compounds 21: 81-85.
Needham, LL: Gerthoux, PM: Patterson, DG: Brambilla, P: Turner, WE: Beretta, C: Pirkle, JL:
       Colombo, L: Sampson, EJ: Tramacere, PL; Signorini, S: Meazza, L: Carreri, V: Jackson,
       RJ: Mocarelli, P. (1997). Serum dioxin levels in Seveso, Italy, population in 1976.
       Teratog Carcinog Mutagen 17: 225-240. http://dx.doi.org/10.1002/(SICI) 1520-
       6866(1997)17:4/5<225::AID-TCM5>3.0.CO:2-K.
Needham, LL: Barr, DB: Caudill, SP: Pirkle, JL: Turner, WE: Osterloh, J: Jones, RL: Sampson,
       EJ. (2005). Concentrations of environmental chemicals associated with
       neurodevelopmental effects in the US population. Neurotoxicology 26: 531-545.
       http://dx.doi.0rg/10.1016/j.neuro.2004.09.005.
NTP (National Toxicology Program). (2006). NTP technical report on the toxicology and
       carcinogenesis studies of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (CAS no. 1746-
       01-6) in female harlan Sprague-Dawley rats (gavage studies). (NTP TR 521; NIH
       Publication No. 06-4468). Research Triangle Park, NC.
                                        C-222

-------
Okura, Y; Urban, LH; Mahoney, DW: Jacobsen, SJ: Rodeheffer, RJ. (2004). Agreement between
       self-report questionnaires and medical record data was substantial for diabetes,
       hypertension, myocardial infarction and stroke but not for heart failure. J Clin Epidemiol
       57: 1096-1103. http://dx.doi.0rg/10.1016/i.iclinepi.2004.04.005.
Ott MG: Olson. RA: Cook. RR: Bond. GG. (1987). Cohort mortality study of chemical workers
       with potential exposure to the higher chlorinated dioxins. J Occup Environ Med 29: 422-
       429.
Ott MG: Messerer, P; Zober, A. (1993). Assessment of past occupational exposure to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin using blood lipid analyses. Int Arch Occup Environ Health
       65: 1-8.
Ott, MG: Zober, A. (1996a). Cause specific mortality and cancer incidence among employees
       exposed to 2,3,7,8-TCDD after a 1953 reactor accident. Occup Environ Med 53: 606-
       612. http://dx.doi.0rg/10.1136/oem.53.9.606.
Ott, MG: Zober, A. (1996b). Morbidity study of extruder personnel with potential exposure to
       brominated dioxins and furans. II. Results of clinical laboratory studies.  Occup Environ
       Med 53: 844-846.
Papke. O: Ball M: Lis. A. (1994). PCDD/PCDF in humans, a 1993-update of background data.
       Chemosphere 29: 2355-2360.
Pavuk. M: Schecter. AJ:  Akhtar. FZ: Michalek. JE. (2003).  Serum 2,3,7,8-tetrachlorodibenzo-p-
       dioxin (TCDD) levels and thyroid function in Air Force veterans of the Vietnam War.
       Ann Epidemiol 13:  335-343. http://dx.doi.org/10.1016/51047-2797(02)00422-2.
Percy, C: Stanek, E, III; Gloeckler, L.  (1981). Accuracy of cancer death certificates and its effect
       on cancer mortality statistics. Am J Public Health 71: 242-250.
       http://dx.doi.Org/10.2105/AJPH.71.3.242.
Pesatori, AC: Zocchetti, C: Guercilena, S: Consonni, D; Turrini, D; Bertazzi, PA. (1998). Dioxin
       exposure and non-malignant health effects: A mortality study. Occup Environ Med 55:
       126-131. http://dx.doi.0rg/10.1136/oem.55.2.126.
Pesatori, AC: Consonni, D: Bachetti, S: Zocchetti, C: Bonzini, M: Baccarelli, A; Bertazzi, PA.
       (2003). Short- and long-term morbidity and mortality in the population exposed to dioxin
       after the "Seveso accident". Ind Health 41: 127-138.
Piacitelli, LA; Sweeney,  MH; Fingerhut, MA; Patterson, DG: Turner, WE;  Connally, LB; Wille,
       KK: Tompkins, B. (1992). Serum levels of PCDDS  and PCDFS among workers exposed
       to 2,3,7,8-TCDD contaminated chemicals. Chemosphere 25: 251-254.
Pirkle, JL: Wolfe. WH: Patterson. DG: Needham, LL: Michalek, JE: Miner. JC: Peterson. MR:
       Phillips, PL. (1989). Estimates of the half-life of 2,3,7,8-tetrachlorodibenzo-p-dioxin in
       Vietnam Veterans of Operation Ranch Hand.  J Toxicol Environ Health 27: 165-171.
Puga, A: Barnes. SJ: Dalton, TP: Chang. C: Knudsen, ES: Maier, MA. (2000). Aromatic
       hydrocarbon receptor interaction with the retinoblastoma protein potentiates repression of
       E2F-dependent transcription and cell  cycle arrest. J Biol Chem 275: 2943-2950.
       http://dx.doi.0rg/10.1074/ibc.275.4.2943.
Revich, B; Aksel, E; Ushakova, T; Ivanova, I: Zhuchenko, N; Klyuev, N; Brodsky, B; Sotskov,
       Y^ (2001). Dioxin exposure and public health in Chapaevsk, Russia. Chemosphere 43:
       951-966. http://dx.doi.org/10.1016/S0045-6535(W)00456-2.
Revich, B: Sergeyev, O: Zeilert, V: Hauser, R. (2005).  Chapaevsk, Russia:  40 years of dioxins
       exposure on the human health and 10 years of Russian ?USA epidemiological studies.
       Almaty, Kazakhstan: Institute for Health and the Environment.

                                         C-223

-------
Rigon, F; Bianchin, L; Bernasconi, S: Bona, G: Bozzola, M; Buzi, F; Cicognani, A; De Sanctis,
       C: De Sanctis, V: Radetti, G: Tato, L; Tonini, G: Perissinotto, E. (2010). Update on age
       at menarche in Italy: Toward the leveling off of the secular trend. J Adolesc Health 46:
       238-244. http://dx.doi.0rg/10.1016/i.iadohealth.2009.07.009.
Rothman, KJ. (1986). Modern epidemiology. Boston, MA: Little Brown & Co.
Ryan, JJ: Schecter, A.  (2000). Exposure of Russian phenoxy herbicide producers to dioxin. J
       Occup Environ Med 42: 861-870.
Ryan, JJ: Amirova, Z;  Carrier, G. (2002). Sex Ratios of Children of Russian Pesticide Producers
       Exposed to Dioxin. Environ Health Perspect 110: A699-A701.
Safe, SH. (1986). Comparative toxicology and mechanism of action of polychlorinated dibenzo-
       p-dioxins and dibenzofurans. Annu Rev Pharmacol Toxicol 26: 371-399.
       http://dx.doi.org/10.1146/annurev.pa.26.040186.002103.
Sandau, CD: Ayotte. P: Dewailly, E: Duffe. J: Norstrom, RJ. (2002). Pentachlorophenol and
       hydroxylated polychlorinated biphenyl metabolites in umbilical cord plasma of neonates
       from coastal populations in Quebec. Environ Health Perspect 110: 411-417.
Saracci, R: Kogevinas, M; Bertazzi, PA; Bueno de Mesquita, BH; Coggon, D; Green, LM;
       Kauppinen, T; L'Abbe, KA; Littorin, M; Lynge, E: Mathews, JD: Neuberger, M; Osman,
       J: Pearce, N; Winkelmann, R. (1991). Cancer mortality in workers  exposed to
       chlorophenoxy herbicides and chlorophenols. Lancet 338():: 1027-1032.
Siemiatycki, J: Wacholder, S: Dewar, R: Cardis, E: Greenwood, C: Richardson, L. (1988).
       Degree of confounding bias related to smoking, ethnic  group, and socioeconomic status
       in estimates of the associations between occupation and cancer. J Occup Med 30: 617-
       625.
Slade. BA. (1998). Dioxin and diabetes mellitus. Epidemiology 9: 359-360.
Smith, AH; Fisher, DO: Pearce, N: Chapman, CJ. (1982). Congenital defects and miscarriages
       among New Zealand 2, 4, 5-T sprayers. Arch Environ Health 37: 197-200.
Smith, AH; Lopipero, P.  (2001). Invited commentary:  How do the Seveso  findings affect
       conclusions concerning TCDD as a human carcinogen? Am J Epidemiol 153: 1045-1047.
       http://dx.doi.org/10.1093/aie/153.ll.1045.
Starr, TB. (2003). Significant issues raised by meta-analyses of cancer mortality and dioxin
       exposure. Environ Health Perspect 111: 1443-1447.
Stavner, L: Steenland,  K; Dosemeci, M: Hertz-Picciotto, I. (2003). Attenuation of exposure-
       response curves in occupational cohort studies at high exposure levels. Scand J Work
       Environ Health 29:  317-324.
Steenland, K; Piacitelli, L: Deddens, J: Fingerhut M: Chang, LI. (1999). Cancer, heart disease,
       and diabetes in workers exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin. J Natl Cancer
       Inst91: 779-786.
Steenland, K: Calvert,  G: Ketchum, N: Michalek, J.  (2001a). Dioxin and diabetes mellitus: an
       analysis of the combined NIOSH and Ranch Hand data. Occup Environ Med 58: 641-
       648. http://dx.doi.org/10.1136/oem.58.10.641.
Steenland, K: Deddens, J: Piacitelli, L. (2001b). Risk assessment for 2,3,7,8-tetrachlorodibenzo-
       p-dioxin (TCDD) based  on an epidemiologic study. Am J Epidemiol 154: 451-458.
Steenland, K: Deddens, J. (2003). Dioxin: Exposure-response analyses and risk assessment. Ind
       Health 41: 175-180.
                                         C-224

-------
Stellman, SD: Stellman, JM. (1986). Estimation of exposure to Agent Orange and other
       defoliants among American troops in Vietnam: a methodological approach. Am J Ind
       Med9: 305-321.
f Mannetie, A: McLean.  D: Cheng. S: Boffetta. P: Colin. D: Pearce, N. (2005). Mortality in New
       Zealand workers exposed to phenoxy herbicides and dioxins. Occup Environ Med 62: 34-
       40. http://dx.doi.org/10.1136/oem.2004.015776.
Thiess, AM; Frentzel-Beyme, R. (1977). Mortality study of persons exposed to dioxin following
       an accident which occurred in the BASF on 17 November 1953.5th International
       Conference Medichem, 1977San Francisco, CA.
Thiess, AM; Frentzel-Beyme, R: Link, R. (1982). Mortality study of persons exposed to dioxin
       in a trichlorophenol-process accident that occurred in the BASF AG on November 17,
       1953. Am J Ind Med 3: 179-189.
U.S. EPA (U.S. Environmental Protection Agency). (2003). Exposure and human health
       reassessment of 2,3,7,8 tetrachlorodibenzo-p dioxin  (TCDD) and related compounds
       [NAS review draft]. (EPA/600/P-00/001). Washington, DC: U.S. Environmental
       Protection Agency, National Center for Environmental Assessment.
       http://www.epa.gov/nceawwwl/pdfs/dioxin/nas-review/.
Villeneuve, PJ: Steenland, K. (2010). RE: "Mortality rates among trichlorophenol workers with
       exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin". Am JEpidemiol 171: 129-130.
       http://dx.doi.org/10.1093/aie/kwp380.
Ware. JH: Spengler, JD: Neas, LM: Samet JM: Wagner. GR: Coultas, D: Ozkavnak, H: Schwab.
       M. (1993). Respiratory and irritant health effects of ambient volatile organic compounds:
       the Kanawha County health study. Am J Epidemiol  137:  1287-1301.
Warner, M; Eskenazi, B; Mocarelli, P: Gerthoux, PM: Samuels,  S: Needham, L: Patterson, D:
       Brambilla, P. (2002). Serum dioxin concentrations and breast cancer risk in the seveso
       women's health study. Environ Health Perspect 110: 625-628.
Warner, M: Samuels, S: Mocarelli, P: Gerthoux, PM: Needham, L: Patterson, DG, Jr: Eskenazi,
       R (2004). Serum dioxin concentrations and age at menarche. Environ Health Perspect
       112:  1289-1292. http://dx.doi.org/10.1289/ehp.7004.
Warner, M: Eskenazi, B. (2005). TCDD and puberty: Warner and Eskenazi respond. Environ
       Health Perspect 113: Al8. http://dx.doi.org/10.1289/ehp.113-al8a.
Warner, M: Eskenazi, B: Olive, PL; Samuels, S: Quick-Miles, S: Vercellini, P: Gerthoux, PM:
       Needham, L: Patterson, DG, Jr: Mocarelli, P. (2007). Serum dioxin concentrations and
       quality of ovarian function in women of seveso. Environ Health Perspect  115: 336-340.
       http://dx.doi.org/10.1289/ehp.9667.
Whitlock, JP, Jr. (1999).  Induction of cytochrome P4501A1. Annu Rev Pharmacol Toxicol 39:
       103-125. http://dx.doi.0rg/10.1146/annurev.pharmtox.39.l.103.
WHO (World Health Organization). (1978). International Classification of Diseases: Ninth
       Revision. Geneva, Switzerland, http://www.cdc.gov/nchs/icd/icd9.htm.
WHO (World Health Organization). (1998). Executive summary: Assessment of the health risk
       of dioxins: Re-evaluation of the tolerable daily intake (TDI). Geneva, Switzerland: WHO
       European Centre for Environmental Health and International Programme  on Chemical
       Safety.
                                         C-225

-------
WHO/UNICEF/ICCIDD (World Health Organization/ United Nations Children's Fund/
       International Council for the Control of Iodine Deficiency Disorders). (1994). Indicators
       for assessing iodine deficiency disorders and their control through salt iodization.
       (WHO/NUT/94.6). Geneva: World Health Organization.
Wolfe. W: Michalek. JE: Miner. JC: Rahe. A: Silva. J: Thomas. WF: Grubbs. WD: Lustik. MB:
       Karrison, TG: Roegner, RH: Williams, DE. (1990). Health status of Air Force veterans
       occupationally exposed to herbicides in Vietnam: I physical health. JAMA 264: 1824-
       1831.
Wolff MS: Teitelbaum. SL: Liov. PJ: Santella. RM: Wang. RY: Jones. RL: Caldwell. KL:
       Sjoedin, A; Turner, WE; Li, W: Georgopoulos, P; Berkowitz, GS. (2005). Exposures
       among pregnant women near the World Trade Center site on 11 September 2001.
       Environ Health Perspect 113:  739-748.
Youakim, S. (2006). Risk of cancer among firefighters: A quantitative review of selected
       malignancies [Review]. Arch Environ Occup Health 61: 223-231.
Zack, JA; Suskind, RR. (1980).  The mortality experience of workers exposed to
       tetrachlorodibenzodioxin in a trichlorophenol process accident. J Occup Med 22: 11-14.
Zack, JA; Gaffey, WR. (1983). A mortality study of workers employed at the Monsanto
       Company plant in Nitro, West Virginia. Environ Sci Res 26: 575-591.
Zober, A; Messerer, P; Huber, P. (1990). Thirty-four-year mortality follow-up of BASF
       employees exposed to 2,3,7,8-TCDD after the 1953 accident. Int Arch Occup Environ
       Health 62: 139-157.
Zober, A; Papke, O. (1993). Concentrations of PCDDs and PCDFs in human tissue 36 years after
       accidental dioxin exposure. Chemosphere 27: 413-418.
Zober, A; Ott MG: Messerer, P. (1994). Morbidity follow up study of BASF employees exposed
       to 2,3,7, 8-tetrachlorodibenzo-p-dioxin (TCDD) after a 1953 chemical reactor incident.
       Occup Environ Med 51: 479-486. http://dx.doi.Org/10.1136/oem.51.7.479.
Zober. A: Schilling. D: Ott MG: Schauwecker. P: Riemann, JF: Messerer. P. (1998).
       Helicobacter pylori infection:  Prevalence and clinical relevance in a large company. J
       Occup Environ Med 40: 586-594.
                                        C-226

-------
                                     EPA/600/R-10/038F
                                       www.epa.gov/iris
               APPENDIX D

Summaries and Evaluations of Cancer and
 Noncancer In Vivo Animal Bioassays for
    Inclusion in TCDD Dose-Response
                 Assessment
                  January 2012
           National Center for Environmental Assessment
              Office of Research and Development
             U.S. Environmental Protection Agency
                   Cincinnati, OH

-------
CONTENTS—Appendix D: Summaries and Evaluations of Cancer and Noncancer In Vivo
          Animal Bioassays for Inclusion in TCDD Dose-Response Assessment
LIST OF TABLES	D-v
LIST OF FIGURES	D-v

APPENDIX D.   SUMMARIES AND EVALUATIONS OF CANCER AND
               NONCANCER IN VIVO ANIMAL BIOASSAYS FOR INCLUSION
               IN TCDD DOSE-RESPONSE ASSESSMENT	D-l
      D. 1.   SUMMARY OF ANIMAL BIO ASSAY STUDIES INCLUDED FOR
            TCDD DOSE-RESPONSE MODELING	D-l
            D.I.I.  Reproductive Studies	D-2
                   D.I.1.1. Bowman et al. (1989a;  1989b) [and related Schantz and
                          Bowman (1989); Schantz et al. (1986); Schantz et al.
                          (1992)]	D-2
                         D. 1.1.1.1.  Supplemental published information on these
                                   rhesus monkeys [Rieretal. (1995; 1993)]	D-4
                   D.I.1.2. Franc etal. (2001)	D-5
                   D.I.1.3. Hochstein etal. (2001)	D-7
                   D.I.1.4. Hutt et al. (2008)	D-9
                   D.I.1.5. Ikeda et al. (2005b)	D-10
                   D.I.1.6. Ishihara et al. (2007)	D-ll
                   D. 1.1.7. Latchoumycandane and Mathur (2002) [and related:
                          Latchoumycandane et al. (2003,  2002a; 2002b)]	D-l2
                   D.I.1.8. Murray etal. (1979)	D-13
                   D.I.1.9. Shi et al. (2007)	D-14
                   D.I. 1.10. Yang etal. (2000)	D-15
            D.I.2.  Developmental Studies	D-17
                   D. 1.2.1. Aminetal. (2000)	D-17
                   D.l.2.2. Bell et al. (2007c)	D-18
                   D.l.2.3. Franczak et al. (2006)	D-20
                   D.l.2.4. Hojo et al. (2002) [and related: Zareba et al. (2002)]	D-21
                   D.I.2.5. Kattainen et al. (2001)	D-22
                   D.l.2.6. Keller et al. (2008a; 2008b; 2007c)	D-23
                   D.l.2.7. Kuchiiwa et al. (2002)	D-27
                   D.l.2.8. Li et al. (2006)	D-28
                   D.l.2.9. Markowski etal. (2001)	D-29
                   D.1.2.10.Miettinenetal. (2006)	D-29
                   D.1.2.11.Noharaetal. (2000b)	D-30
                   D.1.2.12.Ohsakoetal. (2001)	D-31
                   D.l.2.13.Schantz etal. (1996)	D-31
                   D. 1.2.14. Seo etal. (1995)	D-33
                   D.1.2.15.Sparschu etal. (1971)	D-34
                   D.I.2.16. Smith etal. (1976)	D-35
                   D.1.2.17.Simanainenetal. (2004b)	D-36

                                        D-ii

-------
                  CONTENTS (continued)

       D.1.2.18.Sugita-Konishietal. (2003)	D-38
D.1.3.  Acute Studies	D-39
       D.l.3.1.  Burlesonetal. (1996)	D-39
       D.l.3.2.  Crofton et al. (2005)	D-40
       D.1.3.3.  Kitchin and Woods (1979)	D-40
       D.l.3.4.  Lietal. (1997)	D-41
       D.1.3.5.  Lucieretal. (1986)	D-42
       D.1.3.6.  Nohara et al. (2002a)	D-42
       D.I.3.7.  Simanainen et al. (2003)	D-42
       D.1.3.8.  Simanainen et al. (2002)	D-43
       D.1.3.9.  Smialowicz et al. (2004)	D-44
       D.1.3.10.VandenHeuveletal. (1994)	D-45
       D.1.3.1 I.Weber etal. (1995)	D-46
D.1.4.  Subchronic Studies	D-48
       D.l.4.1.  Chu etal. (2001)	D-48
       D.l.4.2.  Chu et al. (2007)	D-49
       D.l.4.3.  DeCaprio etal. (1986)	D-50
       D.l.4.4.  Devito etal. (1994)	D-51
       D.l.4.5.  Fattore et al. (2000)	D-52
       D.l.4.6.  Fox etal. (1993)	D-53
       D. 1.4.7.  Hassounetal. (1998)	D-54
       D. 1.4.8.  Hassounetal. (2000)	D-54
       D.l.4.9.  Hassoun et al. (2003)	D-55
       D.1.4.10.Kocibaetal. (1976)	D-56
       D.1.4.11.MallyandChipman(2002)	D-58
       D. 1.4.12. Slezak etal. (2000)	D-58
       D.1.4.13.Smialowicz etal. (2008)	D-60
       D.1.4.14.VanBirgelenetal. (1995a; 1995b)	D-60
       D.1.4.15.Vos etal. (1973)	D-61
       D.1.4.16.Whiteetal. (1986)	D-63
D.I.5.  Chronic  Studies (Noncancer Endpoints)	D-63
       D.I.5.1.  Cantonietal. (1981)	D-63
       D.I.5.2.  Croutch et al. (2005)	D-64
       D.I.5.3.  Hassounetal. (2002)	D-65
       D.I.5.4.  Hong etal. (1989)	D-66
       D.I.5.5.  Kociba etal. (1978)	D-67
       D.I.5.6.  Maronpot et al. (1993)	D-68
       D.I.5.7.  National Toxicology Program (1982)	D-69
       D.I.5.8.  National Toxicology Program (2006)	D-70
       D.I.5.9.  Sewall etal. (1993)	D-72
       D.l.S.lO.Sewalletal. (1995a)	D-74
       D.l.S.ll.Tothetal. (1979)	D-75
       D.1.5.12.Tritscheretal. (1992)	D-76
D.I.6.  Chronic  Studies (Cancer Endpoints)	D-77

                           D-iii

-------
                      CONTENTS (continued)
            D.l.6.1.  DeliaPortaetal. (1987)	D-77
            D.l.6.2.  Kocibaetal. (1978)	D-79
            D.l.6.3.  Tothetal. (1979)	D-81
            D.l.6.4.  NTP(1982)	D-81
            D.l.6.5.  NTP(2006)	D-82
D.2.   EVALUATION OF STUDIES	D-84
      D.2.1. Evaluation of Animal Cancer Bioassays	D-84
      D.2.2. Evaluation of Animal Noncancer Bioassays	D-85
D.3.   CROSS-SPECIES CONCORDANCE OF SELECTED HEALTH
      ENDPOINTS	D-85
D.4.   REFERENCES	D-161
                               D-iv

-------
                                    LIST OF TABLES


D-l.   Noncancer animal studies selected for TCDD dose-response analyses	D-88
D-2.   Noncancer animal studies not selected for TCDD dose-response analyses and
       reasons for exclusion	D-92
D-3.   Cross-species concordance of male reproductive effects	D-148
D-4.   Cross-species concordance of female reproductive effects	D-149
D-5.   Cross-species concordance of thyroid effects	D-150
D-6.   Cross-species concordance of developmental dental effects	D-l 51
D-7.   Cross-species concordance of immune system effects	D-152
D-8.   Cross-species concordance of neurological effects	D-154
                                   LIST OF FIGURES
D-l.   Male reproductive effects across species	D-l55
D-2.   Female reproductive effects across species	D-156
D-3.   Thyroid effects across species	D-l57
D-4.   Developmental dental effects across species	D-158
D-5.   Immune system effects across species	D-l59
D-6.   Neurological effects across species	D-160
                                         D-v

-------
APPENDIX D.  SUMMARIES AND EVALUATIONS OF CANCER AND NONCANCER
                       IN VIVO ANIMAL BIOASSAYS FOR INCLUSION
                          IN TCDD DOSE-RESPONSE ASSESSMENT
D.I.  SUMMARY OF ANIMAL BIOASSAY STUDIES INCLUDED FOR TCDD
      DOSE-RESPONSE MODELING
       This appendix summarizes studies that have already met the in vivo animal bioassay
2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD) study inclusion criteria (see Section 2.3.2). These
studies are identified and described in a tabular form in Section 2.4.2 of the main document in
Tables 2-3 and 2-4, for cancer and noncancer, respectively.  Section D.2 of this appendix also
provides lists of the animal bioassays that met and did not meet the study inclusion criteria.
Sections D.2.1 and D.2.2 describe the results for the cancer and noncancer studies, respectively.
Table D-l presents the noncancer studies that met the study inclusion criteria, and Table D-2
identifies the  noncancer studies that were excluded, along with the criteria that were not met for
those studies. The following study summary sections are organized by reproductive studies,
developmental studies, and general toxicity studies (subdivided by duration).  They summarize
the experimental protocol, the results, and the no-observed-adverse-effect levels (NOAELs) and
LOAELs U.S. Environmental Protection Agency (EPA) has identified for each included study.
       To evaluate and discuss studies consistently, doses were converted to nanograms per
kilogram body weight per day (ng/kg-day) and were also adjusted for continuous exposure.
Some doses were adjusted based on daily dietary intake and body weight. For these studies,
EPA uses 10% of an animal's body weight as the  daily feed rate. More commonly, doses were
adjusted from 5 days/week to a 7 days/week standard adjustment, in which case administered
doses were multiplied by 5 and divided by 7 to obtain continuous doses. To adjust for weekly
dosing, the weekly administered doses were multiplied by the administration frequency per week
(in days) and  divided by 7 to give continuous doses.
       Other exposure protocols used a single loading dose followed by weekly maintenance
doses. To adjust these doses, the loading dose was added to the maintenance doses multiplied by
the administration frequency, and this sum was divided by the exposure duration to give a
continuous dosing rate. The doses administered in single dose studies were not averaged over
the observation period.
                                          D-l

-------
D.I.I.   Reproductive Studies
D. 1.1.1.    Bowman et al (1989a; 1989b) [and related Schantz and Bowman (1989); Schantz
           et al (1986); Schantz et al (1992)1
      Female rhesus monkeys (6 to 10 years old; 8 per treatment) were exposed to 0 or 5 ppt
(for 3.5 years), or 25 ppt (for 4 years) TCDD (purity not specified) (Schantz et al., 1992;
Bowman et al.. 1989a: Bowman et al.. 1989b: Schantz and Bowman. 1989: Schantz et al.. 1986).
Female monkeys were mated to unexposed males after 7 months (Cohort I) and 27 months
(Cohort II) of exposure, and, then again 10 months postexposure (Cohort III).  The average daily
doses to mothers were equivalent to 0, 0.12, and 0.67 ng/kg-day. The 0.67 ng/kg-day dose group
had reduced reproductive rates in both Cohorts \(p< 0.001) and II (Bowman et al.,  1989b). The
mean number of days of offspring survival (p < 0.023) also decreased.  No effects on birth
weight or growth, or physical evidence of toxicity (Bowman et al., 1989a) were observed.
Behavioral effects were observed in the offspring (Cohort I: 7, 6, and 0 offspring, respectively;
Cohort II: 3, 5, and 0 offspring, respectively; Cohort III: 6, 7,  and 3, respectively). In the
0.67 ng/kg-day dose group, the number of offspring was insufficient to form a group in either
Cohorts I or II. Offspring in the 0.12 ng/kg-day dose group had alterations in social behavior of
the mother-infant pairs (mothers had increased care giving, which appeared to be an effect of the
infants and not due to the treatment of the mother) and peer group of the offspring after weaning
(Bowman et al., 1989a). The performance of learning tasks was inversely related to the level of
TCDD in the body fat. Schantz and Bowman (1989) examined effects using
discrimination-reversal learning  (RL) and delayed spatial alteration (DSA). RL detected effects
in the 0.12 ng/kg-day group as measured by retarded learning of the shape reversal (p < 0.05),
but DSA did not. In another behavioral study, Schantz et al. (1992) placed two offspring (one
male, one female) from the 0.12  ng/kg-day dose group of Cohort I into each of three peer groups
that also consisted of two control monkeys tested in a large playroom for 1.5 hours/day,
5 days/week.  Patterns of behavior were then watched beginning on the second day of
socialization 4 days/week for 9 weeks.  Play behavior, displacement, and self-directed behavior
were significantly altered in the TCDD-exposed offspring.  In a second experiment by Schantz
et al. (1992) utilizing offspring from Cohort III (i.e., born after the cessation of maternal
exposure to TCDD), four offspring from mixed treatment groups (i.e., control and 0.12 and
0.67 ng/kg-day dose groups; varying numbers of males and females per group) and 3-4 offspring

                                          D-2

-------
from the same treatment groups were placed into peer groups and assessed similarly as described
above. Behavioral changes were observed in peer groups containing only TCDD-exposed
offspring, but behavior was not altered in TCDD-exposed offspring socializing with control
monkeys.  Additionally, Schantz et al. (1986) combined the cohorts and looked at 5, 5, and
3 mother-infant pairs in the 0, 0.12, and 0.67 ng/kg-day groups, respectively.  They found that
TCDD-exposed mother-infant pairs spent more time in close, social contact compared with the
controls (mutual ventral contact,/* < 0.025; nipple contact,/* < 0.01) and infants had reduced
locomotor activity (p < 0.05), but the dose effect was complex. Of note, the control groups
contained fewer males than did the TCDD-exposed groups.
       From these reproductive studies in monkeys, a lowest-observed-adverse-effect level
(LOAEL) of 0.12  ng/kg-day is established for significantly altered social behavior in offspring
from TCDD-exposed females (Schantz et al., 1992). A NOAEL cannot be determined.
However, there are several issues associated with these data that confound their interpretation.
For example, there were a small number of TCDD-exposed offspring (only one male and
one female) in a limited number of observed peer groups (only three).  The subjective  nature of
the experimental design (e.g., observing and scoring the various social  interactions and other
behaviors among the offspring, the schematic of the playroom apparatus, etc.) also contributes
uncertainty to the  data analysis. Additionally, the biological significance of the alteration in
social behaviors among the TCDD-exposed offspring (e.g., increased initiation of social play as
it pertains to overall  social adjustment) is difficult to assess.  Furthermore, in a follow-up report
by Rier et al. (200 Ib), DLC levels were quantified in the sera of some of the maternal monkeys
from the aforementioned studies 13 years after termination of TCDD treatment. Rier et al.
(200Ib) reported that the animals had elevated serum polychlorinated biphenyl (PCB)77 and
PCB126 levels and an increased serum toxicity equivalence (TEQ). Although the cause of the
elevated PCB levels was unclear, the study authors  speculated that "accumulation of PCBs in
TCDD-treated animals may have resulted from PCB exposure during TCDD administration due
to a contaminated TCDD solution or other inadvertent source." They also inferred that all the
animals may have been exposed to PCBs in their feed or other environmental sources. Taken
together, the multitude of confounding factors greatly decreases the confidence in the
dose-response data from aforementioned reproductive studies in monkeys.
                                          D-3

-------
D.I.1.1.1.  Supplemental published information on these rhesus monkeys [Rier et al. (1995;
           1993)1
       Rier et al. (1995; 1993) examined the impact of chronic TCDD exposure on
endometriosis. Female rhesus monkeys (eight animals per treatment group) were exposed to 0,
5, or 25 ppt TCDD (purity not specified) in feed for 4 years.  Previously, Bowman et al. (1989a)
determined that these dietary concentrations were equivalent to 0, 0.12, and 0.67 ng/kg-day,
respectively.  Ten years after termination of TCDD treatment, the presence of endometriosis was
determined via laparoscopic surgical procedure, and the severity of the disease was assessed.
The study authors reported that three monkeys in the 0.67 ng/kg-day exposure group died at 7, 9,
and 10 years after termination of TCDD treatment.  Autopsy results attributed the deaths to
widespread and severe peritoneal endometriosis (all three monkeys) along with obstruction of the
colon (one monkey) and blockage of the jejunum (one monkey).  Other deaths also occurred in
the control group (1 death from birthing complications and another from an unknown cause); in
the 0.12 ng/kg-day dose group (1 death due to natural causes with no endometriosis), and in the
0.67 ng/kg-day dose group (1 death due to a breeding fight with no incidence of endometriosis).
At study termination, 17 live animals and the 3  that had previously died of endometriosis were
evaluated (total n = 20).
       Incidence of endometriosis was significantly (p < 0.05) higher than in the control group
with 71 and 86% incidence rates in the 0.12 and 0.67 ng/kg-day dose groups, respectively,
compared with 33% in the control group. Severity of endometriosis was also significantly
(p < 0.001) correlated with TCDD  dose. Staging by rAFS indicated that untreated control
animals had either minimal or no incidence of endometriosis. In comparison, endometriosis was
absent in 2 of the 7 monkeys in the 0.12 ng/kg-day dose group, while only 1 of the 7 animals in
the high-dose group was disease free.  Moderate-to-severe disease was observed in 3 of the
7 animals in the 0.12 ng/kg-day dose group and 5 of the 7 animals in the 0.67 ng/kg-day dose
group. Moderate-to-severe disease was not observed in the control group.  The authors also
compared the incidence and severity of endometriosis in TCDD-exposed animals with
304 normal, nonneutered females with no dioxin exposure and reported that the disease was not
present in monkeys that were less than 13 years of age, while the disease rate was 30% among
animals 13 years of age or older. The study authors report that these findings are in agreement
with human and rhesus studies demonstrating that the prevalence of detectable endometriosis can
increase with advanced age.
                                         D-4

-------
       In a follow-up report, Rier et al. (200Ib) examined the DLC and TCDD levels in sera
collected from 9 treated (n = 6, 0.12 ng/kg-day dose group; n = 3, 0.67 ng/kg-day dose group)
and 6 control female monkeys surviving from the Rier et al. (1995; 1993) study and 13 years
after termination of TCDD treatment.  Additional studies were conducted on four monkeys that
died 7 to 11 years after TCDD exposure. Rier et al. (200Ib) reported that treated animals in this
study had elevated serum TCDD, PCB77, and PCB126 levels, as well as an increased serum
TEQ; the fractional contribution of serum TCDD levels to total serum TEQ was 30% in treated
animals. Although the severity of endometriosis in the 15 monkeys examined was determined
previously (Rier et al., 1995; Rier et al., 1993), it was reevaluated and disease  status was similar
between laparoscopies.  Endometriosis severity corresponded to the serum PCB77
concentrations rather than total TCDD. As stated previously, the study authors speculated that
"accumulation of PCBs in TCDD-treated animals may have resulted from PCB exposure during
TCDD administration due to a contaminated TCDD solution or other inadvertent source." They
also inferred that all the animals may have been exposed to PCBs in their feed or other
environmental sources.  Thus, in these studies, it is not possible to determine the contribution of
TCDD, alone, to the endometriosis due to the background contamination. These studies (Rier et
al., 1995; Rier et al., 1993), were not selected for TCDD dose-response modeling because
exposures were not to TCDD only.


D. 1.1.2.    Franc et al. (2001)
       To study the effects of subchronic, low-dose exposure to TCDD on the regulation and
expression of the aryl hydrocarbon receptor (AhR), Franc et al. (2001) used rodent models with
varying sensitivities to TCDD. Female Sprague-Dawley rats, inbred Long-Evans rats, and
outbred Han/Wistar rats (eight per dose group) were dosed via oral gavage with 0, 140, 420, or
1,400 ng/kg TCDD (>99% purity) dissolved in corn oil once every 2 weeks for 22 weeks (0, 10,
30, and 100 ng/kg-day average daily doses). Animals were sacrificed 10 days after the final
dosing. Body weights were recorded biweekly and just before sacrifice. After sacrifice, liver
and thymus weights were determined.  Liver tissue samples were removed and either frozen for
RNA isolation followed by semiquantitative reverse transcription polymerase chain reaction
(RT-PCR) or homogenized and prepared for subcellular fraction analysis. Radioligand binding
and immunoblotting techniques were used to measure AhR levels, and RT-PCR analysis was
                                         D-5

-------
used to assess mRNA levels of AhR, aryl hydrocarbon nuclear receptor (ARNT), and
cytochrome P450 (CYP)lAl.
       Long-Evans rats exhibited significant (p < 0.001) decreased weight gain over time as
compared with the  Sprague-Dawley and Han/Wistar rats as determined by repeated measures
analysis of variance (ANOVA).  Because body-weight gain varied indirectly with TCDD
exposure, liver and thymus tissue weights were normalized to body weight for data analysis.
TCDD exposure led to a significant (p < 0.05) increase in relative liver weights at all
three TCDD doses  and in all three rat strains, compared with the control groups. At the upper
end of the TCDD dose range, Sprague-Dawley rats dosed with 100 ng/kg-day showed the
greatest increase in relative liver weights (160% of the control values), while the relative liver
weights in Long-Evans and Han/Wistar rats were similar to each other, and also were elevated
above control values by 10-20%. At the 30 and 100 ng/kg-day doses, the relative thymus
weights were significantly lower (p < 0.05) in all rat strains compared with their corresponding
controls, but the 10 ng/kg-day dose did not produce a statistically  significant effect in any strain.
However, absolute  thymus weight was  higher at all doses in Han/Wistar rats, which also had a
higher control thymus weight.
       Supporting  observed differences in baseline TCDD sensitivity among the rat strains, liver
AhR levels in the control groups as measured by radioligand binding were similar for Sprague
Dawley and Han/Wistar rats, but were approximately twofold higher for Long-Evans rats. A
significant (p < 0.05) twofold, dose-dependent increase in radioligand binding of liver AhR was
observed at all TCDD doses relative to  the control in Sprague-Dawley rats.  At the 30 ng/kg-day
dose, the AhR level for Long-Evans rats was significantly (p < 0.05) increased to approximately
250% of the control level.
       AhR protein levels measured in the liver cytosol by immunoblotting were highest in the
10 and 30 ng/kg-day TCDD dose groups for all three rat strains. Significant (p < 0.05) increases
in AhR levels were observed in the Sprague-Dawley rats that received 30  ng/kg-day, and in
Long-Evans rats that received either  10 or 30 ng/kg-day. A significant (p < 0.05) decrease in
AhR protein level was observed only at the 100 ng/kg-day dose in Han/Wistar rats. Liver AhR
protein was not detectable by immunoblotting in nuclear extracts for any strain or dose. The
study authors assert that AhR levels measured in cytosol correspond to measures in whole-tissue
lysates as demonstrated in their previous work.
                                          D-6

-------
       Based on RT-PCR analysis, all three rat strains showed similar responses in liver AhR
mRNA following TCDD exposure. Liver AhR mRNA levels increased significantly (p < 0.05)
as compared with control levels in all rat strains at 10 and 30 ng/kg-day and in Long-Evans rats
at 100 ng/kg-day. The study authors observed that statistically significant increases in AhR
mRNA levels in the liver were not always associated with statistically significant increases in
AhR levels for a given strain and dose, but that the opposite (increases in AhR levels associated
with increases in AhR mRNA levels) was always true. Changes in liver ARNT mRNA levels
tended to increase with increasing TCDD dose, and the increases were significant (p < 0.05) in
the 30 ng/kg-day dose groups of Long-Evans and Han/Wistar rats.  At the 100 ng/kg-day TCDD
dose, all rat strains showed a decrease in ARNT mRNA in the liver relative to controls with
significant (p < 0.05) differences for the 100 ng/kg-day TCDD dose groups of Sprague-Dawley
and Han/Wistar rats.  Liver CYP1 Al mRNA induction was not detectable in control animals.  A
significant (p < 0.05) increase in liver CYP1 Al mRNA was observed in all rat strains
administered  10 or 30 ng/kg-day TCDD. Liver CYP1A1 mRNA levels also were significantly
(p < 0.05) elevated above controls in the 100 ng/kg-day groups although not to the same extent
as in the 30 ng/kg-day groups. For all rat strains, the largest up-regulation for AhR and ARNT
mRNA levels occurred in the 30 ng/kg-day TCDD dose groups.
       The NOAEL for TCDD identified in this study is 10 ng/kg-day TCDD.  At 10 ng/kg-day
TCDD, the change in relative liver weight, while significantly (p < 0.05) increased in
Sprague-Dawley rats, was determined (Franc et al., 2001) to be less than 10% and judged by
EPA not to be biologically relevant. Also, at 10 ng/kg-day TCDD, the change in relative thymus
weight, was not statistically significantly decreased in Sprague-Dawley, Han-Wistar or
Long-Evans rats. The study LOAEL is 30 ng/kg-day based on statistically and biologically
significant increases in relative liver weight in Sprague-Dawley and Long-Evans rats and
statistically and biologically significant decreases in relative thymus weight in Sprague-Dawley,
Han-Wistar, and Long-Evans rats.
D. 1.1.3.    Hochstein et al (2001)
       Adult female mink (12/treatment group) were administered dietary concentrations of
0.0006 (control), 0.016, 0.053, 0.180, or 1.40 ppb TCDD (purity >99.8%) for 132 days
(Hochstein et al., 2001).  This dose is estimated to be equivalent to 0.03 (control), 0.8, 2.65, 9,
                                          D-7

-------
and 70 ng/kg-day assuming a food consumption of 5% of body weight per day. Females were
mated with unexposed males beginning on treatment Day 35. Females were allowed to mate
every fourth day during a 29-day mating period or until a confirmed mating. Mated females
were presented with a second male either the day after initial mating or 8 days later. In the
70 ng/kg-day group, the treated animals were lethargic after 4 to 5 weeks, with several having
bloody (tarry) stools near the end of the trial. Two animals in the 70 ng/kg-day dose group died
prior to study termination.  These animals had lost a large percentage of their body weight
(24-43%),  and had pale yellow livers and intestinal hemorrhages. Histopathology from both
mink indicated marked diffuse hepatocellular vacuolation.  The mean body weight decreased in
all treatment groups including the control (losing an average of 3.29% of initial body weight),
compared to a dose-dependent loss of up to 26% in the 70 ng/kg-day group. Mating and
reproduction were considered subnormal in all groups.  The number of females that gave birth in
the 0.03 (control), 0.8, 2.65, 9, and 70 ng/kg-day dose groups were 5/12, 0/12, 3/12, 8/12, and
0/11, respectively.  The study authors speculated that the subnormal breeding and reproductive
performances in the control females likely were due to the indoor environment in which the mink
were housed. In the three groups that gave birth, there was a dose-dependent decrease in kit
body weight at birth, which was significant (p < 0.05) in the 9 mg/kg-day group compared with
the controls.  The body weight in the kits was not significantly different at 3 or 6 weeks after
birth.  The 3-week survival rates of 71, 47, and 11% were recorded for kits in the 0.03  (control),
2.65, and 9 ng/kg-day dose groups, respectively.  Six-week kit survival rates were 62, 29, and
11% in the 0.03 (control), 2.65, and 9 ng/kg-day dose groups, respectively.
       In the adult females, clinical signs of toxicity were noted in the 70 ng/kg-day group near
the end of the study and included alopecia and notably thickened, deformed, and elongated
toenails.  There was a dose-dependent decrease in plasma total solids,  total protein, and
osmolality that reached statistical significance (p < 0.05) in the two highest exposure groups.
Anion gap was significantly decreased (p < 0.05) and alanine aminotransferase was significantly
increased in the 70 ng/kg-day group compared to the controls. At terminal sacrifice, there was a
dose-related decrease in body weight. There was a dose-related increase in liver weight that
reached statistical significance (p < 0.05) in the 70 ng/kg-day dose group.  The brains of 42% of
the animals in the 70 ng/kg-day dose group had localized accumulation of lymphatic cells within
the meninges with mild extension into the adjacent neuropil and mild gliosis.  Of the 10 mink
                                          D-8

-------
surviving to study termination in the 70 ng/kg-day group, 3 had periportal hepatocellular
vacuolation. These same brain and liver lesions were not observed in the control mink.
       As there were no litters produced in the low-dose group and pregnancy outcomes were
not dose related, the 0.8 ng/kg-day exposure level does not inform the choice of NOAEL or
LOAEL.  Thus, the LOAEL for this study is 2.65 ng/kg-day (132-day maternal exposure
duration) based on reduced kit survival (47% of control at 6 weeks). A NOAEL cannot be
determined for this  study.


D.l.1.4.    Hutt et al (2008)
       Hutt et al. (2008) conducted a 3-month study investigating changes in morphology and
morphogenesis of preimplantation embryos as a result of chronic exposure to TCDD in female
rats.  The study authors administered 0 or 50 ng/kg TCDD (>99% purity) in corn oil via oral
gavage to groups of three pregnant Sprague-Dawley rats on gestation days (GDs) 14 and 21 and
on postnatal days (PNDs) 7 and 14. The resulting female pups were divided into groups of 3 and
administered 0 or 50 ng/kg TCDD (>99% purity) in corn oil (equivalent TCDD doses of 0 and
7.14 ng/kg-day) on  PND 21 and weekly thereafter until they reached 3 months of age.  Pups
were then mated, fertilization was verified, and preimplantation embryos were harvested
4.5 days later.  Preimplantation embryos were examined using immunofluorescence microscopy
to determine blastomere abnormalities.
       No significant difference as compared with the control in preimplantation embryotoxicity
was observed following exposure to TCDD. Morphologically normal preimplantation embryos
were significantly (p < 0.05) reduced in the 50 ng/kg TCDD exposed rats (15 of 41, 36.6%)
compared with the control group (31 of 39, 79.5%).  Preimplantation embryos  of TCDD-exposed
rats included irregularities in mitotic spindles (13 of 18 were monopolar),  chromosome patterns
in metaphase, blastomere size, and shape, blastomere nuclei shape in interphase, f-actin, and
cytokinesis. The study authors concluded that the compaction stage of preimplantation
embryogenesis is the most sensitive following exposure to TCDD.
       A LOAEL for this study is 50 ng/kg (7.14 ng/kg-day adjusted dose) for a significantly
(p < 0.05) lower proportion of morphologically normal preimplantation embryos during
compaction stage in female Sprague-Dawley pups weekly for 3 months. A NOAEL cannot be
determined for this  study.
                                         D-9

-------
D.l.1.5.    Ikeda et al (2005b)
       Ikeda et al. (2005b) studied the effect of repeated TCDD exposure to FO dams on the
male gonads of Fl generation and sex ratio in the F2 generation. Twelve female Holtzman rats
were treated with a single dose of 400 ng/kg TCDD (>98% purity) orally, via gavage, followed
by weekly treatment doses of 80 ng/kg TCDD (16.5 ng/kg-day adjusted for continuous exposure
of 10 weeks; specified 2 weeks premating, assumed 1 week for successful mating, 3 weeks of
gestation, and specified 4 weeks to weaning) during mating, pregnancy, and lactational periods
(total exposure duration approximately 10 weeks).  Corn oil served as the control in another
group of 12 dams. Four dams were sacrificed on GD 20 to evaluate the in utero toxicity of
TCDD. Litter sizes from the remaining eight dams were examined on PND 2, and some of the
Fl offspring were sacrificed to estimate TCDD tissue concentrations. The remaining offspring
were weaned on PND 28.  Some of the Fl (number not specified) offspring were mated with
untreated females on PND 98, following which, litter size, sex ratio, weight, and anogenital
distance of F2 pups were examined on PND 2. Mated and unmated Fl males were sacrificed and
the testes, epididymis, seminal vesicle, and the ventral prostate were weighed; the cauda
epididymis was weighed and examined for sperm count.
       All fetuses in the control and TCDD group as a result of in utero exposure in the
FO generation survived.  Litter size, sex ratio, and anogenital distance in the Fl generation on
PND 2 were not altered as a result of in utero TCDD exposure.  Pup weight was  significantly
(p < 0.05) lower in the TCDD-treated group than in controls. TCDD concentration in the
adipose tissue of the FO dams on GD 20 was significantly (p < 0.05) higher than  in the liver.
Adipose TCDD was significantly (p < 0.01) reduced at weaning, however, compared to
concentrations on GD 20.  Fl pup liver TCDD concentration increased significantly (p < 0.01)
and was higher on PND 28 than PND2. The liver weight in Fl  males increased by 14-fold at
PND 28 compared to PND 2, implying a transfer of approximately 850 pg of TCDD from the
dam to the Fl pup livers during lactation. TCDD also was detected in pup adipose tissue on
PND 28. Body weight of TCDD-exposed Fl males was significantly (p < 0.001) lower than
control males at weaning (PND 28).  No significant differences in testis and cauda epididymis
weights were observed between the control and  treated groups.  Ventral prostate  weight in the
Fl males exposed to TCDD, however, was  approximately 60% lower than controls.  No change
in weight of the body, brain, testes, cauda epididymis, or seminal vesicle was observed at

                                         D-10

-------
PND 120. Ventral prostate weight, however, was 16% lower than that of the control group
(p < 0.001).  Sperm count in the cauda epididymis of the Fl males was not affected by TCDD
exposure.
       Examination of F2 generation litters indicated no significant differences in litter size, pup
body weight, and anogenital distance between TCDD-treated or vehicle control groups. The
percentage of male F2 pups born to maternally and lactationally TCDD-exposed males was
significantly (p < 0.05) lower (38%) than those sired by control group males (52%).  Every
female mated with maternally TCDD-exposed Fl males delivered more female than male pups.
       A LOAEL for TCDD of 16.5 ng/kg-day for an estimated 10 week exposure duration in
FO rat dams is identified in this study for decreased development of the ventral prostate in the
Fl generation (60% lower than controls) and for significantly (p < 0.05) altered sex ratio
(decreased percentage of males) in the F2 generation.  A NOAEL cannot be determined for this
study.


D.l.1.6.   Ishihara et al (2007)
       Ishihara et al. (2007) examined the effect of repeated TCDD exposure of FO males on the
sex ratio of Fl offspring. Seven-week-old male ICR mice (n = 127) were divided into
three groups and treated via gastric intubation with an initial loading dose of either 2 or 2,000 ng
TCDD/kg body weight(BW) or an equivalent volume of sesame oil (vehicle) as control, followed
by a weekly  maintenance doses of 0, 0.4, or 400 ng/kg until the animals were  12 weeks old.
One week after the last exposure, the animals were mated with untreated female mice. On the
day a vaginal plug was identified, FO male mice were sacrificed and major organs including
testes, epididymis, and liver were removed and weighed.  Organ tissues also were examined for
histopathological and immunohistochemical changes.  Treatment  levels, averaged over the
6 week period from start of treatment to mating (five maintenance doses), were 0, 0.095,  and
950 ng/kg-day for the control, low dose and high dose groups, respectively.
       All TCDD-treated males  successfully impregnated untreated females and yielded viable
offspring. Mortality, pup weights, and mating and fertility indices were not affected by TCDD
exposure. There were no significant differences in body weights or in relative weights of testes,
epididymis, or livers in the TCDD-treated FO males compared to the control group. The livers of
some animals (number not specified) in the high-dose group, however, were larger and heavier
                                         D-ll

-------
than in the controls or the low-dose group.  Hence, tissues from the high-dose animals were
selected for detailed immunohistochemical examination.
       General histopathological findings in the TCDD-treated groups showed no changes in
cell morphology in germ, Sertoli, and Leydig cells of the testes.  Arrangement of the germ cells
was normal and there was no difference in the epididymis spermatozoon number in  either of the
TCDD-treated groups compared to controls. Livers of some of the animals in the high-dose
group however, showed enlarged and vacuolated areas in the centrilobular area when compared
to the low-dose group and the control group.  Immunohistochemical and quantitative
immunohistological findings showed a marked increase in staining intensity for CYP1A1 in the
cytoplasm of the hepatocytes in the centrilobular area of the high-dose TCDD group compared to
the cells in the low-dose and the control groups.  In addition, proportions of immunoreactive
CYP1 Al areas in the liver sections of the high-dose group were higher than in the low-dose and
control groups. The proportions of immunoreactive CYP1A1 also varied across animals (n = 33)
in the high-dose group.
       In addition to the above findings, there was a dose-related decrease in the male/female
sex ratio. The proportion of male offspring of the high-dose group was significantly lower
(p < 0.05) than that observed in controls (46.2% vs.  53.1%, respectively). Hepatic
immunoreactive CYP1A1 staining levels in individual FO males were strongly correlated with
the sex ratio of their offspring.
       A LOAEL for TCDD of 950 ng/kg-day for a 6 week exposure duration of FO male mice
is identified for significantly (p < 0.05) decreased male/female sex ratio (i.e., higher proportion
of female offspring) in the Fl generation. The NOAEL is 0.095 ng/kg-day.
D.I.1.7.    Latchoumycandane andMathur (2002) [andrelated: Latchoumycandane et al.
           (2003, 2002a; 2002b)J
       Latchoumycandane and Mathur (2002) conducted a study to determine whether treatment
with vitamin E protected rat testes from TCDD-induced oxidative stress. Groups of albino male
Wistar rats (n = 6) were administered an oral dose of 0 (vehicle alone) 1, 10, or 100 ng
TCDD/kg-day for 45 days, while another group of animals (n = 6) was coadministered TCDD at
the same doses, along with vitamin E at a therapeutic dose of 20 mg/kg-day for 45 days. At
study termination, animals were fasted overnight, weighed, and sacrificed.  Testis, epididymis,
                                         D-12

-------
seminal vesicles, and ventral prostate were removed, weighed, and preserved for further
examination. The left testis was used to determine daily sperm production, while the right testis
was used for biochemical studies.  Superoxide dismutase (SOD), catalase,  glutathione reductase,
and glutathione peroxidase activity were measured in the testes, along with production of
hydrogen peroxide and lipid peroxidation. In a separate exposure protocol, groups of albino
male Wistar rats (n = 4) were administered an oral dose of 0 (vehicle alone) 100, 1,000, or
10,000 ng/kg-day TCDD for 4 consecutive days (Latchoumycandane et al., 2003 see summary in
Appendix H): .
       Body weights of TCDD-treated rats did not differ significantly from the control group.
Testis, epididymis, seminal vesicle, and ventral prostate weights in the TCDD-treated groups,
however, decreased significantly (p < 0.05) when compared with controls. None of these
changes were observed in the TCDD-exposed groups receiving vitamin E. There was a
dose-related decrease in  daily sperm production (p < 0.05) in all three TCDD-treated groups
when compared with the control group.  In contrast, the TCDD-treatment groups that also
received vitamin E did not show any significant changes in daily sperm production compared to
the controls. The TCDD-treated groups also showed  significantly (p < 0.05) lower activities of
the antioxidant enzymes (superoxide dismutase, catalase, glutathione reductase,  and glutathione
peroxidase) than the control group. Levels of hydrogen peroxide and lipid peroxidation
increased significantly (p <  0.05) in the testes of the rats treated with TCDD compared to the
corresponding controls.  The TCDD-treated groups that had been coadministered vitamin E show
no difference in antioxidant enzyme activities or in reactive oxygen species production when
compared with controls.
       A LOAEL for TCDD of 1.0 ng/kg-day for a 45-day exposure duration in rats is identified
in this study for significantly (p < 0.05) reduced sperm production and significantly (p < 0.05)
decreased reproductive organ weights.  A NOAEL cannot be determined for this study.
D.I.1.8.    Murray et al (1979)
       Male (10-16 per treatment) and female (20-32 per treatment) Sprague-Dawley rats were
administered diets containing TCDD (purity >99%) to achieve daily dosages of 1, 10, or
100 ng/kg-day through three generations. After 90 days of treatment, FO rats were mated to
produce Fla offspring.  Thirty-three days after weaning of the last F la litter, the FO rats were
                                          D-13

-------
mated again to produce Fib offspring. Some FO rats were mated a third time for a cross-mating
study.  The Fib and F2 rats were mated at about 130 days of age to produce the F2 and
F3 generations. No clinical signs of toxicity or changes in body weight or food consumption
were observed in FO rats during the 90 days of treatment before mating. The 100 ng/kg-day
group was discontinued due to the lack of offspring.  In the three surviving offspring (all males),
no changes in appearance, body weight, or food consumption occurred. A dose of 10 ng/kg-day
caused a consistent decreased body weight in both sexes of Fl and F2 rats, which was associated
with decreased food consumption. A significant (p < 0.05) decrease in the fertility in the Fl and
F2 rats occurred in the 10 ng/kg-day group—but not in FO rats.  The number of live pups and
gestational  survival index were significantly (p < 0.05) decreased in the 100 ng/kg-day FO rats
and in the 10 ng/kg-day Fl and F2 rats. The gestational survival index also was significantly
(p < 0.05) decreased in F2 rats administered 1  ng/kg-day. Postnatal survival was significantly
(p < 0.05) reduced only in F2 rats administered 10 ng/kg-day. Growth (as measured by body
weight) was affected in the 10 ng/kg-day group only in the third generation. In the 10 ng/kg-day
group,  a significant (p < 0.05) decrease in relative thymus weight and increase in liver weight
also occurred in F3 rats (weights were not measured  in F2 rats). Additionally,  mating
100 ng/kg-day TCDD-treated females with untreated males increased the percent of implants
resorbed as assessed by uterine histopathology.
       The reproductive LOAEL is 10 ng/kg-day based on a significant (p < 0.05) decrease in
fertility (33-37% lower than controls); decrease in the number of live pups (18-27% lower than
controls); decrease in gestational survival (10-11% lower than controls); decrease in postnatal
survival (32% lower than controls); and decreased postnatal body weight (14-19% lower than
controls at weaning) in one or more generations. The reproductive NOAEL is  1 ng/kg-day.
D.l.1.9.    Shi et al (2007)
       Pregnant Sprague-Dawley rat dams (3 per treatment group) were administered 0, 1,5, 50,
or 200 ng/kg TCDD (purity >99%) in corn oil by gavage on GD 14 and GD 21 and on PND 7
and PND 14 for lactational exposure to pups (Shi et al.. 2007). Ten female pups per treatment
were selected and administered TCDD weekly at the same dose levels through their reproductive
lifespan (approximately 11 months).  The corresponding equivalent daily TCDD doses are 0,
0.14, 0.71, 7.14, and 28.6 ng/kg-day. Vaginal opening was slightly—but significantly
                                         D-14

-------
(p < 0.05)—delayed in the 28.6 ng/kg-day females. Vaginal opening was also delayed—but not
significantly—in the 0.14 and 7.14 ng/kg-day females. Reproductive senescence with normal
cyclicity was significantly (p < 0.05) accelerated beginning at 9 months in 7.14 and
28.6 ng/kg-day females.  Serum estradiol concentrations were decreased at all time points across
the  estrous cycle in a dose-dependent manner with a statistically significant decrease (p < 0.05)
in all but the lowest dose group.  TCDD exposure, however, did not affect the number or size
distribution of ovarian follicles; responsiveness of the pituitary gland to gonadotropin-releasing
hormone, or serum profiles of follicle stimulating hormone (FSH), luteinizing hormone (LH), or
progesterone.
       A LOAEL for TCDD of 0.71 ng/kg-day for an 11-month exposure duration was
identified in this study based on significantly (p < 0.05) decreased estradiol  levels in offspring.
The NOAEL for this study is 0.14 ng/kg-day.


D.l.1.10.   Yang et al (2000)
       Yang et al. (2000) studied the impact of TCDD exposure on the incidence and severity of
endometriosis in female rhesus monkeys.  Groups of 7- to 10-year old nulliparous cynomolgus
monkeys were treated with 0 (n = 5), 1, 5, or 25 (n = 6 per group)  ng/kg BW TCDD 5 days per
week via gelatin capsules for 12 months. Because the monkeys received 1 capsule 5 days per
week, the doses adjusted for continuous exposure were 0, 0.71,  3.57, and 17.86 ng/kg-day. Prior
to TCDD administration, all animals had endometriosis induced during Days 12-14 of the
menstrual  cycle by auto-transplantation of endometrial-strips in multiple abdominal sites.  All
TCDD-treated and control groups were laparoscopically examined during months 1,3, and 6 to
monitor the survival of endometrial implantations and to obtain peritoneal fluid to determine the
concentration and immunotype of endometrial growth regulator cytokines interleukin-6 (IL-6)
and interleukin-6 soluble receptor (IL-6sR). Because insufficient peritoneal fluids were present
in the treated and control monkeys, however, the study authors collected blood samples at 6 and
12 months during laparoscopy for routine hematology and to assess the circulating levels of IL-6
and IL-6sR.  All animals were sacrificed at 12 months, and circulating levels of gonadal steroids
also were measured at the time of necropsy.
       No changes were observed among treatment levels in general toxicological endpoints
such as body weight changes, food consumption, hematological endpoints, general activity
                                          D-15

-------
levels, and caretaker interaction. In addition, TCDD did not impact circulating levels of gonadal
steroids measured during necropsy.  Similarly, there were no differences in the number of
menstrual cycles, the length of the menstrual cycle, or bleeding intervals. Endometrial implants
were found in at least one site in all  TCDD-treated and control monkeys during the
first laparoscopic examination. Follow-up laparoscopies revealed that there was a continuous
loss of endometrial implants over time in each dose group.  At the 1-, 3-, and 6-month
examination, the number of endometrial losses was not significantly different among different
dose groups.  At the 12-month examination, however, a significantly (p < 0.05) higher rate of
survival of endometrial implants was observed in the 3.57 and 17.86 ng/kg-day dose groups
compared to the control group. The highest rate of endometrial implant survival was observed in
the ovaries regardless of the dose group. In contrast, all lesions disappeared from the left broad
ligament, whereas two on the right broad ligament and one on the uterine fundus survived.
There was a dose-dependent divergence in the growth response of endometrial implants
following TCDD exposure.  Both the maximum and minimum implant diameters in the
17.86 ng/kg-day dose group were significantly (p < 0.05) larger compared to controls. In
contrast, the maximum and minimum implant diameters in the 0.71 ng/kg-day dose group were
significantly (p < 0.05) smaller compared to controls. TCDD did not impact implant diameters
in the 3.57 ng/kg-day dose group when compared to controls. Histological examinations
revealed that endometrial glands and stromal cells were present in all surviving implants.
Sections examined in the 17.86 ng/kg-day of TCDD possessed cystic endometrial glands that
were more frequently observed in this dose group compared to other groups including controls.
In addition, the circulating levels of IL-6 were significantly (p < 0.05) lower in monkeys exposed
to 17.86 ng/kg-day TCDD both at 6 and 12 months compared to the control group.  In contrast,
the circulating levels of IL-6sR were significantly (p < 0.05) higher in animals treated with 3.57
and 17.86 ng/kg-day TCDD at 6 months, while the levels were higher only in the
17.86 ng/kg-day TCDD group at 12 months.
       A LOAEL for TCDD of 17.86 ng/kg-day for a 1 year exposure duration was identified in
this study for significantly (p < 0.05) increased endometriosis induced by endometrial implant
survival, significantly (p < 0.05) increased maximum and minimum implant diameters, and
growth regulatory cytokine dysregulation (as assessed by significantly decreased IL-6 levels,
p < 0.05).  ANOAEL of 3.57 ng/kg-day is identified in this study.
                                         D-16

-------
D.1.2.   Developmental Studies
D. 1.2.1.    Amin et al (2000)
      Amin et al. (2000) studied the impact of in-utero TCDD exposure on the reproductive
behavior in male pups.  Groups of pregnant Harlan Sprague-Dawley rats (n = 108 divided into
4 cohorts; number of animals in the TCDD treatment group is ~3 per dose group) were dosed via
gavage with 0, 25, or 100 ng/kg-day TCDD (purity >98%) in corn oil on GDs 10-16. On the
day of birth (PND 0), pups were examined for gross abnormalities and the number of live pups,
their weights, and sex were recorded from each litter.  Litters consisting of more than eight pups
were reduced to eight,  composed of four males and four females when possible.  Litters
consisting of fewer than five pups were excluded from the study to minimize between-litter
differences  in growth rate, maternal behavior, and lactational exposure.  After this exclusion,
approximately 10 to 11 litters per exposure group remained. All pups were weaned on Day 21
and one male and one female were retained to assess reproductive development, play behavior,
reproductive behavior, and saccharin preference behavior. Both male and female pups were
tested for saccharin preference between 189 and 234 days of age.  A saccharin preference test
was conducted for 8 days. For the first 4 days, rats were provided bottles containing tap water,
and on Days 5  and 6 the animals were provided a bottle  containing water and a bottle containing
0.25% saccharin solution. On Days 7  and 8, the animals were provided water and a bottle
containing 0.50% of saccharin solution.  A 0.50% saccharin solution was used because previous
studies have reported that male rats exhibited a greater reduction in preference for this saccharin
concentration compared to females, hence the sex difference in preference is more marked at this
saccharine dose.
      None of the treated dams exhibited any signs of toxicity as a result of exposure to TCDD.
Gestational body weight, liver weight, litter size, and percent live births were all comparable to
the corresponding control group. Birth rate and weaning weight of the pups also were not
affected by  TCDD exposure.  Sex-related water consumption, however, was significantly
(p < 0.001)  affected during the first 4 days with female pups drinking more water per 100 g of
body weight compared to the respective male counterparts.  Saccharin consumption was
significantly (p < 0.001) affected, with females consuming greater amounts of saccharin solution
per 100  g body weight compared with the corresponding males. Additionally, both male and
                                         D-17

-------
female pups drank significantly (p < 0.001) more of the 0.25% saccharin solution compared with
the 0.50% saccharin solution. Females of all exposure groups consumed less of both the
0.25 and 0.50% saccharin solution compared to the same-sex control group.  Comparisons of
each exposure group to the control group indicated that only the high TCDD exposure group
(100 ng/kg-day) different significantly (p < 0.05) compared to control in the consumption of
0.25% saccharin solution. In contrast, for the 0.50% saccharin solution, both the low- and
high-TCDD-dose groups differed significantly (p < 0.05 andp < 0.01, respectively) compared to
the control group.  The saccharin preference of TCDD-exposed male rats did not differ from that
of the male control group. The TCDD-exposed females' preference for saccharin solution,
however, was significantly reduced in both the 25 (p < 0.05) and the 100 ng/kg-day (p < 0.005)
dose group compared to that of the female  controls. The study authors state that the reduction in
saccharin consumption and preference in females could be due to the antiestrogenic action of
TCDD and that recent research reports suggest that TCDD can decrease the level of estrogen
receptor (ER) mRNA by blocking the ability of ER to transactivate from the estrogen response
element.
       A LOAEL for TCDD of 25 ng/kg-day for 7 days of gestational exposure is identified for
significantly (p  < 0.05) decreased preference in the consumption of 0.25% saccharin solution.  A
NOAEL cannot be determined for this study.


D.l.2.2.    Bell et al (2007c)
       Bell et al. (2007c) examined the reproductive effects of TCDD in rats exposed during
development. Female CRL:WI (Han) rats  were treated with TCDD (99% purity; dissolved in
acetone) in the diet at concentrations of 0 (acetone alone; « = 75), 28,  93, or
530 (« = 65/group) ng TCDD/kg diet, which provided average doses of 0, 2.4, 8, or
46 ng/kg-day, respectively. Rats were exposed to TCDD 12 weeks prior to mating, during
mating, and through pregnancy.  Dams were switched to the control diet after parturition.  Litters
from pregnant dams were reduced to a maximum size of eight on PND 4 and to five males (if
possible) on PND 21.  These males were left untreated until sacrificed (25/group, one/litter) on
PND 70, while all remaining animals were sacrificed on PND 120.  All sacrificed animals were
necropsied and received a seminology examination. Prior to sacrifice, during Weeks 12 and 13,
20 animals from each dose group were tested for learning ability and motor activity, and were
                                         D-18

-------
also administered a functional observation battery. During postnatal Week 16, groups of 20 male
Fl rats from each treatment group were paired with untreated virgin females for 7 days, and
mated females were killed on GD 16 and examined for terminal body weights, pregnancy status,
number of corpora lutea, and number of intrauterine implantations.
       The study authors found no evidence of direct maternal toxicity from exposure to TCDD.
In the high-dose groups, 8 of 27 dams suffered complete litter loss compared with 3 dams in the
control group, but the difference was not statistically significant. Pup survival at PND 4 was also
lower in the high-dose group, but the difference again was not statistically significant.
       A dose-related decrease in mean pup body weight was observed on PND  1, and this trend
continued throughout the lactation period. High-dose male pups had lower body weights when
compared to controls at PND 21, with this trend continuing over the course of the study.
Balanopreputial separation (EPS) was significantly (p < 0.05) delayed compared to controls in
all three treatment groups by 1.8, 1.9, and 4.4 days in the low-, medium-, and high-dose groups,
respectively.  The study authors reported that adjustment for lower body weights observed at
PND 21 and PND 42 did not affect the estimate of delay in EPS. No adverse effects from
maternal treatment were observed on learning or in functional observational battery performance.
Offspring in the high-dose group exhibited less activity when compared to controls (p < 0.05)
when they were subjected to a test of motor activity for 30 minutes.
       The median precoital time was 2-3 days for all 20 Fl males that were mated during
postnatal Week 16. The uterine and implantation data were similar in all dose groups and there
were no significant differences in the proportion of male offspring between groups. Epididymal
sperm counts and sperm motility did not differ significantly between dose groups in animals
sacrificed during postnatal Week 10. The mean number of spermatids was significantly lower
(14%; p < 0.05),  and the proportion of abnormal sperm was significantly (p < 0.05) higher in the
high-dose group when compared to controls on PND 70.  These effects, however, were not seen
in animals sacrificed on PND 120.
       Terminal  body weights were significantly (p < 0.05) decreased in the high-dose group
(6.9 %) compared to controls on PND 120, while the depression in body weight in the
medium-dose group (5.5%) was not statistically significant.  At PND 70, the relative and
absolute testis weight of the high-dose group was less than the controls (12 and 18%,
respectively). Absolute spleen weight in the high-dose group was significantly higher (8%) on
                                         D-19

-------
PND 70, and increased significantly (p < 0.05) by 1-3% on PND 120 in all dose groups
compared to controls.  Kidney weight in the low and medium-dose groups was significantly
(p < 0.05) greater than in controls (-2%) at PND 120. In addition to these organs, ventral
prostate (9.4%) and relative liver (-4.5%) weights were significantly (p < 0.05) higher than
controls on PND 120 in the medium- and low- and high-dose groups, respectively. On
PND 120, absolute brain weight was significantly (p < 0.05) less than the control in the
medium-dose group, while relative brain weight was significantly (p < 0.05) higher than the
control in the low- and high-dose group. Histological examination revealed no unusual findings.
       A LOAEL for TCDD of 2.4 ng/kg-day following an estimated 17-week exposure
duration of dams was identified in this study for significantly (p < 0.05) delayed BPS.  A
NOAEL was not identified in this study.
D.I.2.3.    Franczak et al (2006)
       Franczak et al. (2006) examined the impact of chronic TCDD exposure on the onset of
reproductive senescence in female rats. Pregnant Sprague-Dawley rats (n = 2-3/dose group)
were fed 50 or 200 ng/kg TCDD (>99% purity) or corn oil vehicle (4 mL/kg) orally on GD 14
and 21 and PND 7 and 14 to provide in utero and lactational exposure to TCDD. On PND 21,
female pups (n = 7/dose group) were weaned and were subsequently given weekly doses of
either 50 or 200 ng/kg-week TCDD by gavage (7.14 or 28.6 ng/kg-day adjusted for continuous
exposure; administered doses divided by 7) or corn oil vehicle. Exposure continued for up to
8 months, and the animals were observed for changes in estrus cycle at 4, 6, and 8 months. Rats
were sacrificed at 8 months of age when the TCDD-treated animals had entered the transition to
reproductive senescence. Following sacrifice, diestrus concentrations of serum LH, FSH,
progesterone, and estradiol were measured, and the ovaries were collected for examination.
       Estrus cycles at 4 months exhibited normal cyclicity in both TCDD-exposed groups and
did not differ significantly from the control group. At 6 months, however, there was a tendency
(p < 0.1) toward loss of normal estrus cyclicity in animals treated with TCDD. At the 8 month
observation, estrus cyclicity was significantly (p < 0.05) different in both dioxin-exposed groups
compared to controls (cumulative TCDD exposure is reported as 1.7 and 8 ug/kg for the 50 and
200 ng/kg dose groups, respectively).  The study authors noted that although the low-dose
animals showed an increased prevalence of prolonged cycles,  persistent estrus or diestrus was
                                         D-20

-------
observed in only 10% of the rats.  Conversely, approximately 50% of the rats exhibited loss of
cyclicity in the high-dose group. There were no changes in the number and size distribution of
ovarian follicles or the number of corpora lutea at either dose. Progesterone levels at 8 months
tended to be higher (p < 0.08) in animals receiving either 7.14 or 28.6 ng/kg-day TCDD
compared to controls, while serum estradiol concentrations were significantly (p < 0.03) lower at
diestrus.  Serum LH levels in TCDD-treated animals were comparable to those in the control
group, while FSH levels were elevated in rats receiving 7.14 ng/kg-day TCDD—but not in the
28.6 ng/kg-day dose group.
       A LOAEL for TCDD of 7.14 ng/kg-day for an 8-month exposure duration was identified
for significantly (p < 0.03) decreased serum estradiol levels.  A NOAEL cannot be determined
for this study.
D.l.2.4.    Hojo et al (2002) [and related: Zareba et al (2002)1
       Hojo et al. (2002) studied the impact of prenatal exposure to TCDD on sexually
dimorphic behavior in rats. Thirty-six pregnant Sprague-Dawley rats were assigned according to
a randomized block design to groups receiving 0, 20, 60, or 180 ng/kg TCDD (98% purity) on
GD 8.  Litters from pregnant dams were culled to 5 females and 5 males on PND 4 and allowed
to wean normally, at which time 5, 5, 6, and 5 litters from the 0, 20, 60, and 180 ng/kg TCDD
treatment groups, respectively, were maintained for examination of behavioral response.
Offspring were exposed to TCDD (from a single maternal exposure) for about 35 days through
gestation and lactation.  After weaning at PND 21, offspring were fed ad libitum until PND 80, at
which time a fixed amount of food was supplied daily to maintain constant body weights. At
90 days old, the rats in these treatment groups were trained to press a lever to obtain food pellets
using two operant behavior procedures. Initially, each lever press was reinforced. The fixed
ratio (FR) requirement was then increased every fourth session from the initial setting of 1 to
values between 6 and 71.  The responses for 30 days were studied under a multiple schedule
combining FR 11 and another schedule requiring a pause of at least 10 seconds between
responses (differential reinforcement of low rate,  or DRL 10-seconds)
       Pup and dam body weights were not affected by TCDD exposure, and all pups were
successfully trained in the lever-press response within 3-4 days.  Analyses of the FR procedure
data indicated that the male pups responded at a lower rate at all TCDD doses when compared to
                                         D-21

-------
the control group. In case of female pups, all TCDD-treated groups responded at a higher rate
than controls. None of these results was, by itself, however, statistically significant.
Examination of the FR 11 and DRL 10-second data indicated that when considering the FR
component of this multiple procedure, males from all three treatment groups responded at lower
rates when compared to the controls.  Conversely, all female pups responded at higher rates than
controls.  In addition, the treatment-by-sex interaction was significant (p = 0.036), with the
60 ng/kg female pups responding at a higher rate than the 60-ng/kg male pups. Examination of
the delayed response component in the multiple FR 11 and DRL 10-seconds procedures
indicated that almost all TCDD treatment groups were affected.  Like the FR component, male
pups at all TCDD dose groups responded at  a lower rate compared to controls, while female pups
at all dose groups responded at a higher rate than controls. There was also a significant
(p = 0.001) sex-by-treatment interaction for the DRL 10-seconds similar to the FR component.
Following behavioral testing, the animals were sacrificed and cortical depth measurements were
taken in selected right  and left brain regions. Reduced cortical thickness and altered brain
morphometry were observed in both male and female offspring in the 180-ng/kg exposure group
when compared to controls (Zareba et al., 2002).
       A nominal LOAEL for TCDD of 20  ng/kg for a single exposure on GD 8 is established
for this study based on abrogation of sexually dimorphic neurobehavioral responses.  A NOAEL
cannot be derived for this study.

D.l.2.5.   Kattainen  et al. (2001)
       Pregnant Line A, B, and C rats derived from Han/Wistar and Long-Evans rats
(4-8 pregnant dams/strain/treatment group)  were administered a single gavage dose of 0, 30,
100, 300, or  1,000 ng/kg TCDD (purity >99%) in corn oil on GD 15 (Kattainen et al.. 2001).  On
PND 1, the litters were culled to  three males and three females.  Offspring  were weaned on
PND 28.  Female pups were sacrificed on PND 35 and male pups were sacrificed on PND 70.
TCDD treatment did not affect body weight or cause  clinical signs of toxicity in the dams. In
Line B offspring, body weights in the 1,000  ng/kg group were slightly decreased during
PND 1-7, while Line C offspring had slightly decreased body weights throughout the study
period (data were not provided).  The development of the third molar was affected the most in
Line C offspring.  In 5  of 10 Line C females and 6 of 10 Line C males treated with 1,000 ng/kg
                                         D-22

-------
TCDD, the lower third molar did not develop.  In comparison, 1 of 19 Line A females and 1 of
18 Line B females administered 1,000 ng/kg TCDD lacked the third molar at sacrifice. Third
molars were present in all the controls and all male Line A and B offspring administered
1,000 ng/kg. Due to the lack of eruption of the third molar in the majority of Line B and C
control females (only 30% erupted), however, the effects of TCDD on third molar eruption could
only be evaluated in Line A female offspring (with 94% eruption).  There was a dose-dependent
decrease in the eruption of the lower third molar in Line A female offspring with a significant
(p < 0.05) decrease observed in the 300 and 1,000 ng/kg dose groups. In the male offspring, any
third molar that developed erupted by PND 70. The mesiodistal length of the existing lower
third molar was reduced in a dose-dependent manner in both genders of all three rat lines. In
Line A and C females, the decrease was significant (p < 0.05) at all doses.  The size of the
second molars was  also significantly decreased with 1,000 ng/kg (p < 0.05) in  all but Line C
males.
       A developmental LOAEL for TCDD of 30 ng/kg for maternal exposure on GD 15 is
established for this  study, based on impaired tooth development (significantly reduced
mesiodistal length of the lower third molar by approximately 12% to 38% \p < 0.05]). A
NOAEL could not be determined.
D.l.2.6.    Keller et al (2008a; 2008b; 2007c)
       Keller et al. (2008a; 2008b: 2007c) conducted three separate experiments to assess the
impact of TCDD on molar tooth development using different mouse strains. In Experiment 1,
Keller et al. (2007c) used six inbred mouse strains (C57BL/6J, BALB/cByJ, A/J, CBA/J,
C3H/HeJ, and C57BL/10J) known to possess high affinity ligand-binding aryl hydrocarbon
receptor alleles (&), two with bl alleles (C57BL/6J and CBA/J), and four with b2 alleles
(BALB/cByJ, A/J, C3H/HeJ, and CBA/J). Females (number not specified) from each strain
were mated with males of the same strain. On GD 13, each pregnant female was assigned to one
of the four dose groups and treated with 0, 10,  100, or 1,000 ng TCDD/kg BW via oral gavage.
The control group received corn oil. GD 13 was chosen for dosing because the first
morphological signs of tooth development occur on GD 11.  The first visible signs  of the Ml
(molar) occur on GDs 13-14 followed by final cuspal morphology, which is determined on
GD 15.  The Fl  offspring of females from each strain were weaned and separated by sex at PND
                                         D-23

-------
28 and were euthanized at PND 70. Each Fl mouse was examined for the presence or absence
of both maxillary (M3) and mandibular third molars (Ms) on both the left and right sides. In
addition, all mice were scored as either normal or variant in MI morphology for both molar
rows.
      In Experiment 2 (Keller et al., 2008b), dams from six inbred mouse strains (C57BL/6J,
BALB/cByJ, A/J, CBA/J, C3H/HeJ, and C57BL/10J) were orally dosed on GD 13 with 0, 10,
100, or 1,000 ng TCDD/kg BW in corn oil. GD 13 was used as the dosing day because it
coincided with the formation of Meckel's cartilage (a major signal center) in the mouse mandible
that is followed shortly by intramembranous bone formation on GD 15.  The A/J mouse strain
was abandoned because the authors had difficulty rearing the offspring from this strain.  All
offspring (n = 4 or 5 per treatment group) from the remaining strains were euthanized at 70 days
of age. Mandible size and shape from all selected offspring were examined using geometric
morphometric methods to assess the impact of TCDD exposure.
      In Experiment 3 (Keller et al., 2008a), dams from six inbred mouse strains (C57BL/6J,
BALB/cByJ, A/J, C3H/HeJ, CBA/J, and C57BL/10J) were treated with a single  oral dose of 0,
10,  100, or 1,000 ng TCDD/kg-BW in corn oil.  GD 13 was chosen as the dosing day because the
first visible signs of the first molar (Mi) occurs on GDs 13-14 and the final cuspal morphology
(the pattern of projections on the chewing surface of the tooth) is not determined until after
GD 15.  Similar to Experiment 2, the A/J mouse strain was abandoned due to difficulty in rearing
offspring. All offspring (n = 107-110 in each of the five strains for all treatment groups) were
euthanized at 70 days of age and their molar size, shape, and asymmetry traits were examined
using geometric morphometric methods.
      In Experiment 1, all four MSS were present in all dose groups in mice from C57BL/6J,
BALB/cByJ, and C57BL/10J strains. A similar response was observed in the A/J strain mice
with only 3  of 51 Fl mice exhibiting missing third molars. Approximately one-third of the mice
from the CBA/J and C3H/HeJ strains, however, were missing at least one M3 or Mj molar. The
numbers of CBA/J mice missing one or both Mj or M3 molars were 0/29, 2/21, 6/29, and 30/30
in the 0, 10, 100, and 1,000 ng/kg groups, respectively. In the C3H/Hej animals, the numbers
missing one or both molars were 1/24, 3/28, 1/26, and 30/36, respectively.
      Maternal TCDD exposure was also found to affect the frequency of MI variants, but only
in the C57BL/10J strain, and the dose-response relationship was nonmonotonic.  The proportions
                                         D-24

-------
of variants observed in the 0, 10, 100, and 1,000 ng/kg dose groups were 33, 68, 59, and 58%,
respectively.
       A LOAEL for TCDD of 10 ng/kg maternal exposure on GD 13 is identified for this study
for increased incidence (33%) of the MI variant in the C57BL/10J mouse strain.  A NOAEL
cannot be determined in this study.
       In Experiment 2, TCDD exposure of dams did not affect offspring survival or 10-week
body weight in any of the inbred mouse strains used. ANOVA indicated that although mandible
size in both male and female offspring varied significantly (p < 0.0001) among strains, it was not
affected by TCDD exposure. In contrast, analysis of covariance indicated that TCDD exposure
significantly (p = 0.0033) decreased the mandible size in male offspring in the C3H/HeJ strain at
all treatment groups. The mean mandible size was similar across all treatment groups in both
sexes in all strains with male offspring exhibiting larger mandibles compared to females.  Males
in the C3H/HeJ strain exhibited a significant (level not reported) downward trend in mandible
size throughout all treatment groups.  Females in the C3H strain also showed a similar trend in
mandible size—but the trend was not significant.  ANOVA on mandible shape indicated that
males had significantly (p < 0.0001) different mandible shape in strain x treatment groups. In
contrast, in female offspring, although the mandible shape was significantly (p <  0.0001)
different due to strains, treatment groups, and litter, the strain  x treatment interaction was not
significant. Male offspring from the C3H/HeJ and C57BL/6J  mouse strains appear to be more
sensitive to TCDD than BALB/cByJ or CBA/J mice, with the  C57BL/10J strain exhibiting
intermediate sensitivity.  In addition to these analyses, Procrustes distance analysis also indicated
that C3H/HeJ mice had the greatest response to the highest dose of TCDD, followed by the
C57BL/6J strain. Female offspring in the C3H/HeJ and C57BL/6J strains also exhibited the
largest change in Procrustes distance with TCDD exposure. This trend, however, was not
statistically significant (p  = 0.29).
       A LOAEL for TCDD of 10 ng/kg maternal exposure on GD 13 was identified for this
study for significantly (p = 0.0033) decreased mandible shape and size in male C3H/HeJ mice.
A NOAEL cannot be determined in this study.
       In Experiment 3, the effect of TCDD exposure on offspring survival or body weight was
not reported.  Three-way ANOVA results showed significant (p < 0.0001) differences in molar
size among strains, sexes, and litters—but not among treatment groups.  Molar size difference in
                                         D-25

-------
sex x strain interaction was significant (p = 0.03), whereas differences in sex x treatment and
sex x strain x treatment were not significant. Additionally, molar size in treatment x strain
interaction also was not statistically significant.  Based on these results, the authors reported that
molar size varied significantly (p < 0.0001) among all five strains tested, with all strains
exhibiting similar trends in all four treatment groups.  Strain differences in molar size were more
apparent in male offspring. A hormesis-like trend in molar size was observed in all strains
(except in BALBc/ByJ) and sexes with an increase at the 100 ng/kg dose and a decrease in the
1,000 ng/kg dose. In addition to lack of difference in molar size for all treatment groups in all
strains, fluctuating asymmetry in molar size also did not increase with increasing doses of
TCDD.
       In  contrast to these results on molar size, the Procrustes ANOVA indicated that molar
shape was significantly (p < 0.0001) affected by strain, sex, treatment, and litter size. Molar
shape in sex x strain and sex x strain  x treatment interactions was also highly significant
(p < 0.0001). Based on these results, the authors concluded that differences between males and
females varied based on the strain, and that the effect of TCDD exposure on each strain also
differed for male and female offspring.  Because molar shape in treatment x strain interaction
was significant (p < 0.0001), differences in molar shape between the three treatment groups and
the control group were analyzed for each strain using nonorthogonal contrasts.  In male
offspring,  contrasts between the control group and 1,000 ng/kg were statistically significant only
in the C3H/HeJ (p < 0.0001) and CBA/J (p < 0.03) strains. These results suggest that these
two strains are most susceptible to TCDD effect on molar shape, and similar results were
observed in female offspring of these two strains. The contrast in molar shape between the
control and the 100 ng/kg treatment group for the female C57BL/6J mice also was statistically
significant (p = 0.0096).  On the whole, when considering Procrustes distance results for molar
shape, the C3H/HeJ male offspring had the largest response at the low and high doses, while the
female offspring had the largest response at low and mid doses. This observation in male
C3H/HeJ mice is consistent with that of TCDD-induced changes in mandible size from Keller
et al. (2008b).
       A LOAEL for TCDD of 10 ng/kg maternal exposure on GD 13 is identified for this study
for significant (p < 0.0001) differences in molar shape in male C3H/HeJ mice.  A NOAEL
cannot be  determined  in this study.
                                          D-26

-------
D.l.2.7.    Kuchiiwa et al (2002)
       Kuchiiwa et al. (2002) studied the impact of in utero and lactational TCDD exposure on
serotonin-immunoreactive neurons in raphe nuclei on Fl male mouse offspring. Twenty-one
adult female ddY mice (seven per treatment group) were administered TCDD (99.1% purity) by
oral gavage once per week, for 8 weeks, at doses of 0, 4.9, or 490 ng/kg (0, 0.7, or 70 ng/kg-day
average daily dose; administered doses divided by 7) or an equivalent volume of olive oil vehicle
(6.7 mL/kg) by gavage. Immediately following the final treatment, the mice were housed with
untreated male mice for mating.  At approximately 20-21 days after mating, 3 female mice from
each dose group, including the control group gave birth to 10-12 offspring. One day after birth,
each litter was culled  to 10 offspring to accommodate similar lactational TCDD exposure.  On
PND 28, the offspring were weaned, and three offspring from each TCDD exposed group and
the control  group were selected for an immunocytochemical examination at 42 days of age.
Following sacrifice of these offspring, the brain of each animal was removed and every second
serial section of the brain was processed for immunocytochemistry. In addition to the serial
sections of the brain, cells from 18 offspring (6 males per treatment group) were used to assess
the number of cells in the  dorsal and median raphe nucleus, the supralemniscal area, and the
Nucleus raphe magnus.
       Examination of external morphology, birth, and postnatal body weights indicated that
there were no differences between the male TCDD-exposed offspring and the control male
offspring.  TCDD-exposed males, however, were aggressive toward other normal  mice and were
also hypersensitive to soft touch.
       Serotonin-immunoreactive neurons were found to be distributed throughout the entire
brainstem in 42-day-old males, and the general pattern in the TCDD-exposed animals was
consistent with those  observed in control male offspring.  Serotonergic neurons were identified
and counted in the caudal  linear nucleus, the median and dorsal raphe nucleus, Nucleus raphe
pontis, interpeduncular nucleus, supralemniscal area, pedunculopontine segmental nuclei, deep
mensencephalic nucleus, Nucleus raphe magnus, pallidus, and obscurus, dorsal  and medial to the
facial nucleus and the ventrolateral medulla. Results from computerized cell counts (n = 6)
showed an  average of 1,573.3 immunoreactive neurons in the raphe nuclei from the control
group versus 716.3 and 419.8 neurons in the low- and high-dose offspring, respectively.  The
                                         D-27

-------
numbers of immunoreactive neurons in the individual raphe nuclei (dorsalis, medianus, magnus,
and B9) from the TCDD-exposed offspring were significantly (p < 0.01) lower than control
values, with the degree of reduction being dose-related.
       A LOAEL of 0.7 ng/kg-day for an 8-week exposure duration is identified in this study for
a significantly (p < 0.01) lower number of serotonin-immunoreactive neurons in the raphe nuclei
of male offspring. A NOAEL cannot be determined for this study.


D.l.2.8.   Li et al (2006)
       Pregnant and pseudopregnant (obtained by mating normal estrous female mice with
vasectomized male mice) NIH mice (10 per treatment group) were exposed to 0, 2, 50, or
100 ng/kg-day of TCDD (purity 99%) during early gestation (GDs 1-8), preimplantation
(GDs 1-3), or peri-implantation to postimplantation (GDs 4-8) (Li et al.. 2006). On GD 9,
animals were evaluated. The two highest  TCDD doses (50 and 100 ng/kg-day) caused
significant (p < 0.05) early embryo loss independent of gestational exposure time.  At
100 ng/kg-day, however, the embryo loss  was greater when administered during GDs  1-8 or
GDs 1-3 compared to GDs 4-8 (p < 0.01). Uterine weight was significantly decreased in the
pseudopregnant mice when administered 50 or 100 ng/kg-day TCDD during GDs 1-8
(p < 0.001) or 1-3 (p < 0.01), but was only decreased at 100 ng/kg-day in pseudopregnant mice
when administered during GDs 4-8 (p < 0.01). Estradiol levels were increased at all TCDD
treatment levels (100% at the lowest dose), but statistical significance was not indicated. All
doses at all treatment times resulted in a significant reduction (p < 0.01) in  serum progesterone
levels, with a 45% decrease at the lowest dose. Because the hormone effects were observed
following 4 days of treatment, the nominal doses were averaged over the entire test period of
8 days prior to measurement. The resulting average daily doses of TCDD were 0, 1, 25, and
50 ng/kg-day.
       A LOAEL of 2 ng/kg-day administered for 4 to 8 days is established in this study for a
significant (p <  0.01) decrease in progesterone (45% above control) and an approximate twofold
increase in estradiol levels (significance not indicated). A NOAEL cannot be determined.
                                         D-28

-------
D.l.2.9.    Markowski et al (2001)
       Pregnant Holtzman rats (4-7 per treatment group) were administered a single gavage
dose of 0, 20, 60, or 180 ng/kg TCDD (purity not specified) in olive oil on GD 18 (Markowski et
al., 2001). One female rat from each liter (4-7 per treatment group) was assigned to training on
a wheel apparatus to respond on a lever for brief opportunities to run.  Once animals responded
to an FR1 schedule of reinforcement, the requirement for lever pressing was increased to FR2,
FR5, FR10, FR20, and FR30 schedules.  After each training session, the estrous cycle stage was
determined.  Maternal body weight, length of gestation, number of pups per litter, and sex
distribution within litters were unaffected by treatment. For each of the FR schedules, there was
a significant dose-related (p = 0.0001) decrease in the number of earned run opportunities, lever
response  rate, and total number of revolutions in the wheel in the adult female offspring. There
was no correlation between estrous cycle and responding for access to wheel running.
       The developmental LOAEL for this study is a single dose of 20 ng/kg administered on
GD 18 for neurobehavioral effects. A NOAEL cannot be determined for this study.


D. 1.2.10.   Miettinen et al (2006)
       Miettinen et al. (2006) administered a single oral dose of 0, 30, 100, 300, or 1,000 ng/kg
TCDD (purity >99%) in corn oil on GD 15 to pregnant Line C rats. The offspring (24-32 per
treatment group) were assigned to a sugar-rich cariogenic diet (via feed and drinking water) and
were orally inoculated three separate times with fresh cultures of Streptococcus mutans.
Three control groups varied with regard to TCDD exposure and administration of a cariogenic
diet.  Two of the control groups received no TCDD, and the offspring were either maintained on
a normal  diet without inoculation with S. mutans (Cl; n = 48) or were given the cariogenic diet
with S. mutans inoculation (C2; n = 42).  The final control group was maternally exposed to
1,000 ng/kg TCDD with offspring fed a normal diet without S. mutans inoculation (C3;  n = 12).
TCDD did not affect the maternal or offspring body weight.  Survival of the offspring was
reduced in the 1,000 ng/kg dose group (50-58% survival  compared to 83-95% in  Cl and C2,
respectively). All offspring administered 1,000 ng/kg were missing all lower third molars.
Two animals (8%) in the 100 ng/kg group were missing one of their lower third molars.  All
doses—except the 100 ng/kg dose—  caused a significant (p < 0.05) increase in the number of
caries lesions compared to group C2 (60, 79, 76, 83, and 91% in the C2, 30, 100, 300, and
                                         D-29

-------
1,000 ng/kg groups, respectively). Group C3 (1,000 ng/kg TCDD exposure, normal diet)
animals also had increased caries lesions compared to Cl (8 vs. 0%, respectively).  There were
no detectable changes in tooth mineral composition that could explain the increase in caries
susceptibility.
       The developmental LOAEL from this study is a single dose of 30 ng/kg administered on
GD 15 based on the significant (p < 0.05) increase in dental  caries in pups (30% above control).
A NOAEL cannot be determined from this study.


D. 1.2.11.  Nohara et al  (2000b)
       Pregnant Holtzman rats were administered 0, 12.5, 50, 200, or 800 ng/kg TCDD in corn
oil by gavage on GD 15 (Nohara et al., 2000b).  On PND 2,  five males were randomly selected
from each litter and dose group. TCDD was detected in the  thymus, spleen, and bone marrow of
the male pups on PND 21  and PND 49. TCDD was still detected in the thymus and spleen on
PND  120 but the levels decreased over time.  The TCDD concentration was highest in the
thymus at all time points.  There were no changes in the body, thymus, or spleen weights of the
male offspring on PND 5,  PND 21, PND 49, or PND 120. On PND 5, there was a 200-fold
increase in CYP1 Al in the thymus of the high-dose male pups.  CYP1 Al was only slightly
increased in the spleen.  This induction decreased through PND 49.  There was a slight (not
statistically significant) dose-dependent decrease in thymus  cellularity in the male offspring at
PND  120. Spleen cellularity at PND 49 decreased in a dose-dependent manner (15-50% of the
control), with a statistically significant (p < 0.05) decrease observed in the high-dose group. A
slight but not significant reduction in spleen cellularity was noted in the high-dose group at
PND 21.  The same effect was not observed at PND 120, nor was there any change in the percent
of B or T cells  in the spleen. No changes in cytokine levels  were observed in the 800-ng/kg
group.
       Although a change in spleen cellularity on PND 49 (puberty) was observed, this effect
was transient, and there were no coexisting changes in the percentage of splenic lymphocytes,
spleen weight,  and cytokine levels.  Therefore, a developmental NOAEL of a single dose of
800 ng/kg administered on GD 15 is identified for this study. A LOAEL is not established.
                                         D-30

-------
D.l.2.12.   Ohsako et al (2001)
       Pregnant Holtzman rats (6 per treatment group) were administered 0, 12.5, 50, 200, or
800 ng/kg TCDD (purity >99.5%) in corn oil by gavage on GD 15 (Ohsako et al.. 2001). On
PND 2, five males were randomly selected from each litter. Two male offspring from each litter
were sacrificed on PND 49 and PND 120.  Neither maternal nor male offspring body weight was
affected by TCDD treatment. TCDD was detected in both the fat and testes at all dose levels
(including controls) with highest levels found in fat.  There were no apparent treatment-related
effects on testicular weight, epididymal weight, daily sperm production, cauda epididymal sperm
reserves, luteinizing hormone, follicle stimulating hormone, or testosterone levels.  There was,
however, a clear dose-dependent decrease in urogenital complex weight and ventral prostate
weight at both PND 49 and PND 120. For male offspring, statistically significant (p < 0.05)
decreases were noted in urogenital complex weight at PND 120 in the 200 and 800 ng/kg groups,
in ventral prostate weight at PND 49 in 800 ng/kg group, and at PND 120 in the 200 and
800 ng/kg groups.  There was also a dose-dependent decrease in anogenital distance (the length
between the base of the genital tubercle and the anterior edge of the anus); the decrease was not
statistically significant at PND 49. At PND 120, however, male offspring in all but the lowest
dose group had significantly (p < 0.05) reduced anogenital distance compared to the control
animals.  There was also a dose-dependent increase in 5aR-II mRNA expression in the ventral
prostate on PND 49 with significant increases (p < 0.05) in the 200 and 800 ng/kg animals.
There was a significant (p < 0.01) decrease in the androgen receptor mRNA in the ventral
prostate on PND 49 at all doses tested. Similar effects were not observed on PND 120 or in the
caput  epididymis on PND 49.
       The developmental LOAEL for this study is a single dose of 50 ng/kg administered on
GD 15 for significantly (p < 0.01) reduced anogenital distance in male offspring (approximately
14%).  The NOAEL for this study is 12.5 ng/kg.


D.l.2.13.   Schantz et al (1996)
       Schantz et al.  (1996) studied the impact of in utero TCDD exposure on  spatial learning in
male and female pups. Groups of pregnant Harlan Sprague-Dawley rats (n = 108, divided into
4 cohorts; number of animals in each TCDD group approximately 4 per treatment group) were
dosed via gavage with 0, 25, or 100 ng/kg-day TCDD (purity >98%) in corn oil on GDs  10-16.
                                         D-31

-------
On the day of birth (PND 0), the pups were examined for gross abnormalities and the number of
live pups, weight, and sex were recorded for each litter. On PND 2, litters were culled to
eight animals and were balanced to include four males and four females whenever possible. To
minimize litter-size effects, litters with fewer than five pups were excluded from the study.  The
exclusion of these litters resulted in 10-11 litters per treatment group.  Pups were weaned on
PND 21 and one male and one female pup from each litter were maintained for the learning tests.
Pups were tested 5 days per week for spatial learning and memory in a radial arm maze and a
T-maze.  A radial arm maze working memory test and a T-maze DSA task were used a part of
the testing process.
       TCDD treatment did not affect dam gestational weight gain, dam liver weight, gestation
length, litter size, percentage of live births, birth weight, or postnatal growth of the pups
observed during the course of the study. Exposed pups, however, exhibited some signs of
toxicity in all exposure groups.  Thymus weight was decreased and liver weight was increased in
the 100 ng/kg-day TCDD dose group. Also, liver microsomal 7-ethoxyresorufm-O-deethylase
(EROD) activity was markedly induced in pups from both the 25 and 100 ng/kg-day dose
groups. In the radial maze test, rats from all TCDD exposure groups displayed a significant
(p < 0.01) learning behavior as shown by progressively fewer errors from the first block of
sessions through the fourth session. The treatment by sex and treatment by session block
interactions were not significant.  Comparisons between the average number of errors per session
block in the TCDD-exposed and control group indicated that both the 25 and the 100 ng/kg-day
dose groups made significantly (p < 0.05 andp < 0.001, respectively) fewer errors compared to
the control group. TCDD did not significantly affect adjacent arm selection behavior as
measured by C statistic; hence the reduction in errors observed did not appear to be accounted
for by an increased tendency to run into adjacent arms. Female pups had a significant (p < 0.05)
shorter radial arm maze latency, however, compared to the male pups.  In the T-maze test,
TCDD did not significantly affect the percent of correct performance.  All exposure groups
performed best at the shortest delay, which showed a decline as the length of the intertrial delay
interval was increased.  Additionally, all treated groups improved their performance over a
three-block session period. This finding indicated that animals in all groups could learn the task.
These observations were confirmed by a highly significant main effect of delay (p < 0.001) and
highly significant main effect of session blocks (p < 0.001). At the shortest 15-second delay,
                                          D-32

-------
average percent correct performance increased from 75 to 92%, while at the longest 40-second
delay, the average percent correct performance increased from 62 to 82%. A significant
(p < 0.05) main effect of exposure was evident in latency to respond in the T-maze.
Comparisons of the exposed group to control group, however, indicated that none of the
individual exposure groups differed significantly from the controls.  Because no clear pattern
was observed in the various exposure groups, differences in latency to respond had no impact on
learning of the task.
       Based on these results, the study authors state that the fact TCDD seems to have a
facilitatory effect on radial arm maze learning in rats should be interpreted with caution and
needs further evaluation using different and more varied learning tasks.  No lexicologically
adverse endpoints were concurrently examined.  Thus, a LOAEL and a NOAEL cannot be
determined for this study.


D.l.2.14.  Seo et al (1995)
       To study developmental effects of TCDD on thyroid hormone levels, time-mated female
Sprague-Dawley rat dams (n = 10-14/treatment group) were administered either 25 or
100 ng/kg-day of TCDD (>98% pure) in corn oil via gavage from GDs 10-16. Vehicle controls
received equivalent amounts of corn oil. The study also investigated PCB treatment outcomes.
At birth, pups were weighed and grossly examined for abnormalities. At 2 days of age, litters
with fewer than 5 pups were excluded from the analysis and the remaining litters were culled to
4 males and 4 females. Each treatment group contained 10 or 11  litters. Pups remained with the
dams until weaning.  At weaning, 4-6 pups were retained for neurobehavioral tests (which were
not reported as part of this study). The remaining offspring were sacrificed, which provided
5-9 litters per treatment group. Data were collected from one male and one female when
possible. No signs of toxicity  were evident in the dams; measurements on dams included
gestational weight gain, liver weight, litter size,  and live births. Pup birth weight and weaning
weight were unaffected by treatment.  In pups sacrificed at weaning  (21 days old), a significant
(p < 0.05) decrease occurred in thymus weight for the high-dose group, but not in thyroid, liver,
or brain weight. A significant (p < 0.05) decrease (20.4%) was observed in thyroxine (T4) in
high-dose females. Thyroid stimulating hormone and triiodothyroxine (T3) were unaffected by
treatment. Uridine diphosphate (UDP)-glucuronosyltransferase activity towards 4-nitrophenol
                                         D-33

-------
significantly (p < 0.05) increased in both treatment groups over control values, and the increase
in the high-dose group was significantly (p < 0.05) greater than in the low-dose group. Liver
microsomal EROD activity was significantly (p < 0.05) increased in both treatment groups, but
is considered to be an adaptive response and not adverse.
       A LOAEL of 100 ng/kg-day for decreased thymus weights and decreased thyroxine is
identified for this study.  A NOAEL of 25  ng/kg-day is established.


D.l.2.15.   Sparschu et al (1971)
       Sparschu et al.  (1971) studied the teratogenic and developmental effects of TCDD
exposure in rats.  Groups of pregnant Sprague-Dawley rats were dosed via gavage with 0
(n = 31), 30, 125, 500, 2,000, or 8,000 (n = 10-14 per group) ng/kg-day TCDD (purity 91%) in
corn oil on GDs 6-15. Maternal body weights were assessed on GD 0, 6, 13, and 20, and all
dams were observed for  clinical signs of toxicity throughout the test period.  On GD 20, the
dams were sacrificed and evaluated for the numbers of pregnancies, implantation sites, corpora
lutea,  and viable  and dead fetuses. All  removed fetuses were individually weighed, sexed, and
examined for external malformations as well as intestinal hemorrhage.  One-third of the fetuses
were examined for skeletal alterations,  and two-thirds for visceral abnormalities.
       Clinical signs of toxicity in the dams included vaginal hemorrhage at >2,000 ng/kg-day at
various intervals  throughout gestation.  The study authors described dams in the 8,000 ng/kg-day
dose group as "thin" and showing "signs of debilitation."  Maternal body weight gain was
significantly (p < 0.01) reduced compared to control values at doses >500 ng/kg-day on GD 13,
as well as at 500  (p < 0.01), 2,000 (p < 0.001), and 8,000 ng/kg-day (p < 0.001) on GD 20. No
significant differences were observed in fertility or the number of implantation sites or corpora
lutea at any dose tested.  The mean number of viable fetuses per litter was significantly
(p < 0.05) decreased at 500 ng/kg-day compared to control.  Only 7 viable fetuses were found
and occurred in 4 of the  11 total litters examined in the 2,000 ng/kg-day dose group, and there
were no viable fetuses in the 8,000 ng/kg-day dose group. The mean number of resorption sites
per litter was significantly increased at 500 (p < 0.05), 2,000 (p < 0.001),  and 8,000ng/kg-day
(p< 0.001).
       No significant differences were observed in the fetal sex ratios at any dose tested.  Mean
fetal body weight was  significantly decreased compared to control values at 125 (p < 0.01), 500
                                          D-34

-------
(p < 0.05), and 2,000 ng/kg-day (p < 0.001) for males, and at 125 (p < 0.01) and 2,000 ng/kg-day
(p < 0.001) for females. Incidence of intestinal hemorrhage was increased on a per-fetus and
per-litter basis at doses >125 ng/kg-day. The incidence of tail and limb malformations was not
consistently increased over that of control. With respect to soft tissue abnormalities,
subcutaneous edema was observed at doses >125 ng/kg-day on a per fetus basis. Skeletal
abnormalities included delayed ossification of sternebrae and skull bones and wavy thirteenth
ribs, but these findings occurred throughout the various groups independent of dose and also in
controls.
       The developmental LOAEL for TCDD of 125 ng/kg-day was  identified for decreased
body weight in dams and male fetuses,  as well as fetal intestinal hemorrhage and subcutaneous
edema. The developmental NOAEL in this study is 30 ng/kg-day.  The maternal NOAEL and
LOAEL were 125 and 500 ng/kg-day, respectively, for decreased body weight gain.


D.l.2.16.   Smith et al (1976)
       Smith et al. (1976) studied the teratogenic and developmental  effects of TCDD exposure
in mice. Groups of pregnant CF-1 mice were dosed via gavage with 0, 1.0,  10, 100, 1,000, or
3,000 (n = 14-41 per group) ng/kg-day TCDD (purity not specified) in corn oil on GDs 6-15.
Maternal body weights were assessed on GD 6,  10, 16, and 18, and all dams were observed for
clinical signs of toxicity throughout the test period. On GD 18, the dams were sacrificed and
evaluated for the number of live, dead, and resorbed fetuses, and the livers were also removed
and weighed. All removed fetuses were individually weighed, sexed, measured, and examined
for external malformations.  One-third of each litter was examined for soft tissue anomalies, and
all the fetuses were examined for skeletal anomalies. The litter was considered the experimental
unit of treatment and observation.
       No significant differences were  observed in maternal body weight at any time during
gestation at any dose tested. Relative liver weight in dams was significantly (p < 0.05)  increased
in the 3,000 ng/kg-day dose group (13%) compared to control, but absolute liver weights were
not significantly changed at any dose tested. The percentage of resorptions per implantations
was significantly (p < 0.05) increased only at the 1,000 ng/kg-day dose compared to control.
There were no significant differences from control values at any dose in implantation sites per
litter,  percentage of litters with resorptions, sex ratio, fetal body weight, and fetal length.
                                         D-35

-------
       With respect to fetal anomalies among the litters, there was a significantly (p < 0.05)
increased incidence of cleft palate in the 1,000 and 3,000 ng/kg-day dose groups compared to
that of control. Additionally, there was a significantly (p < 0.05) increased incidence of litters
with bilateral dilated renal pelvis in the 3,000 ng/kg-day group compared controls. Although not
statistically significant, the incidence of exencephaly was greatest at the lowest dose level
(1.0 ng/kg-day).  Because of this observation, an additional group of 30 mice were run through
the GD 6-15 treatment protocol at 1.0 ng/kg-day with another control group run concurrently
(n = 24).  In this exposure, the incidence of exencephaly in the litters from treated dams was
comparable to that in the controls.  The percentage of resorptions per implantations was
increased (12%, p = 0.048) over that of controls (8%); however, this effect was not observed in
the original 1.0 ng/kg-day exposure and the incidence was similar to that of the original control
animals (11%).
       A maternal LOAEL of 3,000 ng/kg-day was identified for increased relative liver weight
in mouse dams. The maternal NOAEL is 1,000 ng/kg-day. A developmental LOAEL of
1,000 ng/kg-day was identified for increased incidence of cleft palate.  The developmental
NOAEL is 100 ng/kg-day.


D. 1.2.17.   Simanainen et al (2004b)
       Simanainen et al. (2004b) studied the impact of in utero and lactational TCDD exposure
on the male reproductive system in three rat lines that are differentially sensitive to TCDD.
Groups of 5 to 8 pregnant Line A, B, and C C57BL/6N CYP1A2 dams were given a single dose
of 0, 30, 100, 300, or 1,000 ng/kg of TCDD (purity >99%) in corn oil on GD 15 via oral gavage.
Control animals were similarly dosed with a corn oil vehicle. One day after birth, litters were
randomly culled to include three males and three females to allow uniform postnatal exposure.
Offspring were weaned on PND 28. Dam and pup viabilities were monitored throughout the
study.  Pup body weights were determined on PNDs 1, 4, 7, 14, and 28. Anogenital distance and
crown-to-rump length were measured on PNDs 1 and 4. On Day 70, pups were sacrificed and
trunk blood was collected.  Serum was collected for testosterone analysis. The testes, cauda of
the right epididymis, ventral prostrate, seminal vesicles, and thymus was dissected and weighed.
Absolute and relative organ weights were determined, and cauda epididymis and testes were also
preserved for sperm count analysis.
                                          D-36

-------
       TCDD caused no mortality or overt signs of toxicity to the dams. Pup survival from
implantation to the day after birth also was not affected by TCDD exposure.  Survival from the
day of implantation to the day after birth, however, was uncharacteristically lower in control
Line B rats (41%), resulting in a significant difference compared with the two lowest doses (30
and 100 ng/mg TCDD).  The average survival percentage in the controls for Line A, B, and C
rats was 85% (range 80-86%); 64% (41-86%); and 74% (63-85%); respectively. Percentage of
male pup survival in each line between PND 1 and PND 28 was 99% except for Line B males
exposed to 30 ng/kg TCDD and Line C males exposed to 30 or 100 ng/kg, where male survival
rate averaged 81% (range 81-83%). On PND 70, a significant (p < 0.05) reduction in body
weight was observed only in Line B and C rats at 1,000 ng/kg. In pups exposed to 1,000 ng/kg
TCDD, both absolute and relative weight of the ventral, anterior, and dorsolateral prostrate
decreased in all three lines at most postnatal time points measured. The change was most
consistent and significant (p < 0.05) in the ventral lobe. Animals exposed to 1,000 ng/kg TCDD
had an average decrease in absolute weight of the anterior prostrate of 37, 32, and 34% in
Lines A, B and C, respectively. Additionally, the average dorsolateral prostrate weight was also
decreased by 34,  28, and 39% in Lines A, B, and C, respectively.  The effect on the ventral
prostrate was reversible with the only significant (p < 0.05) decrease in weight observed in
Line B rats at PND 70 in the 1,000 ng/kg TCDD dose group. The authors reported that TCDD
had no consistent effects on the weight of seminal vesicles.  The absolute weights of the testis
and epididymis showed a significant (p < 0.05) increase on PNDs 28-49, but the relative testis,
epididymis, and cauda epididymis weights remained unchanged. In pups exposed to
1,000 ng/kg TCDD,  severe malformation, including small caput and cauda and degeneration of
corpus epididymis, was observed. Malformations in the epididymis were observed in 6 of
44 Line C male rat offspring and 3 of 47 Line A male rat offspring. In Line A, B, and C rats at
PND 70in the 1,000  ng/kg TCDD dose group, daily sperm production was reduced by 9, 25,  and
36% and cauda epididymal sperm reserves were reduced by 18, 42, and 49%, respectively.
Daily sperm  reduction (17%) was significant (p < 0.05) in Line C rats at a TCDD dose of
300 ng/kg and in  Line B  and C rats at 1,000 ng/kg. A reduction in cauda epididymal sperm
reserves (25%) was significant (p < 0.05) in Line C rats at 300 and 1,000 ng/kg TCDD.
                                         D-37

-------
       A LOAEL for TCDD of 300 ng/kg is identified for reduction in daily sperm production
and cauda epididymal sperm reserves in Line C rats.  A NOAEL of 100 ng/kg is identified for
this study.
D. 1.2.18.   Sugita-Konishi et al (2003)
       Sugita-Konishi et al. (2003) examined the immunotoxic effects of lactational exposure to
TCDD in newborn mice. Eight pregnant female C57BL/6NCji mice were administered 0, 1.8, or
18 ng/L of TCDD via drinking water from parturition to weaning of the offspring (for a total of
17 days). Based on an average water intake of 14-16 mL/day, the average daily intake of TCDD
for the dams was 1.14 and  11.3 ng/kg-day in the low- and high-dose groups, respectively. In
male offspring sacrificed at weaning (21 days after birth), there was a statistically significant
(p < 0.05) decrease in relative spleen weight and a statistically significant (p < 0.005) increase in
thymic CD4+ cells in the high-dose group.  The changes in relative spleen weight and thymic
CD4+ cells were dose related, but effects in the low-dose group did not achieve statistical
significance.  Changes in spleen weight and CD4+  cell numbers were not observed in the female
offspring.  In a separate experiment, offspring infected with Listeria monocytogenes following
lactational TCDD exposure exhibited a statistically significant increase in serum tumor necrosis
factor alpha (TNF-a) 2 days after infection  in both  sexes in the low- (p < 0.05) and high-dose
(p < 0.005) groups. There  was also a statistically significant increase in serum interferon gamma
in Listeria-infected high-dose females (p <  0.05). The number of bacteria in the spleen was also
significantly increased (p < 0.05) 2 days after infection in the high-dose females compared to the
controls, but not in males.  Listeria levels in the spleen returned to control levels by 4 days after
infection in both sexes.
       Based on these results, a LOAEL for TCDD of 11.3 ng/kg-day following a  17 day
exposure to dams was identified for significantly (p < 0.05) decreased spleen weight (in male
pups), a significant (p < 0.005) increase in thymic CD4+ cells (in male pups), and for increased
susceptibility to Listeria monocytogenes (in male and female pups).  The NOAEL for this study
is 1.14 ng/kg-day.
                                          D-38

-------
D.1.3.   Acute Studies
D. 1.3.1.    Burleson et al (1996)
       Burleson et al. (1996) studied the impact of TCDD exposure on mice that were
challenged with the influenza virus 7 days after treatment with TCDD. Groups of 8-week-old
female B6C3Fi mice (n = 20, 2 replicate groups) were treated one time with 0, 1,5, 10, 50, 100,
or 6,000 ng/kg TCDD (purity >99%, dissolved in corn oil) via oral gavage.  In addition to the
treated groups, randomly selected animals were assigned as a sentinel group and screened for
numerous pathogens. Results of all tests performed on this sentinel group were negative.
Seven days after TCDD treatment, all animals were lightly anesthetized and infected intranasally
with a highly lethal influenza A/Hong Kong/8/68 virus (H3N1; passage 14). The animals were
infected with sufficient H3N1 virus to achieve a 30% mortality rate in the control animals.
Animals were observed for mortality and morbidity for 21 days following viral infection.
Six mice from  each treatment group were sacrificed on Days 3, 9, and 12 postinfection, and
body, thymus,  and wet lung weights were recorded.  Influenza viral liters were examined by
sacrificing eight mice each at 2 hours and at 1, 4, 6, 7, 8, 9,  10, and 11 days  post infection.
       Exposure to TCDD resulted in significantly (p < 0.05) increased mortality in the 10, 50,
and 100 ng/kg  dose groups.  No statistically significant difference in the percentage alive was
observed between these dose groups.  TCDD doses of 1 and 5 ng/kg did not alter mortality in
influenza infected animals. A time-related increase in the wet weights of the lungs in infected
mice as a result of increased edema also was reflected in an increase in the lung
weight-to-body-weight ratio. The study authors stated that this ratio was not altered as a result of
TCDD exposure.  TCDD-only exposures at 1, 10, or 100 ng/kg did not affect thymus weight.
Similarly, animals infected with the influenza virus following TCDD exposure also showed no
loss in thymic weight. Enhanced mortality in TCDD-treated animals was not correlated with an
increase in influenza virus liters. Additionally, animals treated with 1, 10, 100, or 1,000 ng/kg
did not affect pulmonary viral titer assays on Days 6, 7, and 8 postinfection. The authors also
concluded that TCDD did not alter Hong Kong virus replication or clearance.
       Although these results support immunotoxic effects induced by TCDD, the findings were
not reproduced by Nohara et al. (2002a) using the identical study design, and the translation of
these findings to humans is dubious. Thus, no LOAEL/NOAEL was established. A
lowest-observed-adverse level (LOEL) for TCDD of 10 ng/kg for a single exposure is identified
                                         D-39

-------
for significantly (p < 0.05) increased mortality in mice infected 7 days later with the influenza
virus. The no-observed-effect level (NOEL) for this study is 5 ng/kg.


D.I.3.2.    Crofton et al (2005)
       Crofton et al. (2005) studied the impact of TCDD exposure in addition to the impact of
mixtures of thyroid disrupting chemicals and PCBs on serum total thyroxine (TT4)
concentration. Groups of female Long-Evans rats were dosed via oral gavage with 0, 0.1, 3,  10,
30, 100, 300, 1,000, 3,000, or 10,000 ng/kg-day TCDD (purity >99%) in corn oil (n = 14, 6,  12,
6, 6, 6, 6, 6,  6, and 4, respectively) for 4 consecutive days. On the day following the last dose,
animals were sacrificed, trunk blood was collected, and serum obtained via centrifugation was
assayed for TT4 concentration using standard radioimmunoassay methods.
       No visible signs of toxicity or changes in animal body weight as a result of TCDD
exposure were observed.  Serum T4 levels showed a dose-dependent decrease, with the levels
dropping sharply beginning at 100 ng/kg-day dose.  Percent serum T4 levels were 96.3, 98.6,
99.8, 93.3, 70.9, 62.5, 52.7, 54.7, and 49.1% in the 0.1, 3, 10, 30, 100, 300, 1,000, 3,000, and
10,000 ng/kg-day groups, respectively.
       A LOAEL for TCDD of 100 ng/kg-day for 4 consecutive days of exposure is identified in
this study for a reduction in serum T4 levels (70.9% compared to 100% in controls). The
NOAEL for this study is 30 ng/kg-day.


D.I.3.3.    Kitchin and Woods (1979)
       Female Sprague-Dawley rats (nine per control and four per treatment group) were
administered a single dose of 0, 0.6, 2, 4, 20, 60, 200, 600, 2,000, 5,000, or 20,000 ng/kg TCDD
(purity >99%) in corn oil. Animals were sacrificed 3 days after treatment and CYP  level and
benzo[a]pyrene hydroxylase  activity in the liver were measured. A significant (p <  0.05)
increase in cytochrome P450 levels occurred with doses of 600 ng/kg or greater and in
benzo[a]pyrene hydroxylase  activity with doses of 2 ng/kg or greater.  Cytochrome P450 was
significantly (p < 0.05) higher 1 month after a single exposure of 2,000 ng/kg (the only dose
measured), but not after 3 or  6 months. Aryl hydrocarbon hydralase (AHH; p < 0.05) and EROD
                                         D-40

-------
(p < 0.01) were both significantly increased through 3 months after treatment, and although
elevated at 6 months, the results were not significant.
       CYP induction alone is not considered a significant lexicologically adverse effect given
that CYPs are induced as a means of hepatic processing of xenobiotic agents. Thus, no LOAEL
or NOAEL was established for this study because adverse endpoints (e.g., indicators of
hepatotoxicity) were not measured. The acute LOEL, however, is 2 ng/kg based on a significant
(p < 0.05) increase in benzo[a]pyrene hydroxylase activity (37% above control). The NOEL is
0.6  ng/kg.

D.l.3.4.   Lietal (1997)
       Female Sprague-Dawley rats (22 days old;  10 per treatment) were administered a single
oral dose of TCDD (>98% pure) in corn oil via gavage at doses of 3, 10,  30, 100, 300, 1,000,
3,000,  10,000,  or 30,000 ng/kg. Vehicle controls received equivalent amounts of corn oil, while
naive controls were sham-treated only.  In a preliminary time-course study, animals received a
single dose of 10,000 ng/kg and were sacrificed at  1, 2, 4, 8,  16, 24, 48, and 72 hours.  The
time-course study showed two peaks in LH and FSH levels at 1 hour and 24 hours, with a
decrease to control values by 48 hours.  Thus, in the dose-response study, animals were
sacrificed at  1 or 24 hours after treatment, blood was collected, and serum FSH and LH were
measured. The dose-response study demonstrated  that the peak at 1 hour was related to the
vehicle as the peak also occurred in the vehicle controls, but  did not occur in the naive controls.
At 24 hours, FSH was increased at 10 ng/kg and higher (>fourfold increase at 10 ng/kg). Doses
of 10 to 1,000 ng/kg showed similar increases (not all  reached statistical significance; p< 0.05).
A dose-dependent increase occurred for doses>3,000  (p < 0.05) with a maximum increase of
20-fold over the vehicle control. At 24 hours, the LH  response significantly (p < 0.05) increased
only for doses >300 ng/kg with a maximum increase of 15-fold above the vehicle control.  The
study authors calculated an effective dose eliciting 50  percent response of 500 ng/kg for
gonadotropin increase.  The dose-dependent release of LH was confirmed in in vitro studies, but
did  not occur with the same magnitude.  The increase  did not occur in calcium-free medium and
was unrelated to gonadotropin releasing hormone.
       Based on the increase in serum FSH, the LOAEL  was 10 ng/kg and the NOAEL was
3 ng/kg.
                                         D-41

-------
D.l.3.5.    Lucier et al (1986)
       Adult female Sprague-Dawley rats (six per treatment) were administered a single gavage
dose of TCDD (purity not specified) in either corn oil or contaminated soil at doses of 15, 40,
100, 200, 500,  1,000, 2,000, 5,000 (corn oil), or 5,500 (contaminated soil) ng/kg. Animals were
sacrificed 6 days later and livers were removed for analysis.  No clinical signs of acute toxicity
or changes in body weight were observed at any dose. AHH increased in a dose-dependent
manner with significant (p < 0.05) increases observed at 15 ng/kg or greater in corn oil or
40 ng/kg or greater in contaminated soil.  Cytochrome P450 was significantly (p < 0.05)
increased with doses of 1,000 ng/kg or greater in corn oil or 500 ng/kg or greater in contaminated
soil. A dose-dependent increase was observed for UDP glucoronyltransferase (significance of
individual doses not reported), with the results twice as high with corn oil than with
contaminated soil. The authors state that the results indicate bioavailability from soils is 50%.
       Because the association between AHH activity and TCDD-mediated hepatotoxicity is
unknown and no adverse endpoints were measured, a LOAEL or NOAEL was not determined
for this study.  The acute LOEL for this study is  15 ng/kg, based on the significant (p < 0.05)
increase (80%  above control) in AHH. No NOEL is established.


D.l.3.6.    Nohara et al (2002a)
       Male and female B6C3Fi (C57BL/6 x C3H), BALB/c, C57BL/6N, and DBA2 mice
(10-40 per treatment group) were administered a single dose of 0, 5, 20, 100, or 500 ng/kg
TCDD in corn oil via gavage.  Seven days following TCDD treatment, mice were infected with a
mouse-adapted strain of influenza (A/PR/34/8; H1N1) at a plaque forming unit dose designed  to
target approximately 30% mortality in each strain. TCDD did not affect the body weight or
survival in any of the infected mouse strains at any dose.
       Therefore, no LOAEL is established in this study. The NOAEL is 500  ng/kg.


D.l.3.7.    Simanainen et al. (2003)
       Simanainen et al. (2003) studied the short-term effects of TCDD exposure to determine
the  efficacy and potency relationships among three differentially susceptible rat lines.  The three
rat lines used were A, B, and C, and they were selectively bred from TCDD-resistant Han/Wistar
                                         D-42

-------
and TCDD-sensitive Long-Evans rats. The study authors reported that Line A rats were most
resistant to TCDD acute lethality followed by Line B and C.  Groups of five or six randomly
selected rats (sex not specified) were treated with a single oral dose of TCDD (purity >99%) in
corn oil by oral gavage. The dose of TCDD was reported to range between 30 ng/kg and
3,000 |ig/kg for Line A, 30 ng/kg and 1,000 |ig/kg in Line B, and 30 ng/kg and 100 |ig/kg for
Line C. Control animals were similarly dosed with a corn oil vehicle. Rats were sacrificed on
Day 8 postexposure, and trunk blood was collected and serum separated. Liver and thymus were
removed and weighed, and liver samples were collected and preserved. Liver EROD activity,
serum aspartate aminotransferase (ASAT) activity, free fatty acid (FFA) concentration, and total
bilirubin concentration were determined.  Teeth were also examined.
       Relative thymus weights were reduced 25% at 300 ng/kg relative to controls in Line B
rats. Liver enzyme (CYP1 Al) induction, as measured by EROD activity, was evident at all
exposure levels; CYP induction is considered to be an adaptive effect and not adverse in itself.
No other endpoints were affected below 1 ng/kg in any of the three rat lines.
       A LOAEL for TCDD of 300 ng/kg is identified for decreased relative thymus weight in
Line B rats.  A NOAEL of 100 ng/kg is identified for this study.


D.l.3.8.    Simanainen et al (2002)
       To study the short-term effects of TCDD on hormone levels, adult female Long-Evans
(TCDD-sensitive) and Han/Wistar (TCDD-resistant) rats (n = 9-1 I/treatment) were administered
a single dose of TCDD (>99% pure) in corn oil via gavage at doses ranging from 30 ng/kg to
100 |ig/kg. Vehicle controls received an equivalent amount of corn oil. The study also
examined other polychlorinated dibenzo-p-dioxins outcomes.  Rats were sacrificed on Day 8
postexposure, and trunk blood was collected and serum separated. Liver and thymus were
removed and weighed, and liver samples were collected and preserved. Liver EROD activity,
serum ASAT activity, FFA concentration, and total bilirubin concentration were determined.
Teeth were also examined.
       Neither FFA nor ASAT levels in Han/Wistar rats showed a dose-response relationship.
In Long-Evans rats, however, a significant (p < 0.05) dose-dependent increase in FFA occurred
at 300 ng/kg TCDD. Serum ASAT sharply increased in Long-Evans rats between 3,000 and
10,000 ng/kg. Body weight change and relative thymus weights were significantly decreased
                                         D-43

-------
(p < 0.05) in Han/Wistar rats with doses > 10,000 ng/kg and in Long-Evans rats with doses
> 1,000 ng/kg.  Liver EROD activity was significantly (p < 0.05) increased with all doses in both
strains.  Serum T4 was significantly (p < 0.05) decreased in Long-Evans rats at concentrations
>300 ng/kg, but were not significantly affected in Han/Wistar rats.  Serum bilirubin was
significantly (p < 0.05) increased with doses >10,000 ng/kg in Long-Evans rats and
>30,000 ng/kg in Hans/Wistar rats. Both strains of rat showed  a dose-dependent increase in
mean severity of incisor tooth defects. The results indicate that TCDD was the most potent
congener tested in both rat strains.
       A LOAEL of 300 ng/kg for decreased T4 in the Long-Evans rat is identified for this
study.  A NOAEL of 100 ng/kg is established.


D.l.3.9.    Smialowicz et al (2004)
       Smialowicz et al. (2004) examined the impact of TCDD exposure on immunosuppression
in mice. Groups of female (number not specified) C57BL/6N CYP1A2 (+/+) wild-type mice
were administered a single dose of 0, 30,  100, 300, 1,000, 3,000, or 10,000 ng/kg TCDD (purity
>99%) in corn oil via oral gavage. Control animals were similarly dosed with a corn oil vehicle.
To assess immune function,  7 days after TCDD administration, all mice were immunized with
sheep red blood cells (SRBCs) via injection into the lateral tail  vein. Five days after
immunization, mice were sacrificed, blood was collected, and enzyme-linked immunosorbant
assays were performed.  Additionally, spleen, thymus, and liver weights also were measured.
       Body and spleen weights of the wild-type mice were unaffected by the TCDD exposure.
A decrease in thymus weights of the mice appeared to be dose related. Only mice treated with
10,000 ng/kg TCDD, however, showed a statistically significant (p < 0.05) decrease in thymus
weights compared to corresponding controls. Liver weights also showed a dose-related increase
with only animals treated with 3,000 and  10,000 ng/kg TCDD showing statistical significance
(p < 0.05) compared to the control group. The antibody response to SRBCs indicated a
dose-related suppression in the wild-type mice, with animals treated with 1,000, 3,000,  and
10,000 ng/kg TCDD showing statistically significant (p < 0.05) suppression compared to the
controls.
                                         D-44

-------
       A LOAEL for TCDD of 1,000 ng/kg is identified in female C57BL/6N CYP1A2 (+/+)
wild-type mice for significant (p < 0.05) suppression of SRBCs. The NOAEL for this study is
300 ng/kg.
D. 1.3.10.   Vanden Heuvel et al (1994)
      Vanden Heuvel et al. (1994) examined the dose-response relationship between TCDD
exposure and induction of hepatic mRNA. Groups of 10-week-old female Sprague-Dawley rats
were administered TCDD (purity -99%) in corn oil once at 0, 0.1, 0.05, 1, 10, 100, 1,000, or
10,000 ng/kg-BW. Four days after TCDD treatment, animals were sacrificed and livers were
excised and preserved. Total hepatic RNA was extracted using guanidine thiocyanate and DNA
was removed using standard phenol-chloroform-isoamyl alcohol partitioning procedures.
Quantitative competitive RNA-PCR method was used to analyze CYP1 Al,
UDP-glucuronosyltransferase I (UGT1), plasminogen activator inhibitor 2 (PAI2), p-actin, and
transforming growth factor a (TGFa). In addition to hepatic mRNA levels, microsomal protein
was assayed for EROD activity and livers were tested for TCDD concentration.
      CYP1 Al mRNA induction levels in the TCDD-treated groups were low in the low-dose
region and sharply increased to plateaus at higher doses. The lowest dose that showed a
statistically significant (p < 0.05) difference compared to controls was the 1 ng/kg dose, which
showed a threefold increase in CYP1 Al mRNA levels.  In contrast, a  130-fold increase occurred
at 100 ng/kg and a 4,000- and 7,000-fold increase occurred at 1,000 and 10,000 ng/kg,
respectively. A slight increase in the CYPlAl/p-actin levels was observed in the 0.1 ng/kg
group, but this increase was not significant. EROD activity exhibited a pattern similar to
CYP1A1 activity.  EROD activity, however, was approximately 100-fold less sensitive
compared to mRNA  levels in TCDD-treated groups. Statistical  significance (p-value not
provided) in CYP1A1 level was observed at the 100 ng/kg dose compared to the 1 ng/kg dose.
The study authors reported that, despite this difference in CYP1 Al and EROD activity, the
correlation between CYP1A1 enzyme activity and mRNA levels was good. Dose-response
relationships for the induction of UGT1, PAI2, and TGFa mRNA differed from what had been
observed for CYP1 Al mRNA. UGT1 mRNA was induced, but at the much higher dose of
1,000 ng/kg. Additionally, the fivefold maximum induction of UGT1 mRNA was much less
than the 7,000-fold induction observed for CYP1 Al mRNA at the 10,000 ng/kg dose. The
                                        D-45

-------
authors state that this could be a result of the constitutive level of UGT1, which is much higher
than CYP1 Al, which makes detecting induction of UGT1 in the low dose regions more difficult.
PAI2 and TGFa mRNA were not affected by TCDD in rat liver in the dose range tested. These
results indicate that dioxin-inducible genes have a quite dissimilar dose-response relationship.
      Induction of CYP1A1 expression is not considered an adverse effect, as the role of
CYP1A1 in TCDD-mediated hepatotoxicity is unsettled. Therefore, in the absence of other
indicators of hepatotoxicity, a NOAEL/LOAEL cannot be determined for this study.  A LOEL
for TCDD of 1 ng/kg for a single exposure was identified for statistically significant (p < 0.05)
increase in CYP1 Al mRNA levels. The NOEL for this study is 0.1 ng/kg.


D.l.3.11.   Weber et al (1995)
      Weber et al. (1995) studied the effects of TCDD on intermediary metabolism in inbred
mice.  Following establishment of dose ranges via lethal dose eliciting 50 percent response
(LD50) studies, male C57BL/6 inbred mice (4-7 per dose group) were administered a single
gavage dose of 0, 30, 100, 300, 1,000, 3,000, 9,400, 37,500, 75,000, 100,000, 133,00, or
235,000 ng/kg TCDD (purity not specified) dissolved in corn oil (on Day 0 of the experiment).
Male DBA/2 inbred mice (4-7 per dose group) were treated with 0, 1,000, 10,000, 97,500,
375,000, 1,500,000, 1,950,000, or 3,295,000 ng/kg TCDD delivered in two gavage doses (on
Days -1  and 0). All mice were sacrificed and weighed on Day 8 after dosing, trunk blood was
collected and pooled for each dose group for serum preparation, and livers and kidneys were
removed, weighed, and snap frozen. In both strains of mice, phosphoenolpyruvate
carboxykinase (PEPCK)  and glucose-6-phosphatase (G-6-Pase) activities were measured in the
liver, and EROD activity was measured in the liver and kidneys. Liver tryptophan
2,3-dioxygenase (TdO) activity and serum tryptophan levels were measured in C57BL/6 mice.
Additionally, glucose concentrations and T4 and T3 levels were measured in the pooled serum of
both mouse strains.
      On Day 8 after dosing, the study authors reported that food consumption and body weight
were unchanged from control values in C57BL/6 mice at any dose tested, but a significant
(p < 0.05) reduction in food consumption and body weight at doses > 1,500,000 ng/kg-day in
DBA/2 mice (data not  shown). Relative liver weight was significantly (p < 0.05) increased
above control values at doses > 3,000 ng/kg-day in C57BL/6 mice and>97,500 ng/kg-day in
                                         D-46

-------
DBA/2 mice. Relative kidney weight was not affected by any dose of TCDD in C57BL/6 mice,
but was significantly (p < 0.05) decreased at 1,950,000 and 3,295,000 ng/kg in DBA/2 mice
(data not shown).
       In both mouse strains tested, basal EROD activities in the kidneys were only about
one-tenth of those in the liver.  In the liver of C57BL/6 mice, EROD  activity was significantly
(p < 0.05) induced over control  values at doses >300 ng/kg-day. Maximum induction occurred
at 37,500 ng/kg-day (58-fold), but then decreased by 28% in mice exposed at higher doses.
Kidney EROD activity in C57BL/6 mice was significantly (p < 0.05) induced over control values
at doses > 37,500 ng/kg-day, and no decrease was observed at the higher doses.  In the liver of
DBA/2 mice, EROD activity was significantly (p < 0.05) induced over control values at doses
>10,000 ng/kg-day. Maximum induction occurred at 375,000 ng/kg-day, but then decreased by
57% in mice exposed at higher doses.  Kidney EROD activity in DBA/2 mice was significantly
(p < 0.05) induced over control  values at doses >375,000 ng/kg-day, with a 3% and 29%
decrease below the level  of maximum  induction (1,500,000  ng/kg-day) at the two highest doses,
respectively. Liver PEPCK activity was significantly (p < 0.05) decreased below control values
at doses >100 ng/kg-day  in C57BL/6 mice, and > 10,000 ng/kg-day in DBA/2 mice. In contrast
to the PEPCK dose response, liver G-6-Pase activity was significantly (p < 0.05) decreased
below control values at doses >75,000 ng/kg-day in  C57BL/6 mice, and >375,000 ng/kg-day in
DBA/2 mice. Liver TdO activity in C57BL/6 mice increased by -20% over that of control  at
300 ng/kg-day, and this magnitude of induction did not change throughout doses tested.
       With respect to serum measurements, there were no dose-dependent changes in
tryptophan levels in either mouse strain tested.  Serum glucose levels followed the course of
PEPCK activity in both strains of mice, with sharp decreases observed only in the high dose
range. Thyroid hormone (T3 and T4) levels exhibited a dose-dependent decrease over the entire
dose range in both strains of mice; the lowest T3 and T4 levels were 35% of controls at the
133,000 ng/kg-day dose in C57BL/6 mice, and 40% (T3) and 20% (T4) of controls  at the highest
dose in DBA/2 mice.
       TCDD-induced hepatic and renal enzyme alterations are not considered significant
lexicologically adverse effects in and of themselves.  Additionally, because the serum
determinations were performed in pooled serum samples, statistical analysis could not be
performed.  Thus, this precludes these effects from being used to identify a NOAEL or LOAEL.
                                         D-47

-------
However, a LOAEL for TCDD of 3,000 ng/kg-day was identified for increased relative liver
weight in C57BL/6 mice.  The NOAEL is 1,000 ng/kg-day for C57BL/6 mice in this study.  In
DBA/2 mice, a LOAEL for TCDD of 97,500 ng/kg-day was identified for increased relative liver
weight, and the NOAEL is 10,000 ng/kg-day for this mouse strain.

D.1.4.  Subchronic Studies
D.l.4.1.    Chu et al (2001)
       Adult female Sprague-Dawley rats (five per treatment group) were administered TCDD
(purity >99%) in corn oil by gavage at doses of 0, 2.5, 25, 250, or 1,000 ng/kg-day for 28 days
(Chu et al., 2001).  The 1,000 ng/kg-day dose of TCDD caused a significant (p < 0.05) decrease
in body weight gain (36% lower than the control), increase in relative liver weight (40% greater
than the control), and decrease in relative thymus weight (50% lower than the control).  There
was a significant (p < 0.05) increase in EROD activity, methoxy resoufm-O-deethylase (MROD)
activity, and UDP-glucuronosyltransferase activity in the liver of female rats receiving 250 or
1,000 ng/kg-day TCDD. In addition, significant (p < 0.05) increases in serum cholesterol were
observed in the 250 and 1,000 ng/kg-day dose groups, and liver ascorbic acid (AA) also was
significantly increased in the 1,000 ng/kg-day dose group.  There was ~1.5-fold increase in liver
glutathione-^-transferase (GST), which was not statistically significant. Other significant
(p < 0.05) findings for the 1,000 ng/kg-day group included a decrease in liver vitamin A (51%
lower than the control), an increase in kidney vitamin A (15.5-fold increase above the control),
an increase in liver benzyloxy resoufm-O-deethylase (BROD, 30-fold increase above control), a
decrease  in liver pentoxyresoufin-O-deethylase (PROD, 37% lower than the control), increase in
serum albumin (18% above the control), and a decrease in  mean corpuscular hemoglobin (MCH,
7% below the control) and mean corpuscular volume (MCV, 7% below the control).
       Based on the numerous significant (p < 0.05) liver-related biochemical changes and
significant (p < 0.05) increased relative liver weight, as well as significantly decreased body
weight and relative thymus weight, the LOAEL for 28 days of exposure in this study is
1,000 ng/kg-day and the NOAEL is 250 ng/kg-day.
                                         D-48

-------
D.l.4.2.    Chu et al (2007)
       Chu et al. (2007) examined the potential impact of TCDD on various organs and the
toxicological impacts as a result of interactions between TCDD and PCBs in rats.  Groups of
female Sprague-Dawley rats (n = 5 per treatment group) were treated daily for 28 days via
gavage with 0, 2.5, 25, 250, or 1,000 ng/kg-day TCDD (purity not specified) dissolved in corn
oil. Body weights were determined three times per week, and clinical observations were made
daily. At study termination, all animals were sacrificed and blood was analyzed for various
biochemical and hematological parameters.  Liver, spleen, heart, thymus, brain, and kidneys
were removed and weighed. A small portion of the liver was homogenized and assayed for
BROD; EROD; MROD; and PROD. UGT, GST, and ascorbic acid levels also were measured.
Vitamin A levels in the liver, kidney, and lungs were analyzed as free retinol (vitamin A), and
histopathological analysis was conducted on various tissues.
       Growth rate and thymic weights in rats treated with  1,000 ng/kg-day TCDD were
significantly (p < 0.05) inhibited compared to the control group. Enzyme analysis indicated that
measured levels of TCDD in the liver correlated with hepatic microsomal enzyme activity. The
authors reported that liver microsomal EROD and MROD activities were significantly (p < 0.05
for EROD activity, significance level for MROD not reported) increased in the 250 and
1,000 ng/kg-day TCDD dose groups compared to the control group.  UGT levels were
significantly (significance level not reported) increased in the 250 and 1,000 ng/kg-day TCDD
dose groups compared to the controls. Serum albumin levels were significantly (p < 0.05)
increased in the 1,000 ng/kg-day TCDD dose group compared to the control group. Serum
cholesterol levels were significantly (level not reported) increased compared to the control group
at 250 ng/kg-day TCDD dose,  while liver ascorbic acid concentrations were significantly (level
not reported) increased in the 1,000 ng/kg-day dose group.  Hematological analysis indicated that
hemoglobin, packed cell volume, MCH, MCV, and platelet values were decreased in the
1,000 ng/kg-day TCDD dose group. Significant (p < 0.05) differences were observed only in
MCH and MCV levels compared to the control. Vitamin A levels in the liver and kidney were
significantly (p < 0.05) lower in the 1,000 ng/kg-day TCDD group compared to the control
group. Histopathological evaluation of various tissues indicated that liver, thyroid, and thymus
were the target organs.  No TCDD-related affects were found in other tissues. A dose-dependent
alteration in the thymus consisted of reduced thymic cortex  and increased medullar volume with

                                         D-49

-------
more animals exhibiting these changes at the 250 and 1,000 ng/kg-day dose level compared to
the control group. Alterations in thyroid included reduced follicles, reduced colloid density, and
increased epithelial height. A dose-dependent change in the thyroid was observed, with the
highest impact evident in reduced follicles and reduced colloid density beginning at a dose of
25 ng/kg-day TCDD. Changes in liver were characterized by accentuated hepatic zones,
anisokaryosis of hepatocytes, increased cytoplasmic density, and vacuolation. These changes
were also dose dependent, with more animals exhibiting these histopathological changes with
increasing TCDD dose. Based on these results, the study authors concluded that exposure to
TCDD resulted in a wide range of adverse effects with the thyroid proving to be most sensitive.
       A LOAEL for TCDD of 25 ng/kg for a 28-day exposure is identified for alterations in
thyroid, thymus, and liver histopathology. The NOAEL for this study is 2.5 ng/kg-day.
D.l.4.3.    DeCaprio et al (1986)
       Hartley guinea pigs (10 per sex per dose) were administered TCDD (purity not specified)
in the diet for 90 days at concentrations of 0, 2,  10, 76, or 430 ppt (equivalent to 0, 0.12, 0.61,
4.9, and 26 ng/kg-day in males and 0, 0.12,  0.68, 4.86, and 31 ng/kg-day in females calculated by
the study authors using food consumption and body weights). Other animals were administered
the high-dose diet (i.e., 430 ppt) for 11, 21,  or 35 days and then administered the control diet
(i.e., no exposure) for the remainder of the 90 days for recovery analysis.  Four high-dose males
died and two were sacrificed moribund by Day 45; the remaining four animals were sacrificed on
Day 46 for necropsy. Four high-dose females also died and two were sacrificed moribund by
Day 55 with the remaining females sacrificed on Day 60 for necropsy.  Animals in the 76- and
430-ppt groups had significantly (p < 0.05)  reduced body weights.  Organ weights were not
obtained in the 430-ppt group  due to the early sacrifice, but in the 76-ppt group a significant
decrease in relative thymus weight (p < 0.05) was  observed, and relative liver (p < 0.01) and
brain (p < 0.05) weights  in males increased. Although a similar trend occurred in the females,
the results were not statistically significant.  Males administered 76 ppt in the diet also had a
53% increase in triglycerides (p < 0.05).  The same increase was observed in females, but was
not statistically significant.  In the recovery groups, mortality during the recovery period after
11 or 21 days of treatment was 10% and after 35 days of treatment was 70%. Animals lost
                                          D-50

-------
weight during the treatment period. Although the body weight increased during the recovery
period, the body weight remained low compared to the control for the study duration.
       The LOAEL from this study is 4.9 ng/kg-day for 90 days of exposure, based on
decreased body weight (12-15%; p < 0.05) and changes in organ weights (10-30%, significant
only in the males).  The NOAEL is 0.61 ng/kg-day.


D.l.4.4.    Devito et al (1994)
       Female B6C3Fi mice (5 per treatment) were administered 0, 1.5, 4.5, 15, 45, or
150 ng/kg TCDD (98% pure) in corn oil via gavage, 5 days a week, for 13 weeks.  This dose is
equivalent to 0, 1.07, 3.21, 10.7, 32.1, 107 ng/kg-day (adjusted for continuous exposure,
administered dose multiplied by 5 and divided by 7). Body weight was recorded weekly and
animals were sacrificed 3 days after the last treatment. Examinations were performed on the
lung, skin, uterus, and liver.  No differences were observed in the liver or uterus weights or in the
estrogen receptor levels in these two tissues.  A dose-dependent increase in EROD activity (an
indicator of CYP1A1 [CYP] induction) in the lung, skin, and liver was observed, with significant
(p < 0.05) increases even at the lowest dose.  The TCDD doses used did not achieve maximal
EROD induction. A significant (p < 0.05) increase in liver acetanilide-4-hydroxylase (ACOH;
an indicator of CYP1A2 induction) also was observed with all doses. A maximum induction of
ACOH occurred with doses of 3.21 ng/kg-day and greater. A dose-dependent increase in
specific phosphotyrosyl protein (pp) levels also was observed. Levels of pp34 and pp38 were
significantly (p < 0.05) increased even at the lowest dose, while pp32 reached statistical
significance (p < 0.05) with doses of 4.5 ng/kg-day and above.
       The role of CYPs and phosphorylated pp32, pp34, and pp38 in TCDD-mediated toxicity
is unknown, and  changes in the activity or function of these proteins are not considered adverse.
Therefore, no LOAEL or NOAEL is established.  The 13-week LOEL is 1.07 ng/kg-day, based
on a significant (p < 0.05) increase in EROD, ACOH, pp34, and pp38 levels (all increased by at
least twofold).  No NOEL is established for this study.
                                         D-51

-------
D.l.4.5.    Fattore et al (2000)
       Fattore et al. (2000) examined TCDD-induced reduction of hepatic vitamin A levels in a

subchronic rat bioassay on Sprague-Dawley rats. Four experiments were conducted;

Experiments 1, 2, and 3 were conducted in both male and female rats, while Experiment 4 was

conducted only in female rats.  The dosing regimens for each experiment were as follows:
   Experiment 1—Groups of six Iva:SIV 50 rats (male and female) were maintained on a diet
   consisting of 0, 200, 2,000, or 20,000 ng TCDD/kg diet and 3-ug vitamin A/kg diet for
   13 weeks. Assuming food consumption of 10% of body weight per day, the average daily
   doses are 0, 20, 200, and 2,000 ng/kg-day TCDD.

   Experiment 2—Groups of six male and female rats were treated with 0 or
   200 ng TCDD/kg-day and 3 ug vitamin A/kg diet for 13 weeks.

   Experiment 3—Groups of six male and female rats were fed 0, 200, or
   1,000 ng TCDD/kg-day and 3 ug vitamin A/kg diet for 13 weeks.

   Experiment 4—Groups of female rats (number not specified; IVA;SIV 50 Sprague-Dawley
   strain) were treated with TCDD for 26 and 39 weeks in addition to a 13-week dietary
   treatment with 0 or 100 ng TCDD/kg-day and 3 ug vitamin A/kg diet for 13 weeks.
       For a 13-week exposure duration employed in all four experiments, male and female rats

were treated at 0, 20, 100 (females only), 200, 1,000, or 2,000 ng/kg-day.  In all

four experiments, the livers from the control and treated animals were analyzed at termination

for free retinol content to determine hepatic vitamin A levels.
   Results

   Experiment 1—Liver and body weights in both treated males and females were significantly
   affected at all but the lowest dose tested (20 ng/kg-day).  Liver injury was severe, particularly
   in female rats treated with 2,000 ng TCDD/kg-day. Dietary intake of vitamin A in male rats
   was comparable to intake in controls—except in the 2,000 ng/kg-day group, which showed a
   reduction of 16% in the dietary intake of vitamin A compared to controls. There was no
   effect of TCDD on vitamin A intake in female rats. Hepatic vitamin A levels showed a
   dose-dependent reduction with levels dropping sharply in the 200 and 2,000 ng/kg-day dose
   groups, particularly in treated females.  The reduction was significant at 200 ng/kg-day
   (p < 0.05) and 2,000 ng/kg-day (p < 0.01) in males and at 200 ng/kg-day (p < 0.5) and
   2,000 ng/kg-day (p < 0.001) in females. The reductions ranged from 68-99% in males and
   72-99% in females when compared to corresponding controls.
                                         D-52

-------
   Experiment 2—Changes in liver and body weights were not reported. Hepatic vitamin A
   level in males and females were reduced by 70% and 99%, respectively, compared to
   controls, in rats receiving 20 ng/kg-day (significance level in females: p < 0.01).
   Experiment 3—Similar to the results of Experiments 1 and 2, a dose-related trend of
   significantly (p < 0.001) reduced hepatic vitamin A level was observed in both males and
   females, with males exhibiting a particularly sharp drop at the 1,000 ng/kg-day dose
   compared to controls.
   Experiment 4—Females treated with 100 ng/kg-day showed significant reductions in hepatic
   vitamin A levels (p < 0.05-0.001) at all three treatment durations (13, 26, and 39 weeks).
       A LOAEL for TCDD of 20 ng/kg-day for a 13-week subchronic exposure was identified
in this study for decreased hepatic vitamin A levels (27 and 24% lower than the corresponding
control in female and male rats, respectively). This LOAEL is determined using data from
Experiment 1.  A NOAEL was not identified in this study.


D.l.4.6.    Fox et al (1993)
       Sprague-Dawley rats (6 per sex per dose) were gavaged with TCDD (purity not
specified) in corn oil using a dose-loading regime to achieve and maintain steady-state levels of
0.03, 30, or 150 ng/g in the liver. The regime consisted of an initial loading dose of 5, 2,500, or
12,000 ng/kg followed every 4 days with a maintenance dose of 0.9, 600, or 3,500 ng/kg.
Averaging the doses over the 14 days provides average daily doses of 0.55, 307, and
1,607 ng/kg-day (e.g., 5 ng/kg-day on Day 1 and 0.9 ng/kg-day on Days 5, 9, and 13 is
5 + 0.9 + 0.9 + 0.9/14 = 0.55 ng/kg-day). Body weight, liver weight, and liver gene expression
were measured at 7 and 14 days. A significant (p < 0.05) decrease in body weight occurred in
high-dose males (at 14 weeks only) and females (at 7 and 14 days). A significant (p < 0.05)
increase in absolute and relative liver weights was observed in mid- and high-dose males and
females at both 7 and 14 days. Although the liver of treated animals indicated moderate
vacuolization and  swelling, there was no indication of necrosis. An increase in gene expression
(clone 1, CYP1 Al, CYP1A2, and albumin) was observed in the mid- and high-dose groups. A
significant (p < 0.05) decrease in labeling index  (indication of cell proliferation) occurred in both
females (all doses) and males (high-dose only) during Week 1—but not during Week 2.
                                         D-53

-------
       The 14-day LOAEL is 307 ng/kg-day for significant (p < 0.05) increases in absolute and
relative liver weights (25-34%).  The NOAEL is 0.55 ng/kg-day.


D.l.4.7.    Hassoun et al (1998)
       Female B6C3Fi mice (number not specified) received TCDD (>98% pure) in corn oil
5 days per week for 13 weeks via gavage at doses of 0, 0.45, 1.5, 15, or 150 ng/kg (equivalent to
0, 0.321, 1.07, 10.7, and 107 ng/kg-day adjusted for continuous exposure; administered dose
multiplied by 5 and divided by 7). Three days after the final dose, animals were sacrificed and
their brains were removed for oxidative stress testing. Biomarkers for oxidative stress included
production of superoxide anion (SA), lipid peroxidation, and DNA single-strand breaks (SSBs).
A significant (p < 0.05) increase was observed in superoxide anion production, lipid peroxidation
as measured by thiobarbituric acid-reactive substances (TEARS), and DNA single-strand breaks
with all doses tested.
       No other indicators of brain pathology were assessed, and it is unfeasible to link the
markers of oxidative stress to a TCDD-induced toxicological outcome in the brain. Thus, no
LOAEL/NOAEL was established. The subchronic (13-week) LOEL is 0.32 ng/kg-day, based on
significant (p < 0.05) increases in superoxide anion production (80% above control); lipid
peroxide production (25% above the control); and DNA single-strand breaks (twofold over the
control).  No NOEL is established.
D.l.4.8.    Hassoun et al (2000)
       Hassoun et al. (2000) examined the effect of subchronic TCDD exposure on oxidative
stress in hepatic and brain tissues.  Groups of 8-week-old female Harlan Sprague-Dawley rats
(6 rats/group) were administered TCDD (98% purity, dissolved in 1% acetone in corn oil) via
gavage at 0, 3, 10, 22, 46, or 100 ng/kg-day, 5 days/week, for 13 weeks (0, 2.14, 7.14, 15.7, 32.9,
or 71.4 ng/kg-day adjusted for continuous exposure; administered doses were multiplied by 5
and divided by 7 days/week). Animals were sacrificed at the end of the study period, and the
brain and liver tissues were collected and used to determine the production of reactive oxygen
species, lipid peroxidation, and DNA SSBs.
                                         D-54

-------
       A dose-dependent effect was observed in both the liver and brain tissue as a result of
TCDD treatment.  Based on the maximal induction of superoxide anion by various doses, more
production of superoxide anion was observed in the liver tissue when compared with the brain
tissue with an observed increase of 3.1- and 2.2-fold respectively, when compared to the control
group. A similar dose-dependent effect was observed in the induction of lipid peroxidation in
TCDD-treated animals with an approximately 1.8-fold increase in lipid peroxidation in both
tissues relative to the corresponding controls.  A dose-dependent relationship was also observed
for DNA SSBs in both the hepatic and brain tissues at all TCDD-treated doses compared to
controls. Increases were statistically significant (p < 0.05) beginning at the lowest administered
dose.
       Similar to the statement above, because no adverse endpoints were measured, no
LOAEL/NOAEL was established.  However, a LOEL for TCDD of 2.14 ng/kg-day for a
13-week exposure duration was identified in this study for significant increases (p < 0.05) in
superoxide anion, lipid peroxidation, and DNA SSBs in the liver and brain tissues. A NOEL
cannot be determined for this study.


D.l.4.9.   Hassoun et al (2003)
       Hassoun et al. (2003) examined the role of antioxidant enzymes  in TCDD-induced
oxidative stress in various regions of the rat brain after subchronic exposure.  Groups of
8-week-old female Harlan Sprague-Dawley rats (12 rats/group) were administered TCDD (98%
purity, dissolved in 1% acetone in corn oil) via gavage at 0, 10, 22, or 46 ng/kg-day (0, 7.14,
15.7, or 32.9 ng/kg-day adjusted for continuous exposure; administered  doses were multiplied by
5 and divided by 7) daily for 13 weeks. Animals were sacrificed at the end of the study period
and the brain was immediately removed  and dissected to the following regions: cerebral cortex
(Cc), hippocampus (H), cerebellum (C),  and brain stem including midbrain, pons, and  medulla.
Four pooled samples from each region per dose (i.e., 3 animals/pooled sample) were used in the
study.  Dissected regions were subsequently assayed for lipid peroxidation (TEARS), superoxide
dismutase, catalase,  and glutathione peroxidase. Because the cytochrome c reduction method
was used to determine SA production in  brain tissues, SOD was added to some of the brain tissue
samples that had the highest SA production (tissue homogenates from Cc and H from rats treated
with 46 ng/kg-day TCDD).
                                         D-55

-------
       A dose-dependent increase in the production of SA was observed in the Cc and H, but
significant changes in SA production were not observed in either the C or the mid-brain, pons, or
medulla brain stem cells. Similar to SA production, there was a dose-dependent increase in the
production of TEARS in the Cc and H regions of the brain, but no significant changes were
observed in either the C or the B sections of the brain. The study authors also measured the
activities of various enzymes as a result of TCDD treatment and reported a dose-dependent
increase in SOD activity in the C and B sections, while there was dose-dependent suppression in
SOD activity in Cc and H.  In contrast, catalase activity was significantly (p < 0.05) increased in
H and Cc at the 10 ng/kg-day TCDD dose level compared to controls and the mid- and high-dose
animals.  Catalase activity also was increased in a dose-dependent manner in the C section, but
no significant changes in the activity of this enzyme were observed in the B section at any of the
three TCDD tested doses. The effects of subchronic exposure to different  doses of TCDD on
glutathione stimulating hormone peroxidase (GSH-Px) showed a different response compared to
other enzymes. There was a dose-dependent increase in the activity of this enzyme in the C and
B regions of the brain, while a significant increase in the activity of GSH-Px occurred in Cc and
H only at the 10 ng/kg-day TCDD dose. In addition, the activity of this enzyme was suppressed
in a dose-dependent manner in the Cc and H at 22 and 46 ng/kg-day TCDD doses. Based on
these results, the study authors concluded that induction of oxidative stress by TCDD in the rat
brain occurs mainly in the Cc and H regions.
       Similar to the statement above, because no adverse endpoints were measured, no
LOAEL/NOAEL was established. However, a LOEL for TCDD of 7.14 ng/kg-day for a
13-week exposure duration was identified for this study for increases in superoxide anion and
lipid peroxidation production, as well as increased activity in SOD, catalase, and GSH-Px.
D.l.4.10.   Kodbaetal (1976)
       Adult Sprague-Dawley rats (12 per sex per treatment group) were administered TCDD
(purity not reported) in corn oil via gavage 5 days per week at doses of 0, 1, 10, 100, or
1,000 ng/kg-day (equivalent to 0, 0.71, 7.14, 71.4, or 714 ng/kg-day averaged over 7 days; 5/7 of
dose).  Five animals per group were sacrificed at the end of treatment, and the remaining animals
were observed over 13 weeks post treatment (only initial results for the post-treatment period
were provided in the report).  Body weights and food consumption were measured semiweekly.
                                         D-56

-------
Hematology and clinical chemistry were measured after 36-37 or 85-86 days of treatment and
59-60 days after termination of treatment. Forty-eight hour urine samples were collected from
select rats from 85-89 days of treatment and 52-56 days after cessation of treatment.  Gross and
histopathological exams were conducted on the tissues.
       Four high-dose females died during treatment. Two high-dose females and
two high-dose males died during the post-treatment period. Animals treated with 714 ng/kg-day
were less active during the treatment period, which became less evident during the posttreatment
period. Yellow discoloration of the external pinnae also was noted in this group, both during
treatment and during the post-treatment period. A significant (p < 0.05) reduction in body
weight and food consumption was observed in the 71.4 and 714 ng/kg-day groups. The
following significant (p < 0.05) hematology changes were observed in the high-dose
(714 ng/kg-day) males at all measured time points: decreased packed cell volume, decreased red
blood cells, decreased hemoglobin, increased reticulocytes, and decreased thrombocytes.
Significant (p < 0.05) changes also occurred in the high-dose females, but the only consistent
observation was a decrease in thrombocytes and increased leukocytes.  Significant changes in
clinical chemistry (p < 0.05) and urinalysis (p < 0.05) were more consistent between the sexes in
the high-dose group and included increases in total and direct serum bilirubin; increase in serum
alkaline phosphatase; decreased urinary creatinine; and increased urinary coproporphyrin,
uroporphyrin, and delta-amino-levulinic.  The following significant (p < 0.05) changes were
observed in the 71.4 ng/kg-day group: decreased packed cell volume (4-9%) in males; decreased
red blood cells (2-10%) in males; decreased hemoglobin (2-13%) in males; increased urinary
coproporphyrin (2.2-fold increase during treatment) in females; increased urinary
delta-amino-levulinic (47% increase during treatment) in females; increased total and  direct
serum bilirubin (48-61%) in females; and increased serum alkaline phosphatase (twofold) in
females.  The following significant (p <  0.05) changes in relative organ weights were  observed
increased brain weight in 714 ng/kg-day males and females; increased liver weight in males
(71.4 and 714 ng/kg-day) and females (7.14, 71.4, and 714 ng/kg-day); increased spleen weight
in 714-ng/kg-day males and females; decreased thymus weight in 71.4 and 714 ng/kg  males  and
females; and increased testes weight in 714 ng/kg-day males.  Microscopic changes were
observed in the thymus, and in other lymphoid tissues, and in the liver in rats treated with
71.4 ng/kg-day or greater.

                                          D-57

-------
       The subchronic (13-week) LOAEL is 71.4 ng/kg-day, based on the numerous changes
noted in body weight, hematology, clinical chemistry, urinalysis, and histopathology. The
NOAEL is 7.14 ng/kg-day.
D. 1.4.11.   Matty and Chipman (2002)
       Female F344 rats (3 per treatment group) were administered TCDD at concentrations of
0, 2.5, 25, or 250 ng/kg in corn oil via gavage for either 3 consecutive days or 2 days per week
for 28 days (Mally and Chipman, 2002). The average daily doses for the 28-day study when
adjusted for 7 days a week were 0, 0.71, 7.1, and 71 ng/kg-day (i.e., 2/7 of administered dose).
No clinical signs of toxicity were observed. Histological examination of the liver revealed no
abnormalities.  All doses of TCDD reduced the number of connexin (Cx) 32 plaques and Cx32
plaque area in the liver, which was considered the target tissue.  The reductions were not
statistically significant after the 3-day treatment, but were significant after the 28-day treatment
(p < 0.05).  TCDD also caused a reduction in the Cx32 plaque number and area in the thyroid
after 28 days, but the results were not statistically significant. Although the reduction in Cx32
plaque number and plaque area in the liver and thyroid occurred at all dose levels, there was no
relation to dose.  TCDD did not induce hepatocyte proliferation.
       In the absence of additional indicators of hepatotoxicity, changes in Cx32 plaques are not
clearly linked to TCDD-mediated hepatotoxicity, nor are they considered an adverse effect.
Additionally, no lexicologically relevant endpoints were examined.  Therefore, a NOAEL or
LOAEL cannot be determined. A 28-day LOEL at the lowest dose of 0.71 ng/kg-day for
significantly (p < 0.05) decreased Cx32 plaque area is evident (approximately 70% of the
controls).
D.l.4.12.   Slezak et al (2000)
       Slezak et al. (2000) studied the impact of subchronic TCDD exposure on oxidative stress
in various organs of B6C3Fi female mice. Groups of 8- to 10-week-old female B6C3Fi mice
(number not specified) were administered TCDD (purity >98%, dissolved in corn oil) via gavage
at 0, 0.15, 0.45, 1.5, 15, or 150 ng/kg-day (0, 0.11, 0.32, 1.07, 10.7, or 107.14 ng/kg-day adjusted
for continuous exposure) 5 days per week for 13 weeks. Three days after the last treatment, the
                                         D-58

-------
animals were sacrificed and organs were removed for the measurement of oxidative stress
indicators including SA, lipid peroxidation (TEARS), AA, and total glutathione stimulating
hormone (GSH). Tissue TCDD concentrations also were measured.
       The study authors reported that TCDD dose range resulted in overlapping tissue
concentrations for liver, lung, kidney, and spleen. Liver had the highest TCDD concentration,
with each tissue demonstrating a dose-dependent increase in TCDD concentration. Compared to
controls,  SA production in the liver was significantly (p < 0.05) lower at the 0.15 ng/kg-day
TCDD dose, while it was significantly (p < 0.05) higher at 15 and 150 ng/kg-day.  A dose-
dependent increase in hepatic TEARS production was observed, although the rate of production
was significant (p < 0.05) only at the highest TCDD administered dose (150 ng/kg-day)
compared to controls.  AA also followed the same pattern observed for hepatic SA and TEARS
with AA production significantly (p < 0.05) increased at the 15 and 150 ng/kg-day TCDD doses.
Contrary to the SA, TEARS, and AA responses, liver GSH levels were decreased  at
0.15 ng/kg-day, were increased at 0.45 and 150 ng/kg-day, and did not change at 1.5 or
15 ng/kg-day when compared to the control group. Unlike the liver, there was no significant
increase in SA production in the lung at any of the TCDD tested doses; a dose dependent
reduction, however, was observed at 0.45, 15, and 150 ng/kg-day compared to controls. GSH
and AA production in the lung was decreased at 0.15 ng/kg-day, while AA  production was
significantly (p < 0.05) increased at 15 and 150 ng/kg-day.  Kidney SA production showed a
statistically significant (p < 0.05) increase only at the 15 and 150 ng/kg-day doses. GSH, like in
the liver and the lung, exhibited a decrease in production in the kidney following treatment at
0.15 ng/kg-day with this trend continuing at 0.45 and 1.5 ng/kg-day. AA levels in the kidney
were significantly (p < 0.05) lower at all subchronic doses, except at 1.5 ng/kg-day dose. SA
levels in the spleen differed little from the control group at any of the TCDD doses. Total GSH
in the spleen was higher only at the 150 ng/kg-day dose level, while the AA levels were
significantly (p < 0.05) decreased at 0.15, 1.5, and 150 ng/kg-day.
       Similar to the statements regarding the Hassoun et al. studies above, because no adverse
endpoints were measured, no LOAEL/NOAEL was established. Therefore, a NOAEL or
LOAEL cannot be determined. However, a NOEL and LOEL of 1.07 and 10.7 ng/kg-day,
respectively, are identified in this study for increases in superoxide anion in the liver.
                                         D-59

-------
D.l.4.13.   Smialowicz et al (2008)
       Female B6C3Fi mice (8-15 per treatment group) were administered TCDD (purity
>98%) in corn oil by gavage at doses of 0, 1.5, 15, 150, or 450 ng/kg-day, 5 days a week, for
13 weeks (1.07, 10.7, 107, or 321 ng/kg-day, adjusted for continuous exposure; i.e., 5/7 of the
dose) (Smialowicz et al., 2008). Mice were immunized 3 days after the final TCDD exposure
with an intravenous injection of an optimal concentration of 4 x 107 SRBCs and sacrificed 4 days
later. No TCDD-related effects on body weight were observed. There was a dose-related
decrease in relative spleen weight (9-19% lower than control values) with statistically significant
(p < 0.05) decreases at all but the lowest dose.  Additionally, there was a statistically significant
(p < 0.05) increase in relative liver weight (5-21%) in all treatment groups compared to controls.
Statistically significant dose-dependent decreases were observed in the antibody response to
SRBCs (24-89% lower than control values), as measured by both the number of plaque forming
cells per 106 cells and plaque forming cells per spleen.
       The 13-week LOAEL for this study is 1.07 ng/kg-day based on a significant (p < 0.05)
increase in relative liver weight (10%) and a significant (p < 0.05) decrease in antibody response
to SRBCs (24%). A NOAEL cannot be determined for this study.

D. 1.4.14.   Van Birgelen et al (1995a; 1995b)
       Van Birgelen et al. (1995a: 1995b) studied the impact of TCDD exposure on various
biochemical endpoints in rats. In Van Birgelen et al.(1995b) groups of 7-week-old female
Sprague-Dawley rats (n = 8 per treatment group) were treated with 0, 200, 400, 700, 5,000,  or
20,000 ng/kg TCDD (purity >99%) in the diet for 13 weeks. Daily TCDD intake based on food
consumption, diet level, and mean weight was estimated to be 0, 14, 26, 47, 320, or 1,024 ng/kg-
day. Blood samples were collected from treated  animals and assayed for retinol (vitamin A),
triiodothyronine, and TT4 and free thyroxine (FT4). At study termination, the animals were
sacrificed, and the liver, thymus, spleen, and kidneys were removed and weighed. Parts of the
liver were homogenized and assayed to determine EROD; CYP1 Al; CYP1A2; and UGT
activity. Liver samples also were analyzed for retinol content. Van Birgelen et al. (1995a)
analyzes in greater detail the effects of TCDD on thyroid hormone metabolism, and both  papers
are based on the same materials and methods.
       TCDD-treated animals showed a dose-related decrease in food consumption.  Animals
treated with 1,024 ng/kg-day TCDD consumed 32% less food compared to controls.  Similarly, a
                                        D-60

-------
dose-related decrease in body weight gain was observed in all animals treated with TCDD.
Animals treated with >47 ng/kg-day of TCDD showed a statistically significant (p < 0.05)
decrease in body weight gain.  Relative liver weights were significantly (p < 0.05) increased in
the 320 and 1,024 ng/kg-day TCDD dose groups compared to the controls. Absolute and relative
thymus weights were significantly (p < 0.05) decreased at all TCDD dose groups compared to
the control group. Relative kidney and spleen weights were significantly (p < 0.05) higher in
animals dosed with >47 ng/kg-day of TCDD compared to the control group, with  the greatest
increase occurring in animals treated with 1,024 ng/kg-day TCDD (121  and 173% higher than
controls for kidney and spleen, respectively).  Cytochrome P450 enzymes, including EROD,
CYP1A2,  CYP1A1, and UGT, exhibited statistically significant (p < 0.05) increases in activity at
all TCDD  dose groups compared to the control group.  TT4 and FT4 thyroid hormone
concentrations were statistically significantly (p < 0.05) decreased only  at TCDD  doses
>47 ng/kg-day.  A dose-dependent increase was observed in the plasma retinol concentrations
with significant (p < 0.05) increases occurring at>47 ng/kg-day TCDD  after a 13-week
exposure.  A dose-dependent reduction in liver retinoid levels also was observed after 13 weeks
of TCDD exposure with the levels dropping significantly (p < 0.05) at all TCDD-treated doses
compared  to the control group.
       A LOAEL for TCDD of 14 ng/kg for a 13-week exposure is identified for significantly
(p < 0.05)  decreased absolute and relative thymus weights and significantly (p < 0.05) decreased
liver retinoid levels. A NOAEL cannot be determined for this study.


D.l.4.15.   Vosetal (1973)
       Vos et al. (1973) conducted a study to  examine the immune response in laboratory
animals treated with TCDD. In one experiment, 10 female Hartley strain guinea pigs were orally
treated with 8 weekly doses of 0, 8, 40, 200, and 1,000 ng/kg TCDD in corn oil (purity of TCDD
not specified) (0, 1.14, 5.71, 28.6, and 143 ng/kg-day adjusted for continuous exposure;
administered dose divided by 7). At  study termination, the animals were sacrificed, and heart
blood was used to determine total leukocyte and differential leukocyte counts.  In another
experiment, the effect of TCDD on humoral immunity was determined by injecting 0.1 mL of
tetanus toxoid into the right hind-foot pad on Day 28 (1 left foot tetanus toxoid, aluminum
phosphate-adsorbed) and again on Day 42 (1 left foot tetanus toxoid, unadsorbed). Blood was
                                          D-61

-------
collected (n= 10) on Days 35 and 49, and the serum tetanus-antitoxin concentrations were
determined using a modified single radial immunodiffusion technique.
       All guinea pigs receiving 1,000 ng/kg-day TCDD either died or were killed when
moribund between 24 and 32 days. These animals showed severe weight loss, lymphopenia, and
depletion of the lymphoid organs, especially the thymus. Microscopic observations revealed
severe atrophy of the thymic cortex with substantial destruction of lymphocytes, with the nuclear
debris being engulfed by macrophages. Large cystic Hassall bodies, filled with
polymorphonuclear leukocytes, were observed in the medulla. All animals treated with 0, 8, 40,
or 200 ng/kg-day TCDD survived until study termination. Body weight gain was significantly
(p < 0.01) lower in the 200 ng/kg-day group. Absolute thymus weight was significantly reduced
in the 40 and 200 ng/kg-day treatment groups (p < 0.01  andp < 0.05, respectively).  In contrast,
relative thymus weight was significantly (p < 0.01) reduced only in the 200 ng/kg-day dose
group. The absolute weight of the superficial cervical lymph nodes was significantly (p < 0.05)
decreased in the 200 ng/kg-day group, while the relative adrenal weight was significantly
(p < 0.05) increased in the 200 ng/kg-day dose group. Total leukocyte count was significantly
(p < 0.05) decreased in the 40 ng/kg-day dose group and total lymphocyte count was
significantly decreased at 8, 40, and 200 ng/kg-day (p < 0.0l,p< 0.05, andp < 0.05,
respectively). A significant (p-values not provided) monotonic dose-response relationship was
determined for body weight (decrease), relative thymus weight (decrease), relative adrenal
weight (increase), and total leukocyte and lymphocyte count (decrease).  Microscopic
examination of the lymphoid organs and adrenals showed no effects, while slight cortical atrophy
of the thymus was observed at the 200 ng/kg-day dose.
       Animals receiving the tetanus toxoid injection showed a small but significant increase in
serum tetanus antitoxin concentrations at the 8 and 40 ng/kg-day dose (p < 0.05 andp < 0.01,
respectively). Measurement at Days 49 and 56 indicated that serum antitoxin levels had
decreased sharply and the significant (p < 0.05 on Day 49 andp < 0.01 on Day 56) effect was
seen only at the 200 ng/kg-day dose level.
       A LOAEL for TCDD of 5.71 ng/kg-day for an 8-week exposure is identified in this study
for significantly (p < 0.01) reduced absolute thymus weight, significantly (p < 0.05) reduced
leukocyte and lymphocyte count, and  significantly (p < 0.01) increased serum tetanus antitoxin
concentration.  The NOAEL for this study is 1.14 ng/kg-day.

                                          D-62

-------
D.l.4.16.   White et al (1986)
       White et al. (1986) studied the impact of TCDD exposure on serum complement levels.
Groups of female (C57BL/6 x C3H)Fl(B6C3Fi) mice were treated for 14 consecutive days with
TCDD in corn oil (purity of TCDD not specified) at doses of 0, 10, 50, 100, 500, 1,000 or
2,000 ng/kg-day via gastric intubation (n = 6-8).  At study termination, blood was collected from
anesthetized animals and assayed for serum complement activity and complement component
C3 levels.
       Serum complement activity between the 10 and 100 ng/kg-day doses was between 69 and
59% compared to the vehicle control group, with all treatment groups being significantly
(p < 0.05) low compared to the vehicle control. In contrast, C3 levels were comparable to the
vehicle control with levels ranging between 98 and 94% of the control group.  The higher doses
of 500, 1,000, and 2,000 ng/kg-day, however, produced a marked decrease of the component
hemolytic activity (45, 35, and 19% of the vehicle control) and of C3 levels (91, 81, and 74% of
the vehicle control, respectively; significance level at/? <  0.05).
       A LOAEL for TCDD of 10 ng/kg-day for  a 14-day exposure is identified in this study for
significantly (p < 0.05) lower serum complement activity. A NOAEL cannot be determined for
this study.

D.1.5. Chronic Studies (Noncancer Endpoints)
D. 1.5.1.    Cantoni et al (1981)
       CD-COBS rats (4 per treatment) were orally administered TCDD (purity not specified)
dissolved in acetone:corn oil (1:6) at doses of 0 (vehicle alone), 10, 100, or 1,000 ng/kg per week
(equivalent to 1.43, 14.3, and 143 ng/kg-day adjusted for continuous exposure, administered
dose by dividing the dose by 7) for 45 weeks.  Urine was collected several times during
treatment and tested for porphyrin excretion.  Twenty-four hours after the final dose, animals
were sacrificed and their livers, spleens, and kidneys were removed for analysis of total
porphyrins. All treatment groups had a significant (p < 0.05) increase in coproporphyrin
excretion beginning at 6, 3, or 2 months, respectively.  Uroporphyrin excretion was significantly
(p < 0.05) increased in the 14.3 ng/kg-day group at 10 months and in the  143 ng/kg-day group

                                         D-63

-------
beginning at 6 months.  The high-dose group also had a significant (p < 0.05) increase in
excretion of heptacarboxylic methyl ester beginning at 6 months.  The high-dose group had a
marked porphyric state beginning at 8 months as indicated by a 70-fold increase above controls
in total urinary porphyrin excretion. This group also had a significant (p < 0.05) increase in total
porphyrins in the liver, kidneys, and spleen.
       The 45-week LOAEL for this study is 1.43 ng/kg-day, based on a two- to threefold
increase in urinary coproporphyrin excretion.  No NOAEL was established for this study.


D.l.5.2.    Croutch et al (2005)
       Croutch et al. (2005) examined the impact of TCDD exposure on body weight via
insulin-like growth factor (IGF) signaling. Female  Sprague-Dawley rats were randomly assigned
in groups of five to initial loading doses of TCDD (purity >98.5%, dissolved in corn oil) at  0,
12.5, 50, 200, 800, or 3,200 ng/kg-day, followed by treatment with maintenance doses equivalent
to 10% of the initial loading dose every third day to maintain a pharmacokinetic steady state
throughout the entire study (equivalent to 14-day average = 0, 1.25, 5, 20, 80, or 320 ng/kg-day;
28-day average = 0, 0.85, 3.4,  13.6, 54.3, or 217 ng/kg-day; 63-day average = 0, 0.60, 2.4, 9.5,
38, or 152 ng/kg-day; and 128-day average dose = 0, 0.51, 2.0, 8.1, 32.5, or 130 ng/kg-day).
Following 2, 4,  8, 16, 32, 64, or 128 days of initial dosing, the animals were sacrificed, the livers
were removed and weighed, and the trunk blood was collected to analyze glucose content. Rat
liver PEPCK mRNA and protein levels also were analyzed, and PEPCK activity was measured.
       Body weights of TCDD-treated animals decreased after the second week of the
3,200 ng/kg-day TCDD loading dose, with significant differences beginning at Week 9. There
was also a statistically significant (p < 0.05) difference in body weights at Weeks 10, 11, 13, 18,
and 19 at the highest loading dose (3,200 ng/kg-day). PEPCK activity in the liver was also
decreased in a dose-dependent manner following TCDD administration at approximately
16 days. PEPCK inhibition was statistically significant (p < 0.05) on Day 4 in rats treated with
either 800 or 3,200 ng/kg-day  TCDD when compared to animals treated with a loading dose of
200 ng/kg-day.  A similar statistically significant change was observed in animals treated with
3,200 ng/kg-day on Day 16 when compared to the 200 ng/kg-day treatment group. In contrast,
differences in PEPCK activity at other doses or time points were not statistically significant. In
TCDD-treated animals, there was also a dose-dependent decrease  in PEPCK mRNA expression
                                          D-64

-------
along with a decrease in PEPCK protein levels in the liver. In addition to body weight and
PEPCK activity changes, animals treated with 3,200 ng/kg-day TCDD showed a sharp decline in
circulating IGF-I levels on Day 8 compared to the control group (corn oil) and TCDD-treated
animals at lower doses. In the highest dose animals, IGF-I levels continued to decline to 42% of
the control group by Day 16 of the study. The IGF-I levels at the highest dose plateaued at an
average decrease of 66% through Day 128 when compared to controls. Beginning at Day 8, the
decrease in IGF-I was  statistically significant at every time point through Day 128 compared to
the control group, as well as groups treated with either 12.5 or 50 ng/kg-day TCDD.  Similar
statistically significant decreases also were observed for the 800 ng/kg-day TCDD-treated groups
with an initial decrease of 37% on Day 16 followed by a further decline to approximately 45%
thereafter compared to controls and the 12.5, 50, and 200 ng/kg-day dose groups. In contrast to
these results, circulating levels of insulin and glucose were unaffected by TCDD treatment, while
the active or phosphorylated form of AMPK-a protein increased with dose as a result of TCDD
treatment.
       A LOAEL for TCDD of 217 ng/kg-day for a 28-day exposure duration (because this
represented the most sensitive time for elicitation of effects) was identified in this study for
decreased body weight, significant (p < 0.05) inhibition of PEPCK activity, and reduced IGF-I
levels (42% lower than the control group). A NOAEL of 54.3 ng/kg-day was identified in this
study.


D.l.5.3.   Hassoun et al (2002)
       Hassoun et al. (2002) examined the potential of TCDD and other dioxin-like chemicals to
induce oxidative stress in a chronic rat bioassay.  Groups of six Harlan Sprague-Dawley female
rats were treated with 0, 3, 10, 22, 46, or 100 ng/kg-day TCDD (98% purity), 5 days a week via
gavage for 30 weeks.  The administered doses adjusted for continuous exposure were 0, 2.14,
7.14, 15.7, 32.9, and 71.4 ng/kg-day, respectively (administered doses were multiplied by 5 and
divided by 7). At study termination, hepatic and brain tissues from all treated rats were divided
into two portions and examined for the production of reactive oxygen species and SSBs in DNA.
       When compared to controls, there was a dose-dependent increase in the production of
superoxide anion in TCDD-treated animals ranging from 21-998% and 66-257% in hepatic and
brain tissues, respectively. Hepatic tissues had statistically significant (p < 0.05) increases in
                                         D-65

-------
superoxide anion production at doses >7.14 ng/kg-day, while the brain tissue had a statistically
significant (p < 0.05) increase over controls at all doses.  Similarly, increases in lipid
peroxidation were observed in hepatic and brain tissues with a 481% increase (p < 0.05) at
71.4 ng/kg-day in the hepatic tissue when compared to controls. The increase in lipid oxidation
in brain tissue ranged from 33-188% (p < 0.05) in the 2.14-71.4 ng/kg-day dose groups. DNA
SSBs were also observed in both hepatic and brain tissue in all treated groups.  When compared
to the control group, there was a dose-dependent statistically significant (p < 0.05) increase in
DNA SSBs ranging from 58-322% and 29-137% in hepatic and brain tissues, respectively.
Nonmonotonic dose-response relationships were observed for superoxide production and lipid
peroxidation in liver tissues, with greater-than-linear increases in effect between the two highest
dose levels.
       As stated above, because no adverse endpoints were measured, no LOAEL/NOAEL was
established.  However,  a LOEL for TCDD of 2.14 ng/kg-day for a 30-week exposure duration is
identified in this study for significant (p < 0.05) increases in superoxide anion, lipid peroxidation
production, and DNA SSBs in the liver and brain tissues. A NOEL cannot be determined for this
study.


D.l.5.4.    Hong et al (1989)
       Hong et al. (1989) studied the immunotoxic effects associated with chronic exposure to
TCDD in rhesus monkeys. Female rhesus monkeys (seven to eight animals per treatment group)
were exposed to 0,  5, or 25 ppt TCDD (purity not specified) in feed for 4 years.  As described
previously (Bowman et al.,  1989a: 1989b), these dietary concentrations were equivalent to 0,
0.12, and 0.67 ng/kg-day, respectively.  These adult females were tested for immune
abnormalities 4 years after cessation of exposure.  Additionally, offspring from  exposed mothers
born into Cohort I (n =  7, 6, and 1, respectively), Cohort II (n = 5, 6, and 2, respectively), and
Cohort III (n = 6, 6, and 3, respectively) (as described by Bowman et al. (1989bV) were also
tested. Monoclonal antibodies with flow cytometry were used to enumerate cells in the various
leukocyte populations.  A proliferative response to mitogens (phytohemagglutinin,  pokeweed,
concanavalin A) as well as allo- and xeno-transplantation antigens was measured.  Natural
killing capacity and a T cell dependent response to immunization with tetanus toxoid was also
                                          D-66

-------
assessed. The range of normal immune responses in rhesus monkeys was obtained from
45 healthy animals unrelated to the TCDD exposure studies.
       In adult monkeys, an increased number of T lymphocytes were observed in the
0.67 ng/kg-day dose group. However, there was not a proportional increase in each of the T cells
subsets, which was represented by increased numbers of cytotoxic/suppressor cells and
decreased numbers of helper/inducer cells. Although this resulted in a lower helper/suppressor
ratio in the 0.67 ng/kg-day group, the values were within the measured normal range.  Peak
antibody level and antibody response to tetanus toxoid immunization was not altered compared
to control values at either dose tested.  Macrophage depletion in the 0.12, and 0.67 ng/kg-day
groups resulted in the absence of amplification in a mixed lymphocyte response assay, compared
to a fivefold amplification in control monkeys.  As previously reported, the 0.67 ng/kg-day dose
group had reduced reproductive rates (Bowman et al.,  1989b) and the mean number of days of
offspring survival also decreased.
       The surviving offspring from the TCDD-exposed mothers were examined using the same
immune panel used on the mothers and described above.  The only material finding was an
immune hyperresponsiveness to tetanus toxoid immunization which correlated with TCDD tissue
levels (r = 0.40). However, this effect seems to be driven by only two of the offspring, and its
biological significance is unknown. There was no  correlation between TCDD body burdens in
the offspring with a mother monkey's TCDD dose (i.e., offspring with the  highest TCDD tissue
levels were born as often to mothers exposed to 0.12 ng/kg-day as 0.67 ng/kg-day).
       In the  absence of any relevant immunotoxicity endpoints or functional decrements of
immune function following TCDD exposure, neither a NOAEL nor a LOAEL can be established
for this study.
D.l.5.5.    Kociba et al (1978)
       Sprague-Dawley rats (50 per sex per treatment group) were administered TCDD (purity
>99%) in the diet at doses of 0, 1, 10, or 100 ng/kg-day for 2 years.  Body weights and food
consumption were routinely measured.  Hematology, clinical chemistry, and urinalysis were
measured after 3, 12, or 23 months of treatment. Animals were routinely palpitated for tumors.
Gross and histopathological exams were conducted on the tissues of dead or dying animals or at
terminal sacrifice. Specific organs also were weighed.
                                         D-67

-------
       The high-dose females had a statistically significant (p < 0.05) increase in mortality
compared to the controls during the second half of the study.  Mortality changes in males were
variable and of questionable toxicological significance.  A significant (p < 0.05) reduction in
body weight occurred in the 100 ng/kg-day males and females beginning at 6 months. Mid-dose
females also had reduced body weight, but to a lesser degree during the same time frame. There
were no consistent changes in food consumption. The following significant (p < 0.05)
hematology changes were observed in the high-dose animals: decreased packed cell volume in
males after 3 months and in females after 1 year, decreased red blood cells in females after
1 year and in males at terminal sacrifice, decreased hemoglobin in males after 3 months and in
females after 1 year, and decreased total white blood cell count in females after 1 year. Changes
in clinical chemistry (p < 0.05) occurred only in high-dose females and consisted of an increase
in serum alkaline phosphatase and gamma glutamyl transferase.  Significant changes in
urinalysis occurred only in females and included increased urinary  coproporphyrin in the mid-
and high-dose groups, increased urinary uroporphyrin in the mid- and high-dose groups, and
increased urinary delta-amino-levulinic acid  in the high-dose group. Significant (p < 0.05)
changes in relative organ weights were observed, including increased liver weight in mid- and
high-dose females and decreased thymus weight in high-dose females.  Mid- and high-dose rats
showed hepatocellular degeneration and inflammatory and necrotic changes in the liver.  Thymic
and splenic atrophy were noted in high-dose  females. An increase  in non-neoplastic lung lesions
was noted in mid-dose females and high-dose males and females. High-dose females had an
increase in uterine changes.  High-dose males had a significant (p < 0.05) increase in the
incidence of stratified squamous cell carcinomas of the tongue.  High-dose males and females
had a significant (p < 0.05) increase in the incidence of squamous cell carcinomas of the hard
palate/turbinates.
       The chronic (2-year) LOAEL is 10 ng/kg-day, based on the numerous significant
(p < 0.05) changes noted in coproporphyrin excretion (67% increase above control) and an
increase in liver and lung lesions in female rats. The NOAEL is 1 ng/kg-day.
D.l.5.6.    Maronpot et al (1993)
       An initiation-promotion study was performed in female Sprague-Dawley rats (8-10 rats
per group). The rats were initiated with saline (S) or diethylnitrosamine (DEN), followed
                                         D-68

-------
2 weeks later by promotion with biweekly administration of TCDD (purity not specified) in corn
oil via gavage for 30 weeks.  The doses were stated to be equivalent to 3.5, 10.7, 35.7, or
125 ng/kg-day.  The rats were sacrificed 7 days after the final treatment. A significant (p < 0.05)
decrease in body weight occurred in the 125 ng/kg-day group. A significant (p < 0.05) increase
in relative liver weight occurred in the 35.7 and 125  ng/kg-day groups. There was a significant
(p < 0.05) increase in the labeling index in the  125 ng/kg-day group, but only with DEN
initiation.  In the TCDD-alone group, a twofold increase in labeling index  occurred in the
125 ng/kg-day group that did not reach statistical significance. A significant (p < 0.05) trend test
for increased alkaline phosphatase levels was observed in TCDD-treated animals; despite a
50% increase in the highest dose group, the increase was not statistically significant from
controls via a pairwise comparison. Total cholesterol and triglycerides were significantly
(p < 0.05) higher in the 125 ng/kg-day TCDD-alone group. A significant (p < 0.05) increase in
5'-nucleotidase occurred in the 35.7 and 125 ng/kg-day TCDD-alone groups.  A dose-dependent
increase in the incidence and severity of liver toxicity as measured by microscopic lesions was
observed.
       The 30-week LOAEL is 35.7 ng/kg-day, based on a significant (p < 0.05) increase in
relative liver weight (12%, accompanied by increases in incidence and severity of liver lesions).
The 30-week NOAEL is 10.7 ng/kg-day.


D.I.5.7.    National Toxicology Program (1982)
       National Toxicology Program (NTP, 1982) conducted a carcinogenic bioassay of TCDD
on rats and mice. Fifty male and female Osborne-Mendel rats and male and female B6C3Fi
mice were treated twice per week with TCDD (purity not specified) in corn oil via oral gavage at
doses of 0, 5, 25, or 250 ng/kg for rats and male mice (1.4, 7.1, 71  ng/kg-day adjusted for
continuous exposure; administered doses multiplied by 2 and divided by 7) and 0, 20, 100, or
1,000 ng/kg for female mice (5.7, 28.6, or 286  ng/kg-day adjusted  for continuous dosing;
administered doses multiplied by 2 and divided by 7) for 104 weeks. Seventy-five rats and mice
of each sex served as vehicle controls.  One untreated control group of 25 rats and mice of each
sex was present in the TCDD treatment room and one untreated control group consisting of
25 rats and mice of each sex were present in the vehicle-control room. Animals surviving until
                                          D-69

-------
study termination were sacrificed at 105 or 108 weeks. A complete histopathological evaluation
was conducted on all animals.
       Survival rates were not affected by TCDD exposure in rats or mice of either sex. Male
rats exhibited a dose-related depression in mean body weight after Week 55, while the females
exhibited a dose-related body-weight depression after 45 weeks of TCDD exposure. However,
the magnitude of the body weight response is not indicated.  Mean body weights in male and
female mice were comparable to the vehicle control group throughout the bioassay. Noncancer
histopathologic findings included increased incidences of liver lesions (termed toxic hepatitis)
from TCDD exposure, and were detected in the high-dose rats and high-dose mice of each sex.
       A LOAEL for TCDD of 1.4 ng/kg-day for a 104-week exposure duration is identified for
increased incidences of liver lesions in mice of both sexes.  A NOAEL cannot be determined for
this study.

D.l.5.8.    National Toxicology Program (2006)
       Female Sprague-Dawley rats (81 control; 82 treatment group) were administered TCDD
(purity >98%) in corn oil:acetone (99:1) via gavage at doses of 0, 3,  10, 22, 46, or
100 ng/kg-day, 5 days  per week for 105 weeks (0, 2.14, 7.14, 15.7, 32.9, or 71.4  ng/kg-day,
adjusted for continuous exposure) (NTP, 2006).  In addition to this primary group, a stop group
of 50 animals was administered 100 ng/kg-day TCDD in  corn oil:acetone (99:1) via gavage for
30 weeks and then just the vehicle for the remainder of the study. Up to  10 rats per dose group
were sacrificed and evaluated at 14, 31, or 53 (n = 8) weeks for biologically noteworthy changes
in the incidences of neoplasms or non-neoplastic lesions in the liver, lung, oral mucosa, uterus,
pancreas, thymus, adrenal cortex, heart, clitoral gland, ovary, kidney, forestomach, bone marrow,
mesentery gland, and pituitary gland.  All interim sacrifice animals also received a complete
necropsy and microscopic examination, and the following organs were weighed:  the left kidney,
liver, lung, left ovary, spleen, thymus (14 weeks only), and thyroid gland. Out of 53 control
animals and 53 or 54 animals per treatment group not used for interim sacrifice analyses, at study
termination the number of surviving animals had declined to 25 in the control group and to 21,
23, 19, 22, and 21 in five treatment groups, respectively, due to accidental deaths, moribund
animals, or death due to natural causes.
       Survival rate was not affected by TCDD treatment.  Mean body weights in the high dose
primary study group and the 100 ng/kg stop group were less than the vehicle control group after
                                         D-70

-------
Week 13 of the study. The mean body weights of animals in the 46 ng/kg-day group were less
than in the vehicle control at study termination (2 years), whereas animals in the 22 ng/kg-day
had lower mean body weights compared to controls during the last 10 weeks of study.  In
addition to body weight changes, liver weights were also impacted as a result of TCDD
exposure. Absolute and relative liver weights were significantly (either/* < 0.01 orp < 0.05)
higher in all dose groups compared to controls at the 14- and 31-week evaluation period, whereas
the relative liver weights were significantly (either/* < 0.01 orp < 0.05) higher only at
>10 ng/kg-day at 53 weeks.
       No clinical findings associated with TCDD treatment were observed. TCDD caused
changes in thyroid hormone levels at 14, 31, and 53 weeks.  The following changes were
statistically significant (p < 0.05) compared to the vehicle control: decrease in TT4 at doses
>22 ng/kg-day at 14 and 31 weeks and at doses >46 ng/kg-day at 53 weeks; decrease in FT4 at
doses >22 ng/kg-day at 14 and 31 weeks; increase in total T3 at doses >46 ng/kg-day at 14 and
31 weeks and at doses >10 ng/kg-day at 53 weeks; and increase in TSH at doses >46 ng/kg-day
at  14 weeks. There was a statistically significant (p < 0.05) increase in  hepatocyte proliferation
at  14 weeks (22 ng/kg-day group only); 31 weeks  (all doses); and 53 weeks (>46  ng/kg-day).
There were statistically significant (p < 0.01) dose-dependent increases in liver (includes EROD
[CYPlAl-associated] activity; 7-PROD [CYP2B-associated] activity; and acetanilide-
4-hydroxylase [CYP1 A2-associated] activity) and lung (EROD) cytochrome P450 enzyme
activities in all treatment groups at all three evaluation periods compared to the vehicle control
group. The largest effect was  an 82-fold induction of hepatic EROD activity in the 46  ng/kg-day
group at 31 weeks.
       TCDD was detected at the greatest concentration in the liver, followed by fat tissue, with
tissue concentration increasing in both of these tissues in a dose-dependent manner. TCDD
tissue levels generally remained constant after the first measurement at  Week 14. Pathological
examination at Week 14 revealed increased incidences of hepatocellular hypertrophy in animals
administered >10 ng/kg-day TCDD.  Examinations at Weeks 31 and 53 indicated that incidence
and or severity of hepatocellular hypertrophy was  increased at all treatment doses although
incidences were statistically significant (p < 0.05)  only at >10 ng/kg-day doses.  The incidence of
non-neoplastic hepatic lesions (including inflammation, necrosis, multiple eosinophilic focus,
diffuse fatty change, pigmentation, toxic hepatopathy) in the liver increased at doses
                                         D-71

-------
>22 ng/kg-day beginning at 14 weeks. The severity of the lesions increased at 14 weeks at doses
>46 ng/kg-day, but lesions were also observed at lower dose levels during later evaluation
periods (31 and 53 weeks). By terminal  sacrifice, numerous non-neoplastic changes were noted
in TCDD treated rats, even at the lowest dose tested.
       Noncancer cardiovascular and pulmonary effects were evident after 2 years of TCDD
exposure.  Significantly increased incidences of minimal to mild cardiomyopathy were seen in
male and female  rats at >10 ng/kg-day. In the lung, there was a significant (p < 0.01)
dose-dependent increase, when compared to the vehicle control, in the incidence of bronchiolar
metaplasia of the alveolar epithelium at all dose groups in the primary study.
       A LOAEL for TCDD of 2.14 ng/kg-day adjusted dose for a 105-week exposure duration
is identified in this study for significantly (either/? < 0.01 or p < 0.05) increased absolute and
relative liver weights, increased incidence of hepatocellular hypertrophy, and increased incidence
of alveolar to bronchiolar epithelial metaplasia. A NOAEL cannot be determined for this study.


D.l.5.9.    Sewall et al (1993)
       Sewall  et al. (1993) examined the impact of TCDD exposure on the hepatic epidermal
growth factor receptor (EGFR)  as a critical effect in hepatocarcinogenicity. In two separate
experiments, groups of 6- to 8-week-old female Sprague-Dawley rats were randomly assigned to
the following groups: control group, receiving saline and  corn oil; a promoted group that
received four different doses of TCDD along with  saline; a DEN-only initiated control group;
and a DEN and TCDD initiated and promoted group that  received four different doses of TCDD.
DEN was administered via intraperitoneal injection at a dose of 175 mg/kg [S vehicle] as the
initiating agent to animals that were 70 days old. The control animals received saline only.  In
the first experiment, each treatment group (S/TCDD and DEN/TCDD) that included
sham-operated or ovariectomized and intact animals were treated with TCDD (purity >98%) at
125 ng/kg-day. In the  second dose-response experiment,  DEN-initiated and saline  control
treatment groups (intact animals, 84 days old) were administered TCDD (purity >98%) in corn
oil via oral gavage once every 2 weeks for 30 weeks at doses equivalent to 0, 3.5, 10.7, 35.7, or
125 ng/kg-day (n = 9). A week after the last treatment, all animals were sacrificed  and livers
were harvested and fixed for immunohistochemistry.  Sections of the fixed liver were tested for
                                         D-72

-------
EGFR binding, EGFR autophosphorylation, immunolocalization of EGFR, and hepatic cell
proliferation.
       In the first experiment, intact animals treated with 125 ng/kg-day TCDD exhibited a
65% reduction in EGFR binding capacity.  In contrast, the EGFR equilibrium maximum binding
capacity (Bmax) of the ovariectomized rats was not statistically different from the ovariectomized
control rats, and no changes in the dissociation constant (Kd) were detected in any treatment
group. In the dose-response experiment with intact animals, a significant (p < 0.05) TCDD dose-
dependent decrease in the Bmax of EGFR was shown. A two-factor, five-level ANOVA
indicated that the effect of TCDD exposure on EGFR Bmax was significant (p = 0.0001),
whereas, the effect of DEN treatment on EGFR Bmax was not significant.  Comparative analysis
using Fisher's protected least significant difference test indicated that the  lowest TCDD dose
resulting in a statistically significant (p < 0.05) decrease in the EGFR Bmax was 10.7 ng/kg-day
S/TCDD group. At the highest TCDD dose of 125 ng/kg-day, the EGFR  Bmax was reduced by
38% compared to controls in both the DEN initiated and noninitiated groups. A two-factor,
five-level ANOVA showed no significant effect on EGFR Kd in either the DEN- or the
TCDD-treated groups. The EGFR autophosphorylation assay indicated that, with increasing
TCDD dose, the amount of EGFR autophosphorylation in DEN/TCDD-treated animals
decreased. The study authors state that this decrease is similar to the dose-response alterations
observed for the EGFR Bmax.  Additionally, EGFR autophosphorylation in control and
125 ng/kg-day noninitiated animals was similar to the corresponding dose levels for the DEN-
treated animals, suggesting that DEN treatment did not affect the EGFR or the EGFR response to
TCDD under the experimental conditions.  The immunolocalization assay indicated that staining
was more apparent in the centrilobular and midzonal regions of the liver in the DEN initiated
control animals, whereas, the amount of hepatocyte plasma membrane staining in DEN/TCDD
treated animals substantially decreased. The cell proliferation assay showed a decrease in the
cell labeling index in the 3.5 ng/kg-day DEN/TCDD dose group that was  statistically less
(p < 0.05) than the labeling index for the control group. In contrast, the labeling index for the
125 ng/kg-day DEN/TCDD treatment group was significantly (p < 0.05) higher compared to
controls. Except for the low-dose (3.5  ng/kg-day) group,  a clear dose-response trend
(two mid-level doses were not statistically significant) was observed in the other three TCDD
treated groups.
                                         D-73

-------
       The role of EGFR in TCDD-mediated hepatotoxicity is unknown, and as such, this
endpoint cannot be unequivocally linked to TCDD-induced hepatotoxicity nor labeled as
adverse. Thus, no LOAEL/NOAEL was established. A LOEL for TCDD of 3.5 ng/kg-day for a
30-week exposure duration was identified in this study for a significant (p = 0.0001 using
ANOVA) decrease in EGFR Bmax levels.  A NOEL cannot be determined for this study.


D.l.5.10.   Sewatt et al (1995a)
       Sewall et al. (1995a) studied the dose-response relationship for thyroid function
alterations in female rats as a result of TCDD exposure.  Groups of female Sprague-Dawley rats
were initiated with DEN at 70 days of age at a dose of 175 mg/kg in a saline vehicle via an i.p.
injection. DEN was administered as a liver-initiating agent for a concurrent study to determine
TCDD promotion of hepatic preneoplastic foci. Saline-treated animals served as controls. At
84 days of age, both the DEN-initiated and the saline-noninitiated groups of animals were
administered TCDD (purity >98%) or corn oil vehicle via oral gavage once every 2 weeks for
30 weeks at dose levels equivalent to 0, 0.1, 0.35, 1.0, 3.5, 10.7, 35.7, or 125 ng/kg-day
(n = 9 per group). One week after the last TCDD treatment, the animals were sacrificed and the
thyroid was removed and fixed for further analysis. Blood was drawn from the  abdominal aortic
vein, and the serum was isolated and preserved for hormone analysis. Liver was also removed
and prepped for further analysis. Thyroid hormone analysis was performed to determine serum
TSH, T3, and T4 levels using radioimmunoassay kits. Histological examination was conducted
on eosin-stained sections of the thyroid tissue. RNA level in the hepatic tissue was determined
using a RT-PCR technique.
       TCDD treatment did not affect thyroid weight. A dose-dependent decrease in serum
T4 levels was observed in both noninitiated and DEN-initiated animals with T4  levels dropping
significantly (p < 0.05) at the 35 and 125 ng/kg-day TCDD doses in the noninitiated group.
Compared to the noninitiated control group, DEN alone  did not significantly affect T4 levels.
Serum T3 level in the 125 ng/kg-day treatment group was slightly elevated but was not
significantly different from levels in the control group.  TSH levels in DEN initiated rats were
increased at a dose of 3.5 ng/kg-day. In the noninitiated group, TSH level in the 125 ng
TCDD/kg-day group was 3.27 ± 0.34 ng/mL (n = 9) compared to 1.3 ± 0.18 ng/mL in the corn
oil control group (n = 7). This result, in conjunction with the T4 data, demonstrates that TCDD
                                         D-74

-------
had a similar effect on thyroid hormone levels in both the noninitiated and DEN initiated groups.
Histological sections examined for nodular lesions or neoplasms exhibited thyroid follicular
adenoma in one DEN/corn oil control animal.  The DEN/TCDD-treated animals exhibited
diffuse follicular hyperplasia, with the size of colloidal follicles decreasing with TCDD
treatment. Other qualitative DEN/TCDD-related changes included increased frequency of
abnormally shaped follicles. The study authors reported that image analysis demonstrated a
significant (p = 0.013) TCDD dose-related decrease in mean follicle size along with a significant
(p = 0.001) TCDD dose-related increase in parenchymal area.  Additionally, like T4 and TSH
levels, DEN treatment alone or in combination with TCDD did not influence thyroid follicular or
C-cell morphology.
       RT-PCR results for UGT1 and CYP1 Al mRNA levels indicated that the amount of
UGT1 mRNA at the 125 ng/kg-day dose was approximately 2.5-fold higher compared to the
concurrent controls. The study authors also stated that the maximal response for the UGT1
mRNA levels was reached at a dose between 1.0 and 3.5 ng TCDD/kg-day.  In contrast, the
maximum induction of CYP1A1  mRNA was 260-fold higher at the 125 ng/kg-day compared to
the concurrent controls.
       A LOAEL for TCDD of 35 ng/kg-day for a 30-week exposure duration was identified in
this study for a significant (p < 0.05) decrease in T4 levels.  The NOAEL for this study is
10.7 ng/kg-day.


D.l.5.11.   Toth et al (1979)
       Toth et al.  (1979) examined the impact of TCDD exposure on the formation of liver
tumors in male mice.  Ten-week-old, outbred Swiss/H/Riop male mice were administered
sunflower oil or TCDD (purity not specified; in sunflower oil) at 0, 7, 700 or 7,000 ng/kg (0, 1,
100, or 1,000 ng/kg-day adjusted for continuous dosing;  administered dose divided by 7; n = 38,
44, 44, and 43, respectively) once per week via gastric tube for 1 year.  Once exposure had
ceased, animals were followed for the rest of their lives.  After spontaneous death or when mice
were moribund, autopsies were performed and all organs were examined histologically.
       Average life span in the 1,000 ng/kg-day dose group decreased considerably (72%) when
compared to the control group. TCDD also caused dose-dependent, severe chronic and ulcerous
skin lesions (12, 30, and 58% in the 1, 100, and 1,000 ng/kg-day dose groups, respectively) that
                                         D-75

-------
was followed by generalized lethal amyloidosis (12, 23, and 40% in the 1,  100, and
1,000 ng/kg-day dose groups, respectively).
       A LOAEL for TCDD of 1 ng/kg-day for 1-year exposure duration was identified in this
study for severe chronic and ulcerous skin lesions (12% higher than controls), and generalized
lethal amyloidosis (12% higher than controls). A NOAEL cannot be determined for this study.


D.l.5.12.   Tritscher et al (1992)
       An initiation-promotion study was performed in female Sprague-Dawley rats (at least
nine rats per group). Rats were initiated with an i.p. injection of diethylnitrosamine (DEN,
175 mg/kg) or saline, followed 2 weeks later by promotion with biweekly administration of
TCDD (purity not specified) in corn oil via gavage for 30 weeks.  The doses were stated to be
equivalent to 3.5, 10.7, 35.7, or 125 ng/kg-day; control animals received corn oil. Rats were
sacrificed 7 days after the final treatment and the livers were removed for further analysis. Liver
TCDD concentrations were analyzed in DEN-initiated rats by gas chromatography-mass
spectrometry. Hepatic cytochrome P450 levels (CYP1A1 and CYP1A2) and EROD activity
were quantified in DEN/TCDD-treated rats, and immunohistochemical detection of CYP1A1
and CYP1A2 in liver was also conducted.
       A linear relationship between administered dose of TCDD and liver TCDD concentration
on a wet weight (r = 0.999) and lipid-adjusted basis (r = 0.993) was observed. A significant
(p < 0.01) dose-response trend for increased CYP1 Al and CYP1A2 protein in the liver (hepatic
microsomes) was observed in initiated and noninitiated rats. However, there were higher
constitutive levels of the two CYP isozymes in nonintiated rats which produced a lower
magnitude of induction by TCDD compared to the TCDD-alone group; there were no
statistically significant differences between initiated and noninitiated rats at any dose tested.  A
strong relationship between liver TCDD concentration and CYP1A1 and CYP1A2 protein levels
and EROD activity was also observed in DEN/TCDD-treated rats.  Immunohistochemical
staining of the serial liver sections for CYP1A1 and CYP1A2 protein from initiated and
noninitiated  rats exhibited a dose-dependent increase consistent with that observed via
microsomal  quantification. Immunolocalization and pattern of induction were also similar for
both CYP isozymes. However, distribution pattern of positive immunoreactivity of the two CYP
isozymes was varying, with the most intense staining observed around central veins.
                                         D-76

-------
       CYP induction alone is not considered a significant lexicologically adverse effect given
that CYPs are induced as a means of hepatic processing of xenobiotic agents. Thus, no LOAEL
or NOAEL was established for this study because adverse endpoints (e.g., indicators of
hepatotoxicity) were not measured.

D.1.6.   Chronic Studies (Cancer Endpoints)
D. 1.6.1.    Delia Porta et al (1987)
       Delia Porta et al. (1987) studied the long-term carcinogenic effects of TCDD in B6C3Fi
(C57BL/6JDp x C3Hf/Dp) mice.  Six-week-old male and female mice (initially about
15/sex/dose, and increased by approximately 30 to 40 per group within a few weeks) were
administered 0, 2,500, and  5,000 ng/kg TCDD (purity not provided) in corn oil by oral gavage
once per week for 52 weeks (0, 357, and 714 ng/kg-day adjusted for continuous exposure).  At
ages 31 to 39 weeks, 41 male mice and 32  female mice in the 2,500 ng/kg dose group were
mistakenly administered a single dose of 25,000 ng/kg TCDD. TCDD treatment for the
2,500 ng/kg dose group was halted for 5 weeks (beginning the week after the 25,000 ng/kg dose
was administered in error) and resumed until exposure was terminated at 57 weeks.  Mortality
was observed and body weights recorded at unspecified intervals until 110 weeks of age, when
all surviving animals were  sacrificed and necropsied. Histopathological analysis was conducted
on the following organs and tissues: Harderian glands, pituitary, thyroid, adrenals, tongue,
esophagus, and trachea; lungs, liver, pancreas; spleen, kidneys, and bladder; testes, ovaries,  and
uterus, mesenteric lymph nodes, small intestine, and all  other organs with presumed pathological
changes.
       The body weights of both male and female mice exposed to 2,500 and 5,000 ng/kg
TCDD were markedly lower than in the corresponding control groups (statistical significance not
reported).  Relative to the controls, a significant (p < 0.001), dose-related decrease in survival
occurred in animals treated with either dose of TCDD. In the subset of animals treated
inadvertently with a single  dose of 25,000  ng/kg TCDD, mortality in male mice increased shortly
after this treatment; females, however, did  not show a mortality increase following the
inadvertent treatment. This mortality in male mice was  associated with subcutaneous edema,
degenerative hepatocyte changes, and  bile  duct hyperplasia.  The incidence of non-neoplastic
lesions (such as amyloidosis of the liver, spleen, adrenals, and pancreas), liver necrosis, and
                                         D-77

-------
nephrosclerosis, was increased in mice exposed to TCDD compared to controls (statistical
significance not reported).
       The study authors used two statistical tests to analyze tumor incidence. Because of the
increased mortality in treated groups compared to controls, one test, which assumes all tumors
are fatal, overestimated the differences between the treated and control groups. The second test
assumes that all tumors are incidental and resulted in an underestimation of TCDD effects.  Both
tests were used to analyze the results for nonthymic lymphomas and hepatic adenomas and
carcinomas. Incidence of nonthymic lymphomas (6/45, 4/51, and 3/50 in the 0, 2,500, and
5,000 ng/kg dose groups, respectively in males and 17/49, 21/42, and 17/48 in the 0, 2,500, and
5,000 ng/kg dose groups, respectively in females) was significantly (p < 0.05 in males and
p < 0.01 in females) higher in TCDD-treated animals compared to the corresponding controls
using the fatal tumor test. However, the incidental tumor test showed that this higher incidence
was not significant.  Similarly, a significantly (p < 0.001) higher incidence of hepatocellular
adenomas occurred in male mice using the fatal tumor test (10/43,  11/51,  and 10/50 in the 0,
2,500, and 5,000 ng/kg dose groups, respectively), but the incidence was not significant when
assessed using the incidental tumor test. Hepatocellular carcinomas in males were significant,
(p < 0.001) using either the fatal or incidental tumor tests (5/43, 15/51, and 33/50 in the 0, 2,500,
and 5,000 ng/kg dose groups,  respectively). In female mice, hepatocellular adenomas were
significant using both the fatal (p < 0.01) and incidental (p < 0.001) tumor tests (2/49, 4/42, and
11/48 in the 0, 2,500, and 5,000 ng/kg dose groups, respectively).  Similar results for female
mice were obtained for incidence of hepatocellular carcinomas (1/49, 12/42, and 9/48 in the 0,
2,500, and 5,000 ng/kg dose groups, respectively), which also were significant using both the
fatal (p < 0.01) and incidental (p < 0.05) tumor tests. TCDD-related incidences of other tumor
types in both sexes were uniformly low and comparable in the treatment and control groups.
       These results indicate that TCDD is carcinogenic in male and female B6C3Fi  mice,
causing hepatocellular adenomas and  carcinomas in both sexes.
       In addition to the long term bioassay results in mice described by Delia Porta et al.
(1987), carcinogenic effects of TCDD in a neonatal bioassay were reported in the same
publication.  Briefly, groups of male and female B6C3Fi and B6CF1 (C57/BL6J x BALB/c)
mice were treated with 0, 1,000, 30,000 or 60,000 ng/kg BW TCDD via i.p. injection beginning
at PND 10.  Animals were treated once weekly for 5 weeks and then observed until 78 weeks of
                                          D-78

-------
age.  However, because this study utilized i.p. injection as the route of TCDD exposure, it does
not qualify for further consideration based on the study selection criterion that the study design
consist of orally administered TCDD.
D.l.6.2.    Kodbaetal (1978)
       As discussed above, Kociba et al. (1978) conducted a lifetime (2-year) feeding study of
male and female Sprague-Dawley rats using doses of 0, 1, 10, and 100 ng/kg-day.  There were
50 males and 50 females in each group.
       With respect to the cancer endpoints examined, the most significant finding was an
increase in hepatocellular hyperplastic nodules and hepatocellular carcinomas in female rats.
The incidence of hepatocellular carcinomas was significantly elevated above the control
incidence at the 100 ng/kg-day dose, whereas increased incidence of hyperplastic nodules was
evident in the 10 ng/kg-day dose group.
       There have been two reevaluations of slides of liver sections from the Kociba et al. study
(Goodman and Sauer, 1992; Sauer,  1990; Squire, 1990). The Squire Review was requested by
EPA as an independent review of the slides. The Sauer Review was carried out using refined
criteria for the diagnosis of proliferative hepatocellular lesions (Maronpot et al.,  1989; Maronpot
et al., 1986). Liver tumor incidences for the three evaluations are compared in Appendix F.
Although there are some quantitative differences between the evaluations, the lowest detectable
effect for liver tumor incidence is consistently observed at 10 ng/kg-day.
       In the 10 ng/kg-day dose group, significant increases in the incidence of hyperplastic
nodules of the liver were observed in female rats (18/50 in the Kociba evaluation, 27/50 in the
Squire evaluation). Two females (2/50) had hepatocellular carcinomas.  In the 1990 reevaluation
(Goodman and Sauer, 1992; Sauer,  1990), nine females (9/50) were identified with
hepatocellular adenomas and none with carcinomas; thus only one-third of the previously
observed "tumors" were identified when using the refined diagnostic criteria.  As discussed
below, the tumor reclassification of Goodman and Sauer (1992) was used in the dose-response
modeling for the Kociba et al. (1978) data set.
       In addition to nodules in the liver, increased incidence of stratified squamous cell
carcinoma of the tongue and nasal turbinates/hard palate, and keratinizing squamous cell
carcinoma of the lung were also observed in female rats in the 100 ng/kg-day dose group.
                                         D-79

-------
One possible cause for the induction of lung tumors in the Kociba feeding study may have been
the aspiration of dosed feed into the lungs. However the promotion of lung tumors has been
observed in mice treated systemically by i.p. injections of TCDD (Beebe etal.,  1995). In
addition the induction of hyperplastic and metaplastic lesions in rats has been observed following
chronic oral gavage treatment with TCDD (Tritscher et al., 2000). More recently, chronic oral
exposure to heptachlorodibenzodioxin (HpCDD) resulted in the induction of lung tumors in
treated female rats (Rozman,  2000). These data indicate that the induction of lung tumors in the
Kociba study was most likely primarily the result of systemic chronic dietary exposure to TCDD
rather than due to a localized  exposure to aspired dosed feed.
       There was no detectable increase in liver tumor incidences in male rats in any of the dose
groups. The mechanism responsible for dioxin-mediated sex specificity for
hepatocarcinogenesis in rats is not clear, but may involve ovarian hormones  (Lucier et al., 1991).
       Although there was no increase in liver tumors in male rats in this study, in the
100 ng/kg-day group, there was an increased incidence of stratified squamous cell carcinoma of
the hard palate/nasal turbinate, stratified squamous cell carcinoma of the tongue, and adenoma of
the adrenal cortex.
       Kociba et al. (1978) had reported that chemically related increases in preneoplastic or
neoplastic lesions were not found in the 1 ng/kg-day dose group. However,  Squire identified two
male rats in the  1 ng/kg-day dose group with squamous cell carcinoma of the nasal
turbinates/hard palate, and one of these male rats had a squamous cell carcinoma of the tongue.
These are both rare tumors in Sprague-Dawley rats, and these sites are targets for TCDD,
implying that 1 ng/kg-day may not represent a NOEL. However, no dose-response relationships
were evident for tumors at these sites (Huff et al., 1991).
       There is considerable  controversy concerning the possibility that TCDD-induced liver
tumors are a consequence of cytotoxicity. Goodman and Sauer (1992) have  extended the
reevaluation of the Kociba slides to include liver toxicity data and have reported a correlation
between the presence of overt hepatotoxicity and the development of hepatocellular neoplasms in
female rats.  With the exception of two tumors in controls and one each in the low- and mid-dose
groups, all liver tumors occurred in livers showing clear signs of toxicity. However, male rat
livers exhibit cytotoxicity in response to high TCDD doses, yet they do not develop liver tumors.
Moreover, both intact and ovariectomized female rats exhibit liver toxicity in response to TCDD,
                                         D-80

-------
yet TCDD is a more potent promoter in intact but not ovariectomized rats (Lucier et al., 1991).
Therefore, if cytotoxicity is playing a role in liver turnorigenesis, other factors must also be
involved. Also, there is little information on the role of cytotoxicity in TCDD-mediated cancer
at other sites such as the lung and thyroid.


D.l.6.3.    Toth et al (1979)
       In a study of 10-week-old outbred male Swiss/H/Riop mice, Toth et al. (1979)
administered oral gavage TCDD doses of 0, 7, 700, and 7,000 ng/kg-day in sunflower oil weekly
for 1 year (0, 1, 100, or 1,000 ng/kg-day adjusted for continuous dosing; see details above). All
mice (100/group) were followed for their entire lives. The study authors identified the effective
number of mice in each group to be the number of surviving animals when the
first tumor-bearing animal was identified. The average lifespan of the control, low, mid and high
dose groups was 588, 649, 633, and 424 days, respectively.
       In the 100 ng/kg-day dose group,  liver tumor incidence was twice that of the control
group and was statistically significant (p < 0.01%). A dose-related increase in liver tumor
incidence was observed (18, 29, 48, and 30% in the control and three TCDD-treated groups,
respectively) in all treated mice.  Increases were not statistically significant, however, at 1 and
1,000 ng/kg-day. The study authors also stated that spontaneous and induced liver tumors were
not histologically different. Additionally, the ratio of benign hepatomas to hepatocellular
carcinomas in the control group was not affected by treatment and an increase was observed only
in the absolute number of liver tumors. Cirrhosis was not observed with the tumors.
D.l.6.4.    NTP (1982)
       As discussed above, the NTP (1982) study was conducted using Osborne-Mendel rats
and B6C3Fi mice (NTP, 1982). Groups of 50 male rats, 50 female rats, and 50 male mice
received TCDD as a suspension in corn oil:acteone (9:1) by gavage twice each week at doses of
0, 5, 25, or 250 ng/kg-day (daily averaged doses of 0, 1.4, 7.1, or 71 ng/kg-day for rats and male
mice and doses of 0, 5.7, 28.6, or 286 ng/kg-day for female mice.
       There were no statistically significant dose-related decreases in survival in any
sex-species group.  TCDD-induced malignant liver tumors occurred in the high-dose female rats
                                          D-81

-------
and in male and female mice.  These can be considered to result from TCDD exposure because
they are relatively uncommon lesions in control Osborne-Mendel rats (male, 1/208; female,
3/208), are seen in female rats and mice of both sexes, and their increasing incidence with
increasing dose is statistically significant (Cochran-Armitage trend test,/? = 0.004). Because
liver tumors were increased in both sexes of mice, this effect is not female-specific as was
observed in rats.  Interestingly, liver tumor incidences were decreased in female rats in both the
NTP and Kociba low doses (not statistically significant compared with controls). For example,
the combined control incidence data were 11/161 (7%) compared with 4/99 (4%) in the low-dose
group.
       The incidences of thyroid gland (follicular cell) tumors were increased in all three dose
groups in male rats.  Because the responses in the two highest dose groups are highly significant,
the statistically significant elevation of incidence in the lowest dose group (Fisher exact
/7-value = 0.042) is considered to be caused by exposure to TCDD, suggesting that thyroid tumor
incidence may be the most sensitive site for TCDD-mediated carcinogenesis. Because
71 ng/kg-day is above the maximum tolerated dose (MTD) (Huff et al., 1991), thyroid tumors
occur at doses more than 50 times lower than the MTD.
       TCDD-induced neoplasms of the adrenal gland  were observed in the 7.1 ng/kg-day/dose
group in male rats and in high-dose female rats.  Fibrosarcomas of the subcutaneous tissue were
significantly elevated in high-dose female mice and female rats. One additional tumor type,
lymphoma, was seen in high-dose female mice.  Lung tumors  were elevated in high-dose female
mice; the increase was not statistically significant when compared with concurrent controls, but
the increase was  dose related (Cochran-Armitage trend test,/?  = 0.004).
       Huff (1992) concluded, based on the NTP bioassay results, that TCDD was a complete
carcinogen and induced neoplasms in rats and mice of both sexes. As was observed in the
Kociba study (1978), liver tumors were observed with greater frequency in treated female rats,
but in male rats the thyroid appears to be the most sensitive (increased tumor incidence at doses
as low as 1.4 ng/kg-day).
D.l.6.5.    NTP (2006)
       As discussed above, female Sprague-Dawley rats (53 control; 53 or 54 animals per
treatment group) were administered TCDD (purity >98%) in corn oil:acetone (99:1) via gavage
                                         D-82

-------
at doses of 0, 3, 10, 22, 46, or 100 ng/kg-day, 5 days per week for 105 weeks (0, 2.14, 7.14, 15.7,
32.9, or 71.4 ng/kg-day, adjusted for continuous exposure) (NTP, 2006).  In addition to this
primary group, a stop-dose group of 50 animals was administered 100 ng/kg-day TCDD in corn
oil:acetone (99:1) via gavage for 30 weeks and then just the vehicle for the remainder of the
study.  At study termination, the number of surviving animals had declined to 25 in the control
group and to 21, 23, 19, 22, and 21 in five treatment groups, respectively, due to accidental
deaths, moribund animals, or death due to natural causes.
       Incidence of hepatocellular adenomas was significantly (p < 0.001) increased in the
100 ng/kg-day dose group in the primary study and exceeded incidences seen in historical
vehicle control range at study termination. A dose-related increase in the incidence of
cholangiosarcoma was seen in the primary study group in animals receiving 22 ng/kg-day or
higher doses of TCDD. The high dose group of 100 ng/kg-day had the highest incidence of
cholangiosarcoma with a significant (p < 0.001) number of animals exhibiting multiple
cholangiosarcomas. Such an incidence was not seen in historical vehicle controls. In contrast,
only two cholangiosarcomas and hepatocellular adenomas were seen in the 100 ng/kg-day group
in the stop-exposure study.
       In the lung, at 2 years, there was a significantly (p = 0.002) increased incidence of cystic
keratinizing epithelioma in the 100 ng/kg-day dose group of the primary study, while there were
no epitheliomas in the 100 ng/kg-day group of the stop-exposure study.  There was also a
significant (p < 0.01) dose-dependent increase, when compared to the vehicle control, in the
incidence of bronchiolar metaplasia of the alveolar epithelium at all dose groups in the primary
study.  Squamous metaplasia was also present in the 46  and 100  ng/kg-day dose groups in the
primary study, and was also observed in the 100 ng/kg-day dose  group in the stop-exposure
study.
       A positive trend in the incidence of gingival squamous cell carcinoma of the oral cavity
was seen at all doses (except 22 ng/kg-day), with the incidence significantly (p = 0.007) high in
the 100 ng/kg-day dose group.  In addition,  the occurrence of this lesion in the 46 and
100 ng/kg-day group of the primary study and 100 ng/kg-day group of the stop-exposure study
exceeded the historical control range. The incidence of gingival  squamous hyperplasia was
significantly (either/? < 0.01 or p < 0.05) increased in all dose groups of the primary study as
well as the 100 ng/kg-day group of the stop-exposure study.
                                          D-83

-------
       In the uterus, at 2 years, there was a significantly (p = 0.032) higher rate of squamous cell
carcinoma in the 46 ng/kg-day group compared to vehicle controls. In addition there were
two squamous cell carcinomas in the 100 ng/kg-day group of the stop-exposure study. No
squamous cell carcinomas have been reported in historical vehicle controls.
       These results indicate that TCDD is carcinogenic to female Sprague-Dawley rats and
causes tumors at multiple sites.

D.2.  EVALUATION OF STUDIES
       Based on the results of EPA's literature search and collection activities (see Section 2.2
and Figure 2-1), a total of 1,441  studies were examined for their potential to be used in TCDD
quantitative  dose-response analysis (see Figure 2-4 of the main document). Of the 1,441 studies,
49 were epidemiologic cancer or noncancer studies (see Appendix C for their summaries and
evaluations). In addition, there were 637 studies eliminated from consideration because they
were not suitable study types; these included, in vitro bioassays, review articles, PBPK modeling
studies, and  studies that evaluated PCBs or other dioxin-like compounds other than TCDD.  A
list of these studies is not provided in this appendix; results of the initial literature review can be
found online at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=199923#Download.
A total of 755 animal studies were evaluated (4 studies contained both cancer and noncancer
endpoints).  The results are shown and discussed in the remainder of this Section D.2.

D.2.1.  Evaluation of Animal Cancer Bioassays
       A total of eight animal cancer bioassays were available for evaluation (see Figure 2-4)
using EPA's study inclusion criteria (see Section 2.3.2 and Figure 2-3).  Table 2-3 of the main
document presents the six studies that met these criteria and comprise the preliminary list of
cancer bioassays considered suitable for quantitative TCDD dose-response modeling. Only two
of the available animal cancer bioassays did not meet EPA's study selection criteria, and,
therefore, are not summarized in this appendix. These include Eastin et al. (1998), because a
genetically altered mouse strain was tested, and Rao et al. (1988), because an intraperitoneal
injection was used instead of oral route of exposure.
                                          D-84

-------
D.2.2.   Evaluation of Animal Noncancer Bioassays
        Table D-l provides the final list of 78 studies that were selected as key studies for
TCDD noncancer dose-response analyses.  These studies are peer-reviewed, noncancer, in vivo
mammalian bioassays that assessed TCDD dose response, and they meet EPA's study inclusion
criteria (see Section 2.3.2 and Figure 2-3).  Information on each of these studies is provided in
Section D.I of this appendix and in Table 2-4 of the main document.
       An additional 637 studies were excluded from analysis based on one or more of the
following reasons (see Figure 2-4): (1) 66 studies used genetically altered animals;
(2) 370  studies had a lowest tested dose that was too high (i.e., greater than 30 ng/kg-day);
(3) 142  studies tested chemicals that were not TCDD-only or used an unspecified TCDD dose;
and (4)  135 studies employed a nonoral dosing method. Table D-2 shows these studies and
identifies the study inclusion criteria that were not met. For many studies, more than one reason
for exclusion was found.  Conversely, in some cases at least one criterion was not met and was
identified, but, given that the study had already been excluded based on one criterion, not all of
the other criteria for exclusion were further evaluated and identified.

D.3. CROSS-SPECIES CONCORDANCE OF SELECTED HEALTH ENDPOINTS
       This appendix presents a cross-species comparison of NOAELs and  LOAELs for selected
endpoints from the animal bioassay and human epidemiology studies that passed the noncancer
study selection criteria outlined in Section 2.  The tables and figures are intended to illustrate the
degree of qualitative and quantitative concordance of effects across species and the consistency
of observation of those effects across studies within species. Tables D-3 through D-8 provide
these comparison for male reproductive, female reproductive, thyroid, developmental dental,
immune system, and neurological effects, respectively (also illustrated in Figures D-l through
D-6). This analysis goes beyond the one presented in Section 4 (see Tables  4-3 and 4-5) in that
effects at doses higher than the study LOAELs (for most  sensitive effect) are included.
Quantitative concordance is considered in terms of modeled equivalent human exposures, as
displayed on the figures, and actual administered doses (tables only).  Results from animal
bioassays that did not pass the low-dose-maximum selection criterion are not included here, but
may provide additional relevant information.
                                         D-85

-------
       The endpoints evaluated here were chosen because they have been observed in both
human epidemiologic studies and animal bioassays (i.e., male and female reproductive effects,
thyroid hormone levels, and developmental dental effects) and quantified by EPA for reference
dose (RfD) point of departure (POD) consideration, or are sensitive effects in animals but not in
humans (i.e., immunological and neurological effects).  Hepatic effects, which are not included
here, are evident in all rodent studies that looked for them and are often severe; hepatic effects
reported for humans were not as severe (Michalek et al., 2001b).  Diabetes may be a sensitive
health effect in humansQVIichalek and Pavuk, 2008), but no animal bioassays included in this
analysis address diabetes or glucose metabolism. Other animal studies that did not meet the
dose-limit selection criterion may show effects of interest at higher doses.
       Male reproductive effects have been reported in all species (mice, rats and humans) in
which they were evaluated (see Table D-3 and Figure D-l).  Sperm  effects, one of the co-critical
effects in humans selected for the RfD, is observed in more than one rat study, but not in mice, in
the studies selected for this analysis.  Altered sex ratios (i.e., decreased proportion of male
offspring) have been reported for both mice and rats and in one human study (Mocarelli et al.,
2000): the human study was not considered for a POD (see Appendix C for study evaluation
details), and thus is not included in Figure D-l.
       Female reproductive effects also have been reported for all species (mice, rats, monkeys
and humans) in which they were evaluated (see Table D-4 and Figure D-2).  Of particular note
are the more severe effects (i.e., reduced fertility, embryo loss, and reduced offspring survival;
see Table D-4) that have been observed in animal species as compared to humans.  Adverse birth
outcomes were not observed for the Seveso Women's Cohort as reported by Eskenazi et al.
(2003).  Other female reproductive effects observed in humans included lengthened menstrual
cycle reported by Eskenazi et al., (2002) which is the only study that passed the selection criteria
(and is shown in Figure D-2).  Other female reproductive  effects were unable to be evaluated for
RfD POD consideration because a critical exposure window could not be identified for these
effects (see Appendix C); these other health outcomes included early menopause (Eskenazi et al.,
2005) and possible anti-estrogenic effects (Eskenazi et al., 2007).
       Effects of TCDD on thyroid hormones have been reported for rats and humans (see
Table D-5 and Figure D-3) but have not been evaluated in other species in the selected data sets.
                                          D-86

-------
Increased neonatal TSH, the other co-critical effect for the RfD, has only been evaluated for
humans; rat studies have reported decreased serum levels of T3 and T4 in adults.
       Developmental dental defects have also been observed in mice, rats and humans (see
Table D-6 and Figure D-4) but are not a particularly sensitive endpoint for humans, as they are
for mice and rats.  Other relatively sensitive endpoints reported in animal bioassays, such as
immunotoxicity (see Table D-7 and Figure D-5) and neurotoxicity (see Table D-8 and
Figure D-6) do not appear to be sensitive human health outcomes associated with TCDD
exposure. Baccarelli et al.  (2004; 2002) reported decreased IgG levels for some individuals in
the Seveso cohort and concluded that the levels were far above those associated with
immunodeficiency disorders. Michalek et al. (200 Ic) found no evidence of peripheral
neuropathy in Vietnam veterans exposed to TCDD during operation Ranch Hand.
       Overall, the analysis presented here  supports the conclusion that there is a substantial
amount of qualitative concordance of effects between laboratory animal  species and humans, but
lower quantitative concordance.
                                          D-87

-------
Table D-l. Noncancer animal studies selected for TCDD dose-response
analyses
Author (year)
Amin et al. (2000)
Bell et al. (2007c)
Bowman et al. (1989a)
Bowman et al. (198%)
iurleson et al. (1996)
Cantoni et al. (1981)
Chu et al. (2001)
Chu et al. (2007)
Crofton et al. (2005)
Croutch et al. (2005)

DeCaprio et al. (1986)
DeVito et al. (1994)
Fattore et al. (2000)
Fox et al. (1993)
Franc et al. (2001)
Franczak et al. (2006)
Hassoun et al. (1998)
Hassoun et al. (2000)
Hassoun et al. (2002)
Title of study
Gestational and Lactational Exposure to TCDD or Coplanar PCBs Alters Adult
Expression of Saccharin Preference Behavior in Female Rats
Toxicity of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in the Developing Male Wistar(Han)
Rat. II: Chronic Dosing Causes Developmental Delay
Behavioral Effects in Monkeys Exposed to 2,3,7,8-TCDD Transmitted Maternally
During Gestation and for Four Months of Nursing
Chronic Dietary Intake of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) at 5 or 25 ppt
in Monkey: TCDD Kinetics and Dose-effect Estimate of Reproductive Toxicology
Effect of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) on Influenza Virus Host
Resistance in Mice
Porphyrogenic Effect of Chronic Treatment with 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
in Female Rats. Dose-Effect Relationship Following Urinary Excretion of
Porphyrins
Mixture Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin and Polychlorinated Biphenyl
Congeners in Rats
Combined Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin and Polychlorinated
Biphenyl Congeners in Rats
Thyroid-Hormone-Disrupting Chemicals: Evidence for Dose-Dependent Additivity
or Synergism
2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and l,2,3,4,7,8-Hexachlorodibenzo-/>-
Dioxin (HxCDD) Alter Body Weight by Decreasing Insulin-Like Growth Factor I
(IGF-I) Signaling
Subchronic Oral Toxicity of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in the Guinea Pig:
Comparisons with a PCB-containing Transformer Fluid Pyrolysate
Dose-response Relationships in Mice Following Subchronic Exposure to
2,3,7,8-Tetrachlorodibenzo-p-dioxin: CYP1A1, CYP1A2, Estrogen Receptor, and
Protein Tyrosine Phosphorylation
Relative Potency Values Derived from Hepatic Vitamin A Reduction in Male and
Female Sprague-Dawley Rats Following Subchronic Dietary Exposure to Individual
Polychlorinated Dibenzo-p-dioxin and Dibenzofuran Congeners and a Mixture
Thereof
Gene Expression and Cell Proliferation in Rat Liver After
2,3,7,8-Tetrachlorodibenzo-p-dioxin Exposure
Persistent, Low-dose 2,3,7,8-Tetrachlorodibenzo-p-dioxin Exposure: Effect on Aryl
Hydrocarbon Receptor Expression in a Dioxin-Resistance Model
Effects of Acute and Chronic Exposure to the Aryl Hydrocarbon Receptor Agonist
2,3,7,8-Tetrachlorodibenzo-p-dioxin on the Transition to Reproductive Senescence
in Female Sprague-Dawley Rats
Induction of Oxidative Stress in Brain Tissues of Mice after Subchronic Exposure to
2,3,7,8-Tetrachlorodibenzo-/>-dioxin
The Relative Abilities of TCDD and its Congeners to Induce Oxidative Stress in the
Hepatic and Brain Tissues of Rats After Subchronic Exposure
Induction of Oxidative Stress in the Tissues of Rats after Chronic Exposure to
TCDD, 2,3,4,7,8-Pentachlorodibenzofuran, and 3,3',4,4',5-Pentachlorobiphenyl
                                D-88

-------
Table D-l. Noncancer animal studies selected for TCDD dose-response
analyses (continued)
Author (year)
Hassoun et al. (2003)
Hochstein et al.
(2001)
Hoio et al. (2002)
Hone et al. (1989)
Hutt et al. (2008)
Ikeda et al. (2005b)
Ishihara et al. (2007)
FCattainen et al. (2001)
Keller et al. (2007)
Keller et al. (2008a)
Keller et al. (2008b)
FCitchin and Woods
(1979)
Kociba et al. (1976)
Kociba et al. (1978)
Kuchiiwa et al. (2002)
Latchoumycandane
and Mathur (2002)
Latchoumycandane et
al.(2002b)
Latchoumycandane et
al. (2002a)
Latchoumycandane et
al. (2003)
Li et al. (1997)
Li et al. (2006)
Lucier et al. (1986)
Mally and Chipman
(2002)
Markowski et al.
(2001)
Title of study
The Role Of Antioxidant Enzymes In TCDD-Induced Oxidative Stress in Various Brain
Regions of Rats After Subchronic Exposure
Chronic Toxicity of Dietary 2,3,7,8-Tetrachlorodibenzo-p-Dioxin to Mink
Sexually Dimorphic Behavioral Responses to Prenatal Dioxin Exposure
Immune Abnormalities Associated With Chronic TCDD Exposure in Rhesus
The Environmental Toxicant 2,3,7,8-Tetrachlorodibenzo-p-dioxin Disrupts
Morphogenesis of the Rat Pre-implantation Embryo
Repeated hi Utero and Lactational 2,3,7,8-TCDD Exposure Affects Male Gonads in
Offspring, Leading to Sex Ratio Changes in F2 Progeny
Does Paternal Exposure to 2,3,7,8-TCDD Affect the Sex Ratio of Offspring?
hi Utero/Lactational 2,3,7,8-Tetrachlorodibenzo-p-dioxin Exposure Impairs Molar Tooth
Development in Rats
Qualitative Effects of Dioxin on Molars Vary Among Inbred Mouse Strains
Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin on Molar Development Among Non-
resistant Inbred Strains of Mice: A Geometric Morphometric Analysis
Genetic Differences in Sensitivity to Alterations of Mandible Structure Caused by the
Teratogen2,3,7,8-Tetrachlorodibenzo-/>-Dioxin
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) Effects on Hepatic Microsomal
Cytochrome P -44 8 -mediated Enzyme Activities
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD): Results of a 13-week Oral Toxicity Study
in Rats
Results of a Two-year Chronic Toxicity and Oncogenicity Study of
2,3,7,8-Tetrachlorodibenzo-/>-dioxin in Rats. Long-term Toxicologic Studies of 2,3,7,8-
Tetrachlorodibenzo-p-dioxin (TCDD) in Laboratory Animals
hi Utero and Lactational Exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin Decreases
Serotonin-immunoreactive Neurons in Raphe Nuclei of Male Mouse Offspring
Effects of Vitamin E on Reactive Oxygen Species-mediated 2,3,7,8-Tetrachlorodibenzo-
p-dioxin Toxicity in Rat Testis
Induction of Oxidation Stress in Rat Epidermal Sperm After Exposure to
2,3,7,8-Tetrachlorodibenzo-p-dioxin
The Effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin on the Antioxidant System in
Mitochondrial and Microsomal Fractions of Rat Testis
2,3,7,8-Tetrachlorodibenzo-p -dioxin (TCDD) Induces Oxidative Stress in the
Epididymis and Epididymal Sperm of Adult Rats
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) Increases Release of Luteinizing Hormone
and Follicle-Stimulating Hormone from the Pituitary of Immature Female Rats In Vivo
and In Vitro
The Early Embryo Loss Caused by 2,3,7,8-Tetrachlorodibenzo-p-dioxin May be Related
to the Accumulation of this Compound in the Uterus
mgestion of Soil Contaminated with 2,3,7,8-Tetrachloro-dibenzo-p-dioxin (TCDD)
Alters Hepatic Enzyme Activities in Rats
Non-genotoxic Carcinogens: Early Effects on Gap Junctions, Cell Proliferation and
Apoptosis in the Rat
Altered operant Responding for Motor Reinforcement and the Determination of
Benchmark Doses Following Perinatal Exposure to Low-level
2,3,7,8-Tetrachlorodibenzo-/>-dioxin
                                D-89

-------
Table D-l. Noncancer animal studies selected for TCDD dose-response
analyses (continued)
Author (year)
Maronpot et al.
(1993)
Miettinen et al. (2006)
Murray et al. (1979)
Nohara et al. (2000b)
Nohara et al. (2002a)
NTP (1982)
NTP (2006)
Ohsako et al. (2001)
Schantz and Bowman
(1989)
Schantz et al. (1986)
Schantz et al. (1992)
Schantz et al. (1996)
Seo et al. (1995)
Sewall et al. (1993)
Sewall et al. (1995 a)
Shi et al. (2007)
Simanainen et al.
(2002)
Simanainen et al.
(2003)
Simanainen et al.
(2004b)
Slezak et al. (2000)
Smialowicz et al.
(2004)
Title of study
Dose Response for TCDD Promotion of Hepatocarcinogenesis in Rats Initiated with
DEN: Histologic, Biochemical, and Cell Proliferation Endpoints
The Effect of Perinatal TCDD Exposure on Caries Susceptibility in Rats
Three-generation Reproduction Study of Rats Given 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
(TCDD) in the Diet
The Effects of Perinatal Exposure to Low Doses of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
on Immune Organs in Rats
Effect of Low-dose 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) on Influenza A Virus-
induced Mortality in Mice
NTP Technical Report on Carcinogenesis Bioassay of 2,3,7,8-tetrachlorodibenzo-p-
dioxin in Osborne-Mendel Rats and B6C3F] Mice (Gavage Study)
NTP Technical Report on the Toxicology and Carcinogenesis Studies of
2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD) in Female Harlan Sprague-Dawley Rats
(Gavage Studies)
Maternal Exposure to a Low Dose of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
Suppressed the Development of Reproductive Organs of Male Rats: Dose-Dependent
Increase of mRNA Levels of 5a-reductase Type 2 in Contrast to Decrease of Androgen
Receptor in the Pubertal Ventral Prostate
Learning in Monkeys exposed Perinatally to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD)
Maternal Care by Rhesus Monkeys of Infant Monkeys Exposed to Either Lead or 2,3,7,8-
Tetrachlorodibenzo-p-dioxin (TCDD)
Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin on Behavior of Monkeys in Peer Groups
Effects of Gestational and Lactational Exposure to TCDD or Coplanar PCBs on Spatial
Learning
Effects of Gestational and Lactational Exposure to Coplanar Polychlorinated Biphenyl
(PCB) Congeners or 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) on Thyroid Hormone
Concentrations in Weanling Rats
TCDD-mediated Changes in Hepatic Epidermal Growth Factor Receptor May be a
Critical Event in the Hepatocarcinogenic Action of TCDD
Alterations in Thyroid Function in Female Sprague-Dawley Rats Following Chronic
Treatment with 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Ovarian Endocrine Disruption Underlies Premature Reproductive Senescence Following
Environmentally Relevant Chronic Exposure to the Aryl Hydrocarbon Receptor Agonist
2,3,7,8-Tetrachlorodibenzo-p-Dioxin
Structure-Activity Relationships and Dose Responses of Polychlorinated Dibenzo-p-
dioxins for Short-Term Effects in 2,3,7,8- Tetrachlorodibenzo-p-dioxin-Resistant and -
Sensitive Rat
Dose-response Analysis of Short-term Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in
Three Differentially Susceptible Rat Lines
Pattern of Male Reproductive System Effects After hi Utero and Lactational
2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD) Exposure in Three Differentially TCDD-
Sensitive Rat Lines
Oxidative Stress in Female B6C3F] Mice Following Acute and Subchronic Exposure to
2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD)
CYP1A2 is Not Required for 2,3,7,8-Tetrachlorodibenzo-p-dioxin-induced
Immunosuppression
                                D-90

-------
Table D-l. Noncancer animal studies selected for TCDD dose-response
analyses (continued)
Author (year)
Smialowicz et al.
(2008)
Smith et al. (1976)
Sparschu et al. (1971)
Sugita-Konishi et al.
(2003)
Tritscher et al. (1992)
Toth et al. (1979)
Van Birgelen et al.
(1995 a)
Van Birgelen et al.
(1995b)
Vanden Heuvel et al.
(1994)
Vos et al. (1973)
Weber et al. (1995)
White et al. (1986)
Yang et al. (2000)
Zareba et al. (2002)
Title of study
Relative Potency Based on Hepatic Enzyme Induction Predicts Immunosuppressive
Effects of a Mixture of PCDDS/PCDFS and PCBS
Teratogenicity of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in CF-1 Mice
Study of the Teratogenicity of 2,3,7,8-Tetrachiorodibenzo-p-dioxin in the Rat
Effect of Lactational Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin on the
Susceptibility to Listeria Infection
Dose-response Relationships for Chronic Exposure to 2,3,7,8-Tetrachlorodibenzo-p-
dioxin in a Rat-tumor Promotion Model: Quantification and Immunolocalization of
CYP1 Aland CYP1A2 in the Liver
Carcinogenicity Testing of Herbicide 2,4,5-Trichlorophenoxyethanol Containing Dioxin
and of Pure Dioxin in Swiss Mice
Subchronic Dose-response Study of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in Female
Sprague-Dawley Rats
Subchronic Effects of 2,3,7,8-TCDD or PCBs on Thyroid Hormone Metabolism: Use in
Risk Assessment
Dioxin-responsive Genes: Examination of Dose-response relationships Using
Quantitative Reverse Transcriptase-polymerase Chain Reaction
Effect of 2,3,7,8-Tetrachlorodibenzo-p-dioxin on the Immune System of Laboratory
Animals
Correlation Between Toxicity and Effects on Intermediary Metabolism in
2,3,7,8-Tetrachlorodibenzo-p-dioxin-treated Male C57BL/6L and DBA/2J Mice
Modulation of Serum Complement Levels Following Exposure to Polychlorinated
Dibenzo-p-dioxins
Subchronic Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin Modulates the
Pathophysiology of Endometriosis in the Cynomolgus Monkey
Sexually Dimorphic Alterations of Brain Cortical Dominance in Rats Prenatally Exposed
to TCDD
                                D-91

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion
Author (year)
Abbott and Birnbaum
(1989)
Abbott and Birnbaum
(1990)
Abbott and Probst
(1995)
Abbott et al. (1987b)
Abbott et al. (1987a)
Abbott et al. (1999a)
Abbott et al. (1999b)
Abbott et al. (20031
Abernethy et al.
(1985)
Abraham et al.
(1988)
Ackermann et al.
(1989)
Adamsson et al.
(2008)
Title of study
rCDD Alters Medial Epithelial Cell Differentiation During
Palatogenesis
Effects of TCDD on Embryonic Ureteric Epithelial EGF
Receptor Expression and Cell Proliferation
Developmental Expression of Two Members of a New Class of
Transcription Factors: II. Expression of Aryl Hydrocarbon
Receptor Nuclear Translocator in the C57BL/6N Mouse Embryo
TCDD Alters the Extracellular Matrix and Basal Lamina of the
Fetal Mouse Kidney
TCDD-Induced Hyperplasia of the Ureteral Epithelium Produces
Hydronephrosis in Murine Fetuses
AhR, ARNT, and CYP1A1 mRNA Quantitation in Cultured
Human Embryonic Palates Exposed to TCDD and Comparison
with Mouse Palate In Vivo and in Culture
RT-PCR Quantification of AHR, ARNT, GR, and CYP1A1
mRNA in Cranio facial Tissues of Embryonic Mice Exposed to
2,3,7,8-Tetrachlorodibenzo-p-dioxin and Hydrocortisone
EGF and TGF-a Expression Influence the Developmental
Toxicity of TCDD: Dose Response and AhR Phenotype in EGF,
TGF-a , and EGF+ TGF- a Knockout Mice
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) Promotes the
Transformation of C3H/10T1/2 Cells
Pharmacokinetics and Biological Activity of
2,3,7,8-Tetrachlorodibenzo-/>-dioxin. 1. Dose-dependent Tissue
Distribution and Induction of Hepatic Ethoxyresorufin
9-deethylase in Rats Following a Single Injection
Selective Inhibition of Polymorphonuclear Activity by
2,3,7,8-Tetracholordibenzo-p-dioxin
The Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin on Fetal
Male Rat Steroidogenesis
Reason for excluding study
Genetically
altered animals
-
-

-
-



-

-
-
Low dose
too high
X
X

X
X
X
X
X
-

X
X
Doses not TCDD only;
unspecified TCDD dose
-
-
X
-
-



-

-
-
Nonoral
dose
-
-

-
-



X
X
-
-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Agrawal et al. (1981)
Aitio et al. (1979)
Albro et al. (1978)
Allen and Carstens
(1967)
Allen and Leamy
(2001)
Alsharif and Hassoun
(2004)
Alsharif etal. (1990)
Alsharif etal. (1994b)
Alsharif etal. (1994c)
Alsharif etal. (1994a)
Altmann et al. (1995)
Altmann et al. (1998)
Andersson et al. (2002)
Title of study
3,4,3N,4N-Tetrachlorobiphenyl Given to Mice Prenatally
Produces Long-term Decreases in Striatal Dopamine and
Receptor Binding Sites in the Caudate Nucleus
Different Effect of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin on
Glucuronide Conjugation of Various Aglycones: Studies in
Wistar and Gunn Rats
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin and Lipid
Profiles in Tissues of the Fischer Rat
Light and Electron Microscopic Observations in Macaco
mulatto Monkeys Fed Toxic Fat
2,3,7,8-Tetrachlorodibenzo-p-dioxin Affects Size and Shape,
but Not Asymmetry, of Mandibles in Mice
Protective Effects of Vitamin A and Vitamin E Succinate
Against 2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD)-induced
Body Wasting, Hepatomegaly, Thymic Atrophy, Production of
Reactive Oxygen Species and DNA Damage in C57BL/6J Mice
2,3,7,8-Tetrachlorodibenzo-p-dioxin(TCDD)-induced
Decrease in the Fluidity of Rat Liver Membranes
Oxidative Stress Induced by 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin is Mediated by the Aryl Hydrocarbon (Ah) Receptor
Complex
Stimulation of NADPH-dependent Reactive Oxygen Species
Formation and DNA Damage by 2,3,7,8-Tetrachlorodibenzo-p-
dioxin TCDD in Rat Peritoneal
The Effects of Ani-TNF-alpha Antibody and Dexamethasone
on 2,3,7,8-Tetrachlorodibenzo-/>-dioxin-induced Oxidative
Stress in Mice
Maternal Exposure to Polychlorinated Biphenyls Inhibits Long-
term Potentiation in the Visual Cortex of Adult Rats
Inhibition of Long-term Potentiation in Developing Rat Visual
Cortex but Not Hippocampus by In Utero Exposure to
Polychlorinated Biphenyls
A Constitutively Active Dioxin/Aryl Hydrocarbon Receptor
(AhR) Induces Stomach Tumors
Reason for excluding study
Genetically altered
animals


-
-
-

-



-

X
Low dose
too high

X
X
X
X
X
X
X
X
X
-

-
Doses not TCDD only;
unspecified TCDD dose
X

-
-
-

-



X
X
-
Nonoral dose


-
-
-

-



-

-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Aoa et al. (2009)
Aragon et al. (2008a)
Aragon et al. (2008b)
Ashida et al. (1996)
Ashida et al. (2000)
Astroffetal. (1987)
Aubert et al. (1985)
Aulerich et al. (2001)
Badawi et al. (2000)
Badesha et al. (1995)
Bagchi et al. (1993)
Title of study
Comparison of Immunotoxicity Among Tetrachloro-,
Pentachloro-, Tetrabromo- and Pentabromo-dibenzo-p-dioxins
in Mice
In Utero and Lactational 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Exposure: Effects on Fetal and Adult Cardiac Gene Expression
and Adult Cardiac and Renal Morphology
Perinatal 2,3,7,8-TCDD Exposure Sensitizes Offspring to
Angiotensin II -induced Hypertension
Protective Action of Dehydroascorbic Acid on the Ah
Receptor-dependent and Receptor-independent Induction of
Lipid Peroxidation in Adipose Tissue of Male Guinea Pig
Caused by TCDD Administration
2,3,7,8-TCDD-induced Changes in Activities of Nuclear
Protein Kinases and Phosphatases Affecting DNA Binding
Activity of c-Myc and AP-1 in the Livers of Guinea Pigs
6-Methyl-l,3,8-Trichlorodibenzofuran as a 2,3,7,8-TCDD
Antagonist: Inhibition of the Induction of Rat Cytochrome
P-450 Isozymes and Related Monooxygenase Activities
Ontogeny of Hypothalamic Luteinizing Hormone-releasing
Hormone (GnRH) and Pituitary GnRH Receptors in Fetal and
Neonatal Rats
Short Communications: Dietary Exposure to
3,3',4,4',5 -Pentachlorobiphenyl (PCB 126) or 2,3,7,8-TCDD
Does Not Induce Proliferation of Squamous Epithelium or
Osteolysis in Jaws of Weanling Rats
Effect of Chlorinated Hydrocarbons on Expression of
Cytochrome P450 1A1, 1A2 and 1B1 and 2- and
4-Hydroxylation of 17p-estradiol in Female Sprague-Dawley
Rats
[mmunotoxic Effects of Prolonged Dietary Exposure of Male
Rats to 2,3,7,8-tetrachlorodibenzo-p-dioxin
Time -dependent Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
on Serum and Urine Levels of Malondialdehyde,
Formaldehyde, Acetaldehyde, and Acetone in Rats
Reason for excluding study
Genetically altered
animals


-






-

Low dose
too high

X
X

X


X
X
X
X
Doses not TCDD only;
unspecified TCDD dose
X

-






-

Nonoral dose


-
X
X
X
X


-


-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Bagchi et al. (2002)
Bars and Elcombe
(19911
Barsotti et al. (1979)
Barter and Klaassen
(1992)
Bastomsky (1977)
Beckett et al. (2005)
Beebe et al. (1995)
Beguinot et al. (1985)
Bell et al. (2007b)
Bell et al. QOOTa)
Bemis et al. (2007)
Besteman et al. (2005)
Title of study
Comparative Effects of TCDD, Endrin, Naphthalene and
Chromium (VI) on Oxidative Stress and Tissue Damage in the
Liver and Brain Tissues of Mice
Dose-dependent Acinar Induction of Cytochromes P450 in Rat
Liver. Evidence for a Differential Mechanism of Induction of
P4501A1 by Beta-naphthaflavone and Dioxin
Hormonal Alterations in Female Rhesus Monkeys Fed a Diet
Containing 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
UDP-glucuronosyltransferase Inducers Reduce Thyroid
Hormone Levels in Rats by an Extrathyroidal Mechanism
Enhanced Thyroxine Metabolism and High Uptake Goiters in
Rats After a Single Dose of 2,3,7,8-Tetrachlorodibenzo-p-
dioxin
Squamous Epithelial Lesion of the Mandibles and Maxillaw of
Wild Mink Naturally Exposed to Polychlorinated Biphenyls
Promotion of N-nitrosodimethylamine-initiated Mouse Lung
Tumors Following Single or Multiple Low Dose Exposure to
2,3,7,8- Tetrachlorodibenzo-p-dioxin
Phorbol Esters Induce Internalization Without Degradation of
Unoccupied Epidermal Growth Factor Receptors
Toxicity of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin in the
Developing Male Wistar(Han) Rat. I: No Decrease in
Epididymal Sperm Count after a Single Acute Dose
Relationships Between Tissue Levels of
2,3,7,8-Tetrachlorodibenzop-dioxin(TCDD), mRNAs, and
Toxicity in the Developing Male Wistar(Han) Rat
TCDD-Induced Alterations in Gene Expression Profiles of the
Developing Mouse Paw Do Not Influence Morphological
Differentiation of This Potential Target Tissue
retrachlorodibenzo-p-Dioxin (TCDD) Inhibits Differentiation
and Increases Apoptotic Cell Death of Precursor T-Cells in the
Fetal Mouse Thymus
Reason for excluding study
Genetically altered
animals


-
-

X

-




Low dose
too high
X

X
-
X
-

-
X
X

X
Doses not TCDD only;
unspecified TCDD dose


-
X

-

X




Nonoral dose

X
-
-

X
X
-


X


-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Besteman et al. (2007)
Biegel et al. (1989)
Birnbaum et al. (1985)
Birnbaum et al. (1986)
Birnbaum et al. (1987a)
Birnbaum et al. (1987b)
Birnbaum et al. (1989)
Birnbaum et al. (1990)
Birnbaum et al. (1991)
Bjerke and Peterson
(1994)
Bjerke et al. (1994a)
Bjerke et al. (1994b)
Title of study
2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) or
Diethylstilbestrol (DBS) Cause Similar Hematopoietic
Hypocellularity and Hepatocellular Changes in Murine Fetal
Liver, but Differentially Affect Gene Expression
2,2N4,4N5,5N-Hexachlorobiphenyl as a
2,3,7,8-Tetrachlorodibenzo-/>-dioxin Antagonist in C57BL/6
Mice
Toxic Interaction of Specific Fob/chlorinated Biphenyls and
2,3,7,8-Tetrachlorodibenzo-/>-dioxin: Increased Incidence of
Cleft Palate in Mice
Synergistic Interaction of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
and Hydrocortisone in the Induction of Cleft Palate in Mice
Teratogenic Effects of Fob/chlorinated Dibenzofurans in
Combination in C57BL/6N Mice
Teratogenicity of Three Polychlorinated Dibenzofurans in
C57BL/6N Mice
Retinoic Acid and 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
(TCDD) Selectively Enhance Teratogenesis in C57BL/6N Mice
Differential toxicity of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD) in C57B1/6 mice congenic at the Ah locus
Teratogenic Effects of 2,3,7,8-Tetrabromodibenzo-p-dioxin and
Three Polybrominated Dibenzofurans in C57BL/6N Mice
Reproductive Toxicity of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
in Male Rats: Different Effects of In Utero Versus Lactational
Exposure
Effects of In Utero and Lactational 2,3,7,8-Tetrachlorodibenzo-
p-dioxin Exposure on Responsiveness of the Male Rat
Reproductive System to Testosterone Stimulation in Adulthood
Partial Demasculinization and Feminization of Sex Behavior in
Male Rats by In Utero and Lactational Exposure to
2,3,7,8-Tetrachlorodibenzo-p-dioxin is Not Associated with
Alterations in Estrogen Receptor Binding or Volumes of
Sexually Differentiated Brain
Reason for excluding study
Genetically altered
animals



-
-
-
-
-
-



Low dose
too high
X
X

X
-
-
X
X
-
X
X
X
Doses not TCDD only;
unspecified TCDD dose


X
-
X
X
-
-
X



Nonoral dose



-
-
-
-
-
-




-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Blaylock et al. (1992)
Bohn et al. (2005)
Boverhofetal. (2005)
Boverhofetal. (2008)
Bowers et al. (2006)
Brewster et al. (1987)
Brewster and
Matsumura (1984)
Brouillette and Quirion
(2008)
Brouwer and van den
Berg (1983)
Brouwer and van den
Berg (1984)
Brouwer et al. (1985)
Brown and
Lamartiniere (1995)
Brunnberg et al. (2006)
Bryant et al. (1997)
Title of study
Exposure to Tetrachlorodibenzo-p-dioxin (TCDD) Alters Fetal
Thymocyte Maturation
Increased Mortality Associated with TCDD Exposure in Mice
Infected with Influenza A Virus is Not Due to Severity of Lung
Injury or Alterations in Clara Cell Protein Content
Temporal and Dose-Dependent Hepatic Gene Expression
Patterns in Mice Provide New Insights into TCDD-Mediated
Hepatotoxicity
Inhibition of Estrogen-Mediated Uterine Gene Expression
Responses by Dioxin
Short Report: 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
Reduces Leishmania Major Burdens In C57B1/6 Mice
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxinonthe Guinea
Pig Heart Muscle
TCDD (2,3,7,8-Tetrachlorodibenzo-/>-dioxin) Reduces
Lipoprotein Lipase Activity in the Adipose Tissue of the
Guinea Pig
The Common Environmental Pollutant Dioxin-induced
Memory Deficits by Altering Estrogen Pathways and a Major
Route of Retinol Transport Involving Transthyretin
Early Decrease in Retinoid Levels in Mice After Exposure to
Low Doses of Polychlorinated Biphenyls
Early and Differential Decrease in Natural Retinoid Levels in
C57Bl/Rij and DBA/2 Mice by 3,4,3N,4N-Tetrachlorobipheny
Time and Dose Responses of the Reduction in Retinoid
Concentrations in C57BL/Rij and DBA/2 Mice Induced by
3,4,3N,4N-Tetrachlorobiphenyl
Xenoestrogens Alter Mammary Gland Differentiation and Cell
Proliferation in the Rat
The Constitutively Active Ah Receptor (CA-AhR) Mouse as a
Potential Model for Dioxin Exposure — Effects in Vital Organs
Effects of TCDD on Ah Receptor, ARNT, EOF, and TGF-
alpha Expression in Embryonic Mouse Urinary Tract
Reason for excluding study
Genetically altered
animals
-

X
-
-
-


-
-

-
-
-
Low dose
too high
X
X

X
X
-

X
-
-

X
X
X
Doses not TCDD only;
unspecified TCDD dose
-


-
-
-


X
X
X
-
-
-
Nonoral dose
-


-
-
X
X
X
-
-

-
-
-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Bryant et al. (2001)
Buchmann et al. (1994)
Bushnell and Rice
(19991
Byers et al. (2006)
Calfee-Mason et al.
(2002)
Camacho et al. (2004)
Cantoni et al. (1984)
Carney et al. (2004)
Chaffm et al. (1996)
Chaffm et al. (1997)
Chahoud et al. (1989)
Chapman and Schiller
(1985)
Title of study
Teratogenicity of 2,3,7,8-Tetrachlorodibenzo-/>-Dioxin
(TCDD) in Mice Lacking the Expression of EOF and/or TGF-
alpha
Effects of 2,3,7,8-Tetrachloro- and
1,2,3,4,6,7,8-Heptachlorodibenzo-p- dioxin on the Proliferation
of Preneoplastic Liver Cells in the Rat
Behavioral Assessments of Learning and Attention in Rats
Exposed Perinatally to 3,3',4,4',5-Pentachlorobiphenyl (PCB
126)
Association Between the Levels of Biogenic Amines and
Superoxide Anion Production in Brain Regions of Rats After
Subchronic Exposure to TCDD
Vitamin E Inhibits Hepatic NF-kB Activation in Rats
Administered the Hepatic Tumor Promoter Phenobarbital
Effect of 2,3,7, 8-TCDD on Maternal Immune Response During
Pregnancy
Different Susceptibility of Mouse Tissues to Porphyrogenic
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
2,3,7,8-TCDD Activation of the AHR/AHR Nuclear
rranslocator Pathway Causes Developmental Toxicity Through
a CYPl-A-independent Mechanism in Zebrafish
In Utero and Lactational Exposure of Female Holtzman Rats to
2,3,7,8-Tetrachlorodibenzo-/>-dioxin: Modulation of the
Estrogen Signal
Alterations to the Pituitary -gonadal Axis in the Female Rat
Exposed In Utero and Through Lactation to
2,3,7,8-Tetrachlorodibenzo-^-dioxin
Reproductive Toxicity and Pharmacokinetics of
2,3,7,8-Tetrachlorodibenzo-/>-dioxin. I. Effects of High Doses
on the Fertility of Male Rats
Dose-related Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD) in C57BL/6J and DBA/2J Mice
Reason for excluding study
Genetically altered
animals
X



-
-
-
X



-
Low dose
too high
X


X
-
X
X

X
X

X
Doses not TCDD only;
unspecified TCDD dose

X
X

X
-
-




-
Nonoral dose




-
-
-



X
-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Chen etal. (1993)
Chen et al. (2001)
Chen etal. (2002)
Chen et al. (2003)
Cheng et al. (2002)
Cho et al. (20061
Choi et al. (2006)
Choi et al. (2008)
Chou et al. (1979)
Clark et al. (19811
Clark et al. (1991a)
Clark et al. (1991b)
Cohen et al. (1979)
Title of study
InUtero Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD) Does Not Impair Testosterone Production by Fetal Rat
Testis
Disposition of Polychlorinated Dibenzo-p-dioxins,
Dibenzofurans, and Non-ortho Plychlorinated Biphenyls in
Pregnant Long Evans Rats and the Transfer to Offspring
A Mixture of Polychlorinated Dibenzo-p-dioxins (PCDDs),
Dibenzofurans (PCDFs), and Non-ortho Polychlorinated
Biphenyls (PCBs) Changed the Lipid Content of Pregnant Long
Evans rats
The Effect of 2,3,7,8-TCDD on Chorionic Gonadotrophin
Activity in Pregnant Macaques
2,3,7,8-TCDD Treatment Induces c-Fos Expression in the
Forebrain of the Long-Evans Rat
Enhanced Expression of Plasma Glutathione Peroxidase in the
Thymus of Mice Treated with TCDD and its Implication for
TCDD-induced Thymic Atrophy
In Utero Exposure to 2,3,7,8-TCDD Induces Amphiregulin
Gene Expression in the Developing Mouse Ureter
Effect of 2,3,7,8-TCDD on Testicular Spermatogenesis-related
Panels and Serum Sex Hormone Levels in Rats
Neuropathology of "Spinning Syndrome" Induced by Prenatal
Intoxication with a PCB in Mice
Enhanced Suppressor Cell Activity as a Mechanism of
Immunosuppression by 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Tumor necrosis Factor involvement in
2,3,7,8-Tetrachlorodibenzo-p-dioxin-mediatedEndotoxin
Hypersensitivity in C57B1/6 Mice Congenic at the Ah Locus
Tumor Promotion by TCDD in Female Rats. In: Biological
Basis for Risk Assessment of Dioxins and Related Compounds
Anticarcinogenic Effects of 2,3,7,8-Tetrachlorodibenzo-p-
dioxinonBenzo[a]pyrene and 7,12-Dimethylbenz[a]anthrene
Tumor Initiation and its Relationship to DNA Binding
Reason for excluding study
Genetically altered
animals



-
-

-
-
-
-

-

Low dose
too high
X


X
X
X
-
X
-
-
X
X

Doses not TCDD only;
unspecified TCDD dose
X
X
X
-
-

-
-
X
-

-

Nonoral dose



-
-

X
-
-
X

X
X

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Collins and Capen
(19801
Collins et al. (2008)
Comer and Norton
(1982)
Courtney (1976)
Courtney and Moore
(1971)
Couture et al. (1989)
Couture et al. (1990)
Crofton and Rice
(1999)
Cummings et al. (1996)
Dalton et al. (2001)
D'Argy et al. (19841
Davies et al. (2008)
Davis et al. (2000)
de Heer et al. (1995)
Title of study
Fine Structural Lesions and Hormonal Alterations in Thyroid
Glands of Perinatal Rats Exposed In Utero and by the Milk to
Polychlorinated Biphenyls
2,3,7,8-Tetracholorodibenzo-p-Dioxin Exposure Disrupts
Granule Neuron Precursor Maturation in the Developing
Mouse Cerebellum
Effects of Perinatal Methimazole Exposure on a Developmental
Test Battery for Neurobehavioral Toxicity in Rats
Mouse Teratology Studies with Chlorodibenzo-p-dioxins
Teratology Studies with 2,4,5-Trichlorophenoxyacetic Acid
and2,3,7,8-Tetrachlorodibenzo-£>-dioxin
Developmental Toxicity of 2,3,4,7,8-Pentachlorodibenzofuran
in the Fischer 344 Rat
Characterization of the Peak Period of Sensitivity for the
Induction of Hydronephrosis in C57BL/6N Mice Following
Exposure to 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
Low-frequency Hearing Loss Following Perinatal Exposure to
3,3',4,4',5-Pentachlorobiphenyl(PCB 126) in Rats
Promotion of Endometriosis by 2,3,7,8- Tetrachlorodibenzo-/>-
dioxin in Rats and Mice: Time-Dose Dependence and Species
Comparison
Dioxin Exposure Is an Environmental Risk Factor for Ischamic
Heart Disease-IP injection
Teratogenicity of TCDD and Congener
3,3N,4,4N-Tetrachloroazoxybenzene in Sensitive and
Nonsensitive Mouse stRains After Reciprocal Blastocyst
Transfer
Essential Role of the AH Receptor in the Dysfunction of Heme
Metabolism Induced by 2,3,7,8-Tetrachlorodibenzo-^-dioxin
Ovarian Tumors in Rats Induced by Chronic
2,3,7,8-Tetrachlorodibenzo-p-dioxin Treatment
Toxicity of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD) to the
Human Thymus after Implantation in SCID Mice
Reason for excluding study
Genetically altered
animals


-
-
-
-

-

-

-
-
-
Low dose
too high

X
-
X
X
-
X
-
X
-
X
X
X
X
Doses not TCDD only;
unspecified TCDD dose
X

X
-
-
X

X

-

-
-
-
Nonoral dose


-
-
X
-

-

X

-
-
-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Dearstyne and
Kerkvliet (2002)
Devito et al. (1992)
Dhar and Setty (1990)
Dienhart et al. (2000)
Diliberto et al. (1999)
Dong et al. (2002)
Dong et al. (2004)
Dragan et al. (1991)
Dragan et al. (1992)
Dragin et al. (2006)
Dunlap and Matsumura
(2000)
Dunlap et al. (1999)
Title of study
Mechanism of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)-
induced Decrease in Anti-CD 3 -activated CD4+ T cells: the
Roles of Apoptosis, Fas, and TNF
Antiestrogenic Action of 2,3,7,8-Tetrachloro- dibenzo-p-
dioxin: Tissue Specific Regulation of Estrogen Receptor in
CD1 Mice
Changes in Testis, Epididymis and Other Accessory Organs of
Male Rats Treated with Anandron During Sexual Maturation
Gestational Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Induces Developmental Defects in the Rat Vagina
Effects of CYP1A2 on Disposition of
2,3,7,8-Tetrachlorodibenzo-/>-dioxin,
2,3,4,7,8-Pentachlorodibenzofuran, and
2,2',4,4',5,5'-Hexachlorobiphenyl in CYP1A2 Knockout and
Parental (C57BL/6N and 129/Sv) Strains of Mice
2,3,7, 8-Tetrachlorodibenzo-p-dioxin in the Zebra Fish Embryo:
Local Circulation Failure in the Dorsal Midbrain is Associated
with Increased Apoptosis
Role of Aryl Hydrocarbon Receptor in Mesencephalic
Circulation Failure and Apoptosis in Zebrafish Embryos
Exposed to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
An initiation-promotion assay in rat liver as a potential
complement to the 2-year carcinogenesis bioassay
Characterization of the Promotion of Altered Hepatic Foci by
2,3,7,8-Tetrachlorodibenzo-/>-dioxin in the Female Rat
For Dioxin-induced Birth Defects, Mouse or Human CYP1 A2
in Maternal Liver Protects whereas Mouse CYP1 Al and
CYP1B1 Are Inconsequential
Analysis of Difference In Vivo Effects of TCDD Between c-src
+Y+ mice, c-src Deficient, -/+ and -/- B6, 129-Srctm 1 sor Mice
and their Wild-type Littermates-IP Injection
Differential Toxicities of TCDD In Vivo Among Normal, c-src
Knockout, Geldanamycin-, and Quercetin-treated Mice
Reason for excluding study
Genetically altered
animals


-
-

X
X
-
-
X
X
X
Low dose
too high
X

-
X
X


-
-
X

-
Doses not TCDD only;
unspecified TCDD dose


X
-



X
-


-
Nonoral dose

X
-
-



-
X


X

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Dunlap et al. (2002)
Ebner et al. (1988)
Eckle et al. (2004)
Elder et al. (1976)
El-Sabeawy et al.
(1998)
El-Tawil and Elsaieed
(2005)
Enan et al. (1992)
Enan et al. (1998)
Eriksson et al. (1991)
Esser et al. (2005)
Evans and Andersen
(2000)
Title of study
Effects of Src -deficiency on the Expression of In Vivo Toxicity
of TCDD in a Strain of c-src Knockout Mice Procured Through
Six Generations of Backcrossings to C57BL/6 Mice-IP
Injection
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin on Serum
Insulin and Glucose Levels in the Rat
Immunohistochemical Detection of Activated Caspases in
Apoptotic Hepatocytes in Rat Liver
The Effect of Porphyrogenic Compound, Hexachlorobenzene,
on the Activity of Hepatic Uroporphyrinogen Decarboxylase in
the Rat
Treatment of Rats during Pubertal Development with
2,3,7,8-Tetrachlorodibenzo-p-dioxin Alters Both Signaling
Kinase Activities and Epidermal Growth Factor Receptor
Binding in the Testis and the Motility and Acrosomal Reaction
of Sperm-IP injection
Induction of Oxidative Stress in the Reproductive System of
Rats after Subchronic Exposure to
2,3,7,8-Tetrachlorodibenzo-p-dioxin
TCDD Causes Reduction in Glucose Uptake Through Glucose
Transporters on the Plasma Membranes of the Guinea Pig
Adipocyte
Mechanism of Gender-Specific TCDD-induced Toxicity in
Guinea Pig Adipose Tissue
Neonatal Exposure to 3,3N,4,4N-Tetrachlorobiphenyl: Changes
in Spontaneous Behavior and Cholinergic Muscarinic
Receptors in the Adult Mouse
Effects of a Single Dose of 2,3,7,8-Tetrachlorodibenzo-p-
dioxin, Given at Post-puberty, in Senescent Mice
Sensitivity Analysis of a Physiological Model for
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD): Assessing the
Impact of Specific Model Parameters on Sequestration in Liver
and Fat in the Rat
Reason for excluding study
Genetically altered
animals
X
-
X




-

-
X
Low dose
too high

-
-


X

X

-

Doses not TCDD only;
unspecified TCDD dose

-
-
X



-
X
-

Nonoral dose
X
X
-

X

X
X

X


-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Faith and Moore (1977)
Fan and Rozman
(19941
Fan et al. (1996)
Faqi et al. (1998)
Fernandez-Salguero et
al. (1995)
Fernandez-Salguero et
al. (1996)
Fetissov et al. (2004)
Fine et al. (1989)
Fine et al. (1990)
Fisher et al. (2005)
Flaws et al. (19971
Fletcher et al. (2001)
Fletcher et al. (2005a)
Title of study
Impairment of Thymus-dependent Immune Function by
Exposure of the Developing Immune System to
2,3,7,8-Tetrachlorodibenzo-p-dioxin(TCDD)
Relationship Between Acute Toxicity of
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) and Distribution
of Intermediary Metabolism in the Long-Evans Rat
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin on Humoral and
Cellmediated Immunity in Sprague-Dawley Rats
Reproductive Toxicity and Tissue Concentrations of Low
Doses of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in Male
Offspring Rats Exposed Throughout Pregnancy and Lactation
Immune System Impairment and Hepatic Fibrosis in Mice
Lacking the Dioxinbinding Ah Receptor
Aryl-hydrocarbon Receptor-Deficient Mice Are Resistant to
2,3,7,8-Tetrachlorodibenzo-p-dioxin-Induced Toxicity
Expression of Hypothalamic Neuropeptides After Acute TCDD
Treatment and Distribution of Ah Receptor Represser
Lymphocyte Stem Cell Alterations Following Perinatal
Exposure to 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
Prothymocyte Activity is Reduced by Perinatal
2,3,7,8-Tetrachlorodibenzo-p-dioxin Exposure
Aryl Hydrocarbon Receptor-dependent Induction of Loss of
Mitochondrial Membrane Potential in Epididydimal
Spermatozoa by 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
In Utero and Lactational Exposure to
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) Induces Genital
Dysmorphogenesis in the Female Rat
Hepatic Vitamin A Depletion is a Sensitive Marker of
2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD) Exposure in Four
Rodent Species
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) Alters the mRNA
Expression of Critical Genes Associated with Cholesterol
Metabolism, Bile Acid Biosynthesis, and Bile Transport in Rat
Liver: A Microarray Study
Reason for excluding study
Genetically altered
animals


-

X
-
-
-
-




Low dose
too high
X
X
X

-
-
X
X
X

X
X
X
Doses not TCDD only;
unspecified TCDD dose


-

-
-
-
-
-




Nonoral dose


-
X
-
X
-
X
X
X




-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Fletcher et al. (2005b)
Flodstrom and Ahlborg
(19921
Foster et al. (1997)
Frericks et al. (2006)
Fritz et al. (2005)
Fujimaki et al. (2002)
Fujiwara et al. (2008)
Funatake et al. (2005)
Funseth et al. (2002a)
Funseth et al. (2002b)
Galijatovic et al. (2004)
Gallo et al. (19861
Title of study
Altered Retinoid Metabolism in Female Long-Evans and
Han/Wistar Rats following Long-Term
2,3,7,8-Tetrachlorodibenzo-p-Dioxin(TCDD)-Treatment-
Subcutaneous administration
Relative Tumor Promoting Activity of Some Fob/chlorinated
Dibenzo-p-dioxin-, Dibenzofuran-, and Biphenyl Congeners in
Female Rats
Morphologic Characteristics of Endometriosis in the Mouse
Model: Application to Toxicology
Transcriptiona! Signatures of Immune Cells in Aryl
Hydrocarbon Receptor (AHR) -proficient and AHR-deficient
Mice
InUtero and Lactational 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Exposure: Effects on the Prostate and Its Response to
Castration in Senescent C57BL/6J Mice
Effect of a Single Oral Dose of 2,3,7,8-Tetrachlorodibenzo-p-
dioxin on Immune Function in Male NC/Nga Mice
Morphological and Immunohistochemical Studies on Cleft
Palates Induced by 2,3,7,8-Tetrachlorodibenzo-p-dioxin in
Mice
Cutting Edge : Activation of the Aryl Hydrocarbon Receptor by
2,3,7,8-Tetrachlorodibenzo-p-dioxin Generates a Population of
CD4+ CD25+ Cells with Characteristics of Regulatory T Cells
Effect of 2,3,7,8-Tetrachlorodibenzo-p-dioxin on Trace
Elements, Inflammation and Viral Clearance in the
Myocardium During Coxsackievirus B3 Infection in Mice
Effects of Coxsackievirus B3 Infection on the Acute-phase
Protein Metallothionein and on Cytochrome P-4501A1
Involved in the Detoxification Processes of TCDD in the
Mouse
The Human CYP1A1 Gene Is Regulated in a Developmental
and Tissue-specific Fashion in Transgenic Mice
Interactive Effects of Estradiol and 2,3,7,8-Tetrachlorodibenzo-
p-dioxin on Hepatic Cytochrome P-450 and Mouse Uterus
Reason for excluding study
Genetically altered
animals


-
X

-

X


-
-
Low dose
too high


-
X
X
X
X
X


-
X
Doses not TCDD only;
unspecified TCDD dose


-


-




-
-
Nonoral dose
X
X
X
X

-


X
X
X
-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Gao et al. (2000)
Gao et al. (2001)
Gao et al. (2004)
Garrett and Gasiewicz
(2006)
Gasiewicz and Rucci
(1984)
Gasiewicz et al. (1983)
Gasiewicz et al. (1986)
Gehrs and Smialowicz
(1999)
Gehrs et al. (19971
Center et al. (2006)
Title of study
Gonadotropin-releasing Hormone (GNRH) Partially Reverses
the Inhibitory Effect of 2,3,7,8-Tetrachlorodibenzo-p-dioxin on
Ovulation in the Immature Gonadotropin-treated Rat
2,3,7,8-Tetrachlorodibenzo-p-dioxin Decreases Responsiveness
of the Hypothalamus to Estradiol as a Feedback Inducer of
Preovulatory Gonadotropin Secretion in the Immature
Gonadotropin-Primed Rat
Lactational Exposure of Han/Wistar Rats to
2,3,7,8-Tetrachlorodibenzo-/>-dioxin Interferes with Enamel
Maturation and Retards Dentin Mineralization
The Aryl Hydrocarbon Receptor Agonist
2,3,7,8-Tetrachlorodibenzo-/>-dioxin Alters the Orcadian
Rhythms, Quiescence, and Expression of Clock Genes in
Murine Hematopoietic Stem and Progenitor Cells
Cytosolic Receptor for 2,3,7,8-Tetrachlorodibenzo-p-dioxin.
Evidence for a Homologous Nature Among Various
Mammalian Species
Distribution, Excretion, and Metabolism of
2,3,7,8-Tetrachlorodibenzo-p-dioxin in C57BL/6J, DBA/2J and
B6D2Fl/JMice
Changes in Hamster Hepatic Cytochrome P-450,
Ethoxycoumarin o-deethylase, and Reduced NAD(P):
Menadione Oxidoreductase Following Treatment with
2,3,7,8-Tetrachlorodibenzo-/>-dioxin. Partial Dissociation of
Temporal and Dose-response Relationships From Elicited
Toxicity
Persistent Suppression of Delayed-type Hypersensitivity in
Adult F344 Rats after Perinatal Exposure to
2,3,7,8-Tetrachlorodibenzo-p-dioxin
Alterations in the Developing; Immune System of the F344 Rat
After Perinatal Exposure to 2,3,7,8-Tetrachlorodibenzo-p-
dioxin. II. Effects on the Pup and the Adult
Comparison of Mouse Hepatic Mitochondria! Versus
Microsomal Cytochromes P450 Following TCDD Treatment
Reason for excluding study
Genetically altered
animals









-
Low dose
too high
X
X
X
X



X
X
-
Doses not TCDD only;
unspecified TCDD dose









-
Nonoral dose




X
X
X


X

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Geusau et al. (2005)
Ghafoorunissa (1980)
Giavini et al. (1982)
Giavini et al. (1983)
Goldey and Crofton
(19981
Goldstein and Linko
(1984)
Goldstein et al. (1973)
Goldstein et al. (1982)
Gonzalez etal. (1995)
Gordon and Miller
(1998)
Gordon et al. (1995)
Gordon et al. (1996)
Gorski and Rozman
(19871
Gorski et al. (1990)
Title of study
2,3,7,8-Tetrachlorodibenzo-p-dioxin Impairs Differentiation of
Normal Human Epidermal Keratinocytes in a Skin Equivalent
Model
Undernutrition and Fertility of Male Rats
Rabbit Teratology Studies With 2,3,7,8-Tetrachlorodibenzo-p-
dioxin
Embryotoxic Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Administered to Female Rats Before Mating
rhyroxine Replacement Attenuates Hypothyroxinemia,
Hearing Loss, and Motor Deficits Following Developmental
Exposure to Aroclor 1254 in Rats
Differential Induction of Two 2,3,7,8-Tetrachlorodibenzo-p-
dioxin-inducible Forms of Cytochrome P-450 in Extrahepatic
Versus Hepatic Tissues
Hepatic Porphyria Induced by 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin in the Mouse
Induction of Porphyria in the Rat by Chronic Versus Acute
Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Xenobiotic Receptor Knockout Mice
rhermoregulation in Rats Exposed Perinatally to Dioxin: Core
Temperature Stability to Altered Ambient Temperature,
Behavioral Thermoregulation, and Febrile Response to
Lipopolysaccharide
Temperature Regulation and Metabolism in Rats Exposed
Perinatally to Dioxin: Permanent Change in Regulated Body
Temperature
Autonomic and Behavioral Thermoregulation in Golden
Hamsters Exposed Perinatally to Dioxin
Dose-response and Time Course of Hypothyroxemia and
Hypoinsulinemia and Characterization of Insulin
Hypersensitivity in 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
(TCDD)-treated Rats
Reduced Gluconeogenesis in2,3,7,8-Tetrachlorodibenzo-p-
dioxin (TCDD) -treated Rats
Reason for excluding study
Genetically altered
animals
X
-
-
-


-
-
X


-

-
Low dose
too high

-
X
X


X
X
-
X
X
X

X
Doses not TCDD only;
unspecified TCDD dose

X
-
-
X

-
-
-


-

-
Nonoral dose

-
-
-

X
-
-
-


-
X
-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Gray et al. (1995b)
Gray et al. (1995a)
Gray et al. (1991 a)
Gray et al. (1997b)
Gray et al. (1997b)
Greenlee et al. (1985)
Greig and DeMatteis
(1973)
Guo et al. (2000)
Guo et al. (2007)
Guo et al. (2008)
Haag-Gronlund et al.
(1997)
Haake et al. (1987)
Title of study
Exposure to TCDD During Development Permanently Alters
Reproductive Function in Male Long Evans Rats and Hamsters:
Reduced Ejaculated and Epididymal Sperm Numbers and Sex
Accessory Gland Weights in Offspring With Normal
Androgenic Status
Functional Developmental Toxicity of Low Doses of
2,3,7,8-Tetrachlorodibenzo-/>-dioxin and a Dioxin-like PCB
(169) in Long Evans Rats and Syrian Hamsters: Reproductive,
Behavioral and Thermoregulatory Alterations
A Dose-response Analysis of the Reproductive Effects of
Single Gestational Dose of 2,3,7,8-Tetrachlorodibenzo-p-
dioxin in Male Long Evans Hooded Rat Offspring
InUtero 2,3,7,8-Tetrachlorodibenzo-p-dioxin(TCDD) Alters
Reproductive Morphology and Function in Female Rat
Offspring
InUtero Exposure to Low Doses of 2,3,7,8-
retrachlorodibenzo-p-dioxin Alters Reproductive Development
of Female Long Evans Hooded Rat Offspring
Evidence for Direct Action of 2,3,7,8-Tetrachlorodibenzo-p-
dioxin (TCDD) on Thymic Epithelium
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin on Drug
Metabolism and Hepatic Microsomes of Rats and Mice
Effect of TCDD on Maternal Toxicity and Chorionic
Gonadotropin: Bioactivity in the Immediate Post-implantation
Period of Macaque
Toxic Effects of TCDD on Osteogenesis Through Altering
1GFBP-6 gene Expression in Osteoblasts
Anti-estrogenic Effect of Dioxin on Rat Skeleton Development
Promotion of Altered Hepatic Foci by
2,3',4,4',5-Pentachlorobiphenyl in Sprague-Dawley Female
Rats
Aroclor 1254 as an Antagonist of the Teratogenicity of
2,3,7,8-Tetrachlorodibenzo-p-dioxin
Reason for excluding study
Genetically altered
animals





X
-

-
-

-
Low dose
too high
X
X
X
X
X
-
X
X
X
X

X
Doses not TCDD only;
unspecified TCDD dose





-
-

-
-

-
Nonoral dose





-
-

X
-
X
-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Haavisto et al. (2001)
Haavisto et al. (2006)
Hahn et al. (1988)
Hakansson and
Hanberg (1989)
Hakansson et al.
(1989a)
Hakansson et al.
(1989b)
Hakansson et al. (1990)
Hakansson et al. (1991)
Hakansson et al. (1994)
Hamm et al. (2000)
Hamm et al. (2003)
Hanson and
Smialowicz (1994)
Title of study
Prenatal Testosterone and Luteinizing Hormone Levels in Male
Rats Exposed During Pregnancy to 2,3,7,8-TCDD and
Diethylstilbestrol
The Effects of Maternal Exposure to 2,3,7,8-TCDD on
Testicular Steroidogenesis in Infantile Male Rats
The Role of the Ah Locus in Hexachlorobenzene-induced
Porphyria: Studies in the Congenic C57BL/6J Mice
The Distribution of [14C]-2,3,7,8-Tetrachlorodibenzo-/>-dioxin
(TCDD) and its Effect on Vitamin A Content in Parenchymal
and Stellate Cells of Rat Liver
2,3,7,8-Tetrachloro-dibenzo-p-dioxin(TCDD)-induced
Alterations in the Vitamin A Homeostasis and in the
7-Ethoxyresorufin o-deethylase (EROD)-activity in SD Rats
and Hartley Guinea Pigs
Hepatic Vitamin A Storage in Relation to Paired Feed
Restriction and TCDD-treatment
Vitamin A Storage in Rats Subchronically Exposed to
PCDDs/PCDFs
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD) on the
Vitamin A Status of Hartley Guinea Pigs, SD Rats, C57B1/6
Mice, DBA/2 Mice, and Golden Syrian Hamsters
Effect of 2,3,7,8-Tetrachlorodibenzo-p-dioxin on the Hepatic
7-Ethoxyresorufin o-deethylase Activity in Four Rodent
Species
InUtero and Lactational Exposure to 2,3,7,8-Tetrachloro-
dibenzo-p-dioxin Alters Postnatal Development of Seminal
Vesicle Epithelium
A Mixture of Dioxins, Furans, and Non-ortho PCBs Based
Upon Consensus TEQ Factors Produces Dioxin-like
Reproductive Effects
Evaluation of the Effect of Low-level
2,3,7,8-Tetrachlorodibenzo-/>-dioxin Exposure on Cell
Mediated Immunity
Reason for excluding study
Genetically altered
animals
X
-
-


-
-





Low dose
too high

X
-
X
X
X
-


X


Doses not TCDD only;
unspecified TCDD dose

-
X


-
X



X

Nonoral dose

-
X


-
-
X
X


X

-------
          Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Many et al. (1999)
Harper et al. (1991)
Harper et al. Q994a)
Harper et al. (1994b)
Harris etal. (1973)
Hart (1972)
Harvey et al. (1993)
Hassoun et al. (1984a)
Hassoun et al. (1984b)
Hassoun et al. (1995)
Title of study
Behavioral Effects Following Single and Combined Maternal
Exposure to PCB 77 (3,4,3',4'-Tetrachlorobiphenyl) andPCB
47 (2,4,2',4'- Tetrachlorobiphenyl) in Rats
Ah Receptor in Mice Genetically "Nonresponsive" for
Cytochrome P4501A1 Induction: Cytosolic Ah Receptor,
Transformation to the Nuclear Binding State, and Induction of
Aryl Hydrocarbon Hydroxylase by Halogenated and
Nonhalogenated Aromatic Hydrocarbons in Embryonic Tissues
and Cells
An Enzyme-linked Immunosorbent Assay (ELIS A) Specific for
Antibodies to TNP-LPS Detects Alterations in Serum
Immunoglobulins and Isotype Switching in C57BL/6 and
DBA/2 Mice Exposed to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
and Related Compounds
Inhibition of Estrogen-induced Progesterone Receptor in MCF-
7 Human Breast Cancer Cells by Aryl Hydrocarbon (Ah)
Receptor Agonists
General Biological Effects of TCDD in Laboratory Animals
Manipulation of Neonatal Androgen: Effects on Sexual
Responses and Penile Development in Male Rats
Spontaneous and Carcinogen-induced Tumorigenesis inP53
Deficient Mice
Teratogenicity of 2,3,7,8-Tetrachloro-dibenzofuran inBXD
Recombinant Inbred Strains
Teratological Studies on the TCDD Congener
3,3N,4,4N-Tetrachloro-azoxybenzene in Sensitive and
Nonsensitive Mouse Strains: Evidence for Direct Effect on
Embryonic Tissues
Evidence of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD)-
Induced Tissue Damage in Fetal and Placental Tissues and
Changes in Amniotic Fluid Lipid Metabolites of Pregnant CF1
Mice
Reason for excluding study
Genetically altered
animals

X
X
X
X
-
X
-


Low dose
too high




X
-
-
-

X
Doses not TCDD only;
unspecified TCDD dose




-
X
-
X
X

Nonoral dose
X



-
-
-
X


o
VO

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Hassoun et al. (1997)
Hassoun et al. (2001)
Hassoun et al. (2004)
Hassoun et al. (2006)
Hebert et al. (1990)
Heimler et al. (1998)
Hemming et al. (1993)
Hemming et al. (1995)
Henck et al. (1981)
Henry and Gasiewicz
(1987)
Henry et al. (2006)
Title of study
Modulation of TCDD -induced Fetotoxicity and Oxidative
Stress in Embryonic and Placenta! Tissues of C57BL/6J Mice
by Vitamin E Succinate and Ellagic Acid
Production of Superoxide Anion, Lipid Peroxidation and DNA
Damage in the Hepatic and Brain Tissues of Rats after
Subchronic Exposure to Mixtures of TCDD and its Congeners
The Modulatory Effects of Ellagic Acid and Vitamin E
Succinate on TCDD-Induced Oxidative Stress in Different
Brain Regions of Rats after Subchronic Exposure
The Effects of Ellagic Acid and Vitamin E Succinate on
Antioxidant Enzymes Activities and Glutathione Levels in
Different Brain Regions of Rats After Subchronic Exposure to
TCDD
Relative Toxicity and Tumor-promoting Ability of
2,3,7,8-Tetrachlorodibenzo-/>-dioxin(TCDD),
2,3,4,7,8-Pentachlorodibenzofuran (PCDF), and
1,2,3,4,7,8-Hexachlorodibenzofuran (HCDF) in Hairless Mice
Dioxin Perturbs, in a Dose- and Time-Dependent Fashion,
Steroid Secretion, and Induces Apoptosis of Human Luteinized
Granule sa Cells
Relative Tumor Promoting Activity of Three Polychlorinated
Biphenyls in Rat Liver
Liver Tumor Promoting Activity of 3,4,5, 3',4'-Pentachloro-
biphenyl and its Interaction with 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin
2,3,7,8-Tetrachlorodibenzo-p-dioxin: Acute Oral Toxicity in
Hamsters
Changes in Thyroid Hormones and Thyroxine Glucuronidation
in Hamsters Compared with Rats Following Treatment with
2,3,7,8-Tetrachlorodibenzo-^-dioxin
A Potential Endogenous Ligand for the Aryl Hydrocarbon
Receptor Has Potent Agonist Activity In Vitro and In Vivo
Reason for excluding study
Genetically altered
animals





X
-

-

X
Low dose
too high
X

X
X


-

X

-
Doses not TCDD only;
unspecified TCDD dose

X




-
X
-

-
Nonoral dose




X

X

-
X
-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Herbet et al. (1990)
Hermsen et al. (2008)
Herr et al. (1996)
Herzke et al. (2002)
Hinsdill et al. (1980)
Hochstein et al. (1998)
Hoegberg et al. (2005)
Hofer et al. (2004)
Hoffer et al. (1996)
Hosaboam et al. (2008)
Hojo et al. (2006)
Holcomb and Safe
(19941
Holene et al. (1995)
Title of study
Relative Toxicity and Tumor-promoting Ability of
2,3,7,8-Tetrachlorodibenzo-p-dioxin(TCDD),
2,3,4,7,8-Pentachlorodibenzofuran (PCDF), and
1,2,3,4,7,8-Hexachorodibenzofuran (HCDF) in Hairless Mice
In Utero and Lactational Exposure to
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) Affects Bone
Tissue in Rhesus Monkeys
Developmental Exposure to Aroclor 1254 Produces Low-
frequency Alterations in Adult Rat Brainstem Auditory Evoked
Responses
Kinetics and Organotropy of Some Polyfluorinated Dibenzo-p-
dioxins and Dibenzofurans (PFDD/PFDF) in Rats
[mmunosuppression in Mice Induced by Dioxin (TCDD) in
Feed
Effects of Dietary Exposure to 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin in Adult Female Mink (Mustela vison)
Retinoid Status and Responsiveness to
2,3,7,8,-Tetrachlorodibenzo-p-dioxin (TCDD) in Mice Lacking
Retinoid Binding Protein or Retinoid Receptor Forms- Exp 3
Simultaneous Exposure of Rats to Dioxin and Carbon
Monoxide Reduces the Xenobiotic but Not the Hypoxic
Response
Dioxin Induces Transcription of Fos and Jun Genes by Ah
Receptor-dependent and -Independent Pathways
The Aryl Hydrocarbon Receptor Affects Distinct Tissue
Sex-specific Alterations of Cerebral Cortical Cell Size in Rats
Exposed Prenatally to Dioxin
Inhibition of 7,12-Dimethylbenzanthracene-induced Rat
Mammary Tumor Growth by 2,3,7,8-Tetrachlorodibenzo-p-
dioxin
Behavioral Effects of Pre- and Postnatal Exposure to Individual
Polychlorinated Biphenyl Congeners in Rats
Reason for excluding study
Genetically altered
animals



-
-
-
X

X
-
-

-
Low dose
too high



-
X
X
X
X
-
X
X

-
Doses not TCDD only;
unspecified TCDD dose


X
-
-
-


-
-
-
X
X
Nonoral dose
X
X

X
-
-


-
-
-

-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Holladay et al. (1991)
Holman et al. (2000)
Hood et al. (2006)
Hook et al. (1975)
House et al. (1990)
Hung et al. (2006)
Hurst et al. (2000)
Hurst et al. (2002)
Hushka et al. (19981
Huuskonen et al.
(1994)
Hwang et al. (2004)
Title of study
Perinatal Thymocyte Antigen Expression and Postnatal
Immune Development Altered by Gestational Exposure to
retrachlorodibenzo-p-dioxin (TCDD)
Low-dose Responses to 2,3,7,8-Tetrachlorodibenzo-p-dioxinin
Single Living Human Cells Measured by Synchrotron Infrared
Spectromicroscopy
Gestational 2,3,7,8-Tetrachlorodibenzo-p-dioxin Exposure
Effects on Sensory Cortex Function
Induction and Suppression of Hepatic and Extrahepatic
Microsomal Foreign-compound-metabolizing Enzyme Systems
by 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Examination of Immune Parameters and Host Resistance
Mechanisms in B6C3FJ Mice Following Adult Exposure to
2,3,7,8-Tetrachlorodibenzo-p-dioxin
Protective Effects of Tea Melanin against
2,3,7,8-Tetrachlorodibenzo-p-dioxin-InducedToxicity:
Antioxidant Activity and Aryl Hydrocarbon Receptor
Suppressive Effect
Acute Administration of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD) in Pregnant Long Evans Rats: Association of
Measured Tissue Concentrations with Developmental Effects
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) Disrupts Early
Morphogenetic Events That Form the Lower Reproductive
Tract in Female Rat Fetuses
Characterization of 2,3,7,8-Tetrachloro-dibenzofuran-
dependent Suppression and AH Receptor Pathway Gene
Expression in the Developing Mouse Mammary Gland
Developmental Toxicity of 2,3,7,8-Tetrachlorodibenzo-p-
dioxin (TCDD) in the Most TCDD-resistant and -Susceptible
Rat Strains
Panax Ginseng Improves Survival and Sperm Quality in
Guinea Pigs exposed to 2,3,7,8-TCDD
Reason for excluding study
Genetically altered
animals

X
-







-
Low dose
too high
X

X
X

X
X
X

X
-
Doses not TCDD only;
unspecified TCDD dose


-





X

-
Nonoral dose


-

X





X

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Iba et al. (2001)
Ikeda et al. (20Q5a)
Inouye et al. (2005)
loannou et al. (1983)
Ishida et al. (2004)
Ishimura et al. (2002)
Ishimura et al. (2006)
Ishizuka et al. (2003)
[to et al. (1980)
[to et al. (2002)
[to et al. (2008)
Jain et al. (1998)
Jamsa et al. (2001)
Title of study
Pulmonary CYP1A1 and CYP1A2 Levels and Activities in
Adult Mate and Female Offspring of Rats Exposed During
Gestation and Lactation to 2,3,7,8-TCDD
In Utero and Lactational Exposure to 2,3,7,8-TCDD in Rats
Disrupts Brain Sexual Differentiation
T cell-derived IL-5 Production is a Sensitive Target of
2,3,7,8-TCDD
Toxicity and Distribution of 2,3,7,8-Tetrachlorodibenzofuran in
Male Guinea Pigs
Reduction of the Toxicity of 2,3,7,8-TCDD in Mice Using an
Antiulcer Drug, Gerany[gerany[acetone
Increased Glycogen Content and Glucose Transporter 3 mRNA
Level in the Placenta of Holtzman rats After Exposure to
2,3,7,8-TCDD
Suppressive Effect of 2,3,7,8-Tetrachlorodibenzo-p-dioxin on
Vascular Remodeling That Takes Place in the Normal
Labyrinth Zone of Rat Placenta during Late Gestation
Perinatal Exposure to Low Doses of 2,3,7,8-
retrachlorodibenzo-p-dioxin Alters Sex-Dependent Expression
of Hepatic CYP2C 11
The Effects of Various Chemicals on the Development of
Hyperplastic Liver Nodules in Hepatectomized Rats Treated
with N-nitrosodiethylamine or N-2-fluorenylacetamide
Mechanism of TCDD-Induced Suppression of Antibody
Production: Effect on T Cell-Derived Cytokine Production in
the Primary Immune Reaction of Mice
TCDD Exposure Exacerbates Atopic Dermatitis-related
Inflammation in NC/Nga Mice
Expression of ARNT, ARNT2, HIF1 Alpha, HIF2 Alpha and
Ah Receptor mRNAs in the Developing Mouse
Effects of 2,3,7,8-tetrachlorodibenzo-p-Dioxin on Bone in Two
Rat Strains with Different Aryl Hydrocarbon Receptor
Structures (subcutaneous exposure)
Reason for excluding study
Genetically altered
animals

-
-
-
-





-
-

Low dose
too high
X
X
X
-
X
X
X



X
-

Doses not TCDD only;
unspecified TCDD dose

-
-
X
-


X
X
X
-
X

Nonoral dose
X
-
-
-
-





-
-
X

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Jang et al. (2007)
Jang et al. (2008)
Janz and Bellward
(1996)
Jean-Faucher et al.
(1982)
Jeong et al. (2008)
Jin et al. (2008a)
Jinetal. (2Q08b)
Jin et al. (2008c)
Jinno et al. (2006)
Johnson et al. (1992)
Johnson et al. (1994)
Johnson et al. (1997)
Title of study
Antiteratogenic Effects of Alpha-naphthoflavone on
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) Exposed Mice In
Utero
Antiteratogenic Effect of Resveratrol in Mice Exposed In Utero
to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
In Ovo 2,3,7,8-Tetrachlorodibenzo-p-dioxin Exposure in Three
Avian Species
The Effect of Preweaning Under-nutrition Upon the Sexual
Development of Male Mice. Biol Neonate 41:45-51
Accumulation of MldG DNA Adducts After Chronic Exposure
to PCBs, but Not From Acute Exposure to Fob/chlorinated
Aromatic Hydrocarbons-mixtures Study
Enhanced TGF-(31 is Involved in 2,3,7,8-Tetrachlorodibenzo-p-
dioxin (TCDD) Induced Oxidative Stress in C57BL/6 Mouse
Testis
In Utero Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Affects the Development of Reproductive System in Mouse-IP
Injection
Toxic Effects of Lactational Exposure to
2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD) on Development
of Male Reproductive System: Involvement of Antioxidants,
Oxidants, and p53 Protein
Induction of Cytochrome P450-1 A by the Equine Estrogen
Equilenin, a New Endogenous Aryl Hydrocarbon Receptor
Ligand
Reduced Leydig Cell Volume and Function in Adult Rats
Exposed to 2,3,7,8-Tetrachlorodibenzo-p-dioxin Without a
Significant Effect on Spermatogenesis. Toxicology
76(2):103-118
2,3,7,8-Tetrachlorodibenzo-p-dioxin Reduces the Number,
Size, and Organelle Content of Leydig Cells in Adult Rat
Testes
Promotion of Endometriosis in Mice by Fob/chlorinated
Dibenzo-p- dioxins, Dibenzofurans, and Biphenyls
Reason for excluding study
Genetically altered
animals

-
X
-







-
Low dose
too high
X
X
-
-

X

X

X
X
X
Doses not TCDD only;
unspecified TCDD dose

-
-
X
X



X


-
Nonoral dose

-
-
-


X

X
X
X
-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Johnson et al. (2000)
Jones and Greig (1975)
Kakeyama et al. (2001)
Kakeyama et al. (2003)
Kakeyama et al. (2008)
Kamath et al. (1997)
Kamath et al. (1999)
Katz et al. (19841
Kedderis et al. (1991)
Keller etal. (2007a)
Keller et al. (2007b)
Title of study
Sensitivity of the SRBC PFC Assay Versus ELISA for
Detection of Immunosuppressionby TCDD and TCDD-like
Congeners
Pathological Changes in the Liver of Mice Given
2,3,7,8-Tetrachlorodibenzo-p-dioxin
Changes in Expression of NMD A Receptor Subunit mRNA by
Perinatal Exposure to Dioxin
Perinatal Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Alters Activity-dependent Expression of BDNF mRNA in the
Neurocortex and Male Rat Sexual Behavior in Adulthood
Perinatal Exposure of Female Rats to
2,3,7,8-Tetrachlorodibenzo-/>-dioxin Induces Central
Precocious Puberty in the Offspring
Evidence for the Induction of Apoptosis in Thymocytes by
2,3,7,8-Tetrachlorodibenzo-p-dioxin In Vivo
Role of Fas-Fas Ligand Interactions in
2,3,7,8-Tetrachlorodibenzo-/>-dioxin(TCDD)-induced
Immunotoxicity: Increased Resistance of Thymocytes From
Fasdeficient (lpr)and Fas Ligand-defective (gld) Mice to
TCDD-induced Toxicity
Characterization of the Enhanced Paw Edema Response to
Carrageenan and Dextran in 2,3,7,8-Tetrachlorodibenzo-p-
dioxin-treated Rats
Disposition of 2,3,7,8-tetrabromodibenzo-p-dioxin and
2,3,7,8-Tetrachlorodibenzo-p-dioxininthe Rat: Biliary
Excretion and Induction of Cytochromes CYP1 Al and
CYP1A2
2,3,7,8-Tetrachlorodibenzo-p-dioxin Affects Fluctuating
Asymmetry of Molar Shape in Mice, and an Epistatic
Interaction of Two Genes for Molar Size
The Effects of 2,3,7, 8-Tetrachlorodibenzo-p-dioxin on Molar
and Mandible Traits in Congenic Mice: A Test of the Role of
the Ahr Locus
Reason for excluding study
Genetically altered
animals

-
-


-





Low dose
too high
X
X
X
X
X
-



X
X
Doses not TCDD only;
unspecified TCDD dose

-
-


-

X



Nonoral dose

-
-


X
X

X



-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Kelley et al. (1998)
Kelley et al. (2000)
Kelling et al. (1985)
Kelling et al. (1987)
Kerkvliet and Brauner
(1990)
Kerkvliet and Oughton
(1993)
Kerkvliet et al. (1990)
Kerkvliet et al. (1996)
Kerkvliet et al. (2002)
Khera (1992)
Title of study
Use of Model-based Compartmental Analysis to Study Effects
of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin on Vitamin A Kinetics
in Rats
Mobilization of Vitamin A Stores in Rats After Administration
of 2,3,7,8-Tetrachlorodibenzo-p-dioxin: a Kinetic Analysis
Hypophagia-induced Weight Loss in Mice, Rats, and Guinea
Pigs Treated with 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin treatment on
Mechanical Function of the Rat Heart
Flow Cytometric Analysis of Lymphocyte Subpopulations in
the Spleen and Thymus of Mice Exposed to an Acute
[mmunosuppressive Dose of 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin
Acute Inflammatory Response to Sheep Red Blood Cell
Challenge in Mice Treated with 2,3,7,8-Tetrachlorodibenzo-p-
dioxin (TCDD): Phenotypic and Functional Analysis of
Peritoneal Exudate Cells
Role of the Ah Locus in Suppression of Cytotoxic T
Lymphocyte (CTL) Activity by Halogenated Aromatic
Hydrocarbons (PCBs and TCDD): Structure-activity
Relationships and Effects in C57B1/6 Mice
Inhibition of TC-1 Cytokine Production, Effector Cytotoxic T
Lymphocyte Development and Alloantibody Production by
2,3,7,8- Tetrachlorodibenzo-p-dioxin (TCDD)
T Lymphocytes Are Direct, Aryl Hydrocarbon Receptor
(AhR)-Dependent Targets of 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin (TCDD): AhR Expression in Both CD4+ and CD8+ T
Cells Is Necessary for Full Suppression of a Cytotoxic T
Lymphocyte Response by TCDD
Extraembryonic Tissue Changes Induced by
2,3,7,8-Tetrachloro-dibenzo-p-dioxinand
2,3,4,7,8-Pentachlorodibenzofuran with a Note on Direction of
Maternal Blood Flow in the Labyrinth of C57BL/6N Mice
Reason for excluding study
Genetically altered
animals

-
-
-






Low dose
too high
X
X
X
X
X
X
X
X
X
X
Doses not TCDD only;
unspecified TCDD dose

-
-
-






Nonoral dose

-
-
-







-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Khera and Ruddick
(19731
Kim et al. (2003a)
Kim et al. (2003b)
Kimmig and Schultz
(1957)
Kitajima et al. (2004a)
Kitajima et al. (2004b)
Kitamura et al. (2006)
Kleeman et al. (1990)
Ko et al. (2002)
Ko et al. (2004)
Title of study
Polychlorodibenzo-p-dioxins: Perinatal Effects and the
Dominant Lethal Test in Wistar rats. In: Chlorodioxins —
Origin and Fate. Blair, EH, ed. Washington, DC: American
Chemical Society; pp. 7084
Area Under the Curve as a Dose Metric for Promotional
Responses Following 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
Exposure
Effects of Benzo[a]pyrene, 2-Bromopropane, phenol and
2,3,7,8-TCDD on IL-6 Production in Mice After Single or
Repeated Exposure-IP Injection
Chlorierte Aromatische Zyklische Ather Als Ursache Der
Sogenannten Chlorakne
Expression of the Arylhydrocarbon Receptor in the Peri-
implantation Period of the Mouse Uterus and the Impact of
Dioxin on Mouse Implantation-subcutaneous Injection
Histomorphometric Alteration and Cell-type Specific
Modulation of Arylhydrocarbon receptor and Estrogen
Receptor Expression by 2,3,7,8-TCDD and 17p-estradiol in
Mouse Experimental Model of Endometriosis-subcutaneous
Injection
Mechanistic Investigation of the Cause for Reduced Toxicity of
TCDD in wa-1 homozygous TGFa Mutant Strain of Mice as
Compared its Matching Wild-type Counterpart, C57BL/6J
Mice-IP Injection
Inhibition of Testicular Steroidogenesis in
2,3,7,8-Tetrachlorodibenzo-p-dioxin-treated Rats: Evidence
That the Key Lesion Occurs Prior to or During Pregnenolone
Formation
In Utero and Lactational Exposure to 2,3,7,8-TCDD in the
C57BL/6J Mouse Prostate: Lobe-specific Effects on Branching
Morphogenesis
Evidence that Inhibited Prostatic Epithelial Bud Formation in
2,3,7,8-TCDD-exposed C57BL/6J Fetal Mice is Not Due to
Interruption of Androgen Signaling in the Urogenital Sinus
Reason for excluding study
Genetically altered
animals



-






Low dose
too high
X
X

-



X
X
X
Doses not TCDD only;
unspecified TCDD dose



-






Nonoral dose


X
X
X
X
X




-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Kopec et al. (2008)
Kopfetal. (2008)
Korenaga et al. (2007)
Korte et al. (1990)
Kozak (1997)
Kransler et al. (2007a)
Kransler et al. (2007b)
Kransler et al. (2008)
Kransler et al. (2009)
Kronenberg et al.
(2000)
Krowke et al. (1989)
Kruger et al. (1990)
Title of study
Comparative Toxicogenomic Examination of the Hepatic
Effects of PCB126 and TCDD in Immature, Ovariectomized
C57BL/6 Mice
Hypertension, Cardiac Hypertrophy, and Impaired Vascular
Relaxation Induced by 2,3,7,8-Tetrachlorodibenzo-p-dioxin are
Associated With Increased Superoxide
Long-term Effects of Subcutaneously Injected
2,3,7,8-Tetrachlorodibenzo-/>-dioxin on the Liver of Rhesus
Monkeys-subcutaneous Injection
Induction of Hepatic Monooxygenases in Female Rats and
Offspring in Correlation with TCDD Tissue Concentrations
After Single Treatment During Pregnancy
ARNT-deficient Mice and Placental Differentiation
Comparative Developmental Toxicity of
2,3,7,8-Tetrachlorodibenzo-p-dioxininthe Hamster, Rat, and
Guinea Pig
Gestational Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Alters Retinoid Homeostasis in Maternal and Perinatal Tissues
of the Holtzman Rat
Effects of Helicobacter infection on Developmental Toxicity of
2,3,7, 8-Tetrachlorodibenzo-p-dioxin in Holtzman rats
Lung Development in the Holtzman rat is Adversely Affected
by Gestational Exposure to 2,3,7,8-Tetrachlorodibenzo-p-
dioxin
Generation of a(3 T-cell receptor+ CD4- CD8+ cells in Major
Histocompatibility Complex Class-I-deficient Mice Upon
Activation of the Aryl Hydrocarbon Receptor by
2,3,7,8-Tetrachlorodibenzo-p-dioxin-IP Injection
Pharmacokinetics and Biological Activity of
2,3,7, S-Tetrachlorodibenzo-p-dioxin. 2. Pharmacokinetics in
Rats Using a Loading-Dose/Maintenance-dose Regime With
High Doses
Induction of Caffeine-demethylations by 2,3,7,8-TCDD in
Marmoset Monkeys Measured with a 14CO2 -breath Test
Reason for excluding study
Genetically altered
animals


X

-


-



-
Low dose
too high
X
X


-
X
X
X
X


-
Doses not TCDD only;
unspecified TCDD dose




X


-



-
Nonoral dose



X
-


-

X
X
X

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Kwon et al. (2004)
Laiosa et al. (2002)
Lakind et al. (2000)
Lakshman et al. (1988)
Lakshman et al. (1989)
Lakshman et al. (1991)
Latchoumycandane and
Mathur (2002)
Laurent et al. (2002)
Lawrence and
Vorderstrasse (2004)
Lawrence et al. (2000)
Lawrence et al. (2006)
Lee et al. (2007)
Title of study
Protective Effects of Ursodeoxycholic Acid Against
2,3,7,8-Tetrachlorodibenzo-p-dioxin-inducedTesticular
Damage in Mice-subcutaneous Injection
2,3,7,8-Tetrachlorodibenzo-p-dioxin Causes Alteration in
Lymphocyte Development and Thymic Atrophy in
Hemopoietic Chimeras Generated from Mice Deficient in
ARNT2-IV Injection
Methodology For Characterizing Distributions Of Incremental
Body Burdens Of 2,3,7,8-TCDD And DDE From Breast Milk
In North American Nursing Infants
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD) onDe
Novo Fatty Acid and Cholesterol Synthesis in the Rat
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin on Lipid
Synthesis and Lipogenic Enzymes in the Rat
Mechanism of Action of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
on Intermediary Metabolism in the Rat
Effects of Vitamin E on Reactive Oxygen Species-mediated
2,3,7,8-Tetrachlorodibenzo-p-dioxin Toxicity in Rat Testis
Portal Absorption of 14C After Ingestion of Spiked Milk With
14C-Phenanthrene, 14C-Benzo[a]pyrene or 14C-TCDD in
Growing Pigs
Activation of the Aryl Hydrocarbon Receptor Diminishes the
Memory Response to Homotypic Influenza Virus Infection but
Does Not Impair Host Resistance
Fewer T lymphocytes and Decreased Pulmonary Influenza
Virus Burden in Mice Exposed to 2,3,7,8-Tetrachlorodibenzo-
p-dioxin (TCDD)
Aryl Hydrocarbon Receptor Activation Impairs the Priming but
Not the Recall of Influenza Virus-Specific CD8_ T Cells in the
Lung
Panax Ginseng Effects on DNA Damage, CYP1A1 Expression
and Histopathological Changes in Testes of Rats Exposed to
2, 3 ,7,8 -Tetrachlorodibenzo -p-dioxin-IP Inj ection
Reason for excluding study
Genetically altered
animals


X
-
-
-
-





Low dose
too high



X
-
X
-
X
X
X
X

Doses not TCDD only;
unspecified TCDD dose



-
-
-
X





Nonoral dose
X
X

-
X
-
-




X

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Lensu et al. (2006)
Lewis et al. (2001)
Li et al. (1995a)
Li et al. (1995b)
Li et al. Q995c)
Lilienthal and Winneke
(1991)
Lilienthal et al. (1997)
Lim et al. (2006)
Lin et al. (1991)
Lin et al. (2001)
Lin et al. (2002a)
Title of study
Assessment by c-Fos Immuno staining of Changes in Brain
Neural Activity Induced by 2,3,7,8-Tetrachlorodibenzo-p-
Dioxin (TCDD) and Leptin in Rats
In Utero and Lactational Treatment with
2,3,7,8-Tetrachlorodibenzo-p-dioxin Impairs Mammary Gland
Differentiation but Does Not Block the Response to Exogenous
Estrogen in the Postpubertal Female Rat
Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) on
Estrous Cyclicity and Ovulation in Female Sprague-Dawley
Rats
Reproductive Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD) in Female Rats: Ovulation, Hormonal Regulation, and
Possible Mechanism(s)
Toxicokinetics of 2,3,7, 8-Tetrachlorodibenzo-p-dioxin in
Female Sprague-Dawley Rats Including Placental and
Lactational Transfer to Fetuses and Neonates
Sensitive Periods for Behavioral Toxicity of Fob/chlorinated
Biphenyls: Determination by Cross-fostering in Rats
Effects of Maternal Exposure to 3,3',4,4'-Tetrachlorobiphenyl
or Propylthiouracil in Rats Trained to Discriminate
Apomorphine From Saline
Dihydroxy-, Hydroxyspirolactone-, and
Dihydroxyspirolactone-urochlorins Induced by 2,3,7,8-
retrachlorodibenzo-p-dioxin in the Liver of Mice
The Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) on
the Hepatic Estrogen and Glucocorticoid Receptors in
Congenic Strains of Ah Responsive and Ah Nonresponsive
C57BL/6 Mice
Role of the Aryl Hydrocarbon Receptor in the Development of
Control and 2,3,7,8- Tetrachlorodibenzo-p-dioxin-Exposed
Male Mice
Critical Window of Vulnerability for Effects of
2,3,7,8-Tetrachlorodibenzo-p-dioxin on Prostate and Seminal
Vesicle Development in C57BL/6 Mice
Reason for excluding study
Genetically altered
animals
X




-





Low dose
too high

X
X
X
X
-

X
X
X
X
Doses not TCDD only;
unspecified TCDD dose





X
X




Nonoral dose





-






-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Lin et al. (2002b)
Linden et al. (2005)
Liu et al. (2003)
Loertscher et al. (2002)
Lucieretal. (1973)
Lucier et al. (1975a)
Lucier et al. (1975b)
Lucier et al. (1991)
Luebeck et al. (2000)
Luebke et al. (1994)
Luebke et al. (1995)
Luebke et al. (1999)
Title of study
Effects of Aryl Hydrocarbon Receptor Null Mutation and In
Utero and Lactational 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Exposure on Prostate and Seminal Vesicle Development in
C57BL/6 Mice
Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) and
Leptin on Hypothalamic mRNA Expression of Factors
Participating in Food Intake Regulation in a TCDD-Sensitive
and a TCDD-Resistant Rat Strain
Induction of Aryl Hydrocarbon Receptor and CYP1 Al mRNA
by 2,3,7,8-Tetrachlorodibenzo-/>-dioxin in Rat Liver-IP
Injection
In Utero Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Causes Accelerated Terminal Differentiation in Fetal Mouse
Skin
TCDD-induced Changes in Rat Liver Microsomal Enzymes
Nature of the Enhancement of Uridine Diphosphate
Glucuronyltransferase Activity by 2,3,7,8-Tetrachlorodibenzo-
o-dioxin in Rats
Postnatal Stimulation of Hepatic Microsomal Enzymes
Following Administration of TCDD to Pregnant Rats
Ovarian Hormones Enhance 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin-mediated Increases in Cell Proliferation and
Preneoplastic Foci in a Two -stage Model for Rat
Hepatocarcinogenesis
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin on Initiation
and Promotion of GST-P -Positive Foci in Rat Liver: A
Quantitative Analysis of Experimental Data Using a Stochastic
Model-subcutaneous injection
Assessment of Host Resistance to Trichinella spiralis in Mice
Following Pre-infection Exposure to 2,3,7,8-TCDD
Host Resistance to T. spiralis infection in Rats Exposed to
2,3,7,8-Tetrachlorodibenzo-p-dioxin(TCDD)
Effects of Aging on Resistance to Trichinella spiralis Infection
in Rodents Exposed to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Reason for excluding study
Genetically altered
animals




-

-


-
-
-
Low dose
too high
X
X

X
X
X
X


-
-
X
Doses not TCDD only;
unspecified TCDD dose




-

-
X

-
-
-
Nonoral dose


X

-

-

X
X
X
-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Luebke et al. (2001)
Luebke et al. (2002)
Lundberg et al. (1990)
Luster et al. (1980)
Luster et al. (1985)
Ma et al. (20071
Mably et al. (19901
Mably et al. (19911
MacLusky et al. (1998)
Madhukar et al. (1984)
Madhukar et al. (1988)
Mann (1997)
Mantovani et al. (1980)
Title of study
Suppression of Allergic Immune Responses to House Dust
Mites in Rats Exposed to 2,3,7,8-TCDD-IP Injection
Mortality in Dioxin-exposed Mice Infected With Influenza:
Mitochondrial Toxicity (Reye's Like Symptoms) Versus
Enhanced Inflammation as a Mode of Action-IP Injection
Effects of 2,3,7,8-Tetrachlorodibenzo-^-dioxin (TCDD)
Treatment In Vivo on Thymocyte Functions in Mice After
Activation In Vitro
Examination of Bone Marrow, Immunologic Parameters and
Host Susceptibility Following Pre- and Postnatal Exposure to
2,3,7,8-Tetrachlorodibenzo-p-dioxin(TCDD)
Acute Myelotoxic Responses in Mice Exposed to
2,3,7,8-Tetrachlorodibenzo-/>-dioxin(TCDD)
Mouse Lung CYP1A1 Catalyzes the Metabolic Activation of
2- Amino - 1 -methyl-6 -phenylimidazo [4,5 -b]pyridine (PhlP)-IP
Injection
Hypergastrinemia is Associated With Decreased Gastric Acid
Secretion in 2,3,7,8-Tetrachlorodibenzo-/>-dioxin Treated Rats
The Male Reproduction System is Highly Sensitive to In Utero
and Lactational TCDD Exposure
Hormonal Interactions in the Effects of Halogenated Aromatic
Hydrocarbons on the Developing Brain
Effects of In Vivo Administered 2,3,7,8-Tetrachloro-dibenzo-/>-
dioxin on Receptor Binding of Epidermal Growth Factor in the
Hepatic Plasma Membrane of Rat, Guinea Pig, Mouse and
Hamster
2,3,7,8-Tetrachlorodibenzo-p-dioxin Causes an Increase in
Protein Kinases Associated With Epidermal Growth Factor
Receptor in the Hepatic Plasma Membrane
Selected Lesions of Dioxin in Laboratory Rodents
Effect of 2,3,7,8-Tetrachlorodibenzo-p-dioxin on Macrophage
and Natural Killer Cell Mediated Cytotoxicity in Mice
Reason for excluding study
Genetically altered
animals
-



-

-
-
-


-
-
Low dose
too high
-

X
X
X

X
X
-


-
-
Doses not TCDD only;
unspecified TCDD dose
-



-

-
-
X


-
-
Nonoral dose
X
X


-
X
-
-
-
X
X
X
X

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Markowski et al.
(20021
Marks (1985)
Marks and Staples
(19801
Marks et al. (1981)
Massart and Meucci
(2007)
Matsumura et al.
(1997)
Max and Silbergeld
(1987)
McConnell and Moore
(1979)
McConnell et al. (1978)
McGrath et al. (1995)
McKinley et al. (1993)
McKinney et al. (1985)
McNulty (1977)
Title of study
Impaired Cued Delayed Alternation Behavior in Adult Rat
Offspring Following Exposure to 2,3,7,8-Tetrachlorodibenzo-
p-dioxinonGD 15
Exposure to Toxic Agents: the Heme Biosynthetic Pathway and
Hemoproteins as Indicator
Teratogenic Evaluation of the Symmetrical Isomers of
Hexachlorobiphenyl (HCB) in the Mouse. In: Proceedings of
the 20th Annual Meeting of the Teratology Society, Portsmouth,
NH, June 1980, p. 54 A
Influence of Symmetrical Polychlorinated Biphenyl Isomers on
Embryo and Fetal Development in Mice
Environmental Thyroid Toxicants and Child Endocrine Health
Altered In Vivo Toxicity of 2,3,7,8-Tetrachlorodibenzo-p-
dioxin (TCDD) in c-src Deficient Mice
Skeletal Muscle Glucocorticoid Receptor and Glutamine
Synthetase Activity in the Wasting Syndrome in Rats Treated
with2,3,7,8-Tetrachlorodibenzo-/>-dioxin
Toxicopathology Characteristics of Halogenated Aromatic
Hydrocarbons
Toxicity of 2,3,7, S-Tetrachlorodibenzo-p-dioxin in Rhesus
Monkeys (Macaco mulatto) Following a Single Oral Dose
Alternative Models for Low Dose-response Analysis of
Biochemical and Immunological Endpoints for
retrachlorodibenzo-p-dioxin
The Effect of Pretreatment on the Biliary Excretion of
2,3,7,8-Tetrachlorodibenzo-/>-dioxin,
2,3,7,8-Tetrachlorodibenzofuran, and
3,3',4,4'-Tetrachlorobiphenyl in the rat
Molecular Interactions of Toxic Chlorinated Dibenzo-p-dioxins
and Dibenzofurans with Thyroxine Binding Prealbumin
Toxicity of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin for Rhesus
Monkeys: Brief Report
Reason for excluding study
Genetically altered
animals

-

-
X
-

-
-


-
-
Low dose
too high
X
-

-
-
-

-
X


-
X
Doses not TCDD only;
unspecified TCDD dose

X
X
X
-
-
X
X
-
X
X
X
-
Nonoral dose

-

-
-
X

-
-


-
-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
McNulty (1984)
McNulty (1985)
McNultvetal. (1982)
Mebus et al. (1987)
Meulenbelt and de
Vries (2005)
Mever (2002)
Michalak (2008)
Michalak et al. (200 la)
Michalak et al. (200 Ic)
Miettinen et al. (2002)
Miettinen et al. (2004)
Miettinen et al. (2005)
Miller (1985)
Miller et al. (1986)
Mimura et al. (1997)
Title of study
Fetotoxicity of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
for Rhesus Macaques (Macaco mulatto)
Toxicity and Fetotoxicity of TCDD, TCDF and PCB Isomers in
Rhesus Macaques (Macaca mulatto)
Persistence of TCDD in Monkey Adipose Tissue
Depression of Rat Testicular 17-Hydroxylase and 17,20-Lyase
After Administration of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
(TCDD)
Toxicity of Dioxins in Humans
Incidence of CTCL in Vietnam Veterans
Diabetes and Cancer in Veterans of Operation Ranch Hand
After Adjustment for Calendar Period, Days of Spraying, and
Time Spent in Southeast Asia
Relation of Serum 2,3,7,8-Tetrachloro-p-dioxin (TCDD) Levels
to Hematological Examination Results in Veterans of
Operation Ranch Hand
Serum Dioxin and Hepatic Abnormalities in Veterans of
Operation Ranch Hand
Effect of In Utero and Lactational 2,3,7,8-Tetrachlorodibenzo-
p-dioxin Exposure on Rat Molar Development: The Role of
Exposure Time
Effects of Epidermal Growth Factor Receptor Deficiency and
2,3,7,8-Tetrachlorodibenzo-p-dioxin on Fetal Development in
Mice
Effects of In Utero and Lactational TCDD Exposure on Bone
Development in Differentially Sensitive Rat Lines
Congenital PCB Poisoning: a Reevaluation
Teratologic Evaluation of Hexabrominated Naphthalenes in
C57BL/6NMice
Loss of Teratogenic Response to 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin (TCDD) in Mice Lacking the Ah (dioxin) Receptor
Reason for excluding study
Genetically altered
animals
-
-
-

-
-


-


-
-
-
-
Low dose
too high
X
-
-
X
-
-


-


-
-
-
-
Doses not TCDD only;
unspecified TCDD dose
-
X
X

X
X
X
X
X
X
X
X
X
X
X
Nonoral dose
-
-
-

-
-


-


-
-
-
-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Mitchell and Lawrence
(2QQ3a)
Mitchell and Lawrence
(2QQ3b)
Mitchell et al. (2006)
Mitrou et al. (2001)
Mitsui et al. (2006)
Mittler et al. (1984)
Mizuyachi et al. (2002)
Mocarelli(2001)
Moennikes et al. (2004)
Moolgavkar et al.
(1996)
Moon et al. (2004)
Moon et al. (2008)
Title of study
Exposure to 2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD)
Renders Influenza Virus-Specific CD8_ T Cells
Hyporesponsive to Antigen
T cell Receptor Transgenic Mice Provide Novel Insights Into
Understanding Cellular Targets of TCDD: Suppression of
Antibody Production, but Not the Response of CD8+ T Cells,
During Infection with Influenza Virus
Sustained Aryl Hydrocarbon Receptor Activity Attenuates
Liver Regeneration
Toxic Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxinand
Related Compounds
Perinatal Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Suppresses Contextual Fear Conditioning-accompanied
Activation of Cyclic AMP Response Element-binding Protein
in the Hippocampal CA1 Region of Male Rats
Changes in Testosterone Hydroxylase Activity in Rat Testis
Following Administration of 2,3,7, 8-Tetrachlorodibenzo-p-
dioxin
Alteration in Ovarian Gene Expression in Response to
2,3,7,8-Tetrachlorodibenzo-/>-dioxin: Reduction of
Cyclooxygenase-2 in the Blockage of Ovulation
Seveso a Teaching Story
A Constitutively Active Dioxin/Aryl Hydrocarbon Receptor
Promotes Hepatocarcinogenesis in Mice
Quantitative Analysis of Enzyme-altered Liver Foci in Rats
Initiated with Diethylnitrosamine and Promoted with
2,3,7,8-Tetrachlorodibenzo-/>-dioxinor
1,2,3,4,6,7,8-Heptachlorodibenzo-p-dioxin
Effect of TCDD on Corpus Cavernosum Histology and Smooth
Muscle Physiology-IP Injection
A Single Administration of 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin that Produces Reduced Food and Water Intake Induces
Long-lasting Expression of Corticotropin-releasing Factor,
Arginine Vasopressin, and Proopiomelanocortin in Rat Brain
Reason for excluding study
Genetically altered
animals

X
-
-



-
-

-

Low dose
too high


-
-



-
-

-
X
Doses not TCDD only;
unspecified TCDD dose
X

X
X
X
X
X
X
X

X

Nonoral dose


-
-



-
-
X
-


-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Moore and Peterson
(19851
Moore et al. (1973)
Moore et al. (1976)
Moore et al. (1979)
Moore et al. (1985)
Moore et al. (1989)
Moore et al. (1991)
Moore et al. (1985)
Moore et al. (1992)
Moos et al. (1994)
Moran et al. (2001)
Moriguchi et al. (2003)
Morris et al. (1992)
Title of study
Enhanced Catabolism and Elimination of Androgens do Not
Cause the Androgenic Deficiency in
2,3,7,8-Tetrachlorodibenzo-p-dioxin-treatedRats
Postnatal Effects of Maternal Exposure to
2,3,7,8-Tetrachlorodibenzo-p-dioxin(TCDD)
Tissue Distribution of [14C] Tetrachlorodibenzo-p-dioxin in
Pregnant and Neonatal Rats
Comparative Toxicity of Three Halogenated Dibenzofurans in
Guinea Pigs, Mice, and Rhesus Monkeys
Enhanced Catabolism and Elimination of Androgens do Not
Cause the Androgenic Deficiency in
2,3,7,8-Tetrachlorodibenzo-p-dioxin-treatedRats
Plasma Concentrations of Pituitary Hormones in
2,3,7,8-Tetrachlorodibenzo-/>-dioxin-treated Male Rats
2,3,7,8-Tetrachlorodibenzo-p-dioxin Inhibits Steroidogenesis in
the Rat Testis by Inhibiting the Mobilization of Cholesterol to
Cytochrome P450 sec 1
Androgenic Deficiency in Male Rats Treated with
2,3,7,8-Tetrachlorodibenzo-p-dioxin
InUtero and Lactational 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
(TCDD) Exposure Decreases Androgenic Responsiveness of
Male Sex Organs and Permanently Inhibits Spermatogenesis
and Demasculinizes Sexual Behavior in Rats
Acute Inflammatory Response to Sheep Red Blood Cells in
Mice Treated with 2,3, 7,8-Tetrachlorodibenzo-/>-dioxin: the
Role of Proinflammatory Cytokines, IL-1 and TNF
Effect of Dioxin on Ovarian Function in the Cynomolgus
Macaque (M. fascicularis)
Distinct Response to Dioxin in an Arylhydrocarbon Receptor
(AHR) -humanized Mouse-IP Injection
Enhanced Suppression of Humoral Immunity in DBA/2 Mice
Following Subchronic Exposure to 2,3,7,8-Tetrachlorodibenzo-
P-dioxin (TCDD)
Reason for excluding study
Genetically altered
animals

-
X
-

-

-


X
-

Low dose
too high

X
-
-

-
X
X
X

X
-
X
Doses not TCDD only;
unspecified TCDD dose
X
-
-
X
X
X

-

X
-
X

Nonoral dose

-
-
-

-

-


-
-


-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Morrissey et al. (1992)
Morse et al. (1993)
Morse et al. (1996)
Moshammer and
Neuberger (2000)
Mukai et al. (2008)
Murante and Gasiewicz
(2000)
Mustafa et al. (2008)
Myllymaki et al. (2005)
Nagarkatti et al. (1984)
Nayyar et al. (2007)
Neff-LaFord et al.
(2003)
Negish et al. (2006)
Ness et al. (1993)
Title of study
Limited PCB Antagonism of TCDD-induced Malformations in
Mice
Interference of polychlorinated biphenyls in hepatic and brain
thyroid hormone metabolism in fetal and neonatal rats
Alterations in rat brain thyroid hormone status following pre-
and postnatal exposure to polychlorinated biphenyls (Aroclor
1254)
Sex ratio in the children of the Austrian chloracne cohort
Behavioral Rhythmicity of Mice Lacking AhR and Attenuation
of Light-Induced Phase Shift by 2,3,7,8-Tetrachlorodibenzo-/>-
Dioxin
Hemopoietic Progenitor Cells Are Sensitive Targets of
2,3,7, 8-Tetrachlorodibenzo-/>-dioxin in C57BL/6J Mice
An Enhanced Postnatal Autoimmune Profile in 24 Week-old
C57BL/6 Mice Developmentally Exposed to TCDD
InUtero and Lactational Exposure to TCDD; Steroidogenic
Outcomes Differ in Male and Female Rat Pups
Sensitivity of Suppression of Cytotoxic T Cell Generation by
2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD) is Dependent on
the Ah Genotype of the Murine Host
Developmental Exposure of Mice to TCDD Elicits a Similar
Uterine Phenotype in Adult Animals as Observed in Women
with Endometriosis
Fewer CTL, Not Enhanced NK Cells, are Sufficient for Viral
Clearance From the Lungs of Immunocompro raised Mice
Gestational and Lactational Exposure to
2,3,7,8-Tetrachlorodibenzo-f-dioxin Affects Social Behaviors
Between Developing Rhesus Monkeys (Macaco mulatto)
Effects of Perinatal Exposure to Specific PCB Congeners on
Thyroid Hormone Concentrations and Thyroid Histology in the
Rat
Reason for excluding study
Genetically altered
animals
-
-

-

-
-
-
X

-


Low dose
too high
X
-

X
X
X
X
X

X
X
X

Doses not TCDD only;
unspecified TCDD dose
-
X
X
-
X
-
-
-


-

X
Nonoral dose
-
-

-

-
-
-


-



-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Neubert et al. (1990)
Nienstedt et al. (1979)
Niittynen et al. (2003)
Niittynen et al. (2007)
Niittynen et al. (2008)
Nikolaidis et al. (1990)
Nilsson et al. (2000)
Nishijo et al. (2007)
Nishimura et al. (2001)
Nishimura et al. (2002)
Nishimura et al. (2003)
Title of study
Polyhalogenated Dibenzo-p-dioxins and Dibenzofurans and the
Immune System 1. Effects on Peripheral Lymphocyte
Subpopulations of a Non-human Primate (Callithrixjacchus)
After Treatment with 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD)
Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin on Hepatic
Metabolism Of Testosterone in the Rat
2,3,7,8-Tetrachlorodibenzo-p-dioxin(TCDD)-Induced
Accumulation of Biliverdin and Hepatic Peliosis in Rats
Differences in Acute Toxicity Syndromes of
2,3,7,8-Tetrachlorodibenzo-p-dioxinand
1,2,3,4,7,8-Hexachlorodibenzo-^-dioxin in Rats
Effect of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) on
Heme Oxygenase-1, Biliverdin IXa Reductase and 5-
aminolevulinic Acid Synthetase 1 in Rats with Wild-type or
Variant AH Receptor
TCDD Inhibits the Support of B -cell Development by the
Bursaof Fabricius
2,3,7,8-Tetrachlorodibenzo-p-dioxin Increases Serum and
Kidney Retinoic Acid Levels and Kidney Retinol Esterification
in the Rat
Effects of Maternal Exposure to 2,3,7,8-Tetrachlorodibenzo-p-
dioxin on Fetal Brain Growth and Motor and Behavioural
Development in Offspring Rats
Induction of Metallothionein in the Livers of Female Sprague-
Dawley Rats Treated with 2,3,7,8-Tetrachlorodibenzo-^-dioxin
Immunohistochemical Localization of Thyroid Stimulating
Hormone Induced by a Low Oral Dose of
2,3,7,8-Tetrachlorodibenzo-p-dioxin in Female Sprague-
Dawley Rats
Rat Thyroid Hyperplasia Induced by Gestational and
Lactational Exposure to 2,3,7,8-Tetrachlorodibenzo-p-Dioxin
Reason for excluding study
Genetically altered
animals

-
-

X
X


-

-
Low dose
too high

X
X
X
X
-
X
X
X
X
X
Doses not TCDD only;
unspecified TCDD dose

-
-


-


-

-
Nonoral dose
X
-
-


-


-

-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Nishimura et al.
(2QQ5a)
Nishimura et al.
(2QQ5b)
Nishimura et al. (2006)
Nishimura et al. (2008)
Nishiumi et al. (2008)
Nohara et al. (2000a)
Nohara et al. (2002b)
Nohara et al. (2008)
Nottebrock et al. (2006)
Novelli et al. (2005)
Ohbayashi et al. (2008)
Title of study
Altered Thyroxin and Retinoid Metabolic Response to
2,3,7,8-Tetrachlorodibenzo-/>-dioxin in Aryl Hydrocarbon
Receptor-null Mice
Disruption of Thyroid Hormone Homeostasis at Weaning of
Holtzman Rats by Lactational but Not In Utero Exposure to
2,3,7,8-Tetrachlorodibenzo-/>-Dioxin
Localization of Cytochrome P450 1 Al in a Specific Region of
Hydronephrotic Kidney of Rat Neonates Lactationally Exposed
to 2,3,7,8-Tetrachlorodibenzo-^-dioxin
Critical Role of Cyclooxygenase-2 Activation in Pathogenesis
of Hydronephrosis Caused by Lactational Exposure of Mice to
Dioxin
Involvement of SREBPs in 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin-induced Disruption of Lipid Metabolism in Male Guinea
Pig-IP Injection
Alterations of Thymocyte Development, Thymic Emigrants
and Peripheral T Cell Population in Rats Exposed to
2,3,7,8-Tetrachlorodibenzo-/>-dioxin
Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) on T
Cell-derived Cytokine Production in Ovalbumin (OVA)-
[mmunized C57B1/6 Mice
Arsenite -Induced Thymus Atrophy is Mediated by Cell Cycle
Arrest: A Characteristic Downregulation of E2F -Related Genes
Revealed by a Microarray Approach-IP injection
Effects of 2,3,7,8-Tetrachloro-dibenzo-p-dioxin on the
Extracellular Matrix of the Thymus in Juvenile Marmosets
(Callithrix /acc/zM5)-Subcutaneous Exposure
2,3,7,8-Tetrachlorodibenzo-p-dioxin-induced Impairment of
Glucose-stimulated Insulin Secretion in Isolated Rat Pancreatic
Islets-IP Injection
Occurrence of Two Different Types of Glutathione S-
Transferase Placental Form-Positive Hepatocytes after a Single
Administration of 2,3,7,8-Tetrabromodibenzo-pdioxin in Rats
Reason for excluding study
Genetically altered
animals







X



Low dose
too high
X
X
X
X

X
X



X
Doses not TCDD only;
unspecified TCDD dose











Nonoral dose




X


X
X
X


-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Ohsako et al. (2002)
Ohyama (2006)
Ohyama et al. (2007)
Okey et al. (1989)
Olson (1980)
Olson and McGarrigle
(1990)
Olson and McGarrigle
(1992)
Olson et al. (1990)
Operana et al. (2007)
Paajarvi et al. (2005)
Pan et al. (2004)
Pande et al. (2005)
Park et al. (2006)
Parkinson et al. (1983)
Title of study
Developmental Stage-Specific Effects of Perinatal
2,3,7,8-Tetrachlorodibenzo-/>-dioxin Exposure on Reproductive
Organs of Male Rat Offspring
Disorders of Sex Differentiation Caused by Exogenous
Hormones
Maternal Exposure of Low Dose of TCDD Modulates the
Expression of Estrogen Receptor Subunits of Male Gonads in
Offspring-subcutaneous Exposure
Detection and Characterization of a Low-affinity Form of
Cytosolic Ah Receptor in Livers of Mice Nonresponsive to
Induction of Cytochrome PI -450 by 3-Methylcholanthrene
Toxicity of 2,3,7,8-Tetrachlorodibenzo-/>-dioxininthe Golden
Syrian Hamster
Characterization of the Developmental Toxicity of
2,3,7,8-TCDD in the Golden Syrian Hamster
Comparative Developmental Toxicity of
2,3,7,8-Tetrachlorodibenzo-/>-dioxin(TCDD)
Developmental Toxicity of 2,3,7,8-TCDD in the Rat and
Hamster
Human CYP1A1AGFP Expression in Transgenic Mice Serves
as a Biomarker for Environmental Toxicant Exposure-IP
Injection
TCDD Activates Mdm2 and Attenuates the P53 Response to
DNA Damaging Agents
Evaluation of Relative Potencies of PCB 126 and PCB 169 for
the Immunotoxicities in Ovalbumin (OVA)-immunized Mice
Aspects of Dioxin Toxicity Are Mediated by Interleukin 1-Like
Cytokines-IP injection
The Therapeutic Effect of Tissue Cultured Root of Wild Panax
ginseng C.A. Mayer on Spermatogenetic Disorder-IP injection
Differential Time Course of Induction of Rat Liver Microsomal
Cytochrome P450 Isozymes and Epoxide Hydrolase by
Arochlor 1254
Reason for excluding study
Genetically altered
animals

-

X
-
-
-
-

-
-
-
-

Low dose
too high
X
-


X
X
X
X

X
X
-
-

Doses not TCDD only;
unspecified TCDD dose

X


-
-
-
-

-
-
-
X
X
Nonoral dose

-
X

-
-
-
-
X
-
-
X
-


-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Partanen et al. (1998)
Patterson et al. (2003)
Peraino et al. (1981)
Perucatti et al. (2006)
Pesonen et al. (2006)
Peters and Wiley
(1995)
Peters et al. (1999)
Petroffetal. (2000)
Petroffetal. (2001)
Petroffetal. (2002)
Pitt et al. (2000)
Pltiess et al. (1988)
Title of study
Epidermal Growth Factor Receptor as a Mediator of
Developmental Toxicity of Dioxin in Mouse Embryonic Teeth
Induction of Apoptosis by 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Following Endotoxin Exposure
Early Appearance of Histochemically Altered Hepatocyte Foci
and Liver Tumors in Female Rats Treated with Carcinogens 1
Day After Birth
Increased Frequencies of Both Chromosome Abnormalities and
SCEs in Two Sheep Flocks Exposed to High Dioxin Levels
During Pasturage
Effects of In Utero and Lactational Exposure to
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) on Rat Follicular
Steroidogenesis
Evidence that Murine Preimplantation Embryos Express Aryl
Hydrocarbon Receptor
Amelioration of TCDD -induced Teratogenesis in Aryl
Hydrocarbon Receptor (AhR)-null Mice
Interaction of Estradiol and 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin (TCDD) in an Ovulation Model: Evidence for Systemic
Potentiation and Local Ovarian Effects
The Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) on
Weight Gain and Hepatic Ethoxyresorufm-o-deethylase
(EROD) Induction Vary with Ovarian Hormonal Status in the
Immature Gonadotropin-primed Rat Model
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD) on
Serum Inhibin Concentrations and Inhibin Immunostaining
During Follicular Development in Female Sprague-Dawley
Rats
Adrenocoricotropin (ACTH) and Corticosterone Secretion by
Perifused Pituitary and Adrenal Glands From Rodents Exposed
to 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
Subchronic Toxicity of Some Chlorinated Dibenzofurans
(PCDFs) and a Mixture of PCDFs and Chlorinated
Dibenzodioxins (PCDDs) in rats
Reason for excluding study
Genetically altered
animals
-
-

X

-
X





Low dose
too high
-
X


X
-
X
X
X
X
X
X
Doses not TCDD only;
unspecified TCDD dose
-
-
X


X
-





Nonoral dose
X
-



-
-






-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Pohjanvirta et al.
(19881
Pohjanvirta et al.
(1989)
Pohjanvirta et al.
(19901
Pohjanvirta et al.
(1993)
Pohjanvirta et al.
(1998)
Pohjanvirta et al.
(2006)
Poland and Glover
(1990)
Poland etal. (1982)
Pollenz et al. (1998)
Porterfield et al. (2000)
Potter et al. (1983)
Potter et al. (1986a)
Title of study
Hepatic Ah-receptor Levels and the Effect of
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) on Hepatic
Microsomal Monooxygenase Activity in a TCDD-susceptible
and -resistant Rat Strain
The Central Nervous System May be Involved in TCDD
Toxicity
Effects of TCDD on Vitamin A Status and Liver Microsomal
Enzyme Activities in a TCDD-susceptible and a TCDD-
resistant Rat Strain
Comparative Acute Lethality of 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin (TCDD), 1,2,3,7,8-Pentachlorodibenzo-p-dioxin and
1,2,3,4,7,8- Hexachlorodibenzo-p-dioxin in the most TCDD-
susceptible and the Most TCDD -resistant Rat Strain
Point Mutation in Intron Sequence Causes Altered Carboxyl-
terminal Structure in the Aryl Hydrocarbon Receptor of the
most 2,3,7,8-Tetrachlorodibenzo-/>-dioxin-resistant Rat Strain
Evaluation of Various Housekeeping Genes for Their
Applicability for Normalization of mRNA Expression in
Dioxin-treated Rats
Characterization and Strain Distribution Pattern of the Murine
Ah Receptor Specified by the Ahd and Ahb-3 Alleles
Tumor Promotion by TCDD in Skin of HRS/J Mice
Female Sprague-Dawley Rats Exposed to a Single Oral Dose of
2,3,7,8-Tetrachlorodibenzo-p-dioxin Exhibit Sustained
Depletion of Aryl Hydrocarbon Receptor Protein in Liver,
Spleen, Thymus, and Lung
Thyroidal Dysfunction and Environmental Chemicals —
Potential Impact on Brain Development
Hypothyroxinemia and Hypothermia in Rats in Response to
2,3,7,8-Tetrachlorodibenzo-p-dioxin Administration
Relationship of Alterations in Energy Metabolism to
Hypophagia in Rats Treated with 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin
Reason for excluding study
Genetically altered
animals
X
-

X
X

-
-

-
-

Low dose
too high

-



X
-
-
X
-
X
X
Doses not TCDD only;
unspecified TCDD dose

-




X
-

X
-

Nonoral dose

X
X



-
X

-
-


-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Potter et al. (1986b)
Powers et al. (2005)
Prell et al. (2000)
Puhvel and Sakamoto
(1988)
Puhvel et al. (1982)
Puhvel et al. (1991)
Ramakrishna et al.
(2002)
Randerath et al. (1988)
Render et al. (2000)
Render et al. (2001)
Rhile et al. (1996)
Rice (1997)
Title of study
Thyroid Status and Thermogenesis in Rats Treated with
2,3,7,8-Tetrachlorodibenzo-p-dioxin
retrachlorodibenzo-p-dioxin Exposure Alters Radial Arm
Maze Performance and Hippocampal Morphology in Female
AhR+/- Mice
CTL Hyporesponsiveness Induced by
2,3,7,8-Tetrachlorodibenzo-/>-dioxin: Role of Cytokines and
Apoptosis
Effect of 2,3,7,8-Tetrachlorodibenzo-p-dioxin on Murine Skin
Hairless Mice as Models for Chloracne: a Study of Cutaneous
Changes Induced by Topical Application of Established
Chloracnegens
Vitamin A Deficiency and the Induction of Cutaneous Toxicity
in Murine Skin by TCDD
Decrease in K-ras p21 and Increase in Rafl and Activated Erkl
and 2 in Murine Lung Tumors Initiated by
N-nitrosodimethylamine and Promoted by 2,3,7,8-TCDD-IP
Injection
Organ-specific Effects of Long-term Feeding of
2,3,7,8-Tetrachlorodibenzo-p-dioxinand
1,2,3,7,8-Pentachlorodibenzo-p-dioxin on I-compounds in
Hepatic and Renal DNA of Female Sprague-Dawley Rats
Proliferation of Periodontal Squamous Epithelium in Mink Fed
2,3,7,8-Tetrachlorodibenzo-/>-dioxin(TCDD)
Squamous Epithelial Proliferation in the Jaws of Mink Fed
Diets Containing 3,3',4,4',5-Pentachlorobiphenyl (PCB 126) or
2,3,7,8-Tetrachlorodibenzo-p-dioxin(TCDD)
Role of Fas Apoptosis and MHC Genes in
2,3,7,8-Tetrachlorodibenzo-/>-dioxin(TCDD)-induced
Immunotoxicity of T Cells
Effect of Postnatal Exposure to a PCB Mixture in Monkeys on
Multiple Fixed Internal-fixed Ratio Performance
Reason for excluding study
Genetically altered
animals
-
X

-
X
-


-


-
Low dose
too high
X
X
X
-

-

X
X
X
X
-
Doses not TCDD only;
unspecified TCDD dose
-


-

-


-


X
Nonoral dose
-


X
X
X
X

-


-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Rice (1999)
Rice and Hayward
(19981
Rice and Hayward
(19991
Riecke et al. (2002)
Rier et al. (1993)
Rier et al. (1995)
Rier et al. (200 la)
Rifkind and Muschick
(1983)
Robv (2001)
Roman and Peterson
(1998)
Roman et al. (1995)
Title of study
Effect of Exposure to 3,3',4,4',5-Pentachlorobiphenyl (PCB
126) Throughout Gestation and Lactation on Development and
Spatial Delayed Alternation Performance in Rats
Lack of Effect of 3,3'4,4',5-Pentachlorobiphenyl (PCB 126)
Throughout Gestation and Lactation on Multiple Fixed
Interval-fixed Ratio and DRL Performance in Rats
Effects of Exposure to 3,3',4,4',5-Pentachlorobiphenyl (PCB
126) Throughout Gestation and Lactation on Behavior
(Concurrent Random Interval-random Interval and Progressive
Ratio Performance) in Rats
Low Doses of 2,3,7,8-Tetrachlorodibenzo-p-dioxin Increase
Transforming Growth Factor [TGF] (3 and Cause Myocardial
Fibrosis In Marmosets (Callithrixjacchus)-S\ibc\itaneo\is
Exposure
Endometriosis in Rhesus Monkeys (Macaca mulata) Following
Chronic Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
[mmunoresponsiveness in Endometriosis: Implications of
Estrogenic Toxicants
Increased Tumor Necrosis Factor-a Production by Peripheral
Blood Leukocytes from TCDD-exposed Rhesus Monkeys
Benoxaprofen Suppression of Fob/chlorinated Biphenyl
Toxicity Without Alteration of Mixed Function Oxidase
Function
Alterations in Follicle Development, Steroidogenesis, and
Gonadotropin Receptor Binding in a Model of Ovulatory
Blockade
In Utero and Lactational Exposure of the Male Rat to
2,3,7,8-Tetrachlorodibenzo-p-dioxin Impairs Prostate
Development
In Utero and Lactational Exposure of the Male Rat to
2,3,7,8-Tetrachlorodibenzo-p-dioxin: Impaired Prostate Growth
and Development Without Inhibited Androgen Production
Reason for excluding study
Genetically altered
animals




-
-
-




Low dose
too high




-
-
X

X
X
X
Doses not TCDD only;
unspecified TCDD dose
X
X
X

X
X
-
X



Nonoral dose



X
-
-
-





-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Roman et al. (1998)
Roman et al. (1998)
Romkes and Safe
(1988)
Rosenthal et al. (1989)
Rozman et al. (1984)
Russell et al. (1988)
Russo and Russo
(1978)
Ryo et al. (2006)
Salisbury and
Marcinkiewicz (2002)
Sanders et al. (1988)
Santostefano et al.
(1998)
Schantz et al. (1979)
Title of study
In Utero and Lactational Exposure of the Male Rat to
2,3,7,8-Tetrachlorodibenzo-/>-dioxin Impairs Prostate
Development. 1. Effects on Gene Expression
In Utero and Lactational Exposure of the Male Rat to
2,3,7,8-Tetrachlorodibenzo-/>-dioxin Impairs Prostate
Development. 2. Effects on Growth and Cytodifferentiation
Comparative Activities of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
and Progesterone as Antiestrogens in the Female Rat Uterus
Characteristics of 2,3,7,8-Tetrachlorodibenzo-p-dioxin Induced
Endotoxin Hypersensitivity: Association with Hepatotoxicity
Effect of Thyroidectomy and Thyroxine on
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)-induced Toxicity
Hypothalamic Site of Action of 2,3,7,8-Tetrachlorodibenzo-p-
dioxin (TCDD)
Developmental Stage of the Rat Mammary Gland as
Determinant of its Susceptibility to
7, 12-Dimethylbenz[a]anthracene
Germ-line Mutations at a Mouse ESTR (Pc-3) Locus and
Human Microsatellite Loci-IP Injection
In Utero and Lactational Exposure to
2,3,7,8-Tetrachlorodibenzo-/>-dioxinand
2,3,4,7,8-Pentachlorodibenzofuran Reduces Growth and
Disrupts Reproductive Parameters in Female Rats
Thyroid and Liver Trophic Changes in Rats Secondary to Liver
Microsomal Enzyme Induction Caused by an Experimental
Leukotriene Antagonist (L-649,923)
A Pharmacodynamic Analysis of TCDD -induced Cytochrome
P450 Gene Expression in Multiple Tissues: Dose- and Time-
dependent Effects
lexicological Effects Produced in Nonhuman Primates
Chronically Exposed to Fifty Parts per Trillion
2,3,7,8-Tetrachlorodibenzo-^-dioxin(TCDD)
Reason for excluding study
Genetically altered
animals


-
-
-
-

-




Low dose
too high
X
X
-
X
-
-

-
X

X
X
Doses not TCDD only;
unspecified TCDD dose


-
-
X
-
X
-

X


Nonoral dose


X
-
-
X

X





-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Schantz et al. (1991)
Schantz et al. (1995)
Schantz et al. (1997)
Schrenk et al. (1994)
Schulz et al. (2000)
Schuur et al. (1997)
Scott et al. (2001)
Seefeld and Peterson
(1984)
Seefeld et al. (1979)
Seefeld et al. (1984a)
Seefeld et al. (1984b1
Seegal et al. (1990)
Seegal et al. (1997)
Title of study
Effects of Perinatal Exposure to 2,3,7,8-Tetrachlorodibenzo-p-
dioxin (TCDD) on Spatial Learning and Memory and
Locomotor Activity in Rats
Spatial Learning Deficits in Adult Rats Exposed to Ortho-
substituted PCB Congeners During Gestation and Lactation
Long-term Effects of Developmental Exposure to
2,2',3,5',6-Pentachlorobiphenyl (PCB 95) on Locomotor
Activity, Spatial Learning and Memory and Brain Ryanodine
Binding
Promotion of Preneoplastic Foci in Rat Liver with
2,3,7,8-Tetrachlorodibenzo-/>-dioxin,
l,2,3,4,6,7,8-Heptachlorodibenzo-/>-dioxin and a Defined
Mixture of 49 Polychlorinated Dibenzo-p-dioxins
Identification of Theta-class Glutathione S-transferase in Liver
Cytosol of the Marmoset Monkey
Extrathyroidal Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
on Thyroid Hormone Turnover in Male Sprague-Dawley Rats
Exposure to the Dioxin 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
(TCDD) Induces Squamous Metaplasia in the Endocervix of
Cynomolgus Macaques
Digestible Energy and Efficiency of Feed Utilization in Rats
Treated with 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin on Indocyanine
Green Blood Clearance in Rhesus Monkeys
Body Weight Regulation in Rats Treated with
2,3,7,8-Tetrachlorodibenzo-p-dioxin
Characterization of the Wasting Syndrome in Rats Treated with
2,3,7,8-Tetrachlorodibenzo-p-dioxin
Lightly Chlorinated Ortho-substituted PCB Congeners
Decrease Dopamine in Nonhuman Primate Brain and in Tissue
Culture
Effects of In Utero and Lactational Exposure of the Laboratory
Rat to 2,4,2',4'- and 3,4,3',4'-Tetrachlorobiphenyl on Dopamine
Function
Reason for excluding study
Genetically altered
animals

-


-
-

-
-
-
-


Low dose
too high
X
-


-
-
X
X
X
X
X


Doses not TCDD only;
unspecified TCDD dose

X
X

X
-

-
-
-
-
X
X
Nonoral dose

-

X
-
X

-
-
-
-



-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Senft et al. (2002)
Seo and Meserve
(19951
Seo et al. (1999)
Seo et al. (2000)
Sewall et al. (1995b)
Shepherd et al. (2000)
Shepherd et al. (2001)
Shirota et al. (2007)
Shiverick and Muther
(1982)
Shiverick and Muther
(1983)
Shonetal. (2002)
Silkworth and Antrim
(1985)
Title of study
Mitochondrial Reactive Oxygen Production is Dependent on
the Aromatic Hydrocarbon Receptor-IP Injection
Effects of Maternal Ingestion of Aroclor 1254 (PCB) on the
Developmental Pattern of Oxygen Consumption and Body
Temperature in Neonatal Rats
Learning and Memory in Rats Gestationally and Lactationally
Exposed to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Radial Arm Maze Performance in Rats Following Gestational
and Lactational Exposure to 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin (TCDD)
TCDD Reduces Rat Hepatic Epidermal Growth Factor
Receptor: Comparison of Binding, Immunodetection, and
Autophosphorylation
The Effects of TCDD on the Activation of Ovalbumin (OVA)-
Specific DO 1 1. 10 Transgenic CD4+ T-cells in Adoptively
Transferred Mice
Anti-CD40 Treatment of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
(TCDD)-Exposed C57B1/6 Mice Induces Activation of Antigen
Presenting Cells Yet Fails to Overcome TCDD-Induced
Internal Dose-effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD) in Gonadotropin-primed Weanling Rat Model
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin on Serum
Concentrations and the Uterotrophic Action of Exogenous
Estrone in Rats
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) Effects on
Hepatic Microsomal Steroid Metabolism and Serum Estradiol
of Pregnant Rats
Effect of Chitosan Oligosaccharide on
2,3,7,8-Tetrachlorodibenzop-dioxin-Induced Oxidative Stress
in Mice
Relationship Between Ah Receptor-mediated Polychlorinated
Biphenyl (PCB)-induced Humoral Immunosuppression and
Thymic Atrophy
Reason for excluding study
Genetically altered
animals
-

-




-




Low dose
too high
-

X
X
X
X
X
X
X
X
X

Doses not TCDD only;
unspecified TCDD dose
-
X
-




-



X
Nonoral dose
X

-




-





-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Silkworth et al. (1984)
Silkworth et al. (1989)
Silkworth et al. (1997)
Sills et al. (19941
Simanainen et al.
(2004a)
Slezak et al. (1999)
Slezak et al. (2002)
Sloop and Lucier
(1987)
Smialowicz et al.
(1997)
Smith et al. (1981)
Smith etal. (1998)
Sommer et al. (2005)
Title of study
Correlations Between Polychlorinated Biphenyl
[mmunotoxicity, the Aromatic Hydrocarbon Locus, and Liver
Microsomal Enzyme Induction in C57B1/6 and DBA/2 Mice
Teratology of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin in a
Complex Environmental Mixture From the Love Canal
Tumor responses, PCB Tissue Concentrations and PCB Hepatic
Binding in S-D Rats Fed Aroclors 1016, 1242, 1254 or 1260
Tumor-Promoting Effects of 2,3,7,8-Tetrachlorodibenzo-p-
Dioxin and Phenobarbital in Initiated Weanling Sprague-
Dawley Rats: A Quantitative, Phenotypic, and ras p21 Protein
Study
Adult 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) Exposure
and Effects on Male Reproductive Organs in Three
Differentially TCDD-Susceptible Rat Lines
2,3,7,8-Tetrachlorodibenzo-p-dioxin-Mediated Oxidative Stress
in CYP1 A2 Knockout (CYP1A2-/-) Mice
TCDD-Mediated Oxidative Stress in Male Rat Pups Following
Perinatal Exposure
Dose-dependent Elevation of Ah Receptor Binding by TCDD
in Rat Liver
Opposite Effects of 2,2',4,4',5,5'-Hexachlorobiphenyl and
2,3,7,8-TCDD on the Antibody Response to Sheep
Erythrocytes in Mice
Hepatic Toxicity and Uroporphyrinogen Decarboxylase
Activity Following a Single Dose of 2,3,7,8-
Tetrachlorodibenzo-p-Dioxin to Mice
Interaction Between Iron Metabolism and
2,3,7,8-Tetrachlorodibenzo-f-dioxin in Mice with Variants of
the AhR Gene: a Hepatic Oxidative Mechanism
Early Developmental 2,3,7,8-Tetrachlorodibenzo-/>-Dioxin
Exposure Decreases Chick Embryo Heart Chronotropic
Response to Isoproterenol but Not to Agents Affecting Signals
Downstream of the Beta-Adrenergic Receptor
Reason for excluding study
Genetically altered
animals

-
-


-
-
-



X
Low dose
too high

-
-

X
X
X
X

X


Doses not TCDD only;
unspecified TCDD dose
X
X
X
X

-
-
-
X

X

Nonoral dose

-
-


-
-
-





-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Staples et al. (1998)
Stohsetal. (1983)
Sugihara et al. (2001)
Sweeney et al. (1979)
Takagi et al. (2000)
Tani et al. (2004)
Teske et al. (2005)
Thackaberry et al.
(2005a)
Thackaberry et al.
(2005b)
Theobald and Peterson
(1997)
Theobald et al. (2000)
Thigpen et al. (1975)
Title of study
Thymic Alterations Induced by 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin are Strictly Dependent on Aryl Hydrocarbon Receptor
Activation in Hematopoietic Cells
Lipid Peroxidation as a Possible Cause of TCDD Toxicity
Aryl Hydrocarbon Receptor (AhR) -Mediated Induction of
Xanthine Oxidase/Xanthine Dehydrogenase Activity by
2,3,7,8-Tetrachlorodibenzo-/>-dioxin
Iron Deficiency Prevents Liver Toxicity of
2,3,7,8-Tetrachlorodibenzo-p-dioxin
Pathogenesis of Cleft Palate in Mouse Embryos Exposed to
2,3,7,8-Tetrachlorodibenzo-^-dioxin(TCDD)
Follicular Epithelial Cell Hypertrophy Induced by Chronic Oral
Administration of 2,3,7, 8-Tetrachlorodibenzo-p-dioxin in
Female Harlan Sprague-Dawley Rats
Activation of the Aryl Hydrocarbon Receptor Increases
Pulmonary Neutrophilia and Diminishes Host Resistance to
Influenza A Virus
Effect of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin on Murine Heart
Development: Alteration in Fetal and Postnatal Cardiac
Growth, and Postnatal Cardiac Chronotropy
Toxicogenomic Profile of 2,3,7, 8-Tetrachlorodibenzo-p-Dioxin
in the Murine Fetal Heart: Modulation of Cell Cycle and
Extracellular Matrix Genes
In Utero and Lactational Exposure to
2,3,7,8-Tetrachlorodibenzo-rho-dioxin: Effects on
Development of the Male and Female Reproductive System of
the Mouse
2,3,7,8-Tetrachlorodibenzo-p-dioxin Inhibits Lumen Cell
Differentiation and Androgen Responsiveness of the Ventral
Prostate Without Inhibiting Prostatic 5a-Dihrdrotestosterone or
Testicular Androgen Production in Rat Offspring
Increased Susceptibility to Bacterial Infection as a Sequela of
Exposure to 2,3,7,8-Tetrachlorodibenzo-/>-dioxin
Reason for excluding study
Genetically altered
animals

-

X
-






-
Low dose
too high

X
X
-
X
X
X
X
X
X
X
X
Doses not TCDD only;
unspecified TCDD dose

-

-
-






-
Nonoral dose
X
-

-
-






-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Thomas and Hinsdill
(1919)
Thornton etal. (2001)
Thornton et al. (2004)
Thunberg (1984)
Thunberg and
Hakansson(1983)
Thunberg et al. (1979)
Tilson et al. (1979)
Timms et al. (2002)
Tomar and Kerkvliet
(1991)
Tritscher et al. (1995)
Tritscher et al. (1996)
Tritscher et al. (1999)
Tritscher et al. (2000)
Truelove et al. (1982)
Tsutsumi (2000)
Title of study
The Effect of Perinatal Exposure to Tetrachlorodibenzo-p-
dioxin on the Immune Response of Young Mice
Mutagenicity of TCDD in Big Blue® Transgenic Rats
The Dioxin TCDD Protects Against Aflatoxin-induced
Mutation in Female Rats, but Not in Male Rats
Effects of TCDD on Vitamin A and its Relation to TCDD
Toxicity
Vitamin A (retinol) Status in the Gunn Rat: the Effect of
2,3,7,8-Tetrachlorodibenzo-/>-dioxin
Vitamin A (Retinol) Status in the Rat After a Single Oral Dose
of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
The Effects of Fob/chlorinated Biphenyls Given Prenatally on
the Neurobehavioral Development of Mice
2,3,7,8-Tetrachlorodibenzo-p-dioxin Interacts with Endogenous
Estradiol to Disrupt Prostate Gland Morphogenesis in Male Rat
Fetuses
Reduced T helper Cell Function in Mice Exposed to
2,3,7,8-Tetrachlorodibenzo-p-dioxin(TCDD)
Persistence of TCDD -induced Hepatic Cell Proliferation and
Growth of Enzyme Altered Foci After Chronic Exposure
Followed by Cessation of Treatment in DEN Initiated Female
Rats
Increased Oxidative DNA Damage in Livers of
2,3,7,8-Tetrachlorodibenzo-/>-dioxin Treated Intact but Not
Ovariectomized Rats
TCDD-induced Lesions in Rat Lung After Chronic Oral
Exposure. Dioxin '99: 19th International Symposium on
Halogenated Environmental Organic Pollutants and POPs
Induction of Lung Lesions in Female Rats following Chronic
Exposure to 2,3,7,8-Tetrachlorodibenzo-^-dioxin
Polychlorinated Biphenyl Toxicity in the Pregnant Cynomolgus
Monkey: A Pilot Study
Effects of Endocrine Disrupters on Preimplantation Embryo
Development
Reason for excluding study
Genetically altered
animals
-
-
-
-
X
-
-

-



-
-
X
Low dose
too high
X
X
X
X
-
X
-
X
X
X
X
X
X
-
-
Doses not TCDD only;
unspecified TCDD dose
-
-
-
-
-
-
X

-



-
X
-
Nonoral dose
-
-
-
-
-
-
-

-



-
-
-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Tucker et al. (1986)
Tuner and Collins
(19831
Unkila et al. (1994a)
Unkila et al. (1994b)
Unkila et al. (1995)
Unkila et al. (19981
Ushinohama et al.
(2001)
Van Birgelen et al.
(1996)
Van Birgelen et al.
(1999b)
Van Birgelen et al.
(1999a)
Van den Berg et al.
(1987)
Vanden Heuvel (1994)
Title of study
Suppression of B Cell Differentiation by
2,3,7,8-Tetrachlorodibenzo-p-dioxin
Liver Morphology in Guinea Pigs Administered Either
Pyrolysis Products of a Fob/chlorinated Biphenyl Transformer
Fluid or 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Characterization of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD) Induced Brain Serotonin Metabolism in Rat
Dose Response and Time Course of Alterations in Tryptophan
Metabolism by 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) in
the Most TCDD- susceptible and the Most TCDD-resistant Rat
Strain: Relationship with TCDD Lethality
Effect of 2,3,7,8-Tetrachlorodibenzo-p-dioxin on Tryptophan
and Glucose Homeostasis in the Most TCDD-susceptible and
the Most TCDD-resistant Species, Guinea Pigs and Hamsters
Body Weight Loss and Changes in Tryptophan Homeostasis by
Chlorinated Dibenzo-p-dioxin Congeners in the Most TCDD-
Susceptible and the Most TCDD-resistant Rat Strain
Impaired Ovulationby 2,3,7,8 Tetrachlorodibenzo-p-dioxin
(TCDD) in Immature Rats Treated with Equine Chorionic
Gonadotropin
Synergistic Effect of 2,2',4,5,5'-Hexachlorobiphenyl and
2,3,7,8-Tetrachlorodibenzo-p-dioxin on Hepatic Porphyrin
Levels in the Rat
Dose and Time-response of TCDD in Tg. AC Mice After
Dermal and Oral Exposure. Dioxin'99: 19th International
Symposium on Halogenated Environmental Organic Pollutants
and POPs
Toxicity of 3,3',4,4'-Tetrachloroazobenzene in Rats and Mice
Transfer of Fob/chlorinated Dibenzo-p-dioxins and
Dibenzofurans to Fetal and Neonatal Rats
Accumulation of Fob/chlorinated Dibenzo-p-dioxins and
Dibenzofurans in Liver of Control Laboratory Rats
Reason for excluding study
Genetically altered
animals
-

-


X



-
-
-
Low dose
too high
X

X
X
X

X
X
X
X
-
-
Doses not TCDD only;
unspecified TCDD dose
-
X
-






-
X
X
Nonoral dose
-

-






-
-
-

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Van derKolk (1992)
Van Logten et al.
(1980)
Van Miller etal. (1977)
Vecchi et al. (1983)
Vezina et al. (2008)
Viluksela et al. (1995)
Viluksela et al. (1997b)
Viluksela et al. (1997a)
Viluksela et al. (19981
Viluksela et al. (1999)
Viluksela et al. (2000)
Title of study
Interactions of 2,2',4,4',5,5'- Hexachlorobiphenyl and
2,3,7,8-Tetrachlorodibenzo-/>-dioxin in a Subchronic Feeding
Study in the Rat
Role of the Endocrine System in the Action of
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) on the Thymus
Increased Incidence of Neoplasms in Rats Exposed to Low
Levels of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Immunosuppressive Effects of 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin in Strains of Mice with Different Susceptibility
Dioxin Causes Ventral Prostate Agenesis by Disrupting
Dorsoventral Patterning in Developing Mouse Prostate
Tissue-specific Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD) on the Activity of Phosphoeno/Pyruvate
Carboxykinase (PEPCK) in Rats
Subchronic/Chronic Toxicity of
1,2,3,4,6,7,8-Heptachlorodibenzop-dioxin (HpCDD) in Rats:
Part I. Design, General Observations, Hematology, and Liver
Concentrations
Subchronic/Chronic Toxicity of
1,2,3,4,6,7,8-Heptachlorodibenzop-dioxin (HpCDD) in Rats:
Part II. Biochemical Effects
Subchronic/Chronic Toxicity of Four Chlorinated Dibenzo-p-
dioxins in Rats. Part I. Design, General Observations,
Hematology, and Liver Concentrations
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD) on
Liver Phosphoenolpyruvate Carboxylase (PEPCK) Activity,
Glucose Homeostasis and Plasma Amino Acid Concentrations
in the Most TCDD-susceptible and the Most TCDD-resistant
Rat Strains
Liver Tumor-promoting Activity of 2,3,7,8-
retrachlorodibenzo-p-dioxin (TCDD) in TCDD-sensitive and
rCDD-resistant Rat Strains
Reason for excluding study
Genetically altered
animals

-
-
-
-





X
Low dose
too high

X
X
X
X
X
X
X

X
X
Doses not TCDD only;
unspecified TCDD dose
X
-
-
-
-



X


Nonoral dose

-
-
X
-







-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Vogel et al. (2003)
Vogel et al. (2007)
Vorderstrasse and
Kerkvliet (2001)
Vorderstrasse and
Lawrence (2006)
Vorderstrasse et al.
(2001)
Vorderstrasse et al.
(2003)
Vorderstrasse et al.
(2004)
Vorderstrasse et al.
(2006)
Vos and Moore (1974)
Vos et al. (19741
Vos et al. (19781
Title of study
The Use of c-src Knockout Mice for the Identification of the
Main Toxic Signaling Pathway of TCDD to Induce Wasting
Syndrome
Modulation of the Chemokines KC and MCP-1 by
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) in Mice
2,3,7,8-Tetrachlorodibenzo-p-dioxin Affects the Number and
Function of Murine Splenic Dendritic Cells and Their
Expression of Accessory Molecules
Protection Against Lethal Challenge with Streptococcus
Pneumoniae is Conferred by Aryl Hydrocarbon Receptor
Activation but is Not Associated with an Enhanced
Inflammatory Response
Aryl Hydrocarbon Receptor-deficient Mice Generate Normal
Immune Responses to Model Antigens and are Resistant to
TCDD-induced Immune Suppression
Examining the Relationship Between Impaired Host Resistance
and Altered Immune Function in Mice Treated with TCDD
Developmental Exposure to the Potent Aryl Hydrocarbon
Receptor Agonist 2,3,7,8-Tetrachlorodibenzo-/>-Dioxin Impairs
the Cell-Mediated Immune Response to Infection with
Influenza A Virus, but Enhances Elements of Innate Immunity
A Dose-response Study of the Effects of Prenatal and
Lactational Exposure to TCDD on the Immune Response to
Influenza A Virus
Suppression of Cellular Immunity in Rats and Mice by
Maternal Treatment with 2,3,7,8-Tetrachlorodibenzo-^-dioxin
Toxicity of 2,3,7, 8-Tetrachlorodibenzo-/>-dioxin (TCDD) in
C57B 1/6 Mice
Studies on 2,3,7,8-Tetrachlorodibenzo-p-dioxin-induced
Immune Suppression and Decreased Resistance to Infection:
Endotoxin Hypersensitivity, Serum Zinc Concentrations and
Effect of Thy mo sin Treatment
Reason for excluding study
Genetically altered
animals

-

X
X
X


-
-

Low dose
too high

-
X

X
-
X
X
X
X
X
Doses not TCDD only;
unspecified TCDD dose

-


X
-


-
-

Nonoral dose
X
X



-


-
-


-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Waern et al. (1991)
Wagner et al. (2001)
Wahba et al. (19881
Wahba et al. (19891
Wahba et al. (1990a)
Wahba et al. (1990^)
Walisser et al. (2004)
Walker et al. (1995)
Walker et al. (1997)
Walker et al. (1998a)
Walker et al. (1998b)
Walker et al. (1999)
Title of study
Effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) in the
Lactating Rat on Maternal and Neonatal Vitamin A Status and
Hepatic Enzyme Induction: A Dose-Response Study
2,3,7,8-Tetrachlorodibenzo-p-dioxin and Natural Immunity:
Lack of an Effect on the Complement System in a Guinea Pig
Model
Induction of Hepatic DNA Single Strand Breaks in Rats by
2,3,7,8-Tetrachlorodibenzo-p-dioxin(TCDD)
Factors Influencing the Induction of DNA Single Strand Breaks
in Rats by 2,3,7,8-Tetrachlorodibenzo-^-dioxin (TCDD)
Altered Hepatic Iron Distribution and Release in Rats After
Exposure to 2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD)
Desferrioxamine -induced Alterations in Hepatic Iron
Distribution, DNA Damage, and Lipid Peroxidation in Control
and 2,3,7,8-Tetrachlorodibenzo-/>-dioxin-treated Rats
Patent Ductus Venosus and Dioxin Resistance in Mice
Harboring a Hypomorphic ARNT Allele
Rat CYP1B1: an Adrenal Cytochrome P450 that Exhibits Sex-
dependent Expression in Livers and Kidneys of TCDD -treated
Animals
Hepatocarcinogenesis in a Sprague-Dawley Rat
Initiation/Promotion Model Following Discontinuous Exposure
to TCDD
Differences in Kinetics of Induction and Reversibility of
TCDD-Induced Changes in Cell Proliferation and CYP1A1
Expression in Female Sprague-Dawley Rat Liver
Induction and Localization of Cytochrome P450 IB 1
(CYP1B 1) Protein in the Livers of TCDD-treated Rats:
Detection Using Polyclonal Antibodies Raised to Histidine-
tagged Fusion Proteins Produced and Purified From Bacteria
Characterization of the Dose-response of CYP1B1, CYP1A1,
and CYP1A2 in the Liver of Female Sprague-Dawley Rats
Following Chronic Exposure to 2,3,7,8-Tetrachlorodibenzo-/>-
dioxin
Reason for excluding study
Genetically altered
animals


-
-
-

-





Low dose
too high


X
X
X

X
X
X
X
X

Doses not TCDD only;
unspecified TCDD dose


-
-
-
X
-





Nonoral dose
X
X
-
-
-

-




X

-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Walker et al. (2004)
Warren et al. (2000)
Weber and Birnbaum
(19851
Weber et al. (1985)
Weber et al. (1994)
Weinand-Harer et al.
(1997)
Weinstein et al. (2008)
Weissberg and Zinkl
(1973)
Wheatley (1968)
Widholm et al. (2003)
Wolfetal. (1999a)
Title of study
Persistent Suppression of Contact Hypersensitivity, and Altered
T-cell Parameters in F344 Rats Exposed Perinatally to
2,3, 7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
Exposure to 2,3,7,8-Tetrachlorodibenzo-/>-dioxin (TCDD)
Suppresses the Humoral and Cell-mediated Immune Responses
to Influenza A Virus Without Affecting Cytolytic Activity in
the Lung
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) and
2,3,7,8-Tetrachlorodibenzofuran (TCDF) in Pregnant C57BL/6
Mice: Distribution to the Embryo and Excretion
Teratogenic Potency of TCDD, TCDF and TCDD-TCDF
Combinations in C57BL/6N Mice
Reduced Activity of Tryptophan 2,3,-Dioxygenase in the Liver
of Rats Treated with Chlorinated Dibenzo-p-dioxins (CDDs):
Dose-responses and Structure-activity Relationship
Behavioral Effects of Maternal Exposure to an Ortho-
chlorinated or a Coplanar PCB Congener in Rats
Mid-gestation Exposure of C57BL/6 Mice to
2,3,7,8-Tetrachlorodibenzo-p-dioxin Causes Postnatal
Morphologic Changes in the Spleen and Liver
Effects of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin Upon
Hemostasis and Hematologic Function in the Rat
Enhancement and Inhibition of the Induction by
7,12-Dimethylbenz(a)anthracene of Mammary Tumors in
Female Sprague-Dawley Rats
Effects of Perinatal Exposure to 2,3,7,8-Tetrachlorodibenzo-p-
Dioxin on Spatial and Visual Reversal Learning in Rats
Administration of Potentially Antiandrogenic Pesticides
(Procymidone, Linuron, Iprodione, Chlozolinate, p,p'-DDE,
and Ketoconazole) and Toxic Substances (Dibutyl- and
Diethylhexyl Phthalate, PCB 169, and Ethane Dimethane
Sulphonate) During Sexual Differentiation Produces Diverse
Profiles of Reproductive Malformations in the Male Rat
Reason for excluding study
Genetically altered
animals

X

-

-

-

-

Low dose
too high
X

X
X
X
-
X
X

X

Doses not TCDD only;
unspecified TCDD dose



-

X

-
X
-
X
Nonoral dose


X
X

-

-

-


-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Wolfetal. Q999b)
Wu et al. (20041
Wyde et al. (1999)
Wyde et al. (20001
Wyde et al. (200 la)
Wyde et al. (200 lb)
Wyde et al. (20021
Wyde et al. (20041
Yang and Foster (1997)
Yang et al. (1983)
Yang et al. (1994)
Title of study
Gestational Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD) Severely Alters Reproductive Function of Female
Hamster Offspring [In Process Citation]
Exposure of Mouse Preimplantation Embryos to
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) Alters the
Methylation Status of Imprinted Genes H19 and Igf2
Influence of Ovariectomy and 17 B-Estradiol on the Promotion
of Altered Hepatocellular Foci by TCDD. Dioxin '99: 19th
International Symposium on Halogenated Environmental
Organic Pollutants and POPs
Toxicity of Chronic Exposure to 2,3,7,8-Tetrachlorodibenzo-p-
dioxin in Diethylnitrosamine-initiated Ovariectomized Rats
Implanted with Subcutaneous 17 Beta-estradiol Pellets
Induction of Hepatic 8-Oxo-deoxyguanosine adducts by
2,3,7,8-Tetrachlorodibenzo-p-dioxin in Sprague-Dawley Rats is
Female-specific and Estrogen-dependent
Regulation of 2,3,7,8-Tetrachlorodibenzo-/>-dioxin-induced
Tumor Promotion by 17 Beta-estradiol in Female Sprague-
Dawley Rats
Promotion of Altered Hepatic Foci by
2,3,7,8-Tetrachlorodibenzo-p-dioxinand 17Beta-estradiol in
Male Sprague-Dawley Rats
Oral and Dermal Exposure to 2,3,7,8-Tetrachlorodibenzo-p-
dioxin (TCDD) Induces Cutaneous Papillomas and Squamous
Cell Carcinomas in Female Hemizygous Tg. AC Transgenic
Mice
Continuous Exposure to 2,3,7,8-Tetrachlorodibenzo-p-dioxin
Inhibits the Growth of Surgically Induced Endometriosis in the
Ovariectomized Mouse Treated with High Dose Estradiol
Effects of Halogenated Dibenzo-p-dioxins on Plasma
Disappearance and Biliary Excretion of Ouabain in Rats
Effect of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) on
Pulmonary Influenza Virus Titer and Natural Killer (NK)
Activity in Rats
Reason for excluding study
Genetically altered
animals

X
X
X
X
X


X
-

Low dose
too high
X






X

X
X
Doses not TCDD only;
unspecified TCDD dose






X


-

Nonoral dose








X
-


-------
Table D-2. Noncancer animal studies not selected for TCDD dose-response analyses and reasons for exclusion (continued)
Author (year)
Yang et al. (2005)
Yasudaetal. (1999)
Ye and Leung (2008)
Yoon et al. (2000)
Yoon et al. (200 la)
Yoon et al. (200 Ib)
Yoon et al. (2006)
Zhu et al. (2008)
Zingeser (1979)
Zinkl et al. (1973)
Title of study
Inhibitory Effects of vitamin A on TCDD -induced Cytochrome
P-450 1 Al Enzyme Activity and Expression
Palatal rugae Anomalies Induced by Dioxins in Mice
Effect of Dioxin Exposure on Aromatase Expression in
Ovariectomized Rats
Teratological Effect of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD): Induction of Cleft Palate in the DDY and C57BL/6
Mouse
Hemopoietic Cell Kinetics After Intraperitoneal Single
Injection of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) in
Mice-IP Injection
Transgene Expression of Thioredoxin (TRX/ADF) Protects
Against 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD)-Induced
Hematotoxicity-IP injection
Gene Expression Profile by 2,3,7,8-Tetrachlorodibenzo-p-
dioxin in the Liver of Wild-type (Ahr +/+) and Aryl
Hydrocarbon Receptor Deficient (Ahr -/-) Mice-IP Injection
Effect of 2,3,7,8-Tetrachlorodibenzo-p-dioxin Administration
and High-fat Diet on the Body Weight and Hepatic Estrogen
Metabolism in Female C3H/HeN Mice-IP Injection
Anomalous Development of the Soft Palate in Rhesus
Macaques (Macaca mulatto) Prenatally Exposed to
3,4,7,8-Tetrachlorodibenzo-p-dioxin
Hematologic and Clinical Chemistry Effects of
2,3,7,8-Tetrachlorodi-benzo-p-dioxin in Laboratory Animals
Totals
Reason for excluding study
Genetically altered
animals
-
-
-






-
66
Low dose
too high
X
X
X
X





X
370
Doses not TCDD only;
unspecified TCDD dose
-
-
-





X
-
140
Nonoral dose
-
-
-

X
X
X
X

-
135

-------
           Table D-3. Cross-species concordance of male reproductive effects
Study
(1) Bell et al. (20070)
(2) Ishihara et al. (2007)
(3) Ikeda et al. (20051))
(4) Kociba et al. (1976)
(5) Latchoumycandane
and Mathur (2002)
(6) Mocarelli et al. (2008)
(7) Ohsako et al. (2001)
(8) Simanainen et al.
(2004b)
Species
Rat
Mouse
Rat
Rat
Rat
Human
Rat
Rat
Specific endpoint
Delayed balanopreputial separation
Increased ventral prostate weight
Higher proportion of abnormal sperm
Altered sex ratio (decreased
percentage of males)
Decreased ventral prostate weight
Altered sex ratio (decreased
percentage of males)
ncreased testes weight
Decreased daily sperm production
Decreased testis, epididymis, seminal
vesicle, and ventral prostate weights
Decreased sperm count, progressive
sperm motility, and total number of
motile sperm
Decreased anogenital distance
Decreased urogenital complex and
ventral prostate weights
Decreased daily sperm production
Decreased ventral prostate weight
Epididymal degeneration
Endpoint category
Altered sexual
development
Organ weight changes
Sperm effects
Altered sex ratio
Organ weight changes
Altered sex ratio
Organ weight changes
Sperm effects
Organ weight changes
Sperm effects
Altered sexual
development
Organ weight changes
Sperm effects
Organ weight changes
Organ toxicity
Administered dose
(ng/kg-day)
NOAEL
2.40E+00
2.40E+00
8.00E+00
l.OOE-01
-
—
7.14E+01
-
—

1.25E+01
5.00E+01
l.OOE+02
3.00E+02
3.00E+02
LOAEL
8.00E+00
8.00E+00
4.60E+01
l.OOE+02
1.65E+01
1.65E+01
7.14E+02
l.OOE+00
l.OOE+00

5.00E+01
2.00E+02
3.00E+02
l.OOE+03
l.OOE+03
Human-equivalent dose
(HED)a (ng/kg-day)
NOAEL
8.85E-02
8.85E-02
3.23E-01
4.91E-05
-
—
3.03E+00
-
—

2.74E-02
1.78E-01
4.33E-01
1.70E+00
1.70E+00
LOAEL
3.23E-01
3.23E-01
2.05E+00
4.96E-01
2.75E+00
2.75E+00
3.19E+01
1.62E-02
1.62E-02
2.01E-02
1.78E-01
1.04E+00
1.70E+00
6.92E+00
6.92E+00
oo
     "Human equivalent dose (HED) for rat and mouse studies based on Emond rodent and human PBPK models described in Section 3.3.6.

-------
            Table D-4. Cross-species concordance of female reproductive effects
Study
(1) Bowman etal.(1989a;
1989b)
(2) Eskenazi et al. (2002).
(3) Franczak et al. (2006)
(4) Hutt et al. (20081
(5) Li et al. (1997)
(6) Li et al. (2006)
(7) Murray et al. (1979)
(8) Shi et al. (2007)
(9) Smith et al. (1976)
(10) Sparschu et al. (2008:
1971)
Species
Monkey
Human
Rat
Rat
Rat
Mouse
Rat
Rat
Mouse
Rat
Specific endpoint
Reduced reproductive rate
Decreased days of offspring survival
Increased length of menstrual period
Altered estrus cyclicity
Lower proportion of morphologically normal
preimplantation embryos
Increased serum FSH
Increased serum LH
Increased serum estradiol, decreased serum
progesterone
Early embryo loss
Decreased uterine weight
Reduced fertility
Reduced neonatal survival
Decreased serum estradiol
Accelerated reproductive senescence with normal
cyclicity
Delayed vaginal opening
Increased percentage of resorptions per implantations
Decreased mean number of viable fetuses per litter
Endpoint category
Reduced fertility
Decreased offspring survival
Altered menstrual cycle
Altered menstrual cycle
Early embryo loss
Altered hormone levels
Altered hormone levels
Altered hormone levels
Early embryo loss
Organ weight changes
Reduced fertility
Decreased offspring survival
Altered hormone levels
Altered menstrual cycle
Altered sexual development
Late embryo loss
Late embryo loss
Administered dose
(ng/kg-day)
NOAEL
1.20E-01
1.20E-01
-
-
—
3.00E+00
l.OOE+02

2.00E+00
2.00E+00
l.OOE+00
l.OOE+00
1.43E-01
7.14E-01
7.14E+00
l.OOE+02
1.25E+02
LOAEL
6.70E-01
6.70E-01
-
7.14E+00
7.14E+00
l.OOE+01
3.00E+02
2.00E+00
5.00E+01
5.00E+01
l.OOE+01
l.OOE+01
7.14E-01
7.14E+00
2.86E+01
l.OOE+03
5.00E+02
Human-equivalent
dose (HED)a
(ng/kg-day)
NOAEL
8.22E-03b
8.22E-03b
-
-
—
2.90E-03
3.78E-01

1.58E-03
1.58E-03
2.89E-02
2.89E-02
4.47E-03
2.69E-02
3.18E-01
5.24E-01
1.73E+00
LOAEL
4.59E-02 b
4.59E-02 b
3.11E+02
3.18E-01
2.52E-01
1.67E-02
1.48E+00
1.58E-03
1.31E-01
1.31E-01
3.79E-01
3.79E-01
2.69E-02
3.18E-01
1.34E+00
7.61E+00
8.03E+00
VO
     aHED for rat and mouse studies based on Emond rodent and human PBPK models described in Section 3.3.6.
     bHED based on 1st order body burden model described in Section 3.3.4.2.

-------
      Table D-5. Cross-species concordance of thyroid effects
Study
(1) Baccarelli et al.
(2008)
(2) Chu et al. (20071
(3) Crofton et al. (2005)
(4) NTP (20061
(5) Seo et al. (19951
(6) Sewall et al. (1995a)
(7) Simanainen et al.
(2002)
(8) VanBirgelen et al.
(1995a)
Species
Human
Rat
Rat
Rat
Rat
Rat
Rat
Rat
Specific endpoint
Elevated blood TSH in male and
female neonates
Reduced follicles, reduced colloid
density, and increased epithelial
height in females
Reduced serum T4 levels in females
Reduced serum free and total T4
levels at 14 and 3 1 weeks
Increased serum total T3 levels at 53
weeks
Follicular cell hypertrophy at 2 years
Increased serum TSH levels in
females
Decreased serum T4 and thymus
weight
Decreased serum T4
Increased serum TSH levels in
females
Decreased serum T4
Reduced serum free and total T4
levels in females
Endpoint category
Altered hormone levels
Histopathological lesions
Altered hormone levels
Altered hormone levels
Altered hormone levels
Histopathological lesions
Altered hormone levels
Altered hormone levels
Altered hormone levels
Altered hormone levels
Altered hormone levels
Altered hormone levels
Administered dose
(ng/kg-day)
NOAEL
—
2.50E+02
3.00E+01
7.14E+00
7.14E+00
7.14E+00
1.57E+01
2.50E+01
5.16E+OOb
3.57E+01
l.OOE+02
2.64E+01
LOAEL
—
l.OOE+03
l.OOE+02
1.57E+01
1.57E+01
1.57E+01
3.29E+01
l.OOE+02
3.57E+01
1.25E+02
3.00E+02
4.69E+01
Human-equivalent dose
(HED)a
(ng/kg-day)
NOAEL
—
7.03E+00
1.69E-01
4.09E-01
4.34E-01
4.53E-01
9.98E-01
1.67E-01
1.80E-01b
1.71E+00
4.26E-01
1.05E+00
LOAEL
2.00E-02
2.96E+01
7.43E-01
9.14E-01
9.63E-01
9.98E-01
2.09E+00
9.15E-01
1.71E+00
6.30E+00
1.67E+00
1.93E+00
JHED for rat and mouse studies based on Emond rodent and human PBPK models described in Section 3.3.6.
3 Benchmark dose lower confidence bound (BMDL) used instead of NOAEL.

-------
      Table D-6.  Cross-species concordance of developmental dental effects
Study
(1) Alaluusua et al.
(2004)
(2) Kattainen et al. (2001)
(3) Keller et al. (2QQ8a;
2008b; 2007c)

Species
Human
Rat
Mouse
Specific endpoints
Developmental dental defects
Reduced mesiodistal length of the
lower third molar in males and
females
Variation in molar morphology and
shape, decreased mandible shape and
size in males and females
Endpoints category
Enamel defects
Altered tooth morphology
Altered tooth morphology
Administered dose
(ng/kg-day)
NOAEL
—


LOAEL
—
3.00E+01
l.OOE+01
Human-equivalent dose
(HED)a
(ng/kg-day)
NOAEL
4.06E-02


LOAEL
9.00E-01
9.01E-02
9.88E-03
aHED for rat and mouse studies based on Emond rodent and human PBPK models described in Section 3.3.6.

-------
          Table D-7. Cross-species concordance of immune system effects
Study
(DChuetal. (2001)
(2) Chu et al. (2007)
(3) DeCaprio et al. (1986)
(4) Franc et al. (2001)
(5)Kocibaetal. (1976)
(6)Kocibaetal. (1978)
(7) Simanainen et al.
(2002)
(8) Simanainen et al.
(2003)
(9) Smialowicz et al.
(2004)
(9) Smialowicz et al.
(2004)
(10) Smialowicz et al.
(2008)
(1 1) VanBirgelen et al.
(1995a)
Species
Rat
Rat
Guinea pig
Rat
Rat
Rat
Rat
Rat
Mouse
Mouse
Mouse
Rat
Specific endpoint
Decreased relative thymus weight in females
Reduced thymic cortex and increased
medullar volume in females
Decreased thymus weight in females
Decreased relative thymus weight in males
Decreased relative thymus weight in females
Increased relative spleen and thymus weights
in males and females
Decreased relative thymus weight
Thymic and splenic atrophy in females
Decreased relative thymus weight in females
Decreased relative thymus weight
Decreased antibody response to SRBCs in
females
Decreased thymus weight in females
Decreased antibody response to SRBCs in
females
Decreased relative spleen weight in females
Decreased absolute and relative thymus
weight in females
Endpoint category
Organ weight changes
Histopathological lesions
Organ weight changes
Organ weight changes
Organ weight changes
Organ weight changes
Organ weight changes
Organ weight changes
Organ weight changes
Organ weight changes
Immunosuppressive effects
Organ weight changes
Immunosuppressive effects
Organ weight changes
Organ weight changes
Administered dose
(ng/kg-day)
NOAEL
2.50E+02
2.50E+01
2.50E+02
6.10E-01
l.OOE+01
7.14E+01
l.OOE+01
l.OOE+01
3.00E+02
l.OOE+02
3.00E+02
3.00E+03
—
1.07E+01
—
LOAEL
l.OOE+03
2.50E+02
l.OOE+03
4.90E+00
3.00E+01
7.14E+02
l.OOE+02
l.OOE+02
l.OOE+03
3.00E+02
l.OOE+03
l.OOE+04
1.07E+00
1.07E+02
1.35E+01
Human-equivalent dose
(HED)a
(ng/kg-day)
NOAEL
7.03E+00
5.63E-01
7.03E+00
4.11E-03b
4.49E-01
3.03E+00
6.34E-01
6.34E-01
1.67E+00
4.26E-01
7.23E-01
1.18E+01
—
9.96E-02
—
LOAEL
2.96E+01
7.03E+00
2.96E+01
3.30E-02 b
1.41E+00
3.19E+01
6.35E+00
6.35E+00
6.80E+00
1.67E+00
3.28E+00
4.35E+01
6.26E-03
1.27E+00
5.14E-01
to

-------
       Table D-7.  Cross-species concordance of immune system effects (continued)
Study
(12) Vosetal. (1973)7
(13) White etal. (1986)
Species
Guinea pig
Mouse
Specific endpoint
Decreased delayed-type
hypersensitivity response to
tuberculin
Decreased relative thymus weight,
relative cervical lymph node weight
Cortical atrophy of the thymus,
lymphopenia and thymic
degeneration
Decreased serum complement
activity in females
Decreased component hemolytic
activity and C3 levels in females
Endpoint category
Immunosuppressive
effects
Organ weight changes
Histopathological lesions
Altered immune system
components
Altered immune system
components
Administered dose
(ng/kg-day)
NOAEL
1.14E+00
5.71E+00
5.71E+00

l.OOE+02
LOAEL
5.71E+00
2.86E+01
2.86E+01
l.OOE+01
5.00E+02
Human-equivalent dose
(HED)a
(ng/kg-day)
NOAEL
6.43E-03
3.22E-02
3.22E-02

5.07E-01 b
LOAEL
3.22E-02
1.61E-01
1.61E-01
2.77E-02 b
3.27E+OOb
JHED for rat and mouse studies based on Emond rodent and human PBPK models described in Section 3.3.6.
bHED based on 1st order body burden model described in Section 3.3.4.2.

-------
      Table D-8. Cross-species concordance of neurological effects
Study
(DSchantzetal. (1992)
(2) Hojo et al. (2002)
(3) Kuchiiwa et al. (2002)
(4) Markowski et al.
(2001)
(5) Schantz et al. (1996)
(6) Zareba et al. (2002)
Species
Monkey
Rat
Mouse
Rat
Rat
Rat
Specific endpoint
Altered social behavior
Food-reinforced operant behavior in
pups
Decreased number of serotonin-
immunoreactive neurons in the raphe
nuclei of males
Neurobehavioral effects in pups
(running, lever press, wheel
spinning)
Maze errors
Reduced cortical thickness and
altered brain morphometry in males
and females
Endpoint category
Neurobehavioral effects
Neurobehavioral effects
Histopathological lesions
Neurobehavioral effects
Neurobehavioral effects
Brain structural alterations
Administered dose
(ng/kg-day)
NOAEL
-
—


-
6.00E+01
LOAEL
1.20E-01
2.00E+01
7.00E-01
2.00E+01
2.50E+01
1.80E+02
Human-equivalent dose
(HED)a
(ng/kg-day)
NOAEL
-
—


-
2.35E-01
LOAEL
8.22E-03 b
5.51E-02
2.75E-03
5.15E-02
1.71E-01
9.54E-01
"HED for rat and mouse studies based on Emond rodent and human PBPK models described in Section 3.3.6.
3HED based on 1st order body burden model described in Section 3.3.4.2.

-------
Mouse
100 — 1 —
-i n
I
|>
I? 1
11
V) (0
-| °'1
1
£
Orn








<




>









Male Reproductive Effects
Rat




































OLOAEL
ANOAEL/BMDL
Human




<










4
I
>
i
L







i ^




4

k

I
/

t
1
i
• *
>
i


» 1
L ^


>
i


i


>



i







L
k













/"p* \5o^ \5o^ ^O vS^ v^ v\^
<^ ^0° ^? ^ y^ ^
fto ^5* ^\f v f ,$* ^
^ a^ 0^ ^ ^




»

\

Figure D-l.  Male reproductive effects across species.
The corresponding data are in Table D-3.  The numbers following the effect designations indicate the corresponding study in Table D-3. Vertical solid
black lines indicate the range of exposures tested below the LOAEL.

-------
        100
         10
  (0 o
  o: TJ


  11
  10 (0
    1C



    3
         0.1
        0.01
       0.001

                                     Female Reproductive Effects
                                   +LOAEL


                                   ANOAEL/BMDL
                     Mouse
               •     A
Rat
Monkey    Human
                          ^   ^
                                      N>    ^
                                                                                                     SV

                                                                                                          ^




                                                          "









                                                         <
Figure D-2. Female reproductive effects across species

The corresponding data are in Table D-4. The numbers following the effect designations indicate the corresponding study in Table D-4. Vertical solid

black lines indicate the range of exposures tested below the LOAEL.

-------
                                                   Thyroid Effects
                                                                                                »LOAEL


                                                                                                ANOAEL/BMDL
           100 -r=
                                                              Rat
            10
      o
      c
    c «.
    o c
    = 01

    p

    n
      V
      c
      n
0.1
          0.01
         0.001
                                                                                       	A_	
                                                                                                                 Human
                                          P
                                             0«r      «r      <
-------
oo
                                                  Developmental  Dental Effects
                                                                                      »LOAEL


                                                                                      ANOAEL/BMDL
                      100 T-
                                       Mouse
                      10
                 O)
                 c
                C «.
                o c
                = 01


                p

                n
                 V
                 c
                 n

                 3
0.1
                     0.01
                    0.001
                                                                        Rat
                                                                                                      Human
                                      ^
           Figure D-4. Developmental dental effects across species.

           The corresponding data are in Table D-6. The numbers following the effect designations indicate the corresponding study in Table D-6. Vertical solid

           black lines indicate the range of exposures tested below the LOAEL.

-------
  U)
<]> O
is S
oc -u
C t:
.2 _§
t? "(5
o> _>
S !T
           100
            10
0.1
           0.01
          0.001
                                                      Immune System Effects
                                                                                                 •••LOAEL

                                                                                                 ANOAEL/BMDL









<










+










t

>


L


i















t 4


j|

A











4


i



t
















>


t





























T T t
'!
• '
A A






















L j
















11






* ^
















^ j






^
















k j




















k 4


i




















k




































| •



A ^ A A

A



              r*S>   X!
           X ^



                                              ^



 Figure D-5. Immune system effects across species.
 The corresponding data are in Table D-7. The numbers following the effect designations indicate the corresponding study in Table D-7. Vertical solid
 black lines indicate the range of exposures tested below the LOAEL.

-------

-i nn





10


T3
ra
-SS
-==,
O)
m di 1
£ w '
(C 0

£ §
O)'5
= 0- 01
¥
(0
E
3
0.01 -






/
 ^> ^
/
^
vS" v\^ v»^ vv^
t<=P ^ ^ ^
^ X y ^
* LOAEL
ANOAEL/BMDL
Monkey
























*





^
/
/
Figure D-6. Neurological effects across species.
The corresponding data are in Table D-8. The numbers following the effect designations indicate the corresponding study in Table D-8. Vertical solid
black lines indicate the range of exposures tested below the LOAEL.

-------
D.4. REFERENCES

Abbott BD: Birnbaum. LS: Pratt RM. (1987a). TCDD-induced hyperplasia of the ureteral
      epithelium produces hydronephrosis in murine fetuses. Teratology 35: 329-334.
      http://dx.doi.org/10.1002/tera.1420350307.
Abbott BD: Morgan. KS: Birnbaum. LS: Pratt RM. (1987b). TCDD alters the extracellular
      matrix and basal lamina of the fetal mouse kidney. Teratology 35: 335-344.
      http://dx.doi.org/10.1002/tera.1420350308.
Abbott BD: Birnbaum, LS. (1989). TCDD alters medial epithelial cell  differentiation during
      palatogenesis. Toxicol Appl Pharmacol 99: 276-286.
Abbott BD: Birnbaum. LS. (1990). Effects of TCDD on embryonic ureteric epithelial EOF
      receptor expression and cell proliferation. Teratology 41: 71-84.
      http://dx.doi.org/10.1002/tera.1420410108.
Abbott BD: Probst MR.  (1995). Developmental expression of two members of a new class of
      transcription factors: II. Expression of aryl hydrocarbon receptor nuclear translocator in
      the C57BL/6N mouse embryo. Dev Dyn 204: 144-155.
      http://dx.doi.org/10.1002/aja.1002040205.
Abbott BD: Held. GA: Wood. CR: Buckalew. AR: Brown. JG: Schmid. J. (1999a). AhR,
      ARNT, and CYP1A1 mRNA quantitation in cultured human embryonic palates exposed
      to TCDD and comparison with mouse palate in vivo and in culture. Toxicol  Sci 47: 62-
      75.
Abbott BD: Schmid. JE:  Brown. JG: Wood. CR: White. RD: Buckalew. AR: Held. GA.
      (1999b). RT-PCR quantification of AHR, ARNT, GR, and CYP1A1 mRNA in
      craniofacial tissues of embryonic mice exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin
      and hydrocortisone. Toxicol Sci 47: 76-85.
Abbott BD: Buckalew. AR: DeVito. MJ: Ross. D: Bryant PL: Schmid. JE. (2003). EGF and
      TGF-alpha expression influence the developmental toxicity of TCDD: dose  response and
      AhR phenotype in EGF,  TGF-alpha, and EGF + TGF-alpha knockout mice.  Toxicol Sci
      71: 84-95.
Abernethy. DJ: Greenlee. WF: Huband. JC: Boreiko. CJ.  (1985). 2,3,7,8-Tetrachlorodibenzo-p-
      dioxin (TCDD) promotes the transformation of C3H/10T1/2 cells. Carcinogenesis 6: 651-
      653.
Abraham, K; Krowke, R: Neubert, D. (1988). Pharmacokinetics  and biological activity of
      2,3,7,8-tetrachlorodibenzo-p-dioxin. 1. Dose-dependent tissue distribution and induction
      of hepatic ethoxyresorufm O-deethylase in rats following a single injection.  Arch Toxicol
      62: 359-368.
Ackermann, MF;  Gasiewicz, TA; Lamm, KR; Germolec, PR: Luster, MI. (1989). Selective
      inhibition of polymorphonuclear neutrophil  activity by 2,3,7,8-tetrachlorodibenzo-p-
      dioxin. Toxicol Appl Pharmacol 101: 470-480.
Adamsson, A; Simanainen, U: Viluksela, M; Paranko, J: Toppari, J. (2008). The effects of 2, 3,
      7, 8-tetrachlorodibenzo-p-dioxin on foetal male rat steroidogenesis. Int J Androl 32: 575-
      585.
Agrawal, AK; Tilson, HA; Bondy, SC. (1981). 3,4,3',4'-Tetrachlorobiphenyl given to mice
      prenatally produces long-term decreases in striatal dopamine and receptor binding sites in
      the caudate nucleus. Toxicol Lett 7: 417-424.
                                        D-161

-------
Aitio. A: Parkki. MG: Marniemi. J. (1979). Different effect of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin on glucuronide conjugation of various aglycones. Studies in Wistar and Gunn rats.
       Toxicol Appl Pharmacol 47: 55-60.
Alaluusua, S: Calderara, P; Gerthoux, PM; Lukinmaa, PL; Kovero, O: Needham, L; Patterson, J,
       r, D. G.: Tuomisto, J: Mocarelli, P. (2004). Developmental dental aberrations after the
       dioxin accident in  Seveso. Environ Health Perspect 112: 1313-1318.
Albro. PW: Corbett JT: Harris. M: Lawson. LD. (1978). Effects of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin on lipid profiles in tissue of the Fischer rat. Chem Biol Interact 23: 315-330.
       http://dx.doi.org/10.1016/0009-2797(78)90093-5.
Allen, DE; Leamy, LJ. (2001). 2,3,7,8-tetrachlorodibenzo-p-dioxin affects size and shape, but
       not asymmetry, of mandibles in mice. Ecotoxicology 10: 167-176.
       http://dx.doi.Org/10.1023/A:1016693911300.
Allen, JR; Carstens, LA. (1967). Light and electron microscopic observations in Macaca mulatta
       monkeys fed toxic fat. Am J Vet Res 28: 1513-1526.
Alsharif. NZ: Grandjean. CJ: Murray. WJ: Stohs. SJ. (1990). 2,3,7,8-Tetrachlorodibenzo-p-
       dioxin (TCDD)-induced decrease in the fluidity of rat liver membranes. Xenobiotica 20:
       979-988. http://dx.doi.org/10.3109/00498259009046913.
Alsharif. NZ: Hassoun.  E: Bagchl M: Lawson. T:  Stohs. SJ. (1994a). The effects of anti-TNF-
       alpha antibody and dexamethasone on TCDD-induced oxidative stress in mice.
       Pharmacology 48:  127-136.
Alsharif. NZ: Lawson. T:  Stohs. SJ. (1994b). Oxidative  stress induced by 2,3,7,8-
       tetrachlorodibenzo-p-dioxin is mediated by the aryl hydrocarbon (Ah) receptor complex.
       Toxicology 92: 39-51.
Alsharif. NZ: Schlueter. WJ: Stohs. SJ. (1994c). Stimulation of NADPH-dependent reactive
       oxygen species formation and DNA damage by 2,3,7,8-tetrachlorodibenzo-p-dioxin in rat
       peritoneal lavage cells. Arch Environ Contam Toxicol 26: 392-397.
Alsharif, NZ: Hassoun,  EA. (2004). Protective effects of vitamin A and vitamin E succinate
       against 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)-induced body wasting,
       hepatomegaly, thymic atrophy, production  of reactive oxygen species and DNA damage
       in C57BL/6J mice. Basic Clin Pharmacol Toxicol 95: 131-138.
       http://dx.doi.0rg/10.llll/i.1742-7843.2004.950305.x.
Altmann, L; Weinand-Haerer, A; Lilienthal, H; Wiegand, H. (1995). Maternal exposure to
       polychlorinated biphenyls inhibits long-term potentiation in the visual cortex of adult
       rats. Neurosci Lett 202:  53-56. http://dx.doi.org/10.1016/0304-3940(95)12197-8.
Altmann, L: Lilienthal, H: Hany, J: Wiegand, H. (1998). Inhibition of long-term potentiation in
       developing rat visual cortex but not hippocampus by in utero exposure to polychlorinated
       biphenyls. Dev Brain Res  110: 257-260. http://dx.doi.org/10.1016/S0165-
       3806(98)00112-6.
Amin, S: Moore, RW: Peterson, RE; Schantz, SL.  (2000).  Gestational and lactational exposure to
       TCDD or coplanar PCBs alters adult expression  of saccharin preference behavior in
       female rats. Neurotoxicol Teratol 22: 675-682. http://dx.doi.org/10.1016/S0892-
       0362(00)00094-5.
Andersson, P; McGuire, J: Rubio, C: Gardin, K; Whitelaw, ML; Pettersson, S: Hanberg, A;
       Poellinger, L. (2002). A constitutively active dioxin/aryl hydrocarbon receptor induces
       stomach tumors. PNAS  99: 9990-9995.  http://dx.doi.org/10.1073/pnas.152706299.
                                         D-162

-------
Aoa, K; Suzukia, T; Murai, H; Matsumotoa, M; Nagai, H; Miyamotoa, Y; Tohyatnab, C:
       Noharaa, K. (2009). Comparison of immunotoxicity among tetrachloro-, pentachloro-,
       tetrabromo- and pentabromo-dibenzo-p-dioxins in mice. Toxicology 256: 25-31.
       http://dx.doi.0rg/10.1016/i.tox.2008.10.024.
Aragon. AC: Kopf, PG: Campen. MJ: Huwe. JK: Walker. MK. (2008a). In utero and lactational
       2,3,7,8-tetrachlorodibenzo-p-dioxin exposure: Effects on fetal and adult cardiac gene
       expression and adult cardiac and renal morphology. Toxicol Sci 101: 321-330.
       http://dx.doi.org/10.1093/toxsci/kfm272.
Aragon. AC: Goens. MB: Carbett E: Walker. MK. (2008b). Perinatal 2,3,7,8-
       tetrachlorodibenzo-p-dioxin exposure sensitizes offspring to angiotensin II-induced
       hypertension. Cardiovasc Toxicol 8: 145-154. http://dx.doi.org/10.1007/sl2012-008-
       9023-1.
Ashida, H: Enan, E; Matsumura, F. (1996). Protective action of dehydroascorbic acid on the Ah
       receptor-dependent and receptor-independent induction of lipid peroxidation in adipose
       tissue of male guinea pig caused by TCDD administration.  J Biochem Toxicol 11: 269-
       278. http://dx.doi.org/10.1002/(SICm522-7146O996)1 l:6<:269::AID-
       JBT2>:3.0.CO:2-I.
Ashida. H: Nagy. S: Matsumura. F. (2000). 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)-
       induced  changes in activities of nuclear protein kinases and phosphatases affecting DNA
       binding activity of c-myc and AP-1 in the livers of guinea pigs.  Biochem Pharmacol 59:
       741-751.
Astroff, B: Zacharewski. T:  Safe. S: Arlotto. MP: Parkinson. A: Thomas. P: Levin. W. (1987). 6-
       methyl-i ,3,8-trichlorodibenzofuran as a 2,3,7,8- tetrachlorodibenzo-p-dioxin antagonist:
       Inhibition of the induction of rat cytochrome P-450 isozymes and related monooxygenase
       activities. Mol Pharmacol 33: 231-236.
Aubeit ML: Begeot M: Winiger. BP: Morel G: Sizonenko. PC: Dubois. PM. (1985). Ontogeny
       of hypothalamic luteinizing hormone-releasing hormone (GnRH) and pituitary GnRH
       receptors in fetal and neonatal rats.  Endocrinology 116: 1565-1576.
Aulerich. RJ: Yamini. B: Bursian. SJ. (2001). Dietary exposure to  3,3',4,4',5-
       pentachlorobiphenyl (PCB 126) or 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)  does not
       induce proliferation of squamous epithelium or osteolysis in the jaws of weanling rats.
       Vet Hum Toxicol 43: 170-171.
Baccarelli, A; Mocarelli, P: Patterson, DG, Jr: Bonzini, M: Pesatori, AC: Caporaso, N; Landi,
       MT. (2002). Immunologic effects of dioxin: new results from Seveso and comparison
       with other studies. Environ Health Perspect 110:  1169-1173.
Baccarelli, A; Pesatori, AC: Masten,  SA; Patterson, DG, Jr: Needham,  LL; Mocarelli, P:
       Caporaso, NE;  Consonni, D; Grassman, JA; Bertazzi, PA; MT,  L. (2004). Aryl-
       hydrocarbon receptor-dependent pathway and toxic effects of TCDD in humans: a
       population-based study in Seveso, Italy. Toxicol Lett 149: 287-293.
       http://dx.doi.0rg/10.1016/i.toxlet.2003.12.062.
Baccarelli, A; Giacomini, SM; Corbetta, C: Landi, MT: Bonzini, M: Consonni, D; Grillo, P:
       Patterson, DG:  Pesatori, AC: Bertazzi, PA. (2008). Neonatal thyroid function in  Seveso
       25 years after maternal exposure to dioxin. PLoS Med 5: e!61.
Badawi, AF; Cavalieri, EL; Rogan, EG. (2000). Effect of chlorinated hydrocarbons on
       expression of cytochrome P450 1A1,  1A2 and  1B1 and 2- and 4-hydroxylation of 17beta-
       estradiol in female Sprague-Dawley rats. Carcinogenesis 21: 1593-1599.

                                         D-163

-------
Badesha, JS: Maliji, G: Flaks, B. (1995). Immunotoxic effects of prolonged dietary exposure of
       male rats to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Eur J Pharmacol 293: 429-437.
Bagchi, D: Shara, MA; Bagchi, M; Hassoun, EA; Stohs, SJ. (1993). Time-dependent effects of
       2,3,7,8-tetrachlorodibenzo-p-dioxin on serum and urine levels of malondialdehyde,
       formaldehyde, acetaldehyde, and acetone in rats. Toxicol Appl Pharmacol 123:  83-88.
       http://dx.doi.org/10.1006/taap.1993.1224.
Bagchi. D: Balmoori, J: Bagchi. M: Ye. X: Williams. CB: Stohs. SJ. (2002). Comparative effects
       of TCDD, endrin, naphthalene and chromium (VI) on oxidative stress and tissue damage
       in the liver and brain tissues of mice.  Toxicology 175: 73-82.
       http://dx.doi.org/10.1016/S0300-483X(02)00062-8.
Bars, RG: Elcombe, CR. (1991). Dose-dependent acinar induction of cytochromes P450 in rat
       liver. Evidence for a differential mechanism of induction of P450IA1 by beta-
       naphthoflavone and dioxin. Biochem  J 277: 577-580.
Barsotti, DA; Abrahamson, LJ: Allen, JR. (1979). Hormonal alterations in female rhesus
       monkeys fed a diet containing 2,3,7,8-tetrachlorodibenzo-p-dioxin.  Bull Environ Contam
       Toxicol 21: 463-469. http://dx.doi.org/10.1007/BF01685454.
Barter, RA; Klaassen, CD. (1992). UDP-glucuronosyltransferase inducers reduce thyroid
       hormone levels in rats by an extrathyroidal mechanism. Toxicol Appl Pharmacol 113: 36-
       42.
Bastomsky, CH. (1977). Enhanced thyroxine metabolism and high  uptake goiters in rats after a
       single dose of 2,3,7,8-tetrachlorodibenzo-p-dioxin. Endocrinology 101: 292-296.
       http://dx.doi.org/10.1210/endo-101-l-292.
Beckett KJ: Millsap, SD: Blankenship, AL: Zwiernik, MJ: Giesv, JP: Bursian, SJ. (2005).
       Squamous epithelial lesion of the mandibles and maxillae of wild mink (Mustela vison)
       naturally exposed to polychlorinated biphenyls. Environ Toxicol Chem 24: 674-677'.
       http://dx.doi.Org/10.1897/04-241R.l.
Beebe, LE: Anver. MR: Riggs,  CW: Fornwald, LW: LM, A. (1995). Promotion of N-
       nitrosodimethylamine-initiated mouse lung tumors following single or multiple low dose
       exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Carcinogenesis 16: 1345-1349.
Beguinot L: Hanover. JA: Ito, S: Richeit ND: Willingham, MC: Pastan, I. (1985). Phorbol
       esters induce transient internalization without degradation of unoccupied epidermal
       growth factor receptors. PNAS 82: 2774-2778.
Bell PR: Clode, S: Fan. MQ: Fernandes, A:  Foster. PM: Jiang.  T: Loizou,  G:  MacNicoll A:
       Miller,  BG: Rose, M:  Tran, L; White, S.  (2007a). Relationships between tissue levels of
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), mRNAs and toxicity in the developing
       male Wistar(Han) rat. Toxicol Sci 99: 591-604. http://dx.doi.org/10.1093/toxsci/kfml79.
Bell PR: Clode. S: Fan. MQ: Fernandes. A:  Foster. PM: Jiang.  T: Loizou.  G:  MacNicoll A:
       Miller.  BG: Rose. M:  Tran. L: White. S.  (2007b). Toxicity of 2,3,7,8-tetrachlorodibenzo-
       p-dioxin in the developing male Wistar(Han) rat. I: No decrease in epididymal sperm
       count after a single acute dose. Toxicol Sci 99: 214-223.
       http://dx.doi.org/10.1093/toxsci/kfml40.
Bell PR: Clode. S: Fan. MQ: Fernandes. A:  Foster. PM: Jiang.  T: Loizou.  G:  MacNicoll A:
       Miller.  BG: Rose. M:  Tran. L: White. S.  (2007c). Toxicity of 2,3,7,8-tetrachlorodibenzo-
       p-dioxin in the developing male Wistar(Han) rat. II: Chronic dosing causes
       developmental delay. Toxicol Sci  99: 224-233. http://dx.doi.org/10.1093/toxsci/kfml41.
                                         D-164

-------
Betnis. JC: Alejandro. NF: Nazarenko. DA: Brooks. AI: Baggs. RB: Gasiewicz. TA. (2007).
       TCDD-induced alterations in gene expression profiles of the developing mouse paw do
       not influence morphological differentiation of this potential target tissue. Toxicol Sci 95:
       240-248. http://dx.doi.org/10.1093/toxsci/kfll32.
Besteman, EG: Zimmerman, KL: Holladay, SD. (2005). Tetrachlorodibenzo-p-Dioxin (TCDD)
       inhibits differentiation and increases apoptotic cell death of precursor T-Cells in the fetal
       mouse thymus. J Immunotoxicol 2: 107-114.
       http://dx.doi.org/10.1080/15476910500182541.
Besteman. EG: Zimmerman. KL: Ruckle. WR: Prater. MR: Gogal RM. Jr: Holladav. SD.
       (2007). 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) or diethylstilbestrol (DBS) cause
       similar hematopoietic hypocellularity and hepatocellular changes in murine fetal liver,
       but differentially affect gene expression. Toxicol Pathol 35: 786-792.
       http://dx.doi.org/10.1080/01926230701584155.
Biegel L: Harris. M: Davis. D: Rosengren. R: Safe. L: Safe. S. (1989). 2,2',4,4',5,5'-
       hexachlorobiphenyl as a 2,3,7,8-tetrachlorodibenzo-p-dioxin antagonist in C57BL/6J
       mice. Toxicol Appl Pharmacol 97: 561-571.
Birnbaum. LS: Weber. H: Harris. MW: Lamb. JC:  McKinnev. JD. (1985). Toxic interaction of
       specific polychlorinated biphenyls and 2,3,7,8-tetrachlorodibenzo-p-dioxin: increased
       incidence of cleft palate in mice. Toxicol Appl Pharmacol 77: 292-302.
Birnbaum. LS: Harris. MW: Miller. CP: Pratt RM: Lamb. JC. (1986). Synergistic interaction of
       2,3,7,8,-tetrachlorodibenzo-p-dioxin and hydrocortisone in the induction of cleft palate in
       mice. Teratology  33: 29-35. http://dx.doi.org/10.1002/tera.1420330106.
Birnbaum, LS: Harris, MW: Crawford, DP: Morrissey, RE. (1987a). Teratogenic effects of
       polychlorinated dibenzofurans in combination in C57BL/6N mice. Toxicol Appl
       Pharmacol 91: 246-255.
Birnbaum, LS: Harris, MW: Barnhart, ER; Morrissey, RE. (1987b). Teratogenicity of three
       polychlorinated dibenzofurans in C57BL/6N mice. Toxicol Appl Pharmacol 90: 206-216.
Birnbaum, LS: Harris, MW: Stocking, LM; Clark,  AM; Morrissey, RE. (1989). Retinoic acid and
       2,3,7,8-tetrachlorodibenzo-p-dioxin selectively enhance teratogenesis in C57BL/6N mice.
       Toxicol Appl Pharmacol 98: 487-500.
Birnbaum. LS: McDonald. MM: Blair. PC: Clark. AM: Harris. MW. (1990). Differential toxicity
       of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in C57BL/6J mice congenic at the Ah
       Locus. Fundam Appl Toxicol 15: 186-200.
Birnbaum. LS: Morrissev. RE: Harris. MW. (1991). Teratogenic effects of 2,3,7,8-
       tetrabromodibenzo-p-dioxin and three polybrominated dibenzofurans in C57BL/6N mice.
       Toxicol Appl Pharmacol 107: 141-152.
Bjerke, PL; Sommer, RJ: Moore, RW: Peterson, RE. (1994a). Effects of in utero and lactational
       2,3,7,8-tetrachlorodibenzo-p-dioxin exposure on responsiveness of the male rat
       reproductive system to testosterone stimulation in adulthood. Toxicol Appl Pharmacol
       127: 250-257. http://dx.doi.org/10.1006/taap.1994.1159.
Bjerke. PL: Brown. TJ: MacLusky. NJ: Hochberg. RB: Peterson. RE. (1994b). Partial
       demasculinization and feminization of sex behavior in male rats by in utero and
       lactational exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin is not associated with
       alterations in estrogen receptor binding or volumes of sexually differentiated brain nuclei.
       Toxicol Appl Pharmacol 127: 258-267.
                                         D-165

-------
Bjerke, PL; Peterson, RE. (1994). Reproductive toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       in male rats: different effects of in utero versus lactational exposure. Toxicol Appl
       Pharmacol 127: 241-249. http://dx.doi.org/10.1006/taap.1994.1158.
Blavlock. BL: Holladav. SD: Comment CE: Heindel JJ: Luster. MI. (1992). Exposure to
       tetrachlorodibenzo-p-dioxin (TCDD) alters fetal thymocyte maturation. Toxicol Appl
       Pharmacol 112: 207-213.
Bonn, AA; Harrod, KS: Teske, S: Lawrence, BP. (2005). Increased mortality associated with
       TCDD exposure in mice infected with influenza A virus is not due to severity of lung
       injury or alterations in Clara cell protein content. Chem Biol Interact 155: 181-190.
       http://dx.doi.0rg/10.1016/i.cbi.2005.06.004.
Boverhof, PR; Burgoon, LD: Williams, KJ: Zacharewski, TR. (2008). Inhibition of estrogen-
       mediated uterine gene expression responses by dioxin. Mol Pharmacol 73: 82-93.
       http://dx.doi.org/10.1124/mol.107.040451.
Boverhoff PR; Burgoon, LD: Tashiro, C: Chittim, B; Harkema, JR: Jump, DB; Zacharewski,
       TR. (2005). Temporal and dose-dependent  hepatic gene expression patterns in mice
       provide new insights into TCDD-mediated  hepatotoxicity. Toxicol Sci 85: 1048-1063.
       http://dx.doi.org/10.1093/toxsci/kfil62.
Bowers. OJ:  Sommersted, KB: Sowell RT:  Boling, GE: Hanneman, WH: Titus. RG: Dekrev,
       GK. (2006). 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) reduces Leishmania major
       burdens in C57BL/6 mice. Am J Trop Med Hyg 75: 749-752.
Bowman, RE; Schantz, SL: Gross, ML; SA, F.  (1989a). Behavioral effects in monkeys exposed
       to 2,3,7,8-TCDD transmitted maternally during gestation and for four months of nursing.
       Chemosphere 18: 235-242.
Bowman. RE: Schantz. SL: Weerasinghe, NCA: Gross. ML: Barsotti, DA. Q989b). Chronic
       dietary intake of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) at 5 or 25 parts per trillion
       in the monkey: TCDD kinetics and dose-effect estimate of reproductive toxicity.
       Chemosphere 18: 243-252.
Brewster, DW: Matsumura, F. (1984). TCDD (2,3,7,8-tetrachlorodibenzo-p-dioxin) reduces
       lipoprotein lipase activity in the adipose tissue of the guinea pig. Biochem Biophys Res
       Commun 122: 810-817.
Brewster, DW: Matsumura, F: Akera, T. (1987). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       on guinea pig heart muscle. Toxicol  Appl Pharmacol 89: 408-417.
Brouillette, J: Quirion, R. (2008).  The common environmental pollutant dioxin-induced memory
       deficits by altering estrogen pathways and a major route of retinol transport involving
       transthyretin. Neurotoxicology 29: 318-327.
       http://dx.doi.0rg/10.1016/i.neuro.2007.12.005.
Brouwer, A; van den berg, K. (1983). Early  decrease in retinoid levels in mice after exposure to
       low doses of polychlorinated biphenyls. Chemosphere 12: 555-557.
       http://dx.doi.org/10.1016/0045-6535(83)90209-6.
Brouwer, A; van den Berg, KJ. (1984). Early and differential decrease in natural retinoid levels
       in C57BL/Rij and DBA/2 mice by 3,4,3',4'-tetrachlorobiphenyl. Toxicol Appl Pharmacol
       73:204-209.
Brouwer, A; van den Berg, KJ: Kukler, A. (1985).  Time and dose responses of the reduction in
       retinoid concentrations in C57BL/Rij and DBA/2 mice induced by 3,4,3',4'-
       tetrachlorobiphenyl. Toxicol Appl Pharmacol 78: 180-189.
                                        D-166

-------
Brown, NM: Lamartiniere, CA. (1995). Xenoestrogens alter mammary gland differentiation and
       cell proliferation in the rat. Environ Health Perspect 103: 708-713.
Brunnberg, S: Andersson, P; Lindstam, M; Paulson, I; Poellinger, L; Hanberg, A. (2006). The
       constitutively active Ah receptor (CA-Ahr) mouse as a potential model for dioxin
       exposure—effects in vital organs. Toxicology 224: 191-201.
       http://dx.doi.0rg/10.1016/i.tox.2006.04.045.
Bryant. PL: Clark. GC: Probst. MR: Abbott. BD. (1997). Effects of TCDD on Ah receptor,
       ARNT, EGF, and TGF-alpha expression in embryonic mouse urinary tract. Teratology
       55: 326-337. http://dx.doi.org/10.1002/(SICB1096-9926(199705)55:5<:326::AID-
       TERA5>:3.0.CO:2-X.
Bryant PL: Schmid. JE: Fenton. SE: Buckalew. AR: Abbott BD. (2001). Teratogenicity of
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in mice lacking the expression of EGF
       and/or TGF-alpha. Toxicol Sci 62: 103-114.
Buchmann, A; Stinchcombe, S: Korner, W; Hagenmaier, H; Bock, KW. (1994). Effects of
       2,3,7,8-tetrachloro- and 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin on the proliferation of
       preneoplastic liver cells in the rat. Carcinogenesis 15:1143-1150.
Burleson. GR: Lebrec. H: Yang. YG: Ibanes. JD: Pennington. KN: Birnbaum. LS. (1996). Effect
       of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on influenza virus host resistance in
       mice. Fundam Appl Toxicol 29: 40-47.
Bushnell, PJ: Rice, DC. (1999). Behavioral assessments of learning and attention in rats exposed
       perinatally to 3,3',4,4',5-pentachlorobiphenyl (PCB 126). Neurotoxicol Teratol 21: 381-
       392. http://dx.doi.org/10.1016/S0892-0362(99)00006-9.
Byers, JP: Masters, K; Sarver, JG: Hassoun, EA. (2006). Association between the levels of
       biogenic amines and superoxide anion production in brain regions of rats after subchronic
       exposure to TCDD. Toxicology 228: 291-298.
       http://dx.doi.0rg/10.1016/i.tox.2006.09.009.
Calfee-Mason. KG: Spear. BT:  Glauert HP. (2002). Vitamin E inhibits hepatic NF-kappaB
       activation in rats administered the hepatic tumor promoter, phenobarbital. J Nutr 132:
       3178-3185.
Camacho. IA: Nagarkattl M: Nagarkattl PS. (2004). Effect of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin (TCDD) on maternal immune response during pregnancy. Arch Toxicol 78: 290-
       300. http://dx.doi.org/10.1007/s00204-003-0538-8.
Cantoni L; Salmona, M: Rizzardini, M. (1981). Porphyrogenic effect of chronic treatment with
       2,3,7,8-tetrachlorodibenzo-p-dioxin in female rats. Dose-effect relationship following
       urinary excretion of porphyrins. Toxicol Appl Pharmacol  57: 156-163.
Cantoni, L: dal Fiume, D: Ferraroli, A; Salmona, M: Ruggieri, R. (1984). Different susceptibility
       of mouse tissues to porphyrogenic effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol
       Lett 20: 201-210.
Carney, SA; Peterson, RE; Heideman, W. (2004). 2,3,7,8-Tetrachlorodibenzo-p-dioxin activation
       of the aryl hydrocarbon receptor/aryl hydrocarbon receptor nuclear translocator pathway
       causes developmental toxicity through a CYP1 A-independent mechanism in zebrafish.
       Mol Pharmacol 66: 512-521. http://dx.doi.Org/10.1124/mol.66.3.66/3/512.
Chaffin, CL; Peterson, RE; Hutz, RJ. (1996). In utero and lactational exposure of female
       Holtzman rats to 2,3,7,8-tetrachlorodibenzo-p-dioxin: modulation of the estrogen signal.
       BiolReprod55: 62-67.
                                         D-167

-------
Chaffin. CL: Trewin. AL: Watanabe. G: Tava. K: Hutz. RJ. (1997). Alterations to the pituitary-
       gonadal axis in the peripubertal female rat exposed in utero and through lactation to
       2,3,7,8-tetrachlorodibenzo-p-dioxin. Biol Reprod 56:  1498-1502.
Chahoud, I; Krowke, R: Schimmel A; Merker, HJ: Neubert, D. (1989). Reproductive toxicity
       and pharmacokinetics of 2,3,7,8-tetrachlorodibenzo-p-dioxin. 1. Effects of high doses on
       the fertility of male rats. Arch Toxicol 63: 432-439.
Chapman, DE; Schiller, CM. (1985). Dose-related effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       (TCDD) in C57BL/6J and DBA/2J mice. Toxicol Appl Pharmacol 78: 147-157.
Chen, CY; Hamm, JT: Hass, JR: Birnbaum, LS. (2001). Disposition of polychlorinated dibenzo-
       p-dioxins, dibenzofurans, and non-ortho polychlorinated biphenyls in pregnant long
       evans rats and the transfer to offspring. Toxicol Appl Pharmacol 173: 65-88.
       http://dx.doi.org/10.1006/taap.2001.9143.
Chen. CY: Hamm. JT: Hass. JR: Albro. PW: Birnbaum. LS. (2002). A mixture of
       polychlorinated dibenzo-p-dioxins (PCDDs), dibenzofurans (PCDFs), and non-ortho
       polychlorinated biphenyls (PCBs) changed the lipid content of pregnant Long Evans rats.
       Chemosphere 46: 1501-1504.
Chen. J: Laughlin. LS: Hendrickx. AG: Natarajan. K: Overstreet JW: Laslev. BL. (2003). The
       effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on chorionic gonadotrophin
       activity in pregnant macaques. Toxicology 186: 21-31.
Chen. SW: Roman. BL: Sarova. SZ: Shinoniya. K: Moore. RW: Peterson. RE. (1993). In utero
       exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) does not impair testosterone
       production by fetal rat testes. Toxicologist 13:  104.
Cheng, SB; Kuchiiwa, S; Nagatomo, I; Akasaki, Y; Uchida, M; Tominaga, M; Hashiguchi, W;
       Kuchiiwa, T; Nakagawa, S. (2002). 2,3,7,8-Tetrachlorodibenzo-p-dioxin treatment
       induces c-Fos expression in the forebrain of the Long-Evans rat. Brain Res 931: 176-180.
Cho, HJ: Hahn, EJ: Hwang. JA: Hong. MS: Kim. SK:  Pak, HR: Park. JH. (2006). Enhanced
       expression of plasma glutathione peroxidase in the thymus of mice treated with TCDD
       and its implication for TCDD-induced thymic atrophy. Molecules and Cells 21: 276-283.
Choi. JS: Kim. IW: Hwang. SY: Shin. BJ: Kim. SK. (2008). Effect of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin on testicular spermatogenesis-related panels and serum sex
       hormone levels in rats. BJU Int 101: 250-255. http://dx.doi.Org/10.l 11 l/j.1464-
       410X.2007.07202.X.
Choi, SS: Miller, MA; Harper, PA. (2006). In utero exposure to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin induces amphiregulin gene expression in the developing mouse ureter. Toxicol Sci
       94: 163-174. http://dx.doi.org/10.1093/toxsci/kfl090.
Chou, SM; Miike, T; Payne, WM: Davis,  GJ. (1979). Neuropathology of "spinning
       syndrome" induced by prenatal intoxication with a PCB in mice. Ann N Y Acad
       Sci 320: 373-395.
Chu. I: Lecavalier. P: Hakansson. H: Yagminas. A: Valli. VE: Poon. P: M. F. (2001). Mixture
       effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin and polychlorinated biphenyl congeners in
       rats. Chemosphere 43: 807-814. http://dx.doi.org/10.1016/S0045-6535(00)00437-9.
Chu, I; Valli, VE: Rousseaux, CG. (2007). Combined effects of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin and polychlorinated biphenyl congeners in rats. Toxicol Environ Chem 89: 71-87.
       http://dx.doi.org/10.1080/02772240600942548.
                                        D-168

-------
Clark, DA; Gauldie, J: Szewczuk, MR; Sweeney, G. (1981). Enhanced suppressor cell activity as
       a mechanism of immunosuppression by 2,3,7,8-tetrachlorodibenzo-p-dioxin. Proc Soc
       Exp Biol Med 168: 290-299.
Clark, GC: Taylor, MJ: Tritscher, AM; Lucier, GW. (199la). Tumor necrosis factor involvement
       in 2,3,7,8-tetrachlorodibenzo-p-dioxin-mediated endotoxin hypersensitivity in C57BL/6J
       mice congenic at the Ah locus. Toxicol Appl Pharmacol 111: 422-431.
Clark. GC: Tritscher. A: Maronpot R: Folev. J: Lucier. G. (1991b). Tumor promotion by TCDD
       in female rats. In Banbury Report 35: Biological basis for risk assessment of dioxin and
       related compounds. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.
Cohen. GM: Bracken. WM: Iyer. RP: Berry. PL: Selkirk. JK: Slaga. TJ. (1979).
       Anticarcinogenic effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin on benzo(a)pyrene and
       7,12-dimethylbenz(a)anthracene tumor initiation and its relationship to DNA binding.
       Cancer Res 39: 4027-4033.
Collins, LL;  Williamson, MA; Thompson, BD; Dever, DP; Gasiewicz, TA; Opanashuk, LA.
       (2008). 2,3,7,8-Tetracholorodibenzo-p-dioxin exposure disrupts granule neuron precursor
       maturation in the developing mouse cerebellum. Toxicol Sci 103: 125-136.
       http://dx.doi.org/10.1093/toxsci/kfn017.
Collins, WT, Jr; Capen, CC. (1980). Fine structural lesions and hormonal alterations in thyroid
       glands of perinatal rats exposed in utero and by the milk to polychlorinated biphenyls.
       Am JPathol99: 125-142.
Comer, CP: Norton, S. (1982). Effects of perinatal methimazole exposure on a developmental
       test battery for neurobehavioral toxicity in rats. Toxicol Appl Pharmacol 63: 133-141.
Courtney,  KD: Moore, JA. (1971). Teratology studies with 2,4,5-trichlorophenoxyacetic acid and
       2,3,7,8-tetrachlorodibenzo-p-dioxin.  Toxicol Appl Pharmacol 20: 396-403.
Courtney,  KD. (1976). Mouse teratology studies with chlorodibenzo-p-dioxins. Bull Environ
       Contam Toxicol 16: 674-681.
Couture. LA: Harris. MW: Birnbaum. LS. (1989). Developmental toxicity of 2,3,4,7,8-
       pentachlorodibenzofuran in the Fischer 344 rat. Fundam Appl Toxicol 12: 358-366.
Couture, LA; Harris, MW: Birnbaum, LS. (1990). Characterization of the peak period of
       sensitivity for the induction of hydronephrosis in C57BL/6N mice following exposure to
       2,3,7, 8-tetrachlorodibenzo-p-dioxin. Fundam Appl Toxicol  15: 142-150.
Crofton, KM; Rice, DC. (1999). Low-frequency hearing loss following perinatal exposure to
       3,3',4,4',5-pentachlorobiphenyl (PCB 126) in rats. Neurotoxicol Teratol 21: 299-301.
Crofton, KM; Craft ES: Hedge, JM: Gennings, C: Simmons, JE: Carchman, RA: Carter, WH, Jr:
       DeVito, MJ. (2005). Thyroid-hormone-disrupting chemicals: Evidence for dose-
       dependent additivity or synergism. Environ Health Perspect 113: 1549-1554.
Croutch, CR: Lebofsky, M: Schramm, KW:  Terranova, PF: Rozman, KK.  (2005). 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin (TCDD) and 1,2,3,4,7,8-hexachlorodibenzo-p-dioxin
       (HxCDD) alter body weight by decreasing insulin-like growth factor I (IGF-I) signaling.
       Toxicol Sci  85: 560-571. http://dx.doi.org/10.1093/toxsci/kfil06.
Cummings, AM; Metcalf, JL: Birnbaum, L.  (1996). Promotion of endometriosis by 2,3,7,8-
       tetrachlorodibenzo-p-dioxin in rats and mice: Time-dose dependence and species
       comparison. Toxicol Appl Pharmacol 138: 131-139.
       http://dx.doi.org/10.1006/taap.1996.0106.
                                        D-169

-------
D'argy, R; Hassoun, E; Dencker, L. (1984). Teratogenicity of tcdd and the congener 3,3',4,4'-
       tetrachloroazoxybenzene in sensitive and non-sensitive mouse strains after reciprocal
       blastocyst transfer. Toxicol Lett 21: 197-202. http://dx.doi.org/10.1016/0378-
       4274(84)90206-6.
Dalton. TP: Kerzee. JK: Wang. B: Miller. M: Dieter. MZ: Lorenz. JN: Shertzer. HG: Nerbeit
       DW: Puga, A. (2001). Dioxin exposure is an environmental risk factor for ischemic heart
       disease. Cardiovasc Toxicol 1: 285-298.
Davies, R: Clothier, B; Robinson, SW: Edwards, RE; Greaves, P; Luo, J: Gant, TW: Chernova,
       T; Smith, AG. (2008). Essential role of the AH receptor in the dysfunction of heme
       metabolism induced by 2,3,7,8-tetrachlorodibenzo-p-dioxin. Chem Res Toxicol 21: 330-
       340. http://dx.doi.org/10.1021/tx700176r.
Davis. BJ: McCurdv. EA: Miller. BD: Lucier. GW: Tritscher. AM. (2000). Ovarian tumors in
       rats induced by chronic 2,3,7,8-tetrachlorodibenzo-p-dioxin treatment. Cancer Res 60:
       5414-5419.
de Heer, C; Schuurman, HJ: Liem, AK; Penninks, AH; Vos, JG: van Loveren, H. (1995).
       Toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) to the human thymus after
       implantation in SCID mice. Toxicol Appl Pharmacol  134: 296-304.
Dearstyne, EA: Kerkvliet NI. (2002). Mechanism of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       (TCDD)-induced decrease in anti-CD3-activated CD4(+) T cells:  The roles of apoptosis,
       Fas, and TNF. Toxicology 170: 139-151.
DeCaprio. AP: McMartin. DN: O'Keefe. PW: Rej. R: Silkworm. JB: Kaminsky. LS. (1986).
       Subchronic  oral toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin in the guinea pig:
       Comparisons with a PCB-containing transformer fluid pyrolysate. Fundam Appl Toxicol
       6: 454-463.  http://dx.doi.Org/10.1093/toxsci/6.3.454.
Delia Porta, G: Dragani,  TA; Sozzi, G. (1987). Carcinogenic effects of infantile and long-term
       2,3,7,8-tetrachlorodibenzo-p-dioxin treatment in the mouse. Tumori 73: 99-107.
DeVito. MJ: Thomas. T: Martin. E: Umbreit TH: Gallo. MA. (1992). Antiestrogenic action of
       2,3,7,8-tetrachlorodibenzo-p-dioxin: Tissue-specific regulation of estrogen receptor in
       CD1  mice. Toxicol  Appl Pharmacol 113: 284-292.
DeVito, MJ: Ma, X; Babish, JG: Menache, M; Birnbaum, LS. (1994). Dose-response
       relationships in mice following subchronic exposure to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin: CYP1A1, CYP1A2, estrogen receptor, and protein tyrosine phosphorylation.
       Toxicol Appl Pharmacol  124: 82-90.
Dhar, JD; Setty, BS. (1990). Changes in testis, epididymis and other accessory organs of male
       rats treated with anandron during sexual maturation. Endocr Res 16: 231-239.
Dienhart MK: Sommer.  RJ: Peterson. RE: Hirshfield. AN: Silbergeld. EK. (2000). Gestational
       exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin induces developmental defects in the rat
       vagina. Toxicol Sci 56: 141-149.
Diliberto. JJ: Burgin. DE: Birnbaum. LS. (1999). Effects of CYP1A2 on Disposition of 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin, 2,3,4,7,8-Pentachlorodibenzofuran, and 2,2',4,4',5,5'-
       Hexachlorobiphenyl in CYP1A2 Knockout and Parental (C57BL/6N and 129/Sv) Strains
       of Mice. Toxicol  Appl Pharmacol 159: 52-64. http://dx.doi.org/10.1006/taap.1999.8720.
Dong, W; Teraoka, H; Yamazaki, K; Tsukiyama, S: Imani, S: Imagawa, T: Stegeman, JJ:
       Peterson, RE; Hiraga, T. (2002). 2,3,7,8-tetrachlorodibenzo-p-dioxin toxicity in the
       zebrafish embryo: Local circulation failure in the dorsal midbrain is associated with
       increased apoptosis. Toxicol Sci 69: 191-201.

                                         D-170

-------
Dong, W; Teraoka, H; Tsujimoto, Y; Stegeman, JJ: Hiraga, T. (2004). Role of aryl hydrocarbon
       receptor in mesencephalic circulation failure and apoptosis in zebrafish embryos exposed
       to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Sci 77: 109-116.
       http://dx.doi.org/10.1093/toxsci/kfh023kfh023.
Dragan. YP: Rizvl T: Xu. YH: Hully. JR: Bawa. N: Campbell HA: Maronpot RR: Pitot HC.
       (1991). An initiation-promotion assay in rat liver as a potential complement to the 2-year
       carcinogenesis bioassay. Fundam Appl Toxicol 16: 525-547.
Dragan. YP: Xu. XH: Goldsworthv. TL: Campbell HA: Maronpot RR: Pitot HC. (1992).
       Characterization of the promotion of altered hepatic foci by 2,3,7,8-tetrachlorodibenzo-p-
       dioxin in the female rat. Carcinogenesis 13: 1389-1395.
Dragin. N: Dalton. TP: Miller. ML: Shertzer. HG: Nebeit DW. (2006). For dioxin-induced birth
       defects, mouse or human CYP1A2 in maternal liver protects whereas mouse CYP1 Al
       and CYP1B1 are inconsequential. J Biol Chem 281: 18591-18600.
       http://dx.doi.org/10.1074/ibc.M601159200.
Dunlap. DY: Moreno-Aliaga. MJ: Wu. Z: Matsumura. F. (1999). Differential toxicities of TCDD
       in vivo among normal, c-src knockout, geldanamycin- and quercetin-treated mice.
       Toxicology 135: 95-107.
Dunlap, DY; Matsumura, F. (2000). Analysis of difference in vivo effects of TCDD between c-
       src +/+ mice, c-src deficient, -/+ and -/- B6, 129-Src(tm 1 sor) mice and their wild-type
       littermates. Chemosphere 40: 1241-1246.
Dunlap. DY: Ikeda. I: Nagashima. H: Vogel CF: Matsumura. F. (2002). Effects of src-
       deficiency on the expression of in vivo toxicity of TCDD in a strain of c-src knockout
       mice procured through six generations  of backcrossings to  C57BL/6 mice. Toxicology
       172: 125-141.
Eastin. WC: Haseman. JK: Mahler. JF: Bucher. JR. (1998). The National Toxicology Program
       evaluation of genetically  altered mice as predictive models for identifying carcinogens.
       Toxicol Pathol 26: 461-473.
Ebner, K; Brewster, DW: Matsumura, F. (1988). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       on serum insulin and glucose levels in the rabbit. J Environ Sci Health B 23: 427-438.
       http://dx.doi.org/10.1080/03601238809372617.
Eckle. VS: Buchmann. A: Bursch. W: Schulte-Hermann. R: Schwarz. M. (2004).
       Immunohistochemical detection of activated caspases in apoptotic hepatocytes in rat
       liver. Toxicol  Pathol 32: 9-15.
el-Sabeawy, F: Wang, S; Overstreet J: Miller, M; Lasley, B;  Enan, E. (1998). Treatment of rats
       during pubertal development with 2,3,7,8-tetrachlorodibenzo-p-dioxin alters both
       signaling kinase activities and epidermal growth factor receptor binding in the testis and
       the motility and acrosomal reaction of sperm. Toxicol Appl Pharmacol 150: 427-442.
       http://dx.doi.org/10.1006/taap.1998.8426.
El-Tawil, OS: Elsaieed, EM. (2005). Induction of oxidative stress in the reproductive system of
       rats after subchronic exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Bull Environ
       Contam Toxicol 75: 15-22. http://dx.doi.org/10.1007/s00128-005-0712-l.
Elder, GH; Evans, JO: Matlin, SA. (1976).  The effect of the porphyrogenic compound,
       hexachlorobenzene, on the activity of hepatic uroporphyrinogen decarboxylase in the rat.
       Clin Sci MolMed 51: 71-80.
                                        D-171

-------
Enan, E; Liu, PC: Matsumura, F. (1992). 2,3,7,8-Tetrachlorodibenzo-p-dioxin causes reduction
       of glucose transporting activities in the plasma membranes of adipose tissue and pancreas
       from the guinea pig. J Biol Chem 267: 19785-19791.
Enan, E; El-Sabeawy, F; Overstreet J: Matsumura, F; Lasley, B. (1998). Mechanisms of gender-
       specific TCDD-induced toxicity in guinea pig adipose tissue. Reprod Toxicol 12: 357-
       369.
Eriksson, P;  Lundkvist U; Fredriksson, A. (1991). Neonatal exposure to 3,3',4,4'-
       tetrachlorobiphenyl: Changes in spontaneous behaviour and cholinergic muscarinic
       receptors in the adult mouse. Toxicology 69:  27-34. http://dx.doi.org/10.1016/0300-
       483X(91)90150-Y.
Eskenazi, B; Warner, M; Mocarelli, P: Samuels, S; Needham, LL: Patterson, DG, Jr: Lippman,
       S: Vercellini, P: Gerthoux. PM: Brambilla, P: Olive. D. (2002). Serum dioxin
       concentrations and menstrual  cycle characteristics. Am J Epidemiol 156:  383-392.
Eskenazi, B: Mocarelli, P: Warner, M: Chee, WY: Gerthoux, PM: Samuels, S: Needham, LL:
       Patterson, DG, Jr. (2003). Maternal serum dioxin levels and birth outcomes in women of
       Seveso, Italy. Environ Health Perspect 111: 947-953.
Eskenazi, B: Warner, M: Marks, AR: Samuels, S: Gerthoux, PM: Vercellini, P: Olive, PL;
       Needham, L; Patterson, D, Jr: Mocarelli, P. (2005). Serum dioxin concentrations and age
       at menopause. Environ Health Perspect 113: 858-862.
Eskenazi, B: Warner, M: Samuels, S: Young, J: Gerthoux, PM: Needham, L: Patterson, D: Olive,
       D: Gavoni, N; Vercellini, P: Mocarelli, P. (2007). Serum  dioxin concentrations and risk
       of uterine leiomyoma in the Seveso Women's Health Study. Am J Epidemiol 166: 79-87.
       http://dx.doi.org/10.1093/aje/kwm048.
Esser, C; Steinwachs, S: Herder, C; Majora, M: Lai,  ZW. (2005). Effects of a single dose of
       2,3,7,8-tetrachlorodibenzo-p-dioxin, given at post-puberty, in senescent mice. Toxicol
       Lett 157: 89-98. http://dx.doi.Org/10.1016/i.toxlet.2005.01.007.
Evans, MV:  Andersen, ME. (2000). Sensitivity analysis of a physiological model for 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD): assessing the impact of specific model parameters
       on sequestration in liver and fat in the rat. Toxicol  Sci 54: 71-80.
Faith, RE; Moore, JA. (1977). Impairment of thymus-dependent immune functions by exposure
       of the developing immune system to 2,3,7,8-tetrachlorodibenzo-p-dioxin  (TCDD). J
       Toxicol Environ Health 3: 451-464. http://dx.doi.org/10.1080/15287397709529578.
Fan, F: Rozman, KK. (1994). Relationship between acute toxicity of 2,3,7,8-tetrachlorodibenzo-
       p-dioxin (TCDD) and disturbance of intermediary metabolism in the Long-Evans rat.
       Arch Toxicol 69: 73-78. http://dx.doi.org/10.1007/s002040050140.
Fana, F: Wierdab, D: Rozman, KK. (1996). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin on
       humoral and cell-mediated immunity in Sprague-Dawley rats. Toxicology 106: 221-228.
Faqi, AS: Dalsenter, PR; Merker, HJ: Chahoud, I. (1998).  Reproductive toxicity  and tissue
       concentrations of low doses of 2,3,7,8-tetrachlorodibenzo-p-dioxin in male offspring rats
       exposed throughout pregnancy and lactation.  Toxicol Appl Pharmacol  150: 383-392.
       http://dx.doi.org/10.1006/taap.1998.8433.
Fattore, E: Trossvik, C: Hakansson, H. (2000). Relative potency values derived from hepatic
       vitamin A reduction in male and female Sprague-Dawley rats following subchronic
       dietary exposure to individual polychlorinated dibenzo-p-dioxin and dibenzofuran
       congeners and a mixture therof. Toxicol Appl Pharmacol  165: 184-194.
       http://dx.doi.org/10.1006/taap.2000.8943.

                                         D-172

-------
Fernandez-Salguero. P: Pineau. T: Hilbeit DM: McPhail. T: Lee. SS: Kitnura. S: Nebert DW:
       Rudikoff, S; Ward, JM: Gonzalez, FJ. (1995). Immune system impairment and hepatic
       fibrosis in mice lacking the dioxin-binding Ah receptor. Science 268: 722-726.
Fernandez-Salguero. PM: Hilbeit DM: Rudikoff. S: Ward. JM: Gonzalez. FJ. (1996). Aryl-
       hydrocarbon receptor-deficient mice are resistant to 2,3,7,8-tetrachlorodibenzo-p-dioxin-
       induced toxicity. Toxicol Appl Pharmacol 140: 173-179.
       http://dx.doi.org/10.1006/taap.1996.0210.
Fetissov, SO: Huang, P; Zhang, 0: Mimura, J: Fujii-Kuriyama, Y; Rannug, A; Hokfelt T:
       Ceccatelli, S. (2004). Expression of hypothalamic neuropeptides after acute TCDD
       treatment and distribution of Ah receptor represser.  Regulatory Peptides 119: 113-124.
       http://dx.doi.Org/10.1016/i.regpep.2004.01.009S0167011504000357.
Fine, JS: Gasiewicz, TA: Silverstone, AE. (1989). Lymphocyte stem cell alterations following
       perinatal exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Mol Pharmacol 35: 18-25.
Fine, JS: Gasiewicz, TA: Fiore, NC: Silverstone, AE. (1990). Prothymocyte activity is reduced
       by perinatal 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure. J Pharmacol Exp Ther 255:
       128-132.
Fisher, MT: Nagarkatti, M; Nagarkatti, PS. (2005). Aryl hydrocarbon receptor-dependent
       induction of loss of mitochondrial membrane potential in epididydimal spermatozoa by
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD).  Toxicol Lett 157: 99-107.
       http://dx.doi.0rg/10.1016/i.toxlet.2005.01.008.
Flaws. JA: Sommer. RJ: Silbergeld, EK: Peterson. RE: Hirshfield, AN. (1997). In utero and
       lactational exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) induces genital
       dysmorphogenesis in the female rat. Toxicol Appl Pharmacol 147: 351-362.
       http://dx.doi.org/10.1006/taap.1997.8295.
Fletcher, N; Hanberg, A; Hakansson, H. (2001). Hepatic vitamin a depletion is a sensitive
       marker of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) exposure in four rodent species.
       Toxicol Sci 62: 166-175.
Fletcher. N: Wahlstrom, D: Lundberg, R: Nilsson, CB: Nilsson, KC: Stockling, K: Hellmold. H:
       Hakansson. H. (2005a). 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) alters the mRNA
       expression of critical genes associated with cholesterol metabolism, bile acid
       biosynthesis, and bile transport in rat liver: a microarray study. Toxicol Appl Pharmacol
       207: 1-24. http://dx.doi.Org/10.1016/i.taap.2004.12.003.
Fletcher. N: Giese, N: Schmidt C: Stern. N: Lind, PM: Viluksela, M: Tuomisto, JT: Tuomisto, J:
       Nau, H: Hakansson, H. (2005b). Altered retinoid metabolism in female Long-Evans and
       Han/Wistar rats following long-term 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)-
       treatment. Toxicol Sci 86: 264-272. http://dx.doi.org/10.1093/toxsci/kfil83.
Flodstrom, S: Ahlborg, U.  (1992). Relative liver tumor promoting activity of some
       polychlorinated dibenzo-p-dioxin-, dibenzofuran- and biphenyl-congeners in female rats.
       Chemosphere 25: 169-172. http://dx.doi.org/10.1016/0045-6535(92)90505-L.
Foster. WG: Ruka. MP: Gareau. P: Foster. RA: Janzen. EG: Yang. JZ. (1997). Morphologic
       characteristics of endometriosis in the mouse model: application to toxicology. Can J
       Physiol Pharmacol  75:  1188-1196.
Fox, TR; Best LL; Goldsworthy, SM; Mills, JJ: Goldsworthy, TL. (1993). Gene expression and
       cell proliferation in rat liver after 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure.  Cancer
       Res 53: 2265-2271.
                                         D-173

-------
Franc. MA: Pohjanvirta. R: Tuomisto. J: Okev. AB. (2001). Persistent, low-dose 2,3,7,8-
       tetrachlorodibenzo-p-dioxin exposure: effect on aryl hydrocarbon receptor expression in
       a dioxin-resistance model. Toxicol Appl Pharmacol 175: 43-53.
       http://dx.doi.org/10.1006/taap.2001.9222.
Franczak, A; Nynca, A; Valdez, KE; Mizinga, KM; Petroff BK. (2006). Effects of acute and
       chronic exposure to the aryl hydrocarbon receptor agonist 2,3,7,8-tetrachlorodibenzo-p-
       dioxin on the transition to reproductive senescence in female Sprague-Dawley rats. Biol
       Reprod 74: 125-130. http://dx.doi.org/10.1095/biolreprod.105.044396.
Frericks, M; Temchura, VV: Majora, M; Stutte, S: Esser, C. (2006). Transcriptional signatures of
       immune cells in aryl hydrocarbon receptor (AHR)-proficient and AHR-deficient mice.
       Biol Chem 387: 1219-1226. http://dx.doi.org/10.1515/BC.2006.151.
Fritz. WA: Lin. TM: Moore. RW: Cooke. PS: Peterson. RE. (2005). In utero and lactational
       2,3,7,8-tetrachlorodibenzo-p-dioxin exposure: effects on the prostate and its response to
       castration in senescent C57BL/6J mice. Toxicol Sci 86: 387-395.
       http://dx.doi.org/10.1093/toxsci/kfil89.
Fujimaki, H; Nohara, K; Kobayashi, T; Suzuki, K; Eguchi-Kasai, K; Tsukumo, S: Kijima, M:
       Tohyama,  C. (2002). Effect of a single oral dose of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       on immune function in male NC/Nga mice. Toxicol Sci 66: 117-124.
Fujiwara, K: Yamada, T; Mishima, K: Imura, H; Sugahara, T. (2008). Morphological and
       immunohistochemical studies on cleft palates induced by 2,3,7,8-tetrachlorodibenzo-p-
       dioxin in mice. Congenit Anom 48: 68-73. http://dx.doi.org/10.llll/U741-
       4520.2008.00181.x.
Funatake. CJ: Marshall NB: Steppan. LB: Mourich. DV: Kerkvliet NI. (2005). Cutting edge:
       activation of the aryl hydrocarbon receptor by 2, 3, 7, 8-tetrachlorodibenzo-p-dioxin
       generates a population of CD4+ CD25+ cells with characteristics of regulatory T cells. J
       Immunol 175: 4184.
Funseth. E: Wesslen. L: Lindh. U:  Friman. G: Ilback. NG. (2002a). Effect of 2, 3, 7, 8-
       tetrachlorodibenzo-p-dioxin on trace elements, inflammation and viral clearance in the
       myocardium during coxsackievirus B3 infection in mice. Sci Total Environ 284: 135-
       147.
Funseth. E: Pahlman. M: Eloranta. ML: Friman. G: nback. NG. (2002b). Effects of
       coxsackievirus B3 infection on the acute-phase protein metallothionein and on
       cytochrome P-4501A1 involved in the detoxification processes of TCDD in the mouse.
       Sci Total Environ 284: 37-47.
Galijatovic, A; Beaton, D; Nguyen, N; Chen, S: Bonzo, J: Johnson, R; Maeda, S: Karin, M:
       Guengerich, FP; Tukey, RH. (2004). The human CYP1A1  gene is regulated in a
       developmental and tissue-specific fashion in transgenic mice. J Biol Chem 279: 23969-
       23976.
Gallo, MA; Hesse, EJ: MacDonald, GJ: Umbreit TH. (1986). Interactive effects of estradiol and
       2, 3, 7, 8-tetrachloro-dibenzo-p-dioxin on hepatic cytochrome P-450 and mouse uterus.
       Toxicol Lett 32: 123-132.
Gao, X; Petroff BK: Rozman, KK; Terranova, PF. (2000). Gonadotropin-releasing hormone
       (GnRH) partially reverses the inhibitory effect of 2, 3, 7, 8-tetrachlorodibenzo-p-dioxin
       on ovulation in the immature gonadotropin-treated rat. Toxicology 147: 15-22.
                                         D-174

-------
Gao, X; Mizuyachi, K; Terranova, PF; Roztnan, KK. (2001). 2, 3, 7, 8-Tetrachlorodibenzo-p-
       dioxin decreases responsiveness of the hypothalamus to estradiol as a feedback inducer of
       preovulatory gonadotropin secretion in the immature gonadotropin-primed rat. Toxicol
       Appl Pharmacol 170: 181-190.
Gao, Y; Sahlberg, C: Kiukkonen, A; Alaluusua, S: Pohjanvirta, R: Tuomisto, J: Lukinmaa, PL.
       (2004). Lactational exposure of Han/Wistar rats to 2, 3, 7, 8-tetrachlorodibenzo-p-dioxin
       interferes with enamel maturation and retards dentin mineralization. J Dent Res 83: 139.
Garrett RW: Gasiewicz, TA. (2006). The aryl hydrocarbon receptor agonist 2, 3, 7, 8-
       tetrachlorodibenzo-p-dioxin alters the circadian rhythms, quiescence, and expression of
       clock genes in murine hematopoietic stem and progenitor cells. Mol Pharmacol 69: 2076.
Gasiewicz, TA: Geiger, LE; Rucci, G: Neal, RA. (1983). Distribution, excretion, and metabolism
       of 2,3,7,8-tetrachlorodibenzo-p-dioxin in C57BL/6J, DBA/2J, and B6D2F1/J mice. Drug
       Metab Dispos 11: 397-403.
Gasiewicz, TA: Rucci, G. (1984). Cytosolic receptor for 2, 3, 7, 8-tetrachlorodibenzo-p-dioxin.
       Evidence for a homologous nature among various mammalian species. Mol Pharmacol
       26: 90.
Gasiewicz, TA: Rucci, G: Henry, EC: Baggs, RB. (1986). Changes in hamster hepatic
       cytochrome P-450, ethoxycoumarin  O-deethylase, and reduced NAD(P): menadione
       oxidoreductase following treatment with 2,3,7,8-tetrachlorodibenzo-p-dioxin. Partial
       dissociation of temporal and dose-response relationships from elicited toxicity. Biochem
       Pharmacol 35: 2737-2742.
Gehrs, BC: Riddle, MM; Williams, WC: Smialowicz, RJ. (1997). Alterations in the developing
       immune system of the F344 rat after perinatal exposure to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin II Effects on the pup and the adult.  Toxicology 122: 229-240.
Gehrs, BC: Smialowicz, RJ. (1999). Persistent suppression of delayed-type hypersensitivity in
       adult F344 rats after perinatal exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin.
       Toxicology 134: 79-88.
Center. MB: Clay. CD: Dalton. TP: Dong. H: Nebeit DW: Shertzer, HG. (2006). Comparison of
       mouse hepatic mitochondrial versus  microsomal  cytochromes P450 following TCDD
       treatment. Biochem Biophys Res Commun 342: 1375-1381.
Geusau, A: Khorchide, M: Mildner, M: Pammer. J: Eckhart L: Tschachler, E. (2005). 2,3,7,8-
       tetrachlorodibenzo-p-dioxin impairs differentiation of normal human epidermal
       keratinocytes in a skin equivalent model. J Invest Dermatol  124: 275-277'.
       http://dx.doi.Org/10.llll/i.0022-202X.2004.23541.x.
Ghafoorunissa. (1980). Undernutrition and fertility of male rats. J Reprod Fertil 59: 317-320.
Giavini, E: Prati. M: Vismara, C. (1982). Rabbit teratology study with 2,3,7,8-
       tetrachlorodibenzo-p-dioxin. Environ Res 27: 74-78.
Giavini, E: Prati, M: Vismara, C. (1983). Embryotoxic effects of 2,3,7,8 tetrachlorodibenzo-p-
       dioxin administered to female rats before mating. Environ Res 31: 105-110.
       http://dx.doi.org/10.1016/0013-9351(83)90066-X.
Goldey, ES:  Crofton, KM. (1998). Thyroxine replacement attenuates hypothyroxinemia, hearing
       loss,  and motor deficits following developmental exposure to Aroclor 1254 in rats.
       Toxicol  Sci 45: 94-105.  http://dx.doi.org/10.1006/toxs.1998.2495.
Goldstein, JA; Hickman, P; Bergman, H; Vos, JG. (1973). Hepatic porphyria induced by 2,3,7,8-
       tetrachlorodibenzo-p-dioxin in the mouse. Res Comm Chem Pathol Pharmacol 6: 919-
       928.

                                         D-175

-------
Goldstein, JA: Linko, P; Bergman, H. (1982). Induction of porphyria in the rat by chronic versus
       acute exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Biochem Pharmacol 31: 1607-
       1613.
Goldstein, JA: Linko, P. (1984). Differential induction of two 2,3,7,8-tetrachlorodibenzo-p-
       dioxin-inducible forms of cytochrome P-450 in extrahepatic versus hepatic tissues. Mol
       Pharmacol 25: 185-191.
Gonzalez, FJ: Fernandez-Salguero, P; Lee, SS: Pineau, T; Ward, JM. (1995). Xenobiotic
       receptor knockout mice. Toxicol Lett 82-83: 117-121.
Goodman, DG: Sauer, RM. (1992). Hepatotoxicity and carcinogenicity in female Sprague-
       Dawley rats treated with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD): a pathology
       working group reevaluation. Regul Toxicol Pharmacol 15: 245-252.
Gordon, CJ: Gray, LE: Monteiro-Riviere, NA: Miller, DB. (1995). Temperature regulation and
       metabolism in rats exposed perinatally to dioxin: permanent change in regulated body
       temperature. Toxicol Appl Pharmacol 133: 172-176.
Gordon, CJ: Yang, Y; Gray, JLE. (1996). Autonomic and Behavioral Thermoregulation in
       Golden Hamsters Exposed Perinatally to Dioxin. Toxicol Appl Pharmacol 137: 120-125.
Gordon, CJ: Miller, DB. (1998).  Thermoregulation in rats exposed perinatally to dioxin: core
       temperature stability to altered ambient temperature, behavioral thermoregulation, and
       febrile response to lipopolysaccharide. J Toxicol Environ Health A 54: 647-662.
Gorski, JR: Rozman, K. (1987). Dose-response and time course of hypothyroxinemia and
       hypoinsulinemia and characterization of insulin hypersensitivity in 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD)-treated rats. Toxicology 44: 297-307.
Gorski, JR: Weber, LW: Rozman, K. (1990). Reduced gluconeogenesis in 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD)-treated rats. Arch Toxicol 64: 66-71.
Gray. EL. Jr: Ostbv, J: Wolf. C: Miller. DB: Kelce, WR: Gordon. CJ: Birnbaum, L. (1995a).
       Functional developmental toxicity of low doses of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
       and a dioxin-like pcb (169) in Long Evans rats and Syrian hamsters: reproductive,
       behavioral and thermoregulatory alterations. Organohalogen Compounds 25: 33-38.
Gray, LE: Ostby, JS: Kelce, WR. (1997a). A dose-response analysis of the reproductive effects
       of a single gestational dose of 2,3,7,8-tetrachlorodibenzo-p-dioxin in male long evans
       hooded rat offspring. Toxicol Appl Pharmacol 146: 11-20.
Gray. LE: Wolf. C: Mann. P: Ostbv, JS. (1997b). In Utero Exposure to Low Doses of 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin Alters Reproductive Development of Female Long Evans
       Hooded Rat Offspring. Toxicol Appl Pharmacol 146: 237-244.
Gray. LE. Jr: Kelce. WR: Monosson, E: Ostbv, JS: Birnbaum, LS. (1995b). Exposure to TCDD
       during development permanently alters reproductive function in male Long Evans rats
       and hamsters: Reduced ejaculated and epididymal sperm numbers and sex accessory
       gland weights in offspring with normal androgenic status.  Toxicol Appl Pharmacol 131:
       108-118. http://dx.doi.org/10.1006/taap.1995.1052.
Greenlee, WF: Dold, KM: Irons. RD: Osborne, R. (1985). Evidence for direct action of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) on thymic epithelium. Toxicol Appl Pharmacol 79:
       112-120.
Greig, JB: De Matteis, F. (1973). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin on drug
       metabolism and hepatic microsomes of rats and mice. Environ Health Perspect 5:211-
       219.
                                        D-176

-------
Guo, L; Zhao, YY; Sun, ZJ: Liu, H; Zhang, SL. (2007). Toxic effects of TCDD on Osteogenesis
       through altering IGFBP-6 gene expression in osteoblasts. Biol Pharm Bull 30: 2018-
       2026.
Guo, L: Zhao, Y; Zhang, S: Liu, K. (2008). Anti-estrogenic effect of dioxin on rat skeleton
       development. Wei Sheng Yan Jiu 37: 563-565.
Guo. YM: Wang. SY: Wang. XR: Laslev. B. (2000). Effect of TCDD on maternal toxicity and
       chorionic gonadotropin—bioactivity in the immediate post-implantation period of
       macaque. Biomed Environ Sci 13: 26-31.
H-YN. H: Goth-Goldstein. R: Martin. MC: Russell ML: McKinnev. WR. (2000). Low-dose
       responses to 2,3,7,8-tetrachlorodibenzo-p-dioxin in single living human cells measured
       by synchrotron infrared spectromicroscopy. Environ Sci Technol 34: 2513-2517.
       http://dx.doi.org/10.1021/es991430w.
Haag-Gronlund, M; Warngard, L: Flodstrom, S: Scheu, G: Kronevi, T; Ahlborg, UG: Fransson-
       Steen, R. (1997). Promotion of altered hepatic foci by 2,3',4,4',5-pentachlorobiphenyl in
       Sprague-Dawley female rats. Fundam Appl Toxicol 35: 120-130.
Haake, JM: Safe. S: Mavura, K: Phillips. TD. (1987). Aroclor 1254 as an antagonist of the
       teratogenicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Lett 38: 299-306.
Haavisto, T; Nurmela, K: Pohjanvirta, R: Huuskonen, H: El-Gehani, F; Paranko, J. (2001).
       Prenatal testosterone and luteinizing hormone levels in male rats exposed during
       pregnancy to 2,3,7,8-tetrachlorodibenzo-p-dioxin and diethylstilbestrol. Mol  Cell
       Endocrinol  178: 169-179.
Haavisto, TE; Mvllymaki, SA; Adamsson, NA; Brokken, LJS: Viluksela, M; Toppari, J:
       Paranko, J. (2006). The effects of maternal exposure to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin on testicular steroidogenesis in infantile male rats. Int J Androl 29: 313-322.
       http://dx.doi.0rg/10.llll/i.1365-2605.2005.00568.x.
Hahn, ME; Gasiewicz, TA; Linko, P; Goldstein, JA. (1988). The role of the Ah locus in
       hexachlorobenzene-induced porphyria: studies in congenic C57BL/6J mice. Biochemistry
       254: 245-254.
Hakansson, H: Johansson, L: Manzoor, E;  Ahlborg, U. (1989a). 2,3,7,8-Tetrachlorodibenzo-p-
       dioxin (TCDD)-induced alterations in the vitamin A homeostasis and in the 7-
       ethoxyresorufm O-deethylase (EROD)-activity in Sprague-Dawley rats and Hartley
       guinea pigs. Chemosphere 18: 299-305. http://dx.doi.org/10.1016/0045-6535(89)90134-
       O
Hakansson, H: Johansson, L: Ahlborg, U: Moore, R: Peterson, R. (1989b). Hepatic vitamin A
       storage in relation to paired feed restriction and to TCDD-treatment. Chemosphere 19:
       919-920. http://dx.doi.org/10.1016/0045-6535(89)90432-3.
Hakansson, H: Ahlborg, U: Johansson, L: Poiger, H. (1990). Vitamin A storage in rats
       subchronically exposed to PCDDs/PCDFs. Chemosphere 20: 1147-1150.
       http://dx.doi.org/10.1016/0045-6535(90)90235-L.
Hakansson, H: Hanberg, A. (1989). The distribution of [14C]-2,3,7,8-tetrachlorodibenzo-p-
       dioxin (TCDD) and its effect on the vitamin A content in parenchymal and stellate cells
       of rat liver. JNutr 119: 573-580.
Hakansson. H: Johansson. L: Manzoor. E:  Ahlborg. UG. (1991). Effects of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) on the vitamin A status of Hartley guinea pigs,
       Sprague-Dawley rats, C57B1/6 mice, DBA/2 mice, and Golden  Syrian hamsters. J Nutr
       Sci Vitaminol 37: 117-138.

                                         D-177

-------
Hakansson. H: Johansson. L: Manzoor. E: Ahlborg. UG. (1994). Effect of 2,3,7,8-
       tetrachlorodibenzoy-p-dioxin on the hepatic 7-ethoxyresorufin O-deethylase activity in
       four rodent species. Eur J Pharmacol 270: 279-284.
Hamm, JT: Sparrow, BR; Wolf D: Birnbaum, LS. (2000). In utero and lactational exposure to
       2,3,7,8-tetrachlorodibenzo-p-dioxin alters postnatal development of seminal vesicle
       epithelium. Toxicol Sci 54: 424-430.
Hamm, JT: Chen, CY; Birnbaum, LS. (2003). A mixture of dioxins, furans, and non-ortho PCBs
       based upon consensus toxic equivalency factors produces dioxin-like reproductive
       effects. Toxicol Sci 74: 182-191.
Hanson. CD: Smialowicz, RJ. (1994). Evaluation of the effect of low-level 2,3,7,8-
       tetrachlorodibenzo-p-dioxin exposure on cell mediated immunity. Toxicology 88: 213-
       224.
Hany, J: Lilienthal H: Roth-Harer, A: Ostendorp, G: Heinzow, B: Winneke, G. (1999).
       Behavioral effects following single and combined maternal exposure to PCB 77
       (3,4,3',4'-tetrachlorobiphenyl) and PCB 47 (2,4,2',4'-tetrachlorobiphenyl) in rats.
       Neurotoxicol Teratol 21: 147-156. http://dx.doi.org/10.1016/S0892-0362(98)00038-5.
Harper, N; Connor, K; Steinberg, M; Safe, S. (1994a). An enzyme-linked immunosorbent assay
       (ELISA) specific for antibodies  to TNP-LPS detects alterations in serum
       immunoglobulins and isotype switching in C57BL/6 and DBA/2 mice exposed to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin and related compounds. Toxicology 92:  155-167.
Harper, N: Wang, X; Liu, H: Safe, S. (1994b). Inhibition of estrogen-induced progesterone
       receptor in MCF-7 human breast cancer cells by aryl hydrocarbon (Ah) receptor agonists.
       Mol Cell Endocrinol 104: 47-55. http://dx.doi.org/10.1016/0303-7207(94)90050-7.
Harper, PA;  Golas, CL; Okey, AB. (1991). Ah receptor in mice genetically
       "nonresponsive" for cytochrome P4501A1 induction: cytosolic Ah receptor,
       transformation to the nuclear binding state, and induction of aryl hydrocarbon
       hydroxylase by halogenated and nonhalogenated aromatic hydrocarbons in embryonic
       tissues and cells. Mol Pharmacol 40: 818-826.
Harris. MW: Moore. JA: Vos, JG: Gupta. BN. (1973). General biological effects of TCDD in
       laboratory animals. Environ Health Perspect 5: 101-109.
Hart, BL. (1972). Manipulation of neonatal androgen:  effects on sexual responses and penile
       development in male rats. Physiol Behav 8: 841-845.
Harvey, M; McArthur, MJ: Montgomery, CA; Butel, JS: Bradley, A; Donehower, LA. (1993).
       Spontaneous and carcinogen-induced tumorigenesis in p53-deficient mice. Nat Genet 5:
       225-229. http://dx.doi.org/10.1038/ngll93-225.
Hassoun, E;  d'Argy, R; Dencker, L; Lundin, LG: Borwell, P. (1984a). Teratogenicity of 2,3,7,8-
       tetrachlorodibenzofuran in BXD recombinant inbred strains. Toxicol Lett 23: 37-42.
Hassoun, E:  d'Argy, R; Dencker, L: Sundstrom, G. (1984b). Teratological studies on the TCDD
       congener 3,3',4,4'-tetrachloroazoxybenzene in sensitive and nonsensitive mouse strains:
       evidence for direct effect on embryonic tissues. Arch Toxicol 55: 20-26.
Hassoun, EA: Bagchi, D: Stohs, SJ. (1995). Evidence of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       (TCDD)-induced tissue damage in fetal and placental tissues and changes in amniotic
       fluid lipid  metabolites of pregnant CF1 mice. Toxicol Lett 76: 245-250.
Hassoun. EA: Walter. AC: Alsharif NZ: Stohs. SJ.  (1997). Modulation of TCDD-induced
       fetotoxicity and oxidative stress in embryonic and placental tissues of C57BL/6J mice by
       vitamin E  succinate and ellagic acid. Toxicology 124: 27-37.

                                         D-178

-------
Hassoun. EA: Wilt SC: Devito. MJ: Van Birgelen. A: Alsharif NZ: Birnbaum. LS: Stohs. SJ.
       (1998). Induction of oxidative stress in brain tissues of mice after subchronic exposure to
       2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Sci 42: 23-27.
       http://dx.doi.0rg/10.1093/toxsci/42.l.23.
Hassoun. EA: Li. F: Abushaban. A: Stohs. SJ. (2000). The relative abilities of TCDD and its
       congeners to induce oxidative stress in the hepatic and brain tissues of rats after
       subchronic exposure. Toxicology 145: 103-113.
Hassoun, EA: Li, F; Abushaban, A; Stohs, SJ. (2001). Production of superoxide anion, lipid
       peroxidation and DNA damage in the hepatic and brain tissues of rats after subchronic
       exposure to mixtures of TCDD and its congeners'. J Appl Toxicol 21: 211-219.
       http://dx.doi.org/10.1002/jat.744.
Hassoun, EA: Wang, H; Abushaban, A;  Stohs, SJ. (2002). Induction of oxidative stress
       following chronic exposure to TCDD, 2,3,4,7,8-pentachlorodibenzofuran, and
       2,3',4,4',5-pentachlorobiphenyl.  J Toxicol Environ Health A 65: 825-842.
Hassoun, EA: Al-Ghafri, M; Abushaban, A. (2003). The role  of antioxidant enzymes in TCDD-
       induced oxidative stress in various brain regions of rats after subchronic exposure. Free
       Radic Biol Med 35: 1028-1036. http://dx.doi.org/10.1016/50891-5849(03)00458-1.
Hassoun, EA: Vodhanel, J: Abushaban, A. (2004). The modulatory effects of ellagic acid and
       vitamin E succinate on TCDD-induced oxidative stress in different brain regions of rats
       after subchronic exposure. J Biochem Mol Toxicol 18: 196-203.
       http://dx.doi.org/10.1002/ibt.20030.
Hassoun, EA: Vodhanel J: Holden, B; Abushaban, A. (2006). The effects of ellagic acid and
       vitamin E succinate on antioxidant enzymes activities and glutathione levels in different
       brain regions of rats after subchronic exposure to TCDD. J Toxicol Environ Health A 69:
       381-393. http://dx.doi.org/10.1080/15287390500246431.
Hebert, CD: Harris, MW: Elwell MR; Birnbaum, LS. (1990). Relative toxicity and tumor-
       promoting ability of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), 2,3,4,7,8-
       pentachlorodibenzofuran (PCDF), and 1,2,3,4,7,8-hexachlorodibenzofuran (HCDF) in
       hairless mice. Toxicol Appl Pharmacol 102: 362-377.
Heimler, I; Rawlins, RG: Owen, H; Hutz, RJ. (1998). Dioxin  perturbs, in a dose- and time-
       dependent fashion, steroid secretion, and induces apoptosis of human luteinized
       granulosa cells. Endocrinology 139: 4373-4379.
Hemming, H: Flodstrom, S; Warngard, L; Bergman, A;  Kronevi, T; Nordgren, I; Ahlborg, UG.
       (1993). Relative tumour promoting activity of three polychlorinated biphenyls in rat liver.
       Eur J Pharmacol 248: 163-174.
Hemming, H: Eager, Y; Flodstrom, S: Nordgren, I; Kronevi, T; Ahlborg, UG: Warngard, L.
       (1995). Liver tumour promoting activity of 3,4,5,3',4'-pentachlorobiphenyl and its
       interaction with 2,3,7,8-tetrachlorodibenzo-p-dioxin. Eur J Pharmacol 292: 241-249.
Henck, JM: New. MA: Kociba, RJ:  Rao. KS. (1981). 2,3,7,8-Tetrachlorodibenzo-p-dioxin: acute
       oral toxicity in hamsters. Toxicol Appl Pharmacol 59: 405-407.
       http://dx.doi.org/10.1016/0041-008X(81)90212-X.
Henry, EC: Gasiewicz, TA. (1987).  Changes in thyroid hormones and thyroxine glucuronidation
       in hamsters compared with rats following treatment with 2,3,7,8-tetrachlorodibenzo-p-
       dioxin. Toxicol Appl Pharmacol  89: 165-174.
                                         D-179

-------
Henry, EC: Betnis, JC: Henry, O: Kende, AS: Gasiewicz, TA. (2006). A potential endogenous
       ligand for the aryl hydrocarbon receptor has potent agonist activity in vitro and in vivo.
       Arch Biochem Biophys 450: 67-77. http://dx.doi.Org/10.1016/j.abb.2006.02.008.
Hermsen, SA; Larsson, S: Arima, A; Muneoka, A; Ihara, T; Sumida, H; Fukusato, T; Kubota, S:
       Yasuda, M; Lind, PM. (2008). In utero and lactational exposure to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) affects bone tissue in rhesus monkeys. Toxicology
       253: 147-152.
Hern DW: Goldev. ES: Crofton. KM. (1996). Developmental exposure to Aroclor 1254
       produces low-frequency alterations in adult rat brainstem auditory evoked responses.
       Fundam Appl Toxicol 33: 120-128. http://dx.doi.org/10.1006/faat.1996.0149.
Herzke, D:  Thiel R: Rotard, WD: Neubert, D. (2002). Kinetics and organotropy of some
       polyfluorinated dibenzo-p-dioxins and dibenzofurans (PFDD/PFDF) in rats. Life Sci 71:
       1475-1486. http://dx.doi.org/10.1016/S0024-3205(02)01924-0.
Hinsdill, RD: Couch, PL; Speirs, RS. (1980). Immunosuppression in mice induced by dioxin
       (TCDD) in feed. J Environ Pathol Toxicol 4: 401-425.
Hochstein, JR; Bursian, SJ: Aulerich, RJ. (1998). Effects of dietary exposure to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin in adult female mink (Mustela vison). Arch Environ Contam
       Toxicol 35: 348-353.
Hochstein, MS, Jr; Render, JA; Bursian, SJ: Aulerich, RJ. (2001). Chronic toxicity of dietary
       2,3,7,8-tetrachlorodibenzo-p-dioxin to mink. Vet Hum Toxicol 43: 134-139.
Hoegberg, P; Schmidt CK; Fletcher, N; Nilsson, CB; Trossvik, C; Gerlienke Schuur, A;
       Brouwer, A; Nau, H; Ghyselinck, NB; Chambon, P; Hakansson, H. (2005). Retinoid
       status and responsiveness to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in mice lacking
       retinoid binding protein or retinoid receptor forms. Chem Biol Interact 156: 25-39.
       http://dx.doi.0rg/10.1016/i.cbi.2005.06.006.
Hofer,  T: Pohjanvirta, R: Spielmann, P: Viluksela, M; Buchmann, DP; Wenger, RH; Gassmann,
       M. (2004). Simultaneous exposure of rats to dioxin and carbon monoxide reduces the
       xenobiotic but not the hypoxic response.  Biol Chem 385: 291-294.
Hoffer, A; Chang, CY; Puga, A.  (1996). Dioxin induces transcription of fos and jun genes by Ah
       receptor-dependent and -independent pathways. Toxicol Appl Pharmacol  141: 238-247.
       http://dx.doi.org/10.1006/taap.1996.0280.
Hogaboam, JP; Moore, AJ: Lawrence, BP. (2008). The aryl hydrocarbon  receptor affects distinct
       tissue compartments during ontogeny of the immune system. Toxicol Sci  102: 160-170.
       http://dx.doi.org/10.1093/toxsci/kfm283.
Hoio. R: Stern. S:  Zareba. G: Markowskl VP: Cox. C: Kost JT: Weiss. B. (2002). Sexually
       dimorphic  behavioral  responses to prenatal dioxin exposure. Environ Health Perspect
       110:247-254.
Hojo, R: Zareba, G: Kai, JW: Baggs, RB: Weiss, B. (2006). Sex-specific  alterations of cerebral
       cortical cell size in rats exposed prenatally to dioxin. J Appl Toxicol 26: 25-34.
       http://dx.doi.org/10.1002/iat.1101.
Holcomb, M: Safe, S. (1994). Inhibition of 7,12-dimethylbenzanthracene-induced rat mammary
       tumor growth by 2,3,7,8-tetrachlorodibenzo-p-dioxin. Cancer Lett 82: 43-47.
Holene. E: Nafstad. I: Skaare. JU: Bernhoft A: Engen. P: Sagvolden. T. (1995). Behavioral
       effects of pre- and postnatal exposure to individual polychlorinated biphenyl  congeners in
       rats. Environ Toxicol  Chem 14:  967-976.
                                        D-180

-------
Holladay, SD: Lindstrom, P; Blaylock, BL; Comment CE; Germolec, PR: Heindell, JJ: Luster,
       MI. (1991). Perinatal thymocyte antigen expression and postnatal immune development
       altered by gestational exposure to tetrachlorodibenzo-p-dioxin (TCDD). Teratology 44:
       385-393. http://dx.doi.org/10.1002/tera.1420440405.
Hong, R; Taylor, K;  Abonour, R. (1989). Immune abnormalities associated with chronic TCDD
       exposure in rhesus. Chemosphere 18: 313-320. http://dx.doi.org/10.1016/0045-
       6535(89)90136-7.
Hood. DB:  Woods. L: UN. B: Johnson. S: Ebner. FF. (2006). Gestational 2,3,7,8-
       tetrachlorodibenzo-p-dioxin exposure effects on sensory cortex function.
       Neurotoxicology 27: 1032-1042. http://dx.doi.Org/10.1016/i.neuro.2006.05.022.
Hook, GE;  Haseman, JK; Lucier, GW. (1975). Induction and suppression of hepatic and
       extrahepatic microsomal foreign-compound-metabolizing enzyme systems by 2,3,7,8-
       tetrachlorodibenzo-p-dioxin. Chem Biol Interact 10: 199-214.
House. RV: Lauer. LD: Murray. MJ: Thomas. PT: Ehrlich. JP: Burleson. GR: Dean. JH. (1990).
       Examination of immune parameters and host resistance mechanisms in B6C3F1  mice
       following adult exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. J Toxicol Environ
       Health 31: 203-215. http://dx.doi.org/10.1080/15287399009531449.
Huff JE; Salmon, AG: Hooper, NK; Zeise, L. (1991). Long-term carcinogenesis  studies on
       2,3,7,8-tetrachlorodibenzo-p-dioxin and hexachlorodibenzo-p-dioxins. Cell Biol Toxicol
       7: 67-94. http://dx.doi.org/10.1007/BF00121331.
Huff JE. (1992). 2,3,7,8-TCDD: A potent and complete carcinogen in experimental animals.
       Chemosphere 25: 173-176.
Hung. YC:  Huang. GS: Sava. VM:  Blagodarsky. VA: Hong. MY. (2006). Protective effects of
       tea melanin against 2,3,7,8-tetrachlorodibenzo-p-dioxin-induced toxicity:  Antioxidant
       activity and aryl hydrocarbon receptor suppressive effect. Biol Pharm Bull 29: 2284-
       2291.
Hurst CH:  DeVito. MJ: Setzer. RW: Birnbaum. LS. (2000). Acute administration of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) in pregnant Long Evans rats: association of
       measured tissue concentrations with  developmental effects. Toxicol Sci 53: 411-420.
Hurst CH:  Abbott B: Schmid. JE:  Birnbaum. LS. (2002). 2,3,7,8-tetrachlorodibenzo-p-dioxin
       (TCDD) disrupts early morphogenetic events that form the lower reproductive tract in
       female rat fetuses. Toxicol Sci 65: 87-98.
Hushka. LJ: Williams. JS: Greenlee. WF. (1998). Characterization of 2,3,7,8-
       tetrachlorodibenzofuran-dependent suppression and AH receptor pathway gene
       expression in the developing mouse mammary gland. Toxicol Appl Pharmacol 152: 200-
       210. http://dx.doi.org/10.1006/taap.1998.8508.
Hutt KJ: Shi. Z: Albertinl DF: Petroff BK. (2008). The environmental toxicant 2,3,7,8-
       tetrachlorodibenzo-p-dioxin disrupts morphogenesis of the rat pre-implantation embryo.
       BMC Dev Biol 8: 1-12. http://dx.doi.org/10.1186/1471-213X-8-l.
Huuskonen, H; Unkila, M; Pohjanvirta, R: Tuomisto, J. (1994). Developmental toxicity of
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in the most TCDD-resistant  and -susceptible
       rat strains. Toxicol Appl Pharmacol 124: 174-180.
       http://dx.doi.org/10.1006/taap.1994.1021.
Hwang, SY; Kim, WJ: Wee, JJ: Choi, JS: Kim, SK. (2004). Panax ginseng improves survival
       and sperm quality in guinea pigs exposed to 2,3,7,8-tetrachlorodibenzo- p-dioxin. BJU
       Int 94: 663-668. http://dx.doi.Org/10.llll/i.1464-410X.2004.05019.x.

                                         D-181

-------
Iba, MM; Fung, J. (2001). Pulmonary CyplAl and CYP1A2 levels and activities in adult male
       and female offspring of rats exposed during gestation and lactation to 2,3,7, 8-
       tetrachlorodibenzo-p-dioxin. Biochem Pharmacol 62: 617-626.
Ikeda, M; Mitsui, T; Setani, K; Tamura, M; Kakeyama, M; Sone, H; Tohyama, C: Tomita, T.
       (2005a). In utero and lactational exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin in rats
       disrupts brain sexual differentiation. Toxicol Appl Pharmacol 205: 98- 105.
       http://dx.doi.0rg/10.1016/j.taap.2004.09.010.
Ikeda, M: Tamura, M: Yamashita, J: Suzuki,  C: Tomita, T. (2005b). Repeated in utero and
       lactational 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure affects male gonads in offspring,
       leading to sex ratio changes in F2 progeny. Toxicol Appl Pharmacol 206: 351-355.
       http://dx.doi.0rg/10.1016/i.taap.2004.ll.019.
Inouve, K: Pan. X: Imai, N: Ito. T: Takei, T: Tohyama, C: Nohara, K. (2005). T cell-derived IL-5
       production is a sensitive target of 2, 3, 7, 8-tetrachlorodibenzo-p-dioxin (TCDD).
       Chemosphere 60: 907-913. http://dx.doi.Org/10.1016/i.chemosphere.2005.01.014.
loannou, YM: Birnbaum, LS: Matthews. HB. (1983). Toxicity and distribution of 2,3,7,8-
       tetrachlorodibenzofuran in male guinea pigs. J Toxicol Environ Health 12: 541-553.
       http://dx.doi.org/10.1080/15287398309530448.
Ishida, T: Oshimo, T: Nishimura, A; Mutoh, J: Ishii, Y; Koga, N: Yamada, H; Hashiguchi, I;
       Akamine, A; Oguri, K. (2004). Reduction of the toxicity of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin in mice using an antiulcer drug, geranylgeranylacetone. Biol Pharm Bull 27: 1397-
       1402. http://dx.doi.org/10.1248/bpb.27.1397.
Ishihara, K: Warita, K: Tanida, T: Sugawara, T: Kitagawa, H: Hoshi, N. (2007). Does paternal
       exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) affect the sex ratio of offspring.
       J Vet MedSci 69: 347-352.
Ishimura, R: Ohsako, S; Miyabara, Y; Sakaue, M: Kawakami, T: Aoki, Y; Yonemoto,  J:
       Tohyama, C. (2002). Increased glycogen content and glucose transporter 3 mRNA level
       in the placenta of Holtzman rats after  exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin.
       Toxicol Appl Pharmacol 178: 161-171. http://dx.doi.org/10.1006/taap.2001.9333.
Ishimura, R: Kawakami, T: Ohsako, S: Nohara, K: Tohyama, C. (2006). Suppressive effect of
       2,3,7,8-tetrachlorodibenzo-p-dioxin on vascular remodeling that takes place in the normal
       labyrinth zone of rat placenta during late gestation. Toxicol  Sci 91: 265-274.
       http://dx.doi.org/10.1093/toxsci/kfil38.
Ishizuka, M: Yonemoto, J: Zaha, H: Tohyama, C: Sone, H. (2003).  Perinatal exposure  to low
       doses of 2,3,7,8-tetrachlorodibenzo-p-dioxin alters sex-dependent expression of hepatic
       CYP2C11. J Biochem Mol Toxicol 17: 278-285. http://dx.doi.org/10.1002/ibt.10090.
Ito, N: Tatematsu, M: Nakanishi, K: Hasegawa, R: Takano, T: Imaida, K: Ogiso, T. (1980). The
       effects of various chemicals on the development of hyperplastic liver nodules in
       hepatectomized rats treated with N-nitrosodiethylamine or N-2-fluorenylacetamide. Gann
       71: 832-842.
Ito, T: Inouye,  K: Fujimaki, H: Tohyama, C: Nohara, K. (2002). Mechanism of TCDD-induced
       suppression of antibody production: effect on T cell-derived cytokine production in the
       primary immune reaction of mice. Toxicol Sci 70: 46-54.
Ito, T: Inouye,  K: Nohara, K: Tohyama, C: Fujimaki, H. (2008). TCDD exposure exacerbates
       atopic dermatitis-related inflammation in NC/Nga mice.  Toxicol Lett 177: 31-37.
       http://dx.doi.0rg/10.1016/i.toxlet.2007.12.011.
                                         D-182

-------
Jain. S: Maltepe. E: Lu. MM: Simon. C: Bradfield. CA. (1998). Expression of ARNT, ARNT2,
      HIF1 alpha, HIF2 alpha and Ah receptor mRNAs in the developing mouse. Mech Dev
      73: 117-123.
Jamsa, T; Viluksela, M; Tuomisto, JT; Tuomisto, J: Tuukkanen, J. (2001). Effects of 2,3,7,8-
      tetrachlorodibenzo-p-dioxin on bone in two rat strains with different aryl hydrocarbon
      receptor structures. J Bone Miner Res 16: 1812-1820.
      http://dx.doi.org/10.1359/ibmr.2001.16.10.1812.
Jang. JY: Shin. S: Choi. BI: Park. D: Jeon, JH: Hwang. SY: Kim.  JC: Kim. YB: Nahm, SS.
      (2007). Antiteratogenic effects of alpha-naphthoflavone on 2,3,7,8-tetrachlorodibenzo-p-
      dioxin (TCDD) exposed mice in utero. Reprod Toxicol 24: 303-309.
      http://dx.doi.0rg/10.1016/j.reprotox.2007.08.002.
Jang. JY: Park. D: Shin. S: Jeon, JH: Choi. BI: Joo, SS: Hwang. SY: Nahm, SS: Kim. YB.
      (2008). Antiteratogenic effect of resveratrol in mice exposed in utero to 2,3,7,8-
      tetrachlorodibenzo-p-dioxin. Eur J Pharmacol 591: 280-283.
      http://dx.doi.0rg/10.1016/j.ejphar.2008.05.033.
Janosek, J: Bittner. M: Hilscherova, K: Blaha, L: Giesv, JP: Holoubek, I. (2007). AhR-mediated
      and antiestrogenic activity of humic substances. Chemosphere 67: 1096-1101.
      http://dx.doi.0rg/10.1016/i.chemosphere.2006.ll.045.
Janz, DM; Bellward, GD. (1996). In ovo 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure in three
      avian species. 2. Effects on estrogen receptor and plasma sex steroid hormones during the
      perinatal period. Toxicol Appl Pharmacol 139: 292-300.
      http://dx.doi.org/10.1006/taap.1996.0168.
Jean-Faucher, C: Berger, M: de Turckheim, M:  Veyssiere, G: Jean, C. (1982). The effect of
      preweaning undernutrition upon the sexual development of male mice. Biol Neonate 41:
      45-51.
Jeong, YC: Walker, NJ: Burgin, DE: Kissling, G: Gupta, M: Kupper, L; Birnbaum, LS:
      Swenberg, JA. (2008). Accumulation of MldG DNA adducts after chronic exposure to
      PCBs, but not from acute exposure to polychlorinated aromatic hydrocarbons. Free Radic
      Biol Med 45: 585-591. http://dx.doi.Org/10.1016/i.freeradbiomed.2008.04.043.
Jin. MH: Hong. CH: Lee. HY: Kang,  HJ: Han. SW. (2008a). Enhanced TGF-fil is involved in
      2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) induced oxidative stress in C57BL/6 mouse
      testis. Toxicol Lett 178: 202-209. http://dx.doi.Org/10.1016/i.toxlet.2008.03.015.
Jin. MH: Ko. HK: Hong. CH: Han. SW. (2008b). In utero exposure to 2,3,7,8-
      Tetrachlorodibenzo-p-Dioxin  affects the development of reproductive system in mouse.
      Yonsei Med J 49: 843-850. http://dx.doi.Org/10.3349/ymi.2008.49.5.843.
Jin. MH: Hong. CH: Lee. HY: Kang,  HJ: Han. SW. (2008c). Toxic effects of lactational
      exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on development of male
      reproductive system: Involvement of antioxidants, oxidants, and p53 protein. Environ
      Toxicol TBD: TBD. http://dx.doi.org/10.1002/tox.20466.
Jinno, A; Maruyama, Y; Ishizuka, M: Kazusaka, A; Nakamura, A; Fujita, S. (2006). Induction of
      cytochrome P450-1A by the equine estrogen equilenin, a new endogenous aryl
      hydrocarbon receptor ligand. J Steroid Biochem Mol Biol  98: 48-55.
      http://dx.doi.0rg/10.1016/i.isbmb.2005.07.003.
Johnson. CW: Williams. WC: Copeland. CB: DeVito. MJ:  Smialowicz. RJ. (2000). Sensitivity of
      the SRBC PFC assay versus ELISA for detection of immunosuppression by TCDD and
      TCDD-like congeners. Toxicology 156: 1-11.

                                         D-183

-------
Johnson, KL; Cummings, AM; Birnbaum, LS. (1997). Promotion of endometriosis in mice by
       polychlorinated dibenzo-p-dioxins, dibenzofurans, and biphenyls. Environ Health
       Perspect 105: 750-755.
Johnson. L: Dickerson. R: Safe. SH: Nvberg. CL: Lewis. RP: Welsh. TH. (1992). Reduced
       Leydig cell volume and function in adult rats exposed to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin without a significant effect on spermatogenesis. Toxicology 76: 103-118.
Johnson. L: Wilker.  CE:  Safe. SH: Scott B: Dean. DP: White. PH. (1994). 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin reduces the number, size, and organelle content of Leydig
       cells in adult rat testes. Toxicology 89: 49-65.
Jones, G: Greig, JB.  (1975). Pathological  changes in the liver of mice given 2,3,7,8-
       tetrachlorodibenzo-p-dioxin. Experientia 31: 1315-1317.
Kakeyama, M; Sone, H; Tohyama, C. (2001). Changes in expression of NMD A receptor subunit
       mRNA by perinatal exposure to dioxin. Neuroreport 12: 4009-4012.
Kakeyama, M; Sone, H; Miyabara, Y; Tohyama, C. (2003). Perinatal exposure to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin alters activity-dependent expression of BDNF mRNA in the
       neocortex and male rat sexual behavior in adulthood. Neurotoxicology 24: 207-217.
       http://dx.doi.org/10.1016/S0161-813X(02)00214-0.
Kakeyama, M: Sone, H: Tohyama, C. (2008). Perinatal exposure of female rats to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin induces central precocious puberty in the offspring. J
       Endocrinol 197: 351-358. http://dx.doi.org/10.1677/JOE-08-0062.
Kamath. AB: Xu. H: Nagarkattl PS: Nagarkattl M. (1997). Evidence for the induction of
       apoptosis in thymocytes by 2,3,7,8-tetrachlorodibenzo-p-dioxin in vivo. Toxicol Appl
       Pharmacol 142: 367-377. http://dx.doi.org/10.1006/taap.1996.8049.
Kamath. AB: Camacho. I: Nagarkattl PS: Nagarkattl M. (1999). Role of Fas-Fas ligand
       interactions in 2,3,7,8-tetrachlorodibenzo- p-dioxin (TCDD)-induced immunotoxicity:
       increased resistance of thymocytes from Fas-deficient (Ipr) and Fas ligand-defective (gld)
       mice to TCDD-induced toxicity. Toxicol Appl Pharmacol 160: 141-155.
       http://dx.doi.org/10.1006/taap.1999.8753.
Kattainen, H: Tuukkanen, J: Simanainen,  U: Tuomisto, JT; Kovero, O: Lukinmaa, PL;
       Alaluusua. S: Tuomisto. J: Viluksela. M. (2001). In Utero/Lactational 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin Exposure Impairs Molar Tooth Development in Rats.
       Toxicol Appl Pharmacol 174: 216-224. http://dx.doi.org/10.1006/taap.2001.9216.
Katz. LB:  Theobald. HM: Bookstaff RC: Peterson. RE. (1984).  Characterization of the enhanced
       paw edema response to carrageenan and dextran in 2,3,7,8-tetrachlorodibenzo-p-dioxin-
       treated rats. J Pharmacol Exp Ther 230: 670-677.
Kedderis. LB: Diliberto.  JJ: Linko. P: Goldstein. JA: Birnbaum. LS. (1991). Disposition of
       2,3,7,8-tetrabromodibenzo-p-dioxin and 2,3,7,8-tetrachlorodibenzo-p-dioxin in the rat:
       Biliary excretion  and induction of cytochromes  CYP1A1 and CYP1A2. Toxicol Appl
       Pharmacol 111: 163-172.
Keller. JM: Allen. DE: Davis. CR: Leamy. LJ. (2007a). 2,3,7,8-Tetrachlorodibenzo-p-dioxin
       affects fluctuating asymmetry of molar shape in mice, and an epistatic interaction of two
       genes for molar size. Heredity 98:  259-267. http://dx.doi.org/10.1038/si.hdv.6800928.
Keller. JM: Huang. JC: Huet-Hudson. Y: Leamy. LJ. (2007b). The effects of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin on molar and mandible traits in congenic mice: a test of the
       role of the Ahr locus.  Toxicology 242: 52-62.
       http://dx.doi.0rg/10.1016/i.tox.2007.09.008.

                                        D-184

-------
Keller, JM: Huet-Hudson, YM: Leamy, LJ. (2007c). Qualitative effects of dioxin on molars vary
       among inbred mouse strains. Arch Oral Biol 52: 450-454.
       http://dx.doi.0rg/10.1016/j.archoralbio.2006.10.017.
Keller, JM: Huet-Hudson, Y; Leamy, LJ. (2008a). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       on molar development among non-resistant inbred strains of mice: A geometric
       morphometric analysis. Growth Development and Aging 71: 3-16.
Keller, JM: Zelditch, ML; Huet YM: Leamy, LJ. (2008b). Genetic differences in sensitivity to
       alterations of mandible structure  caused by the teratogen 2,3,7,8-tetrachlorodibenzo-p-
       dioxin. Toxicol Pathol 36: 1006-1013. http://dx.doi.org/10.1177/0192623308327409.
Kellev. SK: Nilsson, CB: Green. MH: Green. JB: Hakansson, H. (1998). Use of model-based
       compartmental analysis to study  effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin on
       vitamin A kinetics in rats. Toxicol Sci 44: 1-13.
       http://dx.doi.org/10.1006/toxs.1998.2467.
Kellev, SK: Nilsson, CB: Green, MH: Green, JB: Hakansson, H. (2000). Mobilization of vitamin
       A stores in rats after administration of 2,3, 7,8-tetrachlorodibenzo-p-dioxin: a kinetic
       analysis. Toxicol Sci 55: 478-484.
Kelling, CK: Christian, BJ: Inhorn, SL: Peterson, RE. (1985). Hypophagia-induced weight loss
       in mice,  rats, and guinea pigs treated with 2,3,7,8-tetrachlorodibenzo-p-dioxin. Fundam
       Appl Toxicol 5: 700-712.
Kelling, CK: Menahan, LA; Peterson, RE.  (1987). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       treatment on mechanical function of the rat heart.  Toxicol Appl Pharmacol 91: 497-501.
Kerkvliet NI: Brauner, JA. (1990). Flow cytometric analysis of lymphocyte subpopulations in
       the spleen and thymus of mice exposed to an acute immunosuppressive dose of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD). Environ Res 52: 146-154.
       http://dx.doi.org/10.1016/S0013-9351(05)80249-X.
Kerkvliet NI: Baecher-Steppan, L: Smith,  BB: Youngberg, JA: Henderson, MC: Buhler, DR.
       (1990). Role of the Ah locus in suppression of cytotoxic T lymphocyte activity by
       halogenated aromatic  hydrocarbons (PCBs and TCDD): structure-activity relationships
       and effects in C57B1/6 mice congenic at the Ah locus. Fundam Appl Toxicol 14: 532-
       541.
Kerkvliet, NI: Oughton, JA. (1993). Acute inflammatory  response to sheep red blood cell
       challenge in mice treated with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD): phenotypic
       and functional analysis of peritoneal exudate cells. Toxicol Appl Pharmacol 119: 248-
       257.
Kerkvliet NI: Baecher-Steppan, L: Shepherd, DM; Oughton, JA: Vorderstrasse, BA: DeKrey,
       GK. (1996). Inhibition of TC-1 cytokine production, effector cytotoxic T lymphocyte
       development and alloantibody production by 2,3,7,8-tetrachlorodibenzo-p-dioxin. J
       Immunol 157:2310-2319.
Kerkvliet NI: Shepherd, DM; Baecher-Steppan, L. (2002).  T lymphocytes are direct, aryl
       hydrocarbon receptor  (AhR)- dependent targets of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       (TCDD): AhR expression in both CD4+ and CD8+ T cells is necessary for full
       suppression of a cytotoxic T lymphocyte response by TCDD. Toxicol Appl Pharmacol
       185: 146-152. http://dx.doi.org/10.1006/taap.2002.9537.
                                         D-185

-------
Khera, KS. (1992). Extraembryonic tissue changes induced by 2,3,7,8-tetrachlorodibenzo-p-
       dioxin and 2,3,4,7,8-pentachlorodibenzofuran with a note on direction of maternal blood
       flow in the labyrinth of C57BL/6N mice. Teratology 45: 611-627.
       http://dx.doi.org/10.1002/tera.1420450606.
Khera, KS: Ruddick, JA. (1973). Perinatal effects and the dominant lethal test in Wistar rats. In
       EH Blair (Ed.), Chlorodioxins: Orign and fate (pp. 70-84). Washington, DC: American
       Chemical Society.
Kim, AH; Kohn, MC: Nyska, A; Walker, NJ. (2003a). Area under the curve as a dose metric for
       promotional responses following 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure. Toxicol
       Appl Pharmacol 191: 12-21. http://dx.doi.org/10.1016/S0041-008X(03)00225-4.
Kim. HJ: Jeong. KS: Park. SJ: Cho. SW: Son. HY: Kim.  SR: Kim. SH: An. MY: Ryu. SY.
       (2003b). Effects of benzo[alpha]pyrene, 2-bromopropane, phenol and 2,3,7,8-
       tetrachlorodibenzo-p-dioxin on IL-6 production in mice after single or repeated exposure.
       In Vivo 17:269-275.
KIMMIG. J: SCHULZ. KH. (1957). [Occupational acne (so-called chloracne) due to chlorinated
       aromatic cyclic ethers]. Dermatologica 115: 540-546.
Kitajima. M: Khan. KN: Fujishita. A: Masuzaki. H: Koji. T: T. I. (2004a). Expression of the
       arylhydrocarbon receptor in the peri-implantation period of the mouse uterus and the
       impact of dioxin on mouse implantation. Arch Histol Cytol 67: 465-474.
Kitajima, M: Khan, KN: Fujishita, A; Masuzaki, H: Ishimaru, T. (2004b). Histomorphometric
       alteration and cell-type specific modulation of arylhydrocarbon receptor and estrogen
       receptor expression by 2,3,7,8-tetrachlorodibenzo-p-dioxin and 17b-estradiol  in mouse
       experimental model  of endometriosis. Reprod Toxicol 18:  793-801.
       http://dx.doi.0rg/10.1016/i.reprotox.2004.04.012.
Kitamura, N; Wong, P; Matsumura, F. (2006). Mechanistic investigation on the cause for
       reduced toxicity of TCDD in wa-1 homozygous TGFa mutant strain of mice as compared
       its matching wild-type counterpart, C57BL/6J mice. J Biochem Mol Toxicol 20: 151-158.
       http://dx.doi.org/10.1002/ibt.20131.
Kitchin, KT: Woods. JS. (1979). 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) effects on hepatic
       microsomal cytochrome P-448-mediated enzyme  activities. Toxicol Appl Pharmacol 47:
       537-546. http://dx.doi.org/10.1016/0041-008X(79)90524-6.
Kleeman, JM: Moore, RW:  Peterson, RE. (1990). Inhibition of testicular steroidogenesis in
       2,3,7,8-tetrachlorodibenzo-p-dioxin-treated rats: evidence that the key lesion occurs prior
       to or during pregnenolone formation. Toxicol Appl Pharmacol 106: 112-125.
Ko, K; Theobald, HM: Peterson, RE. (2002). In utero and lactational exposure to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin in the C57BL/6J mouse prostate: lobe-specific effects on
       branching morphogenesis. Toxicol Sci 70: 227-237.
Ko. K: Theobald. HM: Moore. RW: Peterson. RE. (2004). Evidence that inhibited prostatic
       epithelial bud formation in 2,3,7,8-tetrachlorodibenzo-p-dioxin-exposed C57BL/6J fetal
       mice is not due to interruption of androgen signaling in the urogenital sinus. Toxicol Sci
       79: 360-369.
Kociba, RJ: Keeler, PA: Park. CN: Gehring, PJ. (1976). 2,3,7,8-Tetrachlorodibenzo-p-dioxin
       (TCDD): Results of a 13-week oral toxicity study in rats. Toxicol Appl Pharmacol 35:
       553-574. http://dx.doi.org/10.1016/0041-008X(76)90078-8.
                                         D-186

-------
Kociba. RJ: Keves. DG: Beyer. JE: Carreon. RM: Wade. CE: Dittenber. DA: Kalnins. RP:
       Frauson. LE: Park. CN: Barnard. SD: Hummel RA: Humiston. CG. (1978). Results of a
       two-year chronic toxicity and oncogenicity study of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       in rats. Toxicol Appl Pharmacol 46:  279-303. http://dx.doi.org/10.1016/0041-
       008X(78)90075-3.
Kopec, AK; Boverhof, PR; Burgoon, LD; Ibrahim-Aibo, D:  Harkema, JR: Tashiro, C: Chittim,
       B; Zacharewski, TR. (2008). Comparative toxicogenomic examination of the hepatic
       effects of PCB126 and TCDD in immature, ovariectomized C57BL/6 mice. Toxicol Sci
       102: 61-75. http://dx.doi.org/10.1093/toxsci/kfm289.
Kopf, PG: Huwe, JK; Walker, MK. (2008).  Hypertension, cardiac hypertrophy, and impaired
       vascular relaxation induced by 2,3,7,8-tetrachlorodibenzo-p-dioxin are associated with
       increased superoxide. Cardiovasc Toxicol 8: 181-193. http://dx.doi.org/10.1007/sl2012-
       008-9027-x.
Korenaga, T;  Fukusato, T; Ohta, M; Asaoka, K; Murata, N; Arima, A; Kubota, S. (2007). Long-
       term effects of subcutaneously injected 2,3,7,8-tetrachlorodibenzo-p-dioxin on the liver
       of rhesus monkeys. Chemosphere 67: S399-S404.
       http://dx.doi.0rg/10.1016/j.chemosphere.2006.05.135.
Korte, M: Stahlmann, R; Neubert, D. (1990). Induction of hepatic monooxygenases in female
       rats and offspring in correlation with TCDD tissue concentrations after single treatment
       during pregnancy. Chemosphere 20: 1193-1198. http://dx.doi.org/10.1016/0045-
       6535(90)90244-N.
Kozak, KR; Abbott B; Hankinson, O. (1997). ARNT-deficient mice and placental
       differentiation. DevBiol 191: 297-305. http://dx.doi.org/10.1006/dbio.1997.8758.
Kransler, KM; McGarrigle, BP; Olson, JR. (2007a). Comparative developmental toxicity of
       2,3,7,8-tetrachlorodibenzo-p-dioxin  in  the hamster, rat and guinea pig. Toxicology 229:
       214-225. http://dx.doi.0rg/10.1016/i.tox.2006.10.019.
Kransler. KM: Tonuccl DA: McGarrigle. BP: Napoll JL: Olson. JR. (2007b). Gestational
       exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin alters retinoid homeostasis in maternal
       and perinatal tissues of the Holtzman rat. Toxicol Appl Pharmacol 224: 29-38.
       http://dx.doi.0rg/10.1016/i.taap.2007.06.006.
Kransler. KM: McGarrigle. BP: Russell RJ: Olson. JR. (2008). Effects of Helicobacter infection
       on developmental toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin in Holtzman rats. Lab
       Anim  37: 171-175. http://dx.doi.org/10.1038/laban0408-171.
Kransler, KM; McGarrigle, BP: Swartz,  DP; Olson, JR. (2009). Lung development in the
       Holtzman rat is adversely affected by gestational exposure to 2,3,7,8-tetrachlorodibenzo-
       p-dioxin. Toxicol Sci 107: 498-511.  http://dx.doi.org/10.1093/toxsci/kfn235.
Kronenberg, S: Lai, Z; Esser, C. (2000).  Generation of alphabeta T-cell receptor+ CD4- CD8+
       cells in major histocompatibility  complex class I-deficient mice upon activation of the
       aryl hydrocarbon receptor by 2,3,7,8-tetrachlorodibenzo-p-dioxin. Immunology 100: 185-
       193.
Krowke, R: Chahoud, I; Baumann-Wilschke, I; Neubert, D. (1989). Pharmacokinetics and
       biological activity of 2,3,7,8-tetrachlorodibenzo-p-dioxin 2: pharmacokinetics in rats
       using  a loading-dose/maintenance-dose regime with high doses. Arch Toxicol 63: 356-
       360. http://dx.doi.org/10.1007/BF00303123.
                                         D-187

-------
Kruger, N; Neubert, B; Helge, H; Neubert D. (1990). Induction of caffeine-demethylations by
       2,3,7,8-TCDD in marmoset monkeys measured with a 14CO2 breath-test. Chemosphere
       20: 1173-1176. http://dx.doi.org/10.1016/0045-6535(90)90240-1.
Kuchiiwa, S: Cheng, SB; Nagatomo, I; Akasaki, Y; Uchida, M; Tominaga, M; Hashiguchi, W;
       Kuchiiwa, T. (2002). In utero and lactational exposure to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin decreases serotonin-immunoreactive neurons in raphe nuclei of male mouse
       offspring. Neurosci Lett 317: 73-76. http://dx.doi.org/10.1016/S0304-3940(01)02434-X.
Kwon, YI; Yeon, JD; Oh, SM; Chung, KH. (2004). Protective effects of ursodeoxycholic acid
       against 2,3,7,8-tetrachlorodibenzo-p-dioxin-induced testicular damage in mice. Toxicol
       Appl Pharmacol 194: 239-247. http://dx.doi.Org/10.1016/i.taap.2003.09.024.
Laiosa, MD; Lai, ZW: Thurmond, TS: Fiore, NC: DeRossi, C; Holdener, BC: Gasiewicz, TA;
       Silverstone, AE. (2002). 2,3,7,8-tetrachlorodibenzo-p-dioxin causes alterations in
       lymphocyte development and thymic atrophy in hemopoietic chimeras generated from
       mice deficient in ARNT2. Toxicol Sci 69: 117-124.
LaKind, JS: Berlin. CM: Park. CN: Naiman, DO: Gudka, NJ. (2000). Methodology for
       characterizing distributions of incremental body burdens of 2,3,7,8-TCDD and DDE from
       breast milk in North American nursing infants. J Toxicol Environ Health A 59: 605-639.
       http://dx.doi.org/10.1080/009841000156628.
Lakshman, MR: Campbell BS: Chirtel SJ: Ekarohita. N. (1988). Effects of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) on de novo fatty acid and cholesterol synthesis in
       the rat.  Lipids 23: 904-906.
Lakshman. MR: Chirtel SJ: Chambers. LL: Coutlakis. PJ. (1989). Effects of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin on lipid synthesis and lipogenic enzymes  in the rat. J
       Pharmacol Exp Ther 248: 62-66.
Lakshman. MR: Ghosh. P: Chirtel SJ. (1991). Mechanism of action of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin on intermediary metabolism in the rat. J Pharmacol Exp Ther
       258: 317-319.
Latchoumvcandane. C: Chitra. KC: Mathur. PP.  (2002a). The effect of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin on the antioxidant system in mitochondrial and microsomal
       fractions of rattestis. Toxicology 171: 127-135. http://dx.doi.org/10.1016/S0300-
       483X0)1)00563-7.
Latchoumycandane, C: Mathur, PP. (2002). Effects of vitamin E on reactive oxygen species-
       mediated 2,3,7,8-tetrachlorodibenzo-p-dioxin toxicity in rat testis. J Appl  Toxicol 22:
       345-351. http://dx.doi.org/10.1002/iat.866.
Latchoumycandane, C: Chitra, C: Mathur, P. (2002b). Induction of oxidative stress  in rat
       epididymal sperm after exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Arch Toxicol
       76: 113-118. http://dx.doi.org/10.1007/s00204-001-0308-4.
Latchoumycandane, C: Chitra, KC: Mathur, PP.  (2003). 2,3,7,8-Tetrachlorodibenzo-p-dioxin
       (TCDD) induces oxidative stress in the epididymis and epididymal sperm of adult rats.
       Arch Toxicol 77: 280-284. http://dx.doi.org/10.1007/s00204-003-0439-x.
Laurent C: Feidt C: Grova. N: Mpassl D: Lichtfouse. E: Laurent F: Rvchen. G. (2002). Portal
       absorption of 14C after ingestion of spiked milk with 14C-phenanthrene, 14C-
       benzo[a]pyrene or 14C-TCDD in growing pigs. Chemosphere 48: 343-348.
       http://dx.doi.org/10.1016/S0045-6535(02)00145-5.
                                        D-188

-------
Lawrence, BP: Warren, TK; Luong, H. (2000). Fewer T lymphocytes and decreased pulmonary
       influenza virus burden in mice exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). J
       Toxicol Environ Health A 61: 39-53.
Lawrence, BP: Vorderstrasse, BA. (2004). Activation of the aryl hydrocarbon receptor
       diminishes the memory response to homotypic influenza virus infection but does not
       impair host resistance. Toxicol Sci 79: 211-213. http://dx.doi.org/10.1093/toxsci/kfh094.
Lawrence. BP: Roberts. AD: Neumiller. JJ: Cundiff, JA: Woodland. PL. (2006). Aryl
       hydrocarbon receptor activation impairs the priming but not the recall of influenza virus-
       specific CD8+ T cells in the lung. J Immunol 177: 5819-5828.
Lee. JH: Sul D: Oh. E: Jung. WW: Hwang. KW: Hwang. TS: Lee. KC: Won. NH. (2007). Panax
       ginseng effects on DNA damage, CYP1 Al expression and histopathological changes in
       testes of rats exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Food Chem Toxicol 45:
       2237-2244. http://dx.doi.Org/10.1016/i.fct.2007.05.019.
Lensu, S: Miettinen, R; Pohjanvirta, R: Linden, J: Tuomisto, J. (2006). Assessment by c-Fos
       immunostaining of changes in brain neural activity induced by 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) and leptin in rats. Basic Clin Pharmacol Toxicol 98:
       363-371. http://dx.doi.org/10.1111/i.l742-7843.20Q6.pto_276.x.
Lewis, BC: Hudgins, S: Lewis, A; Schorr, K; Sommer, R: Peterson,  RE; Flaws, JA: Furth, PA.
       (2001). In utero and lactational treatment with 2,3,7,8-tetrachlorodibenzo-p-dioxin
       impairs mammary gland differentiation but does not block the response to exogenous
       estrogen in the postpubertal female rat. Toxicol Sci 62: 46-53.
Li, B; Liu, HY; Dai, LJ: Lu, JC: Yang, ZM; Huang, L. (2006). The early embryo loss caused by
       2,3,7,8-tetrachlorodibenzo-p-dioxin may be related to the accumulation of this compound
       in the uterus. Reprod Toxicol 21: 301-306.
       http://dx.doi.0rg/10.1016/j.reprotox.2005.09.008.
Li, X; Johnson, DC: Rozman, KK. (1995a). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       (TCDD) on estrous cycllicity and ovulation in female Sprague-Dawley rats. Toxicol Lett
       78: 219-222.
Li, X; Johnson, DC: Rozman, KK. (1995b). Reproductive effects of 2,3,7,8-tetrachlorodibenzo-
       p-dioxin (TCDD) in  female rats: Ovulation, hormonal regulation, and possible
       mechanism(s). Toxicol Appl Pharmacol 133: 321-327.
       http://dx.doi.org/10.1006/taap.1995.1157.
Li, X; Weber, LW: Rozman, KK. (1995c). Toxicokinetics of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       in female Sprague-Dawley rats including placental and lactational transfer to fetuses and
       neonates. Fundam Appl Toxicol 27: 70-76.
Li. X: Johnson. DC: Rozman. KK. (1997). 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
       increases release of luteinizing hormone and follicle-stimulating hormone from the
       pituitary of immature female rats in vivo and in vitro. Toxicol Appl Pharmacol 142: 264-
       269. http://dx.doi.org/10.1006/taap.1996.8044.
Lilienthal, H;  Winneke, G. (1991). Sensitive periods for behavioral toxicity of polychlorinated
       biphenyls: Determination by cross-fostering in rats.  Fundam  Appl Toxicol 17: 368-375.
       http://dx.doi.org/10.1016/0272-0590(91)90226-T.
Lilienthal H:  Weinand-Harer, A: Winterhoff, H: Winneke. G. (1997). Effects of maternal
       exposure to 3,3',4,4'-tetrachlorobiphenyl or propylthiouracil in rats trained to discriminate
       apomorphine from saline.  Toxicol Appl Pharmacol 146: 162-169.
       http://dx.doi.org/10.1006/taap.1997.8245.

                                         D-189

-------
Lim, CK; Danton, M; Clothier, B; Smith, AG. (2006). Dihydroxy-, hydroxyspirolactone-, and
       dihydroxyspirolactone-urochlorins induced by 2,3,7,8-tetrachlorodibenzo-p-dioxin in the
       liver of mice. Chem Res Toxicol 19: 1660-1667.
Lin. FH: Stohs. SJ: Birnbaum. LS: Clark. G: Lucier. GW: Goldstein. JA. (1991). The effects of
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on the hepatic estrogen and glucocorticoid
       receptors in congenic strains of Ah responsive and Ah nonresponsive C57BL/6J mice.
       Toxicol Appl Pharmacol 108:  129-139.
Lin. TM: Ko. K: Moore. RW: Buchanan. PL: Cooke. PS: Peterson. RE. (2001). ROLE of the
       ARYL HYDROCARBON receptor in the development of control and 2,3,7,8-
       TETRACHLORODIBENZO-p-DIOXIN-EXPOSED male mice. J Toxicol Environ
       Health A 64: 327-342. http://dx.doi.org/10.1080/152873901316981312.
Lin, TM; Simanainen, U; Moore, RW: Peterson, RE. (2002a). Critical windows of vulnerability
       for effects of 2,3,7,8-Tetrachlorodibenzo-p-dioxin on prostate and seminal vesicle
       development in C57BL/6 mice. Toxicol Sci 69: 202-209.
Lin. TM: Ko. K: Moore. RW: Simanainen. U: Oberlev. TD: Peterson. RE. (2002b). Effects of
       Aryl Hydrocarbon receptor null mutation and in utero and lactational 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin exposure on prostate and seminal vesicle development in
       C57BL/6 mice. Toxicol Sci 68: 479-487.
Linden. J: Korkalainen. M: Lensu. S: Jo. T: Pohjanvirta. R. (2005). Effects of 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin (TCDD) and Leptin on hypothalamic mRNA expression of
       factors participating in food intake regulation in a TCDD-sensitive and a TCDD-resistant
       rat strain. J Biochem Mol Toxicol 19: 139-148. http://dx.doi.org/10.1002/ibt.20065.
Liu. YR: Tang. NJ: Ren. PL.  (2003). Induction of aryl hydrocarbon receptor CYP1A1 mRNA by
       2,3,7,8-tetrachlorodibenzo-p-dioxin in rat liver. Zhonghua Lao Dong Wei Sheng Zhi Ye
       BingZaZhi 21: 417-419.
Loertscher, JA: Lin, TM: Peterson, RE; Allen-Hoffmann, BL. (2002). In Utero exposure to
       2,3,7,8-tetrachlorodibenzo-p-dioxin causes accelerated terminal differentiation in fetal
       mouse skin. Toxicol Sci 68: 465-472.
Lucier. GW: McDaniel OS: Hook. GE: Fowler. BA: Sonawane. BR: Faeder. E. (1973). TCDD-
       induced changes in rat liver microsomal enzymes. Environ Health Perspect 5: 199-209.
Lucier, GW: McDaniel, OS: Hook, GE. (1975a). Nature of the enhancement of hepatic uridine
       diphosphate glucuronyltransferase activity by 2,3,7,8-tetrachlorodibenzo-p-dioxin in rats.
       Biochem Pharmacol 24: 325-334.
Lucier, GW: Sonawane, BR: McDaniel OS: Hook, GE. (1975b). Postnatal stimulation of hepatic
       microsomal enzymes following administration of TCDD to pregnant rats. Chem Biol
       Interact 11: 15-26.
Lucier. GW: Rumbaugh. RC: McCoy. Z: Hass. R: Harvan.  D: Albro. P. (1986). Ingestion of soil
       contaminated with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) alters hepatic enzyme
       activities in rats. Fundam Appl Toxicol 6: 364-371.
Lucier, GW: Tritscher, A; Goldsworthy, T; Foley, J: Clark, G: Goldstein, J: Maronpot R.
       (1991). Ovarian hormones enhance 2,3,7,8-tetrachlorodibenzo-p-dioxin-mediated
       increases in cell proliferation and preneoplastic foci in a two-stage model for rat
       hepatocarcinogenesis. Cancer Res 51: 1391-1397.
                                        D-190

-------
Luebeck, EG: Buchmann, A; Stinchcombe, S: Moolgavkar, SH; Schwarz, M. (2000). Effects of
       2,3,7,8-tetrachlorodibenzo-p-dioxin on initiation and promotion of GST-P-positive foci in
       rat liver: A quantitative analysis of experimental data using a stochastic model. Toxicol
       Appl Pharmacol 167: 63-73. http://dx.doi.org/10.1006/taap.2000.8980.
Luebke. RW: Copeland. CB: Diliberto. JJ: Akubue. PI: Andrews. PL: Riddle. MM: Williams.
       WC: Birnbaum, LS. (1994). Assessment of host resistance to Trichinella spiralis in mice
       following preinfection exposure to 2,3,7,8-TCDD. Toxicol Appl Pharmacol  125: 7-16.
       http://dx.doi.org/10.1006/taap.1994.1043.
Luebke, RW: Copeland, CB: Andrews, PL. (1995). Host resistance to Trichinella spiralis
       infection in rats exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Fundam Appl
       Toxicol 24: 285-289.
Luebke, RW: Copeland, CB: Andrews, PL. (1999). Effects of aging on resistance to Trichinella
       spiralis infection in rodents exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicology
       136: 15-26.
Luebke, RW: Copeland, CB: Daniels, M; Lambert, AL; Gilmour, MI.  (2001).  Suppression of
       allergic immune responses to house dust mite (HDM) in rats exposed to 2,3,7,8-TCDD.
       Toxicol Sci 62: 71-79.
Luebke. RW: Copeland. CB: Bishop. LR: Daniels. MJ: Gilmour. MI. (2002). Mortality in
       dioxin-exposed mice infected with influenza: mitochondrial toxicity (reye's-like
       syndrome) versus enhanced inflammation as the mode of action. Toxicol Sci 69: 109-
       116.
Lundberg, K; Dencker, L; Gronvik, KO. (1990). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       (TCDD) treatment in vivo on thymocyte functions in mice after activation in vitro.
       International  Journal of Immunopharmacology 12: 459-466.
Luster, MI; Boorman, GA: Dean, JH; Harris,  MW: Luebke, RW: Padarathsingh, ML; Moore, JA.
       (1980). Examination of bone marrow, immunologic parameters and host susceptibility
       following pre- and postnatal exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD).
       International  Journal of Immunopharmacology 2: 301-310.
Luster, MI; Hong, LH; Boorman, GA:  Clark,  G: Hayes, HT; Greenlee, WF; Dold, K; Tucker,
       AN. (1985). Acute myelotoxic  responses in mice exposed to 2,3,7,8-tetrachlorodibenzo-
       p-dioxin (TCDD). Toxicol Appl Pharmacol 81:  156-165.
Ma. X: Idle. JR: Malfatti, MA: Krausz, KW: Nebert DW: Chen. CS: Felton, JS: Waxman, DJ:
       Gonzalez, FJ. (2007). Mouse lung CYP1A1 catalyzes the metabolic activation of 2-
       amino-l-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP). Carcinogenesis 28: 732-737.
Mably, TA: Theobald, HM; Ingall, GB; Peterson, RE. (1990). Hypergastrinemia is associated
       with decreased gastric acid secretion in 2,3,7,8-tetrachlorodibenzo-p-dioxin-treated rats.
       Toxicol Appl Pharmacol 106: 518-528.
Mably, TA: Moore, RW: Bjerke, PL; Peterson, RE. (1991). The male reproductive system is
       highly sensitive to in utero and  lactational TCDD exposure (pp. 69-78). Cold Spring
       Harbor, NY:  Cold Spring Harbor Laboratory Press.
MacLusky, NJ: Brown, TJ: Schantz, S: Seo, BW: Peterson, RE. (1998). Hormonal interactions in
       the effects of halogenated aromatic hydrocarbons on the developing brain. Toxicol Ind
       Health 14: 185-208.
Madhukar, BV:  Brewster, DW: Matsumura, F. (1984). Effects of in vivo-administered 2,3,7,8-
       tetrachlorodibenzo-p-dioxin on receptor binding of epidermal growth factor in the hepatic
       plasma membrane of rat, guinea pig, mouse, and hamster. PNAS 81: 7407-7411.

                                         D-191

-------
Madhukar. BV: Ebner. K: Matsumura. F: Bombick. DW: Brewster. DW: Kawamoto. T. (1988).
       2,3,7,8-Tetrachlorodibenzo-p-dioxin causes an increase in protein kinases associated with
       epidermal growth factor receptor in the hepatic plasma membrane. J Biochem Toxicol 3:
       261-277.
Mally, A;  Chipman, JK. (2002). Non-genotoxic carcinogens: Early effects on gap junctions, cell
       proliferation and apoptosis in the rat. Toxicology 180: 233-248.
Mann, PC. (1997). Selected lesions of dioxin in laboratory rodents. Toxicol Pathol 25:  72-79.
Mantovani, A; Vecchi, A; Luini, W; Sironi, M; Candiani, GP: Spreafico, F; Garattini, S. (1980).
       Effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin on macrophage and natural killer cell-
       mediated cytotoxicity in mice.  32: 200-204.
Markowski, VP: Zareba, G: Stern, S: Cox, C; Weiss, B. (2001). Altered operant responding for
       motor reinforcement and the determination of benchmark doses following perinatal
       exposure to low-level  2,3,7,8-tetrachlorodibenzo-p-dioxin.  Environ Health Perspect 109:
       621-627.
Markowski, VP: Cox, C: Preston, R; Weiss, B. (2002).  Impaired cued delayed alternation
       behavior in adult rat offspring following exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin
       on gestation day 15. Neurotoxicol Teratol 24: 209-218.
Marks, GS. (1985). Exposure to toxic agents: the heme biosynthetic pathway and hemoproteins
       as indicator. Crit Rev Toxicol 15: 151-179.
       http://dx.doi.org/10.3109/10408448509029323.
Marks, TA: Staples, RE. (1980). Teratogenic evaluation of the  symmetrical isomers of
       hexachlorobiphenyl (HCB) in the mouse [Abstract]. Teratology 21: 54A.
Marks, TA: Kimmel, GL:  Staples, RE. (1981). Influence of symmetrical polychlorinated
       biphenyl isomers on embryo and fetal development in mice. I. Teratogenicity of 3, 3', 4,
       4',  5, 5',-hexachlorobiphenyl. Toxicol Appl Pharmacol 61: 269-276.
Maronpot RR: Montgomery, CA: Boorman, GA: McConnell, EE. (1986). National Toxicology
       Program nomenclature for hepatoproliferative lesions of rats. Toxicol Pathol  14: 263-273.
Maronpot RR: Pitot HC: Peraino, C. (1989). Use of rat liver altered focus models for testing
       chemicals that have completed two-year carcinogenicity studies. Toxicol Pathol 17: 651-
       662.
Maronpot RR: Foley, JF:  Takahashi, K: Goldsworthy,  T; Clark, G: Tritscher, A; Portier, C:
       Lucier, G. (1993).  Dose response for TCDD promotion  of hepatocarcinogenesis in rats
       initiated with DEN: histologic, biochemical, and cell proliferation endpoints. Environ
       Health Perspect 101: 643-642.
Massart, F: Meucci, V. (2007). Environmental thyroid toxicants and child endocrine health.
       Pediatr Endocrinol Rev 5: 500-509.
Matsumura, F: Enan, E: Dunlap, DY; Pinkerton, KE; Peake, J.  (1997). Altered in vivo toxicity of
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in C-SRC  deficient mice. Biochem
       Pharmacol 53: 1397-1404.
Max,  SR:  Silbergeld, EK.  (1987). Skeletal muscle glucocorticoid receptor and glutamine
       synthetase activity in the wasting syndrome in rats treated with 2,3,7,8-
       tetrachlorodibenzo-p-dioxin. Toxicol Appl Pharmacol 87: 523-527.
McConnell EE: Moore. JA: Dalgard. DW. (1978). Toxicity of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin in rhesus monkeys (Macaca mulatta) following a single oral dose. Toxicol Appl
       Pharmacol 43: 175-187.
                                         D-192

-------
McConnell, EE; Moore, JA. (1979). Toxicopathology characteristics of the halogenated
       aromatics. Ann N Y Acad Sci 320: 138-150.
Mcgrath, LF; Cooper, KR; Georgopoulos, P; Gallo, MA. (1995). Alternative models for low
       dose-response analysis of biochemical and immunological endpoints for
       Tetrachlorodibenzo-p-dioxin. Regul Toxicol Pharmacol 21: 382-396.
       http://dx.doi.org/10.1006/rtph.1995.1053.
McKinley, MK; Kedderis, LB; Birnbaum, LS. (1993). The effect of pretreatment on the biliary
       excretion of 2,3,7,8-tetrachlorodibenzo-p-dioxin, 2,3,7,8-tetrachlorodibenzofuran, and
       3,3',4,4'-tetrachlorobiphenyl in the rat. Fundam Appl Toxicol 21: 425-432.
McKinney, JD: Chae, K; Oatley, SJ: Blake, CC. (1985). Molecular interactions of toxic
       chlorinated  dibenzo-p-dioxins and dibenzofurans with thyroxine binding prealbumin. J
       Med Chem  28: 375-381. http://dx.doi.org/10.1021/im00381a018.
McNulty, WP. (1977). Toxicity of 2,3,7,8-tetrachlorodibenzo-p-diozin for rhesus monkeys: brief
       report. Bull  Environ Contam Toxicol 18: 108-109.
McNulty, WP: Nielsen-Smith, KA; Lay, JO, Jr; Lippstreu, PL; Kangas, NL; Lyon, PA; Gross,
       ML. (1982). Persistence of TCDD in monkey adipose tissue. Food Chem Toxicol 20:
       985-986.
Mcnultv, WP. (1984). Fetotoxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) for rhesus
       macaques (Macaca mulatta). Am J Primatol 6: 41-47.
       http://dx.doi.org/10.1002/aip.1350060105.
McNultv, WP. (1985). Toxicity and fetotoxicity of TCDD, TCDF and PCB isomers in rhesus
       macaques (Macaca mulatta). Environ Health Perspect 60: 77-88.
Mebus, CA; Reddy, VR: Piper, WN. (1987). Depression of rat testicular 17-hydroxylase and
       17,20-lyase after administration of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD).
       Biochem  Pharmacol 36: 727-731.
Meulenbelt J: de Vries, I. (2005). Toxiciteit van dioxinen voor de mens. Ned Tijdschr Geneeskd
       149: 168-171.
Meyer. KM.  (2002). Incidence of CTCL in Vietnam veterans. Dermatol Nurs 14: 42, 45, 52.
Michalek, JE: Akhtar. FZ: Longnecker. MP: Burton. JE. (200la). Relation of serum 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) level to hematological examination results in
       veterans of  Operation Ranch Hand. Arch Environ Health 56: 396-405.
       http://dx.doi.org/10.1080/00039890109604474.
Michalek, JE: Ketchum, NS: Longnecker, MP. (200Ib). Serum dioxin and hepatic abnormalities
       in veterans of Operation Ranch Hand. Ann Epidemiol 11: 304-311.
       http://dx.doi.org/10.1016/S1047-2797(W)00218-0.
Michalek. JE: Akhtar. FZ: Arezzo, JC: Garabrant DH: Albers,  JW. (200Ic). Serum dioxin and
       peripheral neuropathy in veterans of Operation Ranch Hand. Neurotoxicology 22: 479-
       490.
Michalek, JE: Pavuk, M. (2008). Diabetes and cancer  in veterans of Operation Ranch Hand after
       adjustment for calendar period, days of spraying, and time spent in Southeast Asia. J
       Occup Environ Med 50: 330-340. http://dx.doi.org/10.1097/JOM.Ob013e31815f889b.
Miettinen, HM; Alaluusua, S; Tuomisto, J: Viluksela,  M. (2002). Effect of in utero and
       lactational 2,3,7,8-Tetrachlorodibenzo-p-dioxin exposure on rat molar development: The
       role of exposure time. Toxicol  Appl Pharmacol 184: 57-66.
       http://dx.doi.org/10.1006/taap.2002.9490.
                                        D-193

-------
Miettinen, HM; Huuskonen, H; Partanen, AM; Miettinen, P; Tuomisto, JT; Pohjanvirta, R:
       Tuomisto, J. (2004). Effects of epidermal growth factor receptor deficiency and 2,3,7,8-
       tetrachlorodibenzo-p-dioxin on fetal development in mice. Toxicol Lett 150: 285-291.
       http://dx.doi.0rg/10.1016/i.toxlet.2004.02.009.
Miettinen, HM: Pulkkinen, P; Jamsa, T; Koistinen, J: Simanainen, U; Tuomisto, J: Tuukkanen, J:
       Viluksela, M. (2005). Effects of in utero and lactational TCDD exposure on bone
       development in differentially sensitive rat lines. Toxicol Sci 85: 1003-1012.
       http://dx.doi.org/10.1093/toxsci/kfil36.
Miettinen, HM: Sorvari, R; Alaluusua, S; Murtomaa, M; Tuukkanen, J: Viluksela, M. (2006).
       The Effect of Perinatal TCDD exposure on caries susceptibility in rats. Toxicol Sci 91:
       568-575. http://dx.doi.org/10.1093/toxsci/kfil58.
Miller, CP: Birnbaum, LS. (1986). Teratologic evaluation of hexabrominated naphthalenes in
       C57BL/6N mice. Fundam Appl Toxicol 7: 398-405.
Miller, RW. (1985). Congenital PCB poisoning: a reevaluation. Environ Health Perspect 60:
       211-214.
Mimura, J: Yamashita, K; Nakamura, K; Morita, M: Takagi, TN: Nakao, K; Etna, M: Sogawa,
       K; Yasuda, M: Katsuki, M: Fujii-Kuriyama, Y. (1997). Loss of teratogenic response to
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in mice lacking the Ah (dioxin) receptor.
       Genes Cells 2: 645-654.
Mitchell KA: Lawrence. BP. (2003a). Exposure  to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
       renders influenza virus-specific CD8+ T cells hyporesponsive to antigen. Toxicol Sci 74:
       74-84. http://dx.doi.org/10.1093/toxsci/kfgllO.
Mitchell, KA; Lawrence, BP. (2003b). T cell receptor transgenic mice provide novel insights
       into understanding cellular targets of TCDD: suppression of antibody production, but not
       the response of CD8A+ T cells, during infection with influenza virus. Toxicol Appl
       Pharmacol 192: 275-286. http://dx.doi.org/10.1016/S0041-008X(03)00297-7.
Mitchell KA: Lockhait CA: Huang. G: Elferink, CJ. (2006). Sustained aryl hydrocarbon
       receptor activity attenuates liver regeneration. Mol Pharmacol 70:  163-170.
       http://dx.doi.org/10.1124/mol.106.023465.
Mitrou, PI; Dimitriadis, G: Raptis, SA. (2001). Toxic effects of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin and related compounds. Eur J Intern Med 12: 406-411.
Mitsui, T: Sugiyama, N; Maeda, S: Tohyama, C; Arita, J. (2006). Perinatal exposure to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin suppresses contextual fear conditioning-accompanied
       activation of cyclic AMP response element-binding protein in the hippocampal CA1
       region of male rats. Neurosci Lett 398: 206-210.
       http://dx.doi.0rg/10.1016/i.neulet.2005.12.087.
Mittlen JC: Ertel NH: Peng. RX: Yang. CS: Kiernan. T. (1984). Changes in testosterone
       hydroxylase activity in rattestis following administration of 2,3,7,8-tetrachlorodibenzo-
       p-dioxin. Ann N Y Acad Sci 438: 645-648.
Mizuyachi, K: Son, DS: Rozman, KK; Terranova, PF. (2002). Alteration in ovarian gene
       expression in response to 2,3,7,8-tetrachlorodibenzo-p-dioxin: Reduction of
       cyclooxygenase-2 in the blockage of ovulation. Reprod Toxicol 16: 299-307.
Mocarelli, P: Gerthoux, PM; Ferrari, E; Patterson Jr, DG: Kieszak, SM; Brambilla, P: Vincoli,
       N: Signorini, S: Tramacere, P: Carreri, V: Sampson, EJ: Turner, WE. (2000). Paternal
       concentrations of dioxin and sex ratio of offspring. Lancet 355:  1858-1863.
       http://dx.doi.org/10.1016/S0140-6736(00)02290-X.

                                         D-194

-------
Mocarelli. P. (2001). Seveso: A teaching story. Chemosphere 43: 391-402.
       http://dx.doi.org/10.1016/S0045-6535(00)00386-6.
Mocarelli, P; Gerthoux, PM; Patterson, DG, Jr; Milani, S: Limonata, G: Bertona, M; Signorini,
       S: Tramacere, P; Colombo, L; Crespi, C: Brambilla, P; Sarto, C: Carreri, V:  Sampson,
       EJ: Turner, WE; Needham, LL. (2008). Dioxin exposure, from infancy through puberty,
       produces endocrine disruption and affects human semen quality. Environ Health Perspect
       116: 70-77. http://dx.doi.org/10.1289/ehp.10399.
Moennikes, O: Loeppen, S: Buchmann, A; Andersson, P: Ittrich, C: Poellinger, L; Schwarz, M.
       (2004). A constitutively active dioxin/aryl hydrocarbon receptor promotes
       hepatocarcinogenesis in mice. Cancer Res 64: 4707-4710.
       http://dx.doi.org/10.1158/0008-5472.can-03-0875.
Moolgavkar, SH; Luebeck, EG: Buchmann, A; Bock, KW. (1996). Quantitative analysis of
       enzyme-altered liver foci in rats initiated with diethylnitrosamine and promoted with
       2,3,7,8-tetrachlorodibenzo-p-dioxin or 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin.
       Toxicol Appl Pharmacol 138: 31-42. http://dx.doi.org/10.1006/taap.1996.0094.
Moon. BH: Hong. CG: Kim. SY: Kim. HJ: Shin. SK: Kang, S: Lee. KJ: Kim. YK: Lee. MS:
       Shin, KH.  (2008). A single administration of 2,3,7,8-tetrachlorodibenzo-p-dioxin that
       produces reduced food and water intake induces long-lasting expression of corticotropin-
       releasing factor, arginine vasopressin, and proopiomelanocortin in rat brain.  Toxicol Appl
       Pharmacol 233: 314-322. http://dx.doi.Org/10.1016/i.taap.2008.09.001.
Moon. DG: Lee. KC: Kim. YW: Park. HS: Cho, HY: Kim. JJ. (2004). Effect of TCDD on corpus
       cavernosum histology and smooth muscle physiology. Int J Impot Res 16: 224-230.
       http://dx.doi.org/10.1038/sj.ijir.39010603901060.
Moore, JA: Gupta, BN: Zinkl, JG:  Vos, JG. (1973). Postnatal effects of maternal exposure to
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Environ Health Perspect 5: 81-85.
Moore, JA: Harris, MW: Albro, PW. (1976). Tissue distribution of (14C)tetrachlorodibenzo-p-
       dioxin in pregnant and neonatal rats [Abstract]. Toxicol Appl Pharmacol 37: 146-147.
Moore, JA: McConnell, EE; Dalgard, DW: Harris, MW. (1979). Comparative toxicity of three
       halogenated dibenzofurans  in guinea pigs, mice, and rhesus monkeys. Ann N Y Acad Sci
       320:151-163.
Moore. RW: Potter. CL: Theobald. HM: Robinson. JA: Peterson. RE. (1985). Androgenic
       deficiency in male rats treated with 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Appl
       Pharmacol 79: 99-111.
Moore, RW: Peterson, RE. (1985). Enhanced catabolism and  elimination of androgens do not
       cause the androgenic deficiency in 2,3,7,8-tetrachlorodibenzo-p-dioxin-treated rats
       [Abstract]. Fed Proc 44: 518.
Moore, RW: Parsons, JA: Bookstaff, RC: Peterson, RE. (1989). Plasma concentrations of
       pituitary hormones in 2,3,7,8-tetrachlorodibenzo-p-dioxin-treated male rats.  J Biochem
       Toxicol 4:  165-172.
Moore, RW: Jefcoate, CR: Peterson, RE. (1991). 2,3,7,8-Tetrachlorodibenzo-p-dioxin inhibits
       steroidogenesis in the rat testis by inhibiting the mobilization of cholesterol to
       cytochrome P450scc. Toxicol Appl Pharmacol 109: 85-97.
Moore. RW: Mablv. TA: Bjerke. PL: Peterson. RE. (1992). In utero and lactational 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) exposure decreases androgenic responsiveness of
       male sex organs and permanently inhibits spermatogenesis and demasculinizes sexual
       behavior in rats [Abstract].  Toxicologist 12: 81.

                                        D-195

-------
Moos, AB; Baecher-Steppan, L; Kerkvliet NI. (1994). Acute inflammatory response to sheep
       red blood cells in mice treated with 2,3,7,8-tetrachlorodibenzo-p-dioxin: the role of
       proinflammatory cytokines, IL-1 and TNF. Toxicol Appl Pharmacol  127: 331-335.
Moran. FM: Tarara. R: Chen. J: Santos. S: Cheney. A: Overstreet JW: Laslev. BL. (2001).
       Effect of dioxin  on ovarian function in the cynomolgus macaque (M. fascicularis).
       Reprod Toxicol  15: 377-383.
Moriguchi, T; Motohashi, H; Hosoya, T; Nakajima, O: Takahashi, S: Ohsako, S: Aoki, Y;
       Nishimura, N; Tohyama, C: Fujii-Kuriyama, Y; M, Y. (2003). Distinct response to dioxin
       in an arylhydrocarbon receptor (AHR)-humanized mouse. PNAS 9: 5652-5657.
       http://dx.doi.org/10.1073/pnas.1037886100.
Morris, PL; Snyder, NK; Gokani, V: Blair, RE; Holsapple, MP. (1992). Enhanced suppression
       of humoral immunity in DBA/2 mice following subchronic exposure to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD). Toxicol Appl Pharmacol 112:  128-132.
Morrissey, RE; Harris, MW: Diliberto, JJ: Birnbaum, LS. (1992). Limited PCB antagonism of
       TCDD-induced malformations in mice. Toxicol Lett 60: 19-25.
Morse, DC: Groen, D; Veerman, M: van Amerongen, CJ: Koeter, HB; Smits van Prooije, AE;
       Visser, TJ: Koeman, JH: Brouwer, A. (1993). Interference of polychlorinated biphenyls
       in hepatic and brain thyroid hormone metabolism in fetal and neonatal rats. Toxicol Appl
       Pharmacol 122: 27-33.
Morse, DC: Wehler, EK; Wesseling, W; Koeman, JH: Brouwer, A. (1996). Alterations in rat
       brain thyroid hormone status following pre- and postnatal exposure to polychlorinated
       biphenyls (Aroclor 1254). Toxicol Appl Pharmacol 136: 269-279.
       http://dx.doi.org/10.1006/taap.1996.0034.
Moshammer, H; Neuberger, M. (2000). Sex ratio in the children of the Austrian chloracne
       cohort. Lancet 356: 1271-1272.
Mukai, M: Lin. TM: Peterson. RE: Cooke, PS: Tischkau, SA. (2008). Behavioral rhythmicity of
       mice lacking AhR and attenuation  of light-induced phase shift by 2,3,7,8-
       tetrachlorodibenzo-p-dioxin. J Biol Rhythms 23: 200-210.
       http://dx.doi.org/10.1177/0748730408316022.
Murante, FG:  Gasiewicz, TA. (2000). Hemopoietic progenitor cells are sensitive targets of
       2,3,7,8-tetrachlorodibenzo-p-dioxin in C57BL/6J mice. Toxicol Sci 54: 374-383.
Murray. FJ: Smith. FA:  Nitschke. KD: Humiston. CG: Kociba. RJ: Schwetz. BA. (1979). Three-
       generation reproduction study of rats given 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
       in the diet. Toxicol Appl Pharmacol 50: 241-252. http://dx.doi.org/10.1016/0041-
       008X(79)90149-2.
Mustafa. A: Holladav. SD: Goff. M: Witonsky.  SG: Kern R: Reillv. CM: Sponenberg. DP:
       Gogal, RM, Jr. (2008). An enhanced postnatal autoimmune profile in 24 week-old
       C57BL/6 mice developmentally exposed to TCDD. Toxicol Appl Pharmacol 232:  51-59.
       http://dx.doi.0rg/10.1016/j.taap.2008.04.015.
Myllymakl SA: Haavisto. TE: Brokken. LJ: Viluksela. M: Topparl J: Paranko. J. (2005). In
       utero and lactational exposure to TCDD; steroidogenic outcomes differ in male and
       female rat pups.  Toxicol Sci 88: 534-544. http://dx.doi.org/10.1093/toxsci/kfi308.
Nagarkatti, PS: Sweeney, GD: Gauldie, J: Clark, DA. (1984). Sensitivity to suppression of
       cytotoxic T cell generation by 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) is dependent
       on  the Ah genotype of the murine host. Toxicol Appl Pharmacol 72:  169-176.
                                        D-196

-------
Nayyar, T; Bruner-Tran, KL; Piestrzeniewicz-Ulanska, D: Osteen, KG. (2007). Developmental
       exposure of mice to TCDD elicits a similar uterine phenotype in adult animals as
       observed in women with endometriosis. Reprod Toxicol 23: 326-336.
       http://dx.doi.0rg/10.1016/i.reprotox.2006.09.007.
Neff-LaFord. HP: Vorderstrasse. BA: Lawrence. BP. (2003). Fewer CTL, not enhanced NK
       cells, are sufficient for viral clearance from the lungs of immunocompromised mice. Cell
       Immunol 226: 54-64. http://dx.doi.Org/10.1016/j.cellimm.2003.ll.005.
Negishi, T; Shimomura, H; Koyama, T; Kawasaki, K; Ishii, Y; Kvuwa, S; Yasuda, M; Kuroda,
       Y; Yoshikawa, Y. (2006). Gestational and lactational exposure to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin affects social behaviors between developing rhesus monkeys
       (Macaca mulatta). Toxicol Lett 160: 133-244.
       http://dx.doi.0rg/10.1016/i.toxlet.2005.07.008.
Ness, DK; Schantz, SL; Moshtaghian, J; Hansen, LG. (1993). Effects of perinatal exposure to
       specific PCB congeners on thyroid hormone concentrations and thyroid histology in the
       rat. Toxicol Lett 68:  311-323. http://dx.doi.org/10.1016/0378-4274(93)90023-Q.
Neubeit R: Jacob-Mullen U: Stahlmann. R: Helge. H: Neubeit D. (1990). Polyhalogenated
       dibenzo-p-dioxins and dibenzofurans and the immune system. 1. Effects on peripheral
       lymphocyte subpopulations of a non-human primate (Callithrix jacchus) after treatment
       with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Arch Toxicol 64: 345-359.
Nienstedt W: Parkkl M: Uotila. P: Aitio. A. (1979). Effect of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin on the hepatic metabolism  of testosterone in the rat. Toxicology 13: 233-236.
Niittynen, M: Tuomisto, JT: Auriola, S: Pohjanvirta, R: Svrjala, P: Simanainen, U: Viluksela, M:
       Tuomisto, J. (2003). 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)-induced accumulation
       of biliverdin and hepatic peliosis in rats. Toxicol Sci 71: 112-123.
Niittynen, M: Simanainen, U: Syrjala, P:  Pohjanvirta, R: Viluksela, M:  Tuomisto, J: Tuomisto,
       JT. (2007). Differences in acute toxicity syndromes of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin and 1,2,3,4,7,8-hexachlorodibenzo-p-dioxin in rats. Toxicology 235: 39-51.
       http://dx.doi.0rg/10.1016/i.tox.2007.03.012.
Niittvnen, M: Tuomisto. JT: Pohjanvirta. R. (2008). Effect of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin (TCDD) on heme oxygenase-1, biliverdin IXalpha reductase and delta-
       aminolevulinic acid synthetase 1 in rats with wild-type or variant AH receptor.
       Toxicology 250:  132-142. http://dx.doi.Org/10.1016/i.tox.2008.06.014.
Nikolaidis, E; Brunstrom, B; Dencker, L; Veromaa, T. (1990). TCDD inhibits the support of B-
       cell development by the bursa of Fabricius. Pharmacol Toxicol 67: 22-26.
Nilsson, CB; Hoegberg, P: Trossvik, C; Azais-Braesco, V: Blaner, WS: Fex, G: Harrison, EH;
       Nau. H: Schmidt CK: van Bennekum. AM: Hakansson. H. (2000). 2,3,7,8-
       tetrachlorodibenzo-p-dioxin increases serum and kidney retinoic acid levels and kidney
       retinol esterification in the rat. Toxicol Appl Pharmacol  169: 121-131.
       http://dx.doi.org/10.1006/taap.2000.9059.
Nishijo, M: Kuriwaki, J: Hori, E; Tawara, K; Nakagawa, H: Nishijo, H. (2007). Effects of
       maternal exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin on fetal brain growth and motor
       and behavioral development in offspring rats.  Toxicol Lett 173:  41-47.
       http://dx.doi.0rg/10.1016/j.toxlet.2007.06.007.
Nishimura, N; Miyabara, Y: Suzuki, JS: Sato, M: Aoki, Y: Satoh, M: Yonemoto, J: Tohyama, C.
       (2001). Induction of metallothionein in the livers of female Sprague-Dawley rats treated
       with 2,3,7 ,8-tetrachlorodibenzo-p-dioxin. Life Sci 69: 1291-1303.

                                         D-197

-------
Nishimura, N; Miyabara, Y; Sato, M; Yonemoto, J: Tohyama, C. (2002). Immunohistochemical
       localization of thyroid stimulating hormone induced by a low oral dose of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin in female Sprague-Dawley rats. Toxicology 171: 73-82.
       http://dx.doi.org/10.1016/S0300-483X(OnOQ559-5.
Nishimura, N: Yonemoto, J: Miyabara, Y; Sato, M; Tohyama, C. (2003). Rat thyroid hyperplasia
       induced by gestational and lactational exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin.
       Endocrinology 144: 2075-2083.
Nishimura, N: Yonemoto, J: Miyabara, Y: Fujii-Kuriyama, Y: Tohyama, C. (2005a). Altered
       thyroxin and retinoid metabolic response to 2,3,7,8-tetrachlorodibenzo-p-dioxin in aryl
       hydrocarbon receptor-null mice. Arch Toxicol 79: 260-267.
       http://dx.doi.org/10.1007/s00204-004-0626-4.
Nishimura, N: Yonemoto, J: Nishimura, H; Ikushiro, S; Tohyama, C. (2005b). Disruption of
       thyroid hormone homeostasis at weaning of Holtzman rats by lactational but not in utero
       exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Sci 85: 607-614.
       http://dx.doi.org/10.1093/toxsci/kfil22.
Nishimura, N: Yonemoto, J: Nishimura, H; Tohyama, C. (2006). Localization of cytochrome
       P450 1A1 in a specific region of hydronephrotic kidney of rat neonates lactationally
       exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicology 227: 117-126.
       http://dx.doi.0rg/10.1016/j.tox.2006.07.020.
Nishimura, N: Matsumura, F; Vogel, CF; Nishimura, H: Yonemoto, J: Yoshioka, W; Tohyama,
       C.. (2008). Critical role of cyclooxygenase-2 activation in pathogenesis of hydronephrosis
       caused by lactational exposure of mice to dioxin. Toxicol Appl Pharmacol 231: 374-383.
       http://dx.doi.0rg/10.1016/j.taap.2008.05.012.
Nishiumi, S: Yabushita, Y: Furuyashiki, T; Fukuda, I; Ashida, H. (2008). Involvement of
       SREBPs in 2,3,7,8-tetrachlorodibenzo-p-dioxin-induced disruption of lipid metabolism in
       male guinea pig. Toxicol Appl Pharmacol 229: 281-289.
       http://dx.doi.0rg/10.1016/i.taap.2008.01.030.
Nohara, K; Ushio, H: Tsukumo, S: Kobayashi, T; Kijima, M: Tohyama, C: Fujimaki, H. (2000a).
       Alterations of thymocyte development, thymic emigrants and peripheral T cell population
       in rats exposed to 2,3,7,  8-tetrachlorodibenzo-p-dioxin. Toxicology 145: 227-235.
Nohara, K; Fujimaki, H: Tsukumo, S: Ushio, H: Miyabara, Y: Kijima, M: Tohyama, C:
       Yonemoto, J. (2000b). The effects of perinatal exposure to low doses of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin on immune organs in rats. Toxicology  154: 123-133.
       http://dx.doi.org/10.1016/S0300-483X(00)00323-1.
Nohara. K: Izumi, H: Tamura, S: Nagata, R: Tohyama, C. (2002a). Effect of low-dose 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) on influenza A virus-induced mortality in mice.
       Toxicology 170: 131-138.
Nohara, K: Fujimaki, H: Tsukumo, S: Inouye, K: Sone, H: Tohyama, C. (2002b). Effects of
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on T cell-derived cytokine production in
       ovalbumin (OVA)-immunized C57B1/6 mice. Toxicology 172: 49-58.
       http://dx.doi.org/10.1016/S0300-483X(01)00582-0.
Nohara, K: Ao, K: Miyamoto, Y: Suzuki, T: Imaizumi, S: Tateishi,  Y: Omura, S: Tohyama, C:
       Kobayashi,  T. (2008). Arsenite-Induced Thymus Atrophy is Mediated by Cell Cycle
       Arrest: A Characteristic  Downregulation of E2F-Related Genes Revealed by a
       Microarray Approach. Toxicol Sci 101: 226-238.
       http://dx.doi.org/10.1093/toxsci/kfm268.

                                        D-198

-------
Nottebrock. C: Riecke. K: Kruse. M: Shakibael M: Stahlmann. R. (2006). Effects of 2,3,7,8-
       tetrachloro-dibenzo-p-dioxin on the extracellular matrix of the thymus in juvenile
       marmosets (Callithrix jacchus). Toxicology 226: 197-207.
       http://dx.doi.0rg/10.1016/i.tox.2006.07.010.
Novelli. M: Piaggl S: De Tata. V. (2005). 2,3,7,8-Tetrachlorodibenzo-p-dioxin-induced
       impairment of glucose-stimulated insulin secretion in isolated rat pancreatic islets.
       Toxicol Lett 156: 307-314. http://dx.doi.Org/10.1016/j.toxlet.2004.12.004.
NTP (National Toxicology Program). (1982). NTP Technical Report on carcinogenesis bioassay
       of 2,3,7,8-tetrachlorodibenzo-p-dioxin in Osborne-Mendel rats and B6C3F1 mice
       (gavage study). Research Triangle Park, NC.
       http://ntp.niehs.nih.gov/?obiectid=07060172-DEB2-6542-D7CD537BAB5B2ACD.
NTP (National Toxicology Program). (2006). NTP technical report on the toxicology and
       carcinogenesis studies of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (CAS no. 1746-
       01-6) in female harlan Sprague-Dawley rats (gavage studies). (NTP TR 521; NIH
       Publication No. 06-4468). Research Triangle Park, NC.
Ohbayashi, H; Saito, M:  Senoh, H; Umeda, Y; Aiso, S: Yamazaki, K; Nagano, K; Yamamoto, S:
       Fukushima, S. (2008). Occurrence of two different types of glutathione s-transferase
       placental form-positive hepatocytes after a single administration of 2,3,7,8-
       tetrabromodibenzo-p-dioxin in rats. Ind Health 46: 281-288.
Ohsako, S: Miyabara, Y; Nishimura, N; Kurosawa, S: Sakaue, M: Ishimura, R: Sato, M: Takeda,
       K: Aoki, Y: Sone, H: Tohyama, C: Yonemoto, J. (2001). Maternal exposure to a low
       dose of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) suppressed the development of
       reproductive organs of male rats: Dose-dependent increase of mRNA levels of 5a-
       reductase type 2 in contrast to descrease of androgen receptor in the pubertal ventral
       prostate. Toxicol  Sci 60: 132-143.
Ohsako, S: Miyabara, Y: Sakaue, M: Ishimura, R: Kakeyama, M: Izumi, H: Yonemoto, J:
       Tohyama, C. (2002). Developmental stage-specific effects of perinatal 2,3,7,8-
       tetrachlorodibenzo-p-dioxin exposure on reproductive organs of male rat offspring.
       Toxicol Sci 66: 283-292.
Ohyama, K. (2006). [Disorders of sex differentiation caused by exogenous sex hormones and
       endocrine disrupters]. Nippon Rinsho Suppl 2: 533-538.
Ohvama, K: Ohta, M: Sano, T: Sato. K: Nakagomi, Y: Shimura. Y: Yamano. Y. (2007).
       Maternal exposure of low dose of TCDD modulates the expression of estrogen receptor
       subunits of male gonads in offspring. J Vet Med Sci 69: 619-625.
Okey, AB; Vella, LM; Harper, PA. (1989). Detection and characterization of a low affinity form
       of cytosolic Ah receptor in livers of mice nonresponsive to induction of cytochrome Pl-
       450 by 3-methylcholanthrene. Mol Pharmacol 35: 823-830.
Olson, J: McGarrigle, B; Tonucci, D: Schecter, A; Eichelberger, H. (1990). Developmental
       toxicity of 2,3,7,8-TCDD in the rat and hamster. Chemosphere 20: 1117-1123.
       http://dx.doi.org/10.1016/0045-6535(90)90230-Q.
Olson, J: McGarrigle, B.  (1992). Comparative developmental toxicity of 2,3,7,8-tetrachloro-
       dibenzo-p-dioxin (TCDD). Chemosphere 25: 71-74. http://dx.doi.org/10.1016/0045-
       6535(92)90482-7.
Olson, JR; Holscher, MA; Neal, RA. (1980). Toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin in
       the Golden Syrian hamster. Toxicol Appl Pharmacol 55: 67-78.
       http://dx.doi.org/10.1016/0041-008X(80)90221-5.

                                        D-199

-------
Olson, JR; McGarrigle, BP. (1990). Characterization of the developmental toxicity of 2,3,7,8-
      TCDD in the Golden Syrian hamster [Abstract]. Toxicologist 10: 313.
Operana. T: Nguyen. N: Chen. S: Beaton. D: Tukev. R. (2007). Human CYP1A1GFP Expression
      in Transgenic Mice Serves as a Biomarker for Environmental Toxicant Exposure.
      Toxicol Sci 95: 98-107. http://dx.doi.org/10.1093/toxsci/kfll44.
Paaiarvl G: Viluksela. M: Pohianvirta. R: Stenius. U: J. H. (2005). TCDD activates Mdm2 and
      attenuates the p53 response to DNA damaging agents. Carcinogenesis 26: 201-208.
      http://dx.doi.org/10.1093/carcin/bgh289.
Pan. X: Inouve. K: Ito. T: Nagai. H: Takeuchi. Y: Miyabara. Y: Tohyama. C: Nohara. K. (2004).
      Evaluation of relative potencies of PCB126 and PCB169 for the immunotoxicities in
      ovalbumin (OVA)-immunized mice. Toxicology 204: 51-60.
      http://dx.doi.0rg/10.1016/i.tox.2004.06.024.
Pande, K: Moran,  SM; Bradfield, CA. (2005). Aspects of Dioxin Toxicity Are Mediated by
      Interleukin 1-Like Cytokines.  Mol Pharmacol 67: 1393-1395.
      http://dx.doi.org/10.1124/mol.105.010983.
Park. JS: Hwang. SY: Lee. WS: Yu. KW: Paek. KY: Hwang. BY: Han. K. (2006). The
      therapeutic effect of tissue cultured root of wild Panax ginseng C.A. Mayer on
      spermatogenetic disorder. Arch Pharm Res 29: 800-807.
Parkinson. A: Thomas. PE: Ryan.  DE: Reik. LM: Safe. SH: Robertson. LW: Levin. W. (1983).
      Differential time course of induction of rat liver microsomal cytochrome P-450 isozymes
      and epoxide hydrolase by Aroclor 1254. Arch Biochem Biophys 225: 203-215.
Partanen, AM; Alaluusua, S: Miettinen, PJ: Thesleff, I; Tuomisto, J: Pohjanvirta, R: Lukinmaa,
      PL. (1998). Epidermal growth factor receptor as a mediator of developmental toxicity of
      dioxin in mouse embryonic teeth. Lab Invest 78: 1473-1481.
Patterson, RM; Stachlewitz, R: Germolec, D. (2003). Induction of apoptosis by 2,3,7,8-
      tetrachlorodibenzo-p-dioxin following endotoxin exposure. Toxicol Appl Pharmacol 190:
       120-134. http://dx.doi.org/10.1016/S0041-008X(03)00186-8.
Peraino, C: Staffeldt EF; Ludeman, VA. (1981). Early appearance of histochemically altered
      hepatocyte foci and liver tumors in female rats treated with carcinogens one day after
      birth. Carcinogenesis 2: 463-465.
Perucatti, A; Di Meo, GP: Albarella, S: Ciotola, F; Incarnato, D: Caputi Jambrenghi, A; Peretti,
      V: Vonghia, G: lannuzzi, L. (2006). Increased frequencies of both chromosome
      abnormalities and SCEs in two sheep flocks exposed to high dioxin levels during
      pasturage. Mutagenesis 21: 67-75. http://dx.doi.org/10.1093/mutage/gei076.
Pesonen, SA; Haavisto, TE; Viluksela, M; Toppari, J: Paranko, J. (2006). Effects of in utero and
      lactational  exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on rat follicular
      steroidogenesis. Reprod Toxicol 22: 521-528.
      http://dx.doi.0rg/10.1016/i.reprotox.2006.03.007.
Peters, JM; Wiley, LM. (1995). Evidence that murine preimplantation embryos express aryl
      hydrocarbon receptor. Toxicol Appl Pharmacol 134: 214-221.
      http://dx.doi.org/10.1006/taap.1995.1186.
Peters, JM: Narotsky, MG: Elizondo, G: Fernandez-Salguero, PM; Gonzalez, FJ: Abbott BD.
      (1999). Amelioration of TCDD-induced teratogenesis in aryl hydrocarbon receptor
      (AhR)-null mice. Toxicol Sci 47: 86-92.
                                        D-200

-------
Petroff. BK: Gao. X: Roztnan. KK: Terranova. PF. (2000). Interaction of estradiol and 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) in an ovulation model: Evidence for systemic
       potentiation and local ovarian effects. Reprod Toxicol 14: 247-255.
Petroff BK: Gao. X: Rozman. KK: Terranova. PF. (2001). The effects of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) on weight gain and hepatic ethoxyresorufm-o-
       deethylase (EROD) induction vary with ovarian hormonal status in the immature
       gonadotropin-primed rat model. Reprod Toxicol 15: 269-274.
Petroff BK: Gao. X: Ohshima. K: Shi. F: Son. PS: Robv. KF: Rozman. KK: Watanabe. G:
       Tava. K: Terranova. PF. (2002). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
       on serum inhibin concentrations and inhibin immunostaining during follicular
       development in female Sprague-Dawley rats. Reprod Toxicol 16: 97-105.
Pitt. JA: Buckalew. AR: House. DE: Abbott. BD. (2000). Adrenocorticotropin (ACTH) and
       corticosterone secretion by perifused pituitary and adrenal glands from rodents exposed
       to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Toxicology 151: 25-35.
Pluess, N; Poiger, H; Hohbach, C; Schlatter, C. (1988). Subchronic toxicity of some chlorinated
       dibenzofurans (PCDFs) and a mixture  of PCDFs and chlorinated dibenzodioxins
       (PCDDs) in rats. Chemosphere 17: 973-984. http://dx.doi.org/10.1016/0045-
       6535(88)90068-9.
Pohjanvirta, R: Juvonen, R; Karenlampi, S;  Raunio, H; Tuomisto, J.  (1988). Hepatic Ah-receptor
       levels and the effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on hepatic
       microsomal monooxygenase activities in a TCDD-susceptible and -resistant rat strain.
       Toxicol Appl Pharmacol 92: 131-140.
Pohjanvirta, R: Tuomisto, L; Tuomisto, J. (1989). The central nervous system may be involved
       in TCDD toxicity.  Toxicology 58: 167-174.
Pohjanvirta, R: Hakansson, H: Juvonen, R: Tuomisto, J. (1990). Effects of TCDD on vitamin A
       status and liver microsomal enzyme  activities in a TCDD-susceptible and a TCDD-
       resistant rat strain.  Food Chem Toxicol 28: 197-203.
Pohjanvirta, R: Unkila,  M; Tuomisto, J. (1993). Comparative  acute lethality of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD),  1,2,3,7,8-pentachlorodibenzo-p-dioxin and
       1,2,3,4,7,8-hexachlorodibenzo-p-dioxin in the most  TCDD-susceptible and the most
       TCDD-resistant rat strain. Pharmacol Toxicol 73: 52-56.
Pohianvirta, R: Wong. JM: Li. W: Harper. PA: Tuomisto. J: Okev, AB. (1998). Point mutation in
       intron sequence causes altered carboxyl-terminal structure in the aryl hydrocarbon
       receptor of the most 2,3,7,8-tetrachlorodibenzo-p-dioxin-resistant rat strain. Mol
       Pharmacol 54: 86-93.
Pohjanvirta, R: Niittynen,  M; Linden, J: Boutros, P; Moffat I; Okev, A. (2006). Evaluation of
       various housekeeping genes for their applicability for normalization of mRNA expression
       in dioxin-treated rats. Chem Biol Interact 160:  134-149.
       http://dx.doi.0rg/10.1016/i.cbi.2006.01.001.
Poland. A: Palen, D: Glover. E. (1982). Tumour promotion  by TCDD in skin of HRS/J hairless
       mice. Nature 300:  271-273. http://dx.doi.org/10.1038/300271aO.
Poland, A; Glover, E. (1990). Characterization and strain distribution pattern of the murine Ah
       receptor specified by the Ahd and Ahb-3 alleles. Mol Pharmacol 38: 306-312.
                                         D-201

-------
Pollenz. RS: Santostefano. MJ: Klett E: Richardson. VM: Necela. B: Birnbaum. LS. (1998).
       Female Sprague-Dawley rats exposed to a single oral dose of 2,3,7,8-tetrachlorodibenzo-
       p-dioxin exhibit sustained depletion of aryl hydrocarbon receptor protein in liver, spleen,
       thymus, and lung. Toxicol Sci 42:  117-128.
Porterfield, SP. (2000). Thyroidal dysfunction and environmental chemicals — potential impact
       on brain development. Environ Health Perspect 108: 433-438.
Potter, CL; Sipes, IG: Russell DH. (1983). Hypothyroxinemia and hypothermia in rats in
       response to 2,3,7,8-tetrachlorodibenzo-p-dioxin administration. Toxicol Appl Pharmacol
       69: 89-95.
Potter, CL; Menahan, LA; Peterson, RE. (1986a). Relationship of alterations in energy
       metabolism to hypophagia in rats treated with 2,3,7,8-tetrachlorodibenzo-p-dioxin.
       Fundam Appl Toxicol 6: 89-97. http://dx.doi.Org/10.1093/toxsci/6.l.89.
Potter. CL: Moore. RW: Inhorn. SL: Hagen. TC: Peterson. RE. (1986b). Thyroid status and
       thermogenesis in rats treated with 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Appl
       Pharmacol 84: 45-55. http://dx.doi.org/10.1016/0041-008X(86)90415-l.
Powers. BE: Lin. TM: Vanka. A: Peterson. RE: Juraska. JM: Schantz. SL. (2005).
       Tetrachlorodibenzo-p-dioxin exposure alters radial arm maze performance  and
       hippocampal morphology in female AhR+/? mice. Genes Brain Behav 4: 51-59.
       http://dx.doi.Org/10.llll/i.1601-183X.2004.00098.x.
Prell R: Dearstyne. E: Steppan. LG: Vella. AT: Kerkvliet NI. (2000). CTL hyporesponsiveness
       induced by 2,3,7, 8-tetrachlorodibenzo-p-dioxin: role of cytokines and apoptosis. Toxicol
       Appl Pharmacol 166: 214-221. http://dx.doi.org/10.1006/taap.2000.8971.
Puhvel SM; Sakamoto, M; Ertl DC: Reisner, RM. (1982). Hairless mice as models for
       chloracne: A study of cutaneous changes induced by topical application of established
       chloracnegens. Toxicol Appl Pharmacol 64: 492-503. http://dx.doi.org/10.1016/0041-
       008X(82)90247-2.
Puhvel, SM; Sakamoto, M. (1988). Effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin on murine skin.
       J Invest Dermatol 90: 354-358.
Puhvel, SM: Connor, MJ:  Sakamoto, M. (1991). Vitamin A deficiency and the induction of
       cutaneous toxicity in murine skin by TCDD. Toxicol Appl Pharmacol 107: 106-116.
Ramakrishna, G: Perella, C: Birely, L; Diwan, BA; Fornwald, LW: Anderson, LM. (2002).
       Decrease in K-ras p21 and increase in Rafl and activated Erk 1 and 2 in  murine lung
       tumors initiated by N-nitrosodimethylamine and promoted by 2,3,7,8-tetrachlorodibenzo-
       p-dioxin.  Toxicol Appl Pharmacol 179: 21-34.
       http://dx.doi.org/10.1006/taap.2001.9344S0041008X01993445.
Randerath. K: Putman. KL: Randerath. E: Mason. G: Kellev. M: Safe. S. (1988). Organ-specific
       effects of long term feeding of 2,3,7,8-tetrachlorodibenzo-p-dioxin and 1,2,3,7,8-
       pentachlorodibenzo-p-dioxin on I-compounds in hepatic and renal DNA of female
       Sprague-Dawley rats. Carcinogenesis 9: 2285-2289.
Rao. MS:  Subbarao. V: Prasad. JD: Scarpelll DG. (1988). Carcinogenicity of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin in the Syrian golden hamster. Carcinogenesis 6: 1677-1679.
Render, JA; Hochstein, JR; Aulerich, RJ: Bursian, SJ. (2000). Proliferation of periodontal
       squamous epithelium in mink fed 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Vet Hum
       Toxicol 42: 85-86.
                                        D-202

-------
Render, JA; Bursian, SJ: Rosenstein, DS: Aulerich, RJ. (2001). Squamous epithelial proliferation
       in the jaws of mink fed diets containing 3,3',4,4',5-pentachlorobiphenyl (PCB 126) or
       2,3,7,8-tetrachlorodibenzo-P-dioxin (TCDD). Vet Hum Toxicol 43: 22-26.
Rhile, MJ: Nagarkatti, M; Nagarkatti, PS. (1996). Role of Fas apoptosis and MHC genes in
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)-induced immunotoxicity of T cells.
       Toxicology 110: 153-167.
Rice, DC. (1997). Effect of postnatal exposure to a PCB mixture in monkeys on multiple fixed
       interval-fixed ratio performance. Neurotoxicol Teratol 19: 429-434.
Rice. DC: Havward, S. (1998). Lack of effect of 3,3'4,4',5-pentachlorobiphenyl (PCB 126)
       throughout gestation and lactation on multiple fixed interval-fixed ratio and DRL
       performance in rats. Neurotoxicol Teratol 20: 645-650. http://dx.doi.org/10.1016/S0892-
       0362(98)00024-5.
Rice. DC. (1999). Effect of exposure to 3,3',4,4',5-pentachlorobiphenyl (PCB 126) throughout
       gestation and lactation on development and spatial delayed alternation performance in
       rats. Neurotoxicol Teratol 21: 59-69. http://dx.doi.org/10.1016/50892-0362(98)00031-2.
Rice. DC: Havward, S. (1999). Effects of exposure to 3,3',4,4',5-pentachlorobiphenyl (PCB 126)
       throughout gestation and lactation on behavior (concurrent random interval-random
       interval and progressive ratio performance) in rats. Neurotoxicol Teratol 21: 679-687.
       http://dx.doi.org/10.1016/S0892-0362(99)00021-5.
Riecke, K; Grimm, D; Shakibaei, M; Kossmehl, P; Schulze-Tanzil, G: Paul, M; Stahlmann, R.
       (2002). Low doses of 2,3,7,8-tetrachlorodibenzo-p-dioxin increase transforming growth
       factor B and cause myocardial fibrosis in marmosets (Callithrix jacchus). Arch Toxicol
       76: 360-366. http://dx.doi.org/10.1007/s00204-002-0338-6.
Rier, SE: Martin, DC: Bowman, RE; Dmowski, WP: Becker, JL. (1993). Endometriosis in
       Rhesus Monkeys (Macaca mulatta) Following Chronic Exposure to 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin. Fundam Appl Toxicol 21:  433-441.
Rier, SE: Martin, DC: Bowman, RE; Becker, JL. (1995). Immunoresponsiveness in
       endometriosis: Implications of estrogenic toxicants. Environ Health Perspect 103:  151-
       156.
Rier. SE: Coe, CL: Lemieux, AM: Martin. DC: Morris. R: Lucier, GW: Clark. GC. (200la).
       Increased tumor necrosis factor-alpha production by peripheral blood leukocytes from
       TCDD-exposed rhesus monkeys. Toxicol Sci 60: 327-337.
Rier. SE: Turner. WE: Martin. DC: Morris. R: Lucier, GW: Clark. GC. (200Ib). Serum levels of
       TCDD and dioxin-like chemicals in Rhesus monkeys chronically exposed to dioxin:
       Correlation of increased serum PCB levels with endometriosis. Toxicol Sci 59: 147-159.
Rifkind, AB; Muschick, H. (1983). Benoxaprofen suppression of polychlorinated biphenyl
       toxicity without alteration of mixed function oxidase function. Nature 303: 524-526.
Roby, KF. (2001). Alterations in follicle development, steroidogenesis, and gonadotropin
       receptor binding in a Model of ovulatory blockade. Endocrinology 142: 2328-2335.
Roman, BL: Sommer, RJ: Shinomiya, K: Peterson, RE.  (1995).  In utero and lactational exposure
       of the male rat to 2,3,7,8-tetrachlorodibenzo-p-dioxin: Impaired prostate growth and
       development without inhibited androgen production. Toxicol Appl Pharmacol 134: 241-
       250. http://dx.doi.org/10.1006/taap.1995.1190.
Roman, BL: Peterson, RE. (1998). In utero and lactational exposure of the male rat to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin impairs prostate development. 1. Effects on gene expression.
       Toxicol Appl Pharmacol 150: 240-253. http://dx.doi.org/10.1006/taap.1997.8362.

                                         D-203

-------
Roman, BL; Timms, BG: Prins, GS: Peterson, RE. (1998). In utero and lactational exposure of
       the male rat to 2,3,7,8-tetrachlorodibenzo-p-dioxin impairs prostate development. 2.
       Effects on growth and cytodifferentiation. Toxicol Appl Pharmacol 150: 254-270.
       http://dx.doi.org/10.1006/taap.1998.8395.
Romkes, M; Safe, S. (1988). Comparative activities of 2,3,7,8-tetrachlorodibenzo-p-dioxin and
       progesterone as antiestrogens in the female rat uterus. Toxicol Appl Pharmacol 92: 368-
       380.
Rosenthal GJ: Lebetkin. E: Thigpen. JE: Wilson. R: Tucker. AN: Luster. MI. (1989).
       Characteristics of 2,3,7,8-tetrachlorodibenzo-p-dioxin induced endotoxin
       hypersensitivity: association with hepatotoxicity. Toxicology 56: 239-251.
Rozman, K; Rozman, T; Greim, H. (1984). Effect of thyroidectomy and thyroxine on 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) induced toxicity. Toxicol Appl Pharmacol 72: 372-
       376.
Rozman, KK. (2000). The role of time in toxicology or Haber's c x t product. Toxicology 149:
       35-42. http://dx.doi.org/10.1016/S0300-483X(00)00230-4.
Russell DH: Buckley. AR: Shah. GN: Sipes, IG: Blask, DE: Benson. B. (1988). Hypothalamic
       site of action of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Toxicol Appl Pharmacol
       94: 496-502.
Russo, IH; Russo, J. (1978). Developmental stage of the rat mammary gland as determinant of its
       susceptibility to 7,12-dimethylbenz[a]anthracene. J Natl Cancer Inst 61: 1439-1449.
Ryo, H: Nakajima, H; Nomura, T. (2006). Germ-line mutations at a mouse ESTR (Pc-3) locus
       and human microsatellite loci. J Rubber Res 47: B31-B37.
Salisbury, TB; Marcinkiewicz, JL.  (2002). In utero and lactational exposure to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin and 2,3,4,7,8-pentachlorodibenzofuran reduces growth and
       disrupts reproductive parameters in female rats. Biol Reprod 66: 1621-1626.
Sanders, JE: Eigenberg, DA; Bracht, U: Wang, WR: van Zwieten, MJ. (1988). Thyroid and liver
       trophic changes in rats secondary to liver microsomal enzyme induction caused by an
       experimental leukotriene  antagonist (L-649,923). Toxicol Appl Pharmacol 95: 378-387.
Santostefano, MJ: Wang. X: Richardson. VM: Ross. DG: DeVito, MJ: Birnbaum, LF. (1998). A
       pharmacodynamic analysis  of TCDD-Induced Cytochrome 450 gene expression in
       multiple tissues: Dose and time-dependent effects. Toxicol Appl Pharmacol 151: 294-
       310.
Sauer, RM. (1990). 2,3,7,8-Tetrachlorodibenzo-p-dioxin in sprague-dawley rats. Maryland:
       PATHCO, INC.
Schantz, SL: Barsotti, DA; Allen, JR. (1979). Toxicological effects produced in nonhuman
       primates chronically exposed to fifty parts per trillion 2,3,7,8-tetrachlorodibenzo-Mp-
       dioxin (TCDD) [Abstract]. Toxicol Appl Pharmacol 48: A180.
       http://dx.doi.org/10.1016/0041-008X(79)90509-X.
Schantz, SL: Laughlin, NK; Van Valkenberg, HC: Bowman, RE. (1986).  Maternal care by
       rhesus monkeys  of infant monkeys exposed to either lead or 2,3,7,8-tetrachlorodibenzo-
       P-dioxin. Neurotoxicology 7: 637-650.
Schantz, SL: Bowman, RE. (1989). Learning in monkeys exposed perinatally to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD). Neurotoxicol Teratol 11:  13-19.
       http://dx.doi.org/10.1016/0892-0362(89)90080-9.
                                        D-204

-------
Schantz. SL: Mablv. TA: Peterson. RE. (1991). Effects of perinatal exposure to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) on spatial learning and memory and locomotor
       activity in rats [Abstract]. Teratology 43: 497.
Schantz, SL: Ferguson, SA; Bowman, RE. (1992). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       on behavior of monkeys in peer groups. Neurotoxicol Teratol 14: 433-446.
Schantz, SL: Moshtaghian, J: Ness, DK. (1995). Spatial learning deficits in adult rats exposed to
       ortho-substituted PCB congeners during gestation and lactation. Fundam Appl Toxicol
       26: 117-126. http://dx.doi.org/10.1006/faat.1995.1081.
Schantz. SL: Seo. BW: Moshtaghian. J: Peterson. RE: Moore. RW. (1996). Effects of gestational
       and lactational exposure to TCDD or coplanar PCBs on spatial  learning. Neurotoxicol
       Teratol 18:305-313.
Schantz. SL: Seo. BW: Wong. PW: Pessah. IN. (1997). Long-term effects of developmental
       exposure to 2,2',3,5',6-pentachlorobiphenyl (PCB  95) on locomotor activity, spatial
       learning and memory  and brain ryanodine binding. Neurotoxicology 18: 457-467.
Schrenk, D; Buchmann, A; Dietz, K; Lipp, HP; Brunner, H; Sirma, H;  Miinzel, P; Hagenmaier,
       H; Gebhardt R: Bock, KW. (1994). Promotion of preneoplastic foci in rat liver with
       2,3,7,8-tetrachlorodibenzo-p-dioxin, 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin and a
       defined mixture of 49 polychlorinated dibenzo-p-dioxins. Carcinogenesis 15:  509-515.
       http://dx.doi.0rg/10.1093/carcin/15.3.509.
Schulz. TG: Wiebel FA: Thier. R:  Neubert D: Davies. PS: Edwards. RJ. (2000). Identification
       of theta-class glutathione S-transferase in liver cytosol of the marmoset monkey. Arch
       Toxicol 74: 133-138. http://dx.doi.org/10.1007/s002040050665.
Schuur. AG: Boekhorst FM:  Brouwer. A: Visser. TJ. (1997). Extrathyroidal effects of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin on thyroid hormone turnover in male Sprague-Dawley rats.
       Endocrinology 138: 3727-3734.
Scott MA: Tarara. RP: Hendrickx. AG: Benirschke. K: Overstreet JW: Laslev. BL. (2001).
       Exposure to the dioxin 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) induces squamous
       metaplasia in the endocervix of cynomolgus macaques. J Med Primatol 30: 156-160.
Seefeld. MD: Albrecht RM: Peterson. RE. (1979). Effects of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin on indocyanine green blood clearance in rhesus monkeys.  Toxicology  14: 263-
       272.
Seefeld, MD: Keesey, RE; Peterson, RE. (1984a). Body weight regulation in rats treated with
       2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Appl Pharmacol 76: 526-536.
Seefeld. MD: Corbett SW: Keesev, RE: Peterson. RE. (1984b). Characterization of the wasting
       syndrome in rats treated with 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Appl
       Pharmacol 73: 311-322.
Seefeld, MD: Peterson, RE. (1984). Digestible energy and efficiency of feed utilization in rats
       treated with 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Appl  Pharmacol 74: 214-222.
Seegal. RF: Bush. B: Shain, W. (1990). Lightly chlorinated ortho- sub stituted PCB congeners
       decrease dopamine in nonhuman primate brain and in tissue culture. Toxicol Appl
       Pharmacol 106: 136-144.
Seegal, RF: Brosch, KO: Okoniewski, RJ. (1997). Effects of in utero and lactational exposure of
       the laboratory rat to 2,4,2',4'- and 3,4,3',4'-tetrachlorobiphenyl on dopamine function.
       Toxicol Appl Pharmacol 146: 95-103. http://dx.doi.org/10.1006/taap.1997.8226.
                                         D-205

-------
Senft AP: Dalton. TP: Nebeit DW: Center. MB: Puga. A: Hutchinson. RJ: Kerzee. JK: Uno. S:
       Shertzer, HG. (2002). Mitochondrial reactive oxygen production is dependent on the
       aromatic hydrocarbon receptor.  33: 1268-1278.
Seo. BW: Li. MH: Hansen. LG: Moore. RW: Peterson. RE: Schantz. SL. (1995). Effects of
       gestational and lactational exposure to coplanar polychlorinated biphenyl (PCB)
       congeners or 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on thyroid hormone
       concentrations in weanling rats.  Toxicol Lett 78: 253-262.
Seo. BW: Meserve. LA. (1995). Effects of maternal ingestion of Aroclor 1254 (PCB) on the
       developmental pattern of oxygen consumption and body temperature in neonatal rats.
       Bull Environ Contam Toxicol 55: 22-28.
Seo, BW: Sparks, AJ: Medora, K; Amin, S: Schantz, SL. (1999). Learning and memory in rats
       gestationally and lactationally exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD).
       Neurotoxicol Teratol 21: 231-239.
Seo, BW: Powers, BE; Widholm, JJ: Schantz, SL. (2000). Radial arm maze performance in rats
       following gestational and lactational exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin
       (TCDD). Neurotoxicol  Teratol 22: 511-519.
Sewall, C: Lucier, G: Tritscher, A; Clark, G. (1993). TCDD-mediated changes in hepatic
       epidermal growth factor receptor may be a critical event in the hepatocarcinogenic action
       of TCDD. Carcinogenesis 14: 1885-1893.
Sewall CH: Flagler. N: Vanden Heuvel JP: Clark. GC: Tritscher. AM: Maronpot RM: Lucier.
       GW. (1995a). Alterations in thyroid function in female Sprague-Dawley rats following
       chronic treatment with 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Appl Pharmacol
       132: 237-244.
Sewall CH: Clark. GC: Lucier. GW. (1995b). TCDD reduces rat hepatic epidermal growth
       factor receptor: comparison of binding, immunodetection, and autophosphorylation.
       Toxicol Appl Pharmacol 132: 263-272. http://dx.doi.org/10.1006/taap.1995.1107.
Shepherd. D: Steppan, L: Hedstrom, O: Kerkvliet N. (2001). Anti-CD40 treatment of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD)-exposed C57B1/6 mice induces activation of
       antigen presenting cells yet fails to overcome TCDD-induced suppression of allograft
       immunity. Toxicol Appl Pharmacol 170: 10-22.
       http://dx.doi.org/10.1006/taap.2000.9080.
Shepherd. DM: Dearstyne, EA: Kerkvliet. NI. (2000).  The effects of TCDD on the activation of
       ovalbumin (OVA)-specific DO 11.10 transgenic CD4(+) T cells in adoptively transferred
       mice. Toxicol Sci 56: 340-350.
Shi. Z: Valdez, KE: Ting. AY: Franczak, A: Gum.  SL: Petroff, BK. (2007). Ovarian endocrine
       disruption underlies premature reproductive senescence following environmentally
       relevant chronic exposure to the aryl hydrocarbon receptor agonist 2,3,7,8-
       tetrachlorodibenzo-p-dioxin. Biol Reprod 76: 198-202.
       http://dx.doi.org/10.1095/biolreprod.106.053991.
Shirota, M: Kaneko, T: Okuvama, M: Sakurada, Y: Shirota, K: Matsuki, Y. (2007). Internal
       dose-effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in gonadotropin-primed
       weanling rat model. Arch Toxicol 81: 261-269. http://dx.doi.org/10.1007/s00204-006-
       0146-5.
Shiverick, KT; Muther, TF. (1982). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin on serum
       concentrations and the uterotrophic action of exogenous estrone in rats. Toxicol Appl
       Pharmacol 65: 170-176.

                                        D-206

-------
Shiverick. KT: Muther. TF. (1983). 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) effects on
       hepatic microsomal steroid metabolism and serum estradiol of pregnant rats. Biochem
       Pharmacol 32: 991-995.
Shon. YH: Park. IK: Moon. IS: Chang. HW: Nam. KS. (2002). Effect of chitosan
       oligosaccharide on 2,3,7,8-tetrachlorodibenzo-p-dioxin-induced oxidative stress in mice.
       BiolPharmBull25: 1161-1164. http://dx.doi.org/10.1248/bpb.25.1161.
Silkworth, JB; Antrim, L; Kaminsky, LS. (1984). Correlations between polychlorinated biphenyl
       immunotoxicity, the aromatic hydrocarbon locus, and liver microsomal enzyme induction
       in C57BL/6 and DBA/2 mice. Toxicol Appl Pharmacol 75: 156-165.
       http://dx.doi.org/10.1016/0041-008X(84)90086-3.
Silkworth, JB; Antrim, L. (1985). Relationship between Ah receptor-mediated polychlorinated
       biphenyl (PCB)-induced humoral immunosuppression and thymic atrophy. J Pharmacol
       ExpTher 235: 606-611.
Silkworth, JB: Cutler, PS: Antrim, L: Houston, D; Tumasonis, C: Kaminsky, LS. (1989).
       Teratology of 2,3,7,8-tetrachlorodibenzo-p-dioxin in a complex environmental mixture
       from the Love Canal. Fundam Appl Toxicol 13:1-15. http://dx.doi.org/10.1016/0272-
       0590(89)90302-3.
Silkworth. JB: Maves, BA: Fish. KM: Brown. JF, Jr. (1997). Tumor responses, PCB tissue
       concentrations and PCB hepatic binding in S-D rats fed Aroclors 1016, 1242,  1254, or
       1260. Organohalogen Compounds 34:  164-166.
Sills. RC: Goldsworthv, TL: Sleight SD. (1994). Tumor-promoting effects of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin and phenobarbital in initiated weanling Sprague-Dawley
       rats: a quantitative, phenotypic, and ras p21 protein study. Toxicol Pathol 22: 270-281.
       http://dx.doi.org/10.1177/019262339402200305.
Simanainen, U: Tuomisto, JT:  Tuomisto, J: Viluksela, M. (2002). Structure-Activity
       relationships and  dose responses of Polychlorinated Dibenzo-p-dioxins for short-term
       effects in 2,3,7,8- Tetrachlorodibenzo-p-dioxin-Resistant and sensitive rat strains.
       Toxicol Appl Pharmacol  181: 38-47. http://dx.doi.org/10.1006/taap.2002.9386.
Simanainen, U: Tuomisto, JT:  Tuomisto, J: Viluksela, M. (2003). Dose-response analysis of
       short-term effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin in three differentially
       susceptible rat lines. Toxicol Appl Pharmacol 187: 128-136.
       http://dx.doi.org/10.1016/S0041-008X(02)00068-6.
Simanainen, U: Adamsson, A; Tuomisto, JT: Miettinen, FDVI;  Toppari, J: Tuomisto, J: Viluksela,
       M. (2004a). Adult 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) exposure and effects on
       male reproductive organs in three differentially TCDD-susceptible rat lines. Toxicol Sci
       81:401-407.
Simanainen, U: Haavisto, T; Tuomisto, JT: Paranko, J: Toppari, J: Tuomisto, J: Peterson, RE;
       Viluksela, M. (2004b).  Pattern of male reproductive system effects after in utero and
       lactational 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) exposure in three differentially
       TCDD-sensitive rat lines  Pattern of male reproductive system effects after in utero and
       lactational 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) exposure in three differentially
       TCDD-sensitive rat lines. Toxicol Sci 80: 101-108.
       http://dx.doi.org/10.1093/toxsci/kfhl42.
Slezak, BP; Diliberto, JJ: Birnbaum, LS. (1999). 2,3,7,8-Tetrachlorodibenzo-p-dioxin-mediated
       oxidative stress in CYP1A2 knockout (CYP1A2-/-)  mice. Biochem Biophys Res
       Commun 264: 376-379.

                                        D-207

-------
Slezak. BP: Hatch. GE: DeVito. MJ: Diliberto. JJ: Slade. R: Crissman. K: Hassoun. E:
      Birnbaum, LS. (2000). Oxidative stress in female B6C3F1 mice following acute and
      subchronic exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Toxicol Sci 54:
      390-398.
Slezak. BP: Hamm. JT: Revna. J: Hurst CH: Birnbaum. LS. (2002). TCDD-mediated oxidative
      stress in male rat pups following perinatal exposure. J Biochem Mol Toxicol 16: 49-52.
      http://dx.doi.org/10.1002/jbt.10024.
Sloop, TC: Lucier, GW. (1987). Dose-dependent elevation of Ah receptor binding by TCDD in
      rat liver. Toxicol Appl Pharmacol 88: 329-337.
Smialowicz. RJ: DeVito. MJ: Riddle. MM: Williams. WC: Birnbaum. LS.  (1997). Opposite
      effects of 2,2',4,4',5,5'-hexachlorobiphenyl and 2,3,7,8-tetrachlorodibenzo-p-dioxin on
      the antibody response to sheep erythrocytes in mice. Fundam Appl  Toxicol 37: 141-149.
      http://dx.doi.org/10.1006/faat.1997.2323.
Smialowicz. RJ: Burgin. DE: Williams. WC: Diliberto. JJ: Setzer. RW: Birnbaum. LS. (2004).
      CYP1A2 is not required for 2,3,7,8-tetrachlorodibenzo-p-dioxin-induced
      immunosuppression. Toxicology  197: 15-22. http://dx.doi.Org/10.1016/i.tox.2003.ll.016.
Smialowicz, RJ: DeVito, MJ: Williams, WC: Birnbaum, LS. (2008). Relative potency based on
      hepatic enzyme induction predicts immunosuppressive effects  of a mixture of
      PCDDS/PCDFS andPCBS. Toxicol Appl Pharmacol 227: 477-484.
      http://dx.doi.0rg/10.1016/i.taap.2007.ll.018.
Smith, AG: Francis, JE: Kay, SJ: Greig, JB. (1981). Hepatic toxicity and uroporphyrinogen
      decarboxylase activity following a single dose of 2,3,7,8-tetrachlorodibenzo-p-dioxin to
      mice. Biochem Pharmacol 30: 2825-2830.
Smith. AG: Clothier. B: Robinson. S: Scullion. MJ: Carthew. P: Edwards. R: Luo. J: Lim. CK:
      Toledano, M. (1998). Interaction between iron metabolism and 2,3,7,8-
      tetrachlorodibenzo-p-dioxin in mice with variants of the Ahr gene: a hepatic oxidative
      mechanism. Mol Pharmacol 53: 52-61.
Smith. FA:  Schwetz. BA: Nitschke. KD.  (1976). Teratogenicity of 2,3,7,8-tetrachlorodibenzo-p-
      dioxin in CF-1 mice. Toxicol Appl Pharmacol 38:  517-523.
      http://dx.doi.org/10.1016/0041-008X(76)90183-6.
Sommer, RJ: Hume,  AJ: Ciak, JM; Vannostrand, JJ: Friggens, M; Walker,  MK. (2005). Early
      developmental 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure decreases chick embryo
      heart chronotropic response to isoproterenol but not to agents affecting signals
      downstream of the beta-adrenergic receptor. Toxicol Sci 83: 363-371.
      http://dx.doi.org/10.1093/toxsci/kfi041.
Sparschu. G. . L.: Dunn. F.  . L.: Rowe. V. . K. (1971).  Study of the teratogenicity of 2,3,7,8-
      tetrachiorodibenzo-p-dioxin in the rat. Food Cosmet Toxicol 9: 405-412.
      http://dx.doi.org/10.1016/0015-6264(7n90045-9.
Squire, RA. (1990). Pathologic evaluations of selected tissues from the Dow Chemical TCDD
      and 2,4,5-T rat studies. Washington, DC: Submitted to Carcinogen  Assessment Group,
      U.S. Environmental Protection Agency.
Staples.  JE: Murante. FG: Fiore. NC: Gasiewicz. TA: Silverstone. AE. (1998). Thymic
      alterations induced by 2,3,7,8-tetrachlorodibenzo-p-dioxin are  strictly dependent on aryl
      hydrocarbon receptor activation in hemopoietic cells. J Immunol 160: 3844-3854.
                                        D-208

-------
Stohs, SJ: Hassan, MQ: Murray, WJ. (1983). Lipid peroxidation as a possible cause of TCDD
       toxicity. Biochem Biophys Res Commun 111: 854-859. http://dx.doi.org/10.1016/0006-
       291X(83)91377-3.
Sugihara, K; Kitamura, S: Yamada, T; Ohta, S: Yamashita, K; Yasuda, M; Fujii-Kuriyama, Y.
       (2001). Aryl hydrocarbon receptor (AhR)-mediated induction of xanthine
       oxidase/xanthine dehydrogenase activity by 2,3,7,8-tetrachlorodibenzo-p-dioxin.
       Biochem Biophys Res Commun 281: 1093-1099.
       http://dx.doi.org/10.1006/bbrc.2001.4464.
Sugita-Konishi, Y: Kobavashi, K: Naito. H: Miura, K: Suzuki. Y. (2003). Effect of lactational
       exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin on the susceptibility to Listeria infection.
       Biosci Biotechnol Biochem 67: 89-93.
Sweeney, GD: Jones, KG: Cole, FM; Basford, D: Krestynski, F. (1979). Iron deficiency prevents
       liver toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin. Science 204: 332-335.
Takagi, TN: Matsui, KA; Yamashita, K: Ohmori, H; Yasuda, M. (2000). Pathogenesis of cleft
       palate in mouse embryos exposed to 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD).
       Teratog Carcinog Mutagen 20: 73-86.
Tani, Y: Maronpot RR: Folev, JF: Haseman, JK: Walker. NJ: Nvska, A. (2004). Follicular
       epithelial cell hypertrophy induced by chronic oral administration of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin in female Harlan Sprague-Dawley rats. Toxicol Pathol 32:
       41-49. http://dx.doi.org/10.1080/01926230490260952.
Teske,  S: Bonn. AA: Regal JF: Neumiller, JJ: Lawrence. BP. (2005). Activation of the aryl
       hydrocarbon receptor increases pulmonary neutrophilia and diminishes host resistance to
       influenza A virus. Am J Physiol Lung Cell Mol Physiol 289: 111-124.
       http://dx.doi.org/10.1152/aiplung.00318.2004.
Thackaberry, EA: Nunez. BA: Ivnitski-Steele. ID: Friggins, M: Walker. MK. (2005a). Effect of
       2,3,7,8-tetrachlorodibenzo-p-dioxin on murine heart development: alteration in fetal and
       postnatal cardiac growth, and postnatal cardiac chronotropy. Toxicol Sci 88: 242-249.
       http://dx.doi.org/10.1093/toxsci/kfi302.
Thackaberry, EA: Jiang, Z; Johnson, CD: Ramos, KS: Walker, MK. (2005b). Toxicogenomic
       profile of 2,3,7,8-tetrachlorodibenzo-p-dioxin in the murine fetal heart:  modulation of
       cell cycle and extracellular matrix genes.  Toxicol Sci 88: 231-241.
       http://dx.doi.org/10.1093/toxsci/kfi301.
Theobald, HM: Peterson, RE. (1997). In uteroand lactational exposure to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin: Effects on development of the male and female reproductive
       system of the mouse. Toxicol Appl Pharmacol  145: 124-135.
       http://dx.doi.org/10.1006/taap.1997.8173.
Theobald. HM: Roman. BL: Lin. TM: Ohtanl S: Chen. SW: Peterson. RE. (2000). 2,3,7,8-
       tetrachlorodibenzo-p-dioxin inhibits luminal  cell differentiation and androgen
       responsiveness of the ventral prostate without inhibiting prostatic Salpha-
       dihydrotestosterone formation or testicular androgen production in rat offspring. Toxicol
       Sci 58: 324-338.
Thigpen, JE; Faith, RE; McConnell, EE; Moore,  JA. (1975). Increased  susceptibility to bacterial
       infection as a sequela of exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Infect Immun
       12:1319-1324.
                                         D-209

-------
Thomas, PT; Hinsdill, RD. (1979). The effect of perinatal exposure to tetrachlorodibenzo-p-
       dioxin on the immune response of young mice. Drug Chem Toxicol 2: 77-98.
       http://dx.doi.org/10.3109/01480547908993183.
Thornton. AS: Oda. Y: Stuart. GR: Glickman. BW: de Boer. JG. (2001). Mutagenicity of TCDD
       in Big Blue transgenic rats. Mutat Res 478: 45-50. http://dx.doi.org/10.1016/S0027-
       5107(01)00105-1.
Thornton. AS: Oda. Y: Stuart. GR: Holcroft J:  de Boer. JG. (2004). The dioxin TCDD protects
       against aflatoxin-induced mutation in female rats, but not in male rats. Mutat Res 561:
       147-152. http://dx.doi.0rg/10.1016/j.mrgentox.2004.05.001.
Thunberg, T; Ahlborg, UG: Johnsson, H. (1979). Vitamin A (retinol) status in the rat after a
       single oral dose of 2,3,7,8-tetrachlorodibenzo-p-dioxin. Arch Toxicol 42: 265-274.
Thunberg, T; Hakansson, H. (1983). Vitamin A (retinol) status in the Gunn rat. The effect of
       2,3,7,8-tetrachlorodibenzo-p-dioxin. Arch Toxicol 53: 225-233.
Thunberg. T. (1984). Effects of TCDD on Vitamin A and its relation to TCDD toxicity. In A
       Poland; RD Kimbrough (Eds.), Biological mechanisms of dioxin action (pp. 333-344).
       Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.
Tilson, HA; Davis, GJ: McLachlan, JA: Lucier, GW. (1979). The effects of polychlorinated
       biphenyls given prenatally on the neurobehavioral development of mice. Environ Res 18:
       466-474. http://dx.doi.org/10.1016/0013-9351(79)90122-1.
Timms, BG: Peterson, RE; vom Saal, FS. (2002). 2,3,7,8-tetrachlorodibenzo-p-dioxin interacts
       with endogenous estradiol to disrupt prostate gland morphogenesis in male rat fetuses.
       Toxicol Sci 67: 264-274.
Tomar, RS: Kerkvliet NI. (1991). Reduced T-helper cell function in mice exposed to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD). Toxicol Lett 57: 55-64.
Toth, K; Somfai-Relle, S: Sugar, J: Bence, J. (1979). Carcinogenicity testing of herbicide 2,4,5-
       trichlorophenoxyethanol containing dioxin and of pure dioxin in Swiss mice. Nature 278:
       548-549.
Tritscher. A: Mahler.  J: Portier. CJ: Lucier. GW: Walker. NJ.  (1999). TCDD-induced lesions in
       rat lung after chronic oral exposure. In Endocrine disruption, toxicokinetics, toxicology,
       mechanism of action. Milano, Italy: EMMEZETA Congresi.
Tritscher. AM: Goldstein. JA: Portier. CJ: McCoy. Z: Clark. GC: Lucier. GW. (1992). Dose-
       response relationships for chronic exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin in a rat
       tumor promotion model: Quantification and immunolocalization of CYP1A1 and
       CYP1A2 in the liver. Cancer Res 52: 3436-3442.
Tritscher. AM: Clark. GC: Sewall. C: Sills. RC: Maronpot. R: Lucier. GW. (1995). Persistence
       of TCDD-induced hepatic cell proliferation and growth of enzyme altered foci after
       chronic exposure followed by cessation of treatment in DEN initiated female rats.
       Carcinogenesis 16: 2807-2811.
Tritscher. AM: Seacat AM: Yager. JD: Groopman. JD: Miller. BD: Bell D: Sutter. TR: Lucier.
       GW. (1996). Increased oxidative DNA damage in livers of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin treated intact but not ovariectomized rats. Cancer Lett 98: 219-225.
Tritscher. AM: Mahler. J: Portier. CJ: Lucier. GW: Walker. NJ. (2000). Induction of lung lesions
       in female rats  following chronic exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol
       Pathol 28: 761-769. http://dx.doi.org/10.1177/019262330002800601.
                                         D-210

-------
Truelove, J: Grant, D: Mes, J: Tryphonas, H; Tryphonas, L; Zawidzka, Z. (1982).
       Polychlorinated biphenyl toxicity in the pregnant cynomolgus monkey: A pilot study.
       Arch Environ Contam Toxicol 11: 583-538. http://dx.doi.org/10.1007/BF01056366.
Tsutsumi, O. (2000). [Effects of endocrine disrupters on preimplantation embryo development].
       Nippon Rinsho 58: 2464-2468.
Tucker, AN: Vore, SJ: Luster, MI. (1986). Suppression of B cell differentiation by 2,3,7,8-
       tetrachlorodibenzo-p-dioxin. Mol Pharmacol 29: 372-377.
Turner, JN: Collins, DN. (1983). Liver morphology in guinea pigs administered either pyrolysis
       products of a polychlorinated biphenyl transformer fluid or 2,3,7,8-tetrachlorodibenzo-p-
       dioxin. Toxicol Appl Pharmacol 67: 417-429. http://dx.doi.org/10.1016/0041-
       008X(83)90326-5.
Unkila, M: Pohianvirta, R: MacDonald, E: Tuomisto, J. (1994a). Characterization of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin-induced brain serotonin metabolism in the rat. Eur J
       Pharmacol 270:  157-166.
Unkila, M; Pohianvirta, R: MacDonald, E; Tuomisto, JT: Tuomisto, J. (1994b). Dose response
       and time course of alterations in tryptophan metabolism by 2,3,7,8-tetrachlorodibenzo-p-
       dioxin (TCDD) in the most TCDD-susceptible and the most TCDD-resistant rat strain:
       relationship with TCDD lethality. Toxicol Appl Pharmacol 128: 280-292.
       http://dx.doi.org/10.1006/taap.1994.1208.
Unkila, M: Ruotsalainen, M: Pohjanvirta, R: Viluksela, M: MacDonald, E: Tuomisto, JT:
       Rozman, K: Tuomisto, J. (1995). Effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
       on tryptophan and glucose homeostasis in the most TCDD-susceptible and the most
       TCDD-resistant species, guinea pigs and  hamsters. Arch Toxicol 69: 677-683.
Unkila, M: Pohjanvirta, R: Tuomisto, J. (1998). Body weight loss and changes in tryptophan
       homeostasis by chlorinated dibenzo-p-dioxin congeners in the most TCDD-susceptible
       and the most TCDD-resistant rat strain. Arch Toxicol 72: 769-776.
Ushinohama, K: Son, D: Roby, KF: Rozman, KK: Terranova, PF. (2001). Impaired ovulation by
       2,3,7,8 tetrachlorodibenzo-p-dioxin (TCDD) in immature rats treated with equine
       chorionic gonadotropin. Reprod Toxicol  15: 275-280.
van Birgelen,  AP: Hebert, CD: Wenk, ML: Grimes. LK: Chapin. RE: Mahler. J: Travlos. GS:
       Bucher, JR. (1999a). Toxicity of 3,3',4,4'-tetrachloroazobenzene in rats and mice. Toxicol
       Appl Pharmacol 156: 147-159. http://dx.doi.org/10.1006/taap.1999.8640.
van Birgelen.  APJ. M: Johnson. JD:  Fuciarelll AF: Toft JD. II: Mahler. J: Bucher. JR. (1999b).
       Dose-  and time-response of TCDD in Tg.AC mice after dermal and oral exposure.
       Organohalogen Compounds 42: 235-239.
Van den Berg, M: Heeremans, C: Veenhoven, E: Olie, K. (1987). Transfer of polychlorinated
       dibenzo-p-dioxins and dibenzofurans to fetal and neonatal rats. Fundam Appl Toxicol 9:
       635-644.
van der Kolk. J: van Birgelen. APJ. M: Poigen H: Schlatter. C. (1992). Interactions of
       2,2',4,4',5,5'-hexachlorobiphenyl and 2,3,7,8-tetrachlorodibenzo-p-dioxin in a subchronic
       feeding study in the rat. Chemosphere 25: 2023-2027. http://dx.doi.org/10.1016/0045-
       6535(92)90040-X.
van Logten, MJ: Gupta, BN: McConnell, EE; Moore, JA. (1980). Role of the endocrine system
       in the action of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on the thymus. Toxicology
       15: 135-144.
                                        D-211

-------
Van Miller, JP: Lalich, JJ: Allen, JR. (1977). Increased incidence of neoplasms in rats exposed to
       low levels of 2,3,7,8-tetrachlorodibenzo-p-dioxin. Chemosphere 6: 625-632.
       http://dx.doi.org/10.1016/0045-6535(77)90073-X.
Van Birgelen, AP: Van der Kolk, J: Fase, KM; Bol, I; Poiger, H: Brouwer, A; Van den Berg, M.
       (1995a). Subchronic dose-response study of 2,3,7,8-tetrachlorodibenzo-p-dioxin in
       female Sprague-Dawley rats. Toxicol Appl Pharmacol 132: 1-13.
       http://dx.doi.org/10.1006/taap.1995.1080.
Van Birgelen, AP: Smit, EA: Kampen, EVI: Groeneveld, CN: Fase, KM; Van der Kolk, J: Poiger,
       H: Van den Berg. M: Koeman, JH: Brouwer. A. (1995b). Subchronic effects of 2,3,7,8-
       TCDD or PCBs on thyroid hormone metabolism: Use in risk assessment. Eur J
       Pharmacol 293: 77-85. http://dx.doi.org/10.1016/0926-6917(95)90021-7.
van Birgelen, AP: Fase, KM; van der Kolk, J: Poiger, H: Brouwer, A; Seinen, W; van den Berg,
       M (1996). Synergistic effect of 2,2',4,4',5,5'-hexachlorobiphenyl and 2,3,7,8-
       tetrachlorodibenzo-p-dioxin on hepatic porphyrin levels in the rat. Environ Health
       Perspect 104:  550-557.
Vanden Heuvel, JP: Clark, GC: Kohn, MC: Tritscher, AM: Greenlee, WF: Lucier, GW: Bell,
       DA. (1994). Dioxin-responsive genes: examination of dose-response relationships using
       quantitative reverse transcriptase-polymerase chain reaction. Cancer Res 54: 62-68.
Vanden Heuvel, JP: Clark, GC: Tritscher, A: Lucier, GW. (1994). Accumulation of
       polychlorinated dibenzo-p-dioxins and dibenzofurans in liver of control laboratory rats.
       Fundam Appl Toxicol 23: 465-469. http://dx.doi.Org/10.1093/toxsci/23.3.465.
Vecchi, A; Sironi, M: Canegrati, MA; Recchia, M: Garattini, S. (1983). Immunosuppressive
       effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin in strains of mice with different
       susceptibility  to induction of aryl hydrocarbon hydroxylase. Toxicol Appl Pharmacol 68:
       434-441.
Vezina, CM: Allgeier, SH: Moore, RW: Lin, TM: Bemis, JC: Hardin, HA: Gasiewicz, TA:
       Peterson, RE.  (2008). Dioxin causes ventral prostate agenesis by disrupting dorsoventral
       patterning in developing mouse prostate. Toxicol Sci 106:  488-496.
       http://dx.doi.org/10.1093/toxsci/kfnl83.
Viluksela, M: Stahl, BU: Rozman, KK. (1995). Tissue-specific effects of 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin (TCDD) on the activity of phosphoenolpyruvate
       carboxykinase (PEPCK) in rats. Toxicol Appl Pharmacol 135: 308-315.
       http://dx.doi.org/10.1006/taap.1995.1237.
Viluksela, M: Stahl, BU: Birnbaum, LS: Rozman, KK. (1997a). Sub chronic/chronic toxicity of
       1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin (HpCDD) in rats. Part II. Biochemical effects.
       Toxicol Appl  Pharmacol 146: 217-226.
Viluksela, M: Stahl, BU: Birnbaum, LS: Schramm, KW: Kettrup,  A: Rozman, KK. (1997b).
       Sub chronic/chronic toxicity of 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin (HpCDD) in
       rats: Part I. Design, general observations, hematology, and liver concentrations. Toxicol
       Appl Pharmacol 146: 207-216.
Viluksela, M: Stahl, BU: Birnbaum, LS: Schramm, KW: Kettrup,  A: Rozman, KK. (1998).
       Sub chronic/Chronic  Toxicity of a Mixture of Four Chlorinated Dibenzo-p-dioxins in
       Rats: I. Design, General Observations, Hematology, and Liver Concentrations. Toxicol
       Appl Pharmacol 151: 57-69.
                                         D-212

-------
Viluksela, M; Unkila, M; Pohjanvirta, R: Tuomisto, JT: Stahl, BU: Rozman, KK; Tuomisto, J.
       (1999). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on liver
       phosphoenolpyruvate carboxykinase (PEPCK) activity, glucose homeostasis and plasma
       amino acid concentrations in the most TCDD-susceptible and the most TCDD-resistant
       rat strains. Arch Toxicol 73: 323-336.
Viluksela, M: Eager, Y; Tuomisto, JT: Scheu, G: Unkila, M: Pohjanvirta, R: Flodstrom, S;
       Kosma, VM; Maki-Paakkanen, J: Vartiainen, T; Klimm, C: Schramm, KW: Warngard, L;
       Tuomisto, J. (2000). Liver tumor-promoting activity of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin (TCDD) in TCDD-sensitive and TCDD-resistant rat strains. Cancer Res 60: 6911-
       6920.
Vogel, CF; Zhao, Y; Wong, P; Young, NF; Matsumura, F. (2003). The use of c-src knockout
       mice for the identification of the main toxic signaling pathway of TCDD to induce
       wasting syndrome. J Biochem Mol Toxicol 17: 303-315.
       http://dx.doi.org/10.1002/ibt.10096.
Vogel CFA: Nishimura, N: Sciullo. E: Wong. P: Li. W: Matsumura. F. (2007). Modulation of
       the chemokines KC and MCP-1 by 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in mice.
       Arch Biochem Biophys 461: 169-175. http://dx.doi.Org/10.1016/j.abb.2007.01.015.
Vorderstrasse, BA: Kerkvliet NI. (2001). 2,3,7,8-Tetrachlorodibenzo-p-dioxin affects the
       number and function of murine splenic dendritic cells and their expression of accessory
       molecules. Toxicol Appl Pharmacol  171: 117-125.
       http://dx.doi.org/10.1006/taap.2000.9119.
Vorderstrasse, BA: Steppan, LB; Silverstone, AE; Kerkvliet NI. (2001).  Aryl hydrocarbon
       receptor-deficient mice generate normal immune responses to model  antigens  and are
       resistant to TCDD-induced immune suppression. Toxicol Appl Pharmacol 171: 157-164.
       http://dx.doi.org/10.1006/taap.2000.9122.
Vorderstrasse, BA: Bonn, AA; Lawrence, BP. (2003). Examining the relationship between
       impaired host resistance and altered immune function in mice treated with TCDD.
       Toxicology 188: 15-28. http://dx.doi.org/10.1016/S0300-483X(02)00749-7.
Vorderstrasse, BA: Cundiff, JA: Lawrence, BP. (2004). Developmental Exposure to the Potent
       Aryl Hydrocarbon Receptor Agonist 2,3,7,8-Tetrachlorodibenzo-p-Dioxin Impairs the
       Cell-Mediated Immune Response to  Infection with Influenza A Virus, but Enhances
       Elements of Innate Immunity. J Immunotoxicol 1: 103-112.
       http://dx.doi.org/10.1080/15476910490509244.
Vorderstrasse, BA: Cundiff JA: Lawrence, BP. (2006). A dose-response study of the  effects of
       prenatal and lactational exposure to TCDD on the immune response to influenza a virus.
       J Toxicol Environ Health A 69:  445-463. http://dx.doi.org/10.1080/15287390500246985.
Vorderstrasse, BA: Lawrence, BP. (2006). Protection against lethal challenge with Streptococcus
       pneumoniae is  conferred by aryl hydrocarbon receptor activation but  is not associated
       with an enhanced inflammatory response. Infect Immun  74: 5679-5686.
       http://dx.doi.org/10.1128/IAI.00837-06.
Vos, JG: Moore, JA: Zinkl, JG. (1973). Effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin  on the
       immune system of laboratory animals. Environ Health Perspect 5: 149-162.
Vos, JG: Moore, JA. (1974). Suppression of cellular immunity in rats and mice by maternal
       treatment with 2,3,7,8-tetrachlorodibenzo-p-dioxin. International Arch Allergy Appl
       Immunol 47: 777-794.
                                        D-213

-------
Vos. JG: Moore. JA: Zinkl. JG. (1974). Toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
       in C57B1/6 mice. Toxicol Appl Pharmacol 29: 229-241.
Vos. JG: Kreeftenberg. JG: Engel. HW: Minderhoud. A: Van Noorle Jansen. LM. (1978).
       Studies on 2,3,7,8-tetrachlorodibenzo-p-dioxin induced immune suppression and
       decreased resistance to infection: endotoxin hypersensitivity, serum zinc concentrations
       and effect of thymosin treatment. Toxicology 9:  75-86.
Wagner, E; Frank, MM; Smialowicz, RJ. (2001). 2,3,7,8-Tetrachlorodibenzo-p-dioxin and
       natural immunity: lack of an effect on the complement system in a guinea pig model.
       Toxicology 159: 107-113. http://dx.doi.org/10.1016/S0300-483X(00)00386-3.
Wahba, ZZ: Lawson, TA; Stohs, SJ. (1988). Induction of hepatic DNA single strand breaks in
       rats by 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Cancer Lett 39: 281-286.
Wahba, ZZ: Lawson, TW: Murray, WJ: Stohs, SJ. (1989). Factors influencing the induction of
       DNA single strand breaks in rats by 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD).
       Toxicology 58: 57-69.
Wahba, ZZ: Murray, WJ: Stohs, SJ. (1990a). Altered hepatic iron distribution and release in rats
       after exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Bull Environ Contam
       Toxicol 45: 436-445.
Wahba, ZZ: Murray, WJ: Stohs, SJ. (1990b). Desferrioxamine-induced alterations in hepatic iron
       distribution, DNA damage and lipid peroxidation in control and 2,3,7,8-
       tetrachlorodibenzo-p-dioxin-treated rats. J Appl Toxicol 10: 119-124.
Walisser. JA: Bunger, MK: Glover. E: Harstad, EB: Bradfield, CA. (2004). Patent ductus
       venosus and dioxin resistance in mice harboring a hypomorphic Arnt allele. J Biol Chem
       279: 16326-16331. http://dx.doi.org/10.1074/ibc.M400784200.
Walker, DB: Williams, WC: Copeland, CB: Smialowicz, RJ. (2004). Persistent suppression of
       contact hypersensitivity, and altered T-cell parameters in F344 rats exposed perinatally to
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Toxicology 197: 57-66.
       http://dx.doi.0rg/10.1016/i.tox.2003.12.012.
Walker. NJ: Gastel JA: Costa. LT: Clark. GC: Lucier, GW: Sutter. TR. (1995). Rat CYP1B1: an
       adrenal cytochrome P450 that exhibits sex-dependent expression in livers and kidneys of
       TCDD-treated animals.  Carcinogenesis 16: 1319-1327.
Walker. NJ: Miller. BD: Kohn, MC: Lucier, GW: Tritscher, AM. (1998a). Differences in kinetics
       of induction and reversibility of TCDD-induced  changes in cell proliferation and
       CYP1A1 expression in female Sprague-Dawley rat liver. Carcinogenesis 19: 1427-1435.
       http://dx.doi.0rg/10.1093/carcin/19.8.1427.
Walker. NJ: Crofts. FG: Li. Y: Lax. SF: Haves. CL: Strickland. PT: Lucier, GW: Sutter. TR.
       (1998b). Induction and localization of cytochrome P450 1B1 (CYP1B1) protein in the
       livers of TCDD-treated rats: detection using polyclonal antibodies raised to histidine-
       tagged fusion proteins produced and purified from bacteria. Carcinogenesis 19: 395-402.
Walker. NJ: Portier, CJ: Lax. SF: Crofts. FG: Li. Y: Lucier,  GW: Sutter. TR. (1999).
       Characterization of the dose-response of CYP1B1, CYP1A1, and CYP1A2 in the liver of
       female Sprague-Dawley rats following chronic exposure to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin. Toxicol Appl Pharmacol 154: 279-286. http://dx.doi.org/10.1006/taap.1998.8595.
Walker, NJ: Tritscher, AM; Sills, RC. (1997). Hepatocarcinogenesis in a Sprague-Dawley rat
       initiation/promotion model following discontinuous exposure to TCDD. In R Kites (Ed.),
       Risk assessment, toxicology, endocrine disrupters, epidemiology (pp. 150-153).
       Bloomington, IN: Symposium Secretariat, DIOXIN '97.

                                         D-214

-------
Warn. F: Manzoor. E: Ahlborg. UG: Hakansson. H. (1991). Effects of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) in the lactating rat on maternal and neonatal vitamin
       A status and hepatic enzyme induction: a dose-response study. Chemosphere 23: 1951-
       1956. http://dx.doi.org/10.1016/0045-6535(91)90043-D.
Warren. TK: Mitchell KA: Lawrence. BP. (2000). Exposure to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin (TCDD) suppresses the humoral and cell-mediated immune responses to influenza
       A virus without affecting cytolytic activity in the lung. Toxicol Sci 56: 114-123.
Weber. H: Birnbaum. LS. (1985). 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) and 2,3,7,8-
       tetrachlorodibenzofuran (TCDF) in pregnant C57BL/6N mice: Distribution to the embryo
       and excretion. Arch Toxicol 57: 159-162. http://dx.doi.org/10.1007/BF00290880.
Weber, H: Harris, MW: Haseman, JK; Birnbaum, LS. (1985). Teratogenic potency of TCDD,
       TCDF and TCDD-TCDF combinations in C57BL/6N mice. Toxicol Lett 26: 159-167.
Weber, LW: Palmer, CD: Rozman, K. (1994). Reduced activity of tryptophan 2,3-dioxygenase in
       the liver of rats treated with chlorinated dibenzo-p-dioxins (CDDs): dose-responses and
       structure-activity relationship. Toxicology 86: 63-69. http://dx.doi.org/10.1016/0300-
       483X(94)90053-1.
Weber. LW: Lebofsky. M: Stahl BU: Smith. S: Rozman. KK. (1995). Correlation between
       toxicity and effects on intermediary metabolism in 2,3,7,8-tetrachlorodibenzo-p-dioxin-
       treated male C57BL/6J and DBA/2J mice. Toxicol Appl Pharmacol 131: 155-162.
       http://dx.doi.org/10.1006/taap.1995.1057.
Weinand-Harer, A; Lilienthal H:  Bucholski, KA: Winneke, G. (1997). Behavioral effects of
       maternal exposure to an ortho-chlorinated or a coplanar PCB congener in rats. Environ
       Toxicol Pharmacol 3: 97-103. http://dx.doi.org/10.1016/S1382-6689(96)00145-7.
Weinstein. DA: Gogal RM. Jr:  Mustafa. A: Prater. MR: Holladav. SD. (2008). Mid-gestation
       exposure of C57BL/6 mice to 2,3,7,8-tetrachlorodibenzo-p-dioxin causes postnatal
       morphologic changes in the spleen and liver. Toxicol Pathol 36: 705-713.
       http://dx.doi.org/10.1177/0192623308320276.
Weissberg, JB; Zinkl, JG. (1973). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin upon
       hemostasis and hematologic function in the rat. Environ Health Perspect 5: 119-123.
Wheatley, DN. (1968). Enhancement and inhibition of the induction by 7,12-
       dimethylbenz(a)anthracene of mammary tumours in female Sprague-Dawley rats. Br J
       Cancer 22: 787-797.
White, KL, Jr: Lysy, HH; McCay, JA; Anderson,  AC. (1986). Modulation of serum complement
       levels following exposure  to polychlorinated dibenzo-p-dioxins. Toxicol Appl Pharmacol
       84: 209-219. http://dx.doi.org/10.1016/0041-008X(86)90128-6.
Widholm. JJ: Seo. BW: Strupp. BJ: Seegal RF: Schantz. SL. (2003). Effects of perinatal
       exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin on spatial and visual reversal learning in
       rats. Neurotoxicol Teratol  25: 459-471. http://dx.doi.org/10.1016/S0892-0362(03)00014-
       X
Wolf C: Lambright C:  Mann. P:  Price. M: Cooper. RL: Ostbv. J: LE Jr.  G. (1999a).
       Administration of potentially antiandrogenic pesticides (procymidone, linuron, iprodione,
       chlozolinate, p,p'_DDE, and ketoconazole) and toxic substances (dibutyl- and
       diethylhexyl phthalate, PCB 169, and ethane dimethane sulphonate) during sexual
       differentiation produces diverse profiles of reproductive malformations in the male rat.
       Toxicol Ind Health 15: 94-118. http://dx.doi.org/10.1177/074823379901500109.
                                         D-215

-------
Wolf. CJ: Ostbv. JS: Gray. LE. (1999b). Gestational exposure to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin (TCDD) severely alters reproductive function of female hamster offspring.
       Toxicol Sci 51: 259-264.
Wu, 0: Ohsako, S: Ishimura, R: Suzuki, JS: C, T. (2004). Exposure of mouse preimplantation
       embryos to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) alters the methylation status of
       imprinted genes H19 and Igf2. Biol Reprod 70: 1790-1797.
       http://dx.doi.org/10.1095/biolreprod.103.025387.
Wyde, ME; Lucier, GW: Walker, NJ. (1999). Influence of ovariectomy and 17 B-estradiol on the
       promotion of altered hepatocellular foci by TCDD. In Endocrine disruption,
       toxicokinetics, toxicology, mechanism of action. Milano, Italy: EMMEZETA Congress!.
Wvde. ME: Seelv. J: Lucier. GW: Walker. NJ. (2000). Toxicity of chronic exposure to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin in diethylnitrosamine-initiated ovariectomized rats implanted
       with subcutaneous 17 beta-estradiol pellets. Toxicol Sci 54: 493-499.
Wvde. ME: Wong. VA: Kim. AH: Lucier. GW: Walker. NJ. (200la). Induction of hepatic 8-
       Oxo-deoxyguanosine adducts by 2,3,7,8-Tetrachlorodibenzo-p-dioxin in Sprague-Dawley
       rats is female-specific and estrogen-dependent. Chem Res Toxicol 14:  849-855.
       http://dx.doi.org/10.1021/tx000266i.
Wvde. ME: Eldridge. SR: Lucier. GW: Walker. NJ. (2001b). Regulation of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin-induced tumor promotion by 17 beta-estradiol in female
       Sprague-Dawley rats. Toxicol Appl Pharmacol 173: 7-17.
       http://dx.doi.org/10.1006/taap.2001.9166.
Wvde. ME: Cambre. T: Lebetkin. M: Eldridge. SR: Walker. NJ. (2002). Promotion of altered
       hepatic foci by 2,3,7,8-Tetrachlorodibenzo-p-dioxin and 1713-estradiol in male Sprague-
       Dawley rats. Toxicol Sci 68: 295-303. http://dx.doi.Org/10.1093/toxsci/68.2.295.
Wvde. ME: Braen. AP: Hejtmancik. M: Johnson. JD: Toft JD: Blake. JC: Cooper. SD: Mahler.
       J: Valiant. M: Buchen JR: Walker. NJ. (2004). Oral and dermal exposure to 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) induces cutaneous papillomas and squamous cell
       carcinomas in female hemizygous Tg.AC transgenic mice. Toxicol Sci 82: 34-45.
       http://dx.doi.org/10.1093/toxsci/kfh233.
Yang, JZ; Foster, WG. (1997). Continuous exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin
       inhibits the growth of surgically induced endometriosis in the  ovariectomized mouse
       treated with high dose estradiol. Toxicol Ind Health 13: 15-25.
Yang, JZ; Agarwal, SK; Foster, WG. (2000). Subchronic exposure to 2,3,7,8-tetrachlorodibenzo-
       p-dioxin modulates the pathophysiology of endometriosis in the cynomolgus monkey.
       Toxicol Sci 56: 374-381.
Yang, KH;  Yoo, BS: Choe, SY. (1983). Effects of halogenated dibenzo-p-dioxins on plasma
       disappearance and biliary  excretion of ouabain in rats. Toxicol Lett 15: 259-264.
Yang. YG:  Lebrec. H: Burleson. GR. (1994). Effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       (TCDD) on pulmonary influenza virus titer and natural killer (NK) activity in rats.
       Fundam Appl Toxicol 23: 125-131.
Yang. YM: Huang. DY: Liu. GF: Zhong. JC: Du. K: Li. YF: Song. XH. (2005). Inhibitory
       effects of vitamin A on TCDD-induced cytochrome P-450 1 Al enzyme activity and
       expression. Toxicol Sci 85: 727-734. http://dx.doi.org/10.1093/toxsci/kfil30.
Yasuda, M: Matsui, KA; TN, T. (1999). Palatal rugae anomalies induced by dioxins in mice.
       Organohalogen Compounds.
                                         D-216

-------
Ye, L; Leung, LK. (2008). Effect of dioxin exposure on aromatase expression in ovariectomized
       rats. Toxicol Appl Pharmacol 229: 102-108. http://dx.doi.Org/10.1016/i.taap.2008.01.003.
Yoon, BI; Inoue, T; Kaneko, T. (2000). Teratological effect of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin (TCDD): induction of cleft palate in the ddY and C57BL/6 mouse. J Vet Sci 1:
       113-119.
Yoon, BI; Hirabayashi, Y; Ogawa, Y; Kanno, J: Inoue, T: Kaneko, T. (200la). Hemopoietic cell
       kinetics after intraperitoneal single injection of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       (TCDD) in mice. Chemosphere 43: 819-822. http://dx.doi.org/10.1016/S0045-
       6535(00)00439-2.
Yoon, BI; Hirabayashi, Y: Kaneko, T: Kodama, Y: Kanno, J: Yodoi, J: Kim, DY; Inoue, T.
       (2001b). Transgene expression of thioredoxin (TRX/ADF) protects against 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD)-induced hematotoxicity. Arch Environ Contam
       Toxicol 41: 232-236. http://dx.doi.org/10.1007/s002440010242.
Yoon. CY: Park. M: Kim. BH: Park. JY: Park. MS: Jeong, YK: Kwon, H: Jung. HK: Kang, H:
       Lee, YS: Lee, BJ. (2006). Gene expression profile by 2,3,7,8-tetrachlorodibenzo-p-dioxin
       in the liver of wild-type (AhR+/+) and aryl hydrocarbon receptor-deficient (AhR-/-)
       mice. J VetMed Sci 68: 663-668.
Zareba, G: Hoio, R: Zareba, KM: Watanabe, C: Markowski, VP: Baggs, RB: Weiss. B. (2002).
       Sexually dimorphic alterations of brain cortical  dominance in rats prenatally exposed to
       TCDD. J Appl Toxicol 22:  129-137. http://dx.doi.org/10.1002/iat.839.
Zhu, BT: Gallo. MA: Burger CW. J. r: Meeker. RJ: Cai. MX: Xu. S: Connev, AH. (2008). Effect
       of 2,3,7,8-tetrachlorodibenzo-p-dioxin administration and high-fat diet on the body
       weight and hepatic estrogen metabolism in female C3H/HeN mice. Toxicol Appl
       Pharmacol 226: 107-118. http://dx.doi.Org/10.1016/i.taap.2007.08.018.
Zingeser, MR. (1979). Anomalous development of the soft palate in rhesus macaques (Macaca
       mulatta) prenatally exposed to 3,4,7,8-tetrachlorodibenzo-p-dioxin [Abstract]. Teratology
       19: 54A-55A. http://dx.doi.org/10.1002/tera.1420190222.
Zinkl, JG: Vos, JG: Moore, JA; Gupta, BN. (1973). Hematologic and clinical chemistry effects
       of 2,3,7,8-tetrachlorodibenzo-p-dioxin in laboratory animals. Environ Health Perspect 5:
       111-118.
                                         D-217

-------
 O PPA                                      EPA/600/R-10/038F
\Xd / \                                       www.epa.gov/iris
                    APPENDIX E
       Rodent Bioassay Kinetic Modeling
                        January 2012
               National Center for Environmental Assessment
                  Office of Research and Development
                  U.S. Environmental Protection Agency
                         Cincinnati, OH

-------
            CONTENTS—APPENDIX E: Rodent Bioassay Kinetic Modeling
APPENDIX E.   RODENT BIO AS SAY KINETIC MODELING	E-l
   E. 1.   LITERATURE SEARCH STRATEGY AND RESULTS—IDENTIFYING
         RECENT PUBLICATIONS FOR UPDATING
         2,3,7,8-TETRACHLORODffiENZO-^-DIOXIN (TCDD) TOXICOKINETIC
         MODEL INPUT PARAMETERS	E-l
         E.I.I.  Data Bases Searched	E-l
         E.I.2.  Literature Search Strategy and Approach	E-2
               E.l.2.1.   Chemical Search Terms—DIALOG Search	E-2
               E.l.2.2.   Primary Search Terms (Species)—DIALOG Search	E-2
               E.I.2.3.   Secondary Search Terms (Toxicology)—DIALOG Search	E-3
         E.I.3.  Citation Screening Procedures and Results	E-3
         E. 1.4.  References Selected for More Detailed Review for Updating the PBPK
               Models	E-6
         E.I. 5.  Brief Descriptions of DIALOG Bibliographic Data Bases Searched	E-8
   E.2.   TOXICOKINETIC MODELING CODE (Emondetal., 2005)	E-ll
         E.2.1.  Human Standard Model	E-ll
               E.2.1.1.   Model Code	E-ll
               E.2.1.2.   Input File	E-19
         E.2.2.  Human Gestational Model	E-19
               E.2.2.1.   Model Code	E-19
               E.2.2.2.   Input File	E-30
         E.2.3.  Rat Standard Model	E-31
               E.2.3.1.   Model Code	E-31
               E.2.3.2.   Input Files	E-38
         E.2.4.  Rat Gestational Model	E-54
               E.2.4.1.   Model Code	E-54
               E.2.4.2.   Input Files	E-64
         E.2.5.  Mouse Standard Model	E-71
               E.2.5.1.   Model Code	E-71
               E.2.5.2.   Input Files	E-78
         E.2.6.  Mouse Gestational Model	E-84
               E.2.6.1.   Model Code	E-84
               E.2.6.2.   Input Files	E-94
   E.3.   TOXICOKINETIC MODELING RESULTS FOR KEY ANIMAL
         BIOASSAY STUDIES	E-96
         E.3.1.  Nongestational Studies	E-97
               E.3.1.1.   Cantonietal. (1981)	E-97
               E.3.1.2.   Chu et al. (2007) and Chu et al. (2001)	E-98
               E.3.1.3.   Crofton et al. (2005)	E-99
               E.3.1.4.   Croutch et al. (2005)	E-102
               E.3.1.5.   DeliaPortaetal. (1987) Female	E-103
               E.3.1.6.   Delia Portaetal. (1987) Male	E-104
               E.3.1.7.   Fattore et al. (2000)	E-105

                                       E-ii

-------
                      CONTENTS (continued)
       E.3.1.8.   Fox etal. (1993)	E-107
       E.3.1.9.   Franc etal. (2001) Sprague-Dawley Rats	E-108
       E.3.1.10.  Franc etal. (2001) Long-Evans Rats	E-109
       E.3.1.11.  Franc etal. (2001) Hans Wi star Rats	E-110
       E.3.1.12.  Hassoun et al. (2000)	E-lll
       E.3.1.13.  Hutt et al. (2008)	E-113
       E.3.1.14.  Ishihara et al. (2007)	E-114
       E.3.1.15.  Kitchin and Woods (1979)	E-115
       E.3.1.16.  Kociba etal. (1976)	E-117
       E.3.1.17.  Kociba etal. (1978) Female	E-119
       E.3.1.18.  Kociba etal. (1978) Male	E-120
       E.3.1.19.  Kuchiiwa et al. (2002)	E-121
       E.3.1.20.  Latchoumycandane and Mathur (2002)	E-122
       E.3.1.21.  Li etal. (1997)	E-123
       E.3.1.22.  Murray et al. (1979) Adult Portion	E-126
       E.3.1.23.  NTP (1982) Female Rats, Chronic	E-127
       E.3.1.24.  NTP (1982) Male Rats, Chronic	E-128
       E.3.1.25.  NTP (1982) Female Mice, Chronic	E-130
       E.3.1.26.  NTP (1982) Male Mice, Chronic	E-131
       E.3.1.27.  NTP (2006) 14 Weeks	E-132
       E.3.1.28.  NTP (2006) 31 Weeks	E-134
       E.3.1.29.  NTP (2006) 53 Weeks	E-136
       E.3.1.30.  NTP (2006) 2 Years	E-137
       E.3.1.31.  Nohara et al. (2002)	E-139
       E.3.1.32.  Sewall et al. (1995) and Maronpot et al. (1993)	E-140
       E.3.1.33.  Shi etal. (2007) Adult Portion	E-142
       E.3.1.34.  Simanainen et al. (2002)  and Simanainen et al. (2003)	E-143
       E.3.1.35.  Smialowicz et al. (2004)	E-144
       E.3.1.36.  Smialowicz et al. (2008)	E-146
       E.3.1.37.  Toth etal. (1979) 1 Year	E-148
       E.3.1.38.  Van Birgelen etal. (1995)	E-149
       E.3.1.39.  VandenHeuvel etal. (1994)	E-151
       E.3.1.40.  Weber etal. (1995) C57 Mice	E-153
       E.3.1.41.  White etal.  (1986)	E-156
E.3.2.  Gestational Studies	E-158
       E.3.2.1.   Bell et al. (2007)	E-158
       E.3.2.2.   Hojo etal. (2002)	E-159
       E.3.2.3.   Ikeda et al. (2005)	E-160
       E.3.2.4.   Kattainen et al. (2001) and Simanainen et al. (2004)	E-161
       E.3.2.5.   Keller etal. (2007)	E-162
       E.3.2.6.   Li et al. (2006) 3 Day	E-163
       E.3.2.7.   Markowski etal. (2001)	E-164
       E.3.2.8.   Mietinnen et al. (2006)	E-166

-------
                           CONTENTS (continued)
            E.3.2.9.  Nohara et al. (2000)	E-167
            E.3.2.10. Ohsakoetal. (2001)	E-168
            E.3.2.11. Schantz et al. (1996) and Amin et al. (2000)	E-169
            E.3.2.12. Seoetal. (1995)	E-170
            E.3.2.13. Smith et al.  (1976)	E-171
            E.3.2.14. Sparschu et al. (1971)	E-173
E.4.   RESPONSE SURFACE TABLES	E-176
      E.4.1. Nongestational Lifetime	E-177
      E.4.2. Nongestational 5-Year Peak Average	E-184
      E.4.3. Gestational	E-192
E.5.   REFERENCES	E-199
                                     E-iv

-------
          APPENDIX E.    RODENT BIOASSAY KINETIC MODELING

E.I.  LITERATURE SEARCH STRATEGY AND RESULTS—IDENTIFYING RECENT
     PUBLICATIONS FOR UPDATING 2,3,7,8-TETRACHLORODIBENZO-p-DIOXIN
     (TCDD) TOXICOKINETIC MODEL INPUT PARAMETERS
       The purpose of this literature search was to identify recent publications that address the
input parameters for the physiologically based pharmacokinetic (PBPK) models Aylward and
colleagues (described in articles published in 2005 and 2009) and Emond and colleagues
(described in articles published in 2004, 2005, and 2006). This literature search was part of the
U.S. Environmental Protection Agency (EPA)'s preparation of a response to the National
Academy of Sciences' review (Health Risks from Dioxin and Related Compounds: Evaluation of
the EPA Reassessment, (NAS, 2006) of EPA Exposure and Human Health Reassessment of
2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds (U.S. EPA, 2003), herein
called the "2003 Reassessment." English-only references from 2003 to May 2009 were searched
using bibliographic data bases relevant to health effects and toxicology of TCDD.  The search
focused on toxicokinetic data that could be used to update the dynamic disposition of
2,3,7,8-TCDD in mice, rats, guinea pigs, monkeys, and humans.
       In the primary search, EPA identified 775 distinct citations based on the literature search
criteria described below. EPA also performed an independent supplemental search to avoid
missing key studies.  EPA identified 28 papers for further analysis that appeared on first review
to report data to update the input parameters of the Aylward and Emond PBPK models;
considerations for selection are described in Section E.I.3.

E.I.I.   Data  Bases Searched
       EPA used the following DIALOG bibliographic data bases in the primary search. Brief
descriptions of the DIALOG data bases searched are provided in Section E. 1.5.
   1. File6:NTIS
   2. File 41: Pollution Abstracts
   3. File 55: Biosis
   4. File 153: IPA Toxicology
   5. File 155: MEDLINE
   6. File 156: ToxFile
                                         E-l

-------
   7.  File 157: Biosis Toxicology
   8.  File 159: CancerLit
   9.  File 336: RTECS

   NTIS = National Technical Information Service; IPA = International Pharmaceutical Abstracts;
   RTECS = Registry of Toxic Effects of Chemical Substances.
The PubMed data base was used for the supplemental search.


E.1.2.  Literature Search Strategy and Approach

       The primary search used a tiered key-word approach, as documented below. The
principal search term was the Chemical Abstract Service Registry Number (CASRN) or specific
chemical name, 2,3,7,8-tetrachlorodibenzo:p-dioxin or 2,3,7,8-TCDD. The next tier of search
terms was species, and finally toxicokinetic keywords, as listed below. The period of the search
was 2003 through May 2009, and articles were limited to English language.
       The supplemental PubMed search was limited to the most recent five years (2004 to
present) and used four combinations of key words:
   •   TCDD + pharmacokinetic + humans,
   •   TCDD + toxicokinetic + humans,
   •   TCDD + pharmacokinetic + animals, and
   •   TCDD + toxicokinetic + animals.
E.l.2.1. Chemical Search Terms—DIALOG Search

   •   CASRN: 1746-01-6
   •   2,3,7,8-tetrachlorodibenzo-p-dioxin
   •   2,3,7,8-TCDD
E.1.2.2. Primary Search Terms (Species)—DIALOG Search

   •   Guinea pig(s)
   •   Human(s)
   •   Monkey(s)
   •   Mouse
   •   Mice
   •   Rodent(s)
                                         E-2

-------
       Rat(s)
E.1.2.3. Secondary Search Terms (Toxicology)—DIALOG Search
    1.  Absor*
    2.  ADME
    3.  Aryl hydrocarbon
       receptor
    4.  AhR
    5.  Bioavail*
    6.  Biliar*
    7.  Biotransform*
    8.  Cytochrome
    9.  CYP*
    10. CYP1A1
    11.CYP1A2
    12. Diet, dietary, diets
    13. Disposit*
    14. Distrib*
    15. Drink*
16. Elimin*
17. Excret*
18. Epidemiolog*
19. Feces
20. Feed*
21. First order kinetics
22. Food*
23. Gastro*
24. Gavage*
25. Half-life
26. Induct*
27. Ingest*
28. In silico
29. Kinetic*
30. Liver
31. Lymph*
32. Mechanism (Iw)
   action
33.Metabo*
34. Oral*
35.P450
36. Partition coefficient
37. PBPK
38. Pharmacodynamic*
39. Pharmacokinetic*
40. Physiologically
   based
41. Pharmacokinetic
42. Protein bind*
43. Toxicokinetic*
44. Uri
   * = truncated.
   Iw = terms are within one word of each other and in the order specified (see search term 32).

   ADME = absorption, distribution, metabolism, elimination; AhR = aryl hydrocarbon receptor;
   CYP = cytochrome P450.
E.1.3.   Citation Screening Procedures and Results

       Initial DIALOG searches resulted in a very large number of citation hits. Therefore,

some title and key word restrictions were applied iteratively to screen out less relevant citations
(e.g., requiring some search terms in title, requiring 2,3,7,8-TCDD rather than just TCDD).

Then, using reference management software, pooled information obtained from the various

DIALOG data bases was screened to remove duplicates. Citations then were numbered

sequentially (as a unique identifier). Information retrieved included the following (when

available): author(s), publication year, title, source document name, volume, and page numbers.

       The DIALOG search and duplicate removal procedure produced 775 unique citations. In

the next step, all 775 citations were screened for potential applicability to updating parameters in

the Aylward and Emond PBPK models. Of these  775 citations, 26 were selected for more
                                          E-3

-------
detailed review to determine their potential applicability, and full publications were retrieved.
Two citations were added from the supplemental search, giving a total of 28 articles identified
for further review.
       Bibliographic information for the 28 articles selected for full review is provided in the
reference list at the end of this section. Table E-l summarizes the model input parameters
potentially addressed by the selected articles.
       During 2003 to May 2009, the authors of the two kinetic models under consideration
published several articles. For the Emond model, which was first published in 2004 (Emond et
al., 2004), two subsequent papers have been published (Emond et al., 2006; 2005).  The Aylward
model, which originated from the 1995 papers by Carrier et al. (1995a, b), was later updated by
the same group (Aylward et al., 2005a: 2005b).  The major change implemented in the last two
papers was the description of a desorption process in the digestive tract.  The transfer rate
described is  slow, but for a low body burden of TCDD, this process remains significant.  This
concept was reported in 2002 by Moser and McLachlan (2002).  The major modifications
expected to update the Emond model are (1) consideration of the desorption process in the
gastrointestinal tract and (2) rearrangement of the elimination constant, which will have a
negligible impact on the simulation.  These changes are motivated by plausible observations
reported in the literature.
       Because of the body burden found in humans and the importance of selecting an
appropriate dose metric in human risk assessment, the physiological model is an important tool
for assessing the kinetics following exposure to TCDD (Kim et al., 2003).  Based on the
literature identified in this search, the major contributions that should be reviewed with respect to
the Aylward and Emond kinetic models are not modes of action or pharmacokinetic mechanisms,
but rather information for verifying or improving the accuracy of some model parameters.
       Pharmacokinetics typically refers to four distinct steps including absorption, distribution,
metabolism, and excretion.  Physiologically based models consider each step.  In the model each
step is parameterized to reflect better predictions of the real observations. Occasionally,
reviewing these models is essential to determine if any key processes or parameters might be
described with better accuracy. This perspective underlies the review of the literature described
here.  The review indicates TCDD disposition has become recognized as relatively significant
since the publication of the Emond and Aylward  models.  The literature that provides
                                          E-4

-------
information related to improving these models, however, is limited. For the benefit of this
exercise, EPA selected the literature that would likely contribute significantly to model response,
or to clarify or confirm different key issues driving the model results.  Regarding the two TCDD
models, the two major issues that should be evaluated with respect to the recent literature
identified are the elimination profile and the induction of CYP1A2.
       Reviewing the elimination variation in different species and testing variable elimination
with a data set appears to be appropriate. The literature reports that various factors might
influence elimination rate.  Recent publications report the influence of diverse predictors such
age, body fat, or smoking habit on the elimination half-life (Milbrath et al.,  2009; Kerger et al.,
2007; 2006).  Determining whether using the  Milbrath et al. information would help account for
intraspecies variability in elimination rate in the Emond and Aylward kinetic models would be
useful. In 2006, Emond et al. (2006) reviewed the influence of body fat mass and CYP1A2
induction on the pharmacokinetics of TCDD.  These two factors appear to contribute
significantly to elimination and their influences seem to be driven by TCDD body burden.
Mullerova and Kopecky (2007) discussed the influence of adipose tissue and the "yo-yo" effects
on various diseases that might be influenced by persistent organic pollutant distribution. One
group explored the importance of variable elimination and compared these predictions to
first-order elimination using the Aylward and Emond models and supported these approaches for
risk assessment (Heinzl et al., 2007). Two groups of authors considered a one-compartment
model to derive the elimination half-life (Aylward et al., 2009; Nadal et al., 2008). Comparing
the half-life they obtained using this approach for a range  of body burden to the variable
elimination half-life would be interesting.
       The second important mechanism driving the distribution and elimination of TCDD is the
induction of CYP1A2, identified as the major ligand protein in liver (Diliberto et al., 1997).  For
that process, authors suggested different aspects that should be investigated, including the
importance of the dose metrics in the target tissue and the inducible level of CYP1A2 (Wilkes et
al., 2008; Staskal et al.,  2005).  Other papers address the intraspecies variability of lethal potency
in mature species versus the developing fetus  (Kransler et al., 2007; Korkalainen  et al., 2004).
Still others point out pronounced differences among species (namely, guinea pigs, hamsters,
mice, and rats) (Bohonowych and Denison, 2007). as observed in studies of long-term effects of
low TCDD dose in liver and in studies comparing hepatic accumulation  and clearance of TCDD
                                          E-5

-------
(Korenaga et al., 2007; Boverhoff et al., 2005).  The interspecies variation of the binding affinity

constant of AhR also has been reported (Connor and Aylward, 2006; Nohara et al., 2006).

       The articles identified in this literature review should be adequate to update the Aylward

and Emond models, which need to be evaluated according to the same structure of compartments

described in the literature by the two model authors.


E.I.4.  References Selected for More Detailed Review for Updating the PBPK Models

Aylward, LL; Brunei, RC; Carrier, G; Hays, SM; Gushing, CA; Needham, LL; Patterson, DG;
Gerthoux, PM; Brambilla, P; Mocarelli, P. (2005a). Concentration-dependent TCDD elimination
kinetics in humans: Toxicokinetic modeling for moderately to highly exposed adults from
Seveso, Italy, and Vienna, Austria, and impact on dose estimates for the NIOSH cohort. J Expo
Anal Environ Epidemiol 15: 51-65. http://dx.doi.org/10.1038/sj.jea.7500370.

Aylward, LL; Brunet, RC; Starr, TB; Carrier, G; Delzell, E; Cheng, H; Beall, C. (2005b).
Exposure reconstruction for the TCDD-exposed NIOSH cohort using a concentration- and age-
dependent model of elimination. Risk Anal 25: 945-956. http://dx.doi.Org/10.llll/j.1539-
6924.2005.00645.x.

Aylward, LL; Bodner, KM; Collins, JJ; Wilken, M; McBride, D; Burns, CJ; Hays, SM;
Humphry, N.  (2009). TCDD exposure estimation for workers at a New Zealand 2,4,5-T
manufacturing facility based on serum sampling data. J Expo Sci Environ Epidemiol  TEA: 1-10.
http://dx.doi.org/10.1038/jes.2009.31.

Bohonowych, JE; Denison, MS. (2007). Persistent binding of ligands to the aryl hydrocarbon
receptor. Toxicol Sci 98: 99-109. http://dx.doi.org/10.1093/toxsci/kfm085.

Boverhoff, DR; Burgoon, LD; Tashiro, C; Chittim, B; Harkema, JR; Jump, DB; Zacharewski,
TR. (2005). Temporal and dose-dependent hepatic gene expression patterns in mice provide new
insights into TCDD-mediated hepatotoxicity. Toxicol Sci 85: 1048-1063.
http://dx.doi.org/10.1093/toxsci/kfil62.

Connor, KT; Aylward, LL.  (2006). Human response to dioxin: Aryl hydrocarbon receptor (AhR)
molecular structure, function, and dose-response data for enzyme induction indicate an impaired
human AhR. J Toxicol Environ Health B Crit Rev 9: 147-171.
http://dx.doi.org/10.1080/15287390500196487.

Heinzl, H; Mittlbock, M; Edler, L. (2007). On the translation of uncertainty from toxicokinetic to
toxicodynamic models—the TCDD example. Chemosphere 67: S365-S374.
http://dx.doi.0rg/10.1016/j.chemosphere.2006.05.130.

Irigaray, P; Mejean, L; Laurent, F. (2005). Behaviour of dioxin in pig adipocytes. Food Chem
Toxicol 43: 457-460. http://dx.doi.Org/10.1016/j.fct.2004.ll.016.
                                         E-6

-------
Kerger, BD; Leung, HW; Scott, P; Paustenbach, DJ; Needham, LL; Patterson, DG, Jr; Gerthoux,
PM; Mocarelli, P. (2006). Age- and concentration-dependent elimination half-life of 2,3,7,8-
tetrachlorodibenzo-p-dioxin in Seveso children. Environ Health Perspect 114: 1596-1602.
http://dx.doi.org/10.1289/ehp.8884.

Kerger, BD; Leung, HW; Scott, PK; Paustenbach, DJ. (2007). Refinements on the age-dependent
half-life model for estimating child body burdens of polychlorodibenzodioxins and
dibenzofurans. Chemosphere 67: S272-S278.
http://dx.doi.0rg/10.1016/j.chemosphere.2006.05.108.

Kim, AH; Kohn, MC; Nyska, A; Walker, NJ. (2003). Area under the curve as a dose metric for
promotional responses following 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure. Toxicol Appl
Pharmacol 191: 12-21. http://dx.doi.org/10.1016/S0041-008X(03)00225-4.

Korenaga, T; Fukusato, T; Ohta, M; Asaoka, K; Murata, N; Arima, A; Kubota, S. (2007). Long-
term effects  of subcutaneously injected 2,3,7,8-tetrachlorodibenzo-p-dioxin on the liver of rhesus
monkeys. Chemosphere 67: S399-S404. http://dx.doi.Org/10.1016/j.chemosphere.2006.05.135.

Korkalainen, M; Tuomisto, J; Pohjanvirta, R. (2004). Primary structure and inducibility by
2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) of aryl hydrocarbon receptor represser in a TCDD-
sensitive and a TCDD-resistant rat strain. Biochem Biophys Res Commun 315: 123-131.
http://dx.doi.0rg/10.1016/j.bbrc.2004.01.028.

Kransler, KM; McGarrigle, BP; Olson, JR. (2007). Comparative developmental toxicity of
2,3,7,8-tetrachlorodibenzo-p-dioxin in the hamster, rat and guinea pig. Toxicology 229: 214-225.
http://dx.doi.0rg/10.1016/j.tox.2006.10.019.

Maruyama, W; Yoshida, K; Tanaka, T; Nakanishi, J. (2002). Determination of tissue-blood
partition coefficients for a physiological model for humans, and estimation of dioxin
concentration in tissues. Chemosphere 46: 975-985.

Maruyama, W; Yoshida, K; Tanaka, T; Nakanishi, J. (2003). Simulation of dioxin accumulation
in human tissues and analysis of reproductive risk. Chemosphere 53: 301-313.
http://dx.doi.org/10.1016/S0045-6535(03)00015-8.

Maruyama, W; Aoki, Y. (2006). Estimated cancer risk of dioxins to humans using a bioassay and
physiologically based pharmacokinetic model.  Toxicol Appl Pharmacol 214: 188-198.
http://dx.doi.0rg/10.1016/j.taap.2005.12.005.

Milbrath, MO; Wenger, Y; Chang, CW; Emond, C; Garabrant, D; Gillespie, BW; Jolliet, O.
(2009). Apparent half-lives of dioxins, furans, and polychlorinated biphenyls as a function of
age, body fat, smoking status, and breast-feeding. Environ Health Perspect 117: 417-425.
http://dx.doi.org/10.1289/ehp.11781.

Moser, GA;  McLachlan, MS. (2002). Modeling digestive tract absorption and desorption of
lipophilic organic contaminants in humans. Environ Sci Technol 36: 3318-3325.
                                          E-7

-------
Miillerova, D; Kopecky, J. (2007). White adipose tissue: Storage and effector site for
environmental pollutants.  Physiol Res 56: 375-381.

Nadal, M; Perello, G; Schuhmacher, M; Cid, J; Domingo, JL. (2008). Concentrations of
PCDD/PCDFs in plasma of subjects living in the vicinity of a hazardous waste incinerator:
Follow-up and modeling validation. Chemosphere 73: 901-906.
http://dx.doi.0rg/10.1016/j.chemosphere.2008.07.021.

Nohara, K; Ao, K; Miyamoto, Y; Ito, T;  Suzuki, T; Toyoshiba, H; Tohyama, C. (2006).
Comparison of the 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)-induced CYP1A1 gene
expression profile in lymphocytes from mice, rats, and humans: Mst potent induction in humans.
Toxicology 225: 204-213. http://dx.doi.Org/10.1016/j.tox.2006.06.005.

Olsman, H; Engwall, M; Kammann, U; Klempt, M; Otte, J; Bavel, B; H, H. (2007). Relative
differences in aryl hydrocarbon receptor-mediated response for 18 polybrominated and mixed
halogenated dibenzo-p-dioxins and -furans in cell lines from four different species. Environ
Toxicol Chem 26: 2448-2454.

Saghir SA: Lebofsky, M; Pinson, DM; Rozmana,  KK. (2005). Validation of Haber's Rule
(dosextime = constant) in  rats and mice for monochloroacetic acid and 2,3,7,8-
tetrachlorodibenzo-p-dioxin under conditions of kinetic steady state. Toxicology 215: 48-56.
http://dx.doi.0rg/10.1016/j.tox.2005.06.009.

Schecter, A; Pavuk, M; Papke, O; Ryan,  JJ. (2003). Dioxin, dibenzofuran, and coplanar PCB
levels in Laotian blood and milk from agent orange-sprayed and nonsprayed areas, 2001. J
Toxicol Environ Health A 66: 2067-2075.

Staskal, DF; Diliberto, JJ;  DeVito, MJ; Birnbaum, LS. (2005). Inhibition of human and rat
CYP1A2 by TCDD and dioxin-like chemicals. Toxicol Sci 84: 225-231.
http://dx.doi.org/10.1093/toxsci/kfi090.

Toyoshiba, H; Walker, NJ; Bailer , A; Portier, CJ. (2004). Evaluation of toxic equivalency
factors for induction of cytochromes P450 CYP1A1 and CYP1A2 enzyme activity by dioxin-like
compounds. Toxicol Appl Pharmacol 194:  156-168.
http://dx.doi.0rg/10.1016/j.taap.2003.09.015.

Wilkes,  JG; Hass, BS; Buzatu, DA; Pence, LM; Archer, JC; Beger, RD; Schnackenberg, LK;
Halbert, MK; Jennings, L; Kodell, RL. (2008). Modeling and assaying dioxin-like biological
effects for both dioxin-like and certain non-dioxin-like compounds. Toxicol Sci 102: 187-195.
http://dx.doi.org/10.1093/toxsci/kfm294.
E.I.5.   Brief Descriptions of DIALOG Bibliographic Data Bases Searched

       TheNTIS database comprises summaries of U.S. government-sponsored research,

development, and engineering, plus analyses prepared by federal agencies, their contractors, or

grantees. It is the means through which unclassified, publicly available, unlimited distribution
                                          E-8

-------
reports are made available for sale from 240 agencies. Additionally, some state and local
government agencies contribute summaries of their reports to the database.  NTIS also provides
access to the results of government-sponsored research and development from countries outside
the United States. Organizations that currently contribute to the NTIS database include but are
not limited to the following: the Japan Ministry of International Trade and Industry; laboratories
administered by the United Kingdom Department of Industry; the German Federal Ministry of
Research and Technology; and the French National Center for Scientific Research.
       Pollution Abstracts provides access to environmental information that combines
information on scientific research and government policies in a single resource. Topics of
growing concern are extensively covered from the standpoints of atmosphere, emissions,
mathematical models, effects on people and animals,  and environmental action in response to
global pollution issues.  This database also contains material from conference proceedings and
hard-to-fmd summarized documents along with information from primary journals  in the field of
pollution.
       BIOSIS Previews® contains citations from Biological Abstracts® (BA) and Biological
Abstracts/Reports, Reviews, and Meetings® (BA/RRM) (formerly BioResearch Index®), the
major publications of BIOSIS®. These publications constitute the major English-language
service providing comprehensive worldwide coverage of research in the biological  and
biomedical sciences.  Biological Abstracts includes approximately 350,000  accounts of original
research yearly from nearly 5,000 primary journal and monograph titles. BA/RRM includes an
additional 200,000+ citations a year from meeting abstracts, reviews, books, book chapters,
notes, letters, and selected reports.
       IP A Toxicology provides focused toxicology information on all phases of the
development and use of drugs and on professional pharmaceutical practice.  The scope of the
database  ranges from the clinical and practical to the theoretical aspects of toxicology literature.
A unique feature of abstracts reporting clinical studies is the inclusion of the study design,
number of patients, dosage, dosage forms, and dosage schedule.
       Medical Literature, Analysis, and Retrieval System Online (MEDLINE),  produced by the
U.S. National Library of Medicine (NLM), is NLM's premier bibliographic database.  It contains
more than 15  million references to journal articles in life sciences with a concentration on
biomedicine.  The broad coverage of the database includes basic biomedical research and the
                                          E-9

-------
clinical sciences since 1950, including nursing, dentistry, veterinary medicine, pharmacy, allied
health, and preclinical sciences. MEDLINE also covers life sciences that are vital to biomedical
practitioners, researchers, and educators, including some aspects of biology, environmental
science, marine biology, and plant and animal science, as well as biophysics and chemistry.
MEDLINE is indexed using NLM's controlled vocabulary, Medical Subject Headings.
Approximately 400,000 records are added per year, of which more than 76% are in English.
MEDLINE contains AIDSLINE, HealthSTAR, Toxline, In Process (formerly known as
Pre-MEDLINE), In Data Review, and POPLINE.
      ToxFile covers the lexicological, pharmacological, biochemical, and physiological
effects of drugs and other chemicals. Adverse drug reactions, chemically induced diseases,
carcinogenesis, mutagenesis, teratogenesis, environmental pollution, waste disposal, radiation,
and food contamination are typical areas of coverage. The databases Environmental Mutagen
Information  Center, Developmental and Reproductive Toxicology, and Toxic Substances
Control Act  Test Submissions are included in ToxFile. It is not clearly stated whether the
Chemical Carcinogenesis Research Information System, Hazardous Substances Data Bank, or
Genetic Toxicology Data Bank are included in ToxFile.  Consequently,  a separate, online search
was conducted to ensure that these databases were searched.
      BIOSIS® Toxicology contains citations from BA and BA/RRM (formerly BioResearch
Index®), the  major publications of BIOSIS®, that focus on toxicology and related topics.
Records are  drawn from journal articles, conference papers, monographs and book chapters,
notes, letters, and reports, as well  as original research. U.S. patent records are also included.
      CANCERLIT® is produced by the International Cancer Research DataBank Branch of
the U.S. National Cancer Institute. The database consists of bibliographic records referencing
cancer research publications dating from 1963 to 2002.  Most records contain abstracts, and all
records contain citation information and additional descriptive fields such as document type and
language. Beginning with the June  1983 CANCERLIT® update, records from the MEDLINE
database  dealing with cancer topics have been added to CANCERLIT®.
      The RTECS® is a comprehensive database of basic toxicity information for over 150,000
chemical substances including prescription and nonprescription drugs, food additives, pesticides,
fungicides, herbicides, solvents, diluents, chemical wastes, reaction products of chemical waste,
and substances used in both industrial and household situations. Reports of the toxic effects of
                                          E-10

-------
each compound are cited. In addition to toxic effects and general toxicology reviews, data on
skin and/or eye irritation, mutation, reproductive consequences and tumorigenicity are provided.
Federal standards and regulations, National Institute for Occupational Safety and Health
(NIOSH) recommended exposure limits and information on the activities of EPA, NIOSH,
National Toxicology Program (NTP), and Occupational Safety and Health Administration
regarding the substance are also included. The toxic effects are linked to literature citations from
both published and unpublished governmental reports, and published articles from the scientific
literature.  The database corresponds to the print version of the RTECS®, formerly known as the
Toxic Substances List, which was started in 1971.  Originally prepared by the NIOSH, the
RTECS   database is now produced and distributed by Symyx  Technologies, Inc.
E.2.  TOXICOKINETIC MODELING CODE (Emond et al., 2005)
E.2.1.  Human Standard Model
E.2.1.1. Model Code
PROGRAM: Three Compartment PBPK Model for TCDD in Human: Standard Model
(Nongestation)'
INITIAL   !INITIALIZATION  OF PARAMETERS

      !SIMULATION PARAMETERS ====
CONSTANT  EXP_TIME_ON
(HOUR)
CONSTANT  EXP_TIME_OFF
(HOUR)
CONSTANT  DAYJCYCLE
(HOUR)
CONSTANT  BCK_TIME_ON
EXPOSURE  BEGINS  (HOUR)
CONSTANT  BCK_TIME_OFF
EXPOSURE  ENDS  (HOUR)

      ! EXPOSURE DOSES
CONSTANT   MSTOTBCKGR
(NG/KG)
CONSTANT   MSTOT
CONSTANT   DOSEIV
CONSTANT   MW
 MSTOT NM = MSTOT/MW
                                   0.

                                   6.132e5

                                   24.0

                                   6.132e5

                                   6.132e5
  !  TIME AT WHICH EXPOSURE BEGINS

!  TIME AT WHICH  EXPOSURE ENDS

!  NUMBER OF HOURS BETWEEN DOSES

   !  TIME AT WHICH BACKGROUND

   !  TIME AT WHICH BACKGROUND
                                          !  ORAL BACKGROUND  EXPOSURE DOSE
                                   l.OE-7       !  ORAL EXPOSURE DOSE  (NG/KG)
                                   0.0            !  INJECTED  DOSE (NG/KG)
                                 322.0        !  MOLECULAR WEIGHT (G/MOL)
                                              !  CONVERTS THE  DOSE TO NMOL/KG
 MSTOT_NMBCKGR = MSTOTBCKGR/MW  !CONVERTS  THE BACKGROUND DOSE TO NMOL/KG
  DOSEIV_NM = DOSEIV/MW                      !  CONVERTS THE  INJECTED DOSE TO
NMOL/KG

      !INITIAL GUESS OF THE FREE CONCENTRATION IN THE LIGAND (COMPARTMENT
INDICATED  BELOW) ====

                                       E-ll

-------
CONSTANT CFLLIO
0.0
     !  LIVER  (NMOL/L)
      !BINDING CAPACITY (AhR)  FOR NON LINEAR BINDING  (COMPARTMENT  INDICATED
BELOW) ===
CONSTANT LIBMAX         =0.35            !  LIVER   (NMOL/L)

      ! PROTEIN AFFINITY CONSTANTS (1A2 OR AhR, COMPARTMENT  INDICATED BELOW)
CONSTANT KDLI
ET AL.. 1997
CONSTANT KDLI2
AL. 2004
   0.1

   40.0
      !EXCRETION AND ABSORPTION CONSTANTS
CONSTANT KST            =       0.01
1), EMOND ET AL.,  2005
CONSTANT KABS           =       0.06
(HR-1), EMOND ET AL.  2005

      !ELIMINATION  CONSTANTS
CONSTANT CLURI          =      4.17D-8
ET AL., 2005
CONSTANT KELV           =     l.le-3
ELIMINATION CONSTANT  (I/HOUR)
    !  LIVER (AhR)  (NMOL/L) WANG

  !  LIVER (1A2)  (NMOL/L) EMOND ET



    !  GASTRIC RATE CONSTANT  (HR-

!  INTESTINAL ABSORPTION CONSTANT



!  URINARY CLEARANCE  (L/HR), EMOND

   !  INTERSPECIES VARIABLE
      !CONSTANT TO  DIVIDE THE ABSORPTION INTO LYMPHATIC AND  PORTAL FRACTIONS
CONSTANT A              =       0.7
WANG ET AL.  (1997)

     !PARTITION  COEFFICIENTS
CONSTANT PF             =       1.Oe2
WANG ET AL.  1997
CONSTANT PRE            =       1.5
WANG ET AL.  1997
CONSTANT PLI            =       6.0
AL. 1997

     !PARAMETERS  FOR INDUCTION OF CYP1A2
CONSTANT IND_ACTIVE      =       1.0
0 = NO)
CONSTANT CYP1A2_10UTZ   =   1.6e3
OF 1A2  (NMOL/L)
CONSTANT CYP1A2_1A1     =       1.6e3
(NMOL/L)
CONSTANT CYP1A2_1EC50   =       1.3e2
(NMOL/L)
CONSTANT CYP1A2_1A2     =       1.6e3
(NMOL/L)
CONSTANT CYP1A2_1KOUT   =       0.1
(H-l)
CONSTANT CYP1A2_1TAU   =       0.25
CONSTANT CYP1A2_1EMAX   =       9. 3e3
(UNITLESS)
CONSTANT HILL           =       0.6
BINDING EFFECT  CONSTANT (UNITLESS)
     !  DIFFUSIONAL PERMEABILITY FRACTION
CONSTANT PAFF           =      0.12
                           !  LYMPHATIC FRACTION,



                         ! ADIPOSE  TISSUE/BLOOD,

                         ! REST  OF  THE BODY/BLOOD,

                           !  LIVER/BLOOD,  WANG ET



                   ! INCLUDE  INDUCTION? (1 = YES,

           !  DEGRADATION CONCENTRATION CONSTANT

                ! BASAL CONCENTRATION OF 1A1

             !  DISSOCIATION  CONSTANT TCDD-CYP1A2

                ! BASAL CONCENTRATION OF 1A2

             !  FIRST ORDER RATE OF DEGRADATION

               ! HOLDING TIME (H)
           !  MAXIMUM INDUCTION  OVER BASAL EFFECT

            IHILL CONSTANT;  COOPERATIVE LIGAND


                          ! ADIPOSE (UNITLESS)
                                      E-12

-------
CONSTANT PAREF         =0.03                  !  REST OF BODY  (UNITLESS)
CONSTANT PALIF         =0.35                    !  LIVER  (UNITLESS)

     !TISSUE BLOOD FLOW  EXPRESSED AS  A FRACTION OF CARDIAC OUTPUT  =========
CONSTANT QFF           =       0.05       !  ADIPOSE TISSUE BLOOD FLOW  FRACTION
(UNITLESS), KRISHNAN 2008
CONSTANT QLIF          =       0.26       !  LIVER (UNITLESS), KRISHNAN 2008

     !COMPARTMENT TISSUE BLOOD  EXPRESSED AS A FRACTION OF THE TOTAL
COMPARTMENT VOLUME =========
CONSTANT WFBO          =       0.050     !  ADIPOSE TISSUE, WANG ET AL.  1997
CONSTANT WREBO       =        0.030     !  REST OF THE BODY, WANG ET AL.  1997
CONSTANT WLIBO         =       0.266     !  LIVER, WANG ET AL. 1997

     !EXPOSURE SCENARIO  FOR UNIQUE OR REPETITIVE WEEKLY OR MONTHLY EXPOSURE
     !NUMBER OF EXPOSURES  PER WEEK
CONSTANT WEEK_LAG     =       0.0           !  TIME ELAPSED BEFORE EXPOSURE
BEGINS  (WEEK)
CONSTANT WEEK_PERIOD   =       168.0      !  NUMBER OF HOURS IN THE WEEK
(HOURS)
CONSTANT WEEK_FINISH   =       168.0       !  TIME EXPOSURE ENDS  (HOURS)
    !NUMBER OF EXPOSURES PER  MONTH
CONSTANT MONTH_LAG    =       0.0         !  TIME ELAPSED BEFORE EXPOSURE
BEGINS  (MONTH)

     !SET FOR BACKGROUND EXPOSURE===========
     ITIME CONSTANT FOR  BACKGROUND EXPOSURE===========
CONSTANT Day_LAG_BG      =0.0           !  TIME ELAPSED BEFORE EXPOSURE
BEGINS  (HOUR)
CONSTANT Day_PERIOD_BG     =     24.0        !  LENGTH OF EXPOSURE  (HOUR)

     ITIME CONSTANT FOR  WEEKLY  EXPOSURE
CONSTANT WEEK_LAG_BG     =   0.0        !  TIME ELAPSED BEFORE BACKGROUND
EXPOSURE BEGINS  (WEEK)
CONSTANT WEEK_PERIOD_BG    =     168.0      !   NUMBER OF HOURS IN THE WEEK
(HOURS)
CONSTANT WEEK_FINISH_BG    =     168.0        !  TIME EXPOSURE ENDS  (HOURS)

     !  CONSTANT USED IN  CARDIAC OUTPUT EQUATION
CONSTANT QCC             =     15.36                   !  (L/KG-H),  EMOND ET AL.
2004

     !  COMPARTMENT TOTAL LIPID  FRACTION
     IData from Emonds Thesis 2001
CONSTANT F_TOTLIP          =     0.8000               !  ADIPOSE TISSUE
(UNITLESS)
CONSTANT B_TOTLIP          =     0.0057              !  BLOOD  (UNITLESS)
CONSTANT RE_TOTLIP         =     0.0190             !  REST OF THE BODY
(UNITLESS)
CONSTANT LI_TOTLIP         =     0.0670               !  LIVER (UNITLESS)
CONSTANT MEANLIPID         =     974.0

END  !  END OF THE INITIAL  SECTION
DYNAMIC !  DYNAMIC SIMULATION  SECTION
   i
                                      E-13

-------
ALGORITHM  IALG
CINTERVAL  CINT
MAXTERVAL  MAXT
MINTERVAL  MINT
VARIABLE   T
CONSTANT   TIMELIMIT
CONSTANT      YO
SIMULATION
CONSTANT   GROWON
GROWTH?  (1 = YES,  0 =  NO)
  CINTXY  = CINT
  PFUNC   = CINT

  DAY=T/24.0
  WEEK =T/168.0
  MONTH =T/730. 0
  YEAR=YO+T/8760.0
  GYR =YO + growon*T/8760.0
                                2
                                10.0
                               .Oe+10
                               .OE-10
                                0.0
                               1.752e5
                                   0.0

                                  1.0
        !  GEAR METHOD
          !  COMMUNICATION INTERVAL
        !MAXIMUM INTERVAL CALCULATION
        !MINIMUM INTERVAL CALCULATION

       !SIMULATION LIMIT TIME  (HOUR)
           !  AGE (YEARS) AT BEGINNING OF
        i
                                           INCLUDE BODY WEIGHT AND HEIGHT
                                               !  TIME IN DAYS
                                            !  TIME IN WEEKS
                                          !  TIME IN MONTHS
                                       !  TIME IN YEARS
                                    TIME  FOR USE IN GROWTH EQUATION  (YEARS)
DERIVATIVE  ! PORTION  OF  CODE THAT SOLVES DIFFERENTIAL EQUATIONS

      ! CHRONIC OR  SUBCHRONIC EXPOSURE SCENARIO =======
      ! NUMBER OF EXPOSURES  PER DAY
 DAY_LAG
 (HOURS)
 DAY_PERIOD
 DAY FINISH
             = EXP TIME  ON
             = DAY_CYCLE
             = CINTXY
MONTH_PERIOD = TIMELIMIT
MONTH FINISH = EXP TIME OFF
!  TIME ELAPSED BEFORE EXPOSURE BEGINS

 !  EXPOSURE PERIOD (HOURS)
 !  LENGTH OF EXPOSURE (HOURS)
 !  EXPOSURE PERIOD (MONTHS)
 !  LENGTH OF EXPOSURE (MONTHS)
      ! NUMBER OF EXPOSURES  PER DAY AND MONTH
 DAY_FINISH_BG   =  CINTXY
 MONTH_LAG_BG   = BCK_TIME_ON    ITIME ELAPSED BEFORE BACKGROUND EXPOSURE
BEGINS  (MONTHS)
 MONTH_PERIOD_BG =  TIMELIMIT       !  BACKGROUND EXPOSURE PERIOD  (MONTHS)
 MONTH_FINISH_BG =  BCK_TIME_OFF   !  LENGTH OF BACKGROUND EXPOSURE  (MONTHS)

 B =  1.0-A  ! FRACTION  OF DIOXIN ABSORBED IN THE PORTAL FRACTION OF THE  LIVER

      !HUMAN BODY WEIGHT GROWTH EQUATION========
      !  POLYNOMIAL  REGRESSION EXPRESSION WRITTEN
IAPRIL 10 2008, OPTIMIZED WITH DATA OF PELEKIS ET AL.  2001
!  POLYNOMIAL REGRESSION EXPRESSION WRITTEN WITH
!HUH AND BOLCH 2003  FOR BMI CALCULATION

  !  BODY WEIGHT CALCULATION
  WTO = (0.0006*GYR**3 - 0.0912*GYR**2 + 4.32*GYR + 3.652)! BODY WEIGHT IN  KG

   !   BODY MASS INDEX CALCULATION
      BH =  -2D-5*GYR**4+4.2D-3*GYR**3.0-0.315*GYR**2.0+9.7465*GYR+72.098

   !HEIGHT EQUATION  FORMULATED FOR USE FROM 0 TO 70 YEARS
      BHM=  (BH/100.0)                IHUMAN HEIGHT IN METERS  (BHM)
      HBMI= WTO/(BHM**2.0)  ! HUMAN BODY MASS INDEX (BMI)

    !  ADIPOSE TISSUE FRACTION
                                      E-14

-------
    WTOGR= WTO*1.0e3     !  BODY WEIGHT IN GRAMS
    WFO= -6.36D-20*WTOGR**4.0  +1.12D-14*WTOGR**3.0 -5.8D-10*WTOGR**2.0  +1.2D-
5*WTOGR+5.91D-2

    !  LIVER,VOLUME  FRACTION
    !  APPROACH BASED  ON  LUECKE (2007)
    WLIO= (3.59D-2  -(4.76D-7*WTOGR)+(8.50D-12*WTOGR**2.0)-(5.45D-
17*WTOGR**3.0))

 WREO =  (0.91 -(WLIBO*WLIO+WFBO*WFO+WLIO+WFO))/(1.0+WREBO)
                                     IREST OF THE BODY FRACTION; UPDATED FOR
EPA ASSESSMENT
 QREF = 1.0-(QFF+QLIF)                 IREST OF BODY BLOOD FLOW
 QTTQF = QFF+QREF+QLIF               !  SUM MUST EQUAL  1

   !COMPARTMENT VOLUME  (L  OR KG)  =========
 WF  =  WFO  * WTO                    !  ADIPOSE
 WRE =  WREO * WTO                    !  REST OF THE BODY
 WLI =  WLIO * WTO                    !  LIVER
 WB=0.075*WTO                           !  BLOOD

   !COMPARTMENT TISSUE BLOOD (L OR KG)  =========
 WFB  =  WFBO  * WF                   !  ADIPOSE
 WREB =  WREBO * WRE                   !  REST OF THE BODY
 WLIB =  WLIBO * WLI                   !  LIVER
   !CARDIAC OUTPUT  FOR THE GIVEN BODY WEIGHT
QC= QCC*(WTO**0.75)                     !  [L BLOOD/HOUR]

QF  = QFF*QC                            !  ADIPOSE TISSUE BLOOD FLOW  RATE
[L/HR]
QLI = QLIF*QC                           !  LIVER TISSUE BLOOD  FLOW RATE  [L/HR]
QRE = QREF*QC                        IREST OF THE BODY BLOOD  FLOW RATE  [L/HR]

QTTQ = QF+QRE+QLI                  !  TOTAL FLOW RATE [L/HR]

   !PERMEABILITY ORGAN FLOW [L/HR]=======
PAF  = PAFF*QF                           !  ADIPOSE
PARE = PAREF*QRE                         !  REST OF THE BODY
PALI = PALIF*QLI                         !  LIVER TISSUE

   !  ABSORPTION SECTION
   !  INTRAVENOUS
 IV        =    DOSEIV_NM  * WTO          !AMOUNT IN NMOL
 MSTTBCKGR =    MSTOT_NMBCKGR *WTO       !AMOUNT IN NMOL
 MSTT      =    MSTOT_NM * WTO           !AMOUNT IN NMOL

      !REPETITIVE ORAL BACKGROUND EXPOSURE SCENARIOS
DAY_EXPOSURE_BG   = PULSE(DAY_LAG_BG,DAY_PERIOD_BG,DAY_FINISH_BG)
WEEK_EXPOSURE_BG  = PULSE(WEEK_LAG_BG,WEEK_PERIOD_BG,WEEK_FINISH_BG)
MONTH_EXPOSURE_BG = PULSE(MONTH_LAG_BG,MONTH_PERIOD_BG,MONTH_FINISH_BG)

MSTTCH_BG =  (DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG)*MSTTBCKGR
MSTTFR_BG = MSTTBCKGR/CINT

CYCLE BG =DAY EXPOSURE BG*WEEK EXPOSURE BG*MONTH EXPOSURE BG
                                      E-15

-------
     ! CONDITIONAL  ORAL EXPOSURE (BACKGROUND EXPOSURE)
IF  (MSTTCH_BG.EQ.MSTTBCKGR)  THEN
    ABSMSTT_GB= MSTTFR_BG
ELSE
    ABSMSTT_GB =0.0
END IF
     !REPETITIVE  ORAL  MAIN EXPOSURE SCENARIO
DAY_EXPOSURE   = PULSE(DAY_LAG,DAY_PERIOD,DAY_FINISH)
WEEK_EXPOSURE  = PULSE(WEEK_LAG,WEEK_PERIOD,WEEK_FINISH)
MONTH_EXPOSURE = PULSE(MONTH_LAG,MONTH_PERIOD,MONTH_FINISH)

MSTTCH =  (DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE)*MSTT
CYCLE = DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE
MSTTFR=MSTT/CINT

     !CONDITIONAL ORAL EXPOSURE
IF  (MSTTCH.EQ.MSTT) THEN
   ABSMSTT= MSTTFR
ELSE
   ABSMSTT = 0.
END IF

 CYCLETOT=INTEG(CYCLE, 0.0)

      !  MASS Balance  CHANGE IN THE LUMEN
RMSTT= -(KST+KABS)*MST+ABSMSTT +ABSMSTT_GB  ! RATE OF CHANGE  (NMOL/H)
 MST = INTEG(RMSTT,0.)                       !AMOUNT  REMAINING IN GI TRACT
(NMOL)

      !  ABSORPTION  IN LYMPH CIRCULATION
LYRMLUM = KABS*MST*A
 LYMLUM = INTEG(LYRMLUM,0.0)

      !  ABSORPTION  IN PORTAL CIRCULATION
LIRMLUM = KABS*MST*B
  LIMLUM = INTEG(LIRMLUM,0.0)
       !IV ABSORTPION  SCENARIO 	
 IVR=  IV/PFUNC  !  RATE FOR IV INFUSION IN BLOOD
 EXPIV= IVR *  (l.O-STEP(PFUNC))
 IVDOSE = integ(EXPIV,0.0)

       !SYSTEMIC  BLOOD COMPARTMENT
       ! MODIFICATION  OCT  8  2009
CB=(QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM)/(QC+CLURI)  !
 CA =  CB                                    !CONCENTRATION  (NMOL/L)

    !CB=(QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM-RAURI)/QC  !
    !   CA = CB                             !  CONCENTRATION  (NMOL/L)

       !URINARY  EXCRETION BY KIDNEY
       !  MODIFICATION OCT 8 2009
RAURI  = CLURI *CB
  AURI = INTEG(RAURI,0.0)

                                      E-16

-------
        !CONCENTRATION  UNIT

  CBSNGKGLIADJ =  CB*MW/(0.55*B_TOTLIP)  !serum  concentration in lipid adjust
(PG/G LIPID=PPT)
      CBPPT = CBSNGKGLIADJ
  CBNGKG = CB*MW

CBpptRH = CB*MW*10000/(0.55*MEANLIPID)  !SERUM  CONCENTRATION IN LIPID ADJUST
(PG/G LIPID=PPT)

    AUCJCBSNGKGLIADJ=INTEG(CBSNGKGLIADJ,0.0)

      IADIPOSE TISSUE  COMPARTMENT
RAFB= QF*(CA-CFB)-PAF*(CFB-CF/PF)         !(NMOL/HR)
 AFB = INTEG(RAFB,0.0)                      !(NMOL)
 CFB = AFB/WFB                             !(NMOL/KG)
      !TISSUE SUBCOMPARTMENT
RAF = PAF*(CFB-CF/PF)                      !(NMOL/HR)
 AF = INTEG(RAF,0.0)                        !(NMOL)
 CF  = AF/WF                               !(NMOL/KG)

      IPOST SIMULATION UNIT  CONVERSION
CFTOTAL =  (AF + AFB)/(WF  + WFB)   ! TOTAL CONCENTRATION  NMOL/L
CFNGKG =CFTOTAL*MW

      IREST OF THE BODY COMPARTMENT========
RAREB= QRE*(CA-CREB)-PARE*(CREB-CRE/PRE)  !(NMOL/HR)
 AREB = INTEG(RAREB,0.0)                   !(NMOL)
 CREB = AREB/WREB                         !(NMOL/KG)
      !TISSUE SUBCOMPARTMENT
RARE = PARE*(CREB-CRE/PRE)                 !(NMOL/HR)
 ARE = INTEG(RARE,0.0)                   !(NMOL)
 CRE  = ARE/WRE                             !(NMOL/KG)

      IPOST SIMULATION UNIT  CONVERSION
CRETOTAL =  (ARE + AREB)/(WRE + WREB)  ! TOTAL CONCENTRATION IN NMOL/L

      !LIVER COMPARTMENT
      !TISSUE BLOOD  SUBCOMPARTMENT
RALIB = QLI*(CA-CLIB)-PALI*(CLIB-CFLLIR)+LIRMLUM           !(NMOL/HR)
  ALIB = INTEG(RALIB,0.0)                                    !(NMOL)
 CLIB = ALIB/WLIB
      !TISSUE SUBCOMPARTMENT
 RALI =  PALI*(CLIB-CFLLIR)-REXCLI                         !(NMOL/HR)
  ALI = INTEG(RALI,0.0)                  !(NMOL)
  CLI  = ALI/WLI                  !(NMOL/KG)
       !FREE TCDD  IN  LIVER
       ! MODIFICATION OCTOBER 8 2009
CFLLI= IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLIR))  &
        + ( (CYP1A2_103*CFLLIR/(KDLI2+CFLLIR)*IND_ACTIVE) ) )-CFLLI, CFLLIO)   !
CONCENTRATION OF  FREE TCDD IN LIVER
    CFLLIR=DIM(CFLLI,0.0)


                                      E-17

-------
!MODIFIED FROM:
      !PARAMETER  (LIVER_1RMN = l.OE-30)
      ! CFLLI=  IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLIR  &        !
+LIVER_1RMN))+((CYP1A2_103*CFLLIR/(KDLI2+CFLLIR &
      !         +LIVER_1RMN)*IND_ACTIVE)))-CFLLI,CFLLIO)
      !    CFLLIR=DIM(CFLLI,0.0)


CBNDLI= LIBMAX*CFLLIR/(KDLI+CFLLIR)  !CONG OF TCDD BOUDN TO AhR

!CBNDLI= LIBMAX*CFLLIR/(KDLI+CFLLIR+LIVER_1RMN) !CONG BIND

      !POST  SIMULATION UNIT CONVERSION
CLITOTAL =  (ALI + ALIB)/(WLI + WLIB)        !  TOTAL CONCENTRATION IN  NMOL/L
rec_occ_AHR= 100.0*CFLLIR/(KDLI+CFLLIR+1.0)   !  PERCENT  BOUND TO AhR
OCCUPANCY
PROT_occ_lA2=  100.0*CFLLIR/(KDLI2+CFLLIR)     !  PERCENT  BOUND TO 1A2
OCCUPANCY
CLINGKG= CLITOTAL*MW                        ![NG TCDD/KG]
CBNDLINGKG = CBNDLI*MW

     !FRACTION  INCREASE OF INDUCTION OF CYP1A2
fold_ind=CYP!A2_10UT/CYPlA2_lA2
VARIATIONOFAC  =(CYP1A2_10UT-CYP1A2_1A2)/CYP1A2_1A2

     !VARIABLE  ELIMINATION BASED ON THE CYP1A2
KBILE_LI_T = Kelv*VARIATIONOFAC!

 REXCLI = KBILE_LI_T*CFLLIR*WLI !  DOSE-DEPENDENT RATE OF BILLIARY  EXCRETION
OF DIOXIN
    EXCLI =  INTEG(REXCLI,0.0)   !TOTAL AMOUNT OF DIOXIN EXCRETED

     !CHEMICAL  IN CYP450  (1A2)  COMPARTMENT
     !PARAMETER FOR INDUCTION OF CYP1A2

CYP1A2_1KINP = CYP1A2_1KOUT*CYP1A2_10UTZ !  BASAL RATE OF CYP1A2  PRODUCTION
SET EQUAL TO BASAL RATE OF  DEGRDATION AT STEADY STATE

     ! MODIFICATION OCTOBER  8 2009
CYP1A2_10UT  =INTEG(CYP1A2_1KINP *  (1.0 + CYP1A2_1EMAX *(CBNDLI+1.Oe-30)**HILL
&
     /(CYP1A2_1EC50**HILL + (CBNDLI+1.Oe-30)**HILL)) &
      - CYP1A2_1KOUT*CYP1A2_10UT,  CYP1A2_10UTZ) !  LEVELS OF CYP1A2
!  MODEIFIED  FROM:
!PARAMETER  (CYP1A2_1RMN = le-30)
!CYP1A2_10UT =INTEG(CYP1A2_1KINP * (1 + CYP1A2_1EMAX *(CBNDLI &
!      +CYP1A2_1RMN)**HILL/(CYP1A2_1EC50 + (CBNDLI + CYP1A2_1RMN)**HILL)  &
!      +CYP1A2_1RMN) - CYP1A2_1KOUT*CYP1A2_1&
!      OUT, CYP1A2_10UTZ)

!  EQUATIONS  INCORPORATING DELAY OF CYP1A2 PRODUCTION  (NOT USED IN
SIMULATIONS)
CYP1A2_1R02  =  (CYP1A2_10UT  - CYP1A2_102)/ CYP1A2_1TAU
    CYP1A2_102 =INTEG(CYP1A2_1R02, CYP1A2_1A1)
 CYP1A2_1R03 =  (CYP1A2_102  - CYP1A2_103)/ CYP1A2_1TAU
    CYP1A2_103 =INTEG(CYP1A2_1R03, CYP1A2_1A2)


                                      E-18

-------
      !CHECK MASS BALANCE
  BDOSE= LYMLUM+LIMLUM+IVDOSE
  BMASSE = EXCLI+AURI+AFB+AF+AREB+ARE+ALIB+ALI
      BDIFF = BDOSE-BMASSE
      ! BODY BURDEN  IN  TERMS  OF CONCENTRATION  (NG/KG)
  BBNGKG =  (AFB+AF+AREB+ARE+ALIB+ALI)*MW/WTO        !

      !COMMAND END OF THE  SIMULATION
TERMT (T.GE. TIMELIMIT,  'Time  limit has been reached.

END   !  END OF THE  DERIVATIVE  SECTION
END   !  END OF THE  DYTNAMIC  SECTION
END   !  END OF THE  PROGRAM
E.2.1.2. Input File
output @clear
prepare @clear year T  CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG  CBNDLINGKG CBNGKG
% PARAMETERS FOR SIMULATION
CINT = 1 %0.5
                           %  TIME AT WHICH EXPOSURE BEGINS (HOUR)
                           %324120     % HOUR/YEAR ITIME AT WHICH EXPOSURE
EXP_TIME_ON  =    0.
EXP_TIME_OFF = 613200
ENDS  (HOUR)
DAYJCYCLE    =24
BCK_TIME_ON  = 613200
BEGINS  (HOUR)
BCK_TIME_OFF = 613200
ENDS  (HOUR)
TIMELIMIT    = 613200
(HOUR)
MSTOTBCKGR   =    0.
                           %  NUMBER OF HOURS BETWEEN DOSES  (HOUR)
                           %324120      % TIME AT WHICH BACKGROUND EXPOSURE

                           %324120      % TIME AT WHICH BACKGROUND EXPOSURE

                           %324120        %324120     % SIMULATION TIME LIMIT

                           %  ORAL BACKGROUND EXPOSURE DOSE  (UG/KG)
% oral dose oral dose  oral  dose
MSTOT        = 9.97339283634997E-07
DOSEIV       =   0           %NG/KG
% oral dose oral dose  oral  dose
                                           % ORAL DAILY EXPOSURE DOSE  (NG/KG)
MEANLIPID    = 730
IND ACTIVE= 1
                          INDUCTION INCLUDED? (1=YES, 0=NO)
E.2.2.  Human Gestational Model

E.2.2.1. Model Code

PROGRAM: Three Compartment PBPK Model for TCDD in Human (Gestation)'

INITIAL   !
      !SIMULATION PARAMETERS
CONSTANT PARA_ZERO           =  le-30
CONSTANT EXP_TIME_ON      =0.0
CONSTANT EXP_TIME_OFF     =  530.0
CONSTANT DAYJCYCLE        =   24.0
CONSTANT BCK_TIME_ON      =0.0
BEGINS  (HOURS)
                                      ITIME AT WHICH EXPOSURE BEGINS  (HOURS)
                                       ITIME AT WHICH EXPOSURE ENDS  (HOURS)
                                      I NUMBER OF HOURS BETWEEN DOSES  (HOURS)
                                      ITIME AT WHICH BACKGROUND EXPOSURE
                                      E-19

-------
CONSTANT BCK_TIME_OFF    =0.0       ITIME AT WHICH BACKGROUND  EXPOSURE ENDS
(HOURS)
CONSTANT TRANSTIME_ON    =0.0       !CONTROL TRANSFER  FROM MOTHER TO FETUS
AT 9 WEEKS OR  1512  HOURS  OF GESTATION

      ! INTRAVENOUS  SEQUENCY
CONSTANT IV_LAG         =0.0
CONSTANT IV_PERIOD        =0.0

      !PREGNANCY  PARAMETER
CONSTANT CONCEPTION_T         =0.0        ITIME OF CONCEPTION  (HOUR)
CONSTANT PFETUS           =4.0       !PARTITION COEFFICIENT
CONSTANT CLPLA_FET        = l.Oe-3    !CLEARANCE TRANSFER  FOR MOTHER TO FETUS
(L/HR)

      !CONSTANT EXPOSURE CONTROL
      IACUTE, SUBCHRONIC,  CHRONIC EXPOSURE =====
      !OR BACKGROUND EXPOSURE (IN THIS CASE 3 TIMES A DAY)===
CONSTANT MSTOTBCKGR      =0.0       !  ORAL BACKGROUND  EXPOSURE DOSE (NG/KG)
CONSTANT MSTOT            =0.0       !  ORAL EXPOSURE DOSE (NG/KG)

      IORAL ABSORPTION
      i MSTT= MSTOT/1000 *WTO *1/322*1000 IAMOUNT IN NMOL
  MSTOT_NM = MSTOT/MW              !CONVERTS THE DOSE TO  NMOL/KG

      !INTRAVENOUS ABSORPTION
CONSTANT  DOSEIV         =0.0       !  INJECTED DOSE  (NG/KG)
  DOSEIV_NM =  DOSEIV/MW            !  CONVERTS THE INJECTED  DOSE TO NMOL/KG
CONSTANT DOSEIVLATE =0.0            !INJECTED DOSE LATE  (NG/KG)
  DOSEIVNMlate = DOSEIVLATE/MW     IAMOUNT IN NMOL/G

      !INITIAL  GUESS OF THE FREE CONCENTRATION IN THE LIGAND (COMPARTMENT
INDICATED BELOW)====
CONSTANT CFLLIO           =  0.0      !LIVER     (NMOL/L)
CONSTANT CFLPLAO         =  0.0      !PLACENTA  (NMOL/L)

      !BINDING  CAPACITY (AhR)  FOR NON LINEAR BINDING  (COMPARTMENT INDICATED
BELOW)  (NMOL/L)  ===
CONSTANT LIBMAX           =0.35     !  LIVER   (NMOL/L)
CONSTANT PLABMAX         =0.2      !TEMPORARY PARAMETER

      !PROTEIN AFFINITY CONSTANTS (1A2 OR AhR, COMPARTMENT INDICATED BELOW)
(NMOL/ML)===
CONSTANT KDLI             = 0.1      !LIVER (AhR)  (NMOL/L),  WANG ET AL.  1997
CONSTANT KDLI2            = 40.0       !LIVER  (1A2)  (NMOL/L),  EMOND ET AL.
2004
CONSTANT KDPLA            =0.1      IASSUME IDENTICAL TO  KDLI (AhR)

      !EXCRETION  AND ABSORPTION CONSTANT
CONSTANT KST              =0.01     !  GASTRIC RATE CONSTANT (HR-1),  EMOND ET
AL. 2005
CONSTANT KABS             = 0.06     !  INTESTINAL ABSORPTION CONSTANT (HR-1),
EMOND ET AL.  (2005)

      !INTERSPECIES  ELIMINATION CONSTANT
      ITEST ELIMINATION VARIABLE, EMOND ET AL. 2005
                                      E-20

-------
CONSTANT KELV
ELIMINATION CONSTANT  (I/HOUR)
                              l.le-3 I4.0D-3
                           !  INTERSPECIES VARIABLE
      ! ELIMINATION  CONSTANTS
CONSTANT CLURI            =   4.17e-£
2005
           URINARY CLEARANCE  (L/HR),  EMOND ET AL.
      ! CONSTANT TO  DIVIDE  THE ABSORPTION INTO LYMPHATIC AND PORTAL  FRACTIONS
CONSTANT A               =0.7        !  LYMPHATIC FRACTION, WANG  ET AL.  1997
      !PARTITION COEFFICIENTS
CONSTANT PF
CONSTANT PRE
1997
CONSTANT PLI
CONSTANT PPLA
WANG ET AL. 1997
  Oe2     !  ADIPOSE TISSUE/BLOOD,  WANG ET AL.  1997
  5      !  REST OF THE BODY/BLOOD,  WANG ET AL.

  0       !  LIVER/BLOOD, WANG  ET AL.  1997
  5      !  TEMPORARY PARAMETER NOT CONFIGURED,
     !PARAMETER FOR  INDUCTION OF CYP
CONSTANT IND_ACTIVE       =1.0
CONSTANT CYP1A2_10UTZ     = 1.6e3
1A2  (NMOL/L)
CONSTANT CYP1A2_1A1      = 1.6e3
CONSTANT CYP1A2_1EC50     = 1.3e2
(NMOL/L)
CONSTANT CYP1A2_1A2      = 1.6e3
CONSTANT CYP1A2_1KOUT     =0.1
CONSTANT CYP1A2_1TAU      =0.25
CONSTANT CYP1A2_1EMAX     = 9.3e3
(UNITLESS)
CONSTANT HILL             =0.6
BINDING EFFECT CONSTANT  (UNITLESS)
         1A2, WANG ET AL.  1997
             ! INCLUDE INDUCTION?  (1  =  YES,  0 = NO)
          !  DEGRADATION CONCENTRATION  CONSTANT OF

          !  BASAL CONCENTRATION OF  1A1 (NMOL/L)
         !  DISSOCIATION CONSTANT  TCDD-CYP1A2

          !BASAL CONCENTRATION OF 1A2  (NMOL/L)
         !  FIRST ORDER RATE OF DEGRADATION  (H-l)
         !HOLDING TIME  (H)
          !  MAXIMUM INDUCTION OVER  BASAL EFFECT

         IHILL CONSTANT;  COOPERATIVE LIGAND
    IDIFFUSIONAL  PERMEABILITY FRACTION,  WANG ET AL  (1997)
CONSTANT PAFF             =0.12      !  ADIPOSE (UNITLESS)
CONSTANT PAREF            =0.03      !  REST OF THE BODY  (UNITLESS)
CONSTANT PALIF            =0.35      !  LIVER (UNITLESS)
CONSTANT PAPLAF           =0.3      !  OPTIMIZED PARAMETER

!TISSUE BLOOD FLOW  EXPRESSED AS  A FRACTION OF CARDIAC OUTPUT, KRISHNAN  2007
CONSTANT QFF              =0.05      !  ADIPOSE TISSUE BLOOD FLOW  FRACTION
(UNITLESS), KRISHNAN  2008
CONSTANT QLIF             =  0.26      !  LIVER (UNITLESS), KRISHNAN 2008
!===FRACTION OF TISSUE  BLOOD WEIGHT Wang et al
                       [1997;
CONSTANT WFBO
CONSTANT WREBO
CONSTANT WLIBO
CONSTANT WPLABO
0.050    IADIPOSE TISSUE, WANG  ET AL.  1997
0.030    IREST OF THE BODY, WANG ET AL.  1997
0.266    !LIVER, WANG ET AL.  1997
0.500    !ASSUME HIGHLY VASCULARIZED
!  EXPOSURE SCENARIO  FOR UNIQUE OR REPETITIVE WEEKLY OR MONTHLY EXPOSURE
!  NUMBER OF EXPOSURES  PER WEEK
CONSTANT WEEK_LAG        =0.0        ITIME ELAPSED BEFORE EXPOSURE  BEGINS
(WEEK)
CONSTANT WEEK_PERIOD     = 168.0      !  NUMBER OF HOURS IN THE WEEK  (HOURS)
CONSTANT WEEK FINISH     = 168.0      !  TIME EXPOSURE ENDS  (HOURS)
                                      E-21

-------
!  NUMBER OF EXPOSURES PER MONTH
CONSTANT MONTH_LAG      =0.0
(MONTHS)
     ITIME ELAPSED BEFORE EXPOSURE  BEGINS
!======= CONSTANT FOR BACKGROUND  EXPOSURE
CONSTANT Day_LAG_BG     =0.0
(HOURS)
CONSTANT Day PERIOD BG   =24.0
!  NUMBER OF EXPOSURES PER WEEK
CONSTANT WEEK_LAG_BG    =0.0
BEGINS  (WEEK)
CONSTANT WEEK_PERIOD_BG  =  168.0
CONSTANT WEEK FINISH BG  =  168.0
     !  TIME ELAPSED BEFORE EXPOSURE  BEGINS

       !LENGTH OF EXPOSURE  (HOURS)
    ITIME ELAPSED BEFORE BACKGROUND  EXPOSURE

       !  NUMBER OF HOURS IN THE WEEK (HOURS)
       ITIME EXPOSURE ENDS  (HOURS)
!  CONSTANT USED IN CARDIAC  OUTPUT  EQUATION
CONSTANT QCC            =   15.36     ![L/KG-H],  EMOND ET AL.  2004

!  COMPARTMENT LIPID EXPRESSED AS THE  FRACTION OF TOTAL LIPID
IData from Emonds Thesis 2001
CONSTANT F_TOTLIP
CONSTANT B_TOTLIP
CONSTANT RE_TOTLIP
CONSTANT LI_TOTLIP
CONSTANT PLA_TOTLIP
CONSTANT FETUS TOTLIP
0.8000        ! ADIPOSE TISSUE  (UNITLESS)
0.0057        ! BLOOD  (UNITLESS)
0.0190        ! REST OF THE BODY  (UNITLESS)
0.0670        ! LIVER  (UNITLESS)
0.019       !  PLACENTA  (UNITLESS)
0.019     !  FETUS  (UNITLESS)
CONSTANT MEANLIPID
                                974
END
       END OF THE INITIAL  SECTION
DYNAMIC !  DYNAMIC SIMULATION  SECTION
ALGORITHM
CINTERVAL
MAXTERVAL
MINTERVAL
VARIABLE
CONSTANT
CONSTANT
SIMULATION
CONSTANT
IALG
CINT
MAXT
MINT
T
TIMELIMIT
YO
GROWON
GROWTH?  (1=YES, 0=NO)

 CINTXY  = CINT
 PFUNC   = CINT
                                      2
                                     0.1
                                   l.Oe+10
                                   l.OE-10
                                     0.0
                                     100
                                       0.0

                                     1.0
                ! GEAR METHOD
                ! COMMUNICATION  INTERVAL
                ! MAXIMUM CALCULATION  INTERVAL
                ! MINIMUM CALCULATION  INTERVAL

                !SIMULATION LIMIT  TIME (HOUR)
                !  AGE  (YEARS) AT  BEGINNING OF

               ! INCLUDE BODY WEIGHT AND HEIGHT
   ITIME TRANSFORMATION
 DAY= T/24.0
 WEEK  =T/168.0
 YEAR=YO+T/8760.0
  GYR =YO + growon*T/8760.0
EQUATION
                  ! TIME IN YEARS
                  ! TIME FOR USE IN  GROWTH
DERIVATIVE  ! PORTION OF CODE  THAT  SOLVES DIFFERENTIAL EQUATIONS
                                      E-22

-------
!====== CHRONIC OR SUBCHRONIC  EXPOSURE SCENARIO =======
!  NUMBER OF EXPOSURES  PER  DAY

 DAY_LAG         = EXP_TIME_ON    !  TIME ELAPSED BEFORE EXPOSURE BEGINS
(HOURS)
 DAY_PERIOD       = DAYJCYCLE       !  EXPOSURE PERIOD (HOURS)
 DAY_FINISH       = CINTXY         !  LENGTH OF EXPOSURE (HOURS)
 MONTH_PERIOD     = TIMELIMIT       !  EXPOSURE PERIOD (MONTHS)
 MONTH FINISH     = EXP TIME OFF    !  LENGTH OF EXPOSURE (MONTHS)
!  NUMBER OF EXPOSURES  PER  DAY AND  MONTH
 DAY_FINISH_BG    = CINTXY
 MONTH_LAG_BG    = BCK_TIME_ON     ITIME ELAPSED BEFORE BACKGROUND EXPOSURE
BEGINS  (MONTHS)
 MONTH_PERIOD_BG  = TIMELIMIT       !BACKGROUND EXPOSURE PERIOD  (MONTHS)
 MONTH_FINISH_BG  = BCK_TIME_OFF    !LENGTH OF BACKGROUND EXPOSURE  (MONTHS)

!  INTRAVENOUS LATE
 IV_FINISH = CINTXY
 B = 1-A !  FRACTION OF DIOXIN ABSORBED IN THE PORTAL FRACTION OF THE LIVER

!  MOTHER BODY WEIGHT GROWTH  EQUATION
!  MODIFICATION TO ADAPT THIS MODEL AT  HUMAN MODEL
!  BECAUSE LINEAR DESCRIPTION IS NOT GOOD ENOUGH FOR MOTHER GROWTH
!  MOTHER BODY WEIGHT GROWTH
!  HUMAN BODY WEIGHT  (0 TO  45 YEARS)
!  POLYNOMIAL REGRESSION EXPRESSION WRITTEN
IAPRIL 10 2008, OPTIMIZED  WITH DATA OF PELEKIS ET AL.  2001
!  POLYNOMIAL REGRESSION EXPRESSION WRITTEN WITH
!HUH AND BOLCH 2003 FOR BMI  CALCULATION

!  BODY WEIGHT CALCULATION.   UNIT IN KG FOR GESTATIONAL PORTION

     WTO = (0.0006*GYR**3  -  0.0912*GYR**2 + 4.32*GYR + 3.652)

 IBODY MASS INDEX CALCULATION

     BH =  -2D-5*GYR**4+4.2D-3*GYR**3.0-0.315*GYR**2.0+9.7465*GYR+72.098
!HEIGHT EQUATION FORMULATED  FOR USE FROM 0 TO 70 YEARS
     BHM= (BH/100.0)IHUMAN HEIGHT  IN METER (BHM)
     HBMI= WTO/(BHM**2.0)  !  HUMAN  BODY MASS INDEX (BMI)
MODIFICATION IN KG
RTESTGEST= T-CONCEPTION_T  !  TIME  FOR FETAL GROWTH
TESTGEST=DIM(RTESTGEST,0.0)
!  GROWTH OF FETAL TISSUE
GESTATTION_FE= ( (4d-15*TESTGEST**4  -3d-ll*TESTGEST**3 +ld-7*TESTGEST**2 -8d-
5*TESTGEST +0.0608) )
   WTFER= DIM(GESTATTION_FE, 0. 0)  !  FETAL COMPARTMENT WEIGHT
 WTFE= WTFER

\lllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllll
!  FAT GROWTH EXPRESSION LINEAR DURING PREGNANCY
!  FROM 0'FLAHERTY1992
                                      E-23

-------
WTOGR= WTO*1.0e3     ! MOTHER  BODY WEIGHT IN G

WFO =(-6.36D-20*WTOGR**4.0  +1 . 12D-14*WTOGR**3 . 0 &
           -5.8D-10*WTOGR**2.0+1.2D-5*WTOGR+5.91D-2)  !  MOTHER FAT COMPARTMENT
GROWTH

\lllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllll
!  WPLA PLACENTA GROWTH  EXPRESSION,  SINGLE EXPONENTIAL WITH OFFSET
!  FROM 0'FLAHERTY1992   ! FOR EACH PUP
ISAME EQUATION THEN THE  FORST  MODEL.  BODY WEIGHT KEPT IN G
!A CORRECTION FOR THE  BODY WEIGHT  (WTO (KG) *1000 = WTOGR)

WPLAON_HUMAN=  ( 850*exp (-9 . 434* (exp (-5 . 23d-4* (TESTGEST) )  ) ) )
 WPLAOR = WPLAON_HUMAN/WTOGR
 WPLAOW = DIM (WPLAOR, 0.0)  ! PLACENTA WEIGHT
  WPLAO=WPLAOW

\lllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllll
!  QPLA PLACENTA GROWTH EXPRESSION,  DOUBLE EXPONENTIAL WITH OFFSET
!  FROM 0'FLAHERTY1992
QPLAF_HUMAN= SWITCH_trans* ( ( ld-10*TESTGEST**3 . 0 -5D-7*TESTGEST**2 . 0
+0. 0017*TESTGEST+1. 1937) /QC)
  GEST_QPLAF=DIM(QPLAF_HUMAN,0.0)  !  PLACENTA BLOOD FLOW RATE
  QPLAF =GEST_QPLAF

!  LIVER, VOLUME FRACTION   (HUMAN  0  TO 70 YEARS)
!  APPROACH BASED ON LUECKE  (2007)
 WLIO=  (3.59D-2 - ( 4 . 76D-7*WTOGR) + ( 8 . 50D-12*WTOGR**2 . 0 ) - ( 5 . 45D-17*WTOGR**3 . 0 ) )
!  LIVER VOLUME IN GROWING HUMAN

!  VARIABILITY OF REST OF THE  BODY  DEPENDS ON OTHER ORGAN
 WREO =  (0. 91- (WLIBO*WLIO+WFBO*WFO+  WPLABO*WPLAO + WLIO + WFO +
WPLAO) ) / (1+WREBO)
 QREF = 1- (QFF+QLIF+QPLAF)             ! REST BODY BLOOD FLOW  (ML/HR)
 QTTQF = QFF+QREF+QLIF+QPLAF           !  SUM MUST EQUAL 1

!  COMPARTMENT TISSUE BLOOD VOLUME  (L)  =========
 WF  =  WFO  * WTO                     !  ADIPOSE TISSUE
 WRE =  WREO * WTO                     !  REST OF THE BODY
 WLI =  WLIO * WTO                     !  LIVER
 WPLA=  WPLAO* WTO                     !  PLACENTA

!  COMPARTMENT TISSUE VOLUME  (L)  =========
 WFB  =  WFBO  * WF                   !  ADIPOSE TISSUE
 WREB =  WREBO * WRE                   !  REST OF THE BODY
 WLIB =  WLIBO * WLI                   !  LIVER
 WPLAB = WPLABO* WPLA                  !  PLACANTA

!  TOTAL VOLUME OF COMPARTMENT (L) ======
WFT =   WF                             !  TOTAL ADIPOSE TISSUE
WRET =  WRE                            !  TOTAL REST OF THE BODY
WLIT =  WLI                            !  TOTAL LIVER TISSUE
WPLAT=  WPLAB                          !  TOTAL PLACENTA TISSUE

                                      E-24

-------
!  CONSTANT USED  IN  CARDIAC OUTPUT EQUATION
!  UNIT CHANGED ON  JULY 14  2009  (L/HR)
QC= QCC* (WTO) **0.75
QF  = QFF*QC
QLI = QLIF*QC
QRE = QREF*QC
QPLA = QPLAF*QC
QTTQ = QF+QRE+QLI+QPLA
                                      !  ADIPOSE TISSUE BLOOD  FLOW RATE (L/HR)
                                      !  LIVER TISSUE BLOOD  FLOW RATE (L/HR)
                                      ! REST OF THE BODY BLOOD FLOW RATE (L/HR)
                                      ! PLACENTA TISSUE BLOOD  FLOW RATE (L/HR)
                            ! TOTAL FLOW RATE  (L/HR)
!  ========= DIFFUSIONAL  PERMEABILITY FACTORS FRACTION ORGAN  FLOW =========
PAF  = PAFF*QF
PARE = PAREF*QRE
(L/HR)
PALI = PALIF*QLI
PAPLA = PAPLAF*QPLA

! *******************
!  ABSORPTION SECTION
!  ORAL
!  INTRAPERITONEAL
!  SUBCUTANEOUS
!  INTRAVENOUS
                                        ADIPOSE TISSUE BLOOD  FLOW RATE (L/HR)
                                        REST OF THE BODY  BLOOD  FLOW RATE

                                        LIVER TISSUE BLOOD  FLOW RATE (L/HR)
                                        PLACENTA TISSUE BLOOD FLOW RATE (L/HR)
! BACKGROUND EXPOSURE
! EXPOSURE FOR  STEADY  STATE CONSIDERATION
! REPETITIVE EXPOSURE  SCENARIO
MSTOT_NMBCKGR = MSTOTBCKGR/322
MSTTBCKGR =MSTOTNMBCKGR *WTO
                                      IAMOUNT IN NMOL/G
DAY_EXPOSURE_BG    =  PULSE (DAY_LAG_BG, DAY_PERIOD_BG, DAY_FINISH_BG)
WEEK_EXPOSURE_BG   =  PULSE (WEEK_LAG_BG, WEEK_PERIOD_BG, WEEK_FINISH_BG)
MONTH_EXPOSURE_BG  =  PULSE (MONTH_LAG_BG,MONTH_PERIOD_BG,MONTH_FINISH_BG)

MSTTCH_BG =  (DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG) *MSTTBCKGR
MSTTFR_BG = MSTTBCKGR/CINT

CYCLE_BG =DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG

!  CONDITIONAL ORAL EXPOSURE (BACKGROUND EXPOSURE)

IF  (MSTTCH_BG.EQ. MSTTBCKGR)  THEN
    ABSMSTT_GB= MSTTFR_BG
ELSE
    ABSMSTT_GB =0.0
END IF

CYCLETOTBG=INTEG ( CYCLE_BG, 0.0)
IMULTIROUTE  EXPOSURE
! REPETITIVE EXPOSURE SCENARIO
                                      E-25

-------
MSTT= MSTOT_NM * WTO                   !AMOUNT IN NMOL
DAY_EXPOSURE   = PULSE(DAY_LAG,DAY_PERIOD,DAY_FINISH)
WEEK_EXPOSURE  = PULSE(WEEK_LAG,WEEK_PERIOD,WEEK_FINISH)
MONTH_EXPOSURE = PULSE(MONTH_LAG,MONTH_PERIOD,MONTH_FINISH)

MSTTCH =  (DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE)*MSTT

MSTTFR = MSTT/CINT

CYCLE = DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE

SUMEXPEVENT= INTEG  (CYCLE,0.0)  !NUMBER OF CYCLES GENERATED DURING  SIMULATION

!  CONDITIONAL ORAL  EXPOSURE
IF  (MSTTCH.EQ.MSTT) THEN
   ABSMSTT= MSTTFR
ELSE
   ABSMSTT =0.0
END IF


 CYCLETOT=INTEG(CYCLE, 0.0)

!  MASS CHANGE IN THE  LUMEN
 RMSTT= -(KST+KABS)*MST +ABSMSTT +ABSMSTT_GB !  RATE OF CHANGE  (NMOL/H)
  MST = INTEG(RMSTT,0.0)                       !AMOUNT REMAINING  IN DUODENUM
(NMOL)

!  ABSORPTION IN LYMPH CIRCULATION
 LYRMLUM = KABS*MST*A
  LYMLUM = INTEG(LYRMLUM,0.0)

!  ABSORPTION IN PORTAL CIRCULATION
 LIRMLUM = KABS*MST*B
  LIMLUM = INTEG(LIRMLUM,0.0)
     !IV ABSORPTION  SCENARIO	
 IV= DOSEIV_NM * WTO  IAMOUNT  IN NMOL
 IVR= IV/PFUNC  ! RATE  FOR IV  INFUSION IN BLOOD
 EXPIV= IVR *  (1-STEP(PFUNC))
 IVDOSE = integ(EXPIV,0.0)

    !IV LATE IN THE  CYCLE
    MODIFICATION JANUARY  13 2004
 IV_RlateR = DOSEIVNMlate*WTO
 IV_EXPOSURE=PULSE(IV_LAG,IV_PERIOD,IV_FINISH)

 IV_lateT = IV_EXPOSURE *IV_RlateR
 IV_late = IV_lateT/CINT

SUMEXPEVENTIV= integ(IV_EXPOSURE,0.0)  !NUMBER OF CYCLES GENERATED  DURING
SIMULATION

      !SYSTEMIC BLOOD  COMPARTMENT
      !  MODIFICATION OCT  8  2009
CB=(QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM+QPLA*CPLAB+IV_late)/(QC+CLURI)  !

                                      E-26

-------
 CA = CB
       CONCENTRATION  (NMOL/L)
      !CB=(QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM+QPLA*CPLAB+IV_late-RAURI)/QC
!(NMOL/L)

     !URINARY  EXCRETION BY KIDNEY
     ! MODIFICATION OCT 8 2009
RAURI =  CLURI  *CB
  AURI = INTEG(RAURI,0.0)

     !RAURI =  CLURI *  CRE
     IAURI = INTEG(RAURI,0.0)

     IUNIT CONVERSION  POST SIMULATION
CONSTANT MW=322  IMOLECULAR WEIGHT  (NG/NMOL)
CONSTANT SERBLO  =0.55
CONSTANT UNITCORR = 1.Oe3

 CBSNGKGLIADJ  =  CB*MW/(0.55*B_TOTLIP)  ING  SERUM LIPID ADJUSTED/KG
   AUCBS_NGKGLIADJ=integ(CBSNGKGLIADJ,0.)
CBNGKG=  CB*MW    ING/KG
     !ADIPOSE  COMPARMTENT
     !TISSUE BLOOD SUBCOMPARTMENT
RAFB= QF*(CA-CFB)-PAF*(CFB-CF/PF)
 AFB = INTEG(RAFB,0.0)
 CFB = AFB/WFB
     !TISSUE SUBCOMPARTMENT
RAF = PAF*(CFB-CF/PF)
 AF = INTEG(RAF,0.0)
 CF  = AF/WF
!(NMOL/H)
 !(NMOL)
!(NMOL/L)

!(NMOL/H)
!(NMOL)
!(NMOL/L)
     IUNIT CONVERSION POST SIMULATION
CFTOTAL=  (AF  + AFB)/(WF + WFB) !  TOTAL CONCENTRATION IN NMOL/ML
CFNGKG=CFTOTAL*MW !  FAT CONCENTRATION IN NG/KG
AUCF_NGKGH=integ(CFNGKG, 0 . )
     IREST OF  THE  BODY COMPARTMENT
     ! TISSUE BLOOD SUBCOMPARTMENT
RAREB= QRE *(CA-CREB)-PARE*(CREB-CRE/PRE)
 AREB = INTEG(RAREB,0.0)
 CREB = AREB/WREB
     !TISSUE SUBCOMPARTMENT
RARE = PARE*(CREB -  CRE/PRE)
 ARE = INTEG(RARE,0.0)
 CRE  = ARE/WRE
ARETOT = ARE  +AREB
          !(NMOL/H)
           !(NMOL)
          !(NMOL/L)

          !(NMOL/H)
          !(NMOL)
            (NMOL/L)
     !POST  SIMULATION UNIT CONVERSION
CRETOTAL=  (ARE  +  AREB)/(WRE + WREB)
CRENGKG=CRETOTAL*MW
CONCENTRATION  (NG/KG)
          !  TOTAL CONCENTRATION  (NMOL/L)
          !  REST OF THE BODY
     !LIVER  COMPARTMENT
                                      E-27

-------
     !TISSUE BLOOD  SUBCOMPARTMENT
 RALIB = QLI*(CA-CLIB)-PALI*(CLIB-CFLLIR)+LIRMLUM  !  (NMOL/HR)
  ALIB = INTEG(RALIB,0.0)                         !(NMOL)
 CLIB = ALIB/WLIB                                !(NMOL/L)
     !TISSUE SUBCOMPARMTENT
 RALI = PALI*(CLIB -  CFLLIR)-REXCLI             !  (NMOL/HR)
  ALI = INTEG(RALI,0.0)                               !(NMOL)
 CLI  = ALI/WLI                                  !(NMOL/L)

     IFREE TCDD CONCENTRATION IN LIVER
      !  MODIFICATION  OCTOBER 8 2009
 CFLLI= IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLIR))  &
        +((CYP1A2_103*CFLLIR/(KDLI2+CFLLIR)*IND_ACTIVE)))-CFLLI,CFLLIO)
    CFLLIR=DIM(CFLLI,0.0)  !  FREE TCDD CONCENTRATION  IN  LIVER
!MODIFIED FROM:
!PARAMETER  (LIVER_1RMN  = l.OE-30)
!  CFLLI= IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLIR &
!+LIVER_lRMN))+((CYP1A2_103*CFLLIR/(KDLI2 + CFLLIR &
!+LIVER_lRMN)*IND_ACTIVE)))-CFLLI,CFLLIO)
!CFLLIR=DIM(CFLLI,0.0)

!  MODIFICATION OCTOBER  8 2009
CBNDLI= LIBMAX*CFLLIR/(KDLI+CFLLIR) !BOUND CONCENTRATION  (NMOL/L)

     IPOST SIMULATION  UNIT CONVERSION
CLITOTAL=  (ALI + ALIB)/(WLI  +  WLIB) !  TOTAL CONCENTRATION (NMOL/L)
Rec_occ= CFLLIR/(KDLI+CFLLIR)
CLINGKG=CLITOTAL*MW  ! LIVER CONCENTRATION IN NG/KG
  AUCLI_NGKGH=integ(CLINGKG,0.0)
CBNDLINGKG =  CBNDLI*MW  !  BOUND CONCENTRATION IN NG/KG
  AUCBNDLI_NGKGH =INTEG(CBNDLINGKG,0.0)

     !FRACTION INCREASE  OF INDUCTION OF CYP1A2
fold_ind=CYPlA2_10UT/CYP!A2_lA2
VARIATIONOFAC =(CYP1A2_10UT-CYP1A2_1A2)/CYP1A2_1A2

!VARIABLE ELIMINATION BASED ON THE CYP1A2
!  MODIFICATION OCTOBER  8 2009
KBILE_LI_T =  Kelv*VARIATIONOFAC!  !  DOSE-DEPENDENT EXCRETION RATE CONSTANT

 REXCLI = KBILE_LI_T*CFLLIR*WLI !  DOSE-DEPENDENT BILLIARY EXCRETION RATE
    EXCLI = INTEG(REXCLI,0.0)

IKBILE LI T =((CYP1A2 10UT-CYP1A2 1A2)/CYP1A2  lA2)*Kelv !
!CHEMICAL IN CYP450  (1A2)  COMPARTMENT

CYP1A2_1KINP =  CYP1A2_1KOUT*  CYP1A2_10UTZ  ! BASAL PRODCUTION  RATE OF CYP1A2
SET EQUAL TO BASAL DEGREDATION RATE

    !  MODIFICATION OCTOBER 8  2009
CYP1A2_10UT =INTEG(CYP1A2_1KINP * (1.0 + CYP1A2_1EMAX  *(CBNDLI+1.Oe-30)**HILL
&
     /(CYP1A2_1EC50**HILL  + (CBNDLI+1.Oe-30)**HILL)) &
      - CYP1A2_1KOUT*CYP1A2_10UT, CYP1A2_10UTZ)
!MODIFIED FROM:

                                      E-28

-------
!PARAMETER  (CYP1A2_1RMN  =  1E-30)
!CYP1A2_10UT =INTEG(CYP1A2_1KINP  *  (1 + CYP1A2_1EMAX *(CBND&
!LI +CYP1A2_1RMN)**HILL/(CYP1A2_1EC50 + (CBNDLI + CYP1A2_1&
!RMN)**HILL) +CYP1A2_1RMN)  -  CYP1A2_1KOUT*CYP1A2_1&
!OUT, CYP1A2_10UTZ)

!  EQUATIONS INCORPORATING  DELAY OF  CYP1A2  PRODUCTION (NOT USED  IN
SIMULATIONS)
CYP1A2_1R02 =  (CYP1A2_10UT -  CYP1A2_102)/  CYP1A2_1TAU
  CYP1A2_102 =INTEG(CYP1A2_1R02,  CYP1A2_1A1)

CYP1A2_1R03 =  (CYP1A2_102  - CYP1A2_103)/ CYP1A2_1TAU
  CYP1A2_103 =INTEG(CYP1A2_1R03,  CYP1A2_1A2)

     !PLACENTA COMPARTMENT
     !TISSUE BLOOD  SUBCOMPARTMENT
RAPLAB= QPLA*(CA - CPLAB)-PAPLA* (CPLAB -CFLPLAR)    ! NMOL/HR)
 APLAB = INTEG(RAPLAB,0.0)                            !  (NMOL)
 CPLAB = APLAB/(WPLAB+1E-30)                         ! (NMOL/ML)
     !TISSUE SUBCOMPARTMENT
RAPLA = PAPLA*(CPLAB-CFLPLAR)-RAMPF + RAFPM         ! (NMOL/HR)
 APLA = INTEG(RAPLA,0.0)                              !  (NMOL)
 CPLA  = APLA/(WPLA+le-30)                           ! (NMOL/ML)

     ! NEW EQUATION AUGUST  28   2009
PARAMETER  (PARA_ZERO = l.OE-30)
CFLPLA= IMPLC(CPLA-(CFLPLAR*PPLA  +(PLABMAX*CFLPLAR/(KDPLA&
    +CFLPLAR+PARA_ZERO)))-CFLPLA,CFLPLAO)
CFLPLAR=DIM(CFLPLA,0.0)

     !POST SIMULATION UNIT  CONVERSION
CPLATOTAL = ((APLAB+APLA)/(WPLAB+WPLA))

     !FETUS COMPARTMENT
RAFETUS= RAMPF-RAFPM
 AFETUS=INTEG(RAFETUS,0.0)
CFETUS=AFETUS/(WTFE+1.Oe-30)
CFETOTAL= CFETUS
CFETUS_v = CFETUS/PFETUS

     !POST SIMULATION UNIT  CONVERSION
 CFETUSNGKG = CFETUS*MW                     !(NG/KG)
    !TRANSFER OF DIOXIN  FROM PLACENTA TO FETUS
    !FETAL EXPOSURE ONLY DURING EXPOSURE

IF (T.LT.TRANSTIME_ON) THEN
 SWITCH_trans =0.0
ELSE
 SWITCH_trans = 1
END IF

    !TRANSFER OF DIOXIN  FROM PLACENTA TO FETUS
    !  MODIFICATION 26  SEPTEMBER 2003

RAMPF =  (CLPLA_FET*CPLA)*SWITCH_trans

                                      E-29

-------
  AMPF=INTEG(RAMPF,0.0)

     !TRANSFER OF  DIOXIN  FROM FETUS TO PLACENTA
RAFPM =  (CLPLA_FET*CFETUS_v)*SWITCH_trans!
 AFPM =  INTEG(RAFPM,0.0)

     !CHECK MASS BALANCE  	
BDOSE= IVDOSE +LYMLUM+LIMLUM
BMASSE = EXCLI+AURI+AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB+AFETUS  !
BDIFF =  BDOSE-BMASSE

     IBODY BURDEN  (NMOL)
BODY_BURDEN = AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB

     IBODY BURDEN  CONCENTRATION (NG/KG)
 BBNGKG  =(AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB)*MW/WTO

!  END SIMULATION  COMMAND

TERMT (T.GE. TimeLimit,  'Time limit has been reached.')

END   !   END OF THE  DERIVATIVE SECTION
END   !   END OF THE  DYNAMIC SECTION
END   !   END OF THE  PROGRAM
E.2.2.2. Input File

output @clear
prepare gclear T   year  CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG   CBNDLINGKG CBNGKG
CINT = 1
   %EXPOSURE SCENARIO
EXP_TIME_ON        =  0
EXP_TIME_OFF       =  401190
DAYJCYCLE          =  24
BCK_TIME_ON        =  401190
BCK_TIME_OFF       =  401190
IV_LAG             =  401190
IV_PERIOD          =  401190
    %GESTATION CONTROL
CONCEPTION_T
TIMELIMIT
TRANSTIME_ON
GESTATION
    %EXPOSURE DOSE
MSTOT
MSTOTBCKGR
DOSEIV
DOSEIVLATE
393120
399840
394632
9.977E-07
0.
0.
0.
           %TIME EXPOSURE  BEGINS  (HOUR)
           %TIME EXPOSURE  ENDS  (HOUR)
           %HOURS BETWEEN  DOSES  (HOUR)
           %TIME BACKGROUND  EXPOSURE BEGINS (HOUR)
           %TIME BACKGROUND  EXPOSURE ENDS (HOUR)
%TIME OF CONCEPTION AT  45  YEARS  OLD
%SIMULATION DURATION  (HOUR)
%TRANSFER FROM MOTHER TO FETUS AT 1512 HOURS
%NG OF TCDD PER KG OF  BW
%ORAL BACKGROUND EXPOSURE  DOSE  (NG/KG)
     % TRANFER MOTHER  TO  FETUS CLEARANCE
CLPLA FET          =  0.001      %MOTHER TO FETUS TRANFER CLEARANCE  (L/HR)
                                      E-30

-------
E.2.3.  Rat Standard Model
E.2.3.1. Model Code
PROGRAM: Three Compartment PBPK Model in Rat: Standard Model (Nongestation)'
INITIAL
           INITIALIZATION  OF PARAMETERS
      !SIMULATION  PARAMETERS
CONSTANT PARA_ZERO        =       Id-30
CONSTANT EXP_TIME_ON      =       0.0
(HOURS)
CONSTANT EXP_TIME_OFF     =      900.0
(HOURS)
CONSTANT DAYJCYCLE        =      900.0
DOSES  (HOURS)
CONSTANT BCK_TIME_ON      =       0.0
EXPOSURE BEGINS  (HOURS)
CONSTANT BCK_TIME_OFF     =       0.0
EXPOSURE ENDS  (HOURS)

CONSTANT MW=322  IMOLECULAR WEIGHT (NG/NMOL)
CONSTANT SERBLO =0.55
CONSTANT UNITCORR = 1000
                  ! TIME AT  WHICH EXPOSURE BEGINS

                   ! TIME AT WHICH EXPOSURE ENDS

                    ! NUMBER OF HOURS BETWEEN

                  ! TIME AT  WHICH BACKGROUND

                  ! TIME AT  WHICH BACKGROUND
      !EXPOSURE DOSES
CONSTANT MSTOTBCKGR
(UG/KG)
CONSTANT MSTOT
CONSTANT MSTOTsc
(UG/KG)
CONSTANT DOSEIV

      IORAL DOSE
  MSTOT_NM
  MSTOT_NMBCKGR

      !INTRAVENOUS  DOSE
  DOSEIV NM
   0.0

   10
   0.0
     IORAL BACKGROUND EXPOSURE DOSE

IORAL EXPOSURE DOSE  (UG/KG)
     !SUBCUTANEOUS EXPOSURE DOSE

     !  INJECTED DOSE (UG/KG)
MSTOT/MW        IAMOUNT  IN  NMOL/G
MSTOTBCKGR/MW   IAMOUNT  IN  NMOL/G
DOSEIV/MW
   !AMOUNT IN NMOL/G
      !INITIAL GUESS  OF  THE  FREE CONCENTRATION  IN THE LIGAND  (COMPARTMENT
INDICATED BELOW)====
CONSTANT CFLLIO           =      0.0            !LIVER  (NMOL/ML)

      !BINDING CAPACITY  (AhR)  FOR NON LINEAR BINDING  (COMPARTMENT  INDICATED
BELOW)  (NMOL/ML) ===
CONSTANT LIBMAX           =     3.5e-4          !  LIVER  (NMOL/ML),  WANG  ET AL.
1997

      ! PROTEIN AFFINITY CONSTANTS (1A2 OR AhR, COMPARTMENT INDICATED BELOW)
(NMOL/ML)===
CONSTANT KDLI             =     l.Oe-4          !  LIVER  (AhR)  (NMOL/ML), WANG
ET AL. 1997
                                      E-31

-------
CONSTANT KDLI2            =    4.Oe-2           !LIVER  (1A2)  (NMOL/ML),  EMOND
ET AL. 2004

      !EXCRETION AND ABSORPTION CONSTANT [RAT]
CONSTANT KST              =    0.36             ! GASTRIC RATE  CONSTANT  (HR-1),
WANG ET AL.  (1997)
CONSTANT KABS             =    0.48             !INTESTINAL ABSORPTION CONSTANT
(HR-1), WANG ET AL. 1997

      !URINARY ELIMINATION CLEARANCE (ML/HR)
CONSTANT CLURI            =     0.01            IURINARY CLEARANCE  (ML/HR),
EMOND ET AL. 2004

      IINTERSPECIES VARIABLE ELIMINATION
CONSTANT KELV             =     0.15            ! INTERSPECIES  VARIABLE
ELIMINATION CONSTANT  (I/HOUR)  (OPTIMIZED), EMOND ET AL. 2004

      ! CONSTANT TO DIVIDE THE ABSORPTION INTO LYMPHATIC AND PORTAL  FRACTIONS
CONSTANT A                =0.7             ! LYMPHATIC FRACTION,  WANG ET
AL. 1997

      !PARTITION COEFFICIENTS
CONSTANT PF               =     100             ! ADIPOSE TISSUE/BLOOD,  WANG ET
AL. 1997
CONSTANT PRE              =1.5             ! REST OF THE BODY/BLOOD,  WANG
ET AL. 1997
CONSTANT PLI              =6.0              !  LIVER/BLOOD, WANG ET AL.
1997

      !PARAMETER FOR INDUCTION OF CYP 1A2  [MOUSE]  ===
CONSTANT IND_ACTIVE        =     1.0            !  INCLUDE  INDUCTION? (1 = YES,
0 = NO)
CONSTANT CYP1A2_10UTZ     =1.6             ! DEGRADATION CONCENTRATION
CONSTANT OF 1A2  (NMOL/ML),  WANG ET AL.  1997
CONSTANT CYP1A2_1A1       =1.6             ! BASAL CONCENTRATION OF 1A1
(NMOL/ML), WANG ET AL.  1997
CONSTANT CYP1A2_1EC50     =     0.13            ! DISSOCIATION  CONSTANT  TCDD-
CYP1A2 (NMOL/ML)  , WANG ET AL.  1997
CONSTANT CYP1A2_1A2       =     1.6             ! BASAL CONCENTRATION OF 1A2
(NMOL/ML) Wang et al  (1997)
CONSTANT CYP1A2_1KOUT     =0.1             ! FIRST ORDER RATE  OF
DEGRADATION  (H-l), WANG ET AL.  1997
CONSTANT CYP1A2_1TAU      =     0.25            ! HOLDING TIME  (H), WANG ET AL.
1997
CONSTANT CYP1A2_1EMAX     =     600             ! MAXIMUM INDUCTION OVER BASAL
EFFECT (UNITLESS), WANG ET AL.  1997
CONSTANT HILL             =     0.6      IHILL CONSTANT; COOPERATIVE LIGAND
BINDING EFFECT CONSTANT (UNITLESS)

    !TISSUE BLOOD  FLOW  EXPRESSED AS A FRACTION OF CARDIAC  OUTPUT
CONSTANT QFF  = 0.069                           ! ADIPOSE TISSUE BLOOD FLOW
FRACTION  (UNITLESS), WANG ET AL.  1997
CONSTANT QLIF = 0.183                           ! LIVER (UNITLESS), WANG ET AL.
1997

      IDIFFUSIONAL PERMEABILITY FRACTION


                                      E-32

-------
CONSTANT PAFF             = 0.0910             !  ADIPOSE  (UNITLESS),  WANG ET
AL. 1997
CONSTANT PAREF            = 0.0298             !  REST OF THE BODY  (UNITLESS),
WANG ET AL.  1997
CONSTANT PALIF            = 0.35               !  LIVER  (UNITLESS),  WANG ET AL.
1997

     !FRACTION OF  TISSUE  VOLUME (UNITLESS)
CONSTANT WLIO             =  0.0360            !  LIVER, WANG ET AL.  1997
CONSTANT WFO              =  0.069             !  BLOOD, WANG ET AL.  1997

     !COMPARTMENT  TISSUE  BLOOD EXPRESSED AS A FRACTION OF THE TOTAL
COMPARTMENT VOLUME =========
CONSTANT WFBO             =  0.050             !  ADIPOSE TISSUE, WANG ET AL.
1997
CONSTANT WREBO            =  0.030             !  REST OF THE BODY,  WANG ET AL.
1997
CONSTANT WLIBO            =  0.266             !  LIVER  , WANG ET AL.  1997

     !EXPOSURE SCENARIO FOR UNIQUE OR REPETITIVE WEEKLY OR MONTHLY EXPOSURE
     !  NUMBER OF EXPOSURES PER WEEK
CONSTANT WEEK_LAG       =0.0                 !  TIME ELAPSED BEFORE EXPOSURE
BEGINS  (WEEK)
CONSTANT WEEK_PERIOD      = 168.0                !  NUMBER OF HOURS  IN THE WEEK
(HOURS)
CONSTANT WEEK_FINISH      = 168.0                !  TIME EXPOSURE ENDS (HOURS)

     !NUMBER OF EXPOSURES PER MONTH
CONSTANT MONTH_LAG     =0.0                 !  TIME ELAPSED BEFORE EXPOSURE
BEGINS  (MONTH)

     !SET FOR BACKGROUND  EXPOSURE===========
     !CONSTANT FOR BACKGROUND EXPOSURE===========
CONSTANT Day_LAG_BG    =0.0                 !  TIME ELAPSED BEFORE EXPOSURE
BEGINS  (HOURS)
CONSTANT Day_PERIOD_BG    =24.0                 !  LENGTH OF EXPOSURE (HOURS)

     !NUMBER OF EXPOSURES PER WEEK
CONSTANT WEEK_LAG_BG    =0.0                !  DELAY BEFORE BACKGROUND
EXPOSURE  (WEEK)
CONSTANT WEEK_PERIOD_BG   = 168.0                !NUMBER OF HOURS  IN THE WEEK
(HOURS)
CONSTANT WEEK_FINISH_BG   = 168.0                !  TIME EXPOSURE ENDS (HOURS)

     !GROWTH CONSTANT FOR RAT
     !CONSTANT FOR MOTHER BODY WEIGHT GROWTH ======
CONSTANT BW_TO =250.0                           !(IN G) CHANGED FOR
SIMULATION

     !  CONSTANT USED IN CARDIAC OUTPUT EQUATION
CONSTANT QCCAR =311.4                          !CONSTANT  (ML/MIN/KG),  WANG ET
AL.

     !  COMPARTMENT TOTAL  LIPID FRACTION
CONSTANT F_TOTLIP         = 0.855              IADIPOSE TISSUE  (UNITLESS)
CONSTANT B_TOTLIP         = 0.0033             !BLOOD  (UNITLESS)
CONSTANT RE_TOTLIP       = 0.019              !REST OF THE BODY  (UNITLESS)

                                      E-33

-------
CONSTANT LI_TOTLIP        =0.06

END      !END OF THE  INITIAL SECTION
                                    !LIVER (UNITLESS)
DYNAMIC   !DYNAMIC  SIMULATION SECTION
ALGORITHM
CINTERVAL
MAXTERVAL
MINTERVAL
VARIABLE
CONSTANT
(HOURS)
 CINTXY  = CINT
 PFUNC   = CINT
IALG
CINT
MAXT
MINT
T
TIMELIMIT
   2
  0.1
l.Oe+10
l.OE-10
  0.0
  900.0
  GEAR METHOD
  COMMUNICATION INTERVAL
  MAXIMUM CALCULATION INTERVAL
!  MINIMUM CALCULATION INTERVAL

  !SIMULATION TIME LIMIT
          ITIME CONVERSION
  DAY=T/24.0
  WEEK =T/168.0
  MONTH =T/730.0
  YEAR=T/8760.0
                                      !  TIME IN DAYS
                                      !  TIME IN WEEKS
                                      !  TIME IN MONTHS
                                      !  TIME IN YEARS
DERIVATIVE  ! PORTION  OF  CODE THAT SOLVES DIFFERENTIAL EQUATIONS

         !CHRONIC OR SUBCHRONIC EXPOSURE SCENARIO =======
         !NUMBER OF EXPOSURES PER DAY
 DAY_LAG     = EXP_TIME_ON
BEGINS  (HOURS)
 DAY_PERIOD   = DAY_CYCLE
 DAY_FINISH   = CINTXY
 MONTH_PERIOD = TIMELIMIT
 MONTH  FINISH = EXP TIME OFF
                                  !  TIME ELAPSED BEFORE EXPOSURE

                                   !  EXPOSURE PERIOD  (HOURS)
                                   !  LENGTH OF EXPOSURE  (HOURS)
                                   !  EXPOSURE PERIOD  (MONTHS)
                                   !  LENGTH OF EXPOSURE  (MONTHS)
         !NUMBER OF  EXPOSURES  PER DAY AND MONTH
 DAY_FINISH_BG   =  CINTXY
 MONTH_LAG_BG   = BCK_TIME_ON
EXPOSURE BEGINS  (MONTHS)
 MONTH_PERIOD_BG =  TIMELIMIT
(MONTHS)
 MONTH_FINISH_BG =  BCK_TIME_OFF
(MONTHS)
                                   !  LENGTH OF EXPOSURE  (HOURS)
                                  !  TIME ELAPSED BEFORE BACKGROUND

                                   !  BACKGROUND EXPOSURE PERIOD

                                   !  LENGTH OF BACKGROUND EXPOSURE
  B = 1-A
THE PORTAL FRACTION  OF  THE LIVER
                                  !  FRACTION OF DIOXIN ABSORBED  IN
         ! BODY WEIGHT  GROWTH EQUATION========
 PARAMETER  (BW_RMN  =  l.OE-30)
 WTO=  (BW_TO *(1.0+(0.41*T)/(1402.5+T+BW_RMN)))  ! IN GRAMS

         !VARIABILITY  OF  REST OF THE BODY DEPEND OTHERS ORGAN
 WREO =  (0.91 -  (WLIBO*WLIO  + WFBO*WFO + WLIO + WFO))/(1.0+WREBO)  !REST  OF
THE BODY FRACTION;  UPDATED  FOR EPA ASSESSMENT
 QREF = 1.0-(QFF+QLIF)                          !REST OF BODY BLOOD  FLOW
 QTTQF = QFF+QREF+QLIF                       ! SUM MUST EQUAL  1
                                      E-34

-------
         !COMPARTMENT VOLUME  (G OR ML)  =========
 WF  =  WFO  * WTO                           !  ADIPOSE
 WRE =  WREO * WTO                           !  REST OF THE BODY
 WLI =  WLIO * WTO                           !  LIVER

         !COMPARTMENT TISSUE  BLOOD VOLUME (G OR ML) =========
 WFB  =  WFBO  * WF                          !  ADIPOSE
 WREB =  WREBO * WRE                         !  REST OF THE BODY
 WLIB =  WLIBO * WLI                         !  LIVER

         !CARDIAC OUTPUT  FOR  THE GIVEN BODY WEIGHT
  QC= QCCAR*60.0*(WTO/UNITCORR)**0.75

         ! COMPARTMENT  BLOOD  FLOW (ML/HR)
  QF  = QFF*QC                               !  ADIPOSE TISSUE BLOOD  FLOW RATE
  QLI = QLIF*QC                              !  LIVER TISSUE BLOOD  FLOW RATE
  QRE = QREF*QC                              !  REST OF THE BODY  BLOOD  FLOW
RATE
  QTTQ = QF+QRE+QLI                     !  TOTAL FLOW RATE

         !PERMEABILITY  ORGAN  FLOW (ML/HR)
PAF  = PAFF*QF                               !  ADIPOSE
PARE = PAREF*QRE                             !  REST OF THE BODY
PALI = PALIF*QLI                             !  LIVER TISSUE

         !CONDITIONAL ORAL  EXPOSURE (BACKGROUND EXPOSURE)
         !EXPOSURE +  !REPETITIVE EXPOSURE SCENARIO
  IV= DOSEIV_NM * WTO   IAMOUNT IN NMOL
  MSTT= MSTOT_NM * WTO  !AMOUNT IN NMOL
  MSTTBCKGR =MSTOT_NMBCKGR *WTO

         !REPETITIVE ORAL BACKGROUND EXPOSURE SCENARIOS
  DAY_EXPOSURE_BG   =  PULSE(DAY_LAG_BG,DAY_PERIOD_BG,DAY_FINISH_BG)
  WEEK_EXPOSURE_BG  =  PULSE(WEEK_LAG_BG,WEEK_PERIOD_BG,WEEK_FINISH_BG)
  MONTH_EXPOSURE_BG =  PULSE(MONTH_LAG_BG,MONTH_PERIOD_BG,MONTH_FINISH_BG)

  MSTTCH_BG =  (DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG)*MSTTBCKGR
  MSTTFR_BG = MSTTBCKGR/CINT

  CYCLE_BG =DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG

IF  (MSTTCH_BG.EQ.MSTTBCKGR)  THEN
    ABSMSTT_GB= MSTTFR_BG
ELSE
    ABSMSTT_GB =0.0
END IF
         !REPETITIVE ORAL MAIN EXPOSURE SCENARIO
  DAY_EXPOSURE   = PULSE(DAY_LAG,DAY_PERIOD,DAY_FINISH)
  WEEK_EXPOSURE  = PULSE(WEEK_LAG,WEEK_PERIOD,WEEK_FINISH)
  MONTH_EXPOSURE = PULSE(MONTH_LAG,MONTH_PERIOD,MONTH_FINISH)

  MSTTCH =  (DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE)*MSTT
  CYCLE = DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE
  MSTTFR = MSTT/CINT
                                      E-35

-------
  SUMEXPEVENT= integ  (CYCLE,0.0)  !NUMBER OF CYCLES GENERATED  DURING
SIMULATION
         !CONDITIONAL  ORAL  EXPOSURE
IF (MSTTCH.EQ.MSTT) THEN
   ABSMSTT= MSTTFR
ELSE
   ABSMSTT =0.0
END IF

 CYCLETOT=INTEG(CYCLE, 0.0)

         IMASS CHANGE  IN THE LUMEN
 RMSTT  = -(KST+KABS)*MST+ABSMSTT +ABSMSTT_GB  ! RATE OF CHANGE  (NMOL/H)
   MST = INTEG(RMSTT,0.0)   IAMOUNT REMAINING IN DUODENUM  (NMOL)

         IABSORPTION IN LYMPH CIRCULATION
 LYRMLUM = KABS*MST*A
   LYMLUM = INTEG(LYRMLUM,0.0)

         IABSORPTION IN PORTAL CIRCULATION
 LIRMLUM = KABS*MST*B
   LIMLUM = INTEG(LIRMLUM,0.0)

         !PERCENT OF DOSE REMAINING IN THE GI TRACT
         IABSORPTION  of  Dioxin by IV route	
 IVR= IV/PFUNC  ! RATE FOR IV INFUSION IN BLOOD
 EXPIV= IVR *  (1.0-STEP(PFUNC))
   IVDOSE = integ(EXPIV,0.0)

         !SYSTEMIC BLOOD COMPARTMENT
         ! MODIFICATION  ON OCTOBER 6,  2009
CB=(QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM)/(QC+CLURI)  !
   CA = CB

         !URINARY EXCRETION BY KIDNEY
         ! MODIFICATION  ON OCTOBER 6,  2009
RAURI = CLURI *CB
  AURI = INTEG(RAURI,0.0)

         !CONVERSION  EQUATION POST SIMULATION

 CBNGKG = CB*MW*UNITCORR ![NG/KG]
CBSNGKGLIADJ=  (CB*MW*UNITCORR*(1.0/B_TOTLIP)*(1.0/SERBLO))![NG  of  TCDD
Serum/Kg OF LIPID]

        IADIPOSE TISSUE  COMPARTMENT
        !TISSUE BLOOD  SUBCOMPARTMENT
 RAFB = QF*(CA-CFB)-PAF*(CFB-CF/PF)                  !(NMOL/HR)
   AFB = INTEG(RAFB,0.0)                              !(NMOL)
   CFB = AFB/WFB                                     !(NMOL/ML)
        !TISSUE SUBCOMPARTMENT

                                      E-36

-------
 RAF  = PAF*(CFB-CF/PF)
   AF  = INTEG(RAF,0.0)
   CF  = AF/WF
       !(NMOL/HR)
        !(NMOL)
       !(NMOL/ML)
       !CONVERSION  EQUATION POST SIMULATION
   CFTOTAL    =  (AF + AFB)/(WF + WFB)        !TOTAL  CONCENTRATION IN NMOL/ML
   CFNGKG
            =  CFTOTAL*MW*UNITCORR
       IREST OF  THE  BODY COMPARTMENT
       ! TISSUE  BLOOD SUBCOMPARTMENT
 RAREB= QRE*(CA-CREB)-PARE*(CREB-CRE/PRE)
   AREB = INTEG(RAREB,0.0)
   CREB = AREB/WREB
       ! TISSUE  COMPARTMENT
 RARE = PARE*(CREB  - CRE/PRE)
   ARE = INTEG(RARE,0.0)
   CRE  = ARE/WRE
!  CONCENTRATION  [NG/KG]
        !(NMOL/HR)
         !(NMOL)
        !(NMOL/ML)

        !(NMOL/HR)
            !(NMOL)
       !(NMOL/ML)
   !CONVERSION  EQUATION POST SIMULATION
   CRETOTAL=  (ARE  +  AREB)/(WRE + WREB)
NMOL/ML

   CTREPGG= CRETOTAL*MW*UNITCORR !(PG/ML)
    AUC_REPGG = integ(CTREPGG,0.0)

   !LIVER COMPARTMENT
   !TISSUE BLOOD COMPARTMENT
 RALIB = QLI*(CA-CLIB)-PALI*(CLIB-CFLLIR)+LIRMLUM
   ALIB = INTeg(RALIB,0.0)
   CLIB = ALIB/WLIB
   !TISSUE COMPARTMENT
 RALI =  PALI*(CLIB-CFLLIR)-REXCLI
   ALI = integ(RALI,0.0)
   CLI  = ALI/WLI
        !  TOTAL CONCENTRATION  IN
        !(NMOL/HR)
         !(NMOL)
        !(NMOL/HR)
            !(NMOL)
        !(NMOL/ML)
PARAMETER  (LIVER_1RMN = l.OE-30)
CFLLI= IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLIR &
+LIVER_1RMN))+((CYP1A2_103*CFLLIR/(KDLI2+CFLLIR  &
+LIVER_1RMN)*IND_ACTIVE)))-CFLLIR,CFLLIO)  !  FREE TCDD CONCENTRATION IN LIVER
CFLLIR=DIM(CFLLI,0.0)

  CBNDLI=  LIBMAX*CFLLIR/(KDLI+CFLLIR+LIVER  1RMN)  !BOUND CONCENTRATION
       !CONVERSION  EQUATION POST SIMULATION
  CLITOTAL=  (ALI + ALIB)/(WLI + WLIB)
NMOL/ML

  rec_occ_AHR=  (CFLLIR/(KDLI+CFLLIR+1))*100.0
OCCUPANCY
  PROT_occ_lA2=  (CFLLIR/(KDLI2+CFLLIR))*100.0
OCCUPANCY
  CLINGKG =(CLITOTAL*MW*UNITCORR)
  CBNDLINGKG = CBNDLI*MW*UNITCORR
    AUCLI_NGKGH=INTEG(CLINGKG,0.0)
  CLINGG=CLITOTAL*MW
         !  TOTAL CONCENTRATION  IN


           !  PERCENT OF AhR

           !  PERCENT OF 1A2
                                      E-37

-------
        IVARIABLE ELIMINATION HALF-LIFE BASED ON THE CONCENTRATION  OF  CYP1A2
    KBILE_LI_T =((CYP1A2_10UT-CYP1A2_1A2)/CYP1A2_1A2)*Kelv  ! INDUCED  BILIARY
EXCRETION RATE CONSTANT

 REXCLI=  (KBILE_LI_T*CFLLIR*WLI)  !  DOSE-DEPENDENT BILIARY EXCRETION RATE
   EXCLI = INTEG(REXCLI,0.0)

      !CHEMICAL IN CYP450  (1A2)  COMPARTMENT
  i===PARAMETER FOR  INDUCTION OF CYP1A2

 CYP1A2_1KINP = CYP1A2_1KOUT* CYP1A2_10UTZ !  BASAL RATE OF  CYP1A2  PRODUCTION
SET EQUAL TO BASAL RATE OF  DEGREDATION
      ! MODIFICATION ON  OCTOBER 6,  2009
 CYP1A2_10UT =INTEG(CYP1A2_1KINP * (1.0 + CYP1A2_1EMAX *(CBNDLI+1.Oe-
30)**HILL &
     /(CYP1A2_1EC50**HILL  +  (CBNDLI+1.Oe-30)**HILL))  &-
      - CYP1A2_1KOUT*CYP1A2_10UT,  CYP1A2_10UTZ)

!  EQUATIONS INCORPORATING  DELAY OF CYP1A2 PRODUCTION (NOT USED  IN
SIMULATIONS)

 CYP1A2_1R02 =  (CYP1A2_10UT  -  CYP1A2_102)/ CYP1A2_1TAU
    CYP1A2_102 =INTEG(CYP1A2_1R02, CYP1A2_1A1)
 CYP1A2_1R03 =  (CYP1A2_102 - CYP1A2_103)/ CYP1A2_1TAU
    CYP1A2_103 =INTEG(CYP1A2_1R03, CYP1A2_1A2)

i  	CHECK MASS BALANCE	
  BDOSE= LYMLUM+LIMLUM+IVDOSE
  BMASSE = EXCLI+AURI+AFB+AF+AREB+ARE+ALIB+ALI
      BDIFF = BDOSE-BMASSE

i	BODY  BURDEN	
  BBNGKG =(((AFB+AF+AREB+ARE+ALIB+ALI)*MW)/(WTO/UNITCORR))   !
i  	END OF THE  SIMULATION COMMAND	

TERMT  (T.GE. TimeLimit,  'Time  limit has been reached.')

END    ! END OF THE DERIVATIVE  SECTION
END    ! END OF THE DYNAMIC SIMULATION SECTION
END    ! END OF THE PROGRAM.
E.2.3.2. Input Files

E.2.3.2.1.  Cantoni et al (1981)
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

%Cantoni et al.  1981
%protocol:  oral exposure 1  dose/week for 45 weeks; female CD-COBS  rats
%dose levels:  0.01,  0.1,  1  ug/kg 1 dose/week for 45 weeks
%dose levels:  10,  100,  1000  ng/kg 1 dose/week for 45 weeks

                                      E-38

-------
%dose levels equivalent  to:  1.43,  14.3 143 ng/kg 7 days/week  for  45  weeks
MAXT
CINT
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
BW_TO
(G)
0.01
0.1
0.
7560
168
0.
0.
7560
125
%EXPOSURE DOSE  SCENARIOS
  %MSTOT           =0.01
   %MSTOT        =0.1
   MSTOT        =  1
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%SIMULATION DURATION  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE SIMULATION
     (UG/KG)
         %ORAL EXPOSURE  DOSE  (UG/KG)
         %ORAL EXPOSURE  DOSE  (UG/KG)
         %ORAL EXPOSURE  DOSE  (UG/KG)
E.2.3.2.2.  Chu et al (2007) and Chu et al (2001)
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG

% Chu et al. 2007
%protocol:  oral exposure daily for 28 days
                              0.250, 1.0 ug/kg every day  for  28
                              1000 ng/kg every day for 28 days
%dose levels :
%dose levels =
MAXT
CINT
EXP TIME ON
EXP TIME OFF
DAY CYCLE
BCK TIME ON
BCK TIME OFF
TIMELIMIT
BW TO
SIMULATION (G)
0.0025, 0.025,
2.5, 25, 250,
= 0.01
= 0.1
= 0.
= 672.
= 24.
= 0.
= 0.
= 672.
= 200.

                                            days
                                %TIME EXPOSURE BEGINS  (HOUR)
                                %TIME EXPOSURE ENDS  (HOUR)
                                %HOURS BETWEEN DOSES
                                %TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
                                %TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
                                %SIMULATION DURATIOHN  (HOUR)
                                %BODY WEIGHT AT THE BEGINNING OF  THE
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
  %MSTOT         =  0.0025       %ORAL EXPOSURE DOSE  (UG/KG)
  %MSTOT         =  0.025        %ORAL EXPOSURE DOSE  (UG/KG)
  %MSTOT         =  0.250        %ORAL EXPOSURE DOSE  (UG/KG)
  MSTOT         =1.0           %ORAL EXPOSURE DOSE  (UG/KG)
E.2.3.2.3.  Crofton et al (2005)
output @clear
prepare @clear
prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG

% Crofton et al. 2005
%protocol:  oral exposure daily for 4 days
%dose levels: 0.0001,  0.003,  0.01,  0.03, 0.1, 0.3, 1, 3, and  10  ug/kg every
day for four days
                                      E-39

-------
%dose levels:  0.1,  3,  10,  30,  100,  300, 1000, 3000, and  10000  ng/kg every day
for four days
MAXT             =   0.001
CINT             =   0.1
EXP_TIME_ON      =   0.
EXP_TIME_OFF     =   96.
DAY_CYCLE        =   24.
BCK_TIME_ON      =   0.
BCK_TIME_OFF     =   0.
TIMELIMIT        =   96.
BW_TO            =   250
SIMULATION  (G)
                      %TIME  EXPOSURE BEGINS  (HOUR)
                      %TIME  EXPOSURE ENDS  (HOUR)
                      %HOURS BETWEEN DOSES
                      %TIME  BACKGROUND EXPOSURE BEGINS(HOUR)
                      %TIME  BACKGROUND EXPOSURE ENDS  (HOUR)
                      %SIMULATION DURATION (HOUR)
                      %BODY  WEIGHT AT THE BEGINNING OF THE
%EXPOSURE
  MSTOT
  %MSTOT
  %MSTOT
  %MSTOT
  %MSTOT
  %MSTOT
  %MSTOT
  %MSTOT
  MSTOT
DOSE SCENARIOS  (UG/KG)
        = 0.0001
         0.003
         0.01
         0.03
           1
           3
 = 0.
 = 0.
 = 1.
 = 3.
= 10.
            %ORAL EXPOSURE  DOSE  (UG/KG)
            %ORAL EXPOSURE  DOSE  (UG/KG)
            %ORAL EXPOSURE  DOSE  (UG/KG)
            %ORAL EXPOSURE  DOSE  (UG/KG)
            %ORAL EXPOSURE  DOSE  (UG/KG)
            %ORAL EXPOSURE  DOSE  (UG/KG)
            %ORAL EXPOSURE  DOSE  (UG/KG)
            %ORAL EXPOSURE  DOSE  (UG/KG)
            %ORAL EXPOSURE  DOSE  (UG/KG)
E.2.3.2.4.  Croutch et al (2005)
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG  CBNGKG

% Croutch et al. ,  2005
 MAXT
 CINT
 TIMELIMIT
 EXP_TIME_ON
 EXP_TIME_OFF
 DAY_CYCLE
 WEEK_FINISH
 BCK_TIME_ON
 BCK_TIME_OFF
  BW_TO
(G)

%EXPOSURE DOSE
   %MSTOTBCKGR
   %MSTOT
   %MSTOTBCKGR
   %MSTOT
   %MSTOTBCKGR
   %MSTOT
   %MSTOTBCKGR
   %MSTOT
   MSTOTBCKGR
          0.001
          0.1
          672
          72
          672
          72
          672
          0.
          0.02
          250
              %SIMULATION  DURATION (HOUR)
              %TIME  EXPOSURE BEGINS (HOUR)
              %TIME  EXPOSURE ENDS (HOUR)
              %HOURS  BETWEEN DOSES
              %LENGTH OF EXPOSURE (HOUR)
              %TIME  BACKGROUND EXPOSURE BEGINS(HOUR)
              %TIME  BACKGROUND EXPOSURE ENDS  (HOUR)
              %BODY  WEIGHT AT THE BEGINNING OF THE  SIMULATION
     SCENARIOS  (UG/KG)
          = 0.0125   %INITIAL LOADING DOSE [UG/KG]
          = 0.00125  %EXPOSURE DOSE [UG/KG]
                     %INITIAL LOADING DOSE [UG/KG]
                     %EXPOSURE DOSE [UG/KG]
                     %INITIAL LOADING DOSE [UG/KG]
                     %EXPOSURE DOSE [UG/KG]
                     %INITIAL LOADING DOSE [UG/KG]
                     %EXPOSURE DOSE [UG/KG]
                     %INITIAL LOADING DOSE [UG/KG]
 = 0
0.05
0.005
  2
                                      E-40

-------
   MSTOT
                =  0.32
           %EXPOSURE DOSE  [UG/KG]
E.2.3.2.5.  Fattore et al (2000)
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG

% Fattore et al. 2000
%protocol:  oral exposure in diet for 13 weeks; SD rats
%dose levels:   0.02,  0.1,  0.2,  2 ug/kg 7 days/week for 13 weeks
%dose levels equivalent  to:  20,  100,  200, 2000 ng/kg 7 days/week  for 13 weeks
MAXT = 0.01
CINT  =0.1
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
BW_TO
(G)
 0.         %TIME EXPOSURE BEGINS  (HOUR)
 2184       %TIME EXPOSURE ENDS  (HOUR)
 24         %HOURS BETWEEN DOSES
 0.         %TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
 0.         %TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
 2184       %SIMULATION DURATION  (HOUR)
 150        %BODY WEIGHT AT THE BEGINNING OF  THE SIMULATION
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
 %MSTOT             =0.02      %EXPOSURE DOSE IN UG/KG
  %MSTOT           =0.1        %EXPOSURE DOSE IN UG/KG
  %MSTOT           =0.2        %EXPOSURE DOSE IN UG/KG
  MSTOT          = 2           %EXPOSURE DOSE IN UG/KG
E.2.3.2.6.  Fox et al (1993)
output @clear
prepare @clear
prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% Fox 1993
 MAXT
 CINT
 TIMELIMIT
 EXP_TIME_ON
 EXP_TIME_OFF
 DAY_CYCLE
 BCK_TIME_ON
 BCK_TIME_OFF
  BW_TO
(G)
 0.001
 0.1
 336
 96
 336
 96
 0.
 0.02
= 200
%SIMULATION DURATION  (HOUR)
%TIME EXPOSURE BEGINS(HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS(HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE SIMULATION
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
   MSTOTBCKGR       =  0.005     %INITIAL LOADING DOSE  [UG/KG]
   MSTOT            =  0.0009    %EXPOSURE DOSE  [UG/KG]
   %MSTOTBCKGR   =2.5         %INITIAL LOADING DOSE  [UG/KG]
   %MSTOT        =0.6         %EXPOSURE DOSE  [UG/KG]
                                      E-41

-------
   %MSTOTBCKGR   =  12.
   %MSTOT        =3.5
     %INITIAL LOADING DOSE  [UG/KG]
     %EXPOSURE DOSE [UG/KG]
E.2.3.2.7.  Franc et al (2001) Sprague-Dawley rats
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% Franc et al.  2001
% dose levels:  0.140,  0.420,  and 1.400 ug/kg every 2 weeks  for  22  weeks
% dose levels:  140, 420, and  1400 ng/kg every 2 weeks for 22 weeks
% dose levels equivalent to  10,  30,  and 100 ng/kg-day
MAXT            =   0.01
CINT            =   0.1
EXP_TIME_ON     =   0.
EXP_TIME_OFF    =   3696.
DAY_CYCLE       =   336.
BCK_TIME_ON     =   0.
BCK_TIME_OFF    =   0.
TIMELIMIT       =   3696.
BW_TO           =   200.
SIMULATION  (G)
      %TIME EXPOSURE BEGINS  (HOUR)
      %TIME EXPOSURE ENDS  (HOUR)

       %TIME BACKGROUND EXPOSURE BEGINS  (HOUR)
       %TIME OF BACKGROUND EXPOSURE ENDS  (HOUR)
       %SIMULATION DURATION  (HOUR)
       %BODY WEIGHT AT THE BEGINNING OF THE
%EXPOSURE DOSE SCENARIOS
   %MSTOT         =0.14
   %MSTOT         =0.42
   MSTOT         =1.4
(UG/KG)
       %ORAL EXPOSURE DOSE  (UG/KG)
       %ORAL EXPOSURE DOSE  (UG/KG)
       %ORAL EXPOSURE DOSE  (UG/KG)
E.2.3.2.8.  Franc et al. (2001) Long-Evans rats
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% Franc et al.  2001
% dose levels:  0.140,  0.420,  and 1.400 ug/kg every 2 weeks  for  22  weeks
% dose levels:  140, 420, and  1400 ng/kg every 2 weeks for 22 weeks
% dose levels equivalent to  10,  30,  and 100 ng/kg-day
MAXT            =   0.01
CINT            =   0.1
EXP_TIME_ON     =   0.
EXP_TIME_OFF    =   3696.
DAYJCYCLE       =   336.
BCK_TIME_ON     =   0.
BCK_TIME_OFF    =   0.
TIMELIMIT       =   3696.
BW_TO           =   190.
SIMULATION  (G)
        %TIME EXPOSURE BEGINS  (HOUR)
        %TIME EXPOSURE ENDS  (HOUR)
        %HOURS BETWEEN DOSES
        %TIME BACKGROUND EXPOSURE BEGINS  (HOUR)
        %TIME BACKGROUND EXPOSURE ENDS  (HOUR)
        %SIMULATION DURATION  (HOUR)
        %BODY WEIGHT AT THE BEGINNING OF  THE
%EXPOSURE DOSE SCENARIOS
   %MSTOT         =0.14
   %MSTOT         =0.42
(UG/KG)
        %ORAL EXPOSURE DOSE  (UG/KG)
        %ORAL EXPOSURE DOSE  (UG/KG)
                                      E-42

-------
   MSTOT
                 =  1.4
%ORAL EXPOSURE DOSE  (UG/KG)
E.2.3.2.9.  Franc et al (2001) Hans Wistar rats
output @clear
prepare @clear
prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% Franc et al.  2001
% dose levels:  0.140,  0.420,  and 1.400 ug/kg every 2 weeks for 22 weeks
% dose levels:  140, 420, and  1400 ng/kg every 2 weeks for 22 weeks
% dose levels equivalent to  10,  30,  and 100 ng/kg-day
MAXT            =   0.01
CINT            =   0.1
EXP_TIME_ON     =   0.
EXP_TIME_OFF    =   3696.
DAY_CYCLE       =   336.
BCK_TIME_ON     =   0.
BCK_TIME_OFF    =   0.
TIMELIMIT       =   3696.
BW_TO           =   205.
SIMULATION  (G)

%EXPOSURE DOSE SCENARIOS  (UG/KG)
   %MSTOT         =0.14
   %MSTOT         =0.42
   MSTOT         =1.4
 %TIME EXPOSURE BEGINS(HOUR)
 %TIME EXPOSURE ENDS  (HOUR)
 %HOURS BETWEEN DOSES
 %TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
 %TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
 %SIMULATION DURATION  (HOUR)
 %BODY WEIGHT AT THE BEGINNING OF  THE
 %ORAL EXPOSURE DOSE  (UG/KG)
 %ORAL EXPOSURE DOSE  (UG/KG)
 %ORAL EXPOSURE DOSE  (UG/KG)
E.2.3.2.10. Hassoun et al. (2000)
output @clear
prepare @clear
prepare T CLINGKG CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% Hassoun et al. 2000
%protocol:  oral exposure for 13  weeks;  SD rats
%dose levels:   0.003,  0.010,  0.022,  0.046 0.1 ug/kg 5 days/week  for  13  weeks
%dose levels equivalent to:  3,  10,  22,  46 100 ng/kg 5 days/week  for  13  weeks
%dose levels equivalent to:  2.14,  7.14,  15.7, 32.9 71.4 ng/kg 7  days/week  for
13 weeks
MAXT             =  0.01
CINT             =0.1
EXP_TIME_ON      =  0.
EXP_TIME_OFF     =  2184.
DAYJCYCLE        =24.
WEEK_PERIOD      =  168.
WEEK_FINISH      =  119.
BCK_TIME_ON      =  0.
BCK_TIME_OFF     =  0.
TIMELIMIT        =  2184.
BW_TO            =215.
SIMULATION  (G)
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%HOURS IN A WEEK
%LAST HOUR IN WEEK WHEN DOSE  OCCURS
%TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%SIMULATION DURATION  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE
                                      E-43

-------
%EXPOSURE DOSE SCENARIOS  (UG/KG)
     %MSTOT
     %MSTOT
     %MSTOT
     %MSTOT
     MSTOT
  =  0.003
=  0.010
=  0.022
=  0.046
0.1
 %EXPOSURE
 %EXPOSURE
 %EXPOSURE
 %EXPOSURE
 %EXPOSURE
DOSE UG/KG
DOSE UG/KG
DOSE UG/KG
DOSE UG/KG
DOSE UG/KG
E.2.3.2.11. Huttetal (2008)
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% Hutt et al. 2008
% dose levels: 0.050 ug/kg  every week for 13 weeks
% dose levels: 50  ng/kg every week for 13 weeks
% dose levels equivalent  to 7.14 ng/kg-day
MAXT            =   0.01
CINT            =   0.1
EXP_TIME_ON     =   0.
EXP_TIME_OFF    =   2184.
DAY_CYCLE       =   168.
BCK_TIME_ON     =   0.
BCK_TIME_OFF    =   0.
TIMELIMIT       =   2184.
BW_TO           =   4.5
SIMULATION  (G)

%EXPOSURE DOSE SCENARIOS
   MSTOT         =0.05
               %TIME EXPOSURE BEGINS  (HOUR)
               %TIME EXPOSURE ENDS  (HOUR)
               %HOURS BETWEEN DOSES
               %TIME BACKGROUND EXPOSURE BEGINS(HOUR)
               %TIME BACKGROUND EXPOSURE ENDS(HOUR)
               %SIMULATION DURATION  (HOUR)
               %BODY WEIGHT AT THE BEGINNING OF THE
        (UG/KG)
               %ORAL EXPOSURE DOSE  (UG/KG)
E.2.3.2.12. Kitchin and Woods (1979)
output @clear
prepare @clear
prepare T CLINGKG CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG
% Kitchen and Woods  1979
%protocol: single oral  gavage
%dose levels:  0.0006,  0.002,  0
5.000, 20.000 ug/kg  single  oral
% dose levels =  0.6,  2, 4,  20,
oral gavage
MAXT             =  0.001
CINT             =  0.1
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK TIME ON
BCK_TIME_OFF
TIMELIMIT
BW_TO
SIMULATION  (G)
  0
  0.
  24.
  24.
  0.
  0.
  24.
  225.
              .004,  0.020, 0.060, 0.200, 0.600, 2.000,
               gavage
              60,  200,  600, 2000, 5000, 20000 ng/kg  single
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS(HOUR)
%TIME OF BACKGROUND EXPOSURE  ENDS  (HOUR)
%SIMULATION DURATION  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF THE
                                      E-44

-------
%EXPOSURE
%MSTOT
%MSTOT
%MSTOT
%MSTOT
%MSTOT
%MSTOT
%MSTOT
%MSTOT
%MSTOT
MSTOT
DOSE SCENARIOS
= 0
= 0
= 0
= 0
= 0
= 0
= 0
= 2
= 5
= 20
.0006
.002
.004
.020
.060
.200
.600
.000
.000
.000
                          (UG/KG)
                                 %ORAL EXPOSURE DOSE  (UG/KG)
                                 %ORAL EXPOSURE DOSE  (UG/KG)
                                 %ORAL EXPOSURE DOSE  (UG/KG)
                                 %ORAL EXPOSURE DOSE  (UG/KG)
                                 %ORAL EXPOSURE DOSE  (UG/KG)
                                 %ORAL EXPOSURE DOSE  (UG/KG)
                                 %ORAL EXPOSURE DOSE  (UG/KG)
                                 %ORAL EXPOSURE DOSE  (UG/KG)
                                 %ORAL EXPOSURE DOSE  (UG/KG)
                                 %ORAL EXPOSURE DOSE  (UG/KG)
E.2.3.2.13. Kociba et al (1976) 13 weeks
output @clear
prepare @clear
prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% Kociba et al. 1976.
%dose levels:   0.001,  0.01,  0.1,  1  ug/kg 5 days/week for  13 weeks
%dose levels:  1,  10,  100,  1000 ng/kg 5 days/week for 13 weeks
%dose levels equivalent  to:  0.714,  7.14, 71.4, 714 ng/kg-d  (adj)  7  days/week
for 13 weeks
MAXT
CINT
EXP_TIME_ON
EXP_TIME_OFF
WEEK_PERIOD
WEEK_FINISH
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
BW_TO
(G)
%EXPOSURE DOSE
%MSTOT
 %MSTOT
 %MSTOT
 MSTOT
     0.001
     0.1
     0.
     2184
     168
     119
      24
     0.
     0.
     2184
     180
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS IN A WEEK
%LAST HOUR IN WEEK WHEN  DOSE  OCCURS
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS(HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%SIMULATION DURATION  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE SIMULATION
SCENARIOS  (UG/KG)
  = 0.001      %ORAL EXPOSURE  DOSE (UG/KG)
  =0.01       %ORAL EXPOSURE  DOSE (UG/KG)
  =0.1        %ORAL EXPOSURE  DOSE (UG/KG)
 = 1           %ORAL EXPOSURE  DOSE (UG/KG)
E.2.3.2.14. Kociba et al. (1978) female, 104 weeks

output @clear
prepare @clear
prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG

% Kociba et al, 1978.
%protocol:  daily dietary exposure for 104 weeks; SD rats
%dose levels:   0.001,  0.01,  0.1 ug/kg 7 days/week for 104 weeks
%dose levels: 1,  10,  100 ng/kg 7 days/week for 104 weeks
MAXT
CINT
 =  0.01
 =  0.1
                                      E-45

-------
EXP_TIME_ON     =   0.
EXP_TIME_OFF    =   17472
DAY_CYCLE       =   24
BCK_TIME_ON     =   0.
BCK_TIME_OFF    =   0.
TIMELIMIT       =   17472
BW_TO           =   180
SIMULATION  (G)

%EXPOSURE DOSE SCENARIOS
 %MSTOT          =  0.001
 %MSTOT         =0.01
 MSTOT         =0.1
             %TIME EXPOSURE  BEGINS  (HOUR)
             %TIME EXPOSURE  ENDS  (HOUR)
             %HOURS BETWEEN  DOSES
             %TIME BACKGROUND  EXPOSURE BEGINS (HOUR)
             %TIME BACKGROUND  EXPOSURE ENDS (HOUR)
             %SIMULATION DURATION  (HOUR)
             %BODY WEIGHT AT THE BEGINNING OF THE
        (UG/KG)
             %EXPOSURE DOSE  IN  UG/KG
             %EXPOSURE DOSE  IN  UG/KG
             %EXPOSURE DOSE  IN  UG/KG
E.2.3.2.15. Kociba et al (1978) male, 104 weeks
output @clear
prepare @clear
prepare T CLINGKG CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG

% Kociba et al, 1978.
%dose levels:   0.001,  0.01,  0.1  ug/kg 7 days/week for 104 weeks
%dose levels: 1,  10,  100  ng/kg 7 days/week for 104 weeks
MAXT            =   0.01
CINT            =   0.1
EXP_TIME_ON     =   0.
EXP_TIME_OFF    =   17472
DAYJCYCLE       =   24
BCK_TIME_ON     =   0.
BCK_TIME_OFF    =   0.
TIMELIMIT       =   17472
BW_TO           =   250
SIMULATION  (G)

%EXPOSURE DOSE SCENARIOS
 %MSTOT          =  0.001
 %MSTOT         =0.01
 MSTOT         =0.1
              %TIME EXPOSURE  BEGINS  (HOUR)
              %TIME EXPOSURE  ENDS  (HOUR)
              %HOURS BETWEEN  DOSES
              %TIME BACKGROUND  EXPOSURE BEGINS (HOUR)
              %TIME BACKGROUND  EXPOSURE ENDS (HOUR)
              %SIMULATION  DURATION  (HOUR)
              %BODY WEIGHT AT THE BEGINNING OF THE
        (UG/KG)
              %EXPOSURE DOSE  IN  UG/KG
              %EXPOSURE DOSE  IN  UG/KG
              %EXPOSURE DOSE  IN  UG/KG
E.2.3.2.16. Latchoumycandane andMathur (2002)
output @clear
prepare @clear
prepare T CLINGKG CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% Latchoumycandane and Mathur  2002.
%protocol:  1 time per day  for 45  days oral gavage
%dose levels:  0.001, 0.01,  0.1 ug/kg daily for 45 days
%dose levels:  1, 10, 100 ng/kg daily for 45 days
MAXT
CINT
EXP_TIME_ON
EXP TIME OFF
= 0.01
= 0.1
= 0.
= 1080
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
                                      E-46

-------
DAY_CYCLE          =  24
BCK_TIME_ON        =  0.
BCK_TIME_OFF       =  0.
TIMELIMIT          =  1080
BW_TO              =  200
SIMULATION  (G)
                       %HOURS  BETWEEN DOSES
                       %TIME BACKGROUND EXPOSURE BEGINS  (HOUR)
                       %TIME OF BACKGROUND EXPOSURE ENDS  (HOUR)
                       %SIMULATION DURATION (HOUR)
                       %BODY WEIGHT AT THE BEGINNING OF THE
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
  %MSTOT            =  0.001       %EXPOSURE DOSE UG/KG
  %MSTOT           =0.01         %EXPOSURE DOSE UG/KG
  MSTOT          =0.1           %EXPOSURE DOSE UG/KG
E.2.3.2.17. LietaL (1997)
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG

% Li et al 1997
% dose levels: 3,  10,  30,  100,  300,  1000, 3000, 10000, 30000  nkd  one  dose via
gavage, sacrificed 24  hrs  later
MAXT            =   0.1
CINT            =   0.1
EXP_TIME_ON     =   0.
EXP_TIME_OFF    =   24.
DAYJCYCLE       =   24.
BCK_TIME_ON     =   0.
BCK_TIME_OFF    =   0.
TIMELIMIT       =   24.
BW_TO           =   56.5
SIMULATION  (G)
%EXPOSURE
   MSTOT
   %MSTOT
   %MSTOT
   %MSTOT
   %MSTOT
   %MSTOT
   %MSTOT
   %MSTOT
   %MSTOT
DOSE SCENARIOS
       = 0.003
        = 0.01
        = 0.03
        = 0.1
        = 0.3
        = 1.
        = 3.
        = 10.
        = 30.
                       %TIME  EXPOSURE BEGINS (HOUR)
                       %TIME  EXPOSURE ENDS(HOUR)
                       %HOURS BETWEEN DOSES
                       %TIME  BACKGROUND EXPOSURE BEGINS  (HOUR)
                       %TIME  BACKGROUND EXPOSURE ENDS  (HOUR)
                       %SIMULATION DURATION (HOUR)
                       %BODY  WEIGHT AT THE BEGINNING OF THE
(UG/KG)
       %ORAL EXPOSURE DOSE  (UG/KG)
       %ORAL EXPOSURE DOSE  (UG/KG)
       %ORAL EXPOSURE DOSE  (UG/KG)
       %ORAL EXPOSURE DOSE  (UG/KG)
       %ORAL EXPOSURE DOSE  (UG/KG)
       %ORAL EXPOSURE DOSE  (UG/KG)
       %ORAL EXPOSURE DOSE  (UG/KG)
       %ORAL EXPOSURE DOSE  (UG/KG)
       %ORAL EXPOSURE DOSE  (UG/KG)
E.2.3.2.18. Murray et al. (1979)
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG
% Murray et al  1979
%built and check  in August 7  2009
%protocol:  dietary exposure  for 3 generations
each)
                                      (assume 120 day exposure  for
                                      E-47

-------
%dose levels:  0.001  0.01,  0.1  ug/kg-d
%dose levels:  1,  10,  100   ng/kg-d
MAXT            =   0.01
CINT            =   0.1
EXP_TIME_ON     =   0.
EXP_TIME_OFF    =   2880
DAY_CYCLE       =   24.
BCK_TIME_ON     =   0.
BCK_TIME_OFF    =   0.
TIMELIMIT       =   2880
BW_TO           =   4.5
SIMULATION  (G)
               %TIME EXPOSURE  BEGINS (HOUR)
               %TIME EXPOSURE  ENDS  (HOUR)
               %HOURS BETWEEN  DOSES
               %TIME BACKGROUND  EXPOSURE BEGINS (HOUR)
               %TIME BACKGROUND  EXPOSURE ENDS (HOUR)
               %SIMULATION  DURATION  (HOUR)
               % BODY WEIGHT AT  THE  BEGINNING OF THE
%EXPOSURE DOSE SCENARIOS  (UG/KG)
  %MSTOT           =  0.001
  %MSTOT         =0.01
  MSTOT         =0.1
               %ORAL EXPOSURE  DOSE IN UG/KG
               %ORAL EXPOSURE  DOSE IN UG/KG
               %ORAL EXPOSURE  DOSE IN UG/KG
E.2.3.2.19. NTP (1982) female, chronic
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

%NTP 1982
%dose levels:   0.005,  0.025,  0.25 ug/kg twice weekly for 104 weeks
%dose levels:   5,  25,  250 ng/kg twice weekly for 104 weeks
%dose levels equivalent to:  1.43, 7.14, 71.4 ng/kg-day  (adj)
MAXT
CINT
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
BW_TO
SIMULATION  (G)
   0.01
   0.1
   0.
   17472
   84
   0.
   0.
   17472
   250
%EXPOSURE DOSE SCENARIOS  (UG/KG)
  %MSTOT
  %MSTOT
  MSTOT
 = 0.005
 = 0.025
= 0.25
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%SIMULATION DURATION  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE
%EXPOSURE DOSE UG/KG
%EXPOSURE DOSE UG/KG
%EXPOSURE DOSE UG/KG
E.2.3.2.20. NTP (1982) male,chronic
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

%NTP 1982
%dose levels:   0.005,  0.025,  0.25 ug/kg twice weekly for 104 weeks
                                      E-48

-------
%dose levels:  5, 25,  250  ng/kg  twice weekly for 104 weeks
%dose levels equivalent  to:  1.43,  7.14,  71.4 ng/kg-day  (adj'
MAXT               =  0.01
CINT               =0.1
EXP_TIME_ON        =  0.
EXP_TIME_OFF       =  17472
DAY_CYCLE          =84
BCK_TIME_ON        =  0.
BCK_TIME_OFF       =  0.
TIMELIMIT          =  17472
BW_TO              =  350
SIMULATION  (G)
%EXPOSURE DOSE SCENARIOS  (UG/KG)
                  %TIME EXPOSURE BEGINS  (HOUR)
                  %TIME EXPOSURE ENDS(HOUR)
                  %HOURS BETWEEN DOSES
                  %TIME BACKGROUND EXPOSURE BEGINS  (HOUR)
                  %TIME BACKGROUND EXPOSURE ENDS  (HOUR)
                  %SIMULATION DURATION  (HOUR)
                  %BODY WEIGHT AT THE BEGINNING OF  THE
%MSTOT
%MSTOT
MSTOT
=  0.005
=  0.025
0.25
%EXPOSURE DOSE UG/KG
%EXPOSURE DOSE UG/KG
%EXPOSURE DOSE UG/KG
E.2.3.2.21. NTP (2006)14 weeks
output @clear
prepare @clear
prepare T CLINGKG CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% NTP 2006
%dose levels:  0.003,  0.010,  0.022,  0.046 0.1 ug/kg 5 days/week for  14 weeks
%dose levels equivalent to:  3,  10,  22,  46 100 ng/kg 5 days/week for  14 weeks
%dose levels equivalent to:  2.14,  7.14,  15.7, 32.9 71.4 ng/kg-day days/week
MAXT             =  0.01
CINT             =0.1
EXP_TIME_ON      =  0.
EXP_TIME_OFF     =  2352
DAYJCYCLE        =   24
WEEK_PERIOD      =  168
WEEK_FINISH      =  119
BCK_TIME_ON      =  0.
BCK_TIME_OFF     =  0.
TIMELIMIT        =  2352
BW_TO            =215
SIMULATION  (G)
%EXPOSURE DOSE SCENARIOS  (UG/KG)
     %MSTOT         = 0.003
     %MSTOT      =  0.010
     %MSTOT      =  0.022
     %MSTOT      =  0.046
     MSTOT      =0.1
                  %TIME EXPOSURE BEGINS  (HOUR)
                  %TIME EXPOSURE ENDS  (HOUR)
                  %HOURS BETWEEN DOSES
                  %HOURS IN A WEEK
                  %LAST HOUR IN WEEK WHEN DOSE OCCURS
                  %TIME BACKGROUND EXPOSURE BEGINS  (HOUR)
                  %TIME BACKGROUND EXPOSURE ENDS  (HOUR)
                  %SIMULATION DURATION  (HOUR)
                  %BODY WEIGHT AT THE BEGINNING OF  THE
                  %EXPOSURE
                  %EXPOSURE
                  %EXPOSURE
                  %EXPOSURE
                  %EXPOSURE
          DOSE UG/KG
          DOSE UG/KG
          DOSE UG/KG
          DOSE UG/KG
          DOSE UG/KG
E.2.3.2.22. NTP (2006) 31 weeks
output @clear
prepare @clear
prepare T CLINGKG CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% NTP 2006
                                      E-49

-------
%dose levels:  0.003,  0.010,  0.022,  0.046 0.1 ug/kg 5 days/week for 31 weeks
%dose levels equivalent  to:  3,  10,  22,  46 100 ng/kg 5 days/week for 31 weeks
%dose levels equivalent  to:  2.14,  7.14,  15.7, 32.9 71.4 ng/kg 7 days/week for
31 weeks
MAXT             =  0.01
CINT             =0.1
EXP_TIME_ON      =  0.
EXP_TIME_OFF     =5208
DAY_CYCLE        =  24
WEEK_PERIOD      =  168
WEEK_FINISH      =  119
BCK_TIME_ON      =  0.
BCK_TIME_OFF     =  0.
TIMELIMIT        =  5208
BW_TO            =215
SIMULATION  (G)

%EXPOSURE DOSE SCENARIOS  (UG/KG)
     %MSTOT         =  0.003
     %MSTOT      =  0.010
     %MSTOT      =  0.022
     %MSTOT      =  0.046
     MSTOT      =0.1
 %TIME EXPOSURE BEGINS  (HOUR)
 %TIME EXPOSURE ENDS  (HOUR)
 %HOURS BETWEEN DOSES
 %HOURS IN A WEEK
 %LAST HOUR IN WEEK WHEN DOSE OCCURS
 %TIME BACKGROUND EXPOSURE BEGINS(HOUR)
 %TIME BACKGROUND EXPOSURE ENDS  (HOUR)
 %SIMULATION DURATION  (HOUR)
 %BODY WEIGHT AT THE BEGINNING OF  THE
 %EXPOSURE
 %EXPOSURE
 %EXPOSURE
 %EXPOSURE
 %EXPOSURE
DOSE UG/KG
DOSE UG/KG
DOSE UG/KG
DOSE UG/KG
DOSE UG/KG
E.2.3.2.23. NTP (2006) 53 weeks

output @clear
prepare @clear
prepare T CLINGKG CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% NTP 2006
%protocol:  oral exposure for  53  weeks;  SD rats
%dose levels:  0.003,  0.010, 0.022,  0.046 0.1 ug/kg 5 days/week for 53 weeks
%dose levels equivalent to:  3,  10,  22,  46 100 ng/kg 5 days/week for 53 weeks
%dose levels equivalent to:  2.14,  7.14,  15.7, 32.9 71.4 ng/kg 7 days/week for
53 weeks
MAXT             =  0.01
CINT             =0.1
EXP_TIME_ON      =  0.
EXP_TIME_OFF     =  8904
DAYJCYCLE        =  24
WEEK_PERIOD      =  168
WEEK_FINISH      =  119
BCK_TIME_ON      =  0.
BCK_TIME_OFF     =  0.
TIMELIMIT        =  8904
BW_TO            =215
SIMULATION  (G)
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS(HOUR)
%HOURS BETWEEN DOSES
%HOURS IN A WEEK
%LAST HOUR IN WEEK WHEN DOSE OCCURS
%TIME BACKGROUND EXPOSURE BEGINS(HOUR)
%TIME BACKGROUND EXPOSURE ENDS  (HOUR)
%SIMULATION DURATION  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF THE
%EXPOSURE DOSE SCENARIOS  (UG/KG)
     %MSTOT        =  0.003
     %MSTOT      = 0.010
     %MSTOT      = 0.022
     %MSTOT      = 0.046
%EXPOSURE DOSE UG/KG
%EXPOSURE DOSE UG/KG
%EXPOSURE DOSE UG/KG
%EXPOSURE DOSE UG/KG
                                      E-50

-------
     MSTOT
                =  0.1
               %EXPOSURE DOSE UG/KG
E.2.3.2.24. NTP (2006) 2 year
output @clear
prepare @clear
prepare T CLINGKG CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG

% NTP 2006
%protocol:  oral exposure for  105  weeks;  SD rats
%dose levels:  0.003,  0.010,  0.022,  0.046,  0.1 ug/kg 5 days/week for 105
weeks
%dose levels equivalent to:  3,  10,  22,  46,  100 ng/kg 5 days/week for 105
weeks
%dose levels equivalent to:  2.14,  7.14,  15.7,  32.9,  71.4 ng/kg 7 days/week
for 105 weeks
MAXT             =  0.01
CINT             =0.1
EXP_TIME_ON      =  0.
EXP_TIME_OFF     =  17640
DAY_CYCLE        =  24
WEEK_PERIOD      =  168
WEEK_FINISH      =  119
BCK_TIME_ON      =  0.
BCK_TIME_OFF     =  0.
TIMELIMIT        =  17640
BW_TO            =215
SIMULATION  (G)

%EXPOSURE DOSE SCENARIOS  (UG/KG)
     %MSTOT         =  0.003
     %MSTOT      =  0.010
     %MSTOT      =  0.022
     %MSTOT      =  0.046
     MSTOT      =0.1
               %TIME EXPOSURE BEGINS  (HOUR)
               %TIME EXPOSURE ENDS  (HOUR)
               %HOURS BETWEEN DOSES
               %HOURS IN A WEEK
               %LAST HOUR IN WEEK WHEN  DOSE  OCCURS
               %TIME BACKGROUND  EXPOSURE  BEGINS (HOUR)
               %TIME BACKGROUND  EXPOSURE  ENDS  (HOUR)
               %SIMULATION DURATION  (HOUR)
               %BODY WEIGHT AT THE BEGINNING OF THE
               %EXPOSURE DOSE
               %EXPOSURE DOSE
               %EXPOSURE DOSE
               %EXPOSURE DOSE
               %EXPOSURE DOSE
               IN UG/KG
               IN UG/KG
               IN UG/KG
               IN UG/KG
               IN UG/KG
E.2.3.2.25. Sewalletal (1995) and Maronpot et al (1993)
output @clear
prepare @clear
prepare T CLINGKG CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG
% Sewall et al. 1995
%protocol: gavage every 2 weeks  for 30  weeks
%dose levels:  0.049,  0.1498,  0.49,  and  1.75 ug/kg every 2 weeks
%dose levels:  3.5,  10.7,  35,  and 125 ng/kg-d or 49, 149.8, 490, and 1750
ng/kg every 2 weeks
MAXT
CINT
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK TIME OFF
0.01
0.1
0.
5040
336.
0.
0.
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS(HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE BEGINS(HOUR)
%TIME BACKGROUND EXPOSURE ENDS  (HOUR)
                                      E-51

-------
TIMELIMIT       =   5040
BW_TO           =   250
SIMULATION  (G)
%SIMULATION DURATION  (HOUR)
%BODY WEIGHT AT THE BEGINNING  OF  THE
%EXPOSURE DOSE SCENARIOS  (UG/KG)
  %MSTOT           =  0.049
  %MSTOT         = 0.1498
  %MSTOT         = 0.49
  MSTOT         =1.75
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
E.2.3.2.26. Shi et al (2007) adult portion

output @clear
prepare @clear
prepare T CLINGKG  CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG

% Shi et al  2007
%protocol:   gavage once per week for 322 days
%dose levels:  0.001,  0.005, 0.05 and 0.2 ug TCDD:kg body weight by  gavage
once per week
%dose levels:  1, 5,  50  and 200  ng/kg ng TCDD:kg body weight by gavage  once
per week
% dose equivalent  adjusted 0.143,  0.714, 7.14 and 28.6 ng/kg-d
MAXT            =   0.0001
CINT            =   0.1
EXP_TIME_ON     =   504.
EXP_TIME_OFF    =   7728
DAY_CYCLE       =   168.
BCK_TIME_ON     =   0.
BCK_TIME_OFF    =   0.
TIMELIMIT       =   7728
BW_TO           =   4.5
SIMULATION  (G)
 %TIME EXPOSURE BEGINS  (HOUR)
 %TIME EXPOSURE ENDS  (HOUR)
 %HOURS BETWEEN DOSES
 %TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
 %TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
 %SIMULATION DURATION  (HOUR)
 %BODY WEIGHT AT THE BEGINNING OF  THE
%EXPOSURE DOSE SCENARIOS  (UG/KG)
   %MSTOT          =  0.001
   %MSTOT         = 0.005
   %MSTOT         =0.05
   MSTOT         = 0.2
 %ORAL EXPOSURE DOSE IN UG/KG
 %ORAL EXPOSURE DOSE IN UG/KG
 %ORAL EXPOSURE DOSE IN UG/KG
 %ORAL EXPOSURE DOSE IN UG/KG
E.2.3.2.27. Van Birgelen et al. (1995)
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG

% Van Birgelen et  al.  (1995)
%protocol:  daily  dietary exposure for 13 weeks
%dose levels: 0.0135,  0.0264,  0.0469,  0.320, 1.024 ug/kg every day  for  13
weeks
% dose levels =  13.5,  26.4,  46.9,  320, 1024 ng/kg every day for  13  weeks
MAXT             =   0.001
CINT             =   0.1
EXP TIME ON      =0.               %TIME EXPOSURE BEGINS  (HOUR)
                                      E-52

-------
EXP_TIME_OFF    =   2184.
DAY_CYCLE       =   24.
BCK_TIME_ON     =   0.
BCK_TIME_OFF    =   0.
TIMELIMIT       =   2184.
BW_TO           =   150.
SIMULATION  (G)

%EXPOSURE DOSE  SCENARIOS  (UG/KG)
  %MSTOT         =  0.0135
  %MSTOT         =  0.0264
  %MSTOT         =  0.0469
  %MSTOT         =  0.320
  MSTOT         = 1.024
               %TIME EXPOSURE  ENDS  (HOUR)
               %HOURS  BETWEEN  DOSES
               %DELAY  BEFORE BACKGROUND EXPOSURE  (HOUR)
               %TIME OF  BACKGROUND  EXPOSURE STOP  (HOUR)
               %SIMULATION  LIMIT  TIME (HOUR)
               %BODY WEIGHT AT THE  BEGINNING OF THE
               %ORAL EXPOSURE
               %ORAL EXPOSURE
               %ORAL EXPOSURE
               %ORAL EXPOSURE
               %ORAL EXPOSURE
               DOSE  (UG/KG)
               DOSE  (UG/KG)
               DOSE  (UG/KG)
               DOSE  (UG/KG)
               DOSE  (UG/KG)
E.2.3.2.28. Simanainen et al (2002) and Simanainen etal (2003)

output @clear
prepare @clear
prepare T CLINGKG  CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% Simanainen et  al.,  2002  and Simanainen et al., 2003
 MAXT
 CINT
 TIMELIMIT
 EXP_TIME_ON
 EXP_TIME_OFF
 DAY_CYCLE
 BCK_TIME_ON
 BCK_TIME_OFF
 BW_TO
SIMULATION  (G)
0.01
0.1
24
0
24
24
0.
0.
200
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
   %MSTOT        =0.1
   MSTOT        =0.3
%SIMULATION DURATION  (HOUR)
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE
               %EXPOSURE  DOSE  [UG/KG]
               %EXPOSURE  DOSE  [UG/KG]
E.2.3.2.29. Vanden Heuvel et al (1994)
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% Vanden Heuvel et al. 1994.
%protocol: single  gavage
%dose levels:0.00005,  0.0001,  0.001,  0.010, 0.1, 1, 10 ug/kg-d
%dose levels equivalent  to:  0.05,  0.1,  1, 10, 100, 1000,  10000  ng/kg-d
MAXT
CINT
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK TIME ON
 0.001
 0.1
 0.
 24
 24
 0.
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS(HOUR)
                                      E-53

-------
BCK_TIME_OFF        =  0.
TIMELIMIT           =  24
BW_TO               =  250
SIMULATION  (G)

%EXPOSURE DOSE  SCENARIOS  (UG/KG)
  %MSTOT
  %MSTOT
  %MSTOT
  %MSTOT
  %MSTOT
  %MSTOT
  MSTOT
   = 0.00005
 = 0.0001
 = 0.001
 = 0.01
 = 0.1
 = 1
= 10
                 %TIME BACKGROUND  EXPOSURE ENDS (HOUR)
                 %SIMULATION  DURATION (HOUR)
                 %BODY WEIGHT AT THE BEGINNING OF THE
%EXPOSURE
%EXPOSURE
%EXPOSURE
%EXPOSURE
%EXPOSURE
%EXPOSURE
%EXPOSURE
DOSE
DOSE
DOSE
DOSE
DOSE
DOSE
DOSE
UG/KG
UG/KG
UG/KG
UG/KG
UG/KG
UG/KG
UG/KG
E.2.4.  Rat Gestational Model

E.2.4.1. Model Code

PROGRAM: Three Compartment PBPK Model for TCDD in Rat (Gestation)'


INITIAL   ! INITIALIZATION  OF PARAMETERS

      !SIMULATION  PARAMETERS  ====
CONSTANT PARA_ZERO           = 1E-30
CONSTANT EXP_TIME_ON      =0.0
CONSTANT EXP_TIME_OFF     = 530
CONSTANT DAYJCYCLE        = 24.0
CONSTANT BCK_TIME_ON      =0.0
BEGINS  (HOURS)
CONSTANT BCK_TIME_OFF     =0.0
(HOURS)
CONSTANT TRANSTIME_ON     = 144.0
AT GESTATIONAL DAY  6
                    ! TIME AT WHICH EXPOSURE BEGINS (HOURS)
                    ! TIME AT WHICH EXPOSURE ENDS  (HOURS)
                    ! NUMBER OF  HOURS  BETWEEN DOSES (HOURS)
                    ! TIME AT WHICH BACKGROUND EXPOSURE

                    ! TIME AT WHICH BACKGROUND EXPOSURE ENDS

                    !CONTROL TRANSFER  FROM MOTHER TO FETUS
  IUNIT CONVERSION
CONSTANT MW=322  ! MOLECULAR WEIGHT (NG/NMOL)
CONSTANT SERBLO  =0.55
CONSTANT UNITCORR =  1000
     !INTRAVENOUS  SEQUENCE
constant IV_LAG          =0.0
constant IV PERIOD       =0.0
     !PREGNANCY  PARAMETER ====
CONSTANT CONCEPTION_T          =0.0
CONSTANT N FETUS          = 10.0
                         ITIME  OF CONCEPTION(HOUR)
                     !NUMBER  OF FETUS  PRESENT
     !CONSTANT EXPOSURE  CONTROL ===========
     IACUTE, SUBCHRONIC,  CHRONIC EXPOSURE =====
     !OR BACKGROUND  EXPOSURE (IN THIS CASE 3 TIMES A DAY)===
CONSTANT MSTOTBCKGR      =0.0       !  ORAL BACKGROUND EXPOSURE DOSE  (UG/KG)
CONSTANT MSTOT            =0.0       !  ORAL EXPOSURE DOSE  (UG/KG)
                                      E-54

-------
     IORAL ABSORPTION
  MSTOT_NM = MSTOT/MW              !  CONVERTS THE DOSE  TO  NMOL/G

     !INTRAVENOUS ABSORPTION
CONSTANT  DOSEIV          =0.0       !  INJECTED DOSE  (UG/KG)
  DOSEIV_NM =  DOSEIV/MW            !  CONVERTS THE INJECTED DOSE TO NMOL/G
CONSTANT DOSEIVLATE =0.0            !  INJECTED DOSE  LATE  (UG/KG)
  DOSEIVNMlate = DOSEIVLATE/MW     IAMOUNT IN NMOL/G

     !INITIAL GUESS OF THE FREE CONCENTRATION IN THE LIGAND (COMPARTMENT
INDICATED BELOW)====
CONSTANT CFLLIO          =0.0  !LIVER     (NMOL/ML)
CONSTANT CFLPLAO          =0.0  !PLACENTA  (NMOL/ML)

     !BINDING CAPACITY (AhR)  FOR NON LINEAR BINDING  (COMPARTMENT INDICATED
BELOW)  (NMOL/ML) ===
CONSTANT LIBMAX          = 3.5E-4    !  LIVER   (NMOL/ML), WANG ET AL.  1997
CONSTANT PLABMAX          = 2.0E-4    !TEMPORARY PARAMETER

     ! PROTEIN  AFFINITY CONSTANTS (1A2  OR AhR, COMPARTMENT  INDICATED BELOW)
(NMOL/ML)===
CONSTANT KDLI             = l.OE-4    !LIVER  (AhR)  (NMOL/ML),  WANG ET AL.  1997
CONSTANT KDLI2           = 4.OE-2    !LIVER  (1A2)  (NMOL/ML),  EMOND ET AL. 2004
CONSTANT KDPLA           = l.OE-4    !TEMPORARY PARAMETER;  ASSUME IDENTICAL TO
KDLI  (AhR)

     !EXCRETION AND ABSORPTION CONSTANT
CONSTANT KST              = 0.36     !  GASTRIC RATE CONSTANT  (HR-1),  WANG ET
AL.  1997
CONSTANT KABS             = 0.48     !INTESTINAL ABSORPTION CONSTANT (HR-1)  ),
WANG ET AL. 1997

     ! ELIMINATION CONSTANTS
CONSTANT CLURI           =0.01       !  URINARY CLEARANCE  (ML/HR),  EMOND ET
AL.  2004

     IINTERSPECIES ELIMINATION VARIABLE
CONSTANT kelv             =0.15     !  INTERSPECIES VARIABLE  ELIMINATION
CONSTANT  (I/HOUR)

     ! CONSTANT TO DIVIDE THE ABSORPTION INTO LYMPHATIC  AND PORTAL FRACTIONS
CONSTANT A                =0.7        !  LYMPHATIC FRACTION,  WANG ET AL.  1997

     !PARTITION COEFFICIENTS
CONSTANT PF               = 100     !  ADIPOSE TISSUE/BLOOD, WANG ET AL.  1997
CONSTANT PRE              =1.5      !  REST OF THE BODY/BLOOD,  WANG ET AL.
1997
CONSTANT PLI              =6.0       !  LIVER/BLOOD, WANG ET  AL.  1997
CONSTANT PPLA             =1.5      !  TEMPORARY PARAMETER  NOT  CONFIGURED,
WANG ET AL. 1997

     !PARAMETER FOR INDUCTION OF CYP 1A2,  WANG ET AL.  1997
CONSTANT IND_ACTIVE        =1.0        !  INCLUDE INDUCTION?  (1 = YES,  0 = NO)
CONSTANT CYP1A2_10UTZ    =1.6      !  DEGRADATION CONCENTRATION CONSTANT OF
1A2  (NMOL/ML)
CONSTANT CYP1A2_1A1       =1.6      !  BASAL CONCENTRATION  OF 1A1 (NMOL/ML)

                                      E-55

-------
CONSTANT CYP1A2_1EC50     =0.13      !  DISSOCIATION CONSTANT TCDD-CYP1A2
(NMOL/ML)
CONSTANT CYP1A2_1A2       =1.6      ! BASAL CONCENTRATION OF 1A2  (NMOL/ML)
CONSTANT CYP1A2_1KOUT     =0.1      !  FIRST ORDER RATE OF DEGRADATION (H-l)
CONSTANT CYP1A2_1TAU      =0.25     !HOLDING TIME  (H)
CONSTANT CYP1A2_1EMAX     = 600      !  MAXIMUM INDUCTION OVER BASAL  EFFECT
(UNITLESS)
CONSTANT HILL             =0.6      I HILL CONSTANT; COOPERATIVE  LIGAND
BINDING EFFECT  CONSTANT  (UNITLESS)

     IDIFFUSIONAL  PERMEABILITY FRACTION
CONSTANT PAFF             = 0.0910   IADIPOSE (UNITLESS), WANG  ET AL.  1997
CONSTANT PAREF            = 0.0298   !REST OF THE BODY   (UNITLESS),  WANG ET
AL.  1997
CONSTANT PALIF            = 0.3500   !LIVER (UNITLESS), WANG ET AL.  1997
CONSTANT PAPLAF          =0.3      !TEMPORARY PARAMETER NOT CONFIGURED

    !FRACTION OF TISSUE WEIGHT =========
CONSTANT WLIO             = 0.0360   !LIVER, WANG ET AL. 1997

    !TISSUE BLOOD  FLOW EXPRESSED AS  A FRACTION OF CARDIAC OUTPUT
CONSTANT QFF              = 0.069    !  ADIPOSE TISSUE BLOOD FLOW  FRACTION
(UNITLESS), WANG  ET AL.  1997
CONSTANT QLIF             = 0.183    !LIVER (UNITLESS), WANG ET AL.  1997

    !COMPARTMENT TISSUE BLOOD  EXPRESSED AS A FRACTION OF THE TOTAL COMPARTMENT
VOLUME
CONSTANT WFBO             = 0.050    IADIPOSE TISSUE, WANG ET AL.  1997
CONSTANT WREBO            = 0.030    !REST OF THE BODY, WANG ET AL.  1997
CONSTANT WLIBO            = 0.266    !LIVER, WANG ET AL. 1997
CONSTANT WPLABO          = 0.500    !TEMPORARY PARAMETER NOT CONFIGURED

    !EXPOSURE SCENARIO FOR UNIQUE OR REPETITIVE WEEKLY OR MONTHLY EXPOSURE
    !NUMBER OF EXPOSURES  PER WEEK
CONSTANT WEEK_LAG       =0.0       ITIME ELAPSED BEFORE EXPOSURE BEGINS
(WEEK)
CONSTANT WEEK_PERIOD      = 168      !  NUMBER OF HOURS IN THE WEEK  (HOURS)
CONSTANT WEEK_FINISH      = 168      !  TIME EXPOSURE ENDS  (HOURS)

    !NUMBER OF EXPOSURES  PER MONTH
CONSTANT MONTH_LAG       =0.0       ITIME ELAPSED BEFORE EXPOSURE BEGINS
(MONTHS)

    !CONSTANT FOR  BACKGROUND EXPOSURE===========
CONSTANT Day_LAG_BG      =0.0       ITIME ELAPSED BEFORE EXPOSURE BEGINS
(HOURS)
CONSTANT Day_PERIOD_BG    =24       !LENGTH OF EXPOSURE  (HOURS)

    !NUMBER OF EXPOSURES  PER WEEK
CONSTANT WEEK_LAG_BG       =0.0      ITIME ELAPSED BEFORE BACKGROUND  EXPOSURE
BEGINS  (WEEKS)
CONSTANT WEEK_PERIOD_BG     =  168      !NUMBER OF HOURS IN THE  WEEK  (HOURS)
CONSTANT WEEK_FINISH_BG     =  168      ITIME EXPOSURE ENDS  (HOURS)

    I INITIAL BODY  WEIGHT
CONSTANT BW_TO              =  250      I   (IN G)  WANG ET AL. 1997


                                      E-56

-------
CONSTANT RATIO_RATF_MOUSEF =1.0
GESTATIONAL DAY 22
                              !RATIO OF FETUS MOUSE/RAT AT
    ! COMPARTMENT TOTAL  LIPID FRACTION
                              POULIN ET AL 2000
CONSTANT F_TOTLIP
CONSTANT B_TOTLIP
CONSTANT RE_TOTLIP
(UNITLESS)
CONSTANT LI_TOTLIP
CONSTANT PLA_TOTLIP
CONSTANT FETUS TOTLIP
                   0.855
                   0.0023
                   0.019

                   0.060
                   0.019
                   0.019
                ! ADIPOSE TISSUE  (UNITLESS)
                 ! BLOOD  (UNITLESS)
                ! REST OF THE  BODY

                ! LIVER  (UNITLESS)
END
         ! END OF THE  INITIAL SECTION
DYNAMIC  ! DYNAMIC  SIMULATION SECTION
ALGORITHM
CINTERVAL
MAXTERVAL
MINTERVAL
VARIABLE
CONSTANT
 CINTXY  = CINT
 PFUNC   = CINT
IALG
CINT
MAXT
MINT
T
TIMELIMIT
   2
  0.1
l.Oe+10
l.OE-10
  0.0
  100
  GEAR METHOD
  COMMUNICATION INTERVAL
  MAXIMUM CALCULATION INTERVAL
!  MINIMUM CALCULATION INTERVAL

!SIMULATION LIMIT TIME  (HOURS)
    ITIME CONVERSION
  DAY         = T/24
  WEEK        = T/168
  MONTH       = T/730
  YEAR        = T/8760
                     !  TIME  IN  DAYS
                     !  TIME  IN  WEEKS
                     !  TIME  IN  MONTHS
                     !  TIME  IN  YEARS
DERIVATIVE  ! PORTION  OF  CODE THAT SOLVES DIFFERENTIAL EQUATIONS
    !CHRONIC OR  SUBCHRONIC  EXPOSURE SCENARIO =======
    !NUMBER OF EXPOSURES  PER DAY
 DAY_LAG
 (HOURS)
 DAY_PERIOD
 DAY_FINISH
 MONTH_PERIOD
 MONTH FINISH
      = EXP_TIME_ON

       = DAY_CYCLE
       = CINTXY
       = TIMELIMIT
  TIME ELAPSED BEFORE EXPOSURE  BEGINS

 !  EXPOSURE PERIOD  (HOURS)
 !  LENGTH OF EXPOSURE  (HOURS)
 !  EXPOSURE PERIOD  (MONTHS)
       = EXP TIME OFF    !  LENGTH OF EXPOSURE (MONTHS)
    !NUMBER OF EXPOSURES  PER DAY AND MONTH
 DAY_FINISH_BG
 MONTH_LAG_BG
BEGINS  (MONTHS)
 MONTH_PERIOD_BG
 MONTH  FINISH BG
       = CINTXY
       = BCK TIME ON
ITIME ELAPSED BEFORE BACKGROUND  EXPOSURE
       = TIMELIMIT       !BACKGROUND EXPOSURE (MONTHS)
       = BCK TIME OFF    !LENGTH OF BACKGROUND EXPOSURE  (MONTHS)
    !INTRAVENOUS LATE
 IV_FINISH = CINTXY
 B = 1-A  ! FRACTION OF  DIOXIN ABSORBED IN THE PORTAL FRACTION OF THE  LIVER
!FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUM
E
   ! FROM OFLAHERTY  1992
                                      E-57

-------
RTESTGEST= T-CONCEPTION_T
TESTGEST=DIM(RTESTGEST,0.0)

WTFER_RODENT=  (2.3d-3*EXP(1.49d-2*(TESTGEST))+1.3d-2)*Gest_on
WTFER =  (WTFER_RODENT*RATIO_RATF_MOUSEF*N_FETUS)
WTFE = DIM(WTFER,0.0)

  i
FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME
  !  FAT GROWTH EXPRESSION LINEAR DURING PREGNANCY
  !  FROM 0'FLAHERTY_1992

WFO=  (((9.66d-5*(TESTGEST))*gest_on)+0.069)

  !  PLACENTA,VOLUME,  PLACENTA,VOLUME,  PLACENTA,VOLUME,  PLACENTA,VOLUME
  !  WPLA PLACENTA GROWTH   EXPRESSION,  SINGLE EXPONENTIAL WITH  OFFSET
  !  FROM 0'FLAHERTY_1992   !  FOR EACH PUP

WPLAON_RODENT =  (0.6/(1+(5d+3*EXP(-0.0225*(TESTGEST)))))*N_FETUS
WPLAOR =  (WPLAON_RODENT/WTO)*Gest_on
WPLAO = DIM(WPLAOR,0.0)

  !  PLACENTA,FLOW RATE,  PLACENTA,FLOW RATE, PLACENTA,FLOW  RATE, PLACENTA,FLOW
RATE
  !  QPLA PLACENTA GROWTH  EXPRESSION, DOUBLE EXPONENTIAL WITH OFFSET
  !  FROM 0'FLAHERTY_1992

 QPLARF =  (1.67d-7  *exp(9.6d-3*(TESTGEST))  &
   +1.6d-3*exp(7.9d-3*(TESTGEST))+0.0)*Gest_on*SWITCH_trans
 QPLAF=DIM(QPLARF,0.0)                !FRACTION OF FLOW  RATE IN PLACENTA

  !  GESTATION CONTROL
IF  (T.LT.CONCEPTION_T)  THEN
    Gest_off =1.0
    Gest_on=   0.0
ELSE
    Gest_off =0.0
    Gest_on  =1.0
END IF

  !  MOTHER BODY WEIGHT  GROWTH EQUATION========
  !  MODIFICATION  TO ADAPT THIS MODEL AT HUMAN MODEL
  !  BECAUSE LINEAR DESCRIPTION IS NOT GOOD ENOUGH FOR MOTHER GROWTH
  !  MOTHER BODY WEIGHT  GROWTH

  PARAMETER  (BW_RMN =  l.OE-30)
  WTO= BW_TO *(1+(0.41*T)/(1402.5+T+BW_RMN)) !   IN GRAMS

  !  VARIABILITY OF REST  OF THE BODY DEPENDS ON OTHER  ORGANS
  WREO =  (0.91 -  (WLIBO*WLIO  + WFBO*WFO +WPLABO*WPLAO + WLIO + WFO  +
WPLAO))/(1+WREBO)   !  REST OF  THE BODY FRACTION; UPDATED FOR EPA ASSESSMENT
  QREF = 1-(QFF+QLIF+QPLAF)             !REST OF BODY  BLOOD FLOW RATE (ML/HR)
  QTTQF = QFF+QREF+QLIF+QPLAF          !  SUM MUST EQUAL 1

  !  COMPARTMENT VOLUME  (ML OR G)  =========
 WF  =  WFO  * WTO                      ! ADIPOSE TISSUE
 WRE =  WREO * WTO                      ! REST OF THE BODY

                                      E-58

-------
 WLI =  WLIO * WTO
 WPLA=  WPLAO* WTO
                                       !  LIVER
                                       !  PLACENTA
     COMPARTMENT TISSUE  BLOOD (ML OR G)
 WFB  =  WFBO  * WF
 WREB =  WREBO * WRE
 WLIB =  WLIBO * WLI
 WPLAB = WPLABO* WPLA
                                       !  ADIPOSE TISSUE
                                       !  REST OF THE BODY
                                       !  LIVER
                                       !  PLACANTA
    ! CARDIAC OUTPUT  FOR THE GIVEN BODY WEIGHT (ML/H) =========
    !QC= QCCAR*60* (WTO/1000.0) **0. 75
CONSTANT QCC=18684.0                     !  EQUIVALENT TO 311.4 *  60
QC= QCC* (WTO/UNITCORR) **0. 75

    ! COMPARTMENT BLOOD  FLOW RATE (ML/HR)
QF  = QFF*QC
QLI = QLIF*QC
QRE = QREF*QC
QPLA = QPLAF*QC
QTTQ = QF+QRE+QLI+QPLA
                                       IADIPOSE TISSUE BLOOD FLOW  RATE
                                       ! LIVER TISSUE BLOOD FLOW  RATE
                                       ! REST OF THE BODY BLOOD FLOW RATE
                                       ! PLACENTA TISSUE BLOOD FLOW RATE
                          ! TOTAL FLOW RATE
     ! PERMEABILITY ORGAN  FLOW (ML/HR) =========
PAF  = PAFF*QF
PARE = PAREF*QRE
PALI = PALIF*QLI
PAPLA = PAPLAF*QPLA

    ! *******************
    !  ABSORPTION SECTION
    !  ORAL
    !  INTRAPERITONEAL
    !  INTRAVENOUS
                                      !  ADIPOSE TISSUE
                                      !  REST OF THE BODY
                                      !  LIVER TISSUE
                                      !  PLACENTA
     ! REPETITIVE ORAL  BACKGROUND EXPOSURE SCENARIO

MSTOT_NMBCKGR = MSTOTBCKGR/MW       !  CONVERTS THE BACKGROUND DOSE  TO  NMOL/G
MSTTBCKGR =MSTOT_NMBCKGR *WTO

DAY_EXPOSURE_BG   = PULSE (DAY_LAG_BG, DAY_PERIOD_BG, DAY_FINISH_BG)
WEEK_EXPOSURE_BG  = PULSE (WEEK_LAG_BG, WEEK_PERIOD_BG, WEEK_FINISH_BG)
MONTH_EXPOSURE_BG = PULSE (MONTH_LAG_BG,MONTH_PERIOD_BG,MONTH_FINISH_BG)

MSTTCH_BG =  (DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG) *MSTTBCKGR
MSTTFR_BG = MSTTBCKGR/CINT

CYCLE_BG =DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG

     ! CONDITIONAL ORAL  EXPOSURE (BACKGROUND EXPOSURE)

IF  (MSTTCH_BG.EQ. MSTTBCKGR)  THEN
    ABSMSTT_GB= MSTTFR_BG
ELSE
    ABSMSTT_GB =0.0
END IF

CYCLETOTBG=INTEG ( CYCLE_BG, 0.0)
                                      E-59

-------
    !REPETITIVE ORAL  EXPOSURE SCENARIO

MSTT= MSTOT_NM * WTO                   !AMOUNT IN NMOL

DAY_EXPOSURE   = PULSE(DAY_LAG,DAY_PERIOD,DAY_FINISH)
WEEK_EXPOSURE  = PULSE(WEEK_LAG,WEEK_PERIOD,WEEK_FINISH)
MONTH_EXPOSURE = PULSE(MONTH_LAG,MONTH_PERIOD,MONTH_FINISH)

MSTTCH =  (DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE)*MSTT
MSTTFR = MSTT/CINT

CYCLE = DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE
SUMEXPEVENT= INTEG  (CYCLE,0.0)  !NUMBER OF CYCLES GENERATED DURING  SIMULATION

    ! CONDITIONAL ORAL EXPOSURE
IF  (MSTTCH.EQ.MSTT)  THEN
   ABSMSTT= MSTTFR
ELSE
   ABSMSTT =0.0
END IF
 CYCLETOT=INTEG(CYCLE, 0.0)

    ! MASS CHANGE  IN  THE  LUMEN
 RMSTT= -(KST+KABS)*MST  +ABSMSTT +ABSMSTT_GB !  RATE OF CHANGE  (NMOL/H)
  MST = INTEG(RMSTT,0.0)                       !AMOUNT REMAINING IN  DUODENUM
(NMOL)

    ! ABSORPTION IN LYMPH CIRCULATION
 LYRMLUM = KABS*MST*A
  LYMLUM = INTEG(LYRMLUM,0.0)

    ! ABSORPTION IN PORTAL CIRCULATION
 LIRMLUM = KABS*MST*B
  LIMLUM = INTEG(LIRMLUM,0.0)
i  	IV EXPOSURE  	

 IV= DOSEIV_NM * WTO  IAMOUNT  IN NMOL
 IVR= IV/PFUNC  ! RATE  FOR IV  INFUSION IN BLOOD
 EXPIV= IVR *  (1.0-STEP(PFUNC))
 IVDOSE = integ(EXPIV,0.0)

     i	IV LATE  IN  THE CYCLE
     ! MODIFICATION  ON  January 13 2004
 IV_RlateR = DOSEIVNMlate*WTO
 IV_EXPOSURE=PULSE(IV_LAG,IV_PERIOD, IV_FINISH)

 IV_lateT = IV_EXPOSURE  *IV_RlateR
 IV_late = IV_lateT/CINT

SUMEXPEVENTIV= integ  (IV_EXPOSURE, 0.0)  !NUMBER OF CYCLES GENERATED  DURING
SIMULATION


                                      E-60

-------
     !SYSTEMIC  CONCENTRATION OF TCDD
      ! MODIFICATION  ON OCTOBER 6, 2009
 CB=  (QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM+QPLA*CPLAB+IV_late)/(QC+CLURI)  !
  CA = CB  ! CONCENTRATION (NMOL/ML)
     !URINARY EXCRETION BY KIDNEY
     ! MODIFICATION  ON OCTOBER 6, 2009
RAURI = CLURI  *CB
  AURI = INTEG(RAURI,0.0)
  IUNIT CONVERSION  POST SIMULATION
 CBSNGKGLIADJ=(CB*MW*UNITCORR*(1.0/B_TOTLIP)*(1.0/SERBLO))![NG of TCDD
Serum/Kg OF LIPID]
   AUCBS NGKGLIADJ=integ(CBSNGKGLIADJ,0.0)
  CBNGKG= CB*MW*UNITCORR
   !ADIPOSE COMPARTMENT
   ! TISSUE BLOOD  COMPARTMENT
RAFB= QF*(CA-CFB)-PAF*(CFB-CF/PF)
 AFB = INTEG(RAFB,0.0)
  CFB = AFB/WFB
   !TISSUE COMPARTMENT
RAF = PAF*(CFB-CF/PF)
 AF = INTEG(RAF,0.0)
  CF  = AF/WF
!(NMOL/H)
 !(NMOL)
  (NMOL/ML)
!(NMOL/H)
!(NMOL)
 !(NM/ML)
   IUNIT CONVERSION  POST SIMULATION
  CFTOTAL=  (AF  + AFB)/(WF + WFB)  !  TOTAL CONCENTRATION IN NMOL/ML
  CFTFREE = CFB +  CF !TOTAL FREE CONCENTRATION  IN  FAT  (NM/ML)

  CFNGKG=CFTOTAL*MW*UNITCORR !  FAT CONCENTRATION NG/KG
    AUCF_NGKGH=integ(CFNGKG,0.0)

   !REST OF THE BODY COMPARTMENT
RAREB= QRE *(CA-CREB)-PARE*(CREB-CRE/PRE)   !(NMOL/H)
 AREB = INTEG(RAREB,0.0)                      !(NMOL)
  CREB = AREB/WREB                           !(NMOL/H)
   !TISSUE COMPARTMENT
RARE = PARE*(CREB  -  CRE/PRE)                !(NMOL/H)
 ARE = INTEG(RARE,0.0)                       !(NMOL)
  CRE  = ARE/WRE                              !(NMOL/ML)
   IUNIT CONVERSION  POST SIMULATION
  CRETOTAL=  (ARE  + AREB)/(WRE + WREB)
NMOL/ML
           !  TOTAL CONCENTRATION  IN
  CRENGKG=CRETOTAL*MW*UNITCORR !  REST OF THE BODY  CONCENTRATION IN NG/KG
   !LIVER COMPARTMENT
                                      E-61

-------
   !TISSUE BLOOD COMPARTMENT
 RALIB = QLI*(CA-CLIB)-PALI*(CLIB-CFLLIR)+LIRMLUM  !
  ALIB = INTEG(RALIB,0.0)                         !(NMOL)
   CLIB = ALIB/WLIB                              !(NMOL/ML)
   !TISSUE COMPARTMENT
 RALI = PALI*(CLIB -  CFLLIR)-REXCLI             !   (NMOL/HR)
  ALI = INTEG(RALI,0.0)                               !(NMOL)
   CLI  = ALI/WLI                                  !(NMOL/ML)

   IFREE TCDD CONCENTRATION IN LIVER COMPARTMENT
PARAMETER  (LIVER_1RMN = l.OE-30)
 CFLLI= IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLIR  &
        +LIVER_1RMN))+((CYP1A2_103*CFLLIR/(KDLI2 + CFLLIR  &
        +LIVER_1RMN)*IND_ACTIVE)))-CFLLI,CFLLIO)
     CFLLIR=DIM(CFLLI,0.0)  !  FREE CONCENTRATION IN LIVER

  CBNDLI= LIBMAX*CFLLIR/(KDLI+CFLLIR+LIVER_1RMN)  !BOUND CONCENTRATION

  !VARIABLE ELIMINATION BASED ON  THE CYP1A2
 KBILE_LI_T =((CYP1A2_10UT-CYP1A2_1A2)/CYP1A2_1A2)*Kelv  !  INDUCED  BILIARY
EXCRETION RATE CONSTANT IN LIVER
  REXCLI = KBILE_LI_T*CFLLIR*WLI  !  DOSE-DEPENDENT BILIARY  EXCRETION RATE
    EXCLI = INTEG(REXCLI,0.0)

  IUNIT CONVERSION POST SIMULATION
  CLITOTAL= (ALI + ALIB)/(WLI + WLIB)  !  TOTAL CONCENTRATION  IN  NMOL/ML
  Rec_occ= CFLLIR/(KDLI+CFLLIR)
  CLINGKG=CLITOTAL*MW*UNITCORR !  LIVER CONCENTRATION NG/KG
     AUCLI_NGKGH=INTEG(CLINGKG, 0.0)
  CBNDLINGKG = CBNDLI*MW*UNITCORR
     AUCBNDLI NGKGH  =INTEG(CBNDLINGKG,0.0)
   !CHEMICAL IN CYP450  (1A2)  COMPARTMENT
CYP1A2 1KINP = CYP1A2  1KOUT*  CYP1A2 10UTZ
    !  MODIFICATION ON  OCTOBER 6,  2009
CYP1A2_10UT =INTEG(CYP1A2_1KINP * (1.0 + CYP1A2_1EMAX *(CBNDLI+1.Oe-30)**HILL
&
     /(CYP1A2_1EC50**HILL  +  (CBNDLI+1.Oe-30)**HILL))  &
      - CYP1A2_1KOUT*CYP1A2_10UT, CYP1A2_10UTZ)

!  EQUATIONS INCORPORATING  DELAY OF CYP1A2 PRODUCTION (NOT USED  IN
SIMULATIONS)

CYP1A2_1R02 =  (CYP1A2_10UT -  CYP1A2_102)/ CYP1A2_1TAU
  CYP1A2_102 =INTEG(CYP1A2_1R02,  CYP1A2_1A1)

CYP1A2_1R03 =  (CYP1A2_102  -  CYP1A2_103)/ CYP1A2_1TAU
  CYP1A2_103 =INTEG(CYP1A2_1R03,  CYP1A2_1A2)

!  TRANSFER OF DIOXIN FROM  PLACENTA TO FETUS
!  FETAL EXPOSURE ONLY  DURING  EXPOSURE

IF (T.LT.TRANSTIME_ON)  THEN
 SWITCH_trans =0.0

                                      E-62

-------
ELSE
 SWITCH_trans =1.0
END IF

!TRANSFER OF DIOXIN  FROM PLACENTA TO FETUS
!  MODIFICATION  26  SEPTEMBER 2003

CONSTANT PFETUS= 4.0 !
CONSTANT CLPLA  FET =0.17 !
RAMPF =  (CLPLA_FET*CPLA)  *SWITCH_trans
  AMPF=INTEG(RAMPF,0.0)

!TRANSFER OF  DIOXIN FROM FETUS TO PLACENTA
RAFPM =  (CLPLA_FET*CFETUS_v)*SWITCH_trans !
  AFPM = INTEG(RAFPM,0.0)

!  TCDD IN PLACENTA (MOTHER)  COMPARTMENT
RAPLAB= QPLA*(CA - CPLAB)-PAPLA* (CPLAB -CFLPLAR)
 APLAB = INTEG(RAPLAB,0.0)
 CPLAB = APLAB/(WPLAB+1E-30)
RAPLA = PAPLA*(CPLAB-CFLPLAR)-RAMPF + RAFPM
 APLA = INTEG(RAPLA,0.0)
 CPLA  = APLA/(WPLA+le-30)
               !  NMOL/H)
                !  (NMOL)
               !  (NMOL/ML)
               !  (NMOL/H)
                !  (NMOL)
               !  (NMOL/ML)
PARAMETER  (PARA_ZERO  = l.OE-30)
CFLPLA= IMPLC(CPLA-(CFLPLAR*PPLA +(PLABMAX*CFLPLAR/(KDPLA&
    +CFLPLAR+PARA_ZERO)))-CFLPLA,CFLPLAO)
CFLPLAR=DIM(CFLPLA,0.0)
    IUNIT CONVERSION  POST SIMULATION
  CPLATOTAL=  (APLA + APLAB)/((WPLA H
NMOL/ML
WPLAB)+le-30)!  TOTAL  CONCENTRATION IN
    !FETUS COMPARTMENT
RAFETUS= RAMPF-RAFPM
 AFETUS=INTEG(RAFETUS,0.0)
CFETUS=AFETUS/(WTFE+1E-30)
CFETOTAL= CFETUS
CFETUS v = CFETUS/PFETUS
   ! UNIT CONVERSION  POST SIMULATION
CFETUSNGKG =  CFETUS*MW*UNITCORR
AUC FENGKGH = INTEG(CFETUSNGKG,0.0)
                !(NG/KG)
i  	CONTROL MASS BALANCE	
BDOSE= IVDOSE +LYMLUM+LIMLUM
BMASSE = EXCLI+AURI+AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB+AFETUS
BDIFF = BDOSE-BMASSE
       IBODY BURDEN  (NG)
BODY_BURDEN = AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB  !
BBFETUSNG     = AFETUS*MW*UNITCORR    !  UNIT  (NG)

                                      E-63

-------
       ! BODY BURDEN  IN  TERMS OF CONCENTRATION  (NG/KG)
 BBNGKG =(((AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB)/WTO)*MW*UNITCORR)  !
  AUC BBNGKGH=INTEG(BBNGKG,0.0)
i  	COMMAND  OF THE END OF SIMULATION	
TERMT  (T.GE. TimeLimit,  'Time limit has been reached.
END    ! END OF THE  DERIVATIVE SECTION
END    ! END OF THE  DYNAMIC SECTION
END    ! END OF THE  PROGRAM
E.2.4.2. Input Files

E.2.4.2.1.  Bell et al (2007)
output @clear
prepare @clear T  CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH
AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH
CBNGKG AUC CBNGKGH
%Bell et al. 2007  (rat  species)
%protocol:  daily  dietary dose for 12 weeks followed by a two-week  mating
time and 21-day  gestation period
%dose levels:  0.0024,  0.008,  0.046 ug/kg-d with 0.00003 ug/kg-d  background
%dose levels:  2.4,  8,  46  ng/kg-d with 0.03 ng/kg-day background

   %EXPOSURES  SCENARIOS
 MAXT
 CINT
 EXP_TIME_ON
 EXP_TIME_OFF
 DAY_CYCLE
 BCK_TIME_ON
 BCK_TIME_OFF
 TIMELIMIT
 BW_TO
SIMULATION  (G)
 CONCEPTION_T
 TRANSTIME_ON
 N FETUS
0.01
0.1
0
2856
24
0.
2856.
2856
85

2352
2496
10
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%SIMULATION DURATION(HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE

%HOUR OF CONCEPTION  (HOUR)
%HOUR OF CONCEPTION +  6 DAYS(144  HOURS)
%NUMBER OF FETUSES
%EXPOSURE DOSE SCENARIOS  (UG/KG)
   MSTOT            =  0.00243
   %MSTOT           = 0.008
   %MSTOT = 0.0461
                %ORAL  EXPOSURE  DOSE (UG/KG)
                %ORAL  EXPOSURE  DOSE (UG/KG)
                %ORAL  EXPOSURE  DOSE (UG/KG)
E.2.4.2.2.  Hojo et al. (2002)
%clear variable
output @clear
prepare @clear T  CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH
AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH
CBNGKG AUC_CBNGKGH
%Hojo et al. 2002
%protocol:  single  oral  dose at GD8
                                      E-64

-------
%dose levels: 0.02  0.06,  0.18  ug/kg  at GD8
%dose levels: 20, 60,  180   ng/kg  at GD8
% author provided the  body  weight for each group at the beginning of
gestation  (g)
    %20 ng/kg BW =  271g
    %60 ng/kg BW =  275g
    %180 ng/kg BW = 262g

%EXPOSURES SCENARIOS
 MAXT= 0.001
 CINT =0.1
 EXP_TIME_ON     =192
 EXP_TIME_OFF    =  216
 DAY_CYCLE       =24
 BCK_TIME_ON     =  0.
 BCK_TIME_OFF    =  0.
 TIMELIMIT       =  216
 CONCEPTION_T    =  0.
 TRANSTIME_ON    =  144.
 N FETUS         =10
                    %TIME EXPOSURE  BEGINS  (HOUR)
                    %TIME EXPOSURE  ENDS  (HOUR)
                    %HOURS BETWEEN  DOSES
                    %TIME BACKGROUND  EXPOSURE BEGINS(HOUR)
                    %TIME BACKGROUND  EXPOSURE ENDS (HOUR)
                    %SIMULATION  DURATION  (HOUR)
                    %TIME OF  CONCEPTION_(HOUR)
                    %TIME OF  CONCEPTION +  6  DAYS(144 HOURS)
                    %NUMBER OF FETUSES
%EXPOSURE DOSE SCENARIOS  (UG/KG)
   %MSTOT
   %BW_TO

   %MSTOT
   %BW_TO

   MSTOT
   BW TO
  = 0.02
  = 275

 = 0.06
 = 262

= 0.18
= 278
%ORAL EXPOSURE DOSE  (UG/KG)
%20 ng/kg BW = 271g

%ORAL EXPOSURE DOSE  (UG/KG)
%60 ng/kg BW = 275g

%ORAL EXPOSURE DOSE  (UG/KG)
%180 ng/kg BW = 262g
E.2.4.2.3.  Ikeda et al (2005)
output @clear
prepare @clear T CLINGKG  CFNGKG  CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH
AUCF NGKGH AUCBS NGKGLIADJ AUC BBNGKGH AUC FENGKGH CBNDLINGKG AUCBNDLI NGKGH
%Ikeda et al. 2005
%protocol:   loading dose  of  400  ng/kg followed by weekly maintenance doses of
80 ng/kg for 6 weeks,
%dose levels: 0.4 ug/kg-day  followed by weekly 0.08 ug/kg-day
%dose levels: 400 ng/kg-day  followed by weekly 80 ng/kg-day

   %EXPOSURES SCENARIOS
 MAXT
 CINT
 EXP_TIME_ON
 EXP_TIME_OFF
 DAY_CYCLE
 BCK_TIME_ON
 BCK_TIME_OFF
 TIMELIMIT
 BW_TO
SIMULATION  (G)
 CONCEPTION T
   .1
    0.1
    0
    1008
    168
    0.
    167.
    1008
    250
       = 504
 %TIME EXPOSURE BEGINS  (HOUR)
 %TIME EXPOSURE ENDS  (HOUR)
 %HOURS IN A WEEK
 %TIME BACKGROUND EXPOSURE BEGINS  (HOUR)
 %TIME BACKGROUND EXPOSURE ENDS  (HOUR)
 %SIMULATION DURATION  (HOUR)
 %BODY WEIGHT AT THE BEGINNING OF  THE

 %TIME OF CONCEPTION  (HOUR)
                                      E-65

-------
 TRANSTIME_ON      =648
 N FETUS           =10
%TIME OF CONCEPTION +  6  DAYS  (144  HOURS)
%NUMBER OF FETUSES
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
   MSTOT            =0.08
   MSTOTBCKGR       =0.32
%ORAL EXPOSURE DOSE IN UG/KG
%BACKGROUND EXPOSURE IN UG/KG
E.2.4.2.4.  Kattainen et al (2001) and Simanainen et al (2004)
%clear variable
output @clear
prepare @clear T  CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG  CFETUSNGKG AUCLI_NGKGH
AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH
CBNGKG AUC_CBNGKGH

%Kattainen et al.  2001
%protocol:  single  gavage at GDIS
%dose levels: 0.03  0.1,  0.3,  1 ug/kg  at GDIS
%dose levels: 30,  100  300,  1000 ng/kg at GDIS

 MAXT=0.001
 CINT =0.1
   %EXPOSURES  SCENARIOS
 EXP_TIME_ON          =  336
 EXP_TIME_OFF         =  360
 DAYJCYCLE            =24
 BCK_TIME_ON          =  0.
 BCK_TIME_OFF         =  0.
 TIMELIMIT            =  360
 BW_TO                =  190
SIMULATION
 CONCEPTION_T              = 0.
 TRANSTIME_ON         =  144.
 N FETUS              =10
 %TIME EXPOSURE BEGINS  (HOUR)
 %TIME EXPOSURE ENDS  (HOUR)
 %HOURS BETWEEN DOSES
 %TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
 %TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
 %SIMULATION DURATION  (HOUR)
 %BODY WEIGHT AT THE BEGINNING OF  THE

 %TIME OF CONCEPTION  (HOUR)
 %TIME OF CONCEPTION +  6  DAYS(144  HOURS)
 %NUMBER OF FETUSES
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
 %MSTOT             =0.03
 %MSTOT             =0.1
 %MSTOT             =0.3
 MSTOT              =  1
 %ORAL EXPOSURE DOSE  (UG/KG)
 %ORAL EXPOSURE DOSE  (UG/KG)
 %ORAL EXPOSURE DOSE  (UG/KG)
 %ORAL EXPOSURE DOSE  (UG/KG
E.2.4.2.5.  Markowski et al (2001)
%clear variable
output @clear
prepare @clear T  CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH
AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH
CBNGKG AUC_CBNGKGH

%Markowski et al. 2001
%protocol:  single  gavage at GD18
%dose levels: 0.02  0.06,  0.18 ug/kg at GDIS
%dose levels: 20, 60,  180 ng/kg at GDIS
                                      E-66

-------
%EXPOSURES SCENARIOS
 MAXT=0.0001
 CINT =0.1
 EXP_TIME_ON      =408
 EXP_TIME_OFF     =432
 DAY_CYCLE        =24
 BCK_TIME_ON      =  0.
 BCK_TIME_OFF     =  0.
 TIMELIMIT        =  432
 BW_TO            =  190
SIMULATION
 CONCEPTION_T         = 0.
 TRANSTIME_ON     =  144.
 N FETUS          =  10
                      %TIME EXPOSURE BEGINS(HOUR)
                      %TIME EXPOSURE ENDS  (HOUR)
                      %HOURS BETWEEN DOSES
                      %TIME BACKGROUND  EXPOSURE BEGINS (HOUR)
                      %TIME BACKGROUND  EXPOSURE ENDS (HOUR)
                      %SIMULATION DURATION  (HOUR)
                      %BODY WEIGHT AT THE BEGINNING OF THE
                      %TIME OF CONCEPTION
                      %TIME OF CONCEPTION
                      %NUMBER OF FETUSES
                    (HOUR)
                    +  6  DAYS(144  HOURS)
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
   %MSTOT          =0.02
   %MSTOT         =0.06
   MSTOT         =0.18
                      %ORAL EXPOSURE DOSE  (UG/KG)
                      %ORAL EXPOSURE DOSE  (UG/KG)
                      %ORAL EXPOSURE DOSE  (UG/KG)
E.2.4.2.6.  Miettinen et al (2006)
%clear variable
output @clear
prepare gclear T  CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH
AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH
CBNGKG AUCJCBNGKGH

%Miettinen et al. 2006
%protocol:  single  oral  dose at GDIS
%dose levels: 0.03  0.1,  0.3, 1 ug/kg at GDIS
%dose levels: 30, 100,  300,  1000 ng/kg at GDIS

 MAXT=0.01
 CINT =0.1
EXP_TIME_ON
 EXP_TIME_OFF
 DAY_CYCLE
 BCK_TIME_ON
 BCK_TIME_OFF
 TIMELIMIT
 BW_TO
SIMULATION  (G)
 CONCEPTION_T
 TRANSTIME_ON
 N FETUS
 336
= 360
= 24
= 0.
= 0.
= 360
= 180

    = 0.
= 144.
= 10
%TIME EXPOSURE BEGINS(HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%SIMULATION DURATION  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE

%TIME OF CONCEPTION  (HOUR)
%TIME OF CONCEPTION +  6  DAYS(144  HOURS)
%NUMBER OF FETUSES
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
   %MSTOT         =0.03
   %MSTOT       =0.1
   %MSTOT       =0.3
   MSTOT      = 1
                       %ORAL EXPOSURE DOSE  (UG/KG)
                       %ORAL EXPOSURE DOSE  (UG/KG)
                       %ORAL EXPOSURE DOSE  (UG/KG)
                       %ORAL EXPOSURE DOSE  (UG/KG)
                                      E-67

-------
E.2.4.2.7.  Nohara et al (2000)
%clear variable
output @clear
prepare @clear T  CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH
AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH
CBNGKG AUC CBNGKGH
%Nohara et al. 2000
%protocol:  single gavage  at GDIS
%dose levels:  0.0125,  0.050,  0.2, or 0.8 ug TCDD:kg body weight by  gavage  on
GDIS.
%dose levels:  12.5,  50,  200,  or 800 ng TCDD:kg body weight by gavage  on  GDIS.
 MAXT=0.01
 CINT =0.1
 EXP_TIME_ON   =  336
 EXP_TIME_OFF  =  360
 DAY_CYCLE     =24
 BCK_TIME_ON   =  0.
 BCK_TIME_OFF  =  0.
 TIMELIMIT     =  360
 BW_TO         =  180
SIMULATION  (G)
 CONCEPTION_T  =  0.
 TRANSTIME_ON  =  144.
 N FETUS       =10
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%SIMULATION DURATION  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE

%TIME OF CONCEPTION  (HOUR)
%TIME OF CONCEPTION +  6  DAYS(144  HOURS)
%NUMBER OF FETUSES
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
  %MSTOT         =  0.0125
  %MSTOT         =  0.050
  %MSTOT         =0.2
  MSTOT         =0.8
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
E.2.4.2.8.  Ohsako et al (2001)
output @clear
prepare @clear T  CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH
AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH
CBNGKG AUC_CBNGKGH

%0hsako et al. 2001
%protocol:  single oral  dose at GDIS
%dose levels: 0.0125,  0.05,  0.2,  0.8 ug/kg at GDIS
%dose levels: 12.5,  50,  200,  800  ng/kg at GDIS
 MAXT=0.01
 CINT =0.1
 EXP_TIME_ON     =  360
 EXP_TIME_OFF    =  384
 DAYJCYCLE       =24
 BCK_TIME_ON     =  0.
 BCK_TIME_OFF    =  0.
 TIMELIMIT       =  384
 BW_TO           =  200
SIMULATION  (G)
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%SIMULATION DURATION  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE
                                      E-68

-------
 CONCEPTION_T
 TRANSTIME_ON
 N FETUS
       = 0.
  = 144.
  = 10
%TIME OF CONCEPTION_
%TIME OF CONCEPTION
%NUMBER OF FETUSES
(HOUR)
+ 6 DAYS(144 HOURS)
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
 %MSTOT
 %MSTOT
 %MSTOT
 MSTOT
= 0.0125
= 0.05
= 0.20
  = 0.80
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
E.2.4.2.9.  Schantz et al (1996) andAmin et al (2000)
%clear variable
output @clear
prepare gclear T  CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH
AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH
CBNGKG AUC CBNGKGH
%Amin et al. 2000  (rat  species)  and Schantz et al.  1996
%protocol:  daily  doses  on GDs  10 to 16
%dose levels: 25 and  100 ng/kg-day
%dose levels: 0.025 and  0.100 ug/kg-day
 MAXT
 CINT
 EXP_TIME_ON
 EXP_TIME_OFF
 DAY_CYCLE
 BCK_TIME_ON
 BCK_TIME_OFF
 TIMELIMIT
 BW_TO
SIMULATION  (G)
 CONCEPTION_T
 TRANSTIME_ON
 N FETUS
     0.001
     0.1
     240.
     384.
     24
     1000.
     1000.
     384.
     250.

        = 0
     144.
     10
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%SIMULATION DURATION  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE
%TIME OF CONCEPTION_
%TIME OF CONCEPTION
%NUMBER OF FETUSES
 (HOUR)
 - 6 DAYS(144 HOURS)
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
   %MSTOT            =  .025
   MSTOT            = .100
                        %ORAL  EXPOSURE DOSE (UG/KG)
                        %ORAL  EXPOSURE DOSE (UG/KG)
E.2.4.2.10. Seoetal (1995)
%clear variable
output @clear
prepare @clear T  CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH
AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH
CBNGKG AUC_CBNGKGH

%Seo et al.  1995
%protocol:   daily doses  on GDs  10-16
%dose levels: 0.025  and  0.1 ug/kg on GDs 10-16
%dose levels: 25  and 100 ng/kg  on GDs 10-16
                                      E-69

-------
 MAXT = 0.01
 CINT =0.1
 EXP_TIME_ON
 EXP_TIME_OFF
 DAY_CYCLE
 BCK_TIME_ON
 BCK_TIME_OFF
 TIMELIMIT
 BW_TO
SIMULATION  (G)
 CONCEPTION_T
 TRANSTIME_ON
 N FETUS
      240
      384
      24
      0.
      0.
      384
      190

         = 0.
      144.
      10
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)

%TIME BACKGROUND EXPOSURE  BEGINS (HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%SIMULATION DURATION  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF THE

%TIME OF CONCEPTION  (HOUR)
%TIME OF CONCEPTION +  6  DAYS(144 HOURS)
%NUMBER OF FETUSES
%EXPOSURE DOSE  SCENARIOS (UG/KG)
 %MSTOT             =  0.025
 MSTOT             =0.1
                      %ORAL EXPOSURE DOSE  (UG/KG)
                      %ORAL EXPOSURE DOSE  (UG/KG)
E.2.4.2.11. Sparschu etal (1971)
output @clear
prepare gclear  T  CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG  CFETUSNGKG AUCLI_NGKGH
AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH
CBNGKG AUC CBNGKGH
%protocol:  daily  oral  dose from GD6 to GDIS
%EXPOSURES SCENARIOS
 MAXT=0.01
 CINT =0.1
 EXP_TIME_ON      =  120.
 EXP_TIME_OFF     =  337.
 DAYJCYCLE        =24
 BCK_TIME_ON      =  0.
 BCK TIME OFF     =  0.
                      %TIME EXPOSURE BEGINS  (HOUR)
                      %TIME EXPOSURE ENDS  (HOUR)
                      %HOURS BETWEEN DOSES
                      %TIME BACKGROUND EXPOSURE BEGINS  (HOUR)
                      %TIME BACKGROUND EXPOSURE ENDS  (HOUR)
 TIMELIMIT
 BW_TO
SIMULATION  (G)
 ^CONCEPTION
 TRANSTIME_ON
 N FETUS
= 360.
= 295

     = 0.
= 144.
= 10
%SIMULATION DURATION  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF THE

%TIME OF CONCEPTION  (HOUR)
%TIME OF CONCEPTION +  6  DAYS(144 HOURS)
%NUMBER OF FETUSES
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
   %MSTOT
   %MSTOT
   %MSTOT
   %MSTOT
   MSTOT
 = 0.03
= 0.125
= 0.5
= 2.
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
                                      E-70

-------
E.2.5.  Mouse Standard Model
E.2.5.1. Model Code
PROGRAM: Three Compartment PBPK Model for TCDD in Mice: Standard Model
(Nongestation)'
INITIAL   ! INITIALIZATION  OF PARAMETERS

     ! SIMULATION  PARAMETERS ====
CONSTANT  PARA_ZERO       =    1D-30
CONSTANT  EXP_TIME_ON     =     0.0        !  TIME AT WHICH EXPOSURE  BEGINS
(HOURS)
CONSTANT  EXP_TIME_OFF    =    2832        !  TIME AT WHICH EXPOSURE  ENDS
(HOURS)
CONSTANT  DAYJCYCLE       =24         !  NUMBER OF HOURS BETWEEN DOSES
(HOURS)
CONSTANT  BCK_TIME_ON     =      0.0       !  TIME AT WHICH BACKGROUND EXPOSURE
BEGINS  (HOURS)
CONSTANT  BCK_TIME_OFF    =      0.0       !  TIME AT WHICH BACKGROUND EXPOSURE
ENDS (HOURS)

CONSTANT  MW=322  ! MOLECULAR WEIGHT (NG/NMOL)
CONSTANT  SERBLO  =0.55
CONSTANT  UNITCORR = 1000

     ! CONSTANT EXPOSURE  CONTROL ===========
     IACUTE, SUBCHRONIC, CHRONIC EXPOSURE =====
     !OR BACKGROUND EXPOSURE (IN THIS  CASE 3 TIMES A DAY) ===
CONSTANT  MSTOTBCKGR     =        0.0       ! ORAL BACKGROUND EXPOSURE DOSE
(UG/KG)
CONSTANT  MSTOT          =        0.15      ! ORAL EXPOSURE DOSE  (UG/KG)
CONSTANT  MSTOTsc       =        0.0       !  SUBCUTANEOUS EXPOSURE DOSE
(UG/KG)

     IORAL ABSORPTION
  MSTOT_NM              =    MSTOT/MW     IAMOUNT IN NMOL/G

     ! INTRAVENOUS ABSORPTION
CONSTANT  DOSEIV  =0.0                   ! IN JECTED DOSE  (UG/KG)
  DOSEIV_NM = DOSEIV/MW   !  CONVERTS  THE INJECTED DOSE TO NMOL/G

  ! INITIAL GUESS OF THE FREE CONCENTRATION IN THE LIGAND  (COMPARTMENT
INDICATED BELOW) ====
CONSTANT  CFLLIO          =       0.0        ! LIVER (NMOL/ML)

  ! BINDING CAPACITY  (AhR)  FOR NON LINEAR BINDING  (COMPARTMENT  INDICATED
BELOW)   (NMOL/ML)
CONSTANT  LIBMAX          =     3.5e-4     !  LIVER   (NMOL/ML),  WANG  ET AL .
1997

!  PROTEIN AFFINITY CONSTANTS ( 1A2 OR  AhR,  COMPARTMENT INDICATED BELOW)
( NMOL/ML )===


                                      E-71

-------
CONSTANT KDLI
1997
CONSTANT KDLI2
2004
 l.Oe-4     ILIVER (AhR)(NMOL/ML), WANG  ET AL.

 2.0e-2     ILIVER (1A2)(NMOL/ML), EMOND ET AL.
!===EXCRETION AND ABSORPTION CONSTANT (OPTIMIZED)
CONSTANT KST              =    0.3   !  GASTRIC RATE CONSTANT  (HR-1),
CONSTANT KABS             =    0.48  !INTESTINAL ABSORPTION CONSTANT  (HR-1)
WANG ET AL. 1997
!  ELIMINATION CONSTANTS
CONSTANT CLURI
  0.09 !  URINARY CLEARANCE  (ML/HR)
!  ==test elimination  variable
constant kelv             =     0.4
CONSTANT  (I/HOUR)
            !  INTERSPECIES VARIABLE ELIMINATION
!  CONSTANT TO DIVIDE  THE  ABSORPTION INTO LYMPHATIC AND PORTAL  FRACTIONS
CONSTANT A                =0.7       !  LYMPHATIC FRACTION, WANG  ET  AL.
1997
!PARTITION COEFFICIENTS  OPTIMIZED
CONSTANT PF               =   400
CONSTANT PRE              =    3
AL.  2000
CONSTANT PLI              =    6

i===PARAMETER  FOR  INDUCTION OF CYP 1A2
           !  ADIPOSE TISSUE/BLOOD
           !  REST OF THE BODY/BLOOD, WANG  ET

           !  LIVER/BLOOD, WANG ET AL.  1997
                          1.0
                       =  1.6
   !  INCLUDE INDUCTION?  (1 = YES, 0 = NO)
  DEGRADATION CONCENTRATION CONSTANT OF  1A2
                           5
                           13
CONSTANT IND_ACTIVE=
CONSTANT CYP1A2_10UTZ
(NMOL/ML)
CONSTANT CYP1A2_1A1 =
CONSTANT CYP1A2_1EC50
CONSTANT CYP1A2_1A2 =
CONSTANT CYP1A2_1KOUT
CONSTANT CYP1A2_1TAU
CONSTANT CYP1A2_1EMAX =  600
(UNITLESS)
CONSTANT HILL         =0.6
EFFECT CONSTANT  (UNITLESS)
    IDIFFUSIONAL PERMEABILITY  FRACTION
CONSTANT PAFF     = 0.12       ! ADIPOSE (UNITLESS), WANG ET AL.
CONSTANT PAREF    =0.03       ! REST OF THE BODY (UNITLESS)
CONSTANT PALIF    =0.35       ! LIVER (UNITLESS)
                         1.5
                       =  0.1
                       =  1.5
!  BASAL CONCENTRATION OF 1A1  (NMOL/ML)
 !  DISSOCIATION CONSTANT TCDD-CYP1A2  (NMOL/ML)
!  BASAL CONCENTRATION OF 1A2  (NMOL/ML)
!  FIRST ORDER RATE OF DEGRADATION  (H-l)
 HOLDING TIME (H)
!  MAXIMUM INDUCTION OVER BASAL EFFECT

   IHILL CONSTANT; COOPERATIVE LIGAND  BINDING
                                                                  2000
     !COMPARTMENT TISSUE  BLOOD VOLUME  =========
CONSTANT WLIO      =   0.0549    !  LIVER,  ILSI 1994
CONSTANT WFO       =   0.069     !  ADIPOSE

     !TISSUE BLOOD  FLOW EXPRESSED AS A FRACTION OF CARDIAC OUTPUT
CONSTANT QFF  = 0.070          !  ADIPOSE TISSUE BLOOD FLOW FRACTION
(UNITLESS), LEUNG  ET AL.  1990
CONSTANT QLIF = 0.161          !  LIVER (UNITLESS)  ILSI ET AL.  1994

     !COMPARTMENT TISSUE  BLOOD EXPRESSED AS A FRACTION OF THE TOTAL
COMPARTMENT VOLUME
CONSTANT WFBO =    0.050       !  ADIPOSE TISSUE, WANG ET AL.  1997
                                      E-72

-------
CONSTANT WREBO =    0.030
CONSTANT WLIBO =    0.266
                    !  REST  OF THE BODY,  WANG ET AL.  1997
                    !  LIVER,  WANG ET AL.  1997
     ! EXPOSURE  SCENARIO  FOR UNIQUE OR REPETITIVE WEEKLY OR MONTHLY  EXPOSURE
     ! NUMBER OF EXPOSURES  PER WEEK
CONSTANT WEEK_LAG    =0.0      !  TIME ELAPSED BEFORE EXPOSURE BEGINS  (WEEK)
CONSTANT WEEK_PERIOD = 168     !  NUMBER OF HOURS IN THE WEEK  (HOURS)
CONSTANT WEEK_FINISH = 120     !  TIME EXPOSURE ENDS (HOURS)

     ! NUMBER OF EXPOSURES  PER MONTH
CONSTANT MONTH_LAG    =0.0     !  DELAY BEFORE EXPOSURE  (MONTH)

     !SET FOR BACKGROUND  EXPOSURE===========
     !CONSTANT FOR BACKGROUND EXPOSURE===========
CONSTANT Day_LAG_BG  =0.0      !  TIME ELAPSED BEFORE EXPOSURE BEGINS  (HOURS)
CONSTANT Day_PERIOD_BG =24   !  LENGTH OF EXPOSURE (HOURS)

     ! NUMBER OF EXPOSURES  PER WEEK
CONSTANT WEEK_LAG_BG  =0.0 !  TIME ELAPSED BEFORE BACKGROUND EXPOSURE  (WEEK)
CONSTANT WEEK_PERIOD_BG  =  168 !NUMBER OF HOURS IN THE WEEK  (HOURS)
CONSTANT WEEK_FINISH_BG  =  168 !  TIME EXPOSURE ENDS (HOURS)

     !GROWTH CONSTANT FOR RAT AND MOUSE
     !CONSTANT FOR MOTHER BODY WEIGHT GROWTH ======
CONSTANT BW_TO  =20           !CHANGED FOR SIMULATION  (IN G)

     !CONSTANT USED IN CARDIAC OUTPUT EQUATION, HADDAD 2001
CONSTANT QCCAR  =275          !CONSTANT (ML/MIN/KG)
     ! COMPARTMENT TOTAL  LIPID FRACTION
CONSTANT F_TOTLIP =
CONSTANT B_TOTLIP =
CONSTANT RE_TOTLIP =
CONSTANT LI TOTLIP =
                      0
            855
          0.0033
          0.019
          0.06
IADIPOSE TISSUE (UNITLESS)
!BLOOD (UNITLESS)
IREST OF THE BODY (UNITLESS)
!LIVER (UNITLESS)
END
       END OF THE  INITIAL  SECTION
DYNAMIC  ! DYNAMIC  SIMULATION SECTION
ALGORITHM
CINTERVAL
MAXTERVAL
MINTERVAL
VARIABLE
CONSTANT
(HOURS)
 CINTXY  = CINT
 PFUNC   = CINT
IALG
CINT
MAXT
MINT
T
TIMELIMIT
       2        IGEAR METHOD
      1.0       !COMMUNICATION INTERVAL
    l.Oe+10     IMAXIMUM CALCULATION INTERVAL
    l.OE-10     IMINIMUM CALCULATION INTERVAL
      0.0       !HOUR
      2904.0       !SIMULATION TIME LIMIT
      ITIME CONVERSION
  DAY         = T/24 . 0
  WEEK        = T/168.0
  MONTH       = T/730.0
  YEAR        = T/8760.0
                       !  TIME  IN DAYS
                       !  TIME  IN WEEKS
                       !  TIME  IN MONTHS
                       !  TIME  IN YEARS
      INMAX =MAX(T,CTFNGKG)
nmax =max(T,CFNGKG)
                                      E-73

-------
DERIVATIVE  ! PORTION  OF CODE THAT SOLVES DIFFERENTIAL  EQUATIONS

      !CHRONIC OR  SUBCHRONIC EXPOSURE SCENARIO =======
      !NUMBER OF EXPOSURES  PER DAY
 DAY_LAG     = EXP_TIME_ON       !  TIME ELAPSED BEFORE EXPOSURE  BEGINS
(HOURS)
 DAY_PERIOD   = DAYJCYCLE          !  EXPOSURE PERIOD  (HOURS)
 DAY_FINISH   = CINTXY            !  LENGTH OF EXPOSURE (HOURS)
 MONTH_PERIOD = TIMELIMIT          !  EXPOSURE PERIOD  (MONTHS)
 MONTH_FINISH = EXP_TIME_OFF      !  LENGTH OF EXPOSURE (MONTHS)

      !NUMBER OF EXPOSURES  PER DAY AND MONTH
 DAY_FINISH_BG    =  CINTXY
 MONTH_LAG_BG   = BCK_TIME_ON    !  TIME ELAPSED BEFORE BACKGROUND EXPOSURE
BEGINS  (MONTHS)
 MONTH_PERIOD_BG  =  TIMELIMIT      !  BACKGROUND EXPOSURE PERIOD  (MONTHS)
 MONTH_FINISH_BG  =  BCK_TIME_OFF   !  LENGTH OF BACKGROUND EXPOSURE (MONTHS)

      !  FRACTION OF  DIOXIN  ABSORBED IN THE PORTAL FRACTION OF THE LIVER
 B =  1.0-A
      !GROWTH UP  EQUATION (G)

PARAMETER  (BW_RMN  =  l.OE-30)
WTO=  (BW_TO *(1.0+(0.41*T)/(1402.5+T+BW_RMN)))  ! IN GRAMS

      ! VARIABILITY OF REST  OF THE BODY DEPENDS ON OTHER  ORGANS
      IREST OF THE  BODY FRACTION;  UPDATED FOR EPA ASSESSMENT
 WREO =  (0.91 -  (WLIBO*WLIO + WFBO*WFO + WLIO + WFO))/(1+WREBO)

      ! REST OF THE BODY BLOOD FLOW FRACTION
 QREF =  1.0-(QFF+QLIF)             IREST OF BODY BLOOD FLOW  (ML/HR)
      !SUMMATION  OF BLOOD FLOW FRACTION (SHOULD BE EQUAL  TO  1)
 QTTQF = QFF+QREF+QLIF          !  SUM MUST EQUAL 1

      !COMPARTMENT  VOLUME (ML OR G)
 WF  =  WFO  * WTO              !  ADIPOSE
 WRE =  WREO * WTO              !  REST OF THE BODY
 WLI =  WLIO * WTO              !  LIVER

      !COMPARTMENT  TISSUE BLOOD (NL OR G )
 WFB  =  WFBO  * WF              !  ADIPOSE
 WREB =  WREBO * WRE            !  REST OF THE BODY
 WLIB =  WLIBO * WLI            !  LIVER

      !CARDIAC OUTPUT FOR THE GIVEN BODY WEIGHT
  QC= QCCAR*60*(WTO/1000.0)**0.75

QF  = QFF*QC          !  ADIPOSE TISSUE BLOOD FLOW RATE (ML/HR)
QLI = QLIF*QC         !  LIVER TISSUE BLOOD FLOW RATE  (ML/HR)
QRE = QREF*QC         !  REST OF THE BODY BLOOD FLOW RATE  (ML/HR)

QTTQ = QF+QRE+QLI    !TOTAL  FLOW RATE  (ML/HR)

      !PERMEABILITY ORGAN FLOW (ML/HR)  =======

                                      E-74

-------
PAF  = PAFF*QF        ! ADIPOSE TISSUE
PARE = PAREF*QRE      ! REST  OF THE BODY
PALI = PALIF*QLI      ! LIVER TISSUE

     IABSORPTION  SECTION
     IORAL
     !BACKGROUND  EXPOSURE
     !EXPOSURE FOR  STEADY  STATE CONSIDERATION
     !REPETITIVE  EXPOSURE  SCENARIO

MSTOT_NMBCKGR = MSTOTBCKGR/322 IAMOUNT IN NMOL/G
MSTTBCKGR =MSTOT_NMBCKGR *WTO

     !REPETITIVE  ORAL  BACKGROUND EXPOSURE SCENARIOS
DAY_EXPOSURE_BG   = PULSE(DAY_LAG_BG,DAY_PERIOD_BG,DAY_FINISH_BG)
WEEK_EXPOSURE_BG  = PULSE(WEEK_LAG_BG,WEEK_PERIOD_BG,WEEK_FINISH_BG)
MONTH_EXPOSURE_BG = PULSE(MONTH_LAG_BG,MONTH_PERIOD_BG,MONTH_FINISH_BG)

MSTTCH_BG =  (DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG)*MSTTBCKGR
MSTTFR_BG = MSTTBCKGR/CINT

totalBG= integ  (MSTTCH_BG,0.0)
CYCLE BG =DAY EXPOSURE BG*WEEK EXPOSURE BG*MONTH EXPOSURE BG
      !CONDITIONAL ORAL  EXPOSURE (BACKGROUND EXPOSURE)
IF  (MSTTCH_BG.EQ.MSTTBCKGR)  THEN
    ABSMSTT_GB= MSTTFR_BG
ELSE
    ABSMSTT_GB =0.0
END IF

      !EXPOSURE +  !REPETITIVE EXPOSURE SCENARIO
IV= DOSEIV_NM * WTO   IAMOUNT IN NMOL
MSTT= MSTOT_NM * WTO  !AMOUNT IN NMOL

DAY_EXPOSURE   = PULSE(DAY_LAG,DAY_PERIOD,DAY_FINISH)
WEEK_EXPOSURE  = PULSE(WEEK_LAG,WEEK_PERIOD,WEEK_FINISH)
MONTH_EXPOSURE = PULSE(MONTH_LAG,MONTH_PERIOD,MONTH_FINISH)

MSTTCH =  (DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE)*MSTT
CYCLE = DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE

SUMEXPEVENT= integ  (CYCLE,0.0)*cint INUMBER OF CYCLES GENERATED  DURING
SIMULATION

MSTTFR = MSTT/CINT

    !  CONDITIONAL ORAL  EXPOSURE
IF  (MSTTCH.EQ.MSTT) THEN
   ABSMSTT= MSTTFR
ELSE
   ABSMSTT =0.0
END IF

 CYCLETOT=INTEG(CYCLE, 0.0)


                                      E-75

-------
     IMASS CHANGE  IN  THE LUMEN
RMSTT= -(KST+KABS)*MST+ABSMSTT +ABSMSTT_GB  ! RATE OF  CHANGE (NMOL/H)
 MST = INTEG(RMSTT,0.0)   IAMOUNT REMAINING  IN DUODENUM (NMOL)

     IABSORPTION IN LYMPH CIRCULATION
LYRMLUM = KABS*MST*A
 LYMLUM = INTEG(LYRMLUM,0.0)

     !ABSORPTION IN PORTAL CIRCULATION
LIRMLUM = KABS*MST*B
  LIMLUM = INTEG(LIRMLUM,0.0)

     !PERCENT OF DOSE REMAINING IN THE GI TRACT
RFECES = KST*MST  + REXCLI
  FECES = INTEG(RFECES,0.0)
prctFECES =  (FECES/(BDOSE_TOTAL+1E-30))*100
     IABSORPTION OF  DIOXIN BY IV ROUTE	
 IVR= IV/PFUNC  !  RATE  FOR IV INFUSION IN BLOOD
 EXPIV= IVR *  (l.O-STEP(PFUNC))
 IVDOSE = integ(EXPIV,0.0)

     !SYSTEMIC BLOOD CONCENTRATION (NMOL/ML)
     ! MODIFICATION  ON  OCTOBER 6,  2009
CB=(QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM)/(QC+CLURI)  !
 CA = CB

     !URINARY EXCRETION BY KIDNEY
     ! MODIFICATION  ON  OCTOBER 6,  2009
RAURI = CLURI *CB
  AURI = INTEG(RAURI, 0.0)

prctAURI =  (AURI/(BDOSE_TOTAL+1E-30))*100
     IUNIT CONVERSION  POST SIMULATION
CBNGKG=CB*MW*UNITCORR
CBSNGKGLIADJ=  (CB*MW*UNITCORR*(1.0/B_TOTLIP)*(1.0/SERBLO))![NG of TCDD
Serum/Kg OF LIPID]
CBPMOL_KG= CB*UNITCORR*UNITCORR       !CONCENTRATION  IN  PMOL/KG
CBNGG = CB*MW
     IADIPOSE TISSUE COMPARTMENT
     !TISSUE BLOOD  SUBCOMPARTMENT
RAFB = QF*(CA-CFB)-PAF*(CFB-CF/PF)         !(NMOL/HR)
 AFB = INTEG(RAFB,0.0)                      !(NMOL)
 CFB = AFB/WFB                             !(NMOL/ML)
     !TISSUE SUBCOMPARTMENT
RAF  = PAF*(CFB-CF/PF)                     !(NMOL/HR)
 AF  = INTEG(RAF,0.0)                       !(NMOL)
 CF  = AF/WF                               !(NMOL/ML)

     !POST SIMULATION  UNIT CONVERSION
CFTOTAL    =  (AF + AFB)/(WF + WFB)  !  TOTAL CONCENTRATION IN  FAT(NM/ML)
CFNGKG  = CFTOTAL*MW*UNITCORR
CFUGG=(CFTOTAL*MW)/UNITCORR

                                      E-76

-------
CFPMOL_KG= CFTOTAL*UNITCORR*UNITCORR
CFNGG = CFTOTAL*MW
           !CONCENTRATION IN  PMOL/KG
     IREST OF THE  BODY COMPARTMENT
     !TISSUE BLOOD SUBCOMPARTMENT
RAREB= QRE*(CA-CREB)-PARE*(CREB-CRE/PRE)
 AREB = INTEG(RAREB,0.0)
 CREB = AREB/WREB
     !TISSUE SUBCOMPARTMENT
RARE = PARE*(CREB -  CRE/PRE)
 ARE = INTEG(RARE,0.0)
 CRE  = ARE/WRE

     IPOST SIMULATION UNIT CONVERSION
CRETOTAL=  (ARE +  AREB)/(WRE + WREB)
STATE
                 !  (NMOL/HR)
                   !(NMOL)
                 !(NMOL/ML)

                 !(NMOL/HR)
                 !(NMOL)
                   !(NMOL/ML)
                 !  CONCENTRATION AT  STEADY
     !LIVER COMPARTMENT
     !TISSUE BLOOD  SUBCOMPARTMENT
RALIB = QLI*(CA-CLIB)-PALI*(CLIB-CFLLIR)+LIRMLUM
  ALIB = INTeg(RALIB,0.0)
 CLIB = ALIB/WLIB
     !TISSUE SUBCOMPARTMENT
 RALI =  PALI*(CLIB-CFLLIR)-REXCLI
  ALI = integ(RALI,0.0)
  CLI  = ALI/WLI
                    !(NMOL/HR)
                     !(NMOL)
                    !(NMOL/HR)
                           !(NMOL)
                     !(NMOL/ML)
     !FREE TCCD  CONCENTRATION IN LIVER (NMOL/ML)
PARAMETER  (LIVER_1RMN  = l.OE-30)
 CFLLI= IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLI  &
       +LIVER_1RMN))+((CYP1A2_103*CFLLIR/(KDLI2+CFLLIR  &
       +LIVER_1RMN)*IND_ACTIVE)))-CFLLI,CFLLIO)
     CFLLIR=DIM(CFLLI,0.0)  !  FREE CONCENTRATION IN LIVER

CBNDLI= LIBMAX*CFLLIR/(KDLI+CFLLIR+LIVER_1RMN)  !BOUND CONCENTRATION

    IPOST SIMULATION UNIT  CONVERSION
CLITOTAL=  (ALI  + ALIB)/(WLI  + WLIB)!
rec_occ_AHR=  (CFLLIR/(KDLI+CFLLIR+1E-30))*100.0   ! PERCENT  OF AhR OCCUPANCY
PROT_occ_lA2=  (CFLLIR/(KDLI2+CFLLIR))*100.0   ! PERCENT  OF 1A2 OCCUPANCY
CLINGKG =(CLITOTAL*MW*UNITCORR)
CBNDLINGKG = CBNDLI*MW*UNITCORR
CLIUGG=(CLITOTAL*MW)/UNITCORR
CLIPMOL_KG= CLITOTAL*UNITCORR*UNITCORR        !CONCENTRATION IN PMOL/KG
CLINGG = CLITOTAL*MW

    !Fraction increase  of  induction of CYP1A2
fold_ind=(CYP1A2_10UT/CYP1A2_1A2)
VARIATIONOfAC =(CYP1A2_10UT-CYP1A2_1A2)/CYP1A2_1A2

    !VARIABLE ELIMINATION  BASED ON THE CYP1A2
KBILE_LI_T =((CYP1A2_10UT-CYP1A2_1A2)/CYP1A2_1A2)*Kelv  !INDUCED BILIARY
EXCRETION RATE  CONSTANT
REXCLI=  (KBILE_LI_T*CFLLIR*WLi;
 EXCLI = INTEG(REXCLI,0.0)
!DOSE-DEPENDENT EXCRETION RATE
                                      E-77

-------
    !CHEMICAL IN CYP450  (1A2)  COMPARTMENT
    !EQUATION FOR  INDUCTION  OF CYP1A2

CYP1A2_1KINP = CYP1A2_1KOUT*  CYP1A2_10UTZ

    ! MODIFICATION ON  OCTOBER  6,  2009
CYP1A2_10UT =INTEG(CYP1A2_1KINP  * (1.0 + CYP1A2_1EMAX *(CBNDLI+1.Oe-30)**HILL
&
     /(CYP1A2_1EC50**HILL +  (CBNDLI+1.Oe-30)**HILL))  &
      - CYP1A2_1KOUT*CYP1A2_10UT, CYP1A2_10UTZ)
!  EQUATIONS INCORPORATING DELAY  OF CYP1A2 PRODUCTION (NOT USED  IN
SIMULATIONS)

CYP1A2_1R02 =  (CYP1A2_10UT  -  CYP1A2_102)/ CYP1A2_1TAU
  CYP1A2_102 =INTEG(CYP1A2_1R02,  CYP1A2_1A1)
CYP1A2_1R03 =  (CYP1A2_102 - CYP1A2_103)/ CYP1A2_1TAU
  CYP1A2_103 =INTEG(CYP1A2_1R03,  CYP1A2_1A2)

       ! MASS BALANCE  CONTROL
 BDOSE= LYMLUM+LIMLUM+IVDOSE
 BMASSE = EXCLI+AURI+AFB+AF+AREB+ARE+ALIB+ALI
 BDIFF = BDOSE-BMASSE
       ! AMOUNT TOTAL  PRESENT  IN  THE GI TRACT
BDOSE_TOTAL =LYMLUM+LIMLUM+FECES

       IBODY BURDEN  IN NG
 Body_burden =(AFB+AF+AREB+ARE+ALIB+ALI)*MW

       IBODY BURDEN  CONCENTRATION  (NG/KG)
 BBNGKG =(((AFB+AF+AREB+ARE+ALIB+ALI)*MW)/(WTO/UNITCORR))   !

       !COMMAND FOR  END  OF SIMULATION
TERMT  (T.GE. TimeLimit,  'Time limit has been reached.')

END    ! END OF THE  DERIVATIVE SECTION
END    ! END OF THE  DYNAMIC  SECTION
END    ! END OF PROGRAM
E.2.5.2. Input Files

E.2.5.2.1.  Delia Porta (1987) female
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% Delia Porta 1987 for  female mice.
%dose levels:  2.5 and  5 ug/kg/week  for 52 weeks
%dose levels:  2500 and 5000  ng/kg/week for 52 weeks
%dose levels equivalent to:  357 and  714 ng/kg-d
MAXT =  0.01
CINT  =0.1
EXP_TIME_ON
EXP TIME OFF
= 0.
= 8736
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)

  E-78

-------
DAY_CYCLE         =  168
BCK_TIME_ON       =  0.
BCK_TIME_OFF      =  0.
TIMELIMIT         =  8736
BW_TO             =  20
SIMULATION  (G)
               %HOURS BETWEEN  DOSES
               %TIME BACKGROUND  EXPOSURE BEGINS (HOUR)
               %TIME BACKGROUND  EXPOSURE ENDS (HOUR)
               %SIMULATION  DURATION  (HOUR)
               %BODY WEIGHT AT THE BEGINNING OF THE
%EXPOSURE DOSE SCENARIOS  (UG/KG)
    %MSTOT           =  2.5
    MSTOT         =5.0
               %EXPOSURE DOSE UG/KG
               %EXPOSURE DOSE UG/KG
E.2.5.2.2.  Delia Porta (1987) male
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% Delia Porta 1987  for  male  mice.
%dose levels:  2.5  and  5 ug/kg/week for 52 weeks
%dose levels:  2500 and 5000 ng/kg/week for 52 weeks
%dose levels equivalent to:  357  and 714 ng/kg-d
MAXT =  0.01
CINT  =0.1
EXP_TIME_ON       =  0.
EXP_TIME_OFF      =  8736
DAYJCYCLE         =  168
BCK_TIME_ON       =  0.
BCK_TIME_OFF      =  0.
TIMELIMIT         =  8736
BW_TO             =26
SIMULATION  (G)
               %TIME EXPOSURE  BEGINS  (HOUR)
               %TIME EXPOSURE  ENDS  (HOUR)
               %HOURS BETWEEN  DOSES
               %TIME BACKGROUND  EXPOSURE BEGINS (HOUR)
               %TIME BACKGROUND  EXPOSURE ENDS (HOUR)
               %SIMULATION  DURATION  (HOUR)
               %BODY WEIGHT AT THE BEGINNING OF THE
%EXPOSURE DOSE SCENARIOS  (UG/KG)
    %MSTOT           =2.5
    MSTOT         =5.0
               %EXPOSURE DOSE UG/KG
               %EXPOSURE DOSE UG/KG
E.2.5.2.3.  Ishiham et al (2007)
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% Ishihara 2007
%dose levels:  1)  2  ng/kg loading;  0.4 ng/kg weekly
              %2)  2,000 ng/kg loading; 400 ng/kg weekly
 MAXT
 CINT
 TIMELIMIT
 EXP_TIME_ON
 EXP_TIME_OFF
 DAY CYCLE
0.01
0.1
840
168
840
168
%SIMULATION DURATION  (HOUR)
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
                                      E-79

-------
 BCK_TIME_ON      =  0.
 BCK_TIME_OFF     =0.02
  BW_TO            = 23
SIMULATION  (G)

%EXPOSURE DOSE  SCENARIOS  (UG/KG)
   %MSTOTBCKGR       = 0.002
   %MSTOT            = 0.0004
   MSTOTBCKGR   = 2
   MSTOT        =0.4
                %TIME BACKGROUND EXPOSURE BEGINS  (HOUR)
                %TIME BACKGROUND EXPOSURE ENDS  (HOUR)
                %BODY WEIGHT AT  THE BEGINNING OF THE
                %INITIAL  LOADING EXPOSURE DOSE [UG/KG]
                %EXPOSURE DOSE  [UG/KG]
                %INITIAL  LOADING EXPOSURE DOSE [UG/KG]
                %EXPOSURE DOSE  [UG/KG]
E.2.5.2.4.  Kuchiiwa et al (2002)
% Kuchiiwa 2002
%protocol:  oral  exposure  once weekly for 8 weeks
%dose levels:  0.0049,  0.490   ug/kg once weekly for
                                   weeks
 MAXT = 0.01
 CINT  =0.1
 TIMELIMIT       =  1344
 EXP_TIME_ON     =  0.
 EXP_TIME_OFF    =  1344
 DAYJCYCLE       =  168
 BCK_TIME_ON     =  0.
 BCK_TIME_OFF    =0.0
 BW_TO           =  25
SIMULATION  (g)

%EXPOSURE DOSE  SCENARIOS  (UG/KG)
  %MSTOT          = 0.0049
MSTOT       = 0.490
E.2.5.2.5.  NTP (1982) female, chronic
                %SIMULATION  DURATION (HOUR)
                %TIME EXPOSURE  BEGINS (HOUR)
                %TIME EXPOSURE  ENDS (HOUR)
                %HOURS BETWEEN  DOSES
                %TIME BACKGROUND EXPOSURE BEGINS  (HOUR)
                %TIME BACKGROUND EXPOSURE ENDS  (HOUR)
                %BODY WEIGHT AT THE BEGINNING OF THE
                %EXPOSURE  DOSE  [UG/KG]
                %EXPOSURE  DOSE  [UG/KG]
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% NTP 1982.
%protocol: twice weekly  gavage for 104 weeks
%dose levels:   0.02,  0.1,  1  ug/kg twice weekly for 104 weeks
%dose levels:   20,  100,  1000 ng/kg twice weekly for 104 weeks
%dose levels equivalent  to:  5.71,  28.57, 285.1 ng/kg-d
MAXT =  0.01
CINT  =0.1
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
BW_TO
SIMULATION  (G)
0.
17472
84
0.
0.
17472
23
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%SIMULATION DURATION  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
                                      E-80

-------
    %MSTOT
    %MSTOT
    MSTOT
  = 0.02
 = 0.1
= 1.0
%EXPOSURE DOSE UG/KG
%EXPOSURE DOSE UG/KG
%EXPOSURE DOSE UG/KG
E.2.5.2.6.  NTP (1982) male, chronic
output @clear
prepare @clear
prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% NTP 1982.
%protocol: twice weekly  gavage for 104 weeks
%dose levels:   0.005,  0.025,  0.25  ug/kg twice weekly for 104 weeks
%dose levels:   5, 25,  250  ng/kg twice weekly for 104 weeks
%dose levels equivalent  to:  1.4,  7.1, 71 ng/kg-d
MAXT =  0.01
CINT  =0.1
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
BW_TO
SIMULATION  (G)
  0.
  17472
  84
  0.
  0.
  17472
  25
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%SIMULATION DURATION  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE
%EXPOSURE DOSE SCENARIOS  (UG/KG)
    %MSTOT           =  0.005
    %MSTOT         = 0.025
    MSTOT         =  0.25
                   %EXPOSURE  DOSE  UG/KG
                   %EXPOSURE  DOSE  UG/KG
                   %EXPOSURE  DOSE  UG/KG
E.2.5.2.7.  Nohara et al (2002)
%Nohara 2002
%protocol:  single  oral  exposure dose
%dose levels:   0.005,  0.020,  0.100 and 0.500 ug/kg single dose
%dose levels equivalent  5,  20,  100 and 500 ng/kg single dose
MAXT = 0.01
CINT  =0.1
TIMELIMIT        =  24
EXP_TIME_ON      =  0.
EXP_TIME_OFF     =24
DAYJCYCLE        =24
WEEK_FINISH      =  193
BCK_TIME_ON      =  0.
BCK_TIME_OFF     =  0.
BW_TO            =  23
SIMULATION  (G)

%EXPOSURE DOSE SCENARIOS  (UG/KG)
      %MSTOT         =  0.005
      %MSTOT      = 0.020
                   %SIMULATION  DURATION (HOUR)
                   %TIME  EXPOSURE  BEGINS (HOUR)
                   %TIME  EXPOSURE  ENDS (HOUR)
                   %HOURS  BETWEEN  DOSES
                   %LAST  HOUR WHEN DOSE OCCURS  (HOUR)
                   %TIME  BACKGROUND EXPOSURE BEGINS  (HOUR)
                   %TIME  BACKGROUND EXPOSURE ENDS  (HOUR)
                   %BODY  WEIGHT AT THE BEGINNING OF THE
                   %EXPOSURE  DOSE  UG/KG
                   %EXPOSURE  DOSE  UG/KG
                                      E-81

-------
      %MSTOT
      MSTOT
 = 0.100
= 0.500
%EXPOSURE DOSE UG/KG
%EXPOSURE DOSE UG/KG
E.2.5.2.8.  Smialowicz et al (2004)
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG  CBNGKG

% Smialowicz et  al.  2004.
MAXT
CINT
TIMELIMIT
EXP TIME ON
EXP TIME OFF
DAY CYCLE
BCK TIME ON
BCK TIME OFF
BW TO
SIMULATION (G)
%EXPOSURE DOSE
%MSTOT
%MSTOT
%MSTOT
%MSTOT
%MSTOT
= 0.01
=
=
=
=
=
=
=
=

0.1
24.
0.
24.
24
0.
0.
25










SCENARIOS

0
0
1
3
MSTOT = 10
= 0.
.1
.3
.0
.0
.0
03





                          (UG/KG)
                                      %SIMULATION DURATION  (HOUR)
                                      %TIME EXPOSURE  BEGINS  (HOUR)
                                      %TIME EXPOSURE  ENDS  (HOUR)
                                      %HOURS BETWEEN  DOSES
                                      %TIME BACKGROUND  EXPOSURE BEGINS (HOUR)
                                      %TIME BACKGROUND  EXPOSURE ENDS (HOUR)
                                      %BODY WEIGHT AT THE BEGINNING OF THE
                                      %EXPOSURE DOSE  (UG/KG)
                                      %EXPOSURE DOSE  (UG/KG)
                                      %EXPOSURE DOSE  (UG/KG)
                                      %EXPOSURE DOSE  (UG/KG)
                                      %EXPOSURE DOSE  (UG/KG)
                                      %EXPOSURE DOSE  (UG/KG)
E.2.5.2.9.  Smialowicz et al (2008)
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG  CBNGKG

% Smialowicz et  al.  2008.
%protocol:  oral gavage  5  days/week for 13 weeks
%dose levels:  0,  0.0015,  0.015,  0.15, 0.45 ug/kg
%dose levels:  0,  1.5,  15,  150,  450 nkd (0, 1.07, 10.7,  107,  321 nkd adj;
MAXT          =  0.01
CINT          =0.1
TIMELIMIT     =  2184
EXP_TIME_ON   =  0.
EXP_TIME_OFF  =  2184
DAYJCYCLE     =   24
WEEK_PERIOD   =  168
WEEK_FINISH   =  119
BCK_TIME_ON   =  0.
BCK_TIME_OFF  =  0.
BW_TO         =28
SIMULATION  (G)

%EXPOSURE DOSE SCENARIOS  (UG/KG)
   %MSTOT      = 0.0015
                     %SIMULATION DURATION  (HOUR)
                     %TIME  EXPOSURE BEGINS  (HOUR)
                     %TIME  EXPOSURE ENDS  (HOUR)
                     %HOURS BETWEEN DOSES
                     %HOURS IN A WEEK
                     %LAST  HOUR IN WEEK WHERE DOSE OCCURS
                     %TIME  BACKGROUND EXPOSURE BEGINS  (HOUR)
                     %TIME  BACKGROUND EXPOSURE ENDS  (HOUR)
                     %BODY  WEIGHT AT THE BEGINNING OF THE
                     %EXPOSURE DOSE (UG/KG)
                                      E-82

-------
   %MSTOT   =  0.015
   %MSTOT   =  0.150
   MSTOT   = 0.450
%EXPOSURE DOSE  (UG/KG)
%EXPOSURE DOSE  (UG/KG)
%EXPOSURE DOSE  (UG/KG)
E.2.5.2.10. Toth et al (1979) 1 year
output @clear
prepare @clear
prepare T CLINGKG CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

% Toth et al. 1979
%protocol:  weekly  gavage for 1 year
%dose levels:   7, 700,  7000  ng/kg once weekly for 52 weeks  (1 year)
%dose levels:   0.007,  0.7,  7 ug/kg once weekly for 52 weeks  (1  year)
%dose equivalent: 1,  100,  1000 ng/kg-day
MAXT         =  0.01
CINT         =0.1
TIMELIMIT    =  8760
EXP_TIME_ON  =  0.
EXP_TIME_OFF =  8760
DAY_CYCLE    =   168
BCK_TIME_ON  =  0.
BCK_TIME_OFF =  0.
BW_TO        =  27
SIMULATION  (G)
%SIMULATION DURATION  (HOUR)
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS(HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
    %MSTOT    = 0.007
    %MSTOT =0.7
    MSTOT = 7
%EXPOSURE DOSE  (UG/KG)
%EXPOSURE DOSE  (UG/KG)
%EXPOSURE DOSE  (UG/KG)
E.2.5.2.11. Weber et al (1995)
output @clear
prepare @clear
prepare T CLINGKG  CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG

%Weber et al. 1995  C57  strain
%protocol:   single  oral exposure dose
MAXT = 0.01
CINT  =0.1
TIMELIMIT        =  24
EXP_TIME_ON      =  0.
EXP_TIME_OFF     =24
DAYJCYCLE        =24
BCK_TIME_ON      =  0.
BCK_TIME_OFF     =  0.
BW_TO            =24.1
SIMULATION  (G)

%EXPOSURE DOSE  SCENARIOS  (UG/KG)
      %MSTOT        =0.03
%SIMULATION DURATION  (HOUR)
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE
%EXPOSURE DOSE UG/KG
                                      E-83

-------
%MSTOT
%MSTOT
%MSTOT
%MSTOT
%MSTOT
%MSTOT
%MSTOT
%MSTOT
%MSTOT
%MSTOT
MSTOT
= 0.1
= 0.3
= 1.0
= 3.0
= 9.4
= 37.5
= 75.0
= 100.0
= 133.0
= 150.0
= 235.0
%EXPOSURE
%EXPOSURE
%EXPOSURE
%EXPOSURE
%EXPOSURE
%EXPOSURE
%EXPOSURE
%EXPOSURE
%EXPOSURE
%EXPOSURE
%EXPOSURE
DOSE
DOSE
DOSE
DOSE
DOSE
DOSE
DOSE
DOSE
DOSE
DOSE
DOSE
UG/KG
UG/KG
UG/KG
UG/KG
UG/KG
UG/KG
UG/KG
UG/KG
UG/KG
UG/KG
UG/KG
E.2.5.2.12. White et al (1986)
output @clear
prepare @clear
prepare T CLINGKG CFNGKG  CBSNGKGLIADJ BBNGKG CBNDLINGKG

% White et al 1986
%protocol:  oral exposure  single  dose
%dose levels: 10, 50,  100,  500,  1000,  2000 ng /kg-d ug/kg I/day for  14
consecutive days
%dose levels: 0.010,  0.050,  0.100,  0.500,  1.0,  2.0 ug /kg-d ug/kg I/day  for
14 consecutive days
MAXT         = 0.01
CINT         =0.1
TIMELIMIT    = 336
EXP_TIME_ON  = 0.
EXP_TIME_OFF = 336
DAY_CYCLE    =  24
BCK_TIME_ON  = 0.
BCK_TIME_OFF = 0.
BW_TO        = 23
SIMULATION  (G)

%EXPOSURE DOSE SCENARIOS  (UG/KG)
   %MSTOT = 0.010
   %MSTOT = 0.050
   %MSTOT = 0.100
   %MSTOT = 0.500
   %MSTOT =  1
   MSTOT =  2
%SIMULATION DURATION  (HOUR)
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS  (HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%BODY WEIGHT AT THE BEGINNING OF  THE
%EXPOSURE
%EXPOSURE
%EXPOSURE
%EXPOSURE
%EXPOSURE
%EXPOSURE
DOSE IN UG/KG
DOSE IN UG/KG
DOSE IN UG/KG
DOSE IN UG/KG
DOSE IN UG/KG
DOSE IN UG/KG
E.2.6.  Mouse Gestational Model

E.2.6.1. Model Code

PROGRAM: Three Compartment PBPK Model for TCDD in Mice (Gestation)'

INITIAL   !
      !SIMULATION PARAMETERS  ====
CONSTANT PARA_ZERO           =  1E-30
CONSTANT EXP TIME ON      =288.
 !  TIME AT WHICH EXPOSURE BEGINS  (HOURS)

E-84

-------
CONSTANT EXP_TIME_OFF     = 504
CONSTANT DAYJCYCLE        =  504.
CONSTANT BCK_TIME_ON      =0.0
BEGINS  (HOURS)
CONSTANT BCK_TIME_OFF     =0.0
(HOURS)
CONSTANT TRANSTIME_ON     = 144
AT GESTATIONAL  DAY  6
  !  TIME AT WHICH EXPOSURE  ENDS  (HOURS)
   !  NUMBER OF HOURS BETWEEN  DOSES  (HOURS)
   !  TIME AT WHICH BACKGROUND EXPOSURE

   !  TIME AT WHICH BACKGROUND EXPOSURE ENDS

   !CONTROL TRANSFER FROM MOTHER TO FETUS
     IUNIT CONVERSION
CONSTANT MW=322  ! MOLECULAR WEIGHT
CONSTANT SERBLO  =0.55
CONSTANT UNITCORR = 1000
 (NG/NMOL)
     !INTRAVENOUS  SEQUENCY
constant IV_LAG          =0.0
constant IV PERIOD        =0.0
     !PREGNANCY  PARAMETER ====
CONSTANT CONCEPTION_T          =0.0
CONSTANT N FETUS          =10
        ITIME OF CONCEPTION  (HOUR)
  !NUMBER OF FETUS PRESENT
     !CONSTANT EXPOSURE  CONTROL ===========
     IACUTE, SUBCHRONIC,  CHRONIC EXPOSURE =====
     !OR BACKGROUND  EXPOSURE (IN THIS CASE 3 TIMES A DAY)===
CONSTANT MSTOTBCKGR      =0.0       !  ORAL BACKGROUND  EXPOSURE  DOSE (UG/KG)
CONSTANT MSTOT            =0.0       !  ORAL EXPOSURE DOSE  (UG/KG)
     IORAL ABSORPTION
  MSTOT NM = MSTOT/MW
 !CONVERTS THE DOSE TO NMOL/G
     ! INTRAVENOUS ABSORPTION
CONSTANT  DOSEIV          =0.0
  DOSEIV_NM = DOSEIV/MW
CONSTANT DOSEIVLATE  =0.0
  DOSEIVNMlate = DOSEIVLATE/MW
   !  INJECTED DOSE  (UG/KG)
 !  CONVERTS THE INJECTED  DOSE  TO  NMOL/G
   !  INJECTED DOSE LATE  (UG/KG)
 !AMOUNT IN NMOL/G
     !INITIAL GUESS  OF  THE FREE CONCENTRATION IN THE LIGAND  (COMPARTMENT
INDICATED BELOW)====
CONSTANT CFLLIO           =0.0  !LIVER    (NMOL/ML)
CONSTANT CFLPLAO          =0.0  !PLACENTA (NMOL/ML)

     !BINDING CAPACITY  (AhR)  FOR NON LINEAR BINDING  (COMPARTMENT  INDICATED
BELOW)  (NMOL/ML)  ===
CONSTANT LIBMAX           = 3.5E-4    !  LIVER  (NMOL/ML), WANG  ET  AL.  1997
CONSTANT PLABMAX          = 2.OE-4    !TEMPORARY PARAMETER
     ! PROTEIN AFFINITY CONSTANTS
 (NMOL/ML)===
CONSTANT KDLI             = l.OE-4
CONSTANT KDLI2            = 4.OE-2
CONSTANT KDPLA            = l.OE-4
1A2 OR AhR, COMPARTMENT  INDICATED  BELOW)

  ILIVER  (AhR)  (NMOL/ML), WANG ET  AL.  1997
  ILIVER  (1A2)  (NMOL/ML), EMOND ET AL.  2004
  !TEMPORARY PARAMETER  (AhR)
     !EXCRETION AND ABSORPTION CONSTANT
CONSTANT KST              =    0.3   !  GASTRIC RATE CONSTANT  (HR-1)
CONSTANT KABS             =    0.48  !INTESTINAL ABSORPTION CONSTANT  (HR-1)
WANG ET AL. 1997
                                      E-85

-------
!  ELIMINATION CONSTANTS
CONSTANT CLURI
                                 0.09 !  URINARY CLEARANCE  (ML/HR)
    ITEST
constant
CONSTANT
         ELIMINATION
         kelv
          (I/HOUR)
                     VARIABLE
0.4
             INTERSPECIES VARIABLE ELIMINATION
     ! CONSTANT TO  DIVIDE  THE ABSORPTION INTO LYMPHATIC AND  PORTAL  FRACTIONS
CONSTANT A                =0.7        !  LYMPHATIC FRACTION, WANG ET  AL.  1997
     !PARTITION COEFFICIENTS
CONSTANT PF               = 400
CONSTANT PRE              = 3
CONSTANT PLI              = 6
CONSTANT PPLA             = 3
                                   !  ADIPOSE TISSUE/BLOOD
                                   !  REST OF THE BODY/BLOOD, WANG ET  AL.  2000
                                  !  LIVER/BLOOD, WANG ET AL. 1997
                                   !  TEMPORARY PARAMETER NOT CONFIGURED
     !PARAMETER  FOR  INDUCTION OF CYP
CONSTANT IND_ACTIVE       = 1
CONSTANT CYP1A2_10UTZ     = 1.6
1A2  (NMOL/ML)  (OPTIMIZED)
CONSTANT CYP1A2_1A1      =1.5
WANG ET AL  .  (2000)
CONSTANT CYP1A2_1EC50     =0.13
(NMOL/ML)
CONSTANT CYP1A2_1A2      =1.5
(NMOL/ML),WANG  ET AL.  (2000)
CONSTANT CYP1A2_1KOUT     =0.1
CONSTANT CYP1A2_1TAU      =1.5
.   (2000)
CONSTANT CYP1A2_1EMAX     = 600
(UNITLESS)
CONSTANT HILL             =0.6
BINDING EFFECT  CONSTANT  (UNITLESS)
                                     1A2, WANG ET AL. 1997 OR  OPTIMIZED
                                      !  INCLUDE INDUCTION?  (1  =  YES,  0  = NO)
                                     !  DEGRADATION CONCENTRATION CONSTANT OF

                                     !  BASAL CONCENTRATION OF  1A1 (NMOL/ML),

                                      !  DISSOCIATION CONSTANT  TCDD-CYP1A2

                                     !BASAL CONCENTRATION OF 1A2

                                     !  FIRST ORDER RATE OF DEGRADATION  (H-l)
                                     IHOLDING TIME  (H)  (OPTIMIZED), WANG ET  AL

                                     !  MAXIMUM INDUCTION OVER  BASAL EFFECT

                                     IHILL CONSTANT; COOPERATIVELY LIGAND
     IDIFFUSIONAL  PERMEABILITY FRACTION,  WANG ET AL. 1997
CONSTANT PAFF
2000
CONSTANT PAREF
CONSTANT PALIF
CONSTANT PAPLAF
                          =0.12    IADIPOSE (UNITLESS) OPTIMIZED,  WANG ET AL.

                          =0.03    IREST OF THE BODY  (UNITLESS)
                          =0.35    !LIVER (UNITLESS)
                          =0.03    !TEMPORARY PARAMETER NOT  CONFIGURED
    !FRACTION OF TISSUE  WEIGHT =========
CONSTANT WLIO             = 0.0549   !LIVER  ILSI  (1994)

    !TISSUE BLOOD  FLOW EXPRESSED AS A FRACTION OF CARDIAC OUTPUT  CONSTANT QFF
= 0.070     ! ADIPOSE TISSUE BLOOD FLOW FRACTION  (UNITLESS),  LEUNG  ET  AL.  1990
CONSTANT QLIF             = 0.161    !LIVER (UNITLESS), ILSI  1994

    !COMPARTMENT TISSUE  BLOOD EXPRESSED AS A FRACTION OF THE  TOTAL  COMPARTMENT
VOLUME
CONSTANT WFBO
CONSTANT WREBO
CONSTANT WLIBO
CONSTANT WPLABO
                            0.050    IADIPOSE TISSUE, WANG ET AL.  1997
                            0.030    IREST OF THE BODY, WANG ET AL.  1997
                            0.266    !LIVER, WANG ET AL.  1997
                            0.500    !TEMPORARY PARAMETER NOT CONFIGURED
    !EXPOSURE SCENARIO  FOR UNIQUE OR REPETITIVE WEEKLY OR MONTHLY  EXPOSURE
                                      E-86

-------
    !NUMBER OF EXPOSURES  PER WEEK
CONSTANT WEEK_LAG        =0.0
(WEEK)
CONSTANT WEEK_PERIOD      = 168
CONSTANT WEEK_FINISH      = 168

    !NUMBER OF EXPOSURES  PER MONTH
CONSTANT MONTH_LAG       =0.0
(MONTH)
                          ITIME ELAPSED BEFORE EXPOSURE BEGINS

                          !  NUMBER OF HOURS IN THE WEEK  (HOURS)
                          !  TIME EXPOSURE ENDS (HOURS)


                          ITIME ELAPSED BEFORE EXPOSURE BEGINS
    !CONSTANT FOR  BACKGROUND EXPOSURE=
CONSTANT Day_LAG_BG      =0.0
(HOUR)
CONSTANT Day_PERIOD_BG    =24
    !NUMBER OF EXPOSURES  PER WEEK
CONSTANT WEEK_LAG_BG       =0.0
(WEEK)
CONSTANT WEEK_PERIOD_BG     = 168
CONSTANT WEEK_FINISH_BG     = 168

    !INITIAL BODY WEIGHT
CONSTANT BW_TO              =30
CONSTANT RATIO_RATF_MOUSEF =0.2
GESTATIONAL DAY 22
                          !  TIME ELAPSED BEFORE EXPOSURE BEGINS

                          !LENGTH OF EXPOSURE  (HOUR)


                           ITIME ELAPSED BEFORE BACKGROUND EXPOSURE

                            !  NUMBER OF HOURS IN THE WEEK  (HOURS)
                            ITIME EXPOSURE ENDS (HOURS)
                           !  WANG ET AL.  1997 (IN G)
                              !RATIO OF FETUS MOUSE/RAT AT

                              !  FOR RAT (1)  AND FOR MOUSE  (0.2)
    !COMPARTMENT TOTAL  LIPID FRACTION ,  POULIN ET AL. 2000
CONSTANT F_TOTLIP           = 0.855
CONSTANT B_TOTLIP           = 0.0033
CONSTANT RE_TOTLIP          = 0.019
(UNITLESS)
CONSTANT LI_TOTLIP          = 0.060
CONSTANT PLA_TOTLIP         = 0.019
CONSTANT FETUS TOTLIP       = 0.019
                                       !  ADIPOSE TISSUE  (UNITLESS)
                                        !  BLOOD  (UNITLESS)
                                       !  REST OF THE BODY

                                       !  LIVER  (UNITLESS)
                                       !  PLACENTA  (UNITLESS)
                                    !  FETUS  (UNITLESS)
END
         ! END OF  THE  INITIAL SECTION
DYNAMIC  ! DYNAMIC  SIMULATION SECTION
ALGORITHM
CINTERVAL
MAXTERVAL
MINTERVAL
VARIABLE
CONSTANT
 CINTXY  = CINT
 PFUNC   = CINT
IALG
CINT
MAXT
MINT
T
TIMELIMIT
   2
  0.1
l.Oe+10
l.OE-10
  0.0
  313
  GEAR METHOD
  COMMUNICATION INTERVAL
  MAXIMUM CALCULATION INTERVAL
!  MINIMUM CALCULATION INTERVAL

!SIMULATION LIMIT TIME  (HOUR)
    ITIME CONVERSION
  DAY         = T/24
  WEEK        = T/168
  MONTH       = T/730
  YEAR        = T/8760
                     !  TIME  IN DAYS
                     !  TIME  IN WEEKS
                     !  TIME  IN MONTHS
                     !  TIME  IN YEARS
DERIVATIVE  ! PORTION  OF CODE THAT SOLVES DIFFERENTIAL EQUATIONS

    !CHRONIC OR  SUBCHRONIC EXPOSURE SCENARIO =======
                                      E-87

-------
    !NUMBER OF EXPOSURES  PER DAY
 DAY_LAG          =  EXP_TIME_ON    !  TIME ELAPSED BEFORE EXPOSURE  BEGINS
(HOURS)
 DAY_PERIOD       = DAY_CYCLE      !  EXPOSURE PERIOD  (HOURS)
 DAY_FINISH       = CINTXY         !  LENGTH OF EXPOSURE  (HOURS)
 MONTH_PERIOD     = TIMELIMIT      !  EXPOSURE PERIOD  (MONTHS)
 MONTH_FINISH     = EXP_TIME_OFF   !  LENGTH OF EXPOSURE  (MONTHS)

    !NUMBER OF EXPOSURES  PER DAY AND MONTH
 DAY_FINISH_BG    = CINTXY
 MONTH_LAG_BG     =  BCK_TIME_ON    ITIME ELAPSED BEFORE BACKGROUND EXPOSURE
BEGINS  (MONTHS)
 MONTH_PERIOD_BG  = TIMELIMIT      !BACKGROUND EXPOSURE PERIOD  (MONTHS)
 MONTH_FINISH_BG  = BCK_TIME_OFF   !LENGTH OF BACKGROUND  EXPOSURE (MONTHS)

    !INTRAVENOUS LATE
 IV_FINISH = CINTXY
 B = 1-A  ! FRACTION OF DIOXIN ABSORBED IN THE PORTAL  FRACTION OF  THE LIVER
!FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUM
E
  !  FROM OFLAHERTY_1992

RTESTGEST= T-CONCEPTION_T
TESTGEST=DIM(RTESTGEST,0.0)

WTFER_RODENT=  (2.3d-3*EXP(1.49d-2*(TESTGEST))+1.3d-2)*Gest_on
WTFER =  (WTFER_RODENT*RATIO_RATF_MOUSEF*N_FETUS)
WTFE = DIM(WTFER,0.0)

  i
FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME
  !  FAT GROWTH EXPRESSION LINEAR DURING PREGNANCY
  !  FROM 0'FLAHERTY_1992

WFO=  (((9.66d-5*(TESTGEST))*gest_on)+0.069)

  !  PLACENTA,VOLUME,  PLACENTA,VOLUME,  PLACENTA,VOLUME,  PLACENTA,VOLUME
  !  WPLA PLACENTA GROWTH  EXPRESSION, SINGLE EXPONENTIAL WITH OFFSET
  !  FROM 0'FLAHERTY_1992   !  FOR EACH PUP

WPLAON_RODENT  =  (0.6/(1+(5d+3*EXP(-0.0225*(TESTGEST)))))*N_FETUS
WPLAOR =  (WPLAON_RODENT/WTO)*Gest_on
WPLAO = DIM(WPLAOR,0.0)

  !  PLACENTA,FLOW RATE,  PLACENTA,FLOW RATE, PLACENTA,FLOW  RATE, PLACENTA,FLOW
RATE
  !  QPLA PLACENTA GROWTH  EXPRESSION, DOUBLE EXPONENTIAL WITH OFFSET
  !  FROM 0'FLAHERTY_1992

 QPLARF =  (1.67d-7  *exp(9.6d-3*(TESTGEST))  &
   +1.6d-3*exp(7.9d-3*(TESTGEST))+0.0)*Gest_on*SWITCH_trans
 QPLAF=DIM(QPLARF,0.0)                !FRACTION OF FLOW  RATE IN  PLACENTA

  !  GESTATION  CONTROL
IF  (T.LT.CONCEPTION_T)  THEN

                                      E-88

-------
    Gest_off =  1
    Gest_on=    0.0
ELSE
    Gest_off =0.0
    Gest_on  =  1
END IF

  !  MOTHER BODY WEIGHT  GROWTH EQUATION========
  !  MODIFICATION TO ADAPT  THIS MODEL AT HUMAN MODEL
  !  BECAUSE LINEAR DESCRIPTION IS NOT GOOD ENOUGH FOR MOTHER GROWTH
  !  MOTHER BODY WEIGHT  GROWTH

  PARAMETER  (BW_RMN =  l.OE-30)
  WTO= BW_TO * (1.0+ (0.41*T)/ (1402. 5+T+BW_RMN) )   ! IN GRAMS

  !  VARIABILITY OF REST OF THE BODY DEPENDS ON OTHER ORGANS
  WREO =  (0.91  -  (WLIBO*WLIO  + WFBO*WFO +WPLABO*WPLAO + WLIO + WFO +
WPLAO) ) / (1. 0+WREBO)   !  REST OF THE BODY FRACTION; UPDATED FOR EPA ASSESSMENT
  QREF =  1.0- (QFF+QLIF+QPLAF)             ! REST OF BODY BLOOD FLOW RATE
FRACTION
  QTTQF = QFF+QREF+QLIF+QPLAF          !  SUM MUST EQUAL 1
  !  COMPARTMENT VOLUME
 WF  =  WFO  * WTO
 WRE =  WREO * WTO
 WLI =  WLIO * WTO
 WPLA=  WPLAO* WTO
                        (ML  OR G)  =========
                                       !  ADIPOSE TISSUE
                                       !  REST OF THE BODY
                                       !  LIVER
                                       !  PLACENTA
    ! COMPARTMENT TISSUE  BLOOD (ML OR G)  =========
 WFB  =  WFBO  * WF                    !  ADIPOSE TISSUE
 WREB =  WREBO * WRE                   !  REST OF THE BODY
 WLIB =  WLIBO * WLI                   !  LIVER
 WPLAB = WPLABO* WPLA                 !  PLACANTA

    ! CARDIAC OUTPUT FOR  THE GIVEN BODY WEIGHT
    !QC= QCCAR*60* (WTO/1000.0) **0. 75
CONSTANT QCC=16500                     !  EQUIVALENT TO 275 *  60
QC= QCC* (WTO/UNITCORR) **0.75
    ! COMPARTMENT BLOOD  FLOW RATE (ML/HR)
QF  = QFF*QC
QLI = QLIF*QC
QRE = QREF*QC
QPLA = QPLAF*QC
QTTQ = QF+QRE+QLI+QPLA
                                       IADIPOSE TISSUE BLOOD FLOW  RATE
                                       ! LIVER TISSUE BLOOD FLOW  RATE
                                       ! REST OF THE BODY BLOOD FLOW RATE
                                       ! PLACENTA TISSUE BLOOD FLOW RATE
                          ! TOTAL FLOW RATE
     ! PERMEABILITY ORGAN  FLOW (ML/HR) =========
PAF  = PAFF*QF                        !  ADIPOSE TISSUE
PARE = PAREF*QRE                      !  REST OF THE BODY
PALI = PALIF*QLI                      !  LIVER TISSUE
PAPLA = PAPLAF*QPLA                   !  PLACENTA
     ! ABSORPTION  SECTION
     ! ORAL,
     ! INTRAPERITONEAL,
     ! INTRAVENOUS
                                      E-89

-------
     !REPETITIVE ORAL  BACKGROUND EXPOSURE SCENARIO

MSTOT_NMBCKGR = MSTOTBCKGR/322        IAMOUNT IN NMOL/G
MSTTBCKGR =MSTOT_NMBCKGR *WTO

DAY_EXPOSURE_BG   = PULSE(DAY_LAG_BG,DAY_PERIOD_BG,DAY_FINISH_BG)
WEEK_EXPOSURE_BG  = PULSE(WEEK_LAG_BG,WEEK_PERIOD_BG,WEEK_FINISH_BG)
MONTH_EXPOSURE_BG = PULSE(MONTH_LAG_BG,MONTH_PERIOD_BG,MONTH_FINISH_BG)

MSTTCH_BG =  (DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG)*MSTTBCKGR
MSTTFR_BG = MSTTBCKGR/CINT

CYCLE_BG =DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG

     ! CONDITIONAL ORAL  EXPOSURE (BACKGROUND EXPOSURE)

IF  (MSTTCH_BG.EQ.MSTTBCKGR)  THEN
    ABSMSTT_GB= MSTTFR_BG
ELSE
    ABSMSTT_GB =0.0
END IF

CYCLETOTBG=INTEG(CYCLE_BG,0.0)

    !REPETITIVE ORAL EXPOSURE SCENARIO

MSTT= MSTOT_NM * WTO                   !AMOUNT IN NMOL

DAY_EXPOSURE   = PULSE(DAY_LAG,DAY_PERIOD,DAY_FINISH)
WEEK_EXPOSURE  = PULSE(WEEK_LAG,WEEK_PERIOD,WEEK_FINISH)
MONTH_EXPOSURE = PULSE(MONTH_LAG,MONTH_PERIOD,MONTH_FINISH)

MSTTCH =  (DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE)*MSTT
MSTTFR = MSTT/CINT

CYCLE = DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE
SUMEXPEVENT= INTEG  (CYCLE,0.0)/cint !NUMBER OF CYCLES GENERATED  DURING
SIMULATION

    ! CONDITIONAL ORAL EXPOSURE
IF  (MSTTCH.EQ.MSTT) THEN
   ABSMSTT= MSTTFR
ELSE
   ABSMSTT =0.0
END IF
 CYCLETOT=INTEG(CYCLE, 0.0)

    ! MASS CHANGE  IN  THE  LUMEN
 RMSTT= -(KST+KABS)*MST  +ABSMSTT +ABSMSTT_GB !  RATE OF CHANGE  (NMOL/H)
  MST = INTEG(RMSTT,0.0)                       !AMOUNT REMAINING  IN  DUODENUM
(NMOL)

    ! ABSORPTION IN LYMPH CIRCULATION

                                      E-90

-------
 LYRMLUM = KABS*MST*A
  LYMLUM = INTEG(LYRMLUM,0.0)

   !  ABSORPTION IN  PORTAL CIRCULATION
 LIRMLUM = KABS*MST*B
  LIMLUM = INTEG(LIRMLUM,0.0)
i  	IV EXPOSURE  	

 IV= DOSEIV_NM * WTO  IAMOUNT  IN NMOL
 IVR= IV/PFUNC  ! RATE  FOR IV  INFUSION IN BLOOD
 EXPIV= IVR *  (1.0-STEP(PFUNC))
 IVDOSE = integ(EXPIV,0.0)

    !	IV late  in  the cycle
    !  MODIFICATION  ON  January 13 2004
 IV_RlateR = DOSEIVNMlate*WTO
 IV_EXPOSURE=PULSE(IV_LAG,IV_PERIOD, IV_FINISH)

 IV_lateT = IV_EXPOSURE  *IV_RlateR
 IV_late = IV_lateT/CINT

SUMEXPEVENTIV= integ  (IV_EXPOSURE, 0.0)  !NUMBER OF CYCLES GENERATED DURING
SIMULATION

    !SYSTEMIC CONCENTRATION OF TCDD
    !  MODIFICATION  ON  OCTOBER 6, 2009
 CB=(QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM+QPLA*CPLAB+IV_late)/(QC+CLURI)  !
 CA = CB !  CONCENTRATION (NMOL/ML)

    !URINARY EXCRETION BY KIDNEY
    MODIFICATION ON OCTOBER  6,  2009
RAURI = CLURI *CB
  AURI = INTEG(RAURI,0.0)

  IUNIT CONVERSION  POST  SIMULATION
 CBSNGKGLIADJ=(CB*MW*UNITCORR*(1/B_TOTLIP)*(1/SERBLO))![NG  of  TCDD Serum/Kg
OF LIPID]
   AUCBS_NGKGLIADJ=integ(CBSNGKGLIADJ,0.0)
  CBNGKG= CB*MW*UNITCORR
  CBNGG = CB*MW

   !ADIPOSE COMPARTMENT
   !TISSUE BLOOD COMPARTMENT
RAFB= QF*(CA-CFB)-PAF*(CFB-CF/PF)     !(NMOL/H)
 AFB = INTEG(RAFB,0.0)                 !(NMOL)
  CFB = AFB/WFB                        !(NMOL/ML)
   !TISSUE COMPARTMENT
RAF = PAF*(CFB-CF/PF)                 !(NMOL/H)
 AF = INTEG(RAF,0.0)                  !(NMOL)
  CF  = AF/WF                          !(NMOL/ML)

   IUNIT CONVERSION  POST SIMULATION
  CFTOTAL=  (AF + AFB)/(WF + WFB)  !  TOTAL CONCENTRATION IN NMOL/ML

                                      E-91

-------
  CFTFREE = CFB  +  CF !TOTAL FREE CONCENTRATION IN  FAT   (NM/ML)

  CFNGKG=CFTOTAL*MW*UNITCORR !  FAT CONCENTRATION IN  NG/KG
    AUCF_NGKGH=integ(CFNGKG, 0.0)
  CFNGG = CFTOTAL*MW

   IREST OF THE  BODY COMPARTMENT
RAREB= QRE *(CA-CREB)-PARE*(CREB-CRE/PRE)   !(NMOL/H)
 AREB = INTEG(RAREB,0.0)                      !(NMOL)
  CREB = AREB/WREB                           !(NMOL/H)
   !TISSUE COMPARTMENT
RARE = PARE*(CREB  -  CRE/PRE)                 !(NMOL/H)
 ARE = INTEG(RARE,0.0)                       !(NMOL)
  CRE  = ARE/WRE                             !(NMOL/ML)

   IUNIT CONVERSION  POST  SIMULATION
  CRETOTAL=  (ARE + AREB)/(WRE + WREB)            !  TOTAL  CONCENTRATION IN
NMOL/ML
      CRENGKG=CRETOTAL*MW*UNITCORR !  REST OF THE BODY  CONCENTRATION IN NG/KG
   !LIVER COMPARTMENT
   !TISSUE BLOOD  COMPARTMENT
 RALIB = QLI*(CA-CLIB)-PALI*(CLIB-CFLLIR)+LIRMLUM  !
  ALIB = INTEG(RALIB,0.0)                         !(NMOL)
   CLIB = ALIB/WLIB                              !(NMOL/ML)
   !TISSUE COMPARTMENT
 RALI = PALI*(CLIB -  CFLLIR)-REXCLI             !   (NMOL/HR)
  ALI = INTEG(RALI,0.0)                               !(NMOL)
   CLI  = ALI/WLI                                  !(NMOL/ML)

   IFREE TCDD  IN  LIVER  COMPARTMENT
PARAMETER  (LIVER_1RMN = l.OE-30)
 CFLLI= IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLIR &
        +LIVER_1RMN))+((CYP1A2_103*CFLLIR/(KDLI2 +  CFLLIR &
        +LIVER_1RMN)*IND_ACTIVE)))-CFLLI,CFLLIO)
     CFLLIR=DIM(CFLLI,0.0)   ! FREE CONCENTRATION  IN  LIVER

  CBNDLI= LIBMAX*CFLLIR/(KDLI+CFLLIR+LIVER_1RMN)  !BOUND  CONCENTRATION

  !VARIABLE ELIMINATION BASED ON THE CYP1A2
 KBILE_LI_T =((CYP1A2_10UT-CYP1A2_1A2)/CYPlA2_lA2)*Kelv  !  INDUCED BILIARY
EXCRETION RATE CONSTANT
  REXCLI = KBILE_LI_T*CFLLIR*WLI !  DOSE-DEPENDENT EXCRETION  RATE
    EXCLI = INTEG(REXCLI,0.0)

  IUNIT CONVERSION POST SIMULATION
  CLITOTAL=  (ALI  + ALIB)/(WLI +  WLIB) !  TOTAL CONCENTRATION  IN NMOL/ML

  Rec_occ= CFLLIR/(KDLI+CFLLIR)
  CLINGKG=CLITOTAL*MW*UNITCORR !  LIVER CONCENTRATION  IN  NG/KG
     AUCLI_NGKGH=INTEG(CLINGKG, 0.0)
  CBNDLINGKG = CBNDLI*MW*UNITCORR
     AUCBNDLI_NGKGH  =INTEG(CBNDLINGKG,0.0)
  CLINGG = CLITOTAL*MW

   !CHEMICAL IN CYP450  (1A2) COMPARTMENT

                                      E-92

-------
CYP1A2_1KINP = CYP1A2_1KOUT*  CYP1A2_10UTZ !  BASAL RATE OF CYP1A2  PRODUCTION
SET EQUAL TO BASAL  RATE  OF DEGREDATION

   !  MODIFICATION ON  OCTOBER  6,  2009
CYP1A2_10UT =INTEG(CYP1A2_1KINP  * (1.0 + CYP1A2_1EMAX *(CBNDLI+1.Oe-30)**HILL
&
     /(CYP1A2_1EC50**HILL  + (CBNDLI+1.Oe-30)**HILL))  &
      - CYP1A2_1KOUT*CYP1A2_10UT, CYP1A2_10UTZ)

!  EQUATIONS INCORPORATING  DELAY  OF CYP1A2 PRODUCTION (NOT USED  IN
SIMULATIONS)

      CYP1A2_1R02 = (CYP1A2_10UT - CYP1A2_102)/ CYP1A2_1TAU
  CYP1A2_102 =INTEG(CYP1A2_1R02,  CYP1A2_1A1)

CYP1A2_1R03 =  (CYP1A2_102  - CYP1A2_103)/ CYP1A2_1TAU
  CYP1A2_103 =INTEG(CYP1A2_1R03,  CYP1A2_1A2)

!  TRANSFER OF DIOXIN  FROM  PLACENTA TO FETUS
!  FETAL EXPOSURE ONLY DURING  EXPOSURE

IF (T.LT.TRANSTIME_ON) THEN
 SWITCH_trans =0.0
ELSE
 SWITCH_trans = 1
END IF

!TRANSFER OF DIOXIN FROM PLACENTA TO FETUS
!  MODIFICATION 26 SEPTEMBER 2003

CONSTANT PFETUS= 4  !
CONSTANT CLPLA_FET  =0.17  !

RAMPF = (CLPLA_FET*CPLA) *SWITCH_trans
  AMPF=INTEG(RAMPF, 0.0)

!TRANSFER OF DIOXIN FROM FETUS TO PLACENTA
RAFPM = (CLPLA_FET*CFETUS_v)*SWITCH_trans !
  AFPM = INTEG(RAFPM,0.0)
!  TCDD IN PLACENTA MOTHER COMPARTMENT
RAPLAB= QPLAMCA - CPLAB)-PAPLA* (CPLAB -CFLPLAR)
 APLAB = INTEG(RAPLAB,0.0)
 CPLAB = APLAB/(WPLAB+1E-30)
RAPLA = PAPLA*(CPLAB-CFLPLAR)-RAMPF + RAFPM
 APLA = INTEG(RAPLA,0.0)
 CPLA  = APLA/(WPLA+le-30)
!  NMOL/H)
 !  (NMOL)
!  (NMOL/ML)
!  (NMOL/H)
 !  (NMOL)
!  (NMOL/ML)
PARAMETER  (PARA_ZERO  =  l.OE-30)
CFLPLA= IMPLC(CPLA-(CFLPLAR*PPLA +(PLABMAX*CFLPLAR/(KDPLA&
    +CFLPLAR+PARA_ZERO)))-CFLPLA,CFLPLAO)
CFLPLAR=DIM(CFLPLA,0.0)

   IUNIT CONVERSION POST  SIMULATION
  CPLATOTAL=  (APLA  +  APLAB)/((WPLA + WPLAB)+le-30)! TOTAL CONCENTRATION IN
NMOL/ML
                                      E-93

-------
  CPLANGG = CPLATOTAL*MW

    !FETUS COMPARTMENT
RAFETUS= RAMPF-RAFPM
 AFETUS = INTEG(RAFETUS ,0.0)
CFETUS=AFETUS/(WTFE+1E-30)
CFETOTAL= CFETUS
CFETUS v = CFETUS/PFETUS
   ! UNIT CONVERSION  POST SIMULATION
CFETUSNGKG = CFETUS*MW*UNITCORR
AUC_FENGKGH =  INTEG(CFETUSNGKG,0.0)
CFETUSNGG = CFETOTAL*MW
                !(NG/KG)
i  	CONTROL  MASS BALANCE	
BDOSE= IVDOSE +LYMLUM+LIMLUM
BMASSE = EXCLI+AURI+AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB+AFETUS
BDIFF = BDOSE-BMASSE

      IBODY BURDEN  (NG)
BODY_BURDEN = AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB  !
BBFETUSNG     = AFETUS*MW*UNITCORR    !  NG
      !  BODY BURDEN IN TERMS OF CONCENTRATION  (NG/KG)
 BBNGKG =(((AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB)/WTO)*MW*UNITCORR)  !
  AUC BBNGKGH=INTEG(BBNGKG,0.0)
i  	COMMAND  OF THE END OF SIMULATION	
TERMT  (T.GE. TimeLimit,  'Time limit has been reached.')
END    ! END OF THE  DERIVATIVE SECTION
END    ! END OF THE  DYNAMIC SECTION
END    ! END OF THE  PROGRAM
E.2.6.2. Input Files

E.2.6.2.1.  Keller et al (2007)
output @clear
prepare gclear  T  CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH
AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH
CBNGKG AUC CBNGKGH
%Keller et al.  2007
%protocol:  single oral  dose at GDIS
%dose levels:  0.01,  0.100 1 ug/kg at GDIS
%dose levels:    10,  100  1000 ng/kg at GDIS
 MAXT=0.01
 CINT =0.1
 EXP_TIME_ON      =312.
 EXP_TIME_OFF     =  336
 DAYJCYCLE        =24
 BCK_TIME_ON      =  0.
 BCK_TIME_OFF     =  0.
 TIMELIMIT        =  336
%TIME EXPOSURE BEGINS  (HOUR)
%TIME EXPOSURE ENDS  (HOUR)
%HOURS BETWEEN DOSES
%TIME BACKGROUND EXPOSURE  BEGINS (HOUR)
%TIME BACKGROUND EXPOSURE  ENDS  (HOUR)
%SIMULATION DURATION  (HOUR)
                                      E-94

-------
 BW_TO
SIMULATION  (G)
 CONCEPTION_T
 TRANSTIME_ON
 N FETUS
 = 24

      = 0.
 = 144.
 = 10
%EXPOSURE DOSE  SCENARIOS  (UG/KG)
   %MSTOT
   %MSTOT
   MSTOT
  = 0.01
 = 0.1
= 1
%BODY WEIGHT AT THE BEGINNING OF THE

%TIME OF CONCEPTION  (HOUR)
%TIME OF CONCEPTION +  6  DAYS(144 HOURS)
%NUMBER OF FETUSES
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
E.2.6.2.2.  Li et al (2006)
output @clear
prepare gclear T  CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH
AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG  AUCBNDLI_NGKGH
CBNGKG AUC_CBNGKGH
%Li et al.2006
%protocol:  daily oral  dose from GDI to GD3
%dose levels: 0.002,  0.050, 0.10 ug/kg-day at GDI to GD3
%dose levels: 2,  50,  100 ng/kg-day from GDI to GD3
 MAXT=0.001
 CINT =0.1
 EXP_TIME_ON      =  0.
 EXP_TIME_OFF     =  72
 DAYJCYCLE        =24
 BCK_TIME_ON      =  0.
 BCK_TIME_OFF     =  0.
 TIMELIMIT        =  72.
 BW_TO            =  27
SIMULATION  (G)
 CONCEPTION_T          = 0.
 TRANSTIME_ON     =  144.
 N_FETUS          =10

%EXPOSURE DOSE  SCENARIOS (UG/KG)
   %MSTOT
   %MSTOT
   MSTOT
  = 0.002
 = 0.05
= 0.10
                      %TIME  EXPOSURE BEGINS (HOUR)
                      %TIME  EXPOSURE ENDS  (HOUR)
                      %HOURS BETWEEN DOSES
                      %TIME  BACKGROUND EXPOSURE BEGINS  (HOUR)
                      %TIME  BACKGROUND EXPOSURE ENDS  (HOUR)
                      %SIMULATION DURATION (HOUR)
                      %BODY  WEIGHT AT THE BEGINNING OF THE

                      %TIME  OF CONCEPTION  (HOUR)
                      %TIME  OF CONCEPATION + 6 DAYS(144 HOURS)
                      %NUMBER OF FETUSES
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
E.2.6.2.3.  Smith et al (1976)

      output  @clear
      prepare  @clear  T  CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG  CFETUSNGKG
AUCLI_NGKGH AUCF_NGKGH  AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH  CBNDLINGKG
AUCBNDLI_NGKGH CBNGKG AUC_CBNGKGH

 %protocol:  daily oral dose  from GD6 to GDIS

 %EXPOSURES SCENARIOS
  MAXT=0.01
  CINT =0.1
                                      E-95

-------
  EXP_TIME_ON     =120.
  EXP_TIME_OFF    = 337.
  DAY_CYCLE        = 24
  BCK_TIME_ON     = 0.
  BCK_TIME_OFF    = 0.
  TIMELIMIT        = 360.
  BW_TO            =28.5
SIMULATION (G)
  CONCEPTION_T    = 0.
  TRANSTIME_ON    = 144.
  N_FETUS          =10

 %EXPOSURE DOSE SCENARIOS  (UG/KG)
    %MSTOT
    %MSTOT
    %MSTOT
 = 0.001
= 0.01
= 0.10
                    %TIME EXPOSURE  BEGINS (HOUR)
                    %TIME EXPOSURE  ENDS (HOUR)
                    %HOURS BETWEEN  DOSES
                    %TIME BACKGROUND  EXPOSURE BEGINS (HOUR)
                    %TIME BACKGROUND  EXPOSURE ENDS  (HOUR)
                    %SIMULATION DURATION (HOUR)
                    %BODY WEIGHT AT THE BEGINNING OF THE

                    %TIME OF CONCEPTION (HOUR)
                    %TIME OF CONCEPTION + 6 DAYS(144 HOURS)
                    %NUMBER OF FETUSES
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
%ORAL EXPOSURE DOSE  (UG/KG)
E.3.  TOXICOKINETIC MODELING RESULTS FOR KEY ANIMAL BIOASSAY
     STUDIES
       The simulated TCDD serum-adjusted lipid concentrations reported in this appendix for
the rodent bioassays were converted to TCDD concentrations in rodent whole blood. Initially,
EPA multiplied the serum-adjusted lipid concentrations by 0.0033, the ratio of lipid content to
total  serum volume, then by 0.55, the value of the hematocrit. This product yields the TCDD
concentration in whole rodent blood as predicted by the PBPK model. EPA assumed that the
same whole blood TCDD concentration would result in the same effects in humans and rodents.
       This conversion accomplishes the following:
   1.  Allows the human equivalent dose to be based on equivalent blood concentration (that
      represents serum plus erythrocyte TCDD), which is proportional to tissue exposure;

   2.  Avoids criticism that the total blood concentration is normalized to serum lipid alone in
      an unbalanced way (thus EPA does not contradict Centers for Disease Control and
      Prevention data or methods);

   3.  Factors out any impact of the lipid content used in the PBPK model; and

   4.  TCDD concentration in whole blood is encouraged for use in the assessments by the
      National Academy of Sciences (2006, p. 43); see additional information in Section 3.3.
                                        E-96

-------
E.3.1.   Nongestational Studies



E.3.1.1. Cantonietal (1981)
Type:
Strain:
Jody weight:
Sex:
Rat
CD-COBS rats
BW = 125 g
7emale
Dose:
iRoute:
iRegime:
[Simulation time:
10, 100, and 1,000 ng/kg-week
Oral gavage exposure
1 dose/week for 45 weeks
7,560 hours (45 weeks)
BW = body weight.
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
1.43
14.29
142.86
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.85
-
8.84
-
50.0
-
Max.
3.70 (@ 7,392 hours)
-
26.6 (@ 7,392 hours)
-
227 (@ 7,392 hours)
-
Terminal
1.82
-
7.97
-
41.9
-
LIVER CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
1.43
14.29
142.86
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average

382
2,176
3,973
20,500
39,955
Max.
328 (@ 7,398 hours)
431
2,860 (@ 7,231 hours)
4,330
26,978 (@ 7,399 hours)
43,329
Terminal

431
1,928
4,330
17,255
43,329
FAT CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
1.43
14.29
142.86
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
175
256
837
1,237
4,741
10,278
Max.
200 (@ 7,431 hours)
280
937 (@ 7,427 hours)
1,352
5,374 (@ 7,424 hours)
11,224
Terminal
181
244
807
1,167
4,349
9,734
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
1.43
14.29
142.86
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
26.1
32.4
170
230
1,337
2,154
Max.
3 1. 7 (@ 7,3 98 hours)
35.0
2 10 (@ 7,230 hours)
243
1,695 (@ 7,398 hours)
2,266
Terminal
26.3
35.0
156
243
1,151
2,266
                                         E-97

-------
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
1.43
14.29
142.86
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
6.04
-
23.7
-
66.8
-
Max.
7.76 (@ 7,396 hours)
-
29. 1(@ 7,228 hours)
-
80.0 (@ 1 hours)
-
Terminal
6.01
-
22.2
-
63.4
-
Max = maximum; CADM = concentration- and age-dependent elimination model.
E.3.1.2.  Chu et al(2007; 2001)
Type:
Strain:
Jody weight:
Sex:
Rat
Sprague -Dawley
200 g
Female
Dose:
Route:
Regime:
Simulation time:
2.5, 25, 250, and 1,000 ng/kg-day
Oral exposure
1 dose per day for 28 days
672 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
2.5
25
250
1,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.26

7.66

48.8
-
169
-
Max.
2.35 (@ 648 hours)

15.3 (@ 648 hours)

113 (@ 648 hours)
-
4 18 (@ 648 hours)
-
Terminal
1.88

10.4

63.7
-
222
-
LIVER CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
2.5
25
250
1,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
148
337
1,777
4,422
19,232
45,872
77,819
184,076
Max.
268 (@ 652 hours)
505
2,953 (@ 653 hours)
5,786
30,262 (@ 653 hours)
58,681
120,400 (@ 653 hours)
234,992
Terminal
255
505
2,806
5,786
28,668
58,681
113,890
234,992
FA T CON CENTRA TIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
2.5
25
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
108
295
660
1,703
Max.
1 80 (@ 668 hours)
362
1,020 (@ 659 hours)
2,057
Terminal
180
362
1,015
2,057
                                            E-98

-------
250
1,000
Emond
CADM
Emond
CADM
4,210
14,899
14,576
58,824
6,433 (@ 655 hours)
18,210
22,6 10 (@ 655 hours)
72,002
6,354
18,210
22,280
72,002
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
2.5
25
250
1,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
16.1
30.0
138
261
1,239
2,544
4,801
10,150
Max.
27.5 (@ 652 hours)
40.9
222 (@ 652 hours)
336
1,935 (@ 652 hours)
3,243
7,444 (@ 652 hours)
12,930
Terminal
26.9
40.9
214
336
1,842
3,243
7,067
12,930
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
2.5
25
250
1,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
4.15
-
20.5
-
63.3
-
90.2
-
Max.
6.51 (@ 652 hours)
-
28.5 (@ 652 hours)
-
76.0 (@ 652 hours)
-
99.0 (@ 653 hours)
-
Terminal
6.21
-
27.4
-
74.7
-
98.3
-
E.3.1.3. Croftonetal (2005)
Type:
Strain:
Body weight:
Sex:
Rats
Long Evans
BW = 190 g (4 weeks old)
Female
Dose:
Route:
Regime:
Simulation time:
0,0.1,3, 10,30, 100,300, 1,000,
3,000, and 10,000 ng/kg-day
Oral exposure
One dose per day for 4 days
96 hours
The CADM model was not run because the dosing duration is lower than the resolution of the model (1 week).
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.1
3
10
30
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.0202
-
0.488
-
1.38
-
3.46
-
Max.
0.04 1(@ 72 hours)
-
1. 10 (@ 72 hours)
-
3. 40 (@ 72 hours)
-
9.44 (@ 72 hours)
-
Terminal
0.0244
-
0.582
-
1.62
-
3.93
-
                                              E-99

-------
100
300
1,000
3,000
10,000
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
9.26
-
23.1
-
65.7
-
181
-
583
-
29.0 (@ 72 hours)
-
8 1. 8 (@ 72 hours)
-
260 (@ 72 hours)
-
764 (@ 72 hours)
-
2,527 (@ 72 hours)
-
10.2
-
24.5
-
68.2
-
187
-
607
-
LIVER CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.1
3
10
30
100
300
1,000
3,000
10,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.919
-
37.4
-
145
-
494
-
1,839
-
5,925
-
20,717
-
63,511
-
212,890
-
Max.
1. 55 (@ 75 hours)
-
62.6 (@ 76 hours)
-
242 (@ 77 hours)
-
8 18 (@ 78 hours)
-
3,025 (@ 78 hours)
-
9,692 (@ 78 hours)
-
33,738 (@ 79 hours)
-
103, 140 (@ 79 hours)
-
344,9 10 (@ 79 hours)
-
Terminal
1.18
-
53.3
-
214
-
742
-
2,793
-
9,028
-
31,564
-
96,545
-
321,960
-
FA T CONCENTRA TIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.1
3
10
30
100
300
1,000
3,000
10,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.00
-
24.6
-
70.3
-
177
-
480
-
1,206
-
3,452
-
9,522
-
30,657
-
Max.
1. 93 (@ 96 hours)
-
45.9 (@ 96 hours)
-
129 (@ 96 hours)
-
3 17 (@ 96 hours)
-
838 (@ 96 hours)
-
2,065 (@ 96 hours)
-
5,836 (@ 96 hours)
-
16,050 (@ 96 hours)
-
5 1,9 18 (@ 96 hours)
-
Terminal
1.93
-
45.9
-
129
-
317
-
838
-
2,065
-
5,836
-
16,050
-
51,918
-
E-100

-------
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.1
3
10
30
100
300
1,000
3,000
10,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.138
-
4.04
-
13.3
-
39.3
-
129
-
384
-
1,270
-
3,793
-
12,595
-
Max.
0.224 (@ 79 hours)
-
6.56 (@ 78 hours)
-
2 1. 5 (@ 78 hours)
-
63. 5 (@ 78 hours)
-
208 (@ 78 hours)
-
6 18 (@ 77 hours)
-
2,04 1(@ 77 hours)
-
6,094 (@ 77 hours)
-
20,226 (@ 77 hours)
-
Terminal
0.223
-
6.44
-
21.0
-
61.5
-
200
-
590
-
1,942
-
5,784
-
19,154
-
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.1
3
10
30
100
300
1,000
3,000
10,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0
-
2
-
4
-
10
-
22
-
41
-
68
-
90
-
104
-
Max.
0. 1 15 (@ 75 hours)
-
2.47 (@ 76 hours)
-
6.42 (@ 76 hours)
-
14. 1(@ 76 hours)
-
29. 9 (@ 76 hours)
-
5 1. 9 (@ 77 hours)
-
80.2 (@ 1 hours)
-
98.6 (@ 1 hours)
-
108 (@ 1 hours)
-
Terminal
0
-
2
-
5
-
12
-
27
-
49
-
77
-
96
-
107
-
E-101

-------
E.3.1.4. Croutchetal (2005)
Type:
Strain:
Body weight:
Sex:
Rat
Sprague-Dawley
250 g
7emale
Dose:
Route:
Regime:
Simulation time:
12.5, 50, 200, 800, and 3,200 ng/kg initial
and 1.25, 5, 20, 80, and 320 ng/kg
maintenance doses every 4 days (equivalent
to 0.85, 3.4, 13.6, 54.3, and 217 ng/kg-day)
Gavage
One initial dose and maintenance doses
every 3 days for 28 days
672 hours
The CADM model was not run because the dosing protocol includes both initial and maintenance doses, which is
not supported in the model.
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.85
3.4
13.6
54.3
217
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.340
-
1.10
-
3.29
-
9.58
-
28.7
-
Max
0.723 (@ 648 hours)
-
2.44 (@ 648 hours)
-
8.69 (@ 0 hours)
-
34.8 (@0 hours)
-
1 39 (@0 hours)
-
Terminal
0.513
-
1.55
-
4.36
-
12.1
-
35.0
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.85
3.4
13.6
54.3
217
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
25.6
-
119
-
538
-
2,339
-
9,824
-
Max
46.8 (@ 653 hours)
-
206 (@ 654 hours)
-
877 (@ 654 hours)
-
3,617 (@ 655 hours)
-
14,634 (@ 655 hours)
-
Terminal
43.9
-
195
-
834
-
3,444
-
13,931
-
FA T CONCENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.85
Model
Emond
CADM
Metric
Time-weighted average
29.0
-
Max
46. 9 (@ 672 hours)
-
Terminal
46.9
-
                                             E-102

-------
3.4
13.6
54.3
217
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
94.1
-
284
-
828
-
2,480
-
143 (@ 672 hours)
-
409 (@ 672 hours)
-
1, 149 (@ 670 hours)
-
3, 3 89 (@ 666 hours)
-
143
-
409
-
1,149
-
3,384
-
BODY BURDEN (tig/kg)
Dose
(ng/kg-day)
adjusted dose
0.85
3.4
13.6
54.3
217
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
3.67
-
13.5
-
48.9
-
178
-
661
-
Max
6.09 (@ 654 hours)
-
21.6 (@ 653 hours)
-
75.0 (@ 653 hours)
-
264 (@ 653 hours)
-
963 (@ 653 hours)
-
Terminal
6.00
-
21.1
-
72.8
-
254
-
922
-
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.85
3.4
13.6
54.3
217
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.17
-
3.65
-
10.1
-
24.7
-
50.5
-
Max
1. 93 (@ 652 hours)
-
5.59 (@ 652 hours)
-
14.4 (@ 652 hours)
-
35.8 (@ 1 hour)
-
69.9 (@ 1 hour)
-
Terminal
1.77
-
5.18
-
13.4
-
30.6
-
58.6
-
E.3.1.5.  Delia Porta et al (1987) Female
Type:
Strain:
Body weight:
Sex:
VIouse
B6C3
BW = 20 g (6 weeks
old)
Female
Dose:
Route:
Regime:
Simulation time:
2,500 and 5,000 ng/kg-week (equivalent
to 357 and 714 ng/kg-day)
Gavage
Once a week for 52 weeks
8,736 hours
The CADM model was not run because the study duration is longer than the allowed model duration.
                                            E-103

-------
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
357
714
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
67.0
-
37.6
-
Max.
74 1(@ 8,568 hours)
-
374 (@ 8,568 hours)
-
Terminal
46.8
-
27.2
-
LIVER CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
357
714
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
50,269
-
25,422
-
Max.
70,070 (@ 8,577 hours)
-
35,352 (@ 8,577 hours)
-
Terminal
37,389
-
19,105
-
FA T CONCENTRA TIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
357
714
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
25,235
-
14,162
-
Max.
28,559 (@ 8,589 hours)
-
15,914 (@ 8,590 hours)
-
Terminal
22,498
-
12,810
-
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
357
714
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
5,473
-
2,878
-
Max.
7,247 (@ 8,574 hours)
-
3,774 (@ 8,574 hours)
-
Terminal
4,335
-
2,318
-
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
357
714
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
71.5
-
56.4
-
Max.
99.1(@2hours)
-
88.6 (@ 2 hours)
-
Terminal
65.4
-
50.4
-
E.3.1.6. Delia Porta et al  (1987) Male
Type:
Strain:
Jody weight:
Sex:
VIouse
B6C3
26 g (6 weeks old)
Vlale
Dose:
Route:
Regime:
Simulation time:
2,500 and 5,000 ng/kg-week (equivalent
to 357 and 714 ng/kg-day)
Gavage
Once a week for 52 weeks
8,736 hours
The CADM model was not run because the study duration is longer than the allowed model duration.
                                            E-104

-------
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
357
714
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
67.8
-
38.0
-
Max.
787 (@ 8,568 hours)
-
398 (@ 8,568 hours)
-
Terminal
47.0
-
27.3
-
LIVER CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
357
714
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
50,397
-
25,493
-
Max.
70,052 (@ 8,577 hours)
-
35,347 (@ 8,577 hours)
-
Terminal
37,483
-
19,155
-
FA T CONCENTRA TIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
357
714
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
25,516
-
14,306
-
Max.
28,851 (@ 8,589 hours)
-
16,061 (@ 8,590 hours)
-
Terminal
22,861
-
12,999
-
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
357
714
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
5,504
-
2,894
-
Max.
7,282 (@ 8,574 hours)
-
3, 79 1(@ 8,574 hours)
-
Terminal
4,368
-
2,335
-
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
357
714
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
71.6
-
56.4
-
Max.
99.2 (@ 2 hours)
-
88.6 (@ 2 hours)
-
Terminal
65.4
-
50.4
-
E.3.1.7. Fattoreetal (2000)
Type:
Strain:
Body weight:
Sex:
Rat
Sprague-Dawley
BW 1 50 g (7 weeks old)
Female and male
Dose:
Route:
Regime:
Simulation time:
20, 200, 2,000 ng/kg-day
Dietary exposure
Every day for 13 weeks
2, 184 hours
                                         E-105

-------
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
20
200
2,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
9.59
-
57.6
-
476
-
Max.
15.0 (@ 2,160 hours)
-
102 (@ 2, 160 hours)
-
903 (@ 2, 160 hours)
-
Terminal
11.1
-
63.9
-
522
-
LIVER CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
20
200
2,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
2,448
4,815
24,136
48,824
234,170
488,957
Max.
3,228 (@ 2, 164 hours)
5,639
30,245 (@ 2, 164 hours)
56,499
288,020 (@ 2, 164 hours)
565,103
Terminal
3,078
5,639
28,709
56,499
272,590
565,103
FA T CONCENTRA TIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
20
200
2,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
890
1,663
5,355
14,378
44,176
141,356
Max.
1, 113 (@ 2,166 hours)
1,796
6,542 (@ 2, 165 hours)
15,604
54,246 (@ 2, 165 hours)
153,534
Terminal
1,101
1,756
6,430
15,292
53,140
150,516
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
20
200
2,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
187
281
1,556
2,688
14,432
26,746
Max.
242 (@ 2, 164 hours)
324
1,940 (@ 2, 164 hours)
3,084
17,797 (@ 2, 164 hours)
30,674
Terminal
233
324
1,850
3,084
16,891
30,674
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
20
200
2,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
24.9
-
69.4
-
104
-
Max.
29.8 (@ 2, 164 hours)
-
76.0 (@ 2, 164 hours)
-
106 (@ 2, 164 hours)
-
Terminal
28.8
-
74.7
-
106
-
E-106

-------
E.3.1.8. Foxetal (1993)
Type:
Strain:
Body weight:
Sex:
Rat
Sprague-Dawley
200 g (12 weeks old)
Vlale and Female
Dose:
Route:
Regime:
Simulation time:
5, 2,500, and 12,000 ng/kg initial and 0.9,
600, or 3,500 ng/kg maintenance doses
every 4 days (equivalent to 0.55, 307, and
l,607ng/kg-day)
Gavage
One initial dose and maintenance doses
every 4 days for 14 days
336 hours
The CADM model was not run because the dosing protocol includes both initial and maintenance doses, which is
not supported in the model.
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.55
307
1,607
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.119
-
25.4
-
112
-
Max.
0.3 14 (@ 288 hours)
-
143 (@ 288 hours)
-
797 (@ 288 hours)
-
Terminal
0.173
-
32.8
-
150
-
LIVER CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.55
307
1,607
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
6.95
-
8,138
-
46,701
-
Max.
14.3 (@ 292 hours)
-
14,826 (@ 296 hours)
-
86,754 (@ 296 hours)
-
Terminal
11.1
-
12,897
-
75,253
-
FA T CON CENTRA TIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.55
307
1,607
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
9.14
-
1,997
-
8,710
-
Max.
16. 1(@ 336 hours)
-
3, 197 (@ 324 hours)
-
14,7 16 (@ 323 hours)
-
Terminal
16.1
-
3,186
-
14,638
-
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.55
307
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.12
-
545
-
Max.
1. 92 (@ 295 hours)
-
952 (@ 294 hours)
-
Terminal
1.88
-
857
-
                                              E-107

-------
1,607
Emond
CADM
2,890
-
5,239 (@ 294 hours)
-
4,667
-
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.55
307
1,607
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.409
-
45.9
-
82.1
-
Max.
0.803 (@ 292 hours)
-
63.7 (@ 1 hour)
-
95.8 (@ 1 hour)
-
Terminal
0.604
-
56.8
-
92.7
-
E.3.1.9. Franc et al (2001) Sprague-Dawley Rats
Type:
Strain:
Body weight:
Sex:
Rats
Sprague-Dawley
200 g( 10 weeks old)
Female
Dose:
ioute:
Regime:
Simulation time:
140, 420, and 1,400 ng/kg every 2 weeks
(equivalent to 10, 30, and 100 ng/kg-day)
Oral gavage
Once every 2 weeks for 22 weeks
3,696 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
10
30
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
6.59
-
14.5
-
Max.
34.6 (@ 3,360 hours)
-
98. 1(@ 3,360 hours)
-
Terminal
5.52
-
11.3
-
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
100
Model
Emond
CADM
Metric
Time-weighted average
36.4
-
Max.
3 15 (@ 3,360 hours)
-
Terminal
26.4
-
LIVER CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
10
30
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1,447
2,616
4,228
7,936
13,821
26,564
Max.
2,458 (@ 3,368 hours)
3,620
7,161 (@ 3,368 hours)
10,899
23,417 (@ 3,368 hours)
36,361
Terminal
1,150
2,174
3,120
6,510
9,658
21,703
FA T CONCENTRA TIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
10
30
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
619
966
1,362
2,448
Max.
787 (@ 3,417 hours)
1,230
1, 74 1(@ 3,4 15 hours)
3,203
Terminal
560
759
1,161
1,849
                                       E-108

-------
100
Emond
CADM
3,430
7,573
4,464 (@ 3,412 hours)
10,052
2,755
5,606
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
10
30
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
119
159
308
450
921
1,462
Max.
177 (@ 3,366 hours)
212
472 (@ 3,366 hours)
603
1,445 (@ 3,366 hours)
1,969
Terminal
99.5
133
240
367
671
1,181
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
10
30
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
18.6
-
33.7
-
57.5
-
Max.
32.9 (@ 1 hour)
-
59.2 (@ 1 hour)
-
86.9 (@ 1 hour)
-
Terminal
16.4
-
29.0
-
50.4
-
E.3.1.10. Franc etal (2001) Long-Evans Rats
Type:
Strain:
Jody weight:
Sex:
Rats
^ong-Evans
190 g( 10 weeks old)
7emale
Dose:
Route:
Regime:
Simulation time:
140, 420, and 1,400 ng/kg every 2 weeks
(equivalent to 10, 30, and 100 ng/kg-day)
Oral gavage
Once every 2 weeks for 22 weeks
3,696 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
10
30
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
6.58
-
14.5
-
36.4
-
Max.
34.2 (@ 3,360 hours)
-
97.0 (@ 3,360 hours)
-
3 12 (@ 3,360 hours)
-
Terminal
5.52
-
11.3
-
26.4
-
LIVER CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
10
30
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1,447
2,616
4,228
7,936
13,821
26,564
Max.
2,458 (@ 3,368 hours)
3,620
7,161 (@ 3,368 hours)
10,899
23,421 (@ 3,368 hours)
36,361
Terminal
1,150
2,174
3,121
6,510
9,659
21,703
                                       E-109

-------
FA T CON CENTRA TIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
10
30
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
619
966
1,362
2,448
3,429
7,573
Max.
788 (@ 3,417 hours)
1,230
1,742 (@ 3,414 hours)
3,203
4,466 (@ 3,412 hours)
10,052
Terminal
560
759
1,160
1,849
2,752
5,606
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
10
30
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
119
159
308
450
921
1,462
Max.
177 (@ 3,366 hours)
212
472 (@ 3,366 hours)
603
1,445 (@ 3,366 hours)
1,969
Terminal
99.5
133
240
367
671
1,181
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
10
30
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
18.6
-
33.7
-
57.5
-
Max.
32.9 (@ 1 hour)
-
59.2 (@ 1 hour)
-
86.9 (@ 1 hour)
-
Terminal
16.4
-
29.0
-
50.4
-
E.3.1.11. Franc etal (2001) Hans Wistar Rats
Type:
Strain:
Body weight:
Sex:
Rats
Hans Wistar
205 g( 10 weeks old)
Female
Dose:
Route:
Regime:
Simulation time:
140, 420, and 1,400 ng/kg every 2 weeks
(equivalent to 10, 30, and 100 ng/kg-day)
Oral gavage
Once every 2 weeks for 22 weeks
3,696 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
10
30
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
6.59
-
14.5
-
36.4
-
Max.
34.7 (@ 3,360 hours)
-
98.7 (@ 3,360 hours)
-
3 17 (@ 3,360 hours)
-
Terminal
5.52
-
11.3
-
26.4
-
                                        E-110

-------
LIVER CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
10
30
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1,447
2,616
4,228
7,936
13,821
26,564
Max.
2,458 (@ 3,368 hours)
3,620
7, 160 (@ 3,368 hours)
10,899
23,416 (@ 3,368 hours)
36,361
Terminal
1,150
2,174
3,120
6,510
9,658
21,703
FA T CONCENTRA TIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
10
30
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
619
966
1,363
2,448
3,431
7,573
Max.
787 (@ 3,4 18 hours)
1,230
1, 74 1(@ 3,4 15 hours)
3,203
4,463 (@ 3,412 hours)
10,052
Terminal
560
759
1,162
1,849
2,757
5,606
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
10
30
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
119
159
308
450
921
1,462
Max.
177 (@ 3,366 hours)
212
472 (@ 3,366 hours)
603
1,446 (@ 3,366 hours)
1,969
Terminal
99.5
133
240
367
671
1,181
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
10
30
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
18.6
-
33.7
-
57.5
-
Max.
32.9 (@ 1 hour)
-
59.2 (@ 1 hour)
-
86.9 (@ 1 hour)
-
Terminal
16.4
-
29.0
-
50.4
-
E.3.1.12. Hassounetal (2000)
Type:
Strain:
Body weight:
Sex:
Rat
Sprague-Dawley
BW = 215g(8weeks
old)
Female
Dose:
Route:
Regime:
Simulation time:
0, 3, 10, 22, 46, 100 ng/kg-day (2.14, 7.14, 15.7,
32.9, and 71.4 ng/kg-day adjusted doses)
Oral gavage
5 days/week for 13 weeks
2, 184 hours
                                       E-lll

-------
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.94
-
4.6136
-
8.147
-
14.009
-
25.34
-
Max.
3. 12 (@ 2, 112 hours)
-
7.71 (@ 2,1 12 hours)
-
14.2 (@ 2, 112 hours)
-
25.8 (@ 2,1 12 hours)
-
49.7 (@ 2, 112 hours)
-
Terminal
1,303.17
-
2,901.26
-
4,947.3
-
8,277
-
14,637
-
LIVER CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
266.8
470
888
1,678
1,948.499
1,768
4,055.031
7,957
8,774.97
17,387
Max.
399 (@ 2,1 16 hours)
595
1,259 (@ 2, 117 hours)
2,001
2,689 (@ 2, 117 hours)
4,428
5,484 (@ 2, 117 hours)
9,272
1 1,692 (@ 2,1 17 hours)
20,170
Terminal
349
595
1,079
2,001
2,278.182
4,428
4,607.265
9,272
9,754.31
20,170
FA T CONCENTRA TIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
179.2
325
427
730
755
1,356
1,299
2,577
2,349.892
5,304
Max.
243 (@ 2, 126 hours)
355
553 (@ 2, 124 hours)
787
958 (@ 2, 123 hours)
1,463
1,627 (@ 2, 122 hours)
2,787
2,928 (@ 2, 121 hours)
5,748
Terminal
234.9
349
528
769
908
1,430
1,529
2,727
2,727.240
5,630
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
27.425
38.2
76.87
108
153.1
224
295
453
600
970
Max.
38.9 (@ 2,1 16 hours)
45.9
105 (@ 2, 116 hours)
126
205 (@ 2, 116 hours)
258
390 (@ 2,1 16 hours)
522
785 (@ 2, 116 hours)
1,113
Terminal
35.720
45.9
93.67
126
180.2
258
339
522
674
1,113
E-112

-------
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
6
-
13.7242
-
21.9703
-
32.817
-
47.54
-
Max.
8.48 (@ 2, 116 hours)
-
17.5 (@ 2, 116 hours)
-
27. 1(@ 2, 116 hours)
-
39.2 (@ 2,1 16 hours)
-
55.0 (@ 2,1 16 hours)
-
Terminal
8
-
15.7348
-
24.4047
-
35.608
-
50.63
-
E.3.1.13. Huttetal (2008)
Type:
Strain:
Jody weight:
Sex:
Rat
Sprague-Dawley
4.5 g (weight at birth)
7emale
Dose:
Route:
Regime:
Simulation time:
50 ng/kg-week (equivalent to 7.14 ng/kg-day)
Oral gavage
1 per week for 13 weeks
2, 1 84 hours (weekly exposure)
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
7.14
Model
Emond
CADM
Metric
Time-weighted average
4.49
-
Max.
8.86 (@ 2,0 16 hours)
-
Terminal
4.71
-
LIVER CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
7.14
Model
Emond
CADM
Metric
Time-weighted average
867.4
1,678
Max.
1,363 (@ 2,021 hours)
2,007
Terminal
928.1
2,007
FA T CONCENTRA TIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
7.14
Model
Emond
CADM
Metric
Time-weighted average
423.6
730
Max.
555 (@ 2,040 hours)
787.1
Terminal
459.9
769
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
7.14
Model
Emond
CADM
Metric
Time-weighted average
76
108
Max.
108 (@ 2,022 hours)
126
Terminal
81
126
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
7.14
Model
Emond
CADM
Metric
Time-weighted average
14
-
Max.
19.4 (@ 2,020 hours)
-
Terminal
14
-
                                       E-113

-------
E.3.1.14. Ishiharaetal (2007)
Type:
Strain:
Jody weight:
Sex:
VIouse
CR
23 g (7 weeks old)
Male and Female
Dose:
Route:
Regime:
Simulation time:
2 and 2,000 ng/kg-week initial and 0.4 or
400 ng/kg-week maintenance (equivalent
to 0.024 and 2.4 ng/kg-day)
Gavage
One initial dose and weekly maintenance
doses for 5 weeks
840 hours
The CADM model was not run because the dosing protocol includes both initial and maintenance doses, which is
not supported in the model.
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.024
2.4
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.0172
-
7.04
-
Max.
0.076 (@ 672 hours)
-
6 1. 2 (@ 672 hours)
-
Terminal
0.0247
-
6.47
-
LIVER CONCENTRATIONS (tig/kg)
Dose (ng/kg-day)
adjusted dose
0.024
2.4
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.45
-
2,805
-
Max.
3. 65 (@ 677 hours)
-
5,059 (@ 680 hours)
-
Terminal
2.13
-
2,758
-
FA T CON CENTRA TIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.024
2.4
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
5.48
-
2,352
-
Max.
9.88 (@ 749 hours)
-
3,284 (@ 712 hours)
-
Terminal
9.63

2,856
-
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.024
2.4
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.537
-
381
-
Max.
0.964 (@ 680 hours)
-
6 17 (@ 678 hours)
-
Terminal
0.902
-
413
-
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.024
2.4
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.0599
-
18.6
-
Max.
0. 150 (@ 676 hours)
-
43. 6 (@ 2 hours)
-
Terminal
0.0861
-
18.4
-
                                              E-114

-------
E.3.1.15.  Kitchin and Woods (1979)
Type:
Strain:
Jody weight:
Sex:
Rats
Sprague-Dawley
BW = 225 g (200 to 250
g)
Female
Dose:
Route:
Regime:
Simulation time:
0, 0.6, 2, 4, 20, 60, 200, 600, 2,000, 5,000,
20,000 ng/kg-day
Oral exposure
Single dose
24 hours
1 week is the minimum that can be simulated with the CADM model, so the CADM model was not used.
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.6
2
4
20
60
200
600
2,000
5,000
20,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.0645
-
0.202
-
0.384
-
1.61
-
4.15
-
11.6
-
30.3
-
90.9
-
218
-
863
-
Max.
0. 126 (@0 hours)
-
0.42 1(@0 hours)
-
0.841 (@0 hours)
-
4.2 1(@0 hours)
-
12.6 (@ 0 hours)
-
42. 1(@0 hours)
-
126 (@ 0 hours)
-
422 (@ 0 hours)
-
1,056 (@ 0 hours)
-
4,233 (@ 0 hours)
-
Terminal
0.0441
-
0.137
-
0.258
-
1.04
-
2.55
-
6.61
-
15.8
-
42.8
-
96.9
-
365
-
LIVER CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.6
2
4
20
60
200
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
2.95
-
10.5
-
22.2
-
128
-
420
-
1,523
-
Max.
3.81(@4hours)
-
12.9 (@ 4 hours)
-
26.3 (@ 4 hours)
-
143 (@ 6 hours)
-
463 (@ 8 hours)
-
1,666 (@ 9 hours)
-
Terminal
2.31
-
8.69
-
18.9
-
118
-
406
-
1,526
-
                                           E-115

-------
600
2,000
5,000
20,000
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
4,821
-
16,603
-
41,971
-
167,820
-
5,258 (@ 10 hours)
-
1 8,080 (@ 11 hours)
-
45,674 (@ 11 hours)
-
182,580 (@ 11 hours)
-
4,932
-
17,226
-
43,803
-
175,890
-
FA T CONCENTRA TIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.6
2
4
20
60
200
600
2,000
5,000
20,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.60
-
5.07
-
9.68
-
41.7
-
110
-
317
-
851
-
2,620
-
6,361
-
25,401
-
Max
2.47 (@ 24 hours)
-
7.7 1(@ 24 hours)
-
14.6 (@ 24 hours)
-
60.7 (@ 24 hours)
-
155 (@ 24 hours)
-
427 (@ 24 hours)
-
1, 102 (@ 24 hours)
-
3,276 (@ 24 hours)
-
7,816 (@ 24 hours)
-
30,827 (@ 24 hours)
-
Terminal
2.47
-
7.71
-
14.6
-
60.7
-
155
-
427
-
1,102
-
3,276
-
7,816
-
30,827
-
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.6
2
4
20
60
200
600
2,000
5,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.322
-
1.07
-
2.14
-
10.6
-
31.7
-
105
-
315
-
1,049
-
2,621
-
Max
0.34 1(@ 9 hours)
-
1. 14 (@ 8 hours)
-
2.27 (@ 8 hours)
-
11.3(@8hours)
-
33.8 (@ 7 hours)
-
1 12 (@ 7 hours)
-
337 (@ 7 hours)
-
1, 123 (@ 7 hours)
-
2,806 (@ 7 hours)
-
Terminal
0.338
-
1.12
-
2.23
-
11.0
-
32.8
-
108
-
324
-
1,074
-
2,680
-
E-116

-------
20,000
Emond
CADM
10,468
-
1 1,2 1 5 (@ 7 hours)
-
10,693
-
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.6
2
4
20
60
200
600
2,000
5,000
20,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.216
-
0.668
-
1.25
-
4.87
-
11.2
-
25.1
-
45.8
-
73.3
-
90.9
-
106
-
Max
0.309 (@ 3 hours)
-
0.975 (@ 3 hours)
-
1. 86 (@ 3 hours)
-
7.67 (@ 2 hours)
-
18.3 (@ 2 hours)
-
40.8 (@ 1 hours)
-
68.2 (@ 1 hours)
-
93.1 (@ 1 hour)
-
104 (@ 1 hour)
-
110(@lhour)
-
Terminal
0.159
-
0.494
-
0.927
-
3.66
-
8.55
-
19.7
-
37.6
-
64.7
-
84.7
-
104
-
E.3.1.16.  Kodbaetal (1976)
Type:
Strain:
Body weight:
Sex:
Rats
Sprague-Dawley (Spartan)
BW= 180 g (170-190 g)
Female
Dose:
Route:
Regime:
Simulation time:
1, 10, 100, and 1,000 ng/kg-day
Dietary exposure
5 days/week for 13 weeks
2,184 hours (13 weeks exposed)
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.714
7.143
71.43
714.3
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.859
-
4.61
-
25.3
-
181
-
Max
1.38 (@ 2,112 hours)
-
7.62 (@ 2, 112 hours)
-
48.8 (@ 2, 112 hours)
-
403 (@ 2, 112 hours)
-
Terminal
1.13
-
5.27
-
26.6
-
184
-
                                      E-117

-------
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.714
7.143
71.43
714.3
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
88.3
136
888
1,678
8,776
17,387
86,329
174,576
Max
140 (@ 2, 116 hours)
192
1,259 (@ 2,1 17 hours)
2,007
11,693 (@ 2, 117 hours)
20,170
1 12,580 (@ 2, 117 hours)
201,814
Terminal
126
192
1,079
2,007
9,756
20,170
92,835
201,814
FA T CON CENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.714
7.143
71.43
714.3
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
79.4
165
427
730
2,348
5,305
16,815
50,658
Max
1 14 (@ 2, 129 hours)
190
553 (@ 2, 124 hours)
787
2,925 (@ 2, 121 hours)
5,748
2 1, 126 (@ 2, 120 hours)
55,013
Terminal
111
189
528
769
2,720
5,630
19,233
53,928
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.714
7.143
71.43
714.3
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
10.8
15.9
76.9
108
600
969
5,366
9,562
Max
16. 1(@ 2, 116 hours)
20.0
105 (@ 2, 116 hours)
126
785 (@ 2, 116 hours)
1,113
6,960 (@ 2, 116 hours)
10,967
Terminal
15.1
20.0
93.6
126
673
1,113
5,842
10,967
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.714
7.143
71.43
714.3
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
2.89
-
13.7
-
47.5
-
93.4
-
Max
4. 17 (@ 2, 116 hours)
-
17.5 (@ 2, 116 hours)
-
55.0 (@ 2, 116 hours)
-
98.2 (@ 2, 117 hours)
-
Terminal
3.81
-
15.7
-
50.6
-
95.7
-
E-118

-------
E.3.1.17. Kociba et al (1978) Female
Type:
Strain:
Body weight:
Sex:
Rats
Sprague-Dawley (Spartan)
BW= 180 g (170-190 g)
7emale
Dose:
Route:
Regime:
Simulation time:
0, 1, 10, and lOOng/kg-day
Dietary exposure
7 days/week for 104 weeks
17,472 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.55
-
7.15
-
38.6
-
Max Terminal
1.92 (@ 17,448 hours) 1.69
-
9.25 (@ 17,448 hours) 7.16
-
57.5 (@ 17,448 hours) 37.1
-
LIVER CONCENTRATIONS (tig/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
192
295
1,618
3,013
14,892
30.239
Max
226 (@ 17,452 hours)
334
1,742 (@ 17,452 hours)
3,348
15,673 (@ 17,452 hours)
33.488
Terminal
218
334
1,665
3,348
14,907
33.488
FA T CONCENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
147
198
680
869
3,663
6.816
Max
165 (@ 17,457 hours)
229
713 (@ 17,454 hours)
1,015
3,788 (@ 17,454 hours)
7,939
Terminal
164
181
706
788
3,731
6.195
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
21.2
26.1
131
171
989
1,562
Max
24.3 (@ 17,452 hours)
27.0
140 (@ 17,452 hours)
176
1,039 (@ 17,452 hours)
1,601
Terminal
23.8
27.0
136
176
994
1,601
                                       E-119

-------
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
5.11
-
20.0
-
59.9
-
Max
5.77 (@ 17,452 hours)
-
21. 1(@ 17,452 hours)
-
61.5 (@ 17,452 hours)
-
Terminal
5.59
-
20.4
-
60.1
-
E.3.1.18.  Kodbaetal (1978) Male
Type:
Strain:
Jody weight:
Sex:
Rats
Sprague-Dawley (Spartan)
BW approximated to be 250 g
Male
Dose:
Route:
Regime:
Simulation time:
0, 1, 10, and 100 ng/kg-day
Dietary exposure
7 days/week for 104 weeks
17,472 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.56
-
7.16
-
38.7
-
Max
1.96 (@ 17,448 hours)
-
9.35 (@ 17,448 hours)
-
59.3 (@ 17,448 hours)
-
Terminal
1.70
-
7.11
-
37.1
_
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
194
295
1,616
3,013
14,898
30.239
Max
229 (@ 17,452 hours)
334
1,723 (@ 17,452 hours)
3,348
15,671 (@ 17,452 hours)
33.488
Terminal
221
334
1,649
3,348
14,912
33.488
FA T CONCENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
148
198
680
869
3,677
6.816
Max
167 (@ 17,456 hours)
229
709 (@ 17,454 hours)
1,015
3,803 (@ 17,453 hours)
7,939
Terminal
166
181
703
788
3,747
6.195
                                      E-120

-------
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
21.4
26.1
131
171
991
1,562
Max
24.6 (@ 17,452 hours
27.0
139 (@ 17,452 hours)
176
1,041 (@ 17,452 hours)
1,601
Terminal
24.1
27.0
134
176
995
1,601
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
5.15
-
20.0
-
60.0
-
Max
5.83 (@ 17,452 hours)
-
21.0 (@ 17,452 hours)
-
61.5 (@ 17,452 hours)
-
Terminal
5.64
-
20.3
-
60.1
-
E.3.1.19. Kuchiiwaetal (2002)
Type:
Strain:
Jody weight:
Sex:
Mouse
ddy
25 g
7emale
Dose:
Route:
Regime:
Simulation time:
4.9 and 490 ng/kg-week (equivalent to 0.7
and 70 ng/kg-day)
Gavage
Once a week for 8 weeks
1,344 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.7
70
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.257
-
9.12
-
Max.
1.01 (@ 1,176 hours)
-
77.7 (@ 1,176 hours)
-
Terminal
0.323
-
8.10
-
LIVER CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.7
70
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
33.7
28.4
4,033
5,306
Max.
68.0 (@ 1,182 hours)
51.1
6,796 (@ 1,185 hours)
8,597
Terminal
44.7
41.7
3,769
3,914
FA T CONCENTRA TIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.7
Model
Emond
CADM
Metric
Time-weighted average
88.3
92.1
Max.
138 (@ 1,236 hours)
144
Terminal
131
125
                                       E-121

-------
70
Emond
CADM
3,199
2,072
4,252 (@ 1,207 hours)
2,848
3,633
1,739
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.7
70
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
9.32
12.3
533
499
Max.
15.3 (@ 1,182 hours)
19.5
818 (@ 1,182 hours)
749
Terminal
13.3
16.9
544
748
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
0.7
70
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.877
-
22.8
-
Max.
1. 67 (@ 1,181 hours)
-
48.9 (@ 2 hours)
-
Terminal
1.11
-
22.1
-
E.3.1.20. Latchoumycandane andMathur (2002)
Type:
Strain:
Jody weight:
Sex:
Rat
Wistar
BW = 200 g (45 days old)
Male
Dose:
Route:
Regime:
Simulation time:
0, 1, 10, and 100 ng/kg-day
Oral gavage
1 per day for 45 days
1,080 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.785
-
4.65
-
27.3
-
Max
1.37 (@ 1,056 hours)
-
8. 18 (@ 1,056 hours)
-
53.9 (@ 1,056 hours)
-
Terminal
1.18
-
6.18
-
33.8
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
78.5
142
902
1,952
9,579
20,541
Max
138 (@ 1,060 hours)
217
1,423 (@ 1,060 hours)
2,550
14,015 (@ 1,061 hours)
25,915
Terminal
133
182
1,358
1,980
13,306
20,018
                                     E-122

-------
FA T CON CENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
69.8
179
416
861
2,448
6,581
Max
1 13 (@ 1,072 hours)
220
608 (@ 1,065 hours)
1,009
3,425 (@ 1,062 hours)
7,866
Terminal
113
198
604
821
3,380
6,035
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
9.56
16.4
76.7
124
646
1,147
Max
15.9 (@ 1,060 hours)
22.2
1 17 (@ 1,060 hours)
157
933 (@ 1,060 hours)
1,439
Terminal
15.6
19.7
113
125.2
891
1,114
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
2.64
-
13.7
-
48.6
-
Max
4. 12 (@ 1,060 hours)
-
18.8 (@ 1,060 hours)
-
59.0 (@ 1,060 hours)
-
Terminal
3.96
-
18.1
-
57.5
-
E.3.1.21. LietaL (1997)
Type:
Strain:
Body weight:
Sex:
Rats
Sprague -Dawley
BW = 56.5 g (22 days
old, 55 to 58 g)
Female
Dose:
Route:
Regime:
Simulation time:
0, 3, 10, 30, 100, 300, 1,000, 3,000,
10,000, and 30,000 ng/kg-day
Gastric intubation
One dose for one day
24 hours
The CADM model was not run because the dosing duration is lower than the resolution of the model (1 week)
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
3
Model
Emond
CADM
Metric
Time-weighted average
0.266
-
Max
0.470 (@ 1 hour)
-
Terminal
0.180
-
                                            E-123

-------
10
30
100
300
1,000
3,000
10,000
30,000
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
0.799
-
2.10
-
5.87
-
15.0
-
43.3
-
120
-
386
-
1,172
-
1.57 (@ 1 hour)
-
4.68 (@ 1 hour)
-
15.6 (@ 1 hour)
-
46. 8 (@0 hours)
-
1 56 (@0 hours)
-
469 (@ 0 hours)
-
1,570 (@ 0 hours)
-
4,762 (@ 0 hours)
-
0.535
-
1.37
-
3.68
-
8.83
-
23.4
-
59.9
-
182
-
535
-
LIVER CONCENTRATIONS (tig/kg)
Dose
(ng/kg-day)
adjusted dose
3
10
30
100
300
1,000
3,000
10,000
30,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
14.7
-
55.0
-
185
-
690
-
2,248
-
7,938
-
24,474
-
82,349
-
245,610
-
Max
18.6 (@ 4 hours)
-
65.2 (@ 5 hours)
-
2 10 (@ 6 hours)
-
768 (@ 7 hours)
-
2,473 (@ 8 hours)
-
8,671 (@ 9 hours)
-
26,639 (@ 9 hours)
-
89,464 (@ 9 hours)
-
265,670 (@ 10 hours)
-
Terminal
11.9
-
47.6
-
170
-
666
-
2,240
-
8,094
-
25,267
-
85,597
-
255,390
-
FA T CON CENTRA TIONS (tig/kg)
Dose
(ng/kg-day)
adjusted dose
3
10
30
100
300
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
8.75
-
26.6
-
70.8
-
202
-
530
-
Max
12.7 (@ 24 hours)
-
38.0 (@ 24 hours)
-
98.9 (@ 24 hours)
-
273 (@ 24 hours)
-
689 (@ 24 hours)
-
Terminal
12.7
-
38.0
-
98.9
-
273
-
689
-
E-124

-------
1,000
3,000
10,000
30,000
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
1,573
-
4,433
-
14,428
-
44,361
-
1,958 (@ 24 hours)
-
5,358 (@ 24 hours)
-
17, 1 19 (@ 24 hours)
-
5 1,948 (@ 22 hours)
-
1,958
-
5,358
-
17,119
-
51,898
-
BODY BURDEN (tig/kg)
Dose
(ng/kg-day)
adjusted dose
3
10
30
100
300
1,000
3,000
10,000
30,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.60
-
5.33
-
15.9
-
52.8
-
158
-
525
-
1,574
-
5,240
-
15,758
-
Max
1. 70 (@ 8 hours)
-
5.66 (@ 8 hours)
-
16.9 (@ 8 hours)
-
56.2 (@ 7 hours)
-
169 (@ 7 hours)
-
561 (@ 7 hours)
-
1,684 (@ 7 hours)
-
5,610 (@ 7 hours)
-
16,8 15 (@ 7 hours)
-
Terminal
1.68
-
5.56
-
16.5
-
54.5
-
163
-
539
-
1,611
-
5,360
-
16,041
-
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
3
10
30
100
300
1,000
3,000
10,000
30,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.89
-
2.58
-
6.37
-
15.54
-
31.25
-
56.75
-
81.28
-
99.77
-
107.69
-
Max
1. 37 (@ 3 hours)
-
4. 10 (@ 2 hours)
-
10.5 (@ 2 hours)
-
25.9 (@ 2 hours)
-
50. 1 (@ 1 hour)
-
79.8 (@ 1 hour)
-
98.4 (@ 1 hour)
-
108 (@ 1 hour)
-
Ill (@1 hour)
-
Terminal
0.64
-
1.88
-
4.71
-
11.77
-
24.57
-
47.62
-
73.32
-
95.68
-
106.24
-
E-125

-------
E.3.1.22. Murray et al (1979) Adult Portion
Type:
Strain:
Body weight:
Sex:
Rat
Sprague -Dawley
BW = 4.5 g
Female
Dose:
Route:
Regime:
Simulation time:
1, 10, and 100 ng/kg-day
Dietary exposure
Once per day for 120 days
2,880 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.12
-
5.88
-
32.7
-
Max
1. 5 1(@ 2,856 hours)
-
7.59 (@ 2,856 hours)
-
44.3 (@ 2,856 hours)
-
Terminal
1.42
-
6.75
-
36.0
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
128
232
1,273
2,613
12,601
26,609
Max
180 (@ 2,859 hours)
312
1,618 (@ 2,860 hours)
3,179
15,281 (@ 2,860 hours)
31,868
Terminal
173
312
1,540
3,179
14,460
31,868
FA T CON CENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
106
209
556
975
3,095
7,742
Max
139 (@ 2,865 hours)
243
665 (@ 2,864 hours)
1,103
3,604 (@ 2,862 hours)
8,790
Terminal
138
236
657
1,053
3,534
8,427
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
14.8
22.5
105
159
837
1,468
Max
20.0 (@ 2,860 hours)
28.3
130 (@ 2,860 hours)
189
1,003 (@ 2,860 hours)
1,738
Terminal
19.6
28.3
126
189
957
1,738
                                          E-126

-------
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
3.77
-
17.1
-
55.3
-
Max
4.95 (@ 2,859 hours)
-
20.3 (@ 2,859 hours)
-
60.9 (@ 2,860 hours)
-
Terminal
4.77
-
19.5
-
59.4
-
E.3.1.23. NTP (1982) Female Rats, Chronic
Type:
Strain:
Jody weight
Sex:
Rat
Osborne-Mendel
BW = 250 g
6 weeks old)
Female
Dose:
Route:
Regime:
Simulation time
10, 50, and 500 ng/kg-week, 2 doses/week
Oral exposure
2 doses/week
17,472 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.4
7.1
71
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.96
-
5.69
-
29.8
-
Max
3. 11(@ 17,220 hours)
-
1 1.0 (@ 17,388 hours)
-
82.2 (@ 17,388 hours)
-
Terminal
1.94
-
5.40
-
26.9
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.4
7.1
71
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
265
424
1,175
2,150
10,734
21,596
Max
308 (@ 17,226 hours)
477
1,338 (@ 17,394 hours)
2,391
12, 182 (@ 17,395 hours)
23,920
Terminal
265
477
1,117
2,391
9,882
23,920
FAT CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.4
Model
Emond
CADM
Metric
Time-weighted average
186
241
Max
200 (@ 17,328 hours)
280
Terminal
193
220
                                       E-127

-------
7.1
71
Emond
CADM
Emond
CADM
541
673
2,826
4,934
569 (@ 17,409 hours)
787
2,973 (@ 17,404 hours)
5,748
544
610
2,769
4,483
BODY BURDEN (tig/kg)
Dose
(ng/kg-day)
adjusted dose
1.4
7.1
71
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
27.9
33.9
99.4
126.4
729
1,121
Max
3 1.1 (@ 17,225 hours)
35.0
1 10 (@ 17,393 hours)
129.8
814 (@ 17,393 hours)
1,149
Terminal

35.0
96.7
129.8
683
1,149
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.4
7.1
71
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
6.37
-
16.6
-
52.7
-
Max
7.26 (@ 17,224 hours)
-
18.5 (@ 17,392 hours)
-
56.4 (@ 17,393 hours)
-
Terminal
6.38
"
16.1
-
50.9

E.3.1.24. NTP (1982) Male Rats, Chronic
Type:
Strain:
Jody weight
Sex:
Rat
Osborne-Mendel
BW = 350g
6 weeks old)
Male
Dose:
Route:
Regime:
Simulation time
10, 50, and 500 ng/kg-week, 2 doses/week
Oral exposure
2 doses/week
17,472 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.4
7.1
71
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.96
-
5.70
-
29.9
-
Max
3. 18 (@ 17,388 hours)
-
1 1.4 (@ 17,388 hours)
-
87.0 (@ 17,388 hours)
-
Terminal
1.93
-
5.39
-
26.9
-
                                        E-128

-------
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.4
Model
Emond
CADM
Metric
Time-weighted average
265
424
Max
306 (@ 17,394 hours)
477
Terminal
263
477
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
7.1
71
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1,174
2,150
10,736
21,596
Max
1,334 (@ 17,394 hours)
2,391
12, 170 (@ 17,395 hours)
23,920
Terminal
1,114
2,391
9,881
23,920
FA T CON CENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.4
7.1
71
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
186
241
541
673
2,836
4,934
Max
199 (@ 17,412 hours)
280
569 (@ 17,409 hours)
787
2,983 (@ 17,404 hours)
5,748
Terminal
193
220
544
610
2,784
4,483
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.4
7.1
71
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
27.8
33.9
99.5
126.4
730
1,121
Max
30.9 (@ 17,393 hours)
35.0
1 10 (@ 17,393 hours)
129.8
816 (@ 17,393 hours)
1,149
Terminal
28.2
35.0
96.6
129.8
684
1,149
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.4
7.1
71
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
6.36
-
16.6
-
52.7
-
Max
7.22 (@ 17,392 hours)
-
18.4 (@ 17,392 hours)
-
56.3 (@ 17,393 hours)
-
Terminal
6.35
-
16.0
-
50.9
-
E-129

-------
E.3.1.25. NTP (1982) Female Mice, Chronic
Type:
Strain:
Jody weight
Sex:
Slice
B6C3FJ
BW = 23 g (6 weeks
old)
Female
Dose:
Route:
Regime:
Simulation time
40, 200, and 2,000 ng/kg-week, 2
doses/week
Oral exposure
2 doses/week
17,472 hours
The CADM model was not run because the study duration is longer than the allowed model duration.
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
5.7
28.6
286
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.95
-
5.84
-
32.1
-
Max
4.86 (@ 16,800 hours)
-
19.8 (@ 17,388 hours)
-
171 (@ 16,884 hours)
-
Terminal
1.82
-
5.17
-
26.0
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
5.7
28.6
286
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
490
-
2,236
-
20,841
-
Max
582 (@ 16,807 hours)
-
2,629 (@ 17,395 hours)
-
24,353 (@ 17,396 hours)
-
Terminal
463
-
2,025
-
18,182
-
FAT CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
5.7
28.6
286
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
737
-
2,213
-
12,138
-
Max
785 (@ 17,408 hours)
-
2,337 (@ 17,404 hours)
-
12,861 (@ 17,400 hours)
-
Terminal
757
-
2,216
-
11,775
-
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
5.7
28.6
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
91.9
-
329
-
Max
103 (@ 17,393 hours)
-
370 (@ 17,393 hours)
-
Terminal
91.2
-
313
-
                                           E-130

-------
286
Emond
CADM
2,400
-
2,740 (@ 17,393 hours)
-
2,176
-
BOUND LIVER (tig/kg)
Dose
(ng/kg-day)
adjusted dose
5.7
28.6
286
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
6.18
-
16.3
-
52.3
-
Max
7.29 (@ 16,805 hours)
-
18.9 (@ 17,393 hours)
-
67. 8 (@ 2 hours)
-
Terminal
5.93
-
15.3
-
49.3
-
E.3.1.26. NTP (1982) Male Mice, Chronic
Type:
Strain:
Body weight
Sex:
Slice
B6C3FJ
BW = 25 g
6 weeks old)
Male
Dose:
Route:
Regime:
Simulation time
10, 50, and 500 ng/kg-week, 2 doses during the
week
Oral exposure
2 doses/week
17,472 hours (104 week of exposure)
The CADM model was not run because the study duration is longer than the allowed model duration.
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.4
7.1
71
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.767
-
2.27
-
11.2
-
Max
1.53 (@ 17,304 hours)
-
5.99 (@ 17,052 hours)
-
46.7 (@ 17,388 hours)
-
Terminal
0.749
-
2.11
-
9.59
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.4
7.1
71
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
138
-
606
-
5,409
-
Max
165 (@ 17,3 10 hours)
-
722 (@ 17,059 hours)
-
6,328 (@ 17,395 hours)
-
Terminal
136
-
571
-
4,805
-
                                            E-131

-------
FAT CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.4
7.1
71
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
290
-
860
-
4,257
-
Max
3 14 (@ 17,4 11 hours)
-
918 (@ 17,155 hours)
-
4,490 (@ 17,402 hours)
-
Terminal
306
-
883
-
4,204
-
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.4
7.1
71
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
32.3
-
110
-
710
-
Max
36.2 (@ 17,309 hours)
-
123 (@ 17,057 hours)
-
802 (@ 17,393 hours)
-
Terminal
33.3
-
108
-
660
-
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.4
7.1
71
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
2.56
-
7.12
-
27.1
-
Max
3.03 (@ 17,309 hours)
-
8.40 (@ 17,057 hours)
-
32.4 (@ 2 hours)
-
Terminal
2.53
-
6.82
-
25.3
-
E.3.1.27. NTP'(2006) 14 Weeks
Type:
Strain:
Body weight:
Sex:
Rat
Sprague-Dawley
BW = 215g(8weeks
old)
Female and male
Dose:
Route:
Regime:
Simulation time:
0, 3, 10, 22, 46, and 100 ng/kg-day
Oral gavage
5 days/week for 14 weeks
2,352 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.98
-
4.69
-
Max
3. 15 (@ 2,280 hours)
-
7.75 (@ 2,280 hours)
-
Terminal
2.39
-
5.30
-
                                        E-132

-------
15.7
32.9
71.4
Emond
CADM
Emond
CADM
Emond
CADM
8.27
-
14.2
-
25.7
-
14.3 (@ 2,280 hours)
-
25.9 (@ 2,280 hours)
-
49.8 (@ 2,280 hours)
-
9.02
-
15.1
-
26.6
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
275
479
909
1,702
1,988
3,817
4,129
8,054
8,921
17,592
Max
404 (@ 2,284 hours)
599
1, 270 (@ 2,285 hours)
2,017
2,703 (@ 2,285 hours)
4,449
5,508 (@ 2,285 hours)
9,314
1 1,734 (@ 2,285 hours)
20,262
Terminal
354
599
1,089
2,017
2,291
4,449
4,628
9,314
9,792
20,262
FA T CON CENTRA TIONS (tig/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
184
326
436
733
768
1,361
1,319
2,587
2,385
5,326
Max
246 (@ 2,294 hours)
355
557 (@ 2,292 hours)
787
962 (@ 2,291 hours)
1,463
1,63 3 (@ 2,289 hours)
2,787
2,93 8 (@ 2,289 hours)
5,748
Terminal
237
347
532
765
912
1,422
1,535
2,712
2,736
5,599
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
28.2
38.8
78.5
109
156
226
300
459
610
980
Max
39.4 (@ 2,284 hours)
46.1
106 (@ 2,284 hours)
126
206 (@ 2,284 hours)
259
39 1(@ 2,284 hours)
523
788 (@ 2,284 hours)
1,117
Terminal
36.1
46.1
94.4
126
181
259
340
523
676
1,117
E-133

-------
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
6.41
-
13.9
-
22.2
-
33.2
-
47.9
-
Max
8.55 (@ 2,284 hours)
-
17.6 (@ 2,284 hours)
-
27.2 (@ 2,284 hours)
-
39.3 (@ 2,284 hours)
-
55. 1(@ 2,284 hours)
-
Terminal
7.74
-
15.8
-
24.5
-
35.7
-
50.7
-
E.3.1.28. NTP(2006)31 Weeks
Type:
Strain:
Body weight:
Sex:
Rat
Sprague-Dawley
BW = 2 15 g (8 weeks old)
Female and male
Dose:
Route:
Regime:
Simulation time:
0, 3, 10, 22, 46, 100 ng/kg-day
Oral gavage
5 days/week for 3 1 weeks
5,208 hours
WHOLE BLOOD CONCENTRATIONS (tig/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
2.33
-
5.32
-
9.21
-
15.7
-
28.1
-
Max
3.25 (@ 3,960 hours)
-
7.89 (@ 3,960 hours)
-
14.5 (@ 3,960 hours)
-
26.2 (@ 5,136 hours)
-
50.4 (@ 5,136 hours)
-
Terminal
2.48
-
5.40
-
9.15
-
15.3
-
27.0
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
341
555
1,075
1,906
2,296
4,229
4,696
8,880
Max
425 (@ 5, 140 hours)
631
1,308 (@ 3,965 hours)
2,112
2,756 (@ 3,965 hours)
4,652
5,597 (@ 5, 141 hours)
9,732
Terminal
373
631
1,117
2,112
2,336
4,652
4,712
9,732
                                       E-134

-------
71.4
Emond
CADM
10,033
19,347
1 1,905 (@ 5,141 hours)
21,163
9,953
21,163
FA T CONCENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
220
329
501
732
868
1,361
1,476
2,591
2,652
5,344
Max
256 (@ 5, 149 hours)
355
570 (@ 4, 139 hours)
787
978 (@ 4, 138 hours)
1,463
1,657 (@ 5, 145 hours)
2,787
2,978 (@ 5, 144 hours)
5,748
Terminal
246
320
542
706
926
1,315
1,558
2,509
2,775
5,183
BODY BURDEN (tig/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
34.2
43.2
91.6
119
178
246
339
498
682
1,063
Max
4 1. 2 (@ 5, 140 hours)
47.1
108 (@ 3,964 hours)
129
209 (@ 3,964 hours)
264
398 (@ 5,140 hours)
533
799 (@ 5, 140 hours)
1,138
Terminal
37.8
47.1
96.6
129
184
264
346
533
687
1,138
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
7.48
-
15.6
-
24.3
-
35.7
-
50.9
-
Max
8.83 (@ 5,140 hours)
-
17.9 (@ 3,964 hours)
-
27.4 (@ 3,964 hours)
-
39.6 (@ 5,140 hours)
-
55.4 (@ 5,140 hours)
-
Terminal
8.01
-
16.1
-
24.8
-
36.0
-
51.1
-
E-135

-------
E.3.1.29. NTP(2006) 53 Weeks
Type:
Strain:
Body weight:
Sex:
Rat
Sprague-Dawley
BW = 2 15 g (8 weeks old)
Female and male
Dose:
Route:
Regime:
Simulation time:
0, 3, 10, 22, 46, 100 ng/kg-day
Oral gavage
5 days/week for 53 weeks
8,904 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
2.46
-
5.53
-
9.54
-
16.2
-
29.0
-
Max
3. 25 (@ 6,3 12 hours)
-
7.89 (@ 3,960 hours)
-
14.5 (@ 8,832 hours)
-
26.3 (@ 8,832 hours)
-
50.6 (@ 8,832 hours)
-
Terminal
2.48
-
5.41
-
9.17
-
15.3
-
27.1
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
366
593
1,134
2,010
2,406
4,446
4,902
9,318
10,439
20,284
Max
426 (@ 6,3 16 hours)
656
1,308 (@ 3,965 hours)
2,197
2,759 (@ 8,837 hours)
4,836
5,6 12 (@ 8,837 hours)
10,115
1 1,938 (@ 8,837 hours)
21,993
Terminal
373
656
1,121
2,197
2,345
4,836
4,727
10,115
9,985
21,993
FA T CON CENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
233
321
524
711
904
1,323
1,533
2,522
2,749
5,205
Max
256 (@ 6,325 hours)
355
570 (@ 4, 139 hours)
787
980 (@ 8,842 hours)
1,463
1,661 (@ 8,841 hours)
2,787
2,986 (@ 8,840 hours)
5,748
Terminal
247
301
544
663
929
1,236
1,562
2,359
2,784
4,873
                                        E-136

-------
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
36.4
44.9
96.1
123
186
254
353
513
709
1,096
Max
4 1. 2 (@ 6,3 16 hours)
47.4
108 (@ 3,964 hours)
129
2 10 (@ 8,836 hours)
266
399 (@ 8,836 hours)
536
80 1(@ 8,836 hours)
1,144
Terminal
37.8
47.4
96.9
129
185
266
347
536
689
1,144
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
7.87
-
16.2
-
25.1
-
36.6
-
51.9
-
Max
8.84 (@ 6,3 16 hours)
-
17.9 (@ 3,964 hours)
-
27.5 (@ 8,836 hours)
-
39.7 (@ 8,836 hours)
-
55.4 (@ 8,836 hours)
-
Terminal
8.01
-
16.1
-
24.8
-
36.1
-
51.1
-
E.3.1.30. NTP(2006) 2 Years
Type:
Strain:
Jody weight:
Sex:
Rat
Sprague-Dawley
BW = 2 15 g (8 weeks old)
Female and male
Dose:
Route:
Regime:
Simulation time:
0, 3, 10, 22, 46, 100 ng/kg-day
Oral gavage
5 days/week for 105 weeks
17,640 hours
The CADM model simulates for 104 weeks only (17,472 hours).  As a result, the terminal values from the CADM
model may be underestimated compared to the Emond model, which considers the full 105 weeks of exposure.
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
2.56
-
5.69
-
Max
3.47 (@ 17,568 hours)
-
7.97 (@ 17,568 hours)
-
Terminal
2.62
-
5.46
-
                                             E-137

-------
15.7
32.9
71.4
Emond
CADM
Emond
CADM
Emond
CADM
9.79
-
16.6
-
29.7
-
14.6 (@ 17,568 hours)
-
26.4 (@ 17,568 hours)
-
50.8 (@ 17,568 hours)
-
9.22
-
15.4
-
27.1
-
LIVER CONCENTRATIONS (tig/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
385
639
1,177
2,150
2,487
4,742
5,051
9,927
10,734
21,596
Max
460 (@ 17,572 hours)
717
1,320 (@ 17,573 hours)
2,391
2,779 (@ 17,573 hours)
5,261
5,637 (@ 17,573 hours)
11,002
1 1,976 (@17,573hr)
23,920
Terminal
403
717
1,135
2,391
2,361
5,261
4,749
11,002
10,018
23,920
FA T CONCENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
243
304
541
673
930
1,253
1,574
2,390
2,821
4,934
Max
271 (@ 17,581 hours)
355
575 (@ 17,579 hours)
787
985 (@ 17,578 hours)
1,463
1,667 (@ 17,577 hours)
2,787
2,995 (@ 17,576 hours)
5,748
Terminal
261
277
549
610
934
1,137
1,568
2,170
2,792
4,934
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
38.1
46.2
99.5
126
192
260
364
525
729
1,121
Max
44.0 (@ 17,572 hours)
47.6
109 (@ 17,572 hours)
130
211 (@ 17,572 hours)
267
400 (@ 17,572 hours)
538
804 (@ 17,572 hours)
1,149
Terminal
40.4
47.6
97.9
130
186
267
348
538
691
1,149
E-138

-------
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
2.14
7.14
15.7
32.9
71.4
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
8.17
-
16.6
-
25.6
-
37.3
-
52.7
-
Max
9.30 (@ 17,572 hours)
-
18.0 (@ 17,572 hours)
-
27.6 (@ 17,572 hours)
-
39.7 (@ 17,572 hours)
-
55.5 (@ 17,572 hours)
-
Terminal
8.43
-
16.2
-
24.9
-
36.2
-
51.2
-
E.3.1.31. Noharaetal (2002)
Type:
Strain:
Body weight:
Sex:
Mice
Four strains
BW = 23 g (8 weeks old)
[Female
iDose:
[Route:
[Regime:
[Simulation time:
5,20, 100, and500ng/kg
Gavage
Single dose
24 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
5
20
100
500
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.229
-
0.817
-
3.41
-
14.2
-
Max
0.686 (@ 0 hours)
-
2.74 (@ 0 hours)
-
13.7(@0hours)
-
68.6 (@ 0 hours)
-
Terminal
0.135
-
0.448
-
1.65
-
5.70
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
5
20
100
500
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
19.8
6.80
85.7
38.7
472
416
2,541
3,998
Max
23. 6 (@ 5 hours)
6.80
96.3 (@ 6 hours)
38.7
517(@ 10 hours)
416
2,785 (@ 11 hours)
3,998
Terminal
16.8
6.80
77.8
38.7
458
416
2,578
3,998
                                       E-139

-------
FA T CON CENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
5
20
100
500
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
13.5
31.1
49.6
119
217
506
952
1,761
Max
20.4 (@ 24 hours)
31.1
72. 3 (@ 24 hours)
119
299 (@ 24 hours)
506
l,231(@24hours)
1,761
Terminal
20.4
31.1
72.3
119
299
506
1,231
1,761
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
5
20
100
500
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
2.84
4.00
11.3
16.0
55.9
80.0
276
400
Max
3. 03 (@ 8 hours)
4.00
12.1(@8hours)
16.0
60.0 (@ 7 hours)
80.0
298 (@ 7 hours)
400
Terminal
2.96
4.00
11.7
16.0
57.4
80.0
282
400
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
5
20
100
500
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.715
-
2.40
-
8.61
-
25.5
-
Max
1. 07 (@ 3 hours)
-
3. 99 (@ 3 hours)
-
16.4 (@ 2 hours)
-
49.4 (@ 2 hours)
-
Terminal
0.507
-
1.67
-
5.88
-
17.8
-
E.3.1.32.  Sewattetal (1995) and Maronpot et al (1993)
Type:
Strain:
Jody weight:
Sex:
Rat
Sprauge -Dawley
BW = 250 g( 12 weeks
old)
Female
Dose:
Route:
Regime:
Simulation time:
49, 149.8, 490, and 1,750 ng/kg every 2
weeks (equivalent to 3.5, 10.7, 35, and 125
ng/kg-day)
Oral gavage
Once every 2 weeks for 30 weeks
5,040 hours
                                      E-140

-------
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
3.5
10.7
35
125
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
3.29
-
7.11
-
16.6
-
44.7
-
Max
1 3. 7 (@ 4,704 hours)
-
38.7 (@ 4,704 hours)
-
120 (@ 4,704 hours)
-
4 14 (@ 4,704 hours)
-
Terminal
2.88
-
5.79
-
12.6
-
31.4
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
3.5
10.7
35
125
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
550
928
1,605
2,891
5,072
9,534
17,683
34,145
Max
90 1(@ 4,7 11 hours)
1,273
2,632 (@ 4,7 12 hours)
3,940
8,3 50 (@ 4,7 12 hours)
12,926
29,256 (@ 4,7 13 hours)
46,190
Terminal
459
786
1,229
2,373
3,618
7,744
12,011
27,659
FA T CONCENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
3.5
10.7
35
125
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
310
451
670
1,008
1,569
2,786
4,217
9,308
Max
383 (@ 4,765 hours)
560
827 (@ 4,763 hours)
1,300
1, 957 (@ 4,760 hours)
3,693
5,376 (@ 4,757 hours)
12,496
Terminal
290
367
590
774
1,304
2,054
3,303
6,738
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
3.5
10.7
35
125
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
51.4
64.8
130
173
364
534
1,164
1,863
Max
72.5 (@ 4,7 10 hours)
83.25
189 (@ 4,710 hours)
227
546 (@ 4,7 10 hours)
704
1, 793 (@ 4,7 10 hours)
2,468
Terminal
45.3
56.0
106
143
274
429
824
-1,483
E-141

-------
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
3.5
10.7
35
125
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
10.2
-
19.8
-
37.0
-
63.1
-
Max
15.8 (@ 2 hours)
-
34.4 (@ 1 hours)
-
63.2 (@ 1 hours)
-
90.9 (@ 1 hours)
-
Terminal
9.18
-
17.0
-
31.4
-
55.2
-
E.3.1.33. Shi et al (2007) Adult Portion
Type:
Strain:
Body weight:
Sex:
Rat
Sprague -Dawley
BW = 4.5 g
Female
Dose:
Route:
Regime:
Simulation time:
1, 5, 50, and 200 ng/kg-week (equivalent
to 0.143, 0.714, 7. 14, and 28.6
ng/kg-day)
Oral exposure
Weekly doses for 1 1 months
8,040 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.143
0.714
7.14
28.6
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.342
-
1.07
-
5.23
-
13.9
-
Max
0.475 (@ 7,561 hours)
-
1. 53 (@ 7,560 hours)
-
9. 12 (@ 7,560 hours)
-
29.2 (@ 7,560 hours)
-
Terminal
0.380
-
1.09
-
4.86
-
12.4
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.143
0.714
7.14
28.6
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
26.1
33.6
118
189
1,068
1,992
4,119
8,031
Max
36.5 (@ 7,564 hours)
42.6
1 59 (@ 7,564 hours)
216
1,415 (@ 7,565 hours)
2,178
5,450 (@ 7,565 hours)
8,722
Terminal
29.6
42.6
120
216
970
2,178
3,574
8,722
                                         E-142

-------
FA T CON CENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.143
0.714
7.14
28.6
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
32.5
71.0
102
173
497
716
1,322
2,237
Max
40.0 (@ 7,583 hours)
78.6
120 (@ 7,584 hours)
190
571 (@ 7,584 hours)
787
1,527 (@ 7,584 hours)
2,457
Terminal
36.7
73.8
106
167
475
671
1,217
2,104
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.143
0.714
7.14
28.6
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
3.94
6.6
14.0
19.6
90.8
123
300
446
Max
4.99 (@ 7,566 hours)
7.6
17.2 (@ 7,566 hours)
21.2
1 12 (@ 7,566 hours)
129
374 (@ 7,566 hours)
468
Terminal
4.45
7.6
14.5
21.2
84.4
129
266
468
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.143
0.714
7.14
28.6
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.18
-
3.62
-
15.6
-
33.5
-
Max
1. 60 (@ 7,563 hours)
-
4.75 (@ 7,563 hours)
-
19.7 (@ 7,564 hours)
-
40.7 (@ 7,564 hours)
-
Terminal
1.31
-
3.70
-
14.7
-
31.2
-
E.3.1.34. Simanainen et al. (2002) and Simanainen et al. (2003)
Type:
Strain:
Jody weight:
Sex:
Rats
Hans/Wistar and Long-Evans
BW = 200 g
Female
Dose:
Route:
Regime:
Simulation time:
100 and 3 00 ng/kg
Oral gavage
Single dose
24 hours
                                       E-143

-------
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
100
300
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
6.36
-
16.3
-
Max.
20. 5 (@0 hours)
-
61.5(@0hours)
-
Terminal
3.82
-
9.07
-
LIVER CONCENTRATIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
100
300
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
725
-
2,331
-
Max.
796 (@ 8 hours)
-
2,547 (@ 9 hours)
-
Terminal
711
-
2,352
-
FA T CONCENTRA TIONS (ng/kg)
Dose (ng/kg-day)
adjusted dose
100
300
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
174
-
461
-
Max.
241 (@ 24 hours)
-
611 (@ 24 hours)
-
Terminal
241
-
611
-
BODY BURDEN (ng/kg)
Dose (ng/kg-day)
adjusted dose
100
300
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
52.8
-
158
-
Max.
56. 3 (@ 7 hours)
-
169 (@ 7 hours)
-
Terminal
54.5
-
162
-
BOUND LIVER (ng/kg)
Dose (ng/kg-day)
adjusted dose
100
300
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
16.0
-
31.8
-
Max.
26.4 (@ 2 hours)
-
50.6 (@ 1 hour)
-
Terminal
12.3
-
25.3
-
E.3.1.35. Smialowicz et al (2004)
Type:
Strain:
Jody weight:
Sex:
Mice
C57BL/6N
BW = 25 g (Age not
specified)
Female
Dose:
ioute:
Regime:
Simulation time:
30, 100, 300, 1,000, 3,000, and
10,000 ng/kg
Oral gavage
Single dose
24 hours
                                        E-144

-------
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
3,000
10,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.19
-
3.44
-
9.08
-
26.9
-
75.1
-
242
-
Max
4. 19 (@0 hours)
-
14.0 (@ 0 hours)
-
42.0 (@ 0 hours)
-
140 (@ 0 hours)
-
420 (@ 0 hours)
-
1,403 (@ 0 hours)
-
Terminal
0.632
-
1.65
-
3.87
-
9.76
-
23.5
-
66.7
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
3,000
10,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
132
68.6
473
416
1,498
2,039
5,199
9,294
15,934
31,419
53,457
109,703
Max
147 (@ 7 hours)
68.6
5 18 (@ 10 hours)
416
l,641(@llhours)
2,039
5,700 (@ 12 hours)
9,294
17,473 (@ 12 hours)
31,419
58,629 (@ 13 hours)
109,703
Terminal
123
68.6
461
416
1,506
2,039
5,345
9,294
16,586
31,419
56,056
109,703
FA T CONCENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
3,000
10,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
71.4
174
215
506
588
1,201
1,804
3,002
5,165
7,593
16,888
23,319
Max
103 (@ 24 hours)
174
296 (@ 24 hours)
506
776 (@ 24 hours)
1,201
2,278 (@ 24 hours)
3,002
6,3 3 3 (@ 24 hours)
7,593
20,306 (@ 24 hours)
23,319
Terminal
103
174
296
506
776
1,201
2,278
3,002
6,333
7,593
20,306
23,319
E-145

-------
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
3,000
10,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
16.9
24.0
55.9
80.0
166
240
550
800
1,646
2,400
5,469
8,000
Max
18.1(@7hours)
24.0
60.0 (@ 7 hours)
80.0
179 (@ 7 hours)
240
594 (@ 7 hours)
800
1,778 (@ 7 hours)
2,400
5,916 (@ 7 hours)
8,000
Terminal
17.5
24.0
57.4
80.0
170
240
560
800
1,668
2,400
5,528
8,000
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
3,000
10,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
3.37
-
8.63
-
18.6
-
37.6
-
61.3
-
86.5
-
Max
5.79 (@ 3 hours)
-
16.4 (@ 2 hours)
-
36.6 (@ 2 hours)
-
67. 8 (@ 2 hours)
-
9 1. 8 (@ 2 hours)
-
106 (@ 2 hours)
-
Terminal
2.34
-
5.90
-
12.8
-
27.2
-
48.3
-
76.1
-
E.3.1.36. Smialowicz et al (2008)
Type:
Strain:
Body weight:
Sex:
Mice
B6C3FJ
BW = 28g(13weeksold)
Female
pose:
(Route:
[Regime:
(Simulation time:
0, 1.5, 15, 150, and 450 ng/kg-day
Oral gavage
5 days/week for 13 weeks
2, 184 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.07
10.7
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.438
-
2.46
-
Max
0.815 (@ 2,1 12 hours)
-
5. 12 (@ 2, 112 hours)
-
Terminal
0.557
-
2.65
-
                                        E-146

-------
107
321
Emond
CADM
Emond
CADM
13.4
-
31.6
-
36.4 (@ 2, 112 hours)
-
98.6 (@ 2, 112 hours)
-
12.7
-
28.4
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.07
10.7
107
321
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
67.1
59.8
683
776
6,784
8,441
20,218
25.626
Max
107 (@ 2, 116 hours)
91.9
97 1(@ 2, 117 hours)
1,000
9,010 (@ 2,1 17 hours)
10,306
26,379 (@ 2, 117 hours)
31,006
Terminal
91.5
84.2
787
825
7,043
7,863
20,405
23.460
FA T CONCENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.07
10.7
107
321
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
156
153
885
697
4,831
2,802
11,420
6,408
Max
229 (@ 2,130 hours)
210
1, 155 (@ 2,124 hours)
815
5,979 (@ 2, 120 hours)
3,224
14,037 (@ 2, 119 hours)
7,509
Terminal
225
199
1,111
735
5,591
2,684
12,920
5.972
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.07
10.7
107
321
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
17.0
21.1
117
120
852
736
2,304
1.983
Max
25.5 (@ 2, 116 hours)
29.3
159 (@ 2,1 16 hours)
145
1, 103 (@ 2,116 hours)
875
2,958 (@ 2, 116 hours)
2,370
Terminal
23.9
27.7
141
127
923
694
2,419
1.828
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1.07
10.7
107
321
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.48
-
7.60
-
30.3
-
51.1
-
Max
2. 17 (@ 2, 116 hours)
-
9.86 (@ 2, 116 hours)
-
36.0 (@ 2, 117 hours)
-
58. 1(@ 2, 117 hours)
-
Terminal
1.90
-
8.42
-
31.1
-
51.8
-
E-147

-------
E.3.1.37. Toth et al (1979) 1  Year
Type:
Strain:
Jody weight:
Sex:
Slice
Swiss/H/Riop
BW = 27g(10weeks
old)
Female and male
Dose:
Route:
Regime:
Simulation time:
7, 700, and 7,000 ng/kg-week
Oral gavage
In gastric tube
Once per week for 1 year (365 days)
8,760 hours
The CADM model was not run because the study duration is longer than the allowed model duration.
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
100
1,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.573
-
14.2
-
91.2
-
Max
1.61 (@ 8,736 hours)
-
1 16 (@ 8,736 hours)
-
1, 108 (@ 8,736 hours)
-
Terminal
0.682
-
15.7
-
99.3
-
LIVER CONCENTRATIONS (tig/kg)
Dose
(ng/kg-day)
adjusted dose
1
100
1,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
94.2
-
7,343
-
70,243
-
Max
131 (@ 8,743 hours)
-
10,134 (@ 8,745 hours)
-
97,658 (@ 8,745 hours)
-
Terminal
123
-
9,604
-
92,506
-
FA T CONCENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
100
1,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
215
-
5,339
-
34,249
-
Max
247 (@ 8,6 13 hours)
-
5,9 14 (@ 8,760 hours)
-
38,828 (@ 8,756 hours)
-
Terminal
245
-
5,914
-
38,807
-
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
100
Model
Emond
CADM
Emond
CADM
Metric
Time-weighted average
23.4
-
929
-
Max
28.4 (@ 8,742 hours)
-
1,1 89 (@ 8,742 hours)
-
Terminal
27.9
-
1,132
-
                                             E-148

-------
1,000
Emond
CADM
7,569
-
10,045 (@ 8,742 hours)
-
9,471
-
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
1
100
1,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1.93
-
31.8
-
78.6
-
Max
2.65 (@ 8,741 hours)
-
58.4 (@ 2 hours)
-
103 (@ 2 hours)
-
Terminal
2.35
-
36.7
-
84.8
-
E.3.1.38.  Van Birgelen et al (1995)
Type:
Strain:
Body weight:
Sex:
Rat
Sprague-Dawley
BW = 150 g
Female
Dose:
Route:
Regime:
Simulation time:
0, 13.5, 26.4, 46.9, 320, and 1,024 ng/kg-day
Oral gavage
Once per day for 13 weeks
2, 184 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
13.5
26.4
46.9
320
1,024
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
7.20
-
11.8
-
18.1
-
86.4
-
250
-
Max
ll.l(@2,160hours)
-
18.6 (@ 2,160 hours)
-
29.6 (@ 2, 160 hours)
-
156 (@ 2,160 hours)
-
470 (@ 2, 160 hours)
-
Terminal
8.47
-
13.5
-
20.5
-
95.4
-
275
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
13.5
26.4
46.9
320
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
1,655
3,228
3,228
6,379
5,719
11,390
38,484
78,166
Max
2,208 (@ 2, 164 hours)
3,802
4,216 (@ 2,164 hours)
7,447
7,366 (@ 2, 164 hours)
13,240
47,999 (@ 2, 164 hours)
90,406
Terminal
2,107
3,802
4,017
7,447
7,008
13,240
45,537
90,406
                                         E-149

-------
1,024
Emond
CADM
121,640
250,307
150,410 (@ 2,164 hours)
289,326
142,510
289,326
FA T CONCENTRA TIONS (tig/kg)
Dose
(ng/kg-day)
adjusted dose
13.5
26.4
46.9
320
1,024
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
669
1,197
1,092
2,119
1,680
3,572
8,027
22,844
23,234
72,506
Max
843 (@ 2, 167 hours)
1,291
1,357 (@ 2,166 hours)
2,290
2,07 1(@ 2, 166 hours)
3,866
9,816 (@ 2,165 hours)
24,800
28,5 19 (@ 2, 165 hours)
78,746
Terminal
835
1,261
1,342
2,240
2,045
3,785
9,639
24,308
27,954
77,195
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
13.5
26.4
46.9
320
1,024
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
132
194
240
367
404
641
2,437
4,292
7,521
13,702
Max
173 (@ 2, 164 hours)
224
308 (@ 2, 164 hours)
423
513 (@ 2,164 hours)
737
3,03 1(@ 2,164 hours)
4,294
9,3 10 (@ 2, 164 hours)
15,714
Terminal
167
224
296
423
492
737
2,887
4,294
8,846
15,714
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
13.5
26.4
46.9
320
1,024
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
19.9
-
29.0
-
38.8
-
79.1
-
97.5
-
Max
24.2 (@ 2, 164 hours)
-
34.3 (@ 2, 164 hours)
-
45.0 (@ 2, 164 hours)
-
85.2 (@ 2, 164 hours)
-
10 1(@ 2, 164 hours)
-
Terminal
23.4
-
33.2
-
43.7
-
84.1
-
101
-
E-150

-------
E.3.1.39. Vanden Heuvel et al (1994)
Type:
Strain:
Body
weight:
Sex:
Rat
Sprague-Dawley
BW = 250 g (10 weeks
old; BW 225 to 275 g)
Female
Dose:
Route:
Regime:
Simulation time:
0.05, 0.1, 1, 10, 100, 1,000, 1 0,000 ng/kg-day
Oral gavage
Single dose
24 hours
The CADM model was not run because the study duration is longer than the allowed model duration.
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.05
0.1
1
10
100
1,000
10,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted
average
0.01
-
0.0113
-
0.106
-
0.883
-
6.45
-
48.3
-
435
-
Max
0.011 (@0 hours)
-
0.022 (@ 0 hours)
-
0.215 (@0 hours)
-
2. 15 (@0 hours)
-
21.5(@0hours)
-
2 16 (@0 hours)
-
2, 166 (@0 hours)
-
Terminal
0.0039
-
0.008
-
0.0723
-
0.583
-
3.85
-
23.9
-
186
-
LIVER CONCENTRATIONS (tig/kg)
Dose
(ng/kg-day)
adjusted dose
0.05
0.1
1
10
100
1,000
10,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.232
-
0.469
-
5.08
-
60.2
-
730
-
8,186
-
84,254
-
Max
0.315(@3 hours)
-
0.631 (@3 hours)
-
6.42 (@ 4 hours)
-
68.7 (@ 5 hours)
-
800 (@ 9 hours)
-
8,919 (@ 11 hours)
-
9 1, 675 (@ 11 hours)
-
Terminal
0.173
0.0140
0.353
0.0320
4.08
0.950
54.1
52.7
719
1,342
8,442
15,967
88,230
162,773
                                            E-151

-------
FA T CON CENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.05
0.1
1
10
100
1,000
10,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.138
-
0.274
-
2.58
-
22.1
-
170
-
1,348
-
12,500
-
Max Terminal
0.215 (@ 24 hours) 0.215
0.780
0.427 (@ 24 hours) 0.427
1.57
3. 97 (@ 24 hours) 3.97
15.3
32.8 (@ 24 hours) 32.8
125
235 (@ 24 hours) 235
739
1,720 (@ 24 hours) 1,720
5,779
15,265 (@ 24 hours) 15,265
55,825
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.05
0.1
1
10
100
1,000
10,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.0269
-
0.0538
-
0.536
-
5.32
-
52.8
-
525
-
5,238
-
Max
0.028 (@ 9 hours)
-
0.057 (@ 9 hours)
-
0.568 (@ 9 hours)
-
5.65 (@ 8 hours)
-
56.3 (@ 7 hours)
-
562 (@ 7 hours)
-
5,610 (@ 7 hours)
-
Terminal
0.0283
0.0450
0.0565
0.0900
0.562
0.900
5.55
9.00
54.4
90.0
538
900
5,353
9,000
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
0.05
0.1
1
10
100
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
0.0194
-
0.0383
-
0.353
-
2.77
-
16.1
-
Max
0.027 (@ 3 hours)
-
0.054 (@ 3 hours)
-
0.506 (@ 3 hours)
-
4.24 (@ 2 hours)
-
26.4 (@ 2 hours)
-
Terminal
0.0142
-
0.0281
-
0.261
-
2.08
-
12.4
-
E-152

-------
1,000
10,000
Emond
CADM
Emond
CADM
57.4
-
100
-
80.2 (@ 1 hour)
-
108 (@ 1 hour)
-
48.5
-
96.1
-
E.3.1.40. Weber et al (1995) C57 Mice
Type:

Strains:
Body
weight:
Sex:
Mouse

C57BL/6J (C57)
24. Ig (7-8 weeks old)
Male
)ose:

Route:
Regime:
Simulation time:
30, 100, 300, 1,000, 3,000, 9,400, 37,500,
75,000, 100,000, 133,000, 150,000, and 235,000
ng/kg
Gavage
Single dose
24 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
3,000
9,400
37,500
75,000
100,000
133,000
150,000
235,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted
average
1.18
-
3.43
-
9.05
-
26.8
-
74.8
-
226
-
917
-
1,929
-
2,668
-
3,725
-
4,301
-
7,426
-
Max
4. 16 (@0 hours)
-
13.9(@0hours)
-
41.6(@0hours)
-
139 (@ 0 hours)
-
4 17 (@0 hours)
-
1,307 (@0 hours)
-
5,223 (@ 0 hours)
-
10,464 (@ 0 hours)
-
13,967 (@ 0 hours)
-
18,603 (@ 0 hours)
-
21,287 (@ 1 hours)
-
39,404 (@ 1 hours)
-
Terminal
0.630
-
1.65
-
3.86
-
9.74
-
23.5
-
63.0
-
231
-
459
-
612
-
815
-
920
-
1,456
-
                                        E-153

-------
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
3,000
9,400
37,500
75,000
100,000
133,000
150,000
235,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
132
68.6
473
416
1,497
2,039
5,194
9,294
15,923
31,419
50,222
102,986
196,690
417,663
379,350
837,656
491,890
1,117,654
629,230
1,487,253
695,520
1,677,652
993,260
2,629,651
Max
146 (@ 7 hours)
68.6
5 17 (@ 10 hours)
416
1, 63 9 (@ 11 hours)
2,039
5,695 (@ 12 hours)
9,294
17,461 (@ 12 hours)
31,419
55,080 (@ 13 hours)
102,986
216,050 (@ 13 hours)
417,663
418,260 (@ 13 hours)
837,656
544,360 (@ 14 hours)
1,117,654
700,560 (@ 14 hours)
1,487,253
777,030 (@ 15 hours)
1,677,652
1,128,600 (@ 16 hours)
2,629,651
Terminal
122
68.6
460
416
1,503
2,039
5,337
9,294
16,565
31,419
52,624
102,986
207,410
417,663
402,930
837,656
525,670
1,117,654
678,650
1,487,253
753,880
1,677,652
1,101,800
2,629,651
FA T CONCENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
3,000
9,400
37,500
75,000
100,000
133,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
71.8
174
216
506
591
1,201
1,810
3,002
5,183
7,593
15,932
21,974
65,208
84,935
137,960
168,938
191,630
224,938
268,900
298,859
Max
103 (@ 24 hours)
174
297 (@ 24 hours)
506
779 (@ 24 hours)
1,201
2,286 (@ 24 hours)
3,002
6,354 (@ 24 hours)
7,593
19,164 (@ 24 hours)
21,974
77,479 (@ 24 hours)
84,935
162,720 (@ 24 hours)
168,938
224,920 (@ 24 hours)
224,938
3 13,670 (@ 23 hours)
298,859
Terminal
103
174
297
506
779
1,201
2,286
3,002
6,354
7,593
19,164
21,974
77,479
84,935
162,720
168,938
224,920
224,938
313,580
298,859
E-154

-------
150,000
235,000
Emond
CADM
Emond
CADM
311,290
336,939
542,350
527,340
362,150 (@ 22 hours) 361,880
336,939 336,939
625,850 (@ 19 hours) 623,390
527,340 527,340
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
3,000
9,400
37,500
75,000
100,000
133,000
150,000
235,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
16.9
24.0
55.9
80.0
166
240
550
800
1,646
2,400
5,141
7,520
20,411
30,000
40,607
60,000
53,951
80,000
71,431
106,400
80,385
120,000
124,740
188,000
Max
18.1(@7hours)
24.0
60.0 (@ 7 hours)
80.0
179 (@ 7 hours)
240
594 (@ 7 hours)
800
1,778 (@ 7 hours)
2,400
5,561 (@ 7 hours)
7,520
22, 102 (@ 7 hours)
30,000
43,991 (@ 6 hours)
60,000
58,459 (@ 6 hours)
80,000
77,4 11(@ 6 hours)
106,400
87,121 (@ 6 hours)
120,000
135,260 (@ 6 hours)
188,000
Terminal
17.5
24.0
57.4
80.0
170
240
560
800
1,668
2,400
5,197
7,520
20,591
30,000
40,914
60,000
54,329
80,000
71,888
106,400
80,879
120,000
125,340
188,000
BOUND LIVER (tig/kg)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
3,000
9,400
37,500
75,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted average
3.37
-
8.62
-
18.6
-
37.6
-
61.3
-
85.4
-
103.3
-
107.6
-
Max
5.79 (@ 3 hours)
-
16.4 (@ 2 hours)
-
36.6 (@ 2 hours)
-
67. 8 (@ 2 hours)
-
91. 8 (@ 2 hours)
-
105 (@ 2 hours)
-
1 11 (@ 2 hours)
-
1 12 (@ 2 hours)
-
Terminal
2.33
-
5.89
-
12.8
-
27.1
-
48.3
-
74.7
-
98.7
-
105.1
-
E-155

-------
100,000
133,000
150,000
235,000
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
108.7
-
109.6
-
109.9
-
110.7
-
1 12 (@ 2 hours)
-
112(@lhour)
-
112(@lhour)
-
113 (@lhour)
-
106.9
-
108.2
-
108.7
-
110.1
-
E.3.1.41.  White et al (1986)
Type:
Strain:
Body weight:
Sex:
Mice
B6C3FJ
BW = 23 g (7 weeks old)
Female
Dose:
Route:
Regime:
Simulation time:
10, 50, 100, 500, 1,000, 2,000 ng/kg-day
Oral gavage
Once per day for 14 days
336 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
10
50
100
500
1,000
2,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted
average
1.09
-
4.08
-
7.14
-
26.8
-
48.7
-
90.6
-
Max
2.73 (@ 3 12 hours)
-
1 1.6 (@ 3 12 hours)
-
21.7 (@ 3 12 hours)
-
96.5 (@ 3 12 hours)
-
187 (@ 3 12 hours)
-
365 (@ 3 12 hours)
-
Terminal
1.42
-
4.98
-
8.44
-
29.8
-
53.1
-
97.5
-
LIVER CONCENTRATIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
10
50
100
500
1,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted
average
216
232
1,279
1,902
2,707
4,285
14,802
24,327
30,278
49,617
Max
375 (@ 3 17 hours)
463
2, 164 (@ 3 17 hours)
3,261
4,525 (@ 3 17 hours)
6,923
24, 165 (@ 3 17 hours)
36,362
49,034 (@ 3 17 hours)
73,145
Terminal
343
463
1,997
3,261
4,184
6,923
22,383
36,362
45,414
73,145
                                         E-156

-------
2,000
Emond
CADM
61,381
100,261
98,703 (@ 3 17 hours)
146,695
91,363
146,695
FA T CON CENTRA TIONS (ng/kg)
Dose
(ng/kg-day)
adjusted dose
10
50
100
500
1,000
2,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted
average
279
338
1,056
1,103
1,854
1,781
7,008
6,119
12,746
11,248
23,691
21,417
Max
507 (@ 336 hours)
537
1,846 (@ 336 hours)
1,564
3, 195 (@ 333 hours)
2,470
1 1,868 (@ 324 hours)
8,594
2 1,566 (@ 323 hours)
15,993
40, 177 (@ 322 hours)
30,726
Terminal
507
537
1,846
1,564
3,195
2,470
11,816
8,594
21,424
15,993
39,843
30,726
BODY BURDEN (ng/kg)
Dose
(ng/kg-day)
adjusted dose
10
50
100
500
1,000
2,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted
average
37.7
51.3
175
222
338
416
1,597
1,887
3,137
3,702
6,186
7,324
Max
65.9 (@ 3 17 hours)
85.9
297 (@ 3 17 hours)
342
570 (@ 3 16 hours)
624
2,637 (@ 3 16 hours)
2,754
5, 153 (@ 3 16 hours)
5,387
10, 1 18 (@ 3 16 hours)
10,643
Terminal
63.8
85.9
284
342
542
624
2,480
2,754
4,830
5,387
9,459
10,643
BOUND LIVER (ng/kg)
Dose
(ng/kg-day)
adjusted dose
10
50
100
500
1,000
2,000
Model
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Emond
CADM
Metric
Time-weighted
average
3.49
-
11.4
-
18.1
-
44.2
-
59.3
-
74.4
-
Max
5.32 (@ 3 16 hours)
-
16.4 (@ 3 17 hours)
-
25. 1(@ 3 17 hours)
-
56.2 (@ 3 17 hours)
-
71. 9 (@ 3 17 hours)
-
86. 1(@ 3 17 hours)
-
Terminal
4.82
-
15.1
-
23.4
-
53.8
-
69.7
-
84.3
-
E-157

-------
E.3.2.  Gestational Studies



E.3.2.1. Bettetal (2007)
Type:
Strain:
Body weight:
Sex:
Rat
ian/Wistar
BW = 85 g
(6 weeks old)
"emale
Dose:
Route:
Regime:
Simulation
time:
2.4, 8, and 46 ng/kg-day with a 0.03 ng/kg-day
background
Dietary exposure
Once per day for 12 weeks prior to mating, during the 2
week mating period, and during gestation
2,352 hours (98 days) prior to gestation + 504 hours (21
days) during gestation for a total simulation of
2,856 hours
Time averages are computed during the gestation period only.
WHOLE BLOOD CONCENTRATIONS (tig/kg) andAUC ((tig/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
2.43
8.03
46.03
Metric
Time-weighted
average
2.20
5.14
18.4
Area under the curve
6,295
14,674
52,584
Max
3. 10 (@ 2,352 hours)
7.3 1(@ 2,352 hours)
28. 1(@ 2,352 hours)
Terminal
2.20
5.08
18.1
LIVER CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
2.43
8.03
46.03
Metric
Time-weighted
average
320
1,040
5,892
Area under the curve
914,290
2,969,800
16,829,000
Max
437 (@ 2,356 hours)
1,349 (@ 2,356 hours)
7,289 (@ 2,356 hours)
Terminal
321
1,042
6,007
FAT CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
2.43
8.03
46.03
Metric
Time-weighted
average
205
478
1,713
Area under the curve
585,530
1,365,100
4,891,500
Max
263 (@ 2,336 hours)
589 (@ 2,335 hours)
2,045 (@ 2,334 hours)
Terminal
211
486
1,745
BODY BURDEN (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
2.43
8.03
46.03
Metric
Time-weighted
average
33.0
90.4
422
Area under the curve
94,390
258,110
1,206,500
Max
44.4 (@ 2,836 hours)
1 17 (@ 2,836 hours)
53 1(@ 2,836 hours)
Terminal
43.4
114
511
FETUS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
2.43
8.03
Metric
Time-weighted
average
3.03
6.65
Area under the curve
8,648
18,999
Max
39.6 (@ 2,530 hours)
86.7 (@ 2,529 hours)
Terminal
6.48
14.4
                                            E-158

-------
46.03
20.9
59,794
272 (@ 2,527 hours)
46.0
BOUND LIVER (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
2.43
8.03
46.03
Metric
Time-weighted
average
7.10
15.1
39.6
Area under the curve
20,289
43,242
113,070
Max
8.98 (@ 2,356 hours)
18.2 (@ 2,356 hours)
44.8 (@ 2,356 hours)
Terminal
7.23
15.4
40.6
E.3.2.2. Hojo et al (2002)
Type:
Strain:
Jody weight
Sex:
Rat
Sprague-Dawley
20ng/kgBW = 271g
60 ng/kg BW = 275 g
180ng/kgBW = 262g
7emale
Dose:
Route:
Regime:
Simulation time
20, 60, and 180 ng/kg
Oral exposure
Single dose on GD 8
216 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
20
60
180
Metric
Time-weighted
average
1.62
4.17
10.7
Area under the curve
39.1
100
258
Max
4.47 (@ 192 hours)
13.3 (@ 192 hours)
40.3 (@ 192 hours)
Terminal
1.02
2.50
5.96
LIVER CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
20
60
180
Metric
Time-weighted
average
128
420
1,364
Area under the curve
20,554
72,340
250,820
Max
144 (@ 198 hours)
465 (@ 200 hours)
1, 497 (@ 201 hours)
Terminal
43.2
147
497
FAT CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
20
60
180
Metric
Time-weighted
average
32.5
86.4
226
Area under the curve
17,253
44,093
108,730
Max
63. 0(@ 281 hours)
161 (@ 284 hours)
3 98 (@ 286 hours)
Terminal
49.4
124
301
BODY BURDEN (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
20
60
180
Metric
Time-weighted
average
10.6
31.8
95.0
Area under the curve
3,054
8,702
24,747
Max
1 1. 3 (@ 200 hours)
33.8 (@ 199 hours)
101 (@ 199 hours)
Terminal
8.67
23.6
63.4
                                         E-159

-------
FETUS (tig/kg) andAUC ((tig/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
20
60
180
Metric
Time-weighted
average
15.9
39.8
96.3
Area under the curve
2,334
5,829
13,866
Max
18.4 (@ 206 hours)
45. 7 (@ 205 hours)
1 10 (@ 203 hours)
Terminal
1.64
4.10
9.72
BOUND LIVER (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
20
60
180
Metric
Time-weighted
average
4.88
11.2
23.6
Area under the curve
759
1,848
4,157
Max
7.74 (@ 194 hours)
18.5 (@ 194 hours)
38.5 (@ 193 hours)
Terminal
1.75
4.26
9.65
E.3.2.3. Ikeda et al (2005)
Type:
Strain:
Jody weight:
Sex:
Rat
Sprague -Dawley
BW = 250 g
(10 weeks old)
Female
Dose:
Route:
Regime:
Simulation
time:
400 ng/kg single dose and 80 ng/kg weekly
maintenance dose
Oral gavage
400 ng/kg single dose, two weekly maintenance doses
)rior to gestation and weekly maintenance doses during
gestation
504 hours (21 days) prior to gestation + 504 hours
(21 days) during gestation for a total simulation of
1,008 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
16.5
Metric
Time-weighted
average
22.9
Area under the curve
23,086
Max
101 (@ 144 hours)
Terminal
10.1
LIVER CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
16.5
Metric
Time-weighted
average
7,755
Area under the curve
7,817,300
Max
17,016 (@ 150 hours)
Terminal
2,698
FAT CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
16.5
Metric
Time-weighted
average
2,087
Area under the curve
2,103,900
Max
3,663 (@ 184 hours)
Terminal
1,028
                                         E-160

-------
BODY BURDEN (tig/kg) andAUC ((tig/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
16.5
Metric
Time-weighted
average
548
Area under the curve
552,590
Max
1,085 (@ 149 hours)
Terminal
262
FETUS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
16.5
Metric
Time-weighted
average
45.9
Area under the curve
46,290
Max
245 (@ 679 hours)
Terminal
30.2
BOUND LIVER (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
16.5
Metric
Time-weighted
average
44.0
Area under the curve
44,361
Max
63.8 (@ 149 hours)
Terminal
26.8
E.3.2.4. Kattainen et al (2001) and Simanainen et al (2004)
Type:
Strain:
Jody weight:
Sex:
Rat
lan/Wistar (Kuopio)
and Long/Evans
Turku/AB) crossing.
BW=190g(BWnot
specified)*
7emale
Dose:
Route:
Regime:
Simulation time:
30, 100, 300, and 1,000 ng/kg
Oral exposure
Single dose on GD 15
360 hours
*Derelanko and Hollinger (1995).
WHOLE BLOOD CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
Metric
Time-weighted
average
2.23
6.25
16.1
46.9
Area under the curve
53.7
150
387
1,128
Max
5.95 (@ 336 hours)
19.8 (@ 336 hours)
59.8 (@ 336 hours)
200 (@ 336 hours)
Terminal
1.36
3.62
8.62
22.7
LIVER CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
Metric
Time-weighted
average
193
713
2,298
8,055
Area under the curve
4,648
17,141
55,266
193,720
Max
2 19 (@ 342 hours)
793 (@ 344 hours)
2,533 (@ 345 hours)
8,83 1(@ 345 hours)
Terminal
175
680
2,267
8,134
                                          E-161

-------
FAT CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
Metric
Time-weighted
average
42.8
123
327
981
Area under the curve
1,027
2,964
7,853
23,588
Max
62.8 (@ 360 hours)
175 (@ 360 hours)
446 (@ 360 hours)
1,289 (@ 360 hours)
Terminal
62.8
175
446
1,289
BODY BURDEN (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
Metric
Time-weighted
average
15.9
52.7
158
524
Area under the curve
382
1,266
3,791
12,612
Max
16.9 (@ 343 hours)
56.2 (@ 343 hours)
168 (@ 343 hours)
561 (@ 343 hours)
Terminal
16.4
54.3
162
538
FETUS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
Metric
Time-weighted
average
4.86
13.2
31.5
82.2
Area under the curve
117
317
758
1,975
Max
6.66 (@ 360 hours)
17.6 (@ 360 hours)
4 1. 2 (@ 360 hours)
104 (@ 360 hours)
Terminal
6.66
17.6
41.2
104
BOUND LIVER (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
Metric
Time-weighted
average
6.57
15.8
31.6
57.1
Area under the curve
158
381
760
1,373
Max
10.7 (@ 338 hours)
26.3 (@ 338 hours)
50.6 (@ 337 hours)
80. 1(@ 337 hours)
Terminal
4.80
11.9
24.7
47.7
E.3.2.5. Keller et al (2007)
Type:
Strain:
Body weight:
Sex:
Mouse
CBA/J and C3H/HeJ
BW = 24 g (BW not
specified)
7emale
)ose:
Route:
Regime:
Simulation time:
10, 100, and 1,000 ng/kg
Oral
Single dose on GD 13
336 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
10
100
1,000
Metric
Time-weighted
average
0.537
4.29
34.1
Area under the curve
12.9
103
820
Max
1. 43 (@ 3 12 hours)
14.3 (@ 3 12 hours)
143 (@ 3 12 hours)
Terminal
0.269
1.95
12.3
                                          E-162

-------
LIVER CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
10
100
1,000
Metric
Time-weighted
average
30.6
371
4,214
Area under the curve
737
8,922
101,360
Max
39.8 (@ 3 16 hours)
421 (@ 3 19 hours)
4,697 (@ 321 hours)
Terminal
22.2
317
3,940
FAT CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
10
100
1,000
Metric
Time-weighted
average
22.4
188
1,591
Area under the curve
538
4,523
38,233
Max
33.3 (@ 336 hours)
264 (@ 336 hours)
2,080 (@ 336 hours)
Terminal
33.3
264
2,080
BODY BURDEN (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
10
100
1,000
Metric
Time-weighted
average
5.57
54.3
530
Area under the curve
134
1,306
12,747
Max
5.99 (@ 3 19 hours)
59.0 (@ 3 18 hours)
581 (@ 3 18 hours)
Terminal
5.72
54.7
524
FETUS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
10
100
1,000
Metric
Time-weighted
average
2.57
21.7
179
Area under the curve
61.7
522
4,312
Max
3.80 (@ 336 hours)
30.0 (@ 334 hours)
233 (@ 329 hours)
Terminal
3.80
29.9
225
BOUND LIVER (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
10
100
1,000
Metric
Time-weighted
average
1.74
11.5
46.7
Area under the curve
41.8
276
1,123
Max
3. 14 (@ 315 hours)
23. 5 (@ 3 14 hours)
79.8 (@ 3 14 hours)
Terminal
1.01
6.99
32.9
E.3.2.6. Li et al (2006) 3 Day
Type:
Strain:
Body weight:
Sex:
Mouse
HIH
BW = 27 g
[Female


(25-28 g)

toose:
(Route:
[Regime:
[Simulation time:
2, 50, and 100 ng/kg-day
Oral
Daily exposure from GD
1 to GD 3
72 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg) andAUC (fng/kg] • hours)
Dose
(ng/kg-day)
adjusted dose
2
Metric
Time-weighted
average
0.159
Area under the curve
11.4
Max
0.392 (@ 48 hours)
Terminal
0.136
                                         E-163

-------
50
100
2.84
5.12
205
369
8.90 (@ 48 hours)
17.3 (@ 48 hours)
2.38
4.20
LIVER CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
2
50
100
Metric
Time-weighted
average
8.98
333
718
Area under the curve
647
23,971
51,738
Max
15.1(@52hours)
539 (@ 53 hours)
1, 156 (@ 53 hours)
Terminal
9.10
402
888
FAT CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
2
50
100
Metric
Time-weighted
average
17.0
315
576
Area under the curve
1,227
22,704
41,460
Max
31.1(@72hours)
548 (@ 72 hours)
984 (@ 72 hours)
Terminal
31.1
548
984
BODY BURDEN (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
2
50
100
Metric
Time-weighted
average
2.29
53.6
105
Area under the curve
165
3,863
7,598
Max
3. 51 (@ 55 hours)
82.2 (@ 54 hours)
162 (@ 53 hours)
Terminal
3.43
77.1
150
FETUS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
2
50
100
Metric
Time-weighted
average
0.0
0.0
0.0
Area under the curve
0
0
0
Max
0.000 (@ 72 hours)
0.000 (@ 72 hours)
0.000 (@ 72 hours)
Terminal
0.00
0.00
0.00
BOUND LIVER (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
2
50
100
Metric
Time-weighted
average
0.538
8.24
13.6
Area under the curve
38.8
594
981
Max
0.864(@51 hours)
13.5 (@2 hours)
23. 7 (@ 2 hours)
Terminal
0.498
8.16
13.6
E.3.2.7. Markowski et al (2001}
Type:
Strain:
Jody weight:
Sex:
Rat
Holtzman rats
BW = 190 g (BW not specified)*
Female
toose:
(Route:
[Regime:
{Simulation time:
20, 60, and 180 ng/kg
Oral exposure
Single dose on GD 18
432 hours
*Derelanko and Hollinger (1995).
                                           E-164

-------
WHOLE BLOOD CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
20
60
180
Metric
Time-weighted
average
1.56
4.03
10.3
Area under the curve
37.5
97.0
248
Max
3. 82 (@ 408 hours)
1 1. 5 (@ 408 hours)
34.8 (@ 408 hours)
Terminal
0.958
2.38
5.72
LIVER CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
20
60
180
Metric
Time-weighted
average
123
409
1,334
Area under the curve
2,959
9,843
32,086
Max
141 (@ 414 hours)
459 (@ 4 15 hours)
1,479 (@ 416 hours)
Terminal
109
382
1,295
FAT CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
20
60
180
Metric
Time-weighted
average
27.9
74.0
195
Area under the curve
670
1,778
4,685
Max
4 1. 6 (@ 432 hours)
107 (@ 432 hours)
273 (@ 432 hours)
Terminal
41.6
107
273
BODY BURDEN (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
20
60
180
Metric
Time-weighted
average
10.6
31.7
94.7
Area under the curve
254
762
2,278
Max
1 1. 2 (@ 4 15 hours)
33.8 (@ 415 hours)
101 (@ 415 hours)
Terminal
10.9
32.7
97.5
FETUS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
20
60
180
Metric
Time-weighted
average
1.26
3.21
7.81
Area under the curve
30.2
77.2
188
Max
1. 80 (@ 432 hours)
4.49 (@ 432 hours)
10.7 (@ 432 hours)
Terminal
1.80
4.49
10.7
BOUND LIVER (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
20
60
180
Metric
Time-weighted
average
4.74
11.0
23.2
Area under the curve
114
265
559
Max
7.59 (@ 4 10 hours)
18.2 (@ 410 hours)
38. 1(@ 409 hours)
Terminal
3.43
8.16
17.7
E-165

-------
E.3.2.8. Mietinnen et al (2006)
Type:
Strain:
Body weight:
Sex:
Rat
Cross-breeding of
Han/Wistar and
^ong-Evans rats
BW= 180 g (11 weeks
old)
"emale
Dose:
Route:
Regime:
Simulation time:
30, 100, 300, and 1,000 ng/kg
Oral exposure
Single dose on GD 15
360 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
Metric
Time-weighted
average
2.22
6.23
16.0
46.6
Area under the curve
53.4
150
386
1,123
Max
5.87 (@ 336 hours)
19.6 (@ 336 hours)
59.0 (@ 336 hours)
198 (@ 336 hours)
Terminal
1.36
3.61
8.61
22.7
LIVER CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
Metric
Time-weighted
average
193
711
2,294
8,042
Area under the curve
4,631
17,096
55,166
193,410
Max
2 19 (@ 342 hours)
791 (@ 344 hours)
2,530 (@ 345 hours)
8,820 (@ 345 hours)
Terminal
174
677
2,260
8,114
FAT CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
Metric
Time-weighted
average
43.0
124
329
987
Area under the curve
1,034
2,984
7,905
23,729
Max
63. 2 (@ 360 hours)
176 (@ 360 hours)
449 (@ 360 hours)
1,296 (@ 360 hours)
Terminal
63.2
176
449
1,296
BODY BURDEN (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
Metric
Time-weighted
average
15.9
52.6
158
524
Area under the curve
381
1,266
3,791
12,609
Max
16.9 (@ 343 hours)
56. 1(@ 343 hours)
168 (@ 343 hours)
561 (@ 343 hours)
Terminal
16.4
54.3
162
538
FETUS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
100
300
Metric
Time-weighted
average
4.83
13.1
31.3
Area under the curve
116
315
753
Max
6.62 (@ 360 hours)
17.5 (@ 360 hours)
41.0(@360hours)
Terminal
6.62
17.5
41.0
                                        E-166

-------
1,000
81.7
1,963
104 (@ 360 hours)
104
BOUND LIVER (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
100
300
1,000
Metric
Time-weighted
average
6.56
15.8
31.6
57.0
Area under the curve
158
381
760
1,372
Max
10.7 (@ 338 hours)
26.3 (@ 338 hours)
50.5 (@ 337 hours)
80. 1(@ 337 hours)
Terminal
4.78
11.9
24.6
47.6
E.3.2.9. Nohara et al (2000)
Type:
Strain:
Body weight:
Sex:
Rat
ioltzman rats
BW= 190g(BWnot
specified)3
"emale
Dose:
Route:
Regime:
Simulation time:
12.5, 50, 200, or 800 ng TCDD/kg
Oral exposure
Single dose on GD 15
360 hours
 Derelanko and Hollinger (1995).
WHOLE BLOOD CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
12.5
50
200
800
Metric
Time-weighted
average
1.03
3.45
11.3
38.1
Area under the curve
24.8
82.9
271
918
Max
2.44 (@ 336 hours)
9.78 (@ 336 hours)
39.2 (@ 336 hours)
158 (@ 336 hours)
Terminal
0.645
2.07
6.25
18.9
LIVER CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
12.5
50
200
800
Metric
Time-weighted
average
73.8
336
1,492
6,389
Area under the curve
1,776
8,084
35,890
153,640
Max
86. 1(@ 341 hours)
378 (@ 343 hours)
1,65 1(@ 344 hours)
7,012 (@ 345 hours)
Terminal
63.6
311
1,454
6,423
FAT CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
12.5
50
200
800
Metric
Time-weighted
average
19.7
67.6
229
803
Area under the curve
473
1,624
5,504
19,292
Max
29.5 (@ 360 hours)
97.8 (@ 360 hours)
3 17 (@ 360 hours)
1,061 (@ 360 hours)
Terminal
29.5
97.8
317
1,061
                                            E-167

-------
BODY BURDEN (tig/kg) andAUC ((tig/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
12.5
50
200
800
Metric
Time-weighted
average
6.62
26.4
105
420
Area under the curve
159
635
2,528
10,092
Max
7.04 (@ 343 hours)
28. 1(@ 343 hours)
1 12 (@ 343 hours)
449 (@ 343 hours)
Terminal
6.88
27.3
108
430
FETUS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
12.5
50
200
800
Metric
Time-weighted
average
2.25
7.43
22.8
68.1
Area under the curve
54.0
179
548
1,638
Max
3. 14 (@ 360 hours)
10. 1(@ 360 hours)
30. 1(@ 360 hours)
87.0 (@ 360 hours)
Terminal
3.14
10.1
30.1
87.0
BOUND LIVER (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
12.5
50
200
800
Metric
Time-weighted
average
3.24
9.66
24.8
51.9
Area under the curve
77.9
232
597
1,248
Max
5. 12 (@ 338 hours)
16.0 (@ 338 hours)
40.7 (@ 337 hours)
75.0 (@ 337 hours)
Terminal
2.32
7.12
19.0
42.7
E.3.2.10. Ohsako et al (2001)
Type:
Strain:
Body weight
Sex:
|Rat
iHoltzmann
ho weeks old (200 g)
[Female
Dose:
Route:
Regime:
Simulation time
12.5, 50, 200, and 800 ng/kg-day
Oral exposure
Single dose on GD 15
384 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
12.5
50
200
800
Metric
Time-weighted
average
1.04
3.47
11.4
38.4
Area under the curve
25.0
83.6
273
925
Max
2.48 (@ 360 hours)
9.93 (@ 360 hours)
39.9 (@ 360 hours)
16 1(@ 360 hours)
Terminal
0.649
2.07
6.26
18.9
LIVER CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
12.5
50
200
800
Metric
Time-weighted
average
74.3
338
1,497
6,402
Area under the curve
1,788
8,126
36,006
153,960
Max
86.5 (@ 365 hours)
379 (@ 367 hours)
1,655 (@ 368 hours)
7,025 (@ 369 hours)
Terminal
64.2
314
1,461
6,443
                                        E-168

-------
FAT CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
12.5
50
200
800
Metric
Time-weighted
average
19.0
65.3
221
777
Area under the curve
457
1,569
5,321
18,671
Max
28.6 (@ 384 hours)
94.7 (@ 384 hours)
307 (@ 384 hours)
1,029 (@ 384 hours)
Terminal
28.6
94.7
307
1,029
BODY BURDEN (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
12.5
50
200
800
Metric
Time-weighted
average
6.63
26.4
105
420
Area under the curve
159
635
2,529
10,093
Max
7.05 (@ 367 hours)
28.2 (@ 367 hours)
1 12 (@ 367 hours)
449 (@ 367 hours)
Terminal
6.89
27.3
108
430
FETUS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
12.5
50
200
800
Metric
Time-weighted
average
1.65
5.44
16.7
49.9
Area under the curve
39.5
131
401
1,200
Max
2.33 (@ 384 hours)
7.48 (@ 384 hours)
22.3 (@ 384 hours)
64.6 (@ 384 hours)
Terminal
2.33
7.48
22.3
64.6
BOUND LIVER (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
12.5
50
200
800
Metric
Time-weighted
average
3.25
9.69
24.9
51.9
Area under the curve
78.3
233
598
1,249
Max
5. 13 (@ 362 hours)
16.0 (@ 362 hours)
40.7 (@ 361 hours)
75.0 (@ 361 hours)
Terminal
2.34
7.16
19.1
42.8
E.3.2.11. Schantz et al (1996) andAmin et al (2000)
Type:
Strain:
Body weight:
Sex:
Rat
Sprague -Dawley
BW = 250 g (BW not
specified)
Female
Dose:
Route:
Regime:
Simulation time:
25 and 100 ng/kg-day
Oral exposure
Daily doses from GD 10-16
384 hours; time averages are calculated
from the beginning of the dosing
WHOLE BLOOD CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
25
100
Metric
Time-weighted
average
3.38
10.6
Area under the curve
487
1,522
Max
8.63 (@ 360 hours)
31.1(@360hours)
Terminal
4.03
12.3
                                        E-169

-------
LIVER CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
25
100
Metric
Time-weighted
average
512
2,374
Area under the curve
73,686
341,960
Max
87 1(@ 365 hours)
4,0 12 (@ 366 hours)
Terminal
778
3,665
FAT CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
25
100
Metric
Time-weighted
average
169
532
Area under the curve
24,323
76,675
Max
306 (@ 384 hours)
950 (@ 384 hours)
Terminal
306
950
BODY BURDEN (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
25
100
Metric
Time-weighted
average
45.1
177
Area under the curve
6,490
25,438
Max
76.6 (@ 365 hours)
298 (@ 365 hours)
Terminal
74.3
287
FETUS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
25
100
Metric
Time-weighted
average
25.2
74.1
Area under the curve
3,627
10,672
Max
30.4 (@ 343 hours)
88. 1(@ 342 hours)
Terminal
27.3
77.9
BOUND LIVER (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
25
100
Metric
Time-weighted
average
9.99
25.2
Area under the curve
1,439
3,632
Max
14.4 (@ 364 hours)
34.2 (@ 364 hours)
Terminal
12.8
31.6
E.3.2.12. Seo et al (1995)
Type:
Strain:
Body weight:
Sex:
Rat
Sprague -Dawley
BW= 190g(BWnot
specified)
Female
Dose:
Route:
Regime:
Simulation time:
25 and 100 ng/kg-day
Oral exposure
Daily doses from GD 10-16
384 hours; time averages are calculated
from the beginning of the dosing
WHOLE BLOOD CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
25
100
Metric
Time-weighted
average
3.33
10.4
Area under the curve
479
1,498
Max
8.25 (@ 360 hours)
29.6 (@ 360 hours)
Terminal
4.00
12.2
                                         E-170

-------
LIVER CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
25
100
Metric
Time-weighted
average
504
2,347
Area under the curve
72,592
337,970
Max
86 1(@ 365 hours)
3,978 (@ 365 hours)
Terminal
767
3,627
FAT CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
25
100
Metric
Time-weighted
average
172
542
Area under the curve
24,807
78,097
Max
3 10 (@ 384 hours)
962 (@ 384 hours)
Terminal
310
962
BODY BURDEN (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
25
100
Metric
Time-weighted
average
45.0
176
Area under the curve
6,486
25,387
Max
76.5 (@ 365 hours)
298 (@ 365 hours)
Terminal
74.2
287
FETUS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
25
100
Metric
Time-weighted
average
24.7
72.6
Area under the curve
3,551
10,456
Max
29.8 (@ 343 hours)
86.6 (@ 342 hours)
Terminal
26.8
76.8
BOUND LIVER (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
25
100
Metric
Time-weighted
average
9.90
25.0
Area under the curve
1,426
3,607
Max
14.3 (@ 364 hours)
34. 1(@ 364 hours)
Terminal
12.7
31.4
E.3.2.13. Smith et al (1976)
Type:
Strain:
Body weight:
Sex:
Mouse
CF-1
Mean 28-29 g
(GD6)
Female
Dose:
Route:
Regime:
Simulation time:
1, 10, 100, 1,000, and 3,000 ng/kg-day
Gavage
Daily doses from GD 6-15
360 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
1
10
100
1,000
Metric
Time-weighted
average
0.124
1.01
7.11
50.6
Area under the curve
29.8
243
1,707
12,145
Max
0.274 (@ 336 hours)
2.47 (@ 336 hours)
21. 1(@ 336 hours)
188 (@ 336 hours)
Terminal
0.136
1.08
7.16
47.4
                                        E-171

-------
3,000
138
33,142
554 (@ 336 hours)
127
LIVER CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
1
10
100
1,000
3,000
Metric
Time-weighted
average
7.23
101
1,381
16,329
50,491
Area under the curve
1,735
24,194
331,570
3,919,700
12,120,000
Max
12.3 (@ 339 hours)
167 (@ 340 hours)
2,196 (@ 341 hours)
25,189 (@ 341 hours)
77,170 (@ 341 hours)
Terminal
8.71
128
1,788
20,932
64,246
FAT CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
1
10
100
1,000
3,000
Metric
Time-weighted
average
22.8
188
1,344
9,659
26,368
Area under the curve
5,477
45,189
322,580
2,318,300
6,328,900
Max
41. 1(@ 360 hours)
33 1(@ 360 hours)
2,289 (@ 360 hours)
16,123 (@ 357 hours)
44,004 (@ 355 hours)
Terminal
41.1
331
2,289
16,117
43,959
BODY BURDEN (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
1
10
100
1,000
3,000
Metric
Time-weighted
average
3.07
28.1
246
2,211
6,446
Area under the curve
736
6,745
59,076
530,720
1,547,200
Max
5.48 (@ 342 hours)
49. 1(@ 341 hours)
415 (@ 340 hours)
3,626 (@ 340 hours)
10,500 (@ 340 hours)
Terminal
5.40
47.5
390
3,316
9,535
FETUS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
1
10
100
1,000
3,000
Metric
Time-weighted
average
1.90
15.4
105
659
1,663
Area under the curve
456
3,703
25,190
158,110
399,230
Max
2.45 (@ 274 hours)
1 9. 9 (@ 249 hours)
137 (@ 247 hours)
880 (@ 246 hours)
2,254 (@ 246 hours)
Terminal
2.15
16.9
111
686
1,744
BOUND LIVER (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
1
10
100
1,000
3,000
Metric
Time-weighted
average
0.428
3.30
18.5
61.9
85.2
Area under the curve
103
791
4,435
14,855
20,450
Max
0.694 (@ 339 hours)
4.93 (@ 340 hours)
24.9 (@ 340 hours)
79.8 (@ 122 hours)
98.9 (@ 122 hours)
Terminal
0.485
3.77
20.9
67.4
90.1
E-172

-------
E.3.2.14. Sparschu et al (1971)
Type:
Strain:
Body weight:
Sex:
Rat
Sprague -Dawley
BW = 295 g
(290-300 g)
Female
Dose:
Route:
Regime:
Simulation time:
30, 125, 500, 2,000, and 8,000 ng/kg-day
Gavage
Daily doses from GD 6-15
360 hours
WHOLE BLOOD CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
125
500
2,000
8,000
Metric
Time-weighted
average
5.09
16.3
52.9
188
732
Area under the curve
1,222
3,908
12,690
45,188
175,750
Max
12.4 (@ 336 hours)
45.5 (@ 336 hours)
168 (@ 336 hours)
646 (@ 336 hours)
2,572 (@ 336 hours)
Terminal
6.52
20.4
65.6
235
928
LIVER CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
125
500
2,000
8,000
Metric
Time-weighted
average
946
4,480
19,233
79,288
316,550
Area under the curve
227,090
1,075,300
4,616,400
19,031,000
75,979,000
Max
1,636 (@ 341 hours)
7,644 (@ 341 hours)
32,428 (@ 341 hours)
132,390 (@ 341 hours)
522,920 (@ 341 hours)
Terminal
1,507
7,105
30,252
123,500
485,720
FAT CONCENTRATIONS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
125
500
2,000
8,000
Metric
Time-weighted
average
317
1,016
3,295
11,671
45,125
Area under the curve
75,978
243,930
790,910
2,801,200
10,831,000
Max
547 (@ 360 hours)
1,739 (@ 360 hours)
5,663 (@ 360 hours)
20,374 (@ 360 hours)
80,136 (@ 360 hours)
Terminal
547
1,739
5,663
20,374
80,136
BODY BURDEN (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
125
500
2,000
8,000
Metric
Time-weighted
average
80.6
324
1,266
4,996
19,780
Area under the curve
19,348
77,864
303,960
1,199,100
4,747,500
Max
140 (@ 341 hours)
559 (@ 341 hours)
2, 169 (@ 341 hours)
8,527 (@ 341 hours)
33,634 (@ 340 hours)
Terminal
136
537
2,071
8,117
31,926
FETUS (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
Metric
Time-weighted
average
53.8
Area under the curve
12,906
Max
69. 5 (@ 247 hours)
Terminal
54.1
                                        E-173

-------
125
500
2,000
8,000
156
430
1,311
4,694
37,342
103,180
314,680
1,126,700
202 (@ 246 hours)
560 (@ 245 hours)
1, 72 1(@ 269 hours)
6,255 (@ 269 hours)
153
424
1,334
4,943
BOUND LIVER (ng/kg) andAUC ((ng/kg) • hours)
Dose
(ng/kg-day)
adjusted dose
30
125
500
2,000
8,000
Metric
Time-weighted
average
14.4
34.5
64.0
91.2
106
Area under the curve
3,452
8,279
15,367
21,890
25,389
Max
20.7 (@ 340 hours)
46.2 (@ 340 hours)
77.7 (@ 341 hours)
100 (@ 341 hours)
109 (@ 341 hours)
Terminal
19.2
43.9
75.8
99.2
109
E-174

-------
       Table E-l.
       articles
Model input parameters potentially addressed by selected
Articles
Aylward et al. (2005a)
Aylward et al. (2005b)
Avlward et al. (2009)
3ohonowych and Denison
(2007)
Boverhofetal. (2005)
Connor and Aylward (2006)
Heinzl et al. (2007)
Irigarav et al. (2005)
Kerger et al. (2006)
Kerger et al. (2007)
Kim et al. (2003)
Korenaga et al. (2007)
Korkalainen et al. (2004)
Kransler et al. (2007)
Maruyama et al. (2002)
Maruvama et al. (2003)
Vlaruyama and Aoki (2006)
Milbrath et al. (2009)
Moser and McLachlan (2002)
Mullerova and
Kopeckv(2007)
Nadal et al. (2009)
Nohara et al. (2006)
Olsman et al. (2007)
Saghir et al. (2005)
Schecter et al. (2003)
Staskal et al. (2005)
Tovoshiba et al. (2004)
Wilkes et al. (2008)
Model input parameters potentially addressed
Absorption
•
•












•
•
•











Desorption
•
•
















•









Distribution
•
•




•
•
•

•



•
•
•
•

•



•


•

Elimination
•
•
•











•
•
•
•
•

•


•
•



Kinetics
•
•






•








•


•


•




Induction
CYPlAl



•
•






•













•
•
•
Interspecies
differences



•
•
•

•



•
•
•



•



•






Age Differences





•


•
•


•
•










•



j-*.i ji
hydrocarbon



•

•
•














•
•


•
•

Mode of action




























Partition
coefficient




























Partition coefficient estimates and CYP parameter value estimates were derived from Wang et al., (2000: 1997) and
Santostefano et al. (1998).
                                            E-175

-------
E.4. RESPONSE SURFACE TABLES
       In order to calculate human equivalent doses, the human model must be run with a daily
intake which gives average blood concentrations which match the average concentrations in the
rodent models. However, such calculation can require numerous human model runs with
repeated intake adjustments in order to reach the target blood concentrations. To facilitate this
process, a response surface was created for the human model. In the response surface, numerous
intakes were run and the blood, fat, and body burden average concentrations were recorded.
These tables can then be used to  estimate the intake which would give a target blood
concentration.  The two closest intakes are found and the intake is estimated by linearly
interpolating between the two doses.  Then, this intake is run through the human model to
confirm that the average blood concentration is within a specified tolerance of the target blood
concentration.
       For the current analysis, three different response surfaces were created: nongestational
lifetime to be used with long-term animal bioassays, nongestational 5 year average runs to be
used with shorter term animal bioassays, and gestational to be used with gestational animal
bioassays. All three response surfaces are shown in the following tables.
                                         E-176

-------
E.4.1.  Nongestational Lifetime
Nongestational Lifetime Average
Intake (ng/kg-
day)
1.03E-09
1.09E-09
1.16E-09
1.23E-09
1.30E-09
1.38E-09
1.46E-09
1.55E-09
1.64E-09
1.74E-09
1.84E-09
1.95E-09
2.07E-09
2.20E-09
2.33E-09
2.47E-09
2.62E-09
2.77E-09
2.94E-09
3.12E-09
3.30E-09
3.50E-09
3.71E-09
3.93E-09
4.17E-09
4.42E-09
4.68E-09
4.97E-09
5.26E-09
5.58E-09
5.91E-09
6.27E-09
6.65E-09
7.04E-09
7.47E-09
Fat (ng/kg)
2.78E-05
2.95E-05
3.13E-05
3.32E-05
3.52E-05
3.73E-05
3.95E-05
4.19E-05
4.44E-05
4.70E-05
4.99E-05
5.28E-05
5.60E-05
5.94E-05
6.29E-05
6.67E-05
7.07E-05
7.49E-05
7.94E-05
8.42E-05
8.92E-05
9.46E-05
l.OOE-04
1.06E-04
1.13E-04
1.19E-04
1.27E-04
1.34E-04
1.42E-04
1.51E-04
1.60E-04
1.69E-04
1.79E-04
1.90E-04
2.02E-04
Body
burden
(ng/kg)
8.69E-06
9.21E-06
9.77E-06
1.04E-05
1.10E-05
1.16E-05
1.23E-05
1.31E-05
1.38E-05
1.47E-05
1.56E-05
1.65E-05
1.75E-05
1.85E-05
1.96E-05
2.08E-05
2.21E-05
2.34E-05
2.48E-05
2.63E-05
2.79E-05
2.95E-05
3.13E-05
3.32E-05
3.52E-05
3.73E-05
3.95E-05
4.19E-05
4.44E-05
4.70E-05
4.99E-05
5.28E-05
5.60E-05
5.94E-05
6.29E-05
Blood
(ng/kg)
2.93E-07
3.11E-07
3.30E-07
3.49E-07
3.70E-07
3.93E-07
4.16E-07
4.41E-07
4.68E-07
4.96E-07
5.25E-07
5.57E-07
5.90E-07
6.26E-07
6.63E-07
7.03E-07
7.45E-07
7.90E-07
8.37E-07
8.87E-07
9.40E-07
9.97E-07
1.06E-06
1.12E-06
1.19E-06
1.26E-06
1.33E-06
1.41E-06
1.50E-06
1.59E-06
1.68E-06
1.78E-06
1.89E-06
2.00E-06
2.12E-06
Nongestational Lifetime Average
Intake (ng/kg-
day)
7.92E-09
8.39E-09
8.89E-09
9.43E-09
9.99E-09
1.06E-08
1.12E-08
1.19E-08
1.26E-08
1.34E-08
1.42E-08
1.50E-08
1.59E-08
1.69E-08
1.79E-08
1.90E-08
2.01E-08
2.13E-08
2.26E-08
2.39E-08
2.54E-08
2.69E-08
2.85E-08
3.02E-08
3.20E-08
3.40E-08
3.60E-08
3.82E-08
4.05E-08
4.29E-08
4.55E-08
4.82E-08
5.11E-08
5.41E-08
5.74E-08
6.08E-08
6.45E-08
Fat (ng/kg)
2.14E-04
2.26E-04
2.40E-04
2.54E-04
2.70E-04
2.86E-04
3.03E-04
3.21E-04
3.40E-04
3.61E-04
3.82E-04
4.05E-04
4.29E-04
4.55E-04
4.82E-04
5.11E-04
5.42E-04
5.74E-04
6.08E-04
6.45E-04
6.83E-04
7.24E-04
7.67E-04
8.13E-04
8.62E-04
9.13E-04
9.68E-04
1.03E-03
1.09E-03
1.15E-03
1.22E-03
1.29E-03
1.37E-03
1.45E-03
1.54E-03
1.63E-03
1.73E-03
Body
burden
(ng/kg)
6.67E-05
7.07E-05
7.49E-05
7.94E-05
8.42E-05
8.92E-05
9.46E-05
l.OOE-04
1.06E-04
1.13E-04
1.19E-04
1.26E-04
1.34E-04
1.42E-04
1.51E-04
1.60E-04
1.69E-04
1.79E-04
1.90E-04
2.01E-04
2.13E-04
2.26E-04
2.40E-04
2.54E-04
2.69E-04
2.85E-04
3.02E-04
3.21E-04
3.40E-04
3.60E-04
3.81E-04
4.04E-04
4.28E-04
4.54E-04
4.81E-04
5.10E-04
5.40E-04
Blood
(ng/kg)
2.25E-06
2.39E-06
2.53E-06
2.68E-06
2.84E-06
3.01E-06
3.19E-06
3.38E-06
3.58E-06
3.80E-06
4.03E-06
4.27E-06
4.52E-06
4.79E-06
5.08E-06
5.38E-06
5.71E-06
6.05E-06
6.41E-06
6.79E-06
7.20E-06
7.63E-06
8.08E-06
8.57E-06
9.08E-06
9.62E-06
1.02E-05
1.08E-05
1.15E-05
1.21E-05
1.29E-05
1.36E-05
1.44E-05
1.53E-05
1.62E-05
1.72E-05
1.82E-05
Nongestational Lifetime Average
Intake (ng/kg-
day)
6.84E-08
7.25E-08
7.68E-08
8.14E-08
8.63E-08
9.15E-08
9.70E-08
1.03E-07
1.09E-07
1.15E-07
1.22E-07
1.30E-07
1.38E-07
1.46E-07
1.55E-07
1.64E-07
1.74E-07
1.84E-07
1.95E-07
2.07E-07
2.19E-07
2.32E-07
2.46E-07
2.61E-07
2.77E-07
2.93E-07
3.11E-07
3.30E-07
3.49E-07
3.70E-07
3.93E-07
4.16E-07
4.41E-07
4.68E-07
4.96E-07
5.25E-07
5.57E-07
Fat (ng/kg)
1.83E-03
1.94E-03
2.06E-03
2.18E-03
2.31E-03
2.45E-03
2.59E-03
2.75E-03
2.91E-03
3.08E-03
3.27E-03
3.46E-03
3.67E-03
3.88E-03
4.11E-03
4.36E-03
4.62E-03
4.89E-03
5.18E-03
5.49E-03
5.81E-03
6.16E-03
6.52E-03
6.91E-03
7.32E-03
7.75E-03
8.21E-03
8.69E-03
9.21E-03
9.75E-03
1.03E-02
1.09E-02
1.16E-02
1.23E-02
1.30E-02
1.37E-02
1.46E-02
Body
burden
(ng/kg)
5.73E-04
6.07E-04
6.43E-04
6.81E-04
7.22E-04
7.65E-04
8.11E-04
8.59E-04
9.10E-04
9.64E-04
1.02E-03
1.08E-03
1.15E-03
1.22E-03
1.29E-03
1.36E-03
1.45E-03
1.53E-03
1.62E-03
1.72E-03
1.82E-03
1.93E-03
2.04E-03
2.17E-03
2.29E-03
2.43E-03
2.57E-03
2.73E-03
2.89E-03
3.06E-03
3.24E-03
3.43E-03
3.63E-03
3.85E-03
4.08E-03
4.32E-03
4.57E-03
Blood
(ng/kg)
1.93E-05
2.04E-05
2.17E-05
2.30E-05
2.43E-05
2.58E-05
2.73E-05
2.89E-05
3.06E-05
3.25E-05
3.44E-05
3.64E-05
3.86E-05
4.09E-05
4.33E-05
4.59E-05
4.86E-05
5.15E-05
5.46E-05
5.78E-05
6.12E-05
6.49E-05
6.87E-05
7.28E-05
7.71E-05
8.16E-05
8.65E-05
9.16E-05
9.70E-05
1.03E-04
1.09E-04
1.15E-04
1.22E-04
1.29E-04
1.37E-04
1.45E-04
1.53E-04

-------
oo
Nongestational Lifetime Average
Intake (ng/kg-
day)
5.90E-07
6.26E-07
6.63E-07
7.03E-07
7.45E-07
7.90E-07
8.37E-07
8.88E-07
9.41E-07
9.97E-07
1.01E-06
1.03E-06
1.04E-06
1.06E-06
1.07E-06
1.09E-06
1.11E-06
1.12E-06
1.14E-06
1.16E-06
1.17E-06
1.19E-06
1.21E-06
1.23E-06
1.24E-06
1.26E-06
1.28E-06
1.30E-06
1.32E-06
1.34E-06
1.36E-06
1.38E-06
1.40E-06
1.42E-06
1.44E-06
1.46E-06
1.49E-06
Fat (ng/kg)
1.54E-02
1.63E-02
1.73E-02
1.83E-02
1.93E-02
2.05E-02
2.17E-02
2.29E-02
2.43E-02
2.57E-02
2.61E-02
2.64E-02
2.68E-02
2.72E-02
2.76E-02
2.80E-02
2.84E-02
2.88E-02
2.92E-02
2.97E-02
3.01E-02
3.05E-02
3.10E-02
3.14E-02
3.19E-02
3.23E-02
3.28E-02
3.33E-02
3.37E-02
3.42E-02
3.47E-02
3.52E-02
3.57E-02
3.62E-02
3.67E-02
3.73E-02
3.78E-02
Body
burden
(ng/kg)
4.84E-03
5.13E-03
5.43E-03
5.75E-03
6.09E-03
6.45E-03
6.82E-03
7.22E-03
7.65E-03
8.10E-03
8.21E-03
8.33E-03
8.45E-03
8.58E-03
8.70E-03
8.83E-03
8.96E-03
9.09E-03
9.22E-03
9.35E-03
9.49E-03
9.63E-03
9.77E-03
9.91E-03
1.01E-02
1.02E-02
1.03E-02
1.05E-02
1.06E-02
1.08E-02
1.10E-02
1.11E-02
1.13E-02
1.14E-02
1.16E-02
1.18E-02
1.19E-02
Blood
(ng/kg)
1.62E-04
1.72E-04
1.82E-04
1.93E-04
2.04E-04
2.16E-04
2.28E-04
2.42E-04
2.56E-04
2.71E-04
2.75E-04
2.79E-04
2.83E-04
2.87E-04
2.91E-04
2.95E-04
2.99E-04
3.04E-04
3.08E-04
3.12E-04
3.17E-04
3.21E-04
3.26E-04
3.31E-04
3.36E-04
3.40E-04
3.45E-04
3.50E-04
3.55E-04
3.60E-04
3.66E-04
3.71E-04
3.76E-04
3.82E-04
3.87E-04
3.93E-04
3.98E-04
Nongestational Lifetime Average
Intake (ng/kg-
day)
1.53E-06
1.58E-06
1.62E-06
1.67E-06
1.72E-06
1.77E-06
1.83E-06
1.88E-06
1.94E-06
2.00E-06
2.06E-06
2.12E-06
2.18E-06
2.25E-06
2.32E-06
2.39E-06
2.46E-06
2.53E-06
2.61E-06
2.68E-06
2.76E-06
2.85E-06
2.93E-06
3.02E-06
3.11E-06
3.21E-06
3.30E-06
3.40E-06
3.50E-06
3.61E-06
3.72E-06
3.83E-06
3.94E-06
4.06E-06
4.18E-06
4.31E-06
4.44E-06
Fat (ng/kg)
3.89E-02
4.00E-02
4.12E-02
4.24E-02
4.36E-02
4.49E-02
4.61E-02
4.75E-02
4.88E-02
5.02E-02
5.17E-02
5.32E-02
5.47E-02
5.63E-02
5.79E-02
5.95E-02
6.12E-02
6.30E-02
6.48E-02
6.66E-02
6.85E-02
7.05E-02
7.25E-02
7.46E-02
7.67E-02
7.89E-02
8.11E-02
8.34E-02
8.58E-02
8.82E-02
9.07E-02
9.33E-02
9.59E-02
9.86E-02
1.01E-01
1.04E-01
1.07E-01
Body
burden
(ng/kg)
1.23E-02
1.27E-02
1.30E-02
1.34E-02
1.38E-02
1.42E-02
1.46E-02
1.50E-02
1.55E-02
1.59E-02
1.64E-02
1.68E-02
1.73E-02
1.78E-02
1.84E-02
1.89E-02
1.94E-02
2.00E-02
2.06E-02
2.12E-02
2.18E-02
2.24E-02
2.30E-02
2.37E-02
2.44E-02
2.51E-02
2.58E-02
2.65E-02
2.73E-02
2.81E-02
2.89E-02
2.97E-02
3.06E-02
3.14E-02
3.23E-02
3.33E-02
3.42E-02
Blood
(ng/kg)
4.10E-04
4.22E-04
4.34E-04
4.46E-04
4.59E-04
4.72E-04
4.86E-04
5.00E-04
5.14E-04
5.29E-04
5.44E-04
5.60E-04
5.76E-04
5.93E-04
6.10E-04
6.27E-04
6.45E-04
6.64E-04
6.83E-04
7.02E-04
7.22E-04
7.43E-04
7.64E-04
7.86E-04
8.08E-04
8.31E-04
8.54E-04
8.79E-04
9.04E-04
9.29E-04
9.55E-04
9.82E-04
1.01E-03
1.04E-03
1.07E-03
1.10E-03
1.13E-03
Nongestational Lifetime Average
Intake (ng/kg-
day)
4.57E-06
4.71E-06
4.85E-06
4.99E-06
5.14E-06
5.30E-06
5.46E-06
5.62E-06
5.79E-06
5.96E-06
6.14E-06
6.33E-06
6.52E-06
6.71E-06
6.91E-06
7.12E-06
7.33E-06
7.55E-06
7.78E-06
8.01E-06
8.25E-06
8.50E-06
8.76E-06
9.02E-06
9.29E-06
9.57E-06
9.86E-06
1.02E-05
1.05E-05
1.08E-05
1.11E-05
1.14E-05
1.18E-05
1.21E-05
1.25E-05
1.29E-05
1.32E-05
Fat (ng/kg)
1.10E-01
1.13E-01
1.16E-01
1.20E-01
1.23E-01
1.27E-01
1.30E-01
1.34E-01
1.37E-01
1.41E-01
1.45E-01
1.49E-01
1.53E-01
1.58E-01
1.62E-01
1.66E-01
1.71E-01
1.76E-01
1.81E-01
1.86E-01
1.91E-01
1.96E-01
2.01E-01
2.07E-01
2.12E-01
2.18E-01
2.24E-01
2.30E-01
2.37E-01
2.43E-01
2.50E-01
2.56E-01
2.63E-01
2.71E-01
2.78E-01
2.85E-01
2.93E-01
Body
burden
(ng/kg)
3.52E-02
3.62E-02
3.72E-02
3.83E-02
3.94E-02
4.05E-02
4.16E-02
4.28E-02
4.40E-02
4.53E-02
4.65E-02
4.78E-02
4.92E-02
5.06E-02
5.20E-02
5.35E-02
5.50E-02
5.65E-02
5.81E-02
5.97E-02
6.14E-02
6.31E-02
6.49E-02
6.67E-02
6.86E-02
7.05E-02
7.24E-02
7.45E-02
7.65E-02
7.86E-02
8.08E-02
8.31E-02
8.54E-02
8.77E-02
9.01E-02
9.26E-02
9.52E-02
Blood
(ng/kg)
1.16E-03
1.19E-03
1.23E-03
1.26E-03
1.30E-03
1.33E-03
1.37E-03
1.41E-03
1.45E-03
1.49E-03
1.53E-03
1.57E-03
1.62E-03
1.66E-03
1.71E-03
1.75E-03
1.80E-03
1.85E-03
1.90E-03
1.95E-03
2.01E-03
2.06E-03
2.12E-03
2.18E-03
2.24E-03
2.30E-03
2.36E-03
2.43E-03
2.49E-03
2.56E-03
2.63E-03
2.70E-03
2.77E-03
2.85E-03
2.93E-03
3.01E-03
3.09E-03

-------
Nongestational Lifetime Average
Intake (ng/kg-
day)
1.36E-05
1.41E-05
1.45E-05
1.49E-05
1.54E-05
1.58E-05
1.63E-05
1.68E-05
1.73E-05
1.78E-05
1.83E-05
1.89E-05
1.95E-05
2.00E-05
2.06E-05
2.13E-05
2.19E-05
2.25E-05
2.32E-05
2.39E-05
2.46E-05
2.54E-05
2.61E-05
2.69E-05
2.77E-05
2.86E-05
2.94E-05
3.03E-05
3.12E-05
3.21E-05
3.31E-05
3.41E-05
3.51E-05
3.62E-05
3.73E-05
3.84E-05
3.95E-05
Fat (ng/kg)
3.01E-01
3.09E-01
3.17E-01
3.26E-01
3.34E-01
3.43E-01
3.53E-01
3.62E-01
3.72E-01
3.81E-01
3.92E-01
4.02E-01
4.13E-01
4.23E-01
4.35E-01
4.46E-01
4.58E-01
4.70E-01
4.82E-01
4.94E-01
5.07E-01
5.21E-01
5.34E-01
5.48E-01
5.62E-01
5.77E-01
5.92E-01
6.07E-01
6.22E-01
6.38E-01
6.55E-01
6.72E-01
6.89E-01
7.06E-01
7.25E-01
7.43E-01
7.62E-01
Body
burden
(ng/kg)
9.78E-02
l.OOE-01
1.03E-01
1.06E-01
1.09E-01
1.12E-01
1.15E-01
1.18E-01
1.21E-01
1.25E-01
1.28E-01
1.32E-01
1.35E-01
1.39E-01
1.43E-01
1.46E-01
1.50E-01
1.54E-01
1.59E-01
1.63E-01
1.67E-01
1.72E-01
1.76E-01
1.81E-01
1.86E-01
1.91E-01
1.96E-01
2.01E-01
2.06E-01
2.12E-01
2.18E-01
2.23E-01
2.29E-01
2.35E-01
2.42E-01
2.48E-01
2.54E-01
Blood
(ng/kg)
3.17E-03
3.25E-03
3.34E-03
3.43E-03
3.52E-03
3.62E-03
3.71E-03
3.81E-03
3.91E-03
4.02E-03
4.12E-03
4.23E-03
4.34E-03
4.46E-03
4.58E-03
4.70E-03
4.82E-03
4.95E-03
5.07E-03
5.21E-03
5.34E-03
5.48E-03
5.62E-03
5.77E-03
5.92E-03
6.07E-03
6.23E-03
6.39E-03
6.55E-03
6.72E-03
6.90E-03
7.07E-03
7.25E-03
7.44E-03
7.63E-03
7.82E-03
8.02E-03
Nongestational Lifetime Average
Intake (ng/kg-
day)
4.07E-05
4.19E-05
4.32E-05
4.45E-05
4.58E-05
4.72E-05
4.86E-05
5.01E-05
5.16E-05
5.31E-05
5.47E-05
5.64E-05
5.81E-05
5.98E-05
6.16E-05
6.34E-05
6.54E-05
6.73E-05
6.93E-05
7.14E-05
7.36E-05
7.58E-05
7.80E-05
8.04E-05
8.28E-05
8.53E-05
8.78E-05
9.05E-05
9.32E-05
9.60E-05
9.89E-05
1.02E-04
1.05E-04
1.08E-04
1.11E-04
1.15E-04
1.18E-04
Fat (ng/kg)
7.81E-01
8.01E-01
8.21E-01
8.42E-01
8.63E-01
8.84E-01
9.07E-01
9.29E-01
9.53E-01
9.76E-01
l.OOE+00
1.03E+00
1.05E+00
1.08E+00
1.10E+00
1.13E+00
1.16E+00
1.19E+00
1.22E+00
1.25E+00
1.28E+00
1.31E+00
1.34E+00
1.37E+00
1.40E+00
1.44E+00
1.47E+00
1.51E+00
1.55E+00
1.58E+00
1.62E+00
1.66E+00
1.70E+00
1.74E+00
1.78E+00
1.82E+00
1.87E+00
Body
burden
(ng/kg)
2.61E-01
2.68E-01
2.75E-01
2.82E-01
2.90E-01
2.97E-01
3.05E-01
3.13E-01
3.21E-01
3.29E-01
3.38E-01
3.47E-01
3.56E-01
3.65E-01
3.74E-01
3.84E-01
3.94E-01
4.04E-01
4.14E-01
4.25E-01
4.36E-01
4.47E-01
4.58E-01
4.70E-01
4.82E-01
4.94E-01
5.07E-01
5.19E-01
5.33E-01
5.46E-01
5.60E-01
5.74E-01
5.89E-01
6.04E-01
6.19E-01
6.34E-01
6.50E-01
Blood
(ng/kg)
8.22E-03
8.43E-03
8.64E-03
8.86E-03
9.08E-03
9.31E-03
9.55E-03
9.78E-03
l.OOE-02
1.03E-02
1.05E-02
1.08E-02
1.11E-02
1.13E-02
1.16E-02
1.19E-02
1.22E-02
1.25E-02
1.28E-02
1.31E-02
1.34E-02
1.38E-02
1.41E-02
1.44E-02
1.48E-02
1.51E-02
1.55E-02
1.59E-02
1.63E-02
1.67E-02
1.71E-02
1.75E-02
1.79E-02
1.83E-02
1.88E-02
1.92E-02
1.96E-02
Nongestational Lifetime Average
Intake (ng/kg-
day)
1.22E-04
1.25E-04
1.29E-04
1.33E-04
1.37E-04
1.41E-04
1.45E-04
1.50E-04
1.54E-04
1.59E-04
1.63E-04
1.68E-04
1.73E-04
1.79E-04
1.84E-04
1.89E-04
1.95E-04
2.01E-04
2.07E-04
2.13E-04
2.20E-04
2.26E-04
2.33E-04
2.40E-04
2.47E-04
2.55E-04
2.62E-04
2.70E-04
2.78E-04
2.86E-04
2.95E-04
3.04E-04
3.13E-04
3.22E-04
3.32E-04
3.42E-04
3.52E-04
Fat (ng/kg)
1.91E+00
1.96E+00
2.00E+00
2.05E+00
2.10E+00
2.15E+00
2.20E+00
2.25E+00
2.30E+00
2.35E+00
2.41E+00
2.46E+00
2.52E+00
2.58E+00
2.64E+00
2.70E+00
2.76E+00
2.82E+00
2.89E+00
2.96E+00
3.02E+00
3.09E+00
3.16E+00
3.23E+00
3.31E+00
3.38E+00
3.46E+00
3.54E+00
3.62E+00
3.70E+00
3.78E+00
3.86E+00
3.95E+00
4.04E+00
4.13E+00
4.22E+00
4.31E+00
Body
burden
(ng/kg)
6.66E-01
6.83E-01
7.00E-01
7.17E-01
7.35E-01
7.53E-01
7.72E-01
7.91E-01
8.11E-01
8.31E-01
8.51E-01
8.72E-01
8.94E-01
9.16E-01
9.39E-01
9.62E-01
9.85E-01
1.01E+00
1.03E+00
1.06E+00
1.09E+00
1.11E+00
1.14E+00
1.17E+00
1.20E+00
1.23E+00
1.26E+00
1.29E+00
1.32E+00
1.35E+00
1.38E+00
1.42E+00
1.45E+00
1.49E+00
1.52E+00
1.56E+00
1.59E+00
Blood
(ng/kg)
2.01E-02
2.06E-02
2.11E-02
2.16E-02
2.21E-02
2.26E-02
2.31E-02
2.36E-02
2.42E-02
2.48E-02
2.53E-02
2.59E-02
2.65E-02
2.71E-02
2.78E-02
2.84E-02
2.90E-02
2.97E-02
3.04E-02
3.11E-02
3.18E-02
3.25E-02
3.33E-02
3.40E-02
3.48E-02
3.56E-02
3.64E-02
3.72E-02
3.81E-02
3.89E-02
3.98E-02
4.07E-02
4.16E-02
4.25E-02
4.34E-02
4.44E-02
4.54E-02

-------
oo
o
Nongestational Lifetime Average
Intake (ng/kg-
day)
3.63E-04
3.74E-04
3.85E-04
3.97E-04
4.08E-04
4.21E-04
4.33E-04
4.46E-04
4.60E-04
4.74E-04
4.88E-04
5.02E-04
5.17E-04
5.33E-04
5.49E-04
5.65E-04
5.82E-04
6.00E-04
6.18E-04
6.36E-04
6.55E-04
6.75E-04
6.95E-04
7.16E-04
7.38E-04
7.60E-04
7.83E-04
8.06E-04
8.30E-04
8.55E-04
8.81E-04
9.07E-04
9.21E-04
9.35E-04
9.49E-04
9.63E-04
9.69E-04
Fat (ng/kg)
4.41E+00
4.50E+00
4.60E+00
4.71E+00
4.81E+00
4.92E+00
5.02E+00
5.13E+00
5.25E+00
5.36E+00
5.48E+00
5.60E+00
5.72E+00
5.85E+00
5.97E+00
6.10E+00
6.24E+00
6.37E+00
6.51E+00
6.65E+00
6.79E+00
6.94E+00
7.09E+00
7.24E+00
7.39E+00
7.55E+00
7.71E+00
7.87E+00
8.04E+00
8.21E+00
8.38E+00
8.56E+00
8.65E+00
8.74E+00
8.84E+00
8.93E+00
8.97E+00
Body
burden
(ng/kg)
1.63E+00
1.67E+00
1.71E+00
1.75E+00
1.80E+00
1.84E+00
1.89E+00
1.93E+00
1.98E+00
2.03E+00
2.07E+00
2.12E+00
2.18E+00
2.23E+00
2.28E+00
2.34E+00
2.39E+00
2.45E+00
2.51E+00
2.57E+00
2.63E+00
2.69E+00
2.76E+00
2.82E+00
2.89E+00
2.96E+00
3.03E+00
3.10E+00
3.17E+00
3.25E+00
3.33E+00
3.41E+00
3.45E+00
3.49E+00
3.53E+00
3.57E+00
3.59E+00
Blood
(ng/kg)
4.64E-02
4.74E-02
4.85E-02
4.95E-02
5.06E-02
5.17E-02
5.29E-02
5.40E-02
5.52E-02
5.64E-02
5.77E-02
5.89E-02
6.02E-02
6.15E-02
6.29E-02
6.42E-02
6.56E-02
6.71E-02
6.85E-02
7.00E-02
7.15E-02
7.30E-02
7.46E-02
7.62E-02
7.78E-02
7.94E-02
8.11E-02
8.29E-02
8.46E-02
8.64E-02
8.82E-02
9.01E-02
9.11E-02
9.20E-02
9.30E-02
9.40E-02
9.44E-02
Nongestational Lifetime Average
Intake (ng/kg-
day)
9.77E-04
9.84E-04
9.91E-04
9.98E-04
1.01E-03
1.02E-03
1.04E-03
1.05E-03
1.07E-03
1.08E-03
1.10E-03
1.12E-03
1.13E-03
1.15E-03
1.17E-03
1.18E-03
1.20E-03
1.22E-03
1.24E-03
1.26E-03
1.27E-03
1.29E-03
1.31E-03
1.33E-03
1.35E-03
1.37E-03
1.39E-03
1.41E-03
1.43E-03
1.46E-03
1.48E-03
1.50E-03
1.52E-03
1.54E-03
1.57E-03
1.59E-03
1.61E-03
Fat (ng/kg)
9.02E+00
9.07E+00
9.12E+00
9.16E+00
9.21E+00
9.31E+00
9.41E+00
9.50E+00
9.60E+00
9.70E+00
9.81E+00
9.91E+00
l.OOE+01
1.01E+01
1.02E+01
1.03E+01
1.04E+01
1.05E+01
1.07E+01
1.08E+01
1.09E+01
1.10E+01
1.11E+01
1.12E+01
1.13E+01
1.14E+01
1.16E+01
1.17E+01
1.18E+01
1.19E+01
1.21E+01
1.22E+01
1.23E+01
1.24E+01
1.26E+01
1.28E+01
1.31E+01
Body
burden
(ng/kg)
3.61E+00
3.63E+00
3.66E+00
3.68E+00
3.70E+00
3.74E+00
3.79E+00
3.83E+00
3.88E+00
3.92E+00
3.97E+00
4.02E+00
4.06E+00
4.11E+00
4.16E+00
4.21E+00
4.26E+00
4.31E+00
4.36E+00
4.41E+00
4.46E+00
4.52E+00
4.57E+00
4.62E+00
4.68E+00
4.73E+00
4.79E+00
4.85E+00
4.91E+00
4.96E+00
5.02E+00
5.08E+00
5.14E+00
5.20E+00
5.26E+00
5.39E+00
5.54E+00
Blood
(ng/kg)
9.49E-02
9.54E-02
9.59E-02
9.64E-02
9.69E-02
9.80E-02
9.90E-02
l.OOE-01
1.01E-01
1.02E-01
1.03E-01
1.04E-01
1.05E-01
1.06E-01
1.08E-01
1.09E-01
1.10E-01
1.11E-01
1.12E-01
1.13E-01
1.14E-01
1.16E-01
1.17E-01
1.18E-01
1.19E-01
1.20E-01
1.22E-01
1.23E-01
1.24E-01
1.26E-01
1.27E-01
1.28E-01
1.29E-01
1.31E-01
1.32E-01
1.35E-01
1.38E-01
Nongestational Lifetime Average
Intake (ng/kg-
day)
1.64E-03
1.66E-03
1.69E-03
1.71E-03
1.74E-03
1.76E-03
1.79E-03
1.82E-03
1.84E-03
1.87E-03
1.90E-03
1.93E-03
1.96E-03
1.99E-03
2.02E-03
2.08E-03
2.14E-03
2.20E-03
2.27E-03
2.34E-03
2.41E-03
2.48E-03
2.55E-03
2.63E-03
2.71E-03
2.79E-03
2.87E-03
2.96E-03
3.05E-03
3.14E-03
3.23E-03
3.33E-03
3.43E-03
3.53E-03
3.64E-03
3.75E-03
3.81E-03
Fat (ng/kg)
1.33E+01
1.33E+01
1.34E+01
1.35E+01
1.36E+01
1.37E+01
1.38E+01
1.39E+01
1.41E+01
1.43E+01
1.46E+01
1.49E+01
1.49E+01
1.50E+01
1.51E+01
1.54E+01
1.56E+01
1.59E+01
1.62E+01
1.66E+01
1.69E+01
1.72E+01
1.76E+01
1.79E+01
1.83E+01
1.87E+01
1.91E+01
1.94E+01
1.98E+01
2.02E+01
2.07E+01
2.11E+01
2.15E+01
2.19E+01
2.23E+01
2.29E+01
2.31E+01
Body
burden
(ng/kg)
5.60E+00
5.62E+00
5.67E+00
5.73E+00
5.77E+00
5.80E+00
5.87E+00
5.94E+00
6.01E+00
6.11E+00
6.31E+00
6.45E+00
6.42E+00
6.48E+00
6.55E+00
6.66E+00
6.77E+00
6.93E+00
7.09E+00
7.25E+00
7.42E+00
7.60E+00
7.78E+00
7.96E+00
8.15E+00
8.35E+00
8.55E+00
8.75E+00
8.96E+00
9.17E+00
9.41E+00
9.63E+00
9.85E+00
1.01E+01
1.03E+01
1.06E+01
1.08E+01
Blood
(ng/kg)
1.39E-01
1.40E-01
1.41E-01
1.42E-01
1.43E-01
1.44E-01
1.45E-01
1.47E-01
1.48E-01
1.50E-01
1.54E-01
1.57E-01
1.57E-01
1.58E-01
1.59E-01
1.62E-01
1.64E-01
1.68E-01
1.71E-01
1.74E-01
1.78E-01
1.81E-01
1.85E-01
1.89E-01
1.93E-01
1.97E-01
2.00E-01
2.05E-01
2.09E-01
2.13E-01
2.18E-01
2.22E-01
2.26E-01
2.31E-01
2.35E-01
2.41E-01
2.43E-01

-------
w
oo
Nongestational Lifetime Average
Intake (ng/kg-
day)
3.86E-03
3.98E-03
4.10E-03
4.22E-03
4.35E-03
4.48E-03
4.61E-03
4.75E-03
4.89E-03
5.04E-03
5.19E-03
5.35E-03
5.51E-03
5.67E-03
5.84E-03
5.93E-03
6.02E-03
6.20E-03
6.38E-03
6.57E-03
6.77E-03
6.98E-03
7.18E-03
7.40E-03
7.51E-03
7.62E-03
7.85E-03
8.09E-03
8.33E-03
8.58E-03
8.71E-03
8.84E-03
9.10E-03
9.37E-03
9.66E-03
9.94E-03
1.02E-02
Fat (ng/kg)
2.32E+01
2.36E+01
2.40E+01
2.44E+01
2.48E+01
2.53E+01
2.58E+01
2.63E+01
2.68E+01
2.75E+01
2.82E+01
2.89E+01
2.96E+01
3.04E+01
3.10E+01
3.13E+01
3.16E+01
3.22E+01
3.29E+01
3.34E+01
3.40E+01
3.45E+01
3.54E+01
3.61E+01
3.64E+01
3.68E+01
3.75E+01
3.82E+01
3.89E+01
3.96E+01
4.00E+01
4.04E+01
4.12E+01
4.20E+01
4.29E+01
4.37E+01
4.46E+01
Body
burden
(ng/kg)
1.08E+01
1.10E+01
1.12E+01
1.14E+01
1.17E+01
1.19E+01
1.22E+01
1.25E+01
1.28E+01
1.32E+01
1.36E+01
1.41E+01
1.45E+01
1.50E+01
1.53E+01
1.55E+01
1.57E+01
1.61E+01
1.65E+01
1.68E+01
1.72E+01
1.75E+01
1.80E+01
1.85E+01
1.87E+01
1.89E+01
1.93E+01
1.98E+01
2.02E+01
2.07E+01
2.10E+01
2.12E+01
2.17E+01
2.23E+01
2.28E+01
2.34E+01
2.39E+01
Blood
(ng/kg)
2.44E-01
2.48E-01
2.52E-01
2.56E-01
2.61E-01
2.66E-01
2.71E-01
2.77E-01
2.82E-01
2.89E-01
2.97E-01
3.04E-01
3.11E-01
3.20E-01
3.26E-01
3.29E-01
3.32E-01
3.39E-01
3.46E-01
3.51E-01
3.58E-01
3.63E-01
3.72E-01
3.80E-01
3.83E-01
3.87E-01
3.94E-01
4.02E-01
4.09E-01
4.17E-01
4.21E-01
4.25E-01
4.34E-01
4.42E-01
4.51E-01
4.60E-01
4.69E-01
Nongestational Lifetime Average
Intake (ng/kg-
day)
1.06E-02
1.09E-02
1.12E-02
1.15E-02
1.19E-02
1.22E-02
1.26E-02
1.30E-02
1.34E-02
1.38E-02
1.42E-02
1.46E-02
1.50E-02
1.55E-02
1.60E-02
1.64E-02
1.69E-02
1.74E-02
1.80E-02
1.85E-02
1.91E-02
1.96E-02
2.02E-02
2.08E-02
2.14E-02
2.21E-02
2.28E-02
2.34E-02
2.41E-02
2.49E-02
2.56E-02
2.64E-02
2.72E-02
2.80E-02
2.88E-02
2.97E-02
3.06E-02
Fat (ng/kg)
4.54E+01
4.63E+01
4.73E+01
4.83E+01
4.93E+01
5.02E+01
5.11E+01
5.22E+01
5.31E+01
5.42E+01
5.53E+01
5.66E+01
5.76E+01
5.87E+01
5.99E+01
6.10E+01
6.22E+01
6.34E+01
6.46E+01
6.59E+01
6.71E+01
6.88E+01
7.01E+01
7.14E+01
7.26E+01
7.40E+01
7.55E+01
7.69E+01
7.85E+01
8.00E+01
8.16E+01
8.32E+01
8.48E+01
8.64E+01
8.81E+01
8.98E+01
9.15E+01
Body
burden
(ng/kg)
2.45E+01
2.51E+01
2.58E+01
2.65E+01
2.72E+01
2.78E+01
2.84E+01
2.91E+01
2.98E+01
3.06E+01
3.14E+01
3.24E+01
3.31E+01
3.39E+01
3.47E+01
3.56E+01
3.65E+01
3.73E+01
3.83E+01
3.92E+01
4.02E+01
4.15E+01
4.25E+01
4.35E+01
4.44E+01
4.55E+01
4.67E+01
4.78E+01
4.91E+01
5.04E+01
5.16E+01
5.30E+01
5.43E+01
5.56E+01
5.70E+01
5.85E+01
5.99E+01
Blood
(ng/kg)
4.78E-01
4.87E-01
4.98E-01
5.08E-01
5.19E-01
5.28E-01
5.38E-01
5.49E-01
5.59E-01
5.70E-01
5.82E-01
5.95E-01
6.07E-01
6.18E-01
6.30E-01
6.42E-01
6.55E-01
6.67E-01
6.80E-01
6.93E-01
7.06E-01
7.24E-01
7.38E-01
7.52E-01
7.64E-01
7.79E-01
7.94E-01
8.10E-01
8.26E-01
8.42E-01
8.59E-01
8.76E-01
8.93E-01
9.09E-01
9.27E-01
9.45E-01
9.63E-01
Nongestational Lifetime Average
Intake (ng/kg-
day)
3.15E-02
3.24E-02
3.34E-02
3.44E-02
3.54E-02
3.65E-02
3.76E-02
3.87E-02
3.99E-02
4.11E-02
4.23E-02
4.36E-02
4.49E-02
4.63E-02
4.76E-02
4.91E-02
5.05E-02
5.21E-02
5.36E-02
5.52E-02
5.69E-02
5.86E-02
6.03E-02
6.22E-02
6.40E-02
6.59E-02
6.79E-02
7.00E-02
7.21E-02
7.42E-02
7.64E-02
7.87E-02
8.11E-02
8.35E-02
8.60E-02
8.86E-02
9.13E-02
Fat (ng/kg)
9.33E+01
9.51E+01
9.69E+01
9.88E+01
1.01E+02
1.03E+02
1.05E+02
1.07E+02
1.09E+02
1.11E+02
1.13E+02
1.15E+02
1.18E+02
1.20E+02
1.22E+02
1.25E+02
1.27E+02
1.30E+02
1.32E+02
1.34E+02
1.37E+02
1.40E+02
1.43E+02
1.45E+02
1.48E+02
1.51E+02
1.54E+02
1.57E+02
1.60E+02
1.63E+02
1.66E+02
1.70E+02
1.73E+02
1.76E+02
1.80E+02
1.84E+02
1.87E+02
Body
burden
(ng/kg)
6.14E+01
6.30E+01
6.46E+01
6.62E+01
6.79E+01
6.97E+01
7.15E+01
7.33E+01
7.52E+01
7.71 E+01
7.91E+01
8.12E+01
8.33E+01
8.54E+01
8.77E+01
9.00E+01
9.24E+01
9.47E+01
9.71E+01
9.95E+01
1.02E+02
1.05E+02
1.08E+02
1.10E+02
1.13E+02
1.16E+02
1.19E+02
1.22E+02
1.26E+02
1.29E+02
1.32E+02
1.36E+02
1.39E+02
1.43E+02
1.47E+02
1.51E+02
1.55E+02
Blood
(ng/kg)
9.81E-01
l.OOE+00
1.02E+00
1.04E+00
1.06E+00
1.08E+00
1.10E+00
1.12E+00
1.14E+00
1.17E+00
1.19E+00
1.21E+00
1.24E+00
1.26E+00
1.29E+00
1.31E+00
1.34E+00
1.36E+00
1.39E+00
1.41E+00
1.44E+00
1.47E+00
1.50E+00
1.53E+00
1.56E+00
1.59E+00
1.62E+00
1.65E+00
1.69E+00
1.72E+00
1.75E+00
1.79E+00
1.82E+00
1.86E+00
1.89E+00
1.93E+00
1.97E+00

-------
oo
to
Nongestational Lifetime Average
Intake (ng/kg-
day)
9.40E-02
9.68E-02
9.97E-02
1.03E-01
1.06E-01
1.09E-01
1.12E-01
1.16E-01
1.19E-01
1.23E-01
1.26E-01
1.30E-01
1.34E-01
1.38E-01
1.42E-01
1.46E-01
1.51E-01
1.55E-01
1.60E-01
1.65E-01
1.70E-01
1.75E-01
1.80E-01
1.86E-01
1.91E-01
1.97E-01
2.03E-01
2.09E-01
2.15E-01
2.22E-01
2.28E-01
2.35E-01
2.42E-01
2.49E-01
2.57E-01
2.65E-01
2.72E-01
Fat (ng/kg)
1.91E+02
1.95E+02
1.98E+02
2.02E+02
2.06E+02
2.10E+02
2.14E+02
2.19E+02
2.23E+02
2.27E+02
2.32E+02
2.36E+02
2.41E+02
2.45E+02
2.50E+02
2.55E+02
2.60E+02
2.65E+02
2.71E+02
2.76E+02
2.81E+02
2.87E+02
2.93E+02
2.98E+02
3.04E+02
3.10E+02
3.16E+02
3.23E+02
3.29E+02
3.35E+02
3.42E+02
3.49E+02
3.56E+02
3.63E+02
3.70E+02
3.77E+02
3.85E+02
Body
burden
(ng/kg)
1.59E+02
1.63E+02
1.68E+02
1.72E+02
1.77E+02
1.81E+02
1.86E+02
1.91E+02
1.96E+02
2.02E+02
2.07E+02
2.12E+02
2.18E+02
2.24E+02
2.30E+02
2.36E+02
2.43E+02
2.49E+02
2.56E+02
2.63E+02
2.70E+02
2.77E+02
2.85E+02
2.92E+02
3.00E+02
3.08E+02
3.17E+02
3.25E+02
3.34E+02
3.43E+02
3.53E+02
3.62E+02
3.72E+02
3.82E+02
3.93E+02
4.03E+02
4.14E+02
Blood
(ng/kg)
2.01E+00
2.05E+00
2.09E+00
2.13E+00
2.17E+00
2.21E+00
2.26E+00
2.30E+00
2.35E+00
2.39E+00
2.44E+00
2.48E+00
2.53E+00
2.58E+00
2.63E+00
2.69E+00
2.74E+00
2.79E+00
2.85E+00
2.90E+00
2.96E+00
3.02E+00
3.08E+00
3.14E+00
3.20E+00
3.26E+00
3.33E+00
3.39E+00
3.46E+00
3.53E+00
3.60E+00
3.67E+00
3.74E+00
3.82E+00
3.89E+00
3.97E+00
4.05E+00
Nongestational Lifetime Average
Intake (ng/kg-
day)
2.81E-01
2.89E-01
2.98E-01
3.07E-01
3.16E-01
3.25E-01
3.35E-01
3.45E-01
3.56E-01
3.66E-01
3.77E-01
3.89E-01
4.00E-01
4.12E-01
4.25E-01
4.37E-01
4.50E-01
4.64E-01
4.78E-01
4.92E-01
5.07E-01
5.22E-01
5.38E-01
5.54E-01
5.71E-01
5.88E-01
6.05E-01
6.23E-01
6.42E-01
6.61E-01
6.81E-01
7.02E-01
7.23E-01
7.44E-01
7.67E-01
7.90E-01
8.13E-01
Fat (ng/kg)
3.93E+02
4.00E+02
4.08E+02
4.16E+02
4.25E+02
4.33E+02
4.42E+02
4.51E+02
4.60E+02
4.69E+02
4.78E+02
4.88E+02
4.98E+02
5.08E+02
5.18E+02
5.28E+02
5.39E+02
5.50E+02
5.61E+02
5.72E+02
5.84E+02
5.96E+02
6.08E+02
6.20E+02
6.33E+02
6.46E+02
6.59E+02
6.72E+02
6.86E+02
7.00E+02
7.14E+02
7.29E+02
7.44E+02
7.59E+02
7.75E+02
7.91E+02
8.07E+02
Body
burden
(ng/kg)
4.26E+02
4.38E+02
4.50E+02
4.62E+02
4.75E+02
4.88E+02
5.01E+02
5.15E+02
5.29E+02
5.44E+02
5.59E+02
5.74E+02
5.90E+02
6.07E+02
6.23E+02
6.41E+02
6.58E+02
6.77E+02
6.96E+02
7.15E+02
7.35E+02
7.55E+02
7.76E+02
7.98E+02
8.20E+02
8.43E+02
8.67E+02
8.91E+02
9.16E+02
9.42E+02
9.68E+02
9.95E+02
1.02E+03
1.05E+03
1.08E+03
1.11E+03
1.14E+03
Blood
(ng/kg)
4.13E+00
4.21E+00
4.30E+00
4.38E+00
4.47E+00
4.56E+00
4.65E+00
4.74E+00
4.84E+00
4.94E+00
5.03E+00
5.14E+00
5.24E+00
5.34E+00
5.45E+00
5.56E+00
5.67E+00
5.79E+00
5.90E+00
6.02E+00
6.14E+00
6.27E+00
6.40E+00
6.53E+00
6.66E+00
6.79E+00
6.93E+00
7.07E+00
7.22E+00
7.37E+00
7.52E+00
7.67E+00
7.83E+00
7.99E+00
8.15E+00
8.32E+00
8.49E+00
Nongestational Lifetime Average
Intake (ng/kg-
day)
8.38E-01
8.63E-01
8.89E-01
9.16E-01
9.43E-01
9.71E-01
l.OOE+00
1.03E+00
1.06E+00
1.09E+00
1.13E+00
1.16E+00
1.19E+00
1.23E+00
1.27E+00
1.31E+00
1.34E+00
1.38E+00
1.43E+00
1.47E+00
1.51E+00
1.56E+00
1.61E+00
1.65E+00
1.70E+00
1.75E+00
1.81E+00
1.86E+00
1.92E+00
1.97E+00
2.03E+00
2.09E+00
2.16E+00
2.22E+00
2.29E+00
2.36E+00
2.43E+00
Fat (ng/kg)
8.24E+02
8.41E+02
8.58E+02
8.76E+02
8.94E+02
9.13E+02
9.32E+02
9.51E+02
9.71E+02
9.91E+02
1.01E+03
1.03E+03
1.05E+03
1.08E+03
1.10E+03
1.12E+03
1.15E+03
1.17E+03
1.20E+03
1.22E+03
1.25E+03
1.27E+03
1.30E+03
1.33E+03
1.36E+03
1.39E+03
1.42E+03
1.45E+03
1.48E+03
1.51E+03
1.54E+03
1.58E+03
1.61E+03
1.65E+03
1.68E+03
1.72E+03
1.76E+03
Body
burden
(ng/kg)
1.18E+03
1.21E+03
1.24E+03
1.28E+03
1.31E+03
1.35E+03
1.39E+03
1.43E+03
1.47E+03
1.51E+03
1.55E+03
1.60E+03
1.64E+03
1.69E+03
1.74E+03
1.79E+03
1.84E+03
1.89E+03
1.94E+03
2.00E+03
2.06E+03
2.12E+03
2.18E+03
2.24E+03
2.30E+03
2.37E+03
2.44E+03
2.51E+03
2.58E+03
2.65E+03
2.73E+03
2.80E+03
2.89E+03
2.97E+03
3.05E+03
3.14E+03
3.23E+03
Blood
(ng/kg)
8.67E+00
8.85E+00
9.03E+00
9.22E+00
9.41E+00
9.60E+00
9.80E+00
l.OOE+01
1.02E+01
1.04E+01
1.06E+01
1.09E+01
1.11E+01
1.13E+01
1.16E+01
1.18E+01
1.21E+01
1.23E+01
1.26E+01
1.29E+01
1.31E+01
1.34E+01
1.37E+01
1.40E+01
1.43E+01
1.46E+01
1.49E+01
1.52E+01
1.56E+01
1.59E+01
1.62E+01
1.66E+01
1.70E+01
1.73E+01
1.77E+01
1.81E+01
1.85E+01

-------
oo
Nongestational Lifetime Average
Intake (ng/kg-
day)
2.50E+00
2.58E+00
2.65E+00
2.73E+00
2.82E+00
2.90E+00
2.99E+00
3.08E+00
3.17E+00
3.26E+00
3.36E+00
3.46E+00
3.57E+00
3.67E+00
3.78E+00
3.90E+00
4.01E+00
4.13E+00
4.26E+00
4.39E+00
4.52E+00
4.65E+00
4.79E+00
4.94E+00
5.08E+00
5.24E+00
5.39E+00
5.56E+00
5.72E+00
5.89E+00
6.07E+00
6.25E+00
6.44E+00
6.63E+00
6.83E+00
7.04E+00
7.25E+00
Fat (ng/kg)
1.80E+03
1.84E+03
1.88E+03
1.92E+03
1.96E+03
2.00E+03
2.05E+03
2.09E+03
2.14E+03
2.19E+03
2.24E+03
2.29E+03
2.34E+03
2.39E+03
2.45E+03
2.51E+03
2.56E+03
2.62E+03
2.68E+03
2.74E+03
2.81E+03
2.87E+03
2.94E+03
3.01E+03
3.08E+03
3.15E+03
3.22E+03
3.30E+03
3.38E+03
3.46E+03
3.54E+03
3.62E+03
3.71E+03
3.80E+03
3.89E+03
3.98E+03
4.08E+03
Body
burden
(ng/kg)
3.32E+03
3.42E+03
3.52E+03
3.62E+03
3.73E+03
3.83E+03
3.94E+03
4.06E+03
4.17E+03
4.30E+03
4.42E+03
4.55E+03
4.68E+03
4.81E+03
4.95E+03
5.10E+03
5.25E+03
5.40E+03
5.55E+03
5.72E+03
5.88E+03
6.05E+03
6.23E+03
6.41E+03
6.60E+03
6.79E+03
6.99E+03
7.19E+03
7.40E+03
7.61E+03
7.84E+03
8.07E+03
8.30E+03
8.54E+03
8.79E+03
9.05E+03
9.31E+03
Blood
(ng/kg)
1.89E+01
1.93E+01
1.97E+01
2.02E+01
2.06E+01
2.11E+01
2.16E+01
2.20E+01
2.25E+01
2.30E+01
2.36E+01
2.41E+01
2.46E+01
2.52E+01
2.58E+01
2.64E+01
2.70E+01
2.76E+01
2.82E+01
2.89E+01
2.95E+01
3.02E+01
3.09E+01
3.16E+01
3.24E+01
3.31E+01
3.39E+01
3.47E+01
3.55E+01
3.64E+01
3.72E+01
3.81E+01
3.90E+01
3.99E+01
4.09E+01
4.19E+01
4.29E+01
Nongestational Lifetime Average
Intake (ng/kg-
day)
7.47E+00
7.69E+00
7.92E+00
8.16E+00
8.40E+00
8.66E+00
8.92E+00
9.18E+00
9.46E+00
9.74E+00
l.OOE+01
1.06E+01
1.13E+01
1.20E+01
1.27E+01
1.34E+01
1.42E+01
1.51E+01
1.60E+01
1.70E+01
1.80E+01
1.90E+01
2.02E+01
2.14E+01
2.27E+01
2.40E+01
2.55E+01
2.70E+01
2.86E+01
3.04E+01
3.22E+01
3.41E+01
3.62E+01
3.83E+01
4.06E+01
4.31E+01
4.57E+01
Fat (ng/kg)
4.18E+03
4.28E+03
4.38E+03
4.49E+03
4.60E+03
4.71E+03
4.82E+03
4.94E+03
5.07E+03
5.19E+03
5.32E+03
5.58E+03
5.86E+03
6.16E+03
6.47E+03
6.80E+03
7.14E+03
7.51E+03
7.90E+03
8.31E+03
8.74E+03
9.20E+03
9.68E+03
1.02E+04
1.07E+04
1.13E+04
1.19E+04
1.26E+04
1.32E+04
1.40E+04
1.47E+04
1.55E+04
1.64E+04
1.73E+04
1.82E+04
1.92E+04
2.03E+04
Body
burden
(ng/kg)
9.59E+03
9.87E+03
1.02E+04
1.05E+04
1.08E+04
1.11E+04
1.14E+04
1.17E+04
1.21E+04
1.24E+04
1.28E+04
1.35E+04
1.43E+04
1.52E+04
1.61E+04
1.70E+04
1.80E+04
1.91E+04
2.02E+04
2.14E+04
2.26E+04
2.40E+04
2.54E+04
2.69E+04
2.85E+04
3.01E+04
3.19E+04
3.38E+04
3.58E+04
3.79E+04
4.01E+04
4.25E+04
4.50E+04
4.77E+04
5.05E+04
5.34E+04
5.66E+04
Blood
(ng/kg)
4.39E+01
4.50E+01
4.61E+01
4.72E+01
4.84E+01
4.95E+01
5.08E+01
5.20E+01
5.33E+01
5.46E+01
5.60E+01
5.88E+01
6.17E+01
6.48E+01
6.81E+01
7.15E+01
7.52E+01
7.90E+01
8.31E+01
8.74E+01
9.20E+01
9.68E+01
1.02E+02
1.07E+02
1.13E+02
1.19E+02
1.25E+02
1.32E+02
1.39E+02
1.47E+02
1.55E+02
1.63E+02
1.72E+02
1.82E+02
1.92E+02
2.02E+02
2.14E+02
Nongestational Lifetime Average
Intake (ng/kg-
day)
4.84E+01
5.13E+01
5.44E+01
5.76E+01
6.11E+01
6.48E+01
6.86E+01
7.28E+01
7.71E+01
8.18E+01
8.67E+01
9.19E+01
9.74E+01
1.03E+02
1.09E+02
1.16E+02
1.23E+02
1.30E+02
1.38E+02
1.46E+02
Fat (ng/kg)
2.14E+04
2.27E+04
2.39E+04
2.53E+04
2.67E+04
2.82E+04
2.98E+04
3.15E+04
3.33E+04
3.53E+04
3.73E+04
3.94E+04
4.17E+04
4.41E+04
4.67E+04
4.94E+04
5.23E+04
5.54E+04
5.86E+04
6.20E+04
Body
burden
(ng/kg)
5.99E+04
6.34E+04
6.71E+04
7.11E+04
7.52E+04
7.97E+04
8.43E+04
8.93E+04
9.45E+04
l.OOE+05
1.06E+05
1.12E+05
1.19E+05
1.25E+05
1.33E+05
1.40E+05
1.48E+05
1.57E+05
1.66E+05
1.76E+05
Blood
(ng/kg)
2.26E+02
2.38E+02
2.52E+02
2.66E+02
2.81E+02
2.97E+02
3.14E+02
3.32E+02
3.51E+02
3.71E+02
3.92E+02
4.15E+02
4.39E+02
4.64E+02
4.91E+02
5.20E+02
5.50E+02
5.82E+02
6.17E+02
6.53E+02

-------
    E.4.2.  Nongestational 5-Year Peak
           Average
oo
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
1.03E-09
1.09E-09
1.16E-09
1.23E-09
1.30E-09
1.38E-09
1.46E-09
1.55E-09
1.64E-09
1.74E-09
1.84E-09
1.95E-09
2.07E-09
2.20E-09
2.33E-09
2.47E-09
2.62E-09
2.77E-09
2.94E-09
3.12E-09
3.30E-09
3.50E-09
3.71E-09
3.93E-09
4.17E-09
4.42E-09
4.68E-09
4.97E-09
5.26E-09
5.58E-09
5.91E-09
6.27E-09
6.65E-09
Fat (ng/kg)
6.14E-05
6.51E-05
6.90E-05
7.32E-05
7.75E-05
8.22E-05
8.71E-05
9.23E-05
9.79E-05
1.04E-04
1.10E-04
1.17E-04
1.24E-04
1.31E-04
1.39E-04
1.47E-04
1.56E-04
1.65E-04
1.75E-04
1.86E-04
1.97E-04
2.09E-04
2.21E-04
2.34E-04
2.48E-04
2.63E-04
2.79E-04
2.96E-04
3.13E-04
3.32E-04
3.52E-04
3.73E-04
3.95E-04
Body
Burden
(ng/kg)
1.92E-05
2.03E-05
2.15E-05
2.28E-05
2.42E-05
2.56E-05
2.72E-05
2.88E-05
3.05E-05
3.24E-05
3.43E-05
3.64E-05
3.85E-05
4.08E-05
4.33E-05
4.59E-05
4.86E-05
5.15E-05
5.46E-05
5.79E-05
6.14E-05
6.51E-05
6.90E-05
7.31E-05
7.75E-05
8.21E-05
8.70E-05
9.22E-05
9.78E-05
1.04E-04
1.10E-04
1.16E-04
1.23E-04
Blood
(ng/kg)
6.46E-07
6.85E-07
7.26E-07
7.69E-07
8.15E-07
8.64E-07
9.16E-07
9.71E-07
1.03E-06
1.09E-06
1.16E-06
1.23E-06
1.30E-06
1.38E-06
1.46E-06
1.55E-06
1.64E-06
1.74E-06
1.84E-06
1.95E-06
2.07E-06
2.19E-06
2.32E-06
2.46E-06
2.61E-06
2.77E-06
2.93E-06
3.11E-06
3.29E-06
3.49E-06
3.70E-06
3.92E-06
4.16E-06
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
7.04E-09
7.47E-09
7.92E-09
8.39E-09
8.89E-09
9.43E-09
9.99E-09
1.06E-08
1.12E-08
1.19E-08
1.26E-08
1.34E-08
1.42E-08
1.50E-08
1.59E-08
1.69E-08
1.79E-08
1.90E-08
2.01E-08
2.13E-08
2.26E-08
2.39E-08
2.54E-08
2.69E-08
2.85E-08
3.02E-08
3.20E-08
3.40E-08
3.60E-08
3.82E-08
4.05E-08
4.29E-08
4.55E-08
4.82E-08
5.11E-08
5.41E-08
Fat (ng/kg)
4.19E-04
4.44E-04
4.71E-04
4.99E-04
5.29E-04
5.60E-04
5.94E-04
6.29E-04
6.67E-04
7.07E-04
7.49E-04
7.94E-04
8.41E-04
8.91E-04
9.45E-04
l.OOE-03
1.06E-03
1.12E-03
1.19E-03
1.26E-03
1.34E-03
1.42E-03
1.50E-03
1.59E-03
1.69E-03
1.79E-03
1.89E-03
2.01E-03
2.13E-03
2.25E-03
2.39E-03
2.53E-03
2.68E-03
2.84E-03
3.01E-03
3.19E-03
Body
Burden
(ng/kg)
1.31E-04
1.39E-04
1.47E-04
1.56E-04
1.65E-04
1.75E-04
1.85E-04
1.96E-04
2.08E-04
2.21E-04
2.34E-04
2.48E-04
2.63E-04
2.78E-04
2.95E-04
3.13E-04
3.31E-04
3.51E-04
3.72E-04
3.94E-04
4.18E-04
4.43E-04
4.69E-04
4.97E-04
5.27E-04
5.58E-04
5.92E-04
6.27E-04
6.64E-04
7.04E-04
7.46E-04
7.91E-04
8.38E-04
8.88E-04
9.40E-04
9.96E-04
Blood
(ng/kg)
4.41E-06
4.67E-06
4.95E-06
5.24E-06
5.56E-06
5.89E-06
6.24E-06
6.62E-06
7.01E-06
7.43E-06
7.88E-06
8.35E-06
8.84E-06
9.37E-06
9.93E-06
1.05E-05
1.12E-05
1.18E-05
1.25E-05
1.33E-05
1.41E-05
1.49E-05
1.58E-05
1.67E-05
1.77E-05
1.88E-05
1.99E-05
2.11E-05
2.24E-05
2.37E-05
2.51E-05
2.66E-05
2.82E-05
2.99E-05
3.16E-05
3.35E-05
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
5.74E-08
6.08E-08
6.45E-08
6.84E-08
7.25E-08
7.68E-08
8.14E-08
8.63E-08
9.15E-08
9.70E-08
1.03E-07
1.09E-07
1.15E-07
1.22E-07
1.30E-07
1.38E-07
1.46E-07
1.55E-07
1.64E-07
1.74E-07
1.84E-07
1.95E-07
2.07E-07
2.19E-07
2.32E-07
2.46E-07
2.61E-07
2.77E-07
2.93E-07
3.11E-07
3.30E-07
3.49E-07
3.70E-07
3.93E-07
4.16E-07
4.41E-07
Fat (ng/kg)
3.38E-03
3.58E-03
3.79E-03
4.02E-03
4.25E-03
4.51E-03
4.77E-03
5.06E-03
5.36E-03
5.67E-03
6.01E-03
6.37E-03
6.74E-03
7.14E-03
7.56E-03
8.01E-03
8.48E-03
8.98E-03
9.51E-03
1.01E-02
1.07E-02
1.13E-02
1.20E-02
1.27E-02
1.34E-02
1.42E-02
1.50E-02
1.59E-02
1.68E-02
1.78E-02
1.89E-02
2.00E-02
2.11E-02
2.24E-02
2.37E-02
2.51E-02
Body
Burden
(ng/kg)
1.06E-03
1.12E-03
1.19E-03
1.26E-03
1.33E-03
1.41E-03
1.49E-03
1.58E-03
1.68E-03
1.78E-03
1.88E-03
1.99E-03
2.11E-03
2.24E-03
2.37E-03
2.51E-03
2.66E-03
2.82E-03
2.98E-03
3.16E-03
3.34E-03
3.54E-03
3.75E-03
3.97E-03
4.21E-03
4.46E-03
4.72E-03
5.00E-03
5.29E-03
5.60E-03
5.93E-03
6.28E-03
6.65E-03
7.04E-03
7.45E-03
7.89E-03
Blood
(ng/kg)
3.55E-05
3.76E-05
3.99E-05
4.22E-05
4.47E-05
4.74E-05
5.02E-05
5.32E-05
5.63E-05
5.97E-05
6.32E-05
6.69E-05
7.09E-05
7.51E-05
7.95E-05
8.42E-05
8.92E-05
9.45E-05
l.OOE-04
1.06E-04
1.12E-04
1.19E-04
1.26E-04
1.33E-04
1.41E-04
1.49E-04
1.58E-04
1.67E-04
1.77E-04
1.87E-04
1.98E-04
2.10E-04
2.22E-04
2.35E-04
2.49E-04
2.63E-04

-------
oo
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
4.68E-07
4.96E-07
5.25E-07
5.57E-07
5.90E-07
6.26E-07
6.63E-07
7.03E-07
7.45E-07
7.90E-07
8.37E-07
8.88E-07
9.41E-07
9.97E-07
1.01E-06
1.03E-06
1.04E-06
1.06E-06
1.07E-06
1.09E-06
1.11E-06
1.12E-06
1.14E-06
1.16E-06
1.17E-06
1.19E-06
1.21E-06
1.23E-06
1.24E-06
1.26E-06
1.28E-06
1.30E-06
1.32E-06
1.34E-06
1.36E-06
1.38E-06
Fat (ng/kg)
2.65E-02
2.81E-02
2.97E-02
3.14E-02
3.32E-02
3.51E-02
3.72E-02
3.93E-02
4.16E-02
4.40E-02
4.65E-02
4.92E-02
5.20E-02
5.50E-02
5.57E-02
5.65E-02
5.73E-02
5.82E-02
5.90E-02
5.98E-02
6.07E-02
6.15E-02
6.24E-02
6.33E-02
6.42E-02
6.51E-02
6.60E-02
6.69E-02
6.79E-02
6.88E-02
6.98E-02
7.08E-02
7.18E-02
7.28E-02
7.38E-02
7.49E-02
Body
Burden
(ng/kg)
8.35E-03
8.83E-03
9.35E-03
9.90E-03
1.05E-02
1.11E-02
1.17E-02
1.24E-02
1.31E-02
1.39E-02
1.47E-02
1.55E-02
1.64E-02
1.74E-02
1.76E-02
1.79E-02
1.82E-02
1.84E-02
1.87E-02
1.89E-02
1.92E-02
1.95E-02
1.98E-02
2.00E-02
2.03E-02
2.06E-02
2.09E-02
2.12E-02
2.15E-02
2.18E-02
2.21E-02
2.25E-02
2.28E-02
2.31E-02
2.34E-02
2.38E-02
Blood
(ng/kg)
2.79E-04
2.95E-04
3.12E-04
3.30E-04
3.49E-04
3.69E-04
3.91E-04
4.13E-04
4.37E-04
4.62E-04
4.89E-04
5.17E-04
5.47E-04
5.78E-04
5.86E-04
5.94E-04
6.03E-04
6.11E-04
6.20E-04
6.29E-04
6.38E-04
6.47E-04
6.56E-04
6.65E-04
6.75E-04
6.84E-04
6.94E-04
7.04E-04
7.13E-04
7.24E-04
7.34E-04
7.44E-04
7.55E-04
7.65E-04
7.76E-04
7.87E-04
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
1.40E-06
1.42E-06
1.44E-06
1.46E-06
1.49E-06
1.53E-06
1.58E-06
1.62E-06
1.67E-06
1.72E-06
1.77E-06
1.83E-06
1.88E-06
1.94E-06
2.00E-06
2.06E-06
2.12E-06
2.18E-06
2.25E-06
2.32E-06
2.39E-06
2.46E-06
2.53E-06
2.61E-06
2.68E-06
2.76E-06
2.85E-06
2.93E-06
3.02E-06
3.11E-06
3.21E-06
3.30E-06
3.40E-06
3.50E-06
3.61E-06
3.72E-06
Fat (ng/kg)
7.59E-02
7.70E-02
7.81E-02
7.92E-02
8.03E-02
8.25E-02
8.49E-02
8.73E-02
8.97E-02
9.23E-02
9.48E-02
9.75E-02
l.OOE-01
1.03E-01
1.06E-01
1.09E-01
1.12E-01
1.15E-01
1.18E-01
1.22E-01
1.25E-01
1.28E-01
1.32E-01
1.36E-01
1.39E-01
1.43E-01
1.47E-01
1.51E-01
1.55E-01
1.60E-01
1.64E-01
1.69E-01
1.73E-01
1.78E-01
1.83E-01
1.88E-01
Body
Burden
(ng/kg)
2.41E-02
2.44E-02
2.48E-02
2.51E-02
2.55E-02
2.62E-02
2.70E-02
2.77E-02
2.85E-02
2.93E-02
3.02E-02
3.10E-02
3.19E-02
3.28E-02
3.38E-02
3.47E-02
3.57E-02
3.67E-02
3.77E-02
3.88E-02
3.99E-02
4.10E-02
4.22E-02
4.34E-02
4.46E-02
4.58E-02
4.71E-02
4.84E-02
4.98E-02
5.12E-02
5.26E-02
5.41E-02
5.56E-02
5.71E-02
5.87E-02
6.04E-02
Blood
(ng/kg)
7.98E-04
8.09E-04
8.21E-04
8.32E-04
8.44E-04
8.68E-04
8.92E-04
9.17E-04
9.43E-04
9.70E-04
9.97E-04
1.02E-03
1.05E-03
1.08E-03
1.11E-03
1.14E-03
1.18E-03
1.21E-03
1.24E-03
1.28E-03
1.31E-03
1.35E-03
1.39E-03
1.43E-03
1.47E-03
1.51E-03
1.55E-03
1.59E-03
1.63E-03
1.68E-03
1.73E-03
1.77E-03
1.82E-03
1.87E-03
1.92E-03
1.97E-03
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
3.83E-06
3.94E-06
4.06E-06
4.18E-06
4.31E-06
4.44E-06
4.57E-06
4.71E-06
4.85E-06
4.99E-06
5.14E-06
5.30E-06
5.46E-06
5.62E-06
5.79E-06
5.96E-06
6.14E-06
6.33E-06
6.52E-06
6.71E-06
6.91E-06
7.12E-06
7.33E-06
7.55E-06
7.78E-06
8.01E-06
8.25E-06
8.50E-06
8.76E-06
9.02E-06
9.29E-06
9.57E-06
9.86E-06
1.02E-05
1.05E-05
1.08E-05
Fat (ng/kg)
1.93E-01
1.98E-01
2.04E-01
2.09E-01
2.15E-01
2.21E-01
2.27E-01
2.33E-01
2.39E-01
2.45E-01
2.52E-01
2.59E-01
2.66E-01
2.73E-01
2.80E-01
2.87E-01
2.95E-01
3.03E-01
3.11E-01
3.19E-01
3.28E-01
3.36E-01
3.45E-01
3.54E-01
3.63E-01
3.73E-01
3.83E-01
3.93E-01
4.03E-01
4.13E-01
4.24E-01
4.35E-01
4.46E-01
4.58E-01
4.69E-01
4.81E-01
Body
Burden
(ng/kg)
6.20E-02
6.38E-02
6.55E-02
6.73E-02
6.92E-02
7.11E-02
7.31E-02
7.51E-02
7.71E-02
7.92E-02
8.14E-02
8.36E-02
8.59E-02
8.83E-02
9.07E-02
9.31E-02
9.57E-02
9.83E-02
1.01E-01
1.04E-01
1.06E-01
1.09E-01
1.12E-01
1.15E-01
1.18E-01
1.22E-01
1.25E-01
1.28E-01
1.32E-01
1.35E-01
1.39E-01
1.42E-01
1.46E-01
1.50E-01
1.54E-01
1.58E-01
Blood
(ng/kg)
2.03E-03
2.08E-03
2.14E-03
2.20E-03
2.26E-03
2.32E-03
2.38E-03
2.45E-03
2.51E-03
2.58E-03
2.65E-03
2.72E-03
2.79E-03
2.87E-03
2.94E-03
3.02E-03
3.10E-03
3.18E-03
3.27E-03
3.35E-03
3.44E-03
3.53E-03
3.63E-03
3.72E-03
3.82E-03
3.92E-03
4.02E-03
4.12E-03
4.23E-03
4.34E-03
4.45E-03
4.57E-03
4.69E-03
4.81E-03
4.93E-03
5.06E-03

-------
oo
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
1.11E-05
1.14E-05
1.18E-05
1.21E-05
1.25E-05
1.29E-05
1.32E-05
1.36E-05
1.41E-05
1.45E-05
1.49E-05
1.54E-05
1.58E-05
1.63E-05
1.68E-05
1.73E-05
1.78E-05
1.83E-05
1.89E-05
1.95E-05
2.00E-05
2.06E-05
2.13E-05
2.19E-05
2.25E-05
2.32E-05
2.39E-05
2.46E-05
2.54E-05
2.61E-05
2.69E-05
2.77E-05
2.86E-05
2.94E-05
3.03E-05
3.12E-05
Fat (ng/kg)
4.94E-01
5.06E-01
5.19E-01
5.32E-01
5.46E-01
5.60E-01
5.74E-01
5.88E-01
6.03E-01
6.18E-01
6.34E-01
6.49E-01
6.66E-01
6.82E-01
6.99E-01
7.16E-01
7.34E-01
7.52E-01
7.71E-01
7.89E-01
8.09E-01
8.29E-01
8.49E-01
8.69E-01
8.90E-01
9.12E-01
9.34E-01
9.56E-01
9.79E-01
l.OOE+00
1.03E+00
1.05E+00
1.08E+00
1.10E+00
1.13E+00
1.15E+00
Body
Burden
(ng/kg)
1.62E-01
1.67E-01
1.71E-01
1.75E-01
1.80E-01
1.85E-01
1.90E-01
1.94E-01
1.99E-01
2.05E-01
2.10E-01
2.15E-01
2.21E-01
2.27E-01
2.32E-01
2.38E-01
2.45E-01
2.51E-01
2.57E-01
2.64E-01
2.70E-01
2.77E-01
2.84E-01
2.91E-01
2.99E-01
3.06E-01
3.14E-01
3.22E-01
3.30E-01
3.38E-01
3.47E-01
3.55E-01
3.64E-01
3.73E-01
3.82E-01
3.92E-01
Blood
(ng/kg)
5.19E-03
5.32E-03
5.45E-03
5.59E-03
5.74E-03
5.88E-03
6.03E-03
6.18E-03
6.34E-03
6.49E-03
6.66E-03
6.82E-03
6.99E-03
7.17E-03
7.34E-03
7.53E-03
7.71E-03
7.90E-03
8.09E-03
8.29E-03
8.50E-03
8.70E-03
8.91E-03
9.13E-03
9.35E-03
9.58E-03
9.81E-03
l.OOE-02
1.03E-02
1.05E-02
1.08E-02
1.10E-02
1.13E-02
1.16E-02
1.18E-02
1.21E-02
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
3.21E-05
3.31E-05
3.41E-05
3.51E-05
3.62E-05
3.73E-05
3.84E-05
3.95E-05
4.07E-05
4.19E-05
4.32E-05
4.45E-05
4.58E-05
4.72E-05
4.86E-05
5.01E-05
5.16E-05
5.31E-05
5.47E-05
5.64E-05
5.81E-05
5.98E-05
6.16E-05
6.34E-05
6.54E-05
6.73E-05
6.93E-05
7.14E-05
7.36E-05
7.58E-05
7.80E-05
8.04E-05
8.28E-05
8.53E-05
8.78E-05
9.05E-05
Fat (ng/kg)
1.18E+00
1.21E+00
1.24E+00
1.27E+00
1.30E+00
1.33E+00
1.36E+00
1.39E+00
1.42E+00
1.45E+00
1.49E+00
1.52E+00
1.56E+00
1.59E+00
1.63E+00
1.66E+00
1.70E+00
1.74E+00
1.78E+00
1.82E+00
1.86E+00
1.90E+00
1.94E+00
1.99E+00
2.03E+00
2.08E+00
2.12E+00
2.17E+00
2.22E+00
2.26E+00
2.31E+00
2.36E+00
2.42E+00
2.47E+00
2.52E+00
2.58E+00
Body
Burden
(ng/kg)
4.02E-01
4.11E-01
4.22E-01
4.32E-01
4.43E-01
4.54E-01
4.65E-01
4.76E-01
4.87E-01
4.99E-01
5.11E-01
5.24E-01
5.36E-01
5.49E-01
5.62E-01
5.76E-01
5.89E-01
6.04E-01
6.18E-01
6.33E-01
6.48E-01
6.63E-01
6.79E-01
6.95E-01
7.11E-01
7.28E-01
7.45E-01
7.62E-01
7.80E-01
7.98E-01
8.17E-01
8.36E-01
8.55E-01
8.75E-01
8.95E-01
9.16E-01
Blood
(ng/kg)
1.24E-02
1.27E-02
1.30E-02
1.33E-02
1.36E-02
1.39E-02
1.43E-02
1.46E-02
1.49E-02
1.53E-02
1.56E-02
1.60E-02
1.63E-02
1.67E-02
1.71E-02
1.75E-02
1.79E-02
1.83E-02
1.87E-02
1.91E-02
1.95E-02
2.00E-02
2.04E-02
2.09E-02
2.13E-02
2.18E-02
2.23E-02
2.28E-02
2.33E-02
2.38E-02
2.43E-02
2.48E-02
2.54E-02
2.59E-02
2.65E-02
2.70E-02
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
9.32E-05
9.60E-05
9.89E-05
1.02E-04
1.05E-04
1.08E-04
1.11E-04
1.15E-04
1.18E-04
1.22E-04
1.25E-04
1.29E-04
1.33E-04
1.37E-04
1.41E-04
1.45E-04
1.50E-04
1.54E-04
1.59E-04
1.63E-04
1.68E-04
1.73E-04
1.79E-04
1.84E-04
1.89E-04
1.95E-04
2.01E-04
2.07E-04
2.13E-04
2.20E-04
2.26E-04
2.33E-04
2.40E-04
2.47E-04
2.55E-04
2.62E-04
Fat (ng/kg)
2.63E+00
2.69E+00
2.75E+00
2.81E+00
2.87E+00
2.93E+00
2.99E+00
3.05E+00
3.12E+00
3.18E+00
3.25E+00
3.32E+00
3.39E+00
3.46E+00
3.53E+00
3.60E+00
3.68E+00
3.75E+00
3.83E+00
3.91E+00
3.99E+00
4.07E+00
4.16E+00
4.24E+00
4.33E+00
4.42E+00
4.51E+00
4.60E+00
4.69E+00
4.79E+00
4.89E+00
4.99E+00
5.09E+00
5.19E+00
5.29E+00
5.40E+00
Body
Burden
(ng/kg)
9.37E-01
9.58E-01
9.81E-01
l.OOE+00
1.03E+00
1.05E+00
1.07E+00
1.10E+00
1.12E+00
1.15E+00
1.17E+00
1.20E+00
1.23E+00
1.26E+00
1.28E+00
1.31E+00
1.34E+00
1.37E+00
1.40E+00
1.43E+00
1.47E+00
1.50E+00
1.53E+00
1.57E+00
1.60E+00
1.64E+00
1.67E+00
1.71E+00
1.75E+00
1.79E+00
1.83E+00
1.87E+00
1.91E+00
1.95E+00
2.00E+00
2.04E+00
Blood
(ng/kg)
2.76E-02
2.82E-02
2.88E-02
2.95E-02
3.01E-02
3.07E-02
3.14E-02
3.20E-02
3.27E-02
3.34E-02
3.41E-02
3.48E-02
3.55E-02
3.63E-02
3.70E-02
3.78E-02
3.86E-02
3.94E-02
4.02E-02
4.10E-02
4.19E-02
4.27E-02
4.36E-02
4.45E-02
4.55E-02
4.64E-02
4.73E-02
4.83E-02
4.93E-02
5.03E-02
5.13E-02
5.23E-02
5.34E-02
5.45E-02
5.56E-02
5.67E-02

-------
oo
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
2.70E-04
2.78E-04
2.86E-04
2.95E-04
3.04E-04
3.13E-04
3.22E-04
3.32E-04
3.42E-04
3.52E-04
3.63E-04
3.74E-04
3.85E-04
3.97E-04
4.08E-04
4.21E-04
4.33E-04
4.46E-04
4.60E-04
4.74E-04
4.88E-04
5.02E-04
5.17E-04
5.33E-04
5.49E-04
5.65E-04
5.82E-04
6.00E-04
6.18E-04
6.36E-04
6.55E-04
6.75E-04
6.95E-04
7.16E-04
7.38E-04
7.60E-04
Fat (ng/kg)
5.51E+00
5.62E+00
5.73E+00
5.85E+00
5.96E+00
6.08E+00
6.20E+00
6.32E+00
6.45E+00
6.57E+00
6.70E+00
6.84E+00
6.97E+00
7.11E+00
7.25E+00
7.39E+00
7.54E+00
7.68E+00
7.83E+00
7.99E+00
8.15E+00
8.30E+00
8.47E+00
8.63E+00
8.80E+00
8.97E+00
9.14E+00
9.32E+00
9.50E+00
9.68E+00
9.87E+00
1.01E+01
1.03E+01
1.05E+01
1.07E+01
1.09E+01
Body
Burden
(ng/kg)
2.09E+00
2.13E+00
2.18E+00
2.23E+00
2.28E+00
2.33E+00
2.38E+00
2.43E+00
2.48E+00
2.54E+00
2.59E+00
2.65E+00
2.71E+00
2.77E+00
2.83E+00
2.89E+00
2.96E+00
3.02E+00
3.09E+00
3.16E+00
3.23E+00
3.30E+00
3.37E+00
3.45E+00
3.52E+00
3.60E+00
3.68E+00
3.76E+00
3.85E+00
3.93E+00
4.02E+00
4.11E+00
4.20E+00
4.29E+00
4.38E+00
4.48E+00
Blood
(ng/kg)
5.78E-02
5.90E-02
6.01E-02
6.13E-02
6.26E-02
6.38E-02
6.51E-02
6.63E-02
6.76E-02
6.90E-02
7.03E-02
7.17E-02
7.32E-02
7.46E-02
7.61E-02
7.76E-02
7.91E-02
8.06E-02
8.22E-02
8.38E-02
8.55E-02
8.71E-02
8.88E-02
9.06E-02
9.23E-02
9.41E-02
9.59E-02
9.78E-02
9.97E-02
1.02E-01
1.04E-01
1.06E-01
1.08E-01
1.10E-01
1.12E-01
1.14E-01
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
7.83E-04
8.06E-04
8.30E-04
8.55E-04
8.81E-04
9.07E-04
9.21E-04
9.35E-04
9.49E-04
9.63E-04
9.69E-04
9.77E-04
9.84E-04
9.91E-04
9.98E-04
1.01E-03
1.02E-03
1.04E-03
1.05E-03
1.07E-03
1.08E-03
1.10E-03
1.12E-03
1.13E-03
1.15E-03
1.17E-03
1.18E-03
1.20E-03
1.22E-03
1.24E-03
1.26E-03
1.27E-03
1.29E-03
1.31E-03
1.33E-03
1.35E-03
Fat (ng/kg)
1.11E+01
1.13E+01
1.15E+01
1.17E+01
1.19E+01
1.22E+01
1.23E+01
1.24E+01
1.25E+01
1.26E+01
1.27E+01
1.28E+01
1.28E+01
1.29E+01
1.29E+01
1.30E+01
1.31E+01
1.32E+01
1.34E+01
1.35E+01
1.36E+01
1.38E+01
1.39E+01
1.40E+01
1.41E+01
1.43E+01
1.44E+01
1.45E+01
1.47E+01
1.48E+01
1.50E+01
1.51E+01
1.52E+01
1.54E+01
1.55E+01
1.57E+01
Body
Burden
(ng/kg)
4.58E+00
4.68E+00
4.78E+00
4.89E+00
5.00E+00
5.11E+00
5.16E+00
5.22E+00
5.28E+00
5.34E+00
5.36E+00
5.40E+00
5.42E+00
5.45E+00
5.48E+00
5.51E+00
5.58E+00
5.64E+00
5.70E+00
5.76E+00
5.82E+00
5.89E+00
5.95E+00
6.02E+00
6.09E+00
6.15E+00
6.22E+00
6.29E+00
6.36E+00
6.43E+00
6.50E+00
6.57E+00
6.64E+00
6.72E+00
6.79E+00
6.87E+00
Blood
(ng/kg)
1.16E-01
1.18E-01
1.21E-01
1.23E-01
1.25E-01
1.28E-01
1.29E-01
1.30E-01
1.31E-01
1.33E-01
1.33E-01
1.34E-01
1.34E-01
1.35E-01
1.36E-01
1.36E-01
1.38E-01
1.39E-01
1.40E-01
1.42E-01
1.43E-01
1.44E-01
1.46E-01
1.47E-01
1.48E-01
1.50E-01
1.51E-01
1.53E-01
1.54E-01
1.55E-01
1.57E-01
1.58E-01
1.60E-01
1.61E-01
1.63E-01
1.64E-01
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
1.37E-03
1.39E-03
1.41E-03
1.43E-03
1.46E-03
1.48E-03
1.50E-03
1.52E-03
1.54E-03
1.57E-03
1.59E-03
1.61E-03
1.64E-03
1.66E-03
1.69E-03
1.71E-03
1.74E-03
1.76E-03
1.79E-03
1.82E-03
1.84E-03
1.87E-03
1.90E-03
1.93E-03
1.96E-03
1.99E-03
2.02E-03
2.08E-03
2.14E-03
2.20E-03
2.27E-03
2.34E-03
2.41E-03
2.48E-03
2.55E-03
2.63E-03
Fat (ng/kg)
1.58E+01
1.60E+01
1.61E+01
1.63E+01
1.64E+01
1.66E+01
1.67E+01
1.69E+01
1.71E+01
1.72E+01
1.75E+01
1.80E+01
1.83E+01
1.85E+01
1.87E+01
1.90E+01
1.90E+01
1.86E+01
1.87E+01
1.89E+01
1.91E+01
1.93E+01
1.98E+01
2.02E+01
2.03E+01
2.05E+01
2.09E+01
2.10E+01
2.09E+01
2.13E+01
2.17E+01
2.21E+01
2.26E+01
2.30E+01
2.34E+01
2.38E+01
Body
Burden
(ng/kg)
6.94E+00
7.02E+00
7.10E+00
7.18E+00
7.26E+00
7.34E+00
7.42E+00
7.50E+00
7.58E+00
7.67E+00
7.86E+00
8.23E+00
8.35E+00
8.36E+00
8.43E+00
8.54E+00
8.52E+00
8.38E+00
8.47E+00
8.57E+00
8.66E+00
8.80E+00
9.14E+00
9.51E+00
9.42E+00
9.53E+00
9.67E+00
9.70E+00
9.68E+00
9.90E+00
1.01E+01
1.03E+01
1.06E+01
1.08E+01
1.11E+01
1.13E+01
Blood
(ng/kg)
1.66E-01
1.68E-01
1.69E-01
1.71E-01
1.72E-01
1.74E-01
1.76E-01
1.77E-01
1.79E-01
1.81E-01
1.84E-01
1.89E-01
1.92E-01
1.94E-01
1.96E-01
2.00E-01
1.99E-01
1.95E-01
1.96E-01
1.98E-01
2.00E-01
2.03E-01
2.07E-01
2.12E-01
2.13E-01
2.15E-01
2.19E-01
2.20E-01
2.20E-01
2.24E-01
2.28E-01
2.32E-01
2.37E-01
2.41E-01
2.45E-01
2.50E-01

-------
oo
oo
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
2.71E-03
2.79E-03
2.87E-03
2.96E-03
3.05E-03
3.14E-03
3.23E-03
3.33E-03
3.43E-03
3.53E-03
3.64E-03
3.75E-03
3.81E-03
3.86E-03
3.98E-03
4.10E-03
4.22E-03
4.35E-03
4.48E-03
4.61E-03
4.75E-03
4.89E-03
5.04E-03
5.19E-03
5.35E-03
5.51E-03
5.67E-03
5.84E-03
5.93E-03
6.02E-03
6.20E-03
6.38E-03
6.57E-03
6.77E-03
6.98E-03
7.18E-03
Fat (ng/kg)
2.43E+01
2.47E+01
2.52E+01
2.57E+01
2.62E+01
2.66E+01
2.72E+01
2.78E+01
2.82E+01
2.87E+01
2.92E+01
2.99E+01
3.02E+01
3.04E+01
3.09E+01
3.11E+01
3.15E+01
3.20E+01
3.26E+01
3.32E+01
3.39E+01
3.45E+01
3.53E+01
3.63E+01
3.75E+01
3.82E+01
3.93E+01
4.01E+01
4.04E+01
4.08E+01
4.15E+01
4.23E+01
4.29E+01
4.35E+01
4.39E+01
4.50E+01
Body
Burden
(ng/kg)
1.16E+01
1.18E+01
1.21E+01
1.24E+01
1.26E+01
1.29E+01
1.33E+01
1.36E+01
1.38E+01
1.41E+01
1.45E+01
1.49E+01
1.51E+01
1.52E+01
1.54E+01
1.55E+01
1.58E+01
1.61E+01
1.65E+01
1.69E+01
1.73E+01
1.77E+01
1.83E+01
1.91E+01
1.96E+01
2.01E+01
2.08E+01
2.13E+01
2.15E+01
2.18E+01
2.22E+01
2.28E+01
2.31E+01
2.35E+01
2.39E+01
2.47E+01
Blood
(ng/kg)
2.55E-01
2.60E-01
2.64E-01
2.69E-01
2.74E-01
2.79E-01
2.85E-01
2.91E-01
2.95E-01
3.01E-01
3.07E-01
3.13E-01
3.17E-01
3.18E-01
3.24E-01
3.26E-01
3.30E-01
3.36E-01
3.42E-01
3.49E-01
3.55E-01
3.62E-01
3.70E-01
3.81E-01
3.93E-01
4.01E-01
4.12E-01
4.20E-01
4.24E-01
4.28E-01
4.35E-01
4.44E-01
4.50E-01
4.57E-01
4.60E-01
4.71E-01
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
7.40E-03
7.51E-03
7.62E-03
7.85E-03
8.09E-03
8.33E-03
8.58E-03
8.71E-03
8.84E-03
9.10E-03
9.37E-03
9.66E-03
9.94E-03
1.02E-02
1.06E-02
1.09E-02
1.12E-02
1.15E-02
1.19E-02
1.22E-02
1.26E-02
1.30E-02
1.34E-02
1.38E-02
1.42E-02
1.46E-02
1.50E-02
1.55E-02
1.60E-02
1.64E-02
1.69E-02
1.74E-02
1.80E-02
1.85E-02
1.91E-02
1.96E-02
Fat (ng/kg)
4.58E+01
4.63E+01
4.68E+01
4.76E+01
4.83E+01
4.91E+01
5.00E+01
5.05E+01
5.09E+01
5.19E+01
5.28E+01
5.38E+01
5.48E+01
5.58E+01
5.68E+01
5.79E+01
5.91E+01
6.03E+01
6.14E+01
6.24E+01
6.37E+01
6.50E+01
6.61E+01
6.74E+01
6.88E+01
7.02E+01
7.15E+01
7.28E+01
7.42E+01
7.54E+01
7.69E+01
7.82E+01
7.96E+01
8.10E+01
8.24E+01
8.45E+01
Body
Burden
(ng/kg)
2.53E+01
2.56E+01
2.58E+01
2.63E+01
2.68E+01
2.74E+01
2.81E+01
2.84E+01
2.87E+01
2.94E+01
3.01E+01
3.08E+01
3.15E+01
3.22E+01
3.30E+01
3.38E+01
3.47E+01
3.56E+01
3.65E+01
3.72E+01
3.80E+01
3.90E+01
3.98E+01
4.09E+01
4.19E+01
4.32E+01
4.41E+01
4.51E+01
4.62E+01
4.73E+01
4.84E+01
4.96E+01
5.07E+01
5.18E+01
5.30E+01
5.48E+01
Blood
(ng/kg)
4.80E-01
4.85E-01
4.90E-01
4.99E-01
5.06E-01
5.15E-01
5.24E-01
5.29E-01
5.34E-01
5.44E-01
5.54E-01
5.64E-01
5.75E-01
5.85E-01
5.96E-01
6.07E-01
6.20E-01
6.32E-01
6.44E-01
6.54E-01
6.67E-01
6.82E-01
6.93E-01
7.07E-01
7.21E-01
7.36E-01
7.49E-01
7.63E-01
7.78E-01
7.91E-01
8.06E-01
8.20E-01
8.34E-01
8.49E-01
8.64E-01
8.86E-01
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
2.02E-02
2.08E-02
2.14E-02
2.21E-02
2.28E-02
2.34E-02
2.41E-02
2.49E-02
2.56E-02
2.64E-02
2.72E-02
2.80E-02
2.88E-02
2.97E-02
3.06E-02
3.15E-02
3.24E-02
3.34E-02
3.44E-02
3.54E-02
3.65E-02
3.76E-02
3.87E-02
3.99E-02
4.11E-02
4.23E-02
4.36E-02
4.49E-02
4.63E-02
4.76E-02
4.91E-02
5.05E-02
5.21E-02
5.36E-02
5.52E-02
5.69E-02
Fat (ng/kg)
8.61E+01
8.76E+01
8.88E+01
9.05E+01
9.22E+01
9.39E+01
9.57E+01
9.76E+01
9.94E+01
1.01E+02
1.03E+02
1.05E+02
1.07E+02
1.09E+02
1.11E+02
1.13E+02
1.15E+02
1.17E+02
1.19E+02
1.22E+02
1.24E+02
1.26E+02
1.29E+02
1.31E+02
1.34E+02
1.36E+02
1.39E+02
1.41E+02
1.44E+02
1.47E+02
1.50E+02
1.53E+02
1.55E+02
1.58E+02
1.61E+02
1.64E+02
Body
Burden
(ng/kg)
5.60E+01
5.73E+01
5.84E+01
5.98E+01
6.13E+01
6.28E+01
6.43E+01
6.60E+01
6.76E+01
6.93E+01
7.10E+01
7.26E+01
7.44E+01
7.62E+01
7.80E+01
7.99E+01
8.19E+01
8.39E+01
8.60E+01
8.81E+01
9.03E+01
9.26E+01
9.49E+01
9.73E+01
9.97E+01
1.02E+02
1.05E+02
1.07E+02
1.10E+02
1.13E+02
1.16E+02
1.19E+02
1.22E+02
1.24E+02
1.28E+02
1.31E+02
Blood
(ng/kg)
9.02E-01
9.18E-01
9.30E-01
9.48E-01
9.67E-01
9.84E-01
l.OOE+00
1.02E+00
1.04E+00
1.06E+00
1.08E+00
1.10E+00
1.12E+00
1.14E+00
1.16E+00
1.18E+00
1.21E+00
1.23E+00
1.25E+00
1.28E+00
1.30E+00
1.32E+00
1.35E+00
1.38E+00
1.40E+00
1.43E+00
1.45E+00
1.48E+00
1.51E+00
1.54E+00
1.57E+00
1.60E+00
1.63E+00
1.66E+00
1.69E+00
1.72E+00

-------
oo
VO
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
5.86E-02
6.03E-02
6.22E-02
6.40E-02
6.59E-02
6.79E-02
7.00E-02
7.21E-02
7.42E-02
7.64E-02
7.87E-02
8.11E-02
8.35E-02
8.60E-02
8.86E-02
9.13E-02
9.40E-02
9.68E-02
9.97E-02
1.03E-01
1.06E-01
1.09E-01
1.12E-01
1.16E-01
1.19E-01
1.23E-01
1.26E-01
1.30E-01
1.34E-01
1.38E-01
1.42E-01
1.46E-01
1.51E-01
1.55E-01
1.60E-01
1.65E-01
Fat (ng/kg)
1.68E+02
1.71E+02
1.74E+02
1.77E+02
1.80E+02
1.84E+02
1.87E+02
1.91E+02
1.95E+02
1.98E+02
2.02E+02
2.06E+02
2.10E+02
2.14E+02
2.18E+02
2.22E+02
2.26E+02
2.31E+02
2.35E+02
2.40E+02
2.44E+02
2.49E+02
2.54E+02
2.59E+02
2.64E+02
2.69E+02
2.74E+02
2.79E+02
2.84E+02
2.90E+02
2.95E+02
3.01E+02
3.07E+02
3.13E+02
3.19E+02
3.25E+02
Body
Burden
(ng/kg)
1.35E+02
1.38E+02
1.41E+02
1.44E+02
1.48E+02
1.52E+02
1.56E+02
1.60E+02
1.64E+02
1.68E+02
1.73E+02
1.77E+02
1.82E+02
1.87E+02
1.92E+02
1.96E+02
2.01E+02
2.07E+02
2.12E+02
2.18E+02
2.23E+02
2.29E+02
2.35E+02
2.41E+02
2.48E+02
2.54E+02
2.60E+02
2.67E+02
2.74E+02
2.81E+02
2.89E+02
2.96E+02
3.04E+02
3.12E+02
3.20E+02
3.29E+02
Blood
(ng/kg)
1.76E+00
1.79E+00
1.82E+00
1.85E+00
1.89E+00
1.92E+00
1.96E+00
2.00E+00
2.04E+00
2.08E+00
2.12E+00
2.16E+00
2.20E+00
2.24E+00
2.29E+00
2.33E+00
2.37E+00
2.42E+00
2.47E+00
2.51E+00
2.56E+00
2.61E+00
2.66E+00
2.71E+00
2.76E+00
2.82E+00
2.87E+00
2.92E+00
2.98E+00
3.04E+00
3.09E+00
3.15E+00
3.21E+00
3.28E+00
3.34E+00
3.40E+00
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
1.70E-01
1.75E-01
1.80E-01
1.86E-01
1.91E-01
1.97E-01
2.03E-01
2.09E-01
2.15E-01
2.22E-01
2.28E-01
2.35E-01
2.42E-01
2.49E-01
2.57E-01
2.65E-01
2.72E-01
2.81E-01
2.89E-01
2.98E-01
3.07E-01
3.16E-01
3.25E-01
3.35E-01
3.45E-01
3.56E-01
3.66E-01
3.77E-01
3.89E-01
4.00E-01
4.12E-01
4.25E-01
4.37E-01
4.50E-01
4.64E-01
4.78E-01
Fat (ng/kg)
3.31E+02
3.38E+02
3.44E+02
3.51E+02
3.58E+02
3.65E+02
3.72E+02
3.79E+02
3.86E+02
3.94E+02
4.01E+02
4.09E+02
4.17E+02
4.25E+02
4.34E+02
4.42E+02
4.51E+02
4.60E+02
4.69E+02
4.78E+02
4.87E+02
4.97E+02
5.07E+02
5.17E+02
5.27E+02
5.38E+02
5.48E+02
5.59E+02
5.70E+02
5.82E+02
5.93E+02
6.05E+02
6.17E+02
6.29E+02
6.42E+02
6.55E+02
Body
Burden
(ng/kg)
3.37E+02
3.46E+02
3.55E+02
3.65E+02
3.75E+02
3.85E+02
3.95E+02
4.05E+02
4.16E+02
4.27E+02
4.39E+02
4.50E+02
4.62E+02
4.74E+02
4.87E+02
5.00E+02
5.14E+02
5.28E+02
5.42E+02
5.56E+02
5.71E+02
5.87E+02
6.03E+02
6.19E+02
6.36E+02
6.53E+02
6.71E+02
6.89E+02
7.08E+02
7.27E+02
7.47E+02
7.67E+02
7.88E+02
8.10E+02
8.32E+02
8.55E+02
Blood
(ng/kg)
3.47E+00
3.54E+00
3.61E+00
3.68E+00
3.75E+00
3.82E+00
3.90E+00
3.97E+00
4.05E+00
4.13E+00
4.21E+00
4.29E+00
4.37E+00
4.46E+00
4.54E+00
4.63E+00
4.73E+00
4.82E+00
4.91E+00
5.01E+00
5.11E+00
5.21E+00
5.31E+00
5.42E+00
5.52E+00
5.63E+00
5.75E+00
5.86E+00
5.98E+00
6.09E+00
6.22E+00
6.34E+00
6.47E+00
6.60E+00
6.73E+00
6.86E+00
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
4.92E-01
5.07E-01
5.22E-01
5.38E-01
5.54E-01
5.71E-01
5.88E-01
6.05E-01
6.23E-01
6.42E-01
6.61E-01
6.81E-01
7.02E-01
7.23E-01
7.44E-01
7.67E-01
7.90E-01
8.13E-01
8.38E-01
8.63E-01
8.89E-01
9.16E-01
9.43E-01
9.71E-01
l.OOE+00
1.03E+00
1.06E+00
1.09E+00
1.13E+00
1.16E+00
1.19E+00
1.23E+00
1.27E+00
1.31E+00
1.34E+00
1.38E+00
Fat (ng/kg)
6.68E+02
6.81E+02
6.95E+02
7.09E+02
7.23E+02
7.38E+02
7.53E+02
7.68E+02
7.83E+02
7.99E+02
8.16E+02
8.32E+02
8.49E+02
8.66E+02
8.84E+02
9.02E+02
9.21E+02
9.40E+02
9.59E+02
9.78E+02
9.99E+02
1.02E+03
1.04E+03
1.06E+03
1.08E+03
1.11E+03
1.13E+03
1.15E+03
1.18E+03
1.20E+03
1.23E+03
1.25E+03
1.28E+03
1.31E+03
1.33E+03
1.36E+03
Body
Burden
(ng/kg)
8.78E+02
9.02E+02
9.26E+02
9.52E+02
9.78E+02
1.01E+03
1.03E+03
1.06E+03
1.09E+03
1.12E+03
1.15E+03
1.18E+03
1.22E+03
1.25E+03
1.28E+03
1.32E+03
1.36E+03
1.39E+03
1.43E+03
1.47E+03
1.51E+03
1.56E+03
1.60E+03
1.64E+03
1.69E+03
1.74E+03
1.79E+03
1.84E+03
1.89E+03
1.94E+03
1.99E+03
2.05E+03
2.11E+03
2.17E+03
2.23E+03
2.29E+03
Blood
(ng/kg)
7.00E+00
7.14E+00
7.28E+00
7.43E+00
7.58E+00
7.73E+00
7.89E+00
8.05E+00
8.21E+00
8.38E+00
8.55E+00
8.72E+00
8.90E+00
9.08E+00
9.27E+00
9.46E+00
9.65E+00
9.85E+00
l.OOE+01
1.03E+01
1.05E+01
1.07E+01
1.09E+01
1.11E+01
1.14E+01
1.16E+01
1.18E+01
1.21E+01
1.23E+01
1.26E+01
1.29E+01
1.31E+01
1.34E+01
1.37E+01
1.40E+01
1.43E+01

-------
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
1.43E+00
1.47E+00
1.51E+00
1.56E+00
1.61E+00
1.65E+00
1.70E+00
1.75E+00
1.81E+00
1.86E+00
1.92E+00
1.97E+00
2.03E+00
2.09E+00
2.16E+00
2.22E+00
2.29E+00
2.36E+00
2.43E+00
2.50E+00
2.58E+00
2.65E+00
2.73E+00
2.82E+00
2.90E+00
2.99E+00
3.08E+00
3.17E+00
3.26E+00
3.36E+00
3.46E+00
3.57E+00
3.67E+00
3.78E+00
3.90E+00
4.01E+00
Fat (ng/kg)
1.39E+03
1.42E+03
1.45E+03
1.48E+03
1.51E+03
1.55E+03
1.58E+03
1.61E+03
1.65E+03
1.68E+03
1.72E+03
1.76E+03
1.80E+03
1.84E+03
1.88E+03
1.92E+03
1.96E+03
2.00E+03
2.05E+03
2.09E+03
2.14E+03
2.19E+03
2.23E+03
2.28E+03
2.33E+03
2.39E+03
2.44E+03
2.50E+03
2.55E+03
2.61E+03
2.67E+03
2.73E+03
2.79E+03
2.86E+03
2.92E+03
2.99E+03
Body
Burden
(ng/kg)
2.36E+03
2.42E+03
2.49E+03
2.56E+03
2.63E+03
2.71E+03
2.79E+03
2.86E+03
2.95E+03
3.03E+03
3.11E+03
3.20E+03
3.29E+03
3.39E+03
3.48E+03
3.58E+03
3.69E+03
3.79E+03
3.90E+03
4.01E+03
4.12E+03
4.24E+03
4.36E+03
4.49E+03
4.62E+03
4.75E+03
4.89E+03
5.03E+03
5.17E+03
5.32E+03
5.47E+03
5.63E+03
5.79E+03
5.96E+03
6.13E+03
6.30E+03
Blood
(ng/kg)
1.46E+01
1.49E+01
1.52E+01
1.55E+01
1.59E+01
1.62E+01
1.66E+01
1.69E+01
1.73E+01
1.77E+01
1.80E+01
1.84E+01
1.88E+01
1.92E+01
1.97E+01
2.01E+01
2.05E+01
2.10E+01
2.14E+01
2.19E+01
2.24E+01
2.29E+01
2.34E+01
2.39E+01
2.45E+01
2.50E+01
2.56E+01
2.62E+01
2.67E+01
2.74E+01
2.80E+01
2.86E+01
2.93E+01
2.99E+01
3.06E+01
3.13E+01
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
4.13E+00
4.26E+00
4.39E+00
4.52E+00
4.65E+00
4.79E+00
4.94E+00
5.08E+00
5.24E+00
5.39E+00
5.56E+00
5.72E+00
5.89E+00
6.07E+00
6.25E+00
6.44E+00
6.63E+00
6.83E+00
7.04E+00
7.25E+00
7.47E+00
7.69E+00
7.92E+00
8.16E+00
8.40E+00
8.66E+00
8.92E+00
9.18E+00
9.46E+00
9.74E+00
l.OOE+01
1.06E+01
1.13E+01
1.20E+01
1.27E+01
1.34E+01
Fat (ng/kg)
3.06E+03
3.13E+03
3.20E+03
3.28E+03
3.35E+03
3.43E+03
3.51E+03
3.59E+03
3.68E+03
3.77E+03
3.85E+03
3.95E+03
4.04E+03
4.14E+03
4.24E+03
4.34E+03
4.44E+03
4.55E+03
4.66E+03
4.77E+03
4.89E+03
5.01E+03
5.13E+03
5.26E+03
5.39E+03
5.52E+03
5.66E+03
5.80E+03
5.94E+03
6.09E+03
6.24E+03
6.56E+03
6.89E+03
7.24E+03
7.61E+03
8.00E+03
Body
Burden
(ng/kg)
6.49E+03
6.67E+03
6.87E+03
7.06E+03
7.27E+03
7.48E+03
7.69E+03
7.92E+03
8.15E+03
8.38E+03
8.62E+03
8.87E+03
9.13E+03
9.40E+03
9.67E+03
9.95E+03
1.02E+04
1.05E+04
1.08E+04
1.12E+04
1.15E+04
1.18E+04
1.22E+04
1.25E+04
1.29E+04
1.33E+04
1.36E+04
1.40E+04
1.44E+04
1.49E+04
1.53E+04
1.62E+04
1.71E+04
1.81E+04
1.92E+04
2.03E+04
Blood
(ng/kg)
3.21E+01
3.28E+01
3.36E+01
3.43E+01
3.51E+01
3.60E+01
3.68E+01
3.77E+01
3.86E+01
3.95E+01
4.04E+01
4.14E+01
4.23E+01
4.34E+01
4.44E+01
4.55E+01
4.66E+01
4.77E+01
4.88E+01
5.00E+01
5.12E+01
5.25E+01
5.38E+01
5.51E+01
5.65E+01
5.79E+01
5.93E+01
6.08E+01
6.23E+01
6.38E+01
6.54E+01
6.87E+01
7.22E+01
7.58E+01
7.97E+01
8.38E+01
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
1.42E+01
1.51E+01
1.60E+01
1.70E+01
1.80E+01
1.90E+01
2.02E+01
2.14E+01
2.27E+01
2.40E+01
2.55E+01
2.70E+01
2.86E+01
3.04E+01
3.22E+01
3.41E+01
3.62E+01
3.83E+01
4.06E+01
4.31E+01
4.57E+01
4.84E+01
5.13E+01
5.44E+01
5.76E+01
6.11E+01
6.48E+01
6.86E+01
7.28E+01
7.71 E+01
8.18E+01
8.67E+01
9.19E+01
9.74E+01
1.03E+02
1.09E+02
Fat (ng/kg)
8.41E+03
8.84E+03
9.30E+03
9.79E+03
1.03E+04
1.09E+04
1.14E+04
1.20E+04
1.27E+04
1.34E+04
1.41E+04
1.49E+04
1.57E+04
1.65E+04
1.74E+04
1.84E+04
1.94E+04
2.05E+04
2.16E+04
2.28E+04
2.41E+04
2.55E+04
2.69E+04
2.84E+04
3.00E+04
3.17E+04
3.36E+04
3.55E+04
3.75E+04
3.97E+04
4.20E+04
4.44E+04
4.69E+04
4.97E+04
5.25E+04
5.56E+04
Body
Burden
(ng/kg)
2.15E+04
2.28E+04
2.41E+04
2.55E+04
2.70E+04
2.86E+04
3.03E+04
3.21E+04
3.39E+04
3.59E+04
3.80E+04
4.03E+04
4.26E+04
4.52E+04
4.78E+04
5.06E+04
5.36E+04
5.67E+04
6.00E+04
6.36E+04
6.73E+04
7.12E+04
7.54E+04
7.98E+04
8.45E+04
8.94E+04
9.46E+04
l.OOE+05
1.06E+05
1.12E+05
1.19E+05
1.25E+05
1.33E+05
1.40E+05
1.49E+05
1.57E+05
Blood
(ng/kg)
8.81E+01
9.27E+01
9.75E+01
1.03E+02
1.08E+02
1.14E+02
1.20E+02
1.26E+02
1.33E+02
1.40E+02
1.48E+02
1.56E+02
1.64E+02
1.73E+02
1.83E+02
1.93E+02
2.03E+02
2.15E+02
2.27E+02
2.39E+02
2.53E+02
2.67E+02
2.82E+02
2.98E+02
3.15E+02
3.33E+02
3.52E+02
3.72E+02
3.93E+02
4.16E+02
4.40E+02
4.65E+02
4.92E+02
5.20E+02
5.51E+02
5.83E+02

-------
Nongestational 5- Year Peak Average
Intake (ng/kg-
day)
1.16E+02
1.23E+02
1.30E+02
1.38E+02
1.46E+02
Fat (ng/kg)
5.88E+04
6.23E+04
6.59E+04
6.97E+04
7.38E+04
Body
Burden
(ng/kg)
1.66E+05
1.76E+05
1.86E+05
1.96E+05
2.07E+05
Blood
(ng/kg)
6.17E+02
6.53E+02
6.91E+02
7.31E+02
7.74E+02
w

-------
    E.4.3.  Gestational
to
Gestational Average
Intake (ng/kg-
day)
1.03E-09
1.09E-09
1.16E-09
1.23E-09
1.30E-09
1.38E-09
1.46E-09
1.55E-09
1.64E-09
1.74E-09
1.84E-09
1.95E-09
2.07E-09
2.20E-09
2.33E-09
2.47E-09
2.62E-09
2.77E-09
2.94E-09
3.12E-09
3.30E-09
3.50E-09
3.71E-09
3.93E-09
4.17E-09
4.42E-09
4.68E-09
4.97E-09
5.26E-09
5.58E-09
5.91E-09
6.27E-09
6.65E-09
7.04E-09
7.47E-09
Fat (ng/kg)
2.89E-05
3.07E-05
3.25E-05
3.45E-05
3.65E-05
3.87E-05
4.11E-05
4.35E-05
4.61E-05
4.88E-05
5.18E-05
5.49E-05
5.82E-05
6.17E-05
6.53E-05
6.93E-05
7.34E-05
7.79E-05
8.25E-05
8.74E-05
9.27E-05
9.83E-05
1.04E-04
1.10E-04
1.17E-04
1.24E-04
1.31E-04
1.39E-04
1.48E-04
1.57E-04
1.66E-04
1.76E-04
1.86E-04
1.98E-04
2.09E-04
Body
Burden
(ng/kg)
1.14E-05
1.21E-05
1.28E-05
1.36E-05
1.44E-05
1.53E-05
1.62E-05
1.71E-05
1.82E-05
1.92E-05
2.04E-05
2.16E-05
2.29E-05
2.43E-05
2.58E-05
2.73E-05
2.89E-05
3.07E-05
3.25E-05
3.45E-05
3.65E-05
3.88E-05
4.11E-05
4.35E-05
4.61E-05
4.89E-05
5.18E-05
5.49E-05
5.83E-05
6.18E-05
6.55E-05
6.93E-05
7.35E-05
7.79E-05
8.26E-05
Blood
(ng/kg)
3.05E-07
3.23E-07
3.42E-07
3.63E-07
3.89E-07
4.07E-07
4.31E-07
4.54E-07
4.81E-07
5.14E-07
5.45E-07
5.78E-07
6.13E-07
6.49E-07
6.88E-07
7.30E-07
7.73E-07
8.18E-07
8.69E-07
9.21E-07
9.76E-07
1.03E-06
1.09E-06
1.16E-06
1.23E-06
1.31E-06
1.38E-06
1.47E-06
1.55E-06
1.65E-06
1.73E-06
1.85E-06
1.96E-06
2.08E-06
2.21E-06
Gestational Average
Intake (ng/kg-
day)
7.92E-09
8.39E-09
8.89E-09
9.43E-09
9.99E-09
1.06E-08
1.12E-08
1.19E-08
1.26E-08
1.34E-08
1.42E-08
1.50E-08
1.59E-08
1.69E-08
1.79E-08
1.90E-08
2.01E-08
2.13E-08
2.26E-08
2.39E-08
2.54E-08
2.69E-08
2.85E-08
3.02E-08
3.20E-08
3.40E-08
3.60E-08
3.82E-08
4.05E-08
4.29E-08
4.55E-08
4.82E-08
5.11E-08
5.41E-08
5.74E-08
6.08E-08
6.45E-08
Fat (ng/kg)
2.22E-04
2.35E-04
2.49E-04
2.64E-04
2.80E-04
2.97E-04
3.15E-04
3.34E-04
3.54E-04
3.75E-04
3.97E-04
4.21E-04
4.47E-04
4.73E-04
5.01E-04
5.31E-04
5.63E-04
5.97E-04
6.33E-04
6.71E-04
7.11E-04
7.53E-04
7.98E-04
8.46E-04
8.97E-04
9.50E-04
1.01E-03
1.07E-03
1.13E-03
1.20E-03
1.27E-03
1.35E-03
1.43E-03
1.51E-03
1.60E-03
1.70E-03
1.80E-03
Body
Burden
(ng/kg)
8.75E-05
9.27E-05
9.83E-05
1.04E-04
1.10E-04
1.17E-04
1.24E-04
1.32E-04
1.40E-04
1.48E-04
1.57E-04
1.66E-04
1.76E-04
1.86E-04
1.98E-04
2.10E-04
2.22E-04
2.35E-04
2.49E-04
2.65E-04
2.80E-04
2.97E-04
3.15E-04
3.34E-04
3.54E-04
3.75E-04
3.97E-04
4.21E-04
4.46E-04
4.73E-04
5.01E-04
5.31E-04
5.63E-04
5.97E-04
6.32E-04
6.70E-04
7.10E-04
Blood
(ng/kg)
2.34E-06
2.48E-06
2.63E-06
2.78E-06
2.95E-06
3.14E-06
3.31E-06
3.52E-06
3.70E-06
3.95E-06
4.18E-06
4.43E-06
4.70E-06
4.98E-06
5.28E-06
5.59E-06
5.93E-06
6.28E-06
6.66E-06
7.03E-06
7.48E-06
7.93E-06
8.40E-06
8.91E-06
9.44E-06
l.OOE-05
1.06E-05
1.12E-05
1.19E-05
1.26E-05
1.34E-05
1.42E-05
1.50E-05
1.59E-05
1.69E-05
1.79E-05
1.90E-05
Gestational Average
Intake (ng/kg-
day)
6.84E-08
7.25E-08
7.68E-08
8.14E-08
8.63E-08
9.15E-08
9.70E-08
1.03E-07
1.09E-07
1.15E-07
1.22E-07
1.30E-07
1.38E-07
1.46E-07
1.55E-07
1.64E-07
1.74E-07
1.84E-07
1.95E-07
2.07E-07
2.19E-07
2.32E-07
2.46E-07
2.61E-07
2.77E-07
2.93E-07
3.11E-07
3.30E-07
3.49E-07
3.70E-07
3.93E-07
4.16E-07
4.41E-07
4.68E-07
4.96E-07
5.25E-07
5.57E-07
Fat (ng/kg)
1.91E-03
2.02E-03
2.14E-03
2.27E-03
2.41E-03
2.55E-03
2.70E-03
2.86E-03
3.03E-03
3.22E-03
3.41E-03
3.61E-03
3.83E-03
4.05E-03
4.30E-03
4.55E-03
4.82E-03
5.11E-03
5.41E-03
5.74E-03
6.08E-03
6.44E-03
6.82E-03
7.23E-03
7.66E-03
8.11E-03
8.60E-03
9.11E-03
9.65E-03
1.02E-02
1.08E-02
1.15E-02
1.21E-02
1.29E-02
1.36E-02
1.44E-02
1.53E-02
Body
Burden
(ng/kg)
7.53E-04
7.98E-04
8.45E-04
8.96E-04
9.50E-04
1.01E-03
1.07E-03
1.13E-03
1.20E-03
1.27E-03
1.35E-03
1.43E-03
1.51E-03
1 .60E-03
1.70E-03
1.80E-03
1.90E-03
2.02E-03
2.14E-03
2.27E-03
2.40E-03
2.54E-03
2.70E-03
2.86E-03
3.03E-03
3.21E-03
3.40E-03
3.60E-03
3.82E-03
4.04E-03
4.28E-03
4.54E-03
4.81E-03
5.09E-03
5.39E-03
5.72E-03
6.05E-03
Blood
(ng/kg)
2.01E-05
2.13E-05
2.26E-05
2.39E-05
2.53E-05
2.68E-05
2.85E-05
3.01E-05
3.19E-05
3.39E-05
3.59E-05
3.80E-05
4.03E-05
4.27E-05
4.52E-05
4.79E-05
5.08E-05
5.38E-05
5.70E-05
6.04E-05
6.40E-05
6.78E-05
7.18E-05
7.61E-05
8.06E-05
8.54E-05
9.05E-05
9.58E-05
1.02E-04
1.08E-04
1.14E-04
1.21E-04
1 .28E-04
1.35E-04
1.43E-04
1.52E-04
1.61E-04

-------
Gestational Average
Intake (ng/kg-
day)
5.90E-07
6.26E-07
6.63E-07
7.03E-07
7.45E-07
7.90E-07
8.37E-07
8.88E-07
9.41E-07
9.97E-07
1.01E-06
1.03E-06
1.04E-06
1.06E-06
1.07E-06
1.09E-06
1.11E-06
1.12E-06
1.14E-06
1.16E-06
1.17E-06
1.19E-06
1.21E-06
1.23E-06
1.24E-06
1.26E-06
1.28E-06
1.30E-06
1.32E-06
1.34E-06
1.36E-06
1.38E-06
1.40E-06
1.42E-06
1.44E-06
1.46E-06
1.49E-06
Fat (ng/kg)
1.62E-02
1.72E-02
1.82E-02
1.92E-02
2.04E-02
2.16E-02
2.29E-02
2.42E-02
2.56E-02
2.71E-02
2.75E-02
2.79E-02
2.83E-02
2.88E-02
2.92E-02
2.96E-02
3.00E-02
3.05E-02
3.09E-02
3.14E-02
3.18E-02
3.23E-02
3.27E-02
3.32E-02
3.37E-02
3.42E-02
3.47E-02
3.52E-02
3.57E-02
3.62E-02
3.68E-02
3.73E-02
3.78E-02
3.84E-02
3.89E-02
3.95E-02
4.01E-02
Body
Burden
(ng/kg)
6.41E-03
6.79E-03
7.20E-03
7.62E-03
8.08E-03
8.55E-03
9.06E-03
9.60E-03
1.02E-02
1.08E-02
1.09E-02
1.11E-02
1.12E-02
1.14E-02
1.16E-02
1.17E-02
1.19E-02
1.21E-02
1.23E-02
1.24E-02
1.26E-02
1.28E-02
1.30E-02
1.32E-02
1.34E-02
1.36E-02
1.38E-02
1.40E-02
1.42E-02
1.44E-02
1.46E-02
1.48E-02
1.50E-02
1.53E-02
1.55E-02
1.57E-02
1.59E-02
Blood
(ng/kg)
1.70E-04
1.81E-04
1.91E-04
2.02E-04
2.14E-04
2.27E-04
2.40E-04
2.55E-04
2.70E-04
2.86E-04
2.90E-04
2.94E-04
2.98E-04
3.03E-04
3.07E-04
3.11E-04
3.16E-04
3.21E-04
3.25E-04
3.30E-04
3.35E-04
3.40E-04
3.45E-04
3.50E-04
3.55E-04
3.60E-04
3.65E-04
3.71E-04
3.76E-04
3.81E-04
3.87E-04
3.93E-04
3.98E-04
4.04E-04
4.10E-04
4.16E-04
4.22E-04
Gestational Average
Intake (ng/kg-
day)
1.53E-06
1.58E-06
1.62E-06
1.67E-06
1.72E-06
1.77E-06
1.83E-06
1.88E-06
1.94E-06
2.00E-06
2.06E-06
2.12E-06
2.18E-06
2.25E-06
2.32E-06
2.39E-06
2.46E-06
2.53E-06
2.61E-06
2.68E-06
2.76E-06
2.85E-06
2.93E-06
3.02E-06
3.11E-06
3.21E-06
3.30E-06
3.40E-06
3.50E-06
3.61E-06
3.72E-06
3.83E-06
3.94E-06
4.06E-06
4.18E-06
4.31E-06
4.44E-06
Fat (ng/kg)
4.13E-02
4.25E-02
4.37E-02
4.50E-02
4.63E-02
4.77E-02
4.91E-02
5.05E-02
5.20E-02
5.35E-02
5.50E-02
5.66E-02
5.83E-02
6.00E-02
6.17E-02
6.35E-02
6.54E-02
6.73E-02
6.92E-02
7.12E-02
7.33E-02
7.54E-02
7.76E-02
7.98E-02
8.22E-02
8.45E-02
8.70E-02
8.95E-02
9.21E-02
9.47E-02
9.74E-02
l.OOE-01
1.03E-01
1.06E-01
1.09E-01
1.12E-01
1.15E-01
Body
Burden
(ng/kg)
1.64E-02
1.69E-02
1.74E-02
1.79E-02
1.84E-02
1.90E-02
1.95E-02
2.01E-02
2.07E-02
2.13E-02
2.19E-02
2.26E-02
2.32E-02
2.39E-02
2.46E-02
2.53E-02
2.61E-02
2.68E-02
2.76E-02
2.84E-02
2.92E-02
3.01E-02
3.10E-02
3.19E-02
3.28E-02
3.38E-02
3.47E-02
3.57E-02
3.68E-02
3.79E-02
3.90E-02
4.01E-02
4.13E-02
4.25E-02
4.37E-02
4.49E-02
4.63E-02
Blood
(ng/kg)
4.34E-04
4.47E-04
4.60E-04
4.73E-04
4.87E-04
5.02E-04
5.16E-04
5.31E-04
5.47E-04
5.63E-04
5.79E-04
5.96E-04
6.13E-04
6.31E-04
6.50E-04
6.68E-04
6.88E-04
7.08E-04
7.28E-04
7.49E-04
7.71 E-04
7.94E-04
8.17E-04
8.40E-04
8.64E-04
8.89E-04
9.15E-04
9.42E-04
9.69E-04
9.97E-04
1.03E-03
1.05E-03
1 .09E-03
1.12E-03
1.15E-03
1.18E-03
1.22E-03
Gestational Average
Intake (ng/kg-
day)
4.57E-06
4.71E-06
4.85E-06
4.99E-06
5.14E-06
5.30E-06
5.46E-06
5.62E-06
5.79E-06
5.96E-06
6.14E-06
6.33E-06
6.52E-06
6.71E-06
6.91E-06
7.12E-06
7.33E-06
7.55E-06
7.78E-06
8.01E-06
8.25E-06
8.50E-06
8.76E-06
9.02E-06
9.29E-06
9.57E-06
9.86E-06
1.02E-05
1.05E-05
1.08E-05
1.11E-05
1.14E-05
1.18E-05
1.21E-05
1.25E-05
1.29E-05
1.32E-05
Fat (ng/kg)
1.19E-01
1.22E-01
1.26E-01
1.29E-01
1.33E-01
1.37E-01
1.41E-01
1.45E-01
1.49E-01
1.53E-01
1.57E-01
1.62E-01
1.66E-01
1.71E-01
1.76E-01
1.81E-01
1.86E-01
1.91E-01
1.97E-01
2.02E-01
2.08E-01
2.14E-01
2.20E-01
2.26E-01
2.33E-01
2.39E-01
2.46E-01
2.53E-01
2.60E-01
2.67E-01
2.74E-01
2.82E-01
2.90E-01
2.98E-01
3.06E-01
3.15E-01
3.23E-01
Body
Burden
(ng/kg)
4.76E-02
4.90E-02
5.04E-02
5.18E-02
5.33E-02
5.49E-02
5.64E-02
5.81E-02
5.97E-02
6.15E-02
6.32E-02
6.51E-02
6.69E-02
6.88E-02
7.08E-02
7.29E-02
7.49E-02
7.71 E-02
7.93E-02
8.16E-02
8.39E-02
8.63E-02
8.88E-02
9.13E-02
9.39E-02
9.66E-02
9.93E-02
1.02E-01
1.05E-01
1.08E-01
1.11E-01
1.14E-01
1.17E-01
1.21E-01
1.24E-01
1.28E-01
1.31E-01
Blood
(ng/kg)
1.25E-03
1.29E-03
1.32E-03
1.36E-03
1.40E-03
1.44E-03
1.48E-03
1.52E-03
1.57E-03
1.61E-03
1.66E-03
1.70E-03
1.75E-03
1.80E-03
1.85E-03
1.90E-03
1.96E-03
2.01E-03
2.07E-03
2.13E-03
2.19E-03
2.25E-03
2.31E-03
2.38E-03
2.45E-03
2.51E-03
2.59E-03
2.66E-03
2.73E-03
2.81E-03
2.89E-03
2.97E-03
3.05E-03
3.13E-03
3.22E-03
3.31E-03
3.40E-03

-------
Gestational Average
Intake (ng/kg-
day)
1.36E-05
1.41E-05
1.45E-05
1.49E-05
1.54E-05
1.58E-05
1.63E-05
1.68E-05
1.73E-05
1.78E-05
1.83E-05
1.89E-05
1.95E-05
2.00E-05
2.06E-05
2.13E-05
2.19E-05
2.25E-05
2.32E-05
2.39E-05
2.46E-05
2.54E-05
2.61E-05
2.69E-05
2.77E-05
2.86E-05
2.94E-05
3.03E-05
3.12E-05
3.21E-05
3.31E-05
3.41E-05
3.51E-05
3.62E-05
3.73E-05
3.84E-05
3.95E-05
Fat (ng/kg)
3.32E-01
3.42E-01
3.51E-01
3.61E-01
3.71E-01
3.81E-01
3.91E-01
4.02E-01
4.13E-01
4.24E-01
4.36E-01
4.48E-01
4.60E-01
4.73E-01
4.85E-01
4.99E-01
5.12E-01
5.26E-01
5.40E-01
5.55E-01
5.70E-01
5.85E-01
6.01E-01
6.17E-01
6.33E-01
6.50E-01
6.68E-01
6.85E-01
7.04E-01
7.22E-01
7.41E-01
7.61E-01
7.81E-01
8.02E-01
8.24E-01
8.46E-01
8.68E-01
Body
Burden
(ng/kg)
1.35E-01
1.39E-01
1.43E-01
1.47E-01
1.51E-01
1.55E-01
1.59E-01
1.64E-01
1.68E-01
1.73E-01
1.78E-01
1.83E-01
1.88E-01
1.93E-01
1.99E-01
2.04E-01
2.10E-01
2.16E-01
2.22E-01
2.28E-01
2.34E-01
2.40E-01
2.47E-01
2.54E-01
2.61E-01
2.68E-01
2.75E-01
2.83E-01
2.91E-01
2.98E-01
3.07E-01
3.15E-01
3.24E-01
3.32E-01
3.42E-01
3.51E-01
3.61E-01
Blood
(ng/kg)
3.50E-03
3.59E-03
3.69E-03
3.79E-03
3.90E-03
4.01E-03
4.12E-03
4.23E-03
4.34E-03
4.46E-03
4.59E-03
4.71E-03
4.84E-03
4.97E-03
5.11E-03
5.24E-03
5.39E-03
5.53E-03
5.68E-03
5.83E-03
5.99E-03
6.15E-03
6.32E-03
6.49E-03
6.66E-03
6.84E-03
7.02E-03
7.21E-03
7.40E-03
7.60E-03
7.80E-03
8.00E-03
8.21E-03
8.43E-03
8.66E-03
8.89E-03
9.12E-03
Gestational Average
Intake (ng/kg-
day)
4.07E-05
4.19E-05
4.32E-05
4.45E-05
4.58E-05
4.72E-05
4.86E-05
5.01E-05
5.16E-05
5.31E-05
5.47E-05
5.64E-05
5.81E-05
5.98E-05
6.16E-05
6.34E-05
6.54E-05
6.73E-05
6.93E-05
7.14E-05
7.36E-05
7.58E-05
7.80E-05
8.04E-05
8.28E-05
8.53E-05
8.78E-05
9.05E-05
9.32E-05
9.60E-05
9.89E-05
1.02E-04
1.05E-04
1.08E-04
1.11E-04
1.15E-04
1.18E-04
Fat (ng/kg)
8.91E-01
9.14E-01
9.38E-01
9.62E-01
9.87E-01
1.01E+00
1.04E+00
1.07E+00
1.09E+00
1.12E+00
1.15E+00
1.18E+00
1.21E+00
1.24E+00
1.27E+00
1.30E+00
1.34E+00
1.37E+00
1.40E+00
1.44E+00
1.48E+00
1.51E+00
1.55E+00
1.59E+00
1.63E+00
1.67E+00
1.71E+00
1.75E+00
1.80E+00
1.84E+00
1.89E+00
1.94E+00
1.98E+00
2.03E+00
2.08E+00
2.13E+00
2.18E+00
Body
Burden
(ng/kg)
3.70E-01
3.80E-01
3.91E-01
4.01E-01
4.12E-01
4.23E-01
4.34E-01
4.46E-01
4.58E-01
4.70E-01
4.82E-01
4.95E-01
5.08E-01
5.22E-01
5.35E-01
5.49E-01
5.63E-01
5.78E-01
5.93E-01
6.09E-01
6.25E-01
6.41E-01
6.58E-01
6.75E-01
6.92E-01
7.10E-01
7.28E-01
7.48E-01
7.67E-01
7.87E-01
8.08E-01
8.30E-01
8.52E-01
8.74E-01
8.96E-01
9.19E-01
9.41E-01
Blood
(ng/kg)
9.36E-03
9.61E-03
9.86E-03
1.01E-02
1.04E-02
1.07E-02
1.09E-02
1.12E-02
1.15E-02
1.18E-02
1.21E-02
1.24E-02
1.27E-02
1.30E-02
1.34E-02
1.37E-02
1.40E-02
1.44E-02
1.48E-02
1.51E-02
1.55E-02
1.59E-02
1.63E-02
1.67E-02
1.71E-02
1.75E-02
1.80E-02
1.84E-02
1.89E-02
1.94E-02
1.98E-02
2.03E-02
2.09E-02
2.14E-02
2.19E-02
2.24E-02
2.29E-02
Gestational Average
Intake (ng/kg-
day)
1.22E-04
1.25E-04
1.29E-04
1.33E-04
1.37E-04
1.41E-04
1.45E-04
1.50E-04
1.54E-04
1.59E-04
1.63E-04
1.68E-04
1.73E-04
1.79E-04
1.84E-04
1.89E-04
1.95E-04
2.01E-04
2.07E-04
2.13E-04
2.20E-04
2.26E-04
2.33E-04
2.40E-04
2.47E-04
2.55E-04
2.62E-04
2.70E-04
2.78E-04
2.86E-04
2.95E-04
3.04E-04
3.13E-04
3.22E-04
3.32E-04
3.42E-04
3.52E-04
Fat (ng/kg)
2.23E+00
2.29E+00
2.34E+00
2.40E+00
2.46E+00
2.51E+00
2.57E+00
2.64E+00
2.70E+00
2.76E+00
2.83E+00
2.89E+00
2.96E+00
3.04E+00
3.12E+00
3.19E+00
3.25E+00
3.34E+00
3.42E+00
3.50E+00
3.58E+00
3.67E+00
3.75E+00
3.84E+00
3.93E+00
4.02E+00
4.11E+00
4.21E+00
4.32E+00
4.41E+00
4.50E+00
4.60E+00
4.70E+00
4.81E+00
4.92E+00
5.02E+00
5.13E+00
Body
Burden
(ng/kg)
9.65E-01
9.89E-01
1.01E+00
1.04E+00
1.07E+00
1.09E+00
1.12E+00
1.15E+00
1.18E+00
1.21E+00
1.24E+00
1.27E+00
1.30E+00
1.34E+00
1.37E+00
1.41E+00
1.44E+00
1.48E+00
1.51E+00
1.55E+00
1.59E+00
1.63E+00
1.67E+00
1.71E+00
1.76E+00
1.80E+00
1.84E+00
1.89E+00
1.94E+00
1.99E+00
2.03E+00
2.08E+00
2.13E+00
2.18E+00
2.23E+00
2.29E+00
2.34E+00
Blood
(ng/kg)
2.35E-02
2.40E-02
2.46E-02
2.52E-02
2.58E-02
2.64E-02
2.71E-02
2.77E-02
2.83E-02
2.90E-02
2.97E-02
3.04E-02
3.11E-02
3.19E-02
3.27E-02
3.35E-02
3.42E-02
3.51E-02
3.59E-02
3.68E-02
3.77E-02
3.85E-02
3.94E-02
4.04E-02
4.13E-02
4.22E-02
4.32E-02
4.42E-02
4.53E-02
4.63E-02
4.73E-02
4.84E-02
4.94E-02
5.05E-02
5.16E-02
5.28E-02
5.39E-02

-------
Gestational Average
Intake (ng/kg-
day)
3.63E-04
3.74E-04
3.85E-04
3.97E-04
4.08E-04
4.21E-04
4.33E-04
4.46E-04
4.60E-04
4.74E-04
4.88E-04
5.02E-04
5.17E-04
5.33E-04
5.49E-04
5.65E-04
5.82E-04
6.00E-04
6.18E-04
6.36E-04
6.55E-04
6.75E-04
6.95E-04
7.16E-04
7.38E-04
7.60E-04
7.83E-04
8.06E-04
8.30E-04
8.55E-04
8.81E-04
9.07E-04
9.21E-04
9.35E-04
9.49E-04
9.63E-04
9.69E-04
Fat (ng/kg)
5.24E+00
5.37E+00
5.50E+00
5.62E+00
5.75E+00
5.87E+00
6.01E+00
6.14E+00
6.28E+00
6.43E+00
6.57E+00
6.72E+00
6.87E+00
7.02E+00
7.17E+00
7.33E+00
7.49E+00
7.65E+00
7.82E+00
7.99E+00
8.16E+00
8.34E+00
8.52E+00
8.70E+00
8.89E+00
9.08E+00
9.27E+00
9.47E+00
9.67E+00
9.88E+00
1.01E+01
1.03E+01
1.04E+01
1.05E+01
1.06E+01
1.07E+01
1.08E+01
Body
Burden
(ng/kg)
2.40E+00
2.46E+00
2.52E+00
2.58E+00
2.65E+00
2.71E+00
2.78E+00
2.85E+00
2.91E+00
2.99E+00
3.06E+00
3.14E+00
3.21E+00
3.29E+00
3.37E+00
3.45E+00
3.53E+00
3.62E+00
3.71E+00
3.79E+00
3.89E+00
3.98E+00
4.07E+00
4.17E+00
4.27E+00
4.37E+00
4.47E+00
4.58E+00
4.69E+00
4.80E+00
4.91E+00
5.03E+00
5.09E+00
5.15E+00
5.21E+00
5.27E+00
5.30E+00
Blood
(ng/kg)
5.51E-02
5.64E-02
5.77E-02
5.90E-02
6.03E-02
6.17E-02
6.31E-02
6.45E-02
6.60E-02
6.75E-02
6.90E-02
7.05E-02
7.21E-02
7.37E-02
7.53E-02
7.70E-02
7.87E-02
8.04E-02
8.21E-02
8.39E-02
8.57E-02
8.76E-02
8.95E-02
9.14E-02
9.33E-02
9.53E-02
9.74E-02
9.94E-02
1.02E-01
1.04E-01
1.06E-01
1.08E-01
1.09E-01
1.10E-01
1.12E-01
1.13E-01
1.13E-01
Gestational Average
Intake (ng/kg-
day)
9.77E-04
9.84E-04
9.91E-04
9.98E-04
1.01E-03
1.02E-03
1.04E-03
1.05E-03
1.07E-03
1.08E-03
1.10E-03
1.12E-03
1.13E-03
1.15E-03
1.17E-03
1.18E-03
1.20E-03
1.22E-03
1.24E-03
1.26E-03
1.27E-03
1.29E-03
1.31E-03
1.33E-03
1.35E-03
1.37E-03
1.39E-03
1.41E-03
1.43E-03
1.46E-03
1.48E-03
1.50E-03
1.52E-03
1.54E-03
1.57E-03
1.59E-03
1.61E-03
Fat (ng/kg)
1.09E+01
1.09E+01
1.10E+01
1.10E+01
1.11E+01
1.12E+01
1.13E+01
1.14E+01
1.16E+01
1.17E+01
1.18E+01
1.19E+01
1.20E+01
1.22E+01
1.23E+01
1.24E+01
1.25E+01
1.27E+01
1.28E+01
1.29E+01
1.31E+01
1.32E+01
1.33E+01
1.35E+01
1.36E+01
1.38E+01
1.39E+01
1.40E+01
1.42E+01
1.43E+01
1.45E+01
1.46E+01
1.48E+01
1.49E+01
1.51E+01
1.52E+01
1.54E+01
Body
Burden
(ng/kg)
5.33E+00
5.36E+00
5.39E+00
5.42E+00
5.46E+00
5.52E+00
5.58E+00
5.65E+00
5.72E+00
5.78E+00
5.85E+00
5.92E+00
5.99E+00
6.06E+00
6.13E+00
6.20E+00
6.27E+00
6.34E+00
6.42E+00
6.49E+00
6.57E+00
6.64E+00
6.72E+00
6.80E+00
6.88E+00
6.96E+00
7.04E+00
7.12E+00
7.21E+00
7.29E+00
7.37E+00
7.46E+00
7.55E+00
7.63E+00
7.72E+00
7.81E+00
7.90E+00
Blood
(ng/kg)
1.14E-01
1.15E-01
1.15E-01
1.16E-01
1.16E-01
1.18E-01
1.19E-01
1.20E-01
1.21E-01
1.23E-01
1.24E-01
1.25E-01
1.26E-01
1.28E-01
1.29E-01
1.30E-01
1.32E-01
1.33E-01
1.34E-01
1.36E-01
1.37E-01
1.39E-01
1.40E-01
1.41E-01
1.43E-01
1.44E-01
1.46E-01
1.47E-01
1.49E-01
1.50E-01
1.52E-01
1.54E-01
1.55E-01
1.57E-01
1.58E-01
1.60E-01
1.62E-01
Gestational Average
Intake (ng/kg-
day)
1.64E-03
1.66E-03
1.69E-03
1.71E-03
1.74E-03
1.76E-03
1.79E-03
1.82E-03
1.84E-03
1.87E-03
1.90E-03
1.93E-03
1.96E-03
1.99E-03
2.02E-03
2.08E-03
2.14E-03
2.20E-03
2.27E-03
2.34E-03
2.41E-03
2.48E-03
2.55E-03
2.63E-03
2.71E-03
2.79E-03
2.87E-03
2.96E-03
3.05E-03
3.14E-03
3.23E-03
3.33E-03
3.43E-03
3.53E-03
3.64E-03
3.75E-03
3.81E-03
Fat (ng/kg)
1.55E+01
1.57E+01
1.59E+01
1.60E+01
1.62E+01
1.64E+01
1.65E+01
1.67E+01
1.69E+01
1.74E+01
1.92E+01
1.96E+01
1.80E+01
1.79E+01
1.81E+01
1.84E+01
1.87E+01
1.91E+01
1.94E+01
1.98E+01
2.02E+01
2.06E+01
2.10E+01
2.14E+01
2.19E+01
2.23E+01
2.28E+01
2.32E+01
2.37E+01
2.41E+01
2.46E+01
2.51E+01
2.56E+01
2.61E+01
2.66E+01
2.79E+01
2.82E+01
Body
Burden
(ng/kg)
7.99E+00
8.09E+00
8.18E+00
8.28E+00
8.37E+00
8.47E+00
8.57E+00
8.67E+00
8.77E+00
9.10E+00
1.02E+01
1.04E+01
9.44E+00
9.41E+00
9.49E+00
9.67E+00
9.88E+00
1.01E+01
1.03E+01
1.06E+01
1.08E+01
1.11E+01
1.13E+01
1.16E+01
1.18E+01
1.21E+01
1.24E+01
1.27E+01
1.30E+01
1.33E+01
1.36E+01
1.39E+01
1.42E+01
1.46E+01
1.49E+01
1.57E+01
1.59E+01
Blood
(ng/kg)
1.63E-01
1.65E-01
1.67E-01
1.68E-01
1.70E-01
1.72E-01
1.73E-01
1.75E-01
1.77E-01
1.83E-01
2.02E-01
2.06E-01
1.89E-01
1.88E-01
1.89E-01
1.93E-01
1.96E-01
2.00E-01
2.04E-01
2.08E-01
2.12E-01
2.16E-01
2.21E-01
2.25E-01
2.29E-01
2.34E-01
2.39E-01
2.43E-01
2.48E-01
2.53E-01
2.58E-01
2.63E-01
2.69E-01
2.74E-01
2.79E-01
2.92E-01
2.96E-01

-------
Gestational Average
Intake (ng/kg-
day)
3.86E-03
3.98E-03
4.10E-03
4.22E-03
4.35E-03
4.48E-03
4.61E-03
4.75E-03
4.89E-03
5.04E-03
5.19E-03
5.35E-03
5.51E-03
5.67E-03
5.84E-03
5.93E-03
6.02E-03
6.20E-03
6.38E-03
6.57E-03
6.77E-03
6.98E-03
7.18E-03
7.40E-03
7.51E-03
7.62E-03
7.85E-03
8.09E-03
8.33E-03
8.58E-03
8.71E-03
8.84E-03
9.10E-03
9.37E-03
9.66E-03
9.94E-03
1.02E-02
Fat (ng/kg)
2.69E+01
2.73E+01
2.78E+01
2.83E+01
2.88E+01
2.94E+01
2.99E+01
3.05E+01
3.11E+01
3.30E+01
3.41E+01
3.48E+01
3.56E+01
3.63E+01
3.70E+01
3.74E+01
3.78E+01
3.85E+01
3.93E+01
4.01E+01
4.08E+01
4.16E+01
4.25E+01
4.33E+01
4.37E+01
4.35E+01
4.42E+01
4.50E+01
4.59E+01
4.68E+01
4.72E+01
4.77E+01
4.86E+01
4.95E+01
5.05E+01
5.15E+01
5.25E+01
Body
Burden
(ng/kg)
1.51E+01
1.53E+01
1.57E+01
1.60E+01
1.63E+01
1.67E+01
1.71E+01
1.75E+01
1.79E+01
1.92E+01
1.99E+01
2.05E+01
2.10E+01
2.15E+01
2.20E+01
2.23E+01
2.26E+01
2.31E+01
2.36E+01
2.42E+01
2.48E+01
2.54E+01
2.60E+01
2.66E+01
2.69E+01
2.68E+01
2.73E+01
2.79E+01
2.85E+01
2.92E+01
2.96E+01
2.99E+01
3.06E+01
3.13E+01
3.21E+01
3.28E+01
3.36E+01
Blood
(ng/kg)
2.83E-01
2.87E-01
2.91E-01
2.97E-01
3.02E-01
3.08E-01
3.14E-01
3.20E-01
3.26E-01
3.47E-01
3.57E-01
3.65E-01
3.73E-01
3.81E-01
3.88E-01
3.92E-01
3.96E-01
4.04E-01
4.12E-01
4.20E-01
4.28E-01
4.37E-01
4.45E-01
4.54E-01
4.58E-01
4.57E-01
4.64E-01
4.72E-01
4.81E-01
4.90E-01
4.95E-01
5.00E-01
5.10E-01
5.19E-01
5.29E-01
5.40E-01
5.50E-01
Gestational Average
Intake (ng/kg-
day)
1.06E-02
1.09E-02
1.12E-02
1.15E-02
1.19E-02
1.22E-02
1.26E-02
1.30E-02
1.34E-02
1.38E-02
1.42E-02
1.46E-02
1.50E-02
1.55E-02
1.60E-02
1.64E-02
1.69E-02
1.74E-02
1.80E-02
1.85E-02
1.91E-02
1.96E-02
2.02E-02
2.08E-02
2.14E-02
2.21E-02
2.28E-02
2.34E-02
2.41E-02
2.49E-02
2.56E-02
2.64E-02
2.72E-02
2.80E-02
2.88E-02
2.97E-02
3.06E-02
Fat (ng/kg)
5.35E+01
5.45E+01
5.56E+01
5.67E+01
5.74E+01
5.85E+01
5.96E+01
6.11E+01
6.23E+01
6.35E+01
6.48E+01
6.70E+01
6.79E+01
6.86E+01
6.99E+01
7.12E+01
7.26E+01
7.39E+01
7.54E+01
7.68E+01
7.83E+01
8.07E+01
8.20E+01
8.34E+01
8.45E+01
8.61E+01
8.78E+01
8.95E+01
9.12E+01
9.32E+01
9.50E+01
9.68E+01
9.86E+01
l.OOE+02
1.02E+02
1.04E+02
1.06E+02
Body
Burden
(ng/kg)
3.44E+01
3.52E+01
3.61E+01
3.69E+01
3.75E+01
3.84E+01
3.93E+01
4.05E+01
4.15E+01
4.25E+01
4.36E+01
4.55E+01
4.62E+01
4.68E+01
4.79E+01
4.90E+01
5.02E+01
5.14E+01
5.27E+01
5.40E+01
5.53E+01
5.74E+01
5.86E+01
5.98E+01
6.08E+01
6.23E+01
6.38E+01
6.54E+01
6.70E+01
6.88E+01
7.05E+01
7.22E+01
7.40E+01
7.58E+01
7.76E+01
7.95E+01
8.13E+01
Blood
(ng/kg)
5.61E-01
5.72E-01
5.83E-01
5.94E-01
6.02E-01
6.13E-01
6.25E-01
6.40E-01
6.53E-01
6.66E-01
6.80E-01
7.03E-01
7.12E-01
7.20E-01
7.33E-01
7.47E-01
7.61E-01
7.75E-01
7.90E-01
8.06E-01
8.21E-01
8.46E-01
8.60E-01
8.75E-01
8.86E-01
9.03E-01
9.20E-01
9.38E-01
9.57E-01
9.77E-01
9.96E-01
1.01E+00
1.03E+00
1.05E+00
1.07E+00
1.09E+00
1.11E+00
Gestational Average
Intake (ng/kg-
day)
3.15E-02
3.24E-02
3.34E-02
3.44E-02
3.54E-02
3.65E-02
3.76E-02
3.87E-02
3.99E-02
4.11E-02
4.23E-02
4.36E-02
4.49E-02
4.63E-02
4.76E-02
4.91E-02
5.05E-02
5.21E-02
5.36E-02
5.52E-02
5.69E-02
5.86E-02
6.03E-02
6.22E-02
6.40E-02
6.59E-02
6.79E-02
7.00E-02
7.21E-02
7.42E-02
7.64E-02
7.87E-02
8.11E-02
8.35E-02
8.60E-02
8.86E-02
9.13E-02
Fat (ng/kg)
1.08E+02
1.10E+02
1.12E+02
1.15E+02
1.17E+02
1.19E+02
1.21E+02
1.24E+02
1.26E+02
1.28E+02
1.31E+02
1.33E+02
1.36E+02
1.38E+02
1.41E+02
1.44E+02
1.47E+02
1.49E+02
1.52E+02
1.55E+02
1.59E+02
1.62E+02
1.64E+02
1.67E+02
1.70E+02
1.73E+02
1.77E+02
1.81E+02
1.84E+02
1.88E+02
1.91E+02
1.95E+02
1.99E+02
2.03E+02
2.06E+02
2.10E+02
2.14E+02
Body
Burden
(ng/kg)
8.33E+01
8.53E+01
8.74E+01
8.96E+01
9.18E+01
9.40E+01
9.64E+01
9.87E+01
1.01E+02
1.04E+02
1.06E+02
1.09E+02
1.12E+02
1.14E+02
1.17E+02
1.20E+02
1.23E+02
1.26E+02
1.30E+02
1.33E+02
1.37E+02
1.40E+02
1.43E+02
1.46E+02
1.50E+02
1.54E+02
1.58E+02
1.62E+02
1.66E+02
1.70E+02
1.75E+02
1.79E+02
1.84E+02
1.88E+02
1.93E+02
1.98E+02
2.03E+02
Blood
(ng/kg)
1.13E+00
1.16E+00
1.18E+00
1.20E+00
1.22E+00
1.25E+00
1.27E+00
1.30E+00
1.32E+00
1.35E+00
1.37E+00
1.40E+00
1.42E+00
1.45E+00
1.48E+00
1.51E+00
1.54E+00
1.57E+00
1.60E+00
1.63E+00
1.66E+00
1.70E+00
1.72E+00
1.75E+00
1.78E+00
1.82E+00
1.86E+00
1.89E+00
1.93E+00
1.97E+00
2.01E+00
2.05E+00
2.08E+00
2.12E+00
2.16E+00
2.21E+00
2.25E+00

-------
Gestational Average
Intake (ng/kg-
day)
9.40E-02
9.68E-02
9.97E-02
1.03E-01
1.06E-01
1.09E-01
1.12E-01
1.16E-01
1.19E-01
1.23E-01
1.26E-01
1.30E-01
1.34E-01
1.38E-01
1.42E-01
1.46E-01
1.51E-01
1.55E-01
1.60E-01
1.65E-01
1.70E-01
1.75E-01
1.80E-01
1.86E-01
1.91E-01
1.97E-01
2.03E-01
2.09E-01
2.15E-01
2.22E-01
2.28E-01
2.35E-01
2.42E-01
2.49E-01
2.57E-01
2.65E-01
2.72E-01
Fat (ng/kg)
2.19E+02
2.23E+02
2.28E+02
2.32E+02
2.37E+02
2.41E+02
2.46E+02
2.51E+02
2.55E+02
2.60E+02
2.65E+02
2.70E+02
2.75E+02
2.81E+02
2.86E+02
2.92E+02
2.97E+02
3.03E+02
3.09E+02
3.15E+02
3.21E+02
3.27E+02
3.34E+02
3.40E+02
3.47E+02
3.54E+02
3.61E+02
3.68E+02
3.75E+02
3.82E+02
3.90E+02
3.98E+02
4.05E+02
4.13E+02
4.21E+02
4.30E+02
4.38E+02
Body
Burden
(ng/kg)
2.08E+02
2.14E+02
2.20E+02
2.25E+02
2.31E+02
2.37E+02
2.43E+02
2.50E+02
2.56E+02
2.62E+02
2.69E+02
2.76E+02
2.83E+02
2.90E+02
2.98E+02
3.06E+02
3.14E+02
3.22E+02
3.30E+02
3.39E+02
3.48E+02
3.57E+02
3.66E+02
3.76E+02
3.86E+02
3.96E+02
4.07E+02
4.17E+02
4.28E+02
4.40E+02
4.52E+02
4.64E+02
4.76E+02
4.88E+02
5.01E+02
5.14E+02
5.28E+02
Blood
(ng/kg)
2.29E+00
2.34E+00
2.39E+00
2.43E+00
2.48E+00
2.53E+00
2.58E+00
2.63E+00
2.68E+00
2.72E+00
2.78E+00
2.83E+00
2.89E+00
2.94E+00
3.00E+00
3.06E+00
3.12E+00
3.18E+00
3.24E+00
3.30E+00
3.37E+00
3.43E+00
3.50E+00
3.57E+00
3.64E+00
3.71E+00
3.78E+00
3.85E+00
3.93E+00
4.01E+00
4.09E+00
4.17E+00
4.25E+00
4.33E+00
4.42E+00
4.50E+00
4.59E+00
Gestational Average
Intake (ng/kg-
day)
2.81E-01
2.89E-01
2.98E-01
3.07E-01
3.16E-01
3.25E-01
3.35E-01
3.45E-01
3.56E-01
3.66E-01
3.77E-01
3.89E-01
4.00E-01
4.12E-01
4.25E-01
4.37E-01
4.50E-01
4.64E-01
4.78E-01
4.92E-01
5.07E-01
5.22E-01
5.38E-01
5.54E-01
5.71E-01
5.88E-01
6.05E-01
6.23E-01
6.42E-01
6.61E-01
6.81E-01
7.02E-01
7.23E-01
7.44E-01
7.67E-01
7.90E-01
8.13E-01
Fat (ng/kg)
4.47E+02
4.56E+02
4.65E+02
4.74E+02
4.83E+02
4.93E+02
5.02E+02
5.12E+02
5.23E+02
5.33E+02
5.44E+02
5.55E+02
5.65E+02
5.77E+02
5.89E+02
6.00E+02
6.12E+02
6.25E+02
6.37E+02
6.50E+02
6.63E+02
6.76E+02
6.90E+02
7.04E+02
7.18E+02
7.32E+02
7.47E+02
7.62E+02
7.78E+02
7.94E+02
8.10E+02
8.26E+02
8.43E+02
8.61E+02
8.78E+02
8.96E+02
9.15E+02
Body
Burden
(ng/kg)
5.42E+02
5.57E+02
5.72E+02
5.87E+02
6.03E+02
6.19E+02
6.35E+02
6.52E+02
6.70E+02
6.88E+02
7.07E+02
7.26E+02
7.45E+02
7.66E+02
7.87E+02
8.08E+02
8.29E+02
8.52E+02
8.76E+02
8.98E+02
9.23E+02
9.48E+02
9.74E+02
l.OOE+03
1.03E+03
1.06E+03
1.08E+03
1.11E+03
1.14E+03
1.18E+03
1.21E+03
1.24E+03
1.28E+03
1.31E+03
1.35E+03
1.38E+03
1.42E+03
Blood
(ng/kg)
4.68E+00
4.78E+00
4.87E+00
4.97E+00
5.06E+00
5.16E+00
5.26E+00
5.37E+00
5.48E+00
5.59E+00
5.70E+00
5.81E+00
5.93E+00
6.05E+00
6.17E+00
6.29E+00
6.42E+00
6.55E+00
6.68E+00
6.81E+00
6.95E+00
7.09E+00
7.23E+00
7.38E+00
7.53E+00
7.68E+00
7.83E+00
7.99E+00
8.15E+00
8.32E+00
8.49E+00
8.66E+00
8.84E+00
9.02E+00
9.21E+00
9.40E+00
9.59E+00
Gestational Average
Intake (ng/kg-
day)
8.38E-01
8.63E-01
8.89E-01
9.16E-01
9.43E-01
9.71E-01
l.OOE+00
1.03E+00
1.06E+00
1.09E+00
1.13E+00
1.16E+00
1.19E+00
1.23E+00
1.27E+00
1.31E+00
1.34E+00
1.38E+00
1.43E+00
1.47E+00
1.51E+00
1.56E+00
1.61E+00
1.65E+00
1.70E+00
1.75E+00
1.81E+00
1.86E+00
1.92E+00
1.97E+00
2.03E+00
2.09E+00
2.16E+00
2.22E+00
2.29E+00
2.36E+00
2.43E+00
Fat (ng/kg)
9.33E+02
9.53E+02
9.72E+02
9.93E+02
1.01E+03
1.03E+03
1.06E+03
1.08E+03
1.10E+03
1.12E+03
1.15E+03
1.17E+03
1.20E+03
1.22E+03
1.25E+03
1.27E+03
1.30E+03
1.33E+03
1.35E+03
1.38E+03
1.41E+03
1.44E+03
1.47E+03
1.51E+03
1.54E+03
1.57E+03
1.61E+03
1.64E+03
1.68E+03
1.71E+03
1.75E+03
1.79E+03
1.83E+03
1.87E+03
1.91E+03
1.95E+03
1.99E+03
Body
Burden
(ng/kg)
1.46E+03
1.50E+03
1.54E+03
1.59E+03
1.63E+03
1.68E+03
1.72E+03
1.77E+03
1.82E+03
1.87E+03
1.92E+03
1.98E+03
2.03E+03
2.09E+03
2.15E+03
2.21E+03
2.27E+03
2.33E+03
2.40E+03
2.46E+03
2.53E+03
2.60E+03
2.68E+03
2.75E+03
2.83E+03
2.91E+03
2.99E+03
3.08E+03
3.16E+03
3.25E+03
3.34E+03
3.44E+03
3.54E+03
3.64E+03
3.74E+03
3.85E+03
3.95E+03
Blood
(ng/kg)
9.79E+00
9.99E+00
1.02E+01
1.04E+01
1.06E+01
1.08E+01
1.11E+01
1.13E+01
1.15E+01
1.18E+01
1.20E+01
1.23E+01
1.25E+01
1.28E+01
1.31E+01
1.33E+01
1.36E+01
1.39E+01
1.42E+01
1.45E+01
1.48E+01
1.51E+01
1.55E+01
1.58E+01
1.61E+01
1.65E+01
1.68E+01
1.72E+01
1.76E+01
1.79E+01
1.83E+01
1.87E+01
1.91E+01
1.96E+01
2.00E+01
2.04E+01
2.09E+01

-------
oo
Gestational Average
Intake (ng/kg-
day)
2.50E+00
2.58E+00
2.65E+00
2.73E+00
2.82E+00
2.90E+00
2.99E+00
3.08E+00
3.17E+00
3.26E+00
3.36E+00
3.46E+00
3.57E+00
3.67E+00
3.78E+00
3.90E+00
4.01E+00
4.13E+00
4.26E+00
4.39E+00
4.52E+00
4.65E+00
4.79E+00
4.94E+00
5.08E+00
5.24E+00
5.39E+00
5.56E+00
5.72E+00
5.89E+00
6.07E+00
6.25E+00
6.44E+00
6.63E+00
6.83E+00
7.04E+00
7.25E+00
Fat (ng/kg)
2.04E+03
2.08E+03
2.13E+03
2.17E+03
2.22E+03
2.27E+03
2.32E+03
2.38E+03
2.43E+03
2.48E+03
2.54E+03
2.60E+03
2.66E+03
2.72E+03
2.78E+03
2.84E+03
2.91E+03
2.98E+03
3.04E+03
3.12E+03
3.19E+03
3.26E+03
3.34E+03
3.42E+03
3.50E+03
3.58E+03
3.66E+03
3.75E+03
3.84E+03
3.93E+03
4.02E+03
4.12E+03
4.22E+03
4.32E+03
4.42E+03
4.53E+03
4.64E+03
Body
Burden
(ng/kg)
4.07E+03
4.18E+03
4.30E+03
4.42E+03
4.55E+03
4.68E+03
4.81E+03
4.95E+03
5.09E+03
5.24E+03
5.39E+03
5.54E+03
5.70E+03
5.86E+03
6.03E+03
6.20E+03
6.38E+03
6.56E+03
6.75E+03
6.95E+03
7.15E+03
7.35E+03
7.56E+03
7.78E+03
8.01E+03
8.24E+03
8.47E+03
8.72E+03
8.97E+03
9.23E+03
9.50E+03
9.77E+03
1.01E+04
1.03E+04
1.06E+04
1.10E+04
1.13E+04
Blood
(ng/kg)
2.13E+01
2.18E+01
2.23E+01
2.28E+01
2.33E+01
2.38E+01
2.44E+01
2.49E+01
2.55E+01
2.60E+01
2.66E+01
2.72E+01
2.79E+01
2.85E+01
2.91E+01
2.98E+01
3.05E+01
3.12E+01
3.19E+01
3.27E+01
3.34E+01
3.42E+01
3.50E+01
3.58E+01
3.66E+01
3.75E+01
3.84E+01
3.93E+01
4.02E+01
4.12E+01
4.22E+01
4.32E+01
4.42E+01
4.53E+01
4.64E+01
4.75E+01
4.86E+01
Gestational Average
Intake (ng/kg-
day)
7.47E+00
7.69E+00
7.92E+00
8.16E+00
8.40E+00
8.66E+00
8.92E+00
9.18E+00
9.46E+00
9.74E+00
l.OOE+01
1.06E+01
1.13E+01
1.20E+01
1.27E+01
1.34E+01
1.42E+01
1.51E+01
1.60E+01
1.70E+01
1.80E+01
1.90E+01
2.02E+01
2.14E+01
2.27E+01
2.40E+01
2.55E+01
2.70E+01
2.86E+01
3.04E+01
3.22E+01
3.41E+01
3.62E+01
3.83E+01
4.06E+01
4.31E+01
4.57E+01
Fat (ng/kg)
4.75E+03
4.87E+03
4.99E+03
5.11E+03
5.24E+03
5.37E+03
5.50E+03
5.63E+03
5.77E+03
5.92E+03
6.07E+03
6.37E+03
6.69E+03
7.03E+03
7.39E+03
7.76E+03
8.16E+03
8.59E+03
9.03E+03
9.50E+03
l.OOE+04
1.05E+04
1.11E+04
1.17E+04
1.23E+04
1.30E+04
1.37E+04
1.44E+04
1.52E+04
1.60E+04
1.69E+04
1.78E+04
1.88E+04
1.99E+04
2.10E+04
2.21E+04
2.34E+04
Body
Burden
(ng/kg)
1.16E+04
1.19E+04
1.23E+04
1.26E+04
1.30E+04
1.34E+04
1.38E+04
1.42E+04
1.46E+04
1.50E+04
1.54E+04
1.63E+04
1.73E+04
1.83E+04
1.94E+04
2.05E+04
2.17E+04
2.30E+04
2.43E+04
2.57E+04
2.72E+04
2.88E+04
3.05E+04
3.23E+04
3.42E+04
3.62E+04
3.83E+04
4.06E+04
4.30E+04
4.55E+04
4.82E+04
5.10E+04
5.40E+04
5.71E+04
6.05E+04
6.40E+04
6.78E+04
Blood
(ng/kg)
4.98E+01
5.10E+01
5.23E+01
5.36E+01
5.49E+01
5.62E+01
5.76E+01
5.91E+01
6.05E+01
6.20E+01
6.36E+01
6.68E+01
7.01E+01
7.37E+01
7.74E+01
8.14E+01
8.56E+01
9.00E+01
9.47E+01
9.96E+01
1.05E+02
1.10E+02
1.16E+02
1.22E+02
1.29E+02
1.36E+02
1.43E+02
1.51E+02
1.59E+02
1.68E+02
1.77E+02
1.87E+02
1.97E+02
2.08E+02
2.20E+02
2.32E+02
2.45E+02
Gestational Average
Intake (ng/kg-
day)
4.84E+01
5.13E+01
5.44E+01
5.76E+01
6.11E+01
6.48E+01
6.86E+01
7.28E+01
7.71E+01
8.18E+01
8.67E+01
9.19E+01
9.74E+01
1.03E+02
1.09E+02
1.16E+02
1.23E+02
1.30E+02
1.38E+02
1.46E+02
Fat (ng/kg)
2.47E+04
2.61E+04
2.75E+04
2.91E+04
3.08E+04
3.25E+04
3.44E+04
3.63E+04
3.84E+04
4.06E+04
4.30E+04
4.55E+04
4.81E+04
5.09E+04
5.38E+04
5.70E+04
6.03E+04
6.38E+04
6.76E+04
7.15E+04
Body
Burden
(ng/kg)
7.18E+04
7.60E+04
8.04E+04
8.51E+04
9.01E+04
9.53E+04
1.01E+05
1.07E+05
1.13E+05
1.20E+05
1.26E+05
1.34E+05
1.42E+05
1.50E+05
1.58E+05
1.68E+05
1.77E+05
1.87E+05
1.98E+05
2.09E+05
Blood
(ng/kg)
2.59E+02
2.73E+02
2.89E+02
3.05E+02
3.22E+02
3.41E+02
3.60E+02
3.81E+02
4.03E+02
4.26E+02
4.51E+02
4.77E+02
5.04E+02
5.33E+02
5.64E+02
5.97E+02
6.32E+02
6.69E+02
7.08E+02
7.50E+02

-------
E.5. REFERENCES

Amin, S; Moore, RW: Peterson, RE; Schantz, SL. (2000). Gestational and lactational exposure to
       TCDD or coplanar PCBs alters adult expression of saccharin preference behavior in
       female rats. Neurotoxicol Teratol 22: 675-682. http://dx.doi.org/10.1016/S0892-
       0362(00)00094-5.
Aylward, LL: Brunei RC: Carrier, G: Hays, SM: Gushing, CA; Needham, LL: Patterson, DG:
       Gerthoux. PM: Brambilla. P: Mocarelli. P. (2005a). Concentration-dependent TCDD
       elimination kinetics in humans: Toxicokinetic modeling for moderately to highly exposed
       adults from Seveso, Italy, and Vienna, Austria, and impact on dose estimates for the
       NIOSH cohort. J Expo Anal Environ Epidemiol 15: 51-65.
       http://dx.doi.org/10.1038/si.iea.7500370.
Aylward. LL: Brunei RC: Starr. TB: Carrier. G: Delzell. E: Cheng. H: Beall. C. (2005b).
       Exposure reconstruction for the TCDD-exposed NIOSH cohort using a concentration-
       and age-dependent model of elimination. Risk Anal 25: 945-956.
       http://dx.doi.0rg/10.llll/i.1539-6924.2005.00645.x.
Aylward. LL: Bodner. KM: Collins. JJ: Wilken. M: McBride. D: Burns. CJ: Hays. SM:
       Humphry, N. (2009). TCDD exposure estimation for workers at a New Zealand 2,4,5-T
       manufacturing facility based on serum sampling data. J Expo Sci Environ Epidemiol
       TEA: 1-10. http://dx.doi.org/10.1038/ies.2009.31.
Bell PR: Clode. S: Fan. MQ: Fernandes. A: Foster. PM: Jiang. T: Loizou. G: MacNicoll A:
       Miller. BG: Rose. M: Tran. L: White. S. (2007). Toxicity of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin in the developing male Wistar(Han) rat. I: No decrease in epididymal sperm count
       after a single acute dose. Toxicol Sci 99: 214-223.
       http://dx.doi.org/10.1093/toxsci/kfml40.
Bohonowych, JE; Denison, MS. (2007). Persistent binding of ligands to the aryl hydrocarbon
       receptor. Toxicol Sci 98: 99-109. http://dx.doi.org/10.1093/toxsci/kfm085.
Boverhoff, PR: Burgoon, LD; Tashiro, C: Chittim, B; Harkema, JR: Jump, DB; Zacharewski,
       TR. (2005). Temporal and dose-dependent hepatic gene expression patterns in mice
       provide new insights into TCDD-mediated hepatotoxicity. Toxicol Sci 85:  1048-1063.
       http://dx.doi.org/10.1093/toxsci/kfil62.
Cantoni, L; Salmona,  M: Rizzardini, M. (1981). Porphyrogenic effect of chronic treatment with
       2,3,7,8-tetrachlorodibenzo-p-dioxin in female rats. Dose-effect relationship following
       urinary excretion of porphyrins. Toxicol Appl Pharmacol 57: 156-163.
Carrier, G: Brunei RC: Brodeur, J. (1995a). Modeling of the toxicokinetics of polychlorinated
       dibenzo-p-dioxins and dibenzofurans in mammalians, including humans. I. Nonlinear
       distribution of PCDD/PCDF body burden between liver and adipose tissues. Toxicol
       Appl Pharmacol 131: 253-266. http://dx.doi.org/10.1006/taap.1995.1068.
Carrier, G: Brunei RC: Brodeur, J. (1995b). Modeling of the toxicokinetics of polychlorinated
       dibenzo-p-dioxins and dibenzofurans in mammalians, including humans: II. Kinetics of
       absorption and disposition of PCDDs/PCDFs. Toxicol Appl Pharmacol 131: 267-276.
       http://dx.doi.org/10.1006/taap.1995.1069.
Chu. I: Lecavalier. P:  Hakansson. H: Yagminas. A: Valli. VE: Poon. P: M. F. (2001). Mixture
       effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin and polychlorinated biphenyl congeners in
       rats. Chemosphere 43: 807-814. http://dx.doi.org/10.1016/S0045-6535(00)00437-9.
                                         E-199

-------
Chu, I; Valli, VE; Rousseaux, CG. (2007). Combined effects of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin and polychlorinated biphenyl congeners in rats. Toxicol Environ Chem 89: 71-87.
       http://dx.doi.org/10.1080/02772240600942548.
Connor, KT; Avlward, LL. (2006). Human response to dioxin: Aryl hydrocarbon receptor (AhR)
       molecular structure, function, and dose-response data for enzyme induction indicate an
       impaired human AhR. J Toxicol Environ Health B Crit Rev 9:  147-171.
       http://dx.doi.org/10.1080/15287390500196487.
Crofton. KM: Craft ES: Hedge. JM: Gennings. C: Simmons. JE: Carchman. RA: Carter. WH. Jr:
       DeVito, MJ. (2005). Thyroid-hormone-disrupting chemicals: Evidence for dose-
       dependent additivity or synergism. Environ Health Perspect 113: 1549-1554.
Croutch. CR: Lebofsky. M: Schramm. KW: Terranova. PF: Rozman. KK. (2005). 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin (TCDD) and 1,2,3,4,7,8-hexachlorodibenzo-p-dioxin
       (HxCDD) alter body weight by decreasing insulin-like growth factor I (IGF-I) signaling.
       Toxicol Sci  85: 560-571. http://dx.doi.org/10.1093/toxsci/kfil06.
Delia Porta, G: Dragani, TA; Sozzi, G. (1987). Carcinogenic effects of infantile and long-term
       2,3,7,8-tetrachlorodibenzo-p-dioxin treatment in the mouse. Tumori 73: 99-107.
Derelanko. MJ: (Eds.), HM. (1995). Handbook of toxicology (2 ed.). Boca Raton, FL: CRC
       Press.
Diliberto, JJ: Burgin, DE; Birnbaum, LS. (1997). Role of CYP1A2 in hepatic sequestration of
       dioxin: Studies using CYP1A2 knock-out mice. Biochem Biophys Res Commun 236:
       431-433. http://dx.doi.org/10.1006/bbrc.1997.6973.
Emond, C: Birnbaum, LS: DeVito, MJ.  (2004). Physiologically based pharmacokinetic  model for
       developmental exposures to TCDD in the rat. Toxicol Sci 80: 115-133.
       http://dx.doi.org/10.1093/toxsci/kfhll7.
Emond, C: Michalek, JE: Birnbaum, LS: DeVito, MJ. (2005). Comparison of the use of a
       physiologically based pharmacokinetic model and a classical pharmacokinetic model  for
       dioxin exposure assessments. Environ Health Perspect 113: 1666-1668.
Emond, C: Birnbaum, LS: DeVito, MJ.  (2006). Use  of a physiologically based pharmacokinetic
       model for rats to study the influence of body fat mass and induction of CYP1A2 on the
       pharmacokinetics of TCDD. Environ Health  Perspect 114: 1394-1400.
Fattore, E; Trossvik, C: Hakansson, H. (2000). Relative potency values derived from  hepatic
       vitamin A reduction in male and female Sprague-Dawley rats following subchronic
       dietary exposure to individual polychlorinated dibenzo-p-dioxin and dibenzofuran
       congeners and a mixture therof. Toxicol Appl Pharmacol 165: 184-194.
       http://dx.doi.org/10.1006/taap.2000.8943.
Fox, TR; Best LL:  Goldsworthy, SM; Mills, JJ: Goldsworthy,  TL. (1993). Gene expression and
       cell proliferation in rat liver after 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure.  Cancer
       Res 53: 2265-2271.
Franc. MA: Pohjanvirta, R: Tuomisto, J: Okev, AB.  (2001).  Persistent, low-dose 2,3,7,8-
       tetrachlorodibenzo-p-dioxin exposure: effect on aryl  hydrocarbon receptor expression in
       a dioxin-resistance model. Toxicol Appl Pharmacol 175: 43-53.
       http://dx.doi.org/10.1006/taap.2001.9222.
Hassoun, EA: Li. F: Abushaban, A: Stohs, SJ. (2000). The relative abilities of TCDD and its
       congeners to induce oxidative stress in the hepatic and brain tissues of rats after
       subchronic exposure. Toxicology 145: 103-113.
                                        E-200

-------
Heinzl, H; Mittlbock, M; Edler, L. (2007). On the translation of uncertainty from toxicokinetic to
      toxicodynamic models—the TCDD example. Chemosphere 67: S365-S374.
      http://dx.doi.0rg/10.1016/j.chemosphere.2006.05.130.
Hoio. R: Stern. S: Zareba. G: Markowskl VP: Cox. C: Kost JT: Weiss. B. (2002). Sexually
      dimorphic behavioral responses to prenatal dioxin exposure. Environ Health Perspect
       110:247-254.
Hurt KJ: Shi. Z:  Albertini. DF: Petroff BK. (2008). The environmental toxicant 2,3,7,8-
      tetrachlorodibenzo-p-dioxin disrupts morphogenesis of the rat pre-implantation embryo.
      BMC Dev Biol 8: 1-12. http://dx.doi.org/10.1186/1471-213X-8-l.
Ikeda, M; Mitsui, T; Setani,  K; Tamura, M; Kakeyama, M; Sone, H; Tohyama, C: Tomita, T.
      (2005). In utero and lactational exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin in rats
      disrupts brain sexual differentiation. Toxicol Appl Pharmacol 205: 98- 105.
      http://dx.doi.0rg/10.1016/j.taap.2004.09.010.
Irigaray, P: Mejean, L; Laurent F. (2005). Behaviour of dioxin in pig adipocytes. Food Chem
      Toxicol 43: 457-460. http://dx.doi.Org/10.1016/i.fct.2004.ll.016.
Ishihara, K; Warita, K; Tanida, T: Sugawara, T: Kitagawa, H: Hoshi, N. (2007). Does paternal
      exposure  to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) affect the sex ratio of offspring.
      J Vet MedSci 69: 347-352.
Kattainen, H: Tuukkanen, J: Simanainen, U: Tuomisto, JT: Kovero, O: Lukinmaa, PL;
      Alaluusua. S: Tuomisto. J: Viluksela. M. (2001). In Utero/Lactational 2,3,7,8-
      Tetrachlorodibenzo-p-dioxin Exposure Impairs Molar  Tooth Development in Rats.
      Toxicol Appl Pharmacol 174: 216-224. http://dx.doi.org/10.1006/taap.2001.9216.
Keller, JM; Huet-Hudson, YM; Leamy, LJ. (2007). Qualitative effects of dioxin on molars vary
      among inbred mouse strains. Arch Oral Biol 52: 450-454.
      http://dx.doi.0rg/10.1016/j.archoralbio.2006.10.017.
Kergen BD: Leung. HW: Scott. P: Paustenbach. DJ: Needham. LL: Patterson. DG. Jr: Gerthoux.
      PM; Mocarelli, P. (2006). Age- and concentration-dependent elimination half-life of
      2,3,7,8-tetrachlorodibenzo-p-dioxin in Seveso children. Environ Health Perspect 114:
       1596-1602. http://dx.doi.org/10.1289/ehp.8884.
Kerger, BD: Leung, HW: Scott PK; Paustenbach, DJ. (2007). Refinements on the age-dependent
      half-life model for estimating child body burdens of polychlorodibenzodioxins and
      dibenzofurans. Chemosphere 67: S272-S278.
      http://dx.doi.0rg/10.1016/j.chemosphere.2006.05.108.
Kim, AH; Kohn,  MC: Nyska, A; Walker, NJ. (2003). Area under the curve as a dose metric for
      promotional responses following 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure. Toxicol
      Appl Pharmacol 191: 12-21. http://dx.doi.org/10.1016/S0041-008X(03)00225-4.
Kitchin. KT: Woods.  JS. (1979). 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) effects on hepatic
      microsomal cytochrome P-448-mediated enzyme activities. Toxicol Appl Pharmacol 47:
      537-546.  http://dx.doi.org/10.1016/0041-008X(79)90524-6.
Kociba. RJ: Keeler. PA: Park. CN: Gehring. PJ. (1976). 2,3,7,8-Tetrachlorodibenzo-p-dioxin
      (TCDD):  Results of a 13-week oral toxicity study in rats. Toxicol Appl Pharmacol 35:
      553-574.  http://dx.doi.org/10.1016/0041-008X(76)90078-8.
Kociba. RJ: Keves. DG: Beyer. JE: Carreon. RM: Wade. CE: Dittenber. DA: Kalnins. RP:
      Frauson. LE: Park. CN: Barnard. SD: Hummel RA: Humiston. CG. (1978). Results of a
      two-year  chronic toxicity and oncogenicity study of 2,3,7,8-tetrachlorodibenzo-p-dioxin
                                         E-201

-------
       in rats. Toxicol Appl Pharmacol 46: 279-303. http://dx.doi.org/10.1016/0041-
       008X(78)90075-3.
Korenaga, T; Fukusato, T; Ohta, M; Asaoka, K; Murata, N; Arima, A; Kubota, S. (2007). Long-
       term effects of subcutaneously injected 2,3,7,8-tetrachlorodibenzo-p-dioxin on the liver
       of rhesus monkeys. Chemosphere 67: S399-S404.
       http://dx.doi.0rg/10.1016/i.chemosphere.2006.05.135.
Korkalainen, M; Tuomisto, J: Pohjanvirta,  R. (2004). Primary structure and inducibility by
       2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) of aryl hydrocarbon receptor represser in a
       TCDD-sensitive and a TCDD-resistant rat strain. Biochem Biophys Res Commun 315:
       123-131. http://dx.doi.0rg/10.1016/i.bbrc.2004.01.028.
Kransler, KM; McGarrigle, BP; Olson, JR. (2007). Comparative developmental toxicity of
       2,3,7,8-tetrachlorodibenzo-p-dioxin in the hamster, rat and guinea pig. Toxicology 229:
       214-225. http://dx.doi.0rg/10.1016/i.tox.2006.10.019.
Kuchiiwa, S: Cheng, SB; Nagatomo, I; Akasaki, Y; Uchida, M: Tominaga, M: Hashiguchi, W;
       Kuchiiwa, T. (2002). In utero and lactational exposure to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin decreases serotonin-immunoreactive neurons in raphe nuclei of male mouse
       offspring. Neurosci Lett 317: 73-76. http://dx.doi.org/10.1016/S0304-3940(01)02434-X.
Latchoumycandane, C; Mathur, PP. (2002). Effects of vitamin E on reactive oxygen species-
       mediated 2,3,7,8-tetrachlorodibenzo-p-dioxin toxicity in rat testis. J Appl Toxicol 22:
       345-351. http://dx.doi.org/10.1002/iat.866.
Li, B; Liu, HY;  Dai, LJ: Lu, JC: Yang, ZM; Huang, L. (2006).  The early embryo loss caused by
       2,3,7,8-tetrachlorodibenzo-p-dioxin may be related to the accumulation of this compound
       in the uterus. Reprod Toxicol 21: 301-306.
       http://dx.doi.0rg/10.1016/i.reprotox.2005.09.008.
Li. X: Johnson.  DC: Rozman, KK. (1997).  2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
       increases release of luteinizing hormone and follicle-stimulating hormone from the
       pituitary of immature female rats in vivo and in vitro. Toxicol Appl Pharmacol 142: 264-
       269. http://dx.doi.org/10.1006/taap.1996.8044.
Markowski, VP: Zareba, G:  Stern, S: Cox,  C: Weiss, B. (2001). Altered operant responding for
       motor reinforcement and the determination of benchmark doses following perinatal
       exposure to low-level 2,3,7,8-tetrachlorodibenzo-p-dioxin. Environ Health Perspect 109:
       621-627.
Maronpot RR:  Foley, JF; Takahashi, K; Goldsworthy, T: Clark, G: Tritscher, A; Portier, C:
       Lucier, G.  (1993). Dose response for  TCDD promotion of hepatocarcinogenesis in rats
       initiated with DEN: histologic, biochemical, and cell proliferation endpoints. Environ
       Health Perspect 101: 643-642.
Maruyama, W:  Yoshida, K:  Tanaka, T: Nakanishi, J. (2002). Determination  of tissue-blood
       partition coefficients for a physiological model for humans, and estimation of dioxin
       concentration in tissues. Chemosphere 46: 975-985.
Maruyama, W:  Yoshida, K:  Tanaka, T: Nakanishi, J. (2003). Simulation of dioxin accumulation
       in human tissues and analysis of reproductive risk. Chemosphere 53:  301-313.
       http://dx.doi.org/10.1016/S0045-6535(03)00015-8.
Maruyama, W:  Aoki, Y.  (2006). Estimated cancer risk of dioxins to humans using a bioassay and
       physiologically based pharmacokinetic model. Toxicol Appl Pharmacol 214: 188-198.
       http://dx.doi.0rg/10.1016/i.taap.2005.12.005.
                                         E-202

-------
Miettinen, HM; Sorvari, R; Alaluusua, S; Murtotnaa, M; Tuukkanen, J: Viluksela, M. (2006).
       The Effect of Perinatal TCDD exposure on caries susceptibility in rats. Toxicol Sci 91:
       568-575. http://dx.doi.org/10.1093/toxsci/kfil58.
Milbrath. MO: Wenger. Y: Chang. CW: Emond. C: Garabrant D: Gillespie. BW: Jolliet O.
       (2009). Apparent half-lives of dioxins, furans, and polychlorinated biphenyls as a
       function of age, body fat, smoking status, and breast-feeding. Environ Health Perspect
       117: 417-425. http://dx.doi.org/10.1289/ehp.11781.
Moser, GA; McLachlan, MS. (2002). Modeling digestive tract absorption and desorption of
       lipophilic organic contaminants in humans. Environ Sci Technol 36: 3318-3325.
Mullerova, D: Kopecky, J. (2007). White adipose tissue: Storage and effector site for
       environmental pollutants. Physiol Res 56: 375-381.
Murray. FJ: Smith. FA: Nitschke. KD: Humiston. CG: Kociba. RJ: Schwetz. BA. (1979). Three-
       generation reproduction  study of rats given 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
       in the diet. Toxicol Appl Pharmacol 50:  241-252. http://dx.doi.org/10.1016/0041-
       008X(79)90149-2.
Nadal M: Perello, G: Schuhmacher, M: Cid, J: Domingo, JL. (2008). Concentrations of
       PCDD/PCDFs in plasma of subjects living in the vicinity of a hazardous waste
       incinerator: Follow-up and modeling validation. Chemosphere 73: 901-906.
       http://dx.doi.0rg/10.1016/j.chemosphere.2008.07.021.
Nadal M: Mari, M: Schuhmacher, M: Domingo, JL. (2009). Multi-compartmental
       environmental surveillance of a petrochemical area: Levels of micropollutants. Environ
       Int 35: 227-235.  http://dx.doi.Org/10.1016/i.envint.2008.06.001.
NAS (National Academy of Sciences). (2006). Health risks from dioxin and related  compounds:
       Evaluation of the EPA reassessment. Washington, DC: National Academy Press.
       http://www.nap.edu/catalog.php?record_id=l 1688.
Nohara, K; Fujimaki, H; Tsukumo, S: Ushio, H; Miyabara, Y; Kijima, M: Tohyama, C:
       Yonemoto, J. (2000). The effects of perinatal exposure to low doses of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin on immune organs in rats. Toxicology 154: 123-133.
       http://dx.doi.org/10.1016/S0300-483X(00)00323-1.
Nohara. K: Izumi, H: Tamura, S: Nagata, R: Tohyama. C. (2002). Effect of low-dose 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) on influenza A virus-induced mortality in mice.
       Toxicology 170: 131-138.
Nohara. K: Ao. K: Miyamoto. Y: Ito. T: Suzuki. T: Tovoshiba. H: Tohyama. C. (2006).
       Comparison of the 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)-induced CYP1A1 gene
       expression profile in lymphocytes from mice, rats, and humans: Mst potent induction in
       humans. Toxicology 225: 204-213. http://dx.doi.Org/10.1016/i.tox.2006.06.005.
NTP (National Toxicology Program). (1982). Carcinogenesis bioassay of 2,3,7,8-
       tetrachlorodibenzo-p-dioxin in Osberne-Mendel rats and B6C3F1  mice  (gavage study).
       (NTP TR 209). Research Triangle Park, NC.
NTP (National Toxicology Program). (2006). NTP technical report on the toxicology and
       carcinogenesis studies of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (CAS no. 1746-
       01-6) in female harlan Sprague-Dawley rats (gavage studies). (NTP TR 521; NIH
       Publication No. 06-4468). Research Triangle Park, NC.
Ohsako, S: Miyabara, Y: Nishimura, N; Kurosawa, S: Sakaue, M: Ishimura, R:  Sato, M: Takeda,
       K: Aoki, Y: Sone, H: Tohyama, C: Yonemoto, J. (2001). Maternal exposure  to a low
       dose of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) suppressed the development of

                                         E-203

-------
       reproductive organs of male rats: Dose-dependent increase of mRNA levels of 5a-
       reductase type 2 in contrast to descrease of androgen receptor in the pubertal ventral
       prostate. Toxicol Sci 60:  132-143.
Olsman. H: Engwall. M: Kammann. U: Klempt M: Otte. J: Bavel B: H. H.  (2007). Relative
       differences in aryl hydrocarbon receptor-mediated response for 18 polybrominated and
       mixed halogenated dibenzo-p-dioxins and -furans in cell lines from four different species.
       Environ Toxicol Chem 26: 2448-2454.
Saghir SA: Lebofsky. M: Pinson. DM: Rozmana. KK. (2005). Validation of Haber's Rule
       (dosextime = constant) in rats and mice for monochloroacetic acid and 2,3,7,8-
       tetrachlorodibenzo-p-dioxin under conditions of kinetic steady state.  Toxicology 215: 48-
       56. http://dx.doi.0rg/10.1016/i.tox.2005.06.009.
Santostefano. MJ: Wang. X: Richardson. VM: Ross. DG: DeVito. MJ: Birnbaum. LF. (1998). A
       pharmacodynamic analysis of TCDD-Induced Cytochrome 450 gene expression in
       multiple tissues: Dose and time-dependent effects. Toxicol Appl Pharmacol 151: 294-
       310.
Schantz. SL: Seo. BW: Moshtaghian. J: Peterson. RE: Moore. RW. (1996). Effects of gestational
       and lactational exposure to TCDD or coplanar PCBs on spatial learning. Neurotoxicol
       Teratol 18:305-313.
Schecter, A; Pavuk, M: Papke, O: Ryan, JJ. (2003). Dioxin, dibenzofuran, and coplanar PCB
       levels in Laotian blood and milk from agent orange-sprayed and nonsprayed areas, 2001.
       J Toxicol Environ Health A 66: 2067-2075.
Seo. BW: Li. MH: Hansen. LG: Moore. RW: Peterson. RE: Schantz. SL. (1995). Effects of
       gestational and lactational exposure to coplanar polychlorinated biphenyl (PCB)
       congeners or 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on thyroid hormone
       concentrations in weanling rats. Toxicol Lett 78: 253-262.
Sewall CH: Flagler. N: Vanden  Heuvel JP:  Clark. GC: Tritscher. AM: Maronpot. RM: Lucier.
       GW. (1995). Alterations  in thyroid function in female Sprague-Dawley rats following
       chronic treatment with 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Appl Pharmacol
       132: 237-244.
Shi. Z: Valdez. KE: Ting. AY: Franczak. A:  Gum. SL: Petroff, BK. (2007). Ovarian endocrine
       disruption underlies premature reproductive senescence following environmentally
       relevant chronic exposure to the aryl hydrocarbon receptor agonist 2,3,7,8-
       tetrachlorodibenzo-p-dioxin. Biol Reprod 76: 198-202.
       http://dx.doi.org/10.1095/biolreprod.106.053991.
Simanainen, U: Tuomisto, JT: Tuomisto, J: Viluksela, M. (2002). Structure-Activity
       relationships and dose responses of Polychlorinated Dibenzo-p-dioxins for short-term
       effects in 2,3,7,8- Tetrachlorodibenzo-p-dioxin-Resistant and sensitive rat strains.
       Toxicol Appl Pharmacol  181: 38-47.  http://dx.doi.org/10.1006/taap.2002.9386.
Simanainen, U: Tuomisto, JT: Tuomisto, J: Viluksela, M. (2003). Dose-response analysis of
       short-term effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin in three differentially
       susceptible rat lines. Toxicol Appl Pharmacol 187: 128-136.
       http://dx.doi.org/10.1016/S0041-008X(02)00068-6.
Simanainen, U: Haavisto, T; Tuomisto, JT: Paranko, J: Toppari, J: Tuomisto, J: Peterson, RE;
       Viluksela, M. (2004). Pattern of male reproductive system effects after in utero and
       lactational 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) exposure in three differentially
       TCDD-sensitive rat lines Pattern of male reproductive system effects after in utero and

                                         E-204

-------
       lactational 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) exposure in three differentially
       TCDD-sensitive rat lines. Toxicol Sci 80: 101-108.
       http://dx.doi.org/10.1093/toxsci/kfhl42.
Smialowicz. RJ: Burgin. DE: Williams. WC: Diliberto. JJ: Setzer. RW: Birnbaum. LS. (2004).
       CYP1A2 is not required for 2,3,7,8-tetrachlorodibenzo-p-dioxin-induced
       immunosuppression. Toxicology 197: 15-22. http://dx.doi.Org/10.1016/j.tox.2003.ll.016.
Smialowicz, RJ: DeVito, MJ: Williams, WC: Birnbaum, LS. (2008). Relative potency based on
       hepatic enzyme induction predicts immunosuppressive effects of a mixture of
       PCDDS/PCDFS andPCBS. Toxicol Appl Pharmacol 227: 477-484.
       http://dx.doi.0rg/10.1016/i.taap.2007.ll.018.
Smith, FA; Schwetz, BA; Nitschke, KD. (1976). Teratogenicity of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin in CF-1 mice. Toxicol Appl Pharmacol 38: 517-523.
       http://dx.doi.org/10.1016/0041-008X(76)90183-6.
Sparschu. G. . L.: Dunn. F. . L.: Rowe. V. . K. (1971). Study of the teratogenicity of 2,3,7,8-
       tetrachiorodibenzo-p-dioxin in the rat. Food Cosmet Toxicol 9: 405-412.
       http://dx.doi.org/10.1016/0015-6264(71)90045-9.
Staskal DF: Diliberto. JJ: DeVito. MJ: Birnbaum. LS. (2005). Inhibition of human and rat
       CYP1A2 by TCDD and dioxin-like chemicals. Toxicol Sci 84: 225-231.
       http://dx.doi.org/10.1093/toxsci/kfi090.
Toth, K: Somfai-Relle, S: Sugar, J: Bence, J. (1979). Carcinogenicity testing of herbicide 2,4,5-
       trichlorophenoxyethanol containing dioxin and of pure dioxin in Swiss mice. Nature 278:
       548-549.
Toyoshiba, H; Walker, NJ: Bailer , A; Portier, CJ. (2004). Evaluation of toxic equivalency
       factors for induction of cytochromes P450 CYP1A1 and CYP1A2 enzyme activity by
       dioxin-like compounds. Toxicol Appl Pharmacol 194: 156-168.
       http://dx.doi.0rg/10.1016/i.taap.2003.09.015.
U.S. EPA (U.S. Environmental Protection Agency). (2003). Exposure and human health
       reassessment of 2,3,7,8 tetrachlorodibenzo-p dioxin (TCDD) and related compounds
       [NAS review draft]. (EPA/600/P-00/001). Washington, DC: U.S. Environmental
       Protection Agency, National Center for Environmental Assessment.
       http://www.epa.gov/nceawwwl/pdfs/dioxin/nas-review/.
Van Birgelen, AP; Van der Kolk, J: Fase, KM; Bol, I; Poiger, H; Brouwer, A; Van den Berg, M.
       (1995). Subchronic dose-response study of 2,3,7,8-tetrachlorodibenzo-p-dioxin in female
       Sprague-Dawley rats. Toxicol Appl Pharmacol 132: 1-13.
       http://dx.doi.org/10.1006/taap.1995.1080.
Vanden Heuvel JP: Clark. GC: Tritscher. A: Lucier. GW. (1994). Accumulation of
       polychlorinated dibenzo-p-dioxins and dibenzofurans in liver of control laboratory rats.
       Fundam Appl Toxicol 23:  465-469. http://dx.doi.Org/10.1093/toxsci/23.3.465.
Wang. X: Santostefano. MJ: Evans. MV: Richardson. VM: Diliberto. JJ: Birnbaum. LS. (1997).
       Determination of parameters responsible for pharmacokinetic behavior of TCDD in
       female Sprague-Dawley rats. Toxicol Appl Pharmacol 147:  151-168.
       http://dx.doi.org/10.1006/taap.1997.8242.
Wang. X: Santostefano. MJ: DeVito. MJ: Birnbaum. LS. (2000). Extrapolation of a PBPK model
       for dioxins across dosage regimen, gender, strain, and species. Toxicol Sci 56: 49-60.
Weber. LW: Lebofsky. M: Stahl. BU: Smith. S: Rozman. KK. (1995). Correlation between
       toxicity and effects on intermediary metabolism in 2,3,7,8-tetrachlorodibenzo-p-dioxin-

                                        E-205

-------
       treated male C57BL/6J and DBA/2J mice. Toxicol Appl Pharmacol 131: 155-162.
       http://dx.doi.org/10.1006/taap.1995.1057.
White, KL, Jr; Lysy, HH; McCay, JA; Anderson, AC. (1986). Modulation of serum complement
       levels following exposure to polychlorinated dibenzo-p-dioxins. Toxicol Appl Pharmacol
       84: 209-219. http://dx.doi.org/10.1016/0041-008X(86)90128-6.
Wilkes. JG: Hass. BS: Buzatu. DA: Pence. LM: Archer. JC: Beger. RD: Schnackenberg. LK:
       Halbert, MK; Jennings, L; Kodell, RL. (2008). Modeling and assaying dioxin-like
       biological effects for both dioxin-like and certain non-dioxin-like compounds. Toxicol
       Sci 102:  187-195. http://dx.doi.org/10.1093/toxsci/kfm294.
                                        E-206

-------
                                       EPA/600/R-10/038F
                                         www.epa.gov/iris
            APPENDIX F
Epidemiologic Kinetic Modeling
                January 2012
       National Center for Environmental Assessment
          Office of Research and Development
          U.S. Environmental Protection Agency
                 Cincinnati, OH

-------
             CONTENTS—APPENDIX F: Epidemiologic Kinetic Modeling
LIST OF TABLES	F-v

APPENDIX F. EPIDEMIOLOGIC KINETIC MODELING	F-l
     F.I. Derivation of Background Concentration	F-l
         F.I.I. Needham Background Scenario	F-l
                F.I. 1.1. Summary of Modeling Approach	F-l
                F.I.1.2. Input for Continuous Exposure to Measurement	F-2
                F.I.1.3. Needham Background Scenario Results	F-3
         F.I.2. Eskenazi Background Scenario	F-3
                F.I.2.1. Summary of Modeling Approach	F-3
                F.I.2.2. Input for Continuous Exposure to Measurement	F-4
                F.I.2.3. Eskenazi et Background Scenario Results	F-5
     F.2. KINETIC MODELING OF EPIDEMIOLOGIC STUDIES CONSIDERED
         FORRfD	F-6
         F.2.1.Baccarellietal. (2008)	F-6
                F.2.1.1. Input for Exposure During Pregnancy	F-6
                F.2.1.2. Baccarelli etal. (2008) Results	F-6
         F.2.2. Mocarelli et al. (2008)	F-6
                F.2.2.1. Input for Exposure from Event to LASC Measurement	F-6
                F.2.2.2. Input for Exposure from Event to End of Critical Window	F-7
                F.2.2.3. Input for Continuous Exposure over Critical Window	F-7
                F.2.2.4. Mocarelli (2008) Results	F-8
         F.2.3. Alaluusua et al. (2004)	F-9
                F.2.3.1. Input for Exposure from Event to LASC Measurement	F-9
                F.2.3.2. Input for Exposure from Event to the End of the Assumed
                        Critical Exposure Window	F-9
                F.2.3.3. Input for Continuous Exposure over Assumed Critical
                        Exposure Window	F-10
                F.2.3.4. Alaluusua etal. (2004) Results	F-ll
         F.2.4. Eskanazi etal. (2002)	F-ll
                F.2.4.1. Input for Exposure from Event to LASC Measurement	F-ll
                F.2.4.2. Input for Exposure from Event to the End of the Assumed
                        Critical Exposure Window	F-12
                F.2.4.3. Input for Continuous Exposure over Assumed Critical
                        Exposure Window	F-12
                F.2.4.4. Eskenazi et al. (2002) Results	F-13
     F.3. KINETIC MODELING OF EPIDEMIOLOGIC STUDIES FOR
         SENSITIVITY ANALYSIS	F-14
         F.3.1. Alaluusua etal. (2004)	F-14
                F.3.1.1. Summary of Modeling Approach	F-14
                F.3.1.2. Input for Exposure from Event to LASC Measurement	F-14
                F.3.1.3. Input for Exposure from Event to the End of the Assumed
                        Critical Exposure Window	F-l5


                                       F-ii

-------
                     CONTENTS (continued)


       F.3.1.4. Input for Continuous Exposure over Assumed Critical
               Exposure Window	F-16
       F.3.1.5. Alaluusua et al. (2004) Results	F-17
F.3.2. Baccarelli et al. (2008)	F-17
       F.3.2.1. Summary of Modeling Approach	F-17
       F.3.2.2. Baccarelli et al. (2008) Results	F-17
F.3.3.Eskenazietal. (2002)	F-18
       F.3.3.1. Summary of Modeling Approach	F-18
       F.3.3.2. Input for Exposure from Event to LASC Measurement	F-18
       F.3.3.3. Input for Exposure from Event to the End of the Assumed
               Critical Exposure Window	F-19
       F.3.3.4. Input for Continuous Exposure over Assumed Critical
               Exposure Window	F-19
       F.3.3.5. Eskenazi et al. (2002) Results	F-20
F.3.4. Eskenazi et al. (2005)	F-20
       F.3.4.1. Summary of Modeling Approach	F-20
       F.3.4.2. Input for Exposure from Event to LASC Measurement	F-22
       F.3.4.3. Input for Exposure from Event to the End of the Assumed
               Critical Exposure Window	F-22
       F.3.4.4. Input for Continuous Exposure over Assumed Critical
               Exposure Window	F-23
       F.3.4.5. Eskenazi et al. (2005) Results	F-24
F.3.5.Mocarellietal. (2000)	F-24
       F.3.5.1. Summary of Modeling Approach	F-24
       F.3.5.2. Input for Exposure from Event to LASC Measurement	F-25
       F.3.5.3. Input for Exposure from Event to the End of the Assumed
               Critical Exposure Window	F-26
       F.3.5.4. Mocarelli et al. (2000) Results	F-27
F.3.6. Mocarelli et al. (2008)	F-27
       F.3.6.1. Summary of Modeling Approach	F-27
       F.3.6.2. Input for Exposure from Event to LASC Measurement	F-28
       F.3.6.3. Input for Exposure from Event to the End of the Assumed
               Critical Exposure Window	F-29
       F.3.6.4. Input for Continuous Exposure over Assumed Critical
               Exposure Window	F-29
       F.3.6.5. Mocarelli et al. (2008) Results	F-30
F.3.7. Mocarelli etal. (2011)	F-30
       F.3.7.1. Summary of Modeling Approach	F-30
       F.3.7.2. Input for Exposure from Event to LASC Measurement	F-31
       F.3.7.3. Input for Exposure from Event to the Study-Average Age at
               Conception	F-32
       F.3.7.4. Input for Continuous Exposure until Age at Conception for
               General Population	F-33
       F.3.7.5. Mocarelli etal. (2011) Results	F-33

                               F-iii

-------
                          CONTENTS (continued)
     F.3.8. Warner etal. (2004)	F-34
           F.3.8.1. Summary of Modeling Approach	F-34
           F.3.8.2. Input for Exposure from Event to LASC Measurement	F-34
           F.3.8.3. Input for Exposure from Event to the End of the Assumed
                   Critical Exposure Window	F-35
           F.3.8.4. Input for Continuous Exposure over Assumed Critical
                   Exposure Window	F-36
           F.3.8.5. Warner etal. (2004) Results	F-37
     F.3.9. Warner etal. (2007)	F-37
           F.3.9.1. Summary of Modeling Approach	F-37
           F.3.9.2. Input for Exposure from Event to LASC Measurement	F-38
           F.3.9.3. Input for Exposure from Event to the End of the Assumed
                   Critical Exposure Window	F-39
           F.3.9.4. Input for Continuous Exposure over Assumed Critical
                   Exposure Window	F-39
           F.3.9.5. Warner etal. (2007) Results	F-40
F.4. REFERENCES	F-41
                                   F-iv

-------
                                   LIST OF TABLES
F-l.           Estimated background intakes for Needham scenario	F-3
F-2.           Estimated background intakes for Eskenazi background scenario	F-5
F-3.           Estimated continuous TCDD intake corresponding to maternal serum
              concentration	F-6
F-4.           Matching peak and average after pulse to 10-year childhood intake for
              Mocarelli etal. (2008)	F-8
F-5.           Matching peak and average after pulse to chronic intake for Alaluusua et
              al. (2004)	F-ll
F-6.           Matching peak and average after pulse to chronic intake for Eskenazi et al.
              (2002)	F-13
F-7.           Model inputs derived from study details for Alaluusua et al. (2004)	F-14
F-8.           Matching peak and average after pulse to chronic intake for Alaluusua et
              al. (2004) using alternate background value	F-17
F-9.           Estimated continuous intake corresponding to maternal serum
              concentration for TEQ	F-17
F-10.         Model inputs derived from study details for Eskenazi et al. (2002)	F-18
F-ll.         Matching peak and average after pulse to chronic intake for Eskenazi et al.
              (2002) using alternate background value	F-20
F-12.         Model inputs derived from study details for Eskenazi et al. (2005)	F-21
F-13.         Matching peak and average after pulse to chronic intake for Eskenazi et al.
              (2005)	F-24
F-14.         Model inputs derived from study details for Mocarelli et al. (2000)	F-25
F-15.         Matching peak and average after pulse to 5-year average response surface
              for Mocarelli etal. (2000)	F-27
F-16.         Model inputs derived from study details for Mocarelli et al. (2008)	F-28
F-17.         Matching peak and average after pulse to critical-window intake for
              Mocarelli et al. (2008) using alternate background value	F-30
F-18.         Model inputs derived from study details for Mocarelli et al. (2011)	F-31
F-19.         Matching concentration at conception for the study population to chronic
              intake for the general population for Mocarelli et al. (2011)	F-33
F-20.         Model inputs derived from study details for Warner et al. (2004)	F-34
F-21.         Matching peak and average after pulse to chronic intake for Warner et al.
              (2004)	F-37
F-22.         Model inputs derived from study details for Warner et al. (2007)	F-38
F-23.         Matching peak and average after pulse to chronic intake for Warner et al.
              (2007)	F-40
                                           F-v

-------
           APPENDIX F. EPIDEMIOLOGIC KINETIC MODELING

F.I.  DERIVATION OF BACKGROUND CONCENTRATION
       Background intakes for the Seveso cohort were estimated from information from two
separate studies. The details of the modeling and the estimated background intakes are described
in this section.

F.I.I. Needham Background Scenario
F. 1.1.1. Summary of Modeling Approach
       Needham et al. (1998) reported lipid adjusted serum concentrations in 11 pools of
individuals in the non-ABR region near the site of the Seveso TCDD accident in July, 1976. The
individuals in this region did not suffer exposure from the event and represent a reference
(comparison) population in the study.  There were 4-10 individuals per pool, and the median
lipid-adjusted serum concentration (LASC) across the pools was reported by the study authors to
be 15 ppt.
       All subjects in the pooled samples were above age 25, but no further details about age are
given in the study.  Mocarelli et al. (1991) reported details about 10 subjects in the non-ABR
region at the time of serum sample collection in 1976. The oldest individual in this sample
was 46. In the absence of other information, this age was used as an upper bound, suggesting a
median age (between 25 and 46) of approximately 35 years old.
       The Emond model is not coded to allow the background intake to vary in time. Thus,  it
was assumed that the background intake remained constant over the lifetime of the individual.
The Emond model was used to determine the continuous daily TCDD intake which gives a
terminal concentration of 15 ppt at the age of 35 for both women and men.  The background
intakes were then rounded to the nearest 10~5 ng/kg-day. The corresponding male and female
oral intakes were 3.5 x  10~4 ng/kg-day and 3.9 x 10~4 ng/kg-day, respectively.
       For the modeled-TEQ method in the sensitivity analysis, TEQ background intake was
estimated by assuming that TCDD LASC is 10% of total TEQ LASC and that all DLCs are
kinetically-equivalent to TCDD. The TEQ intakes were then modeled as the continuous daily
TCDD-equivalent intake which giving a terminal concentration of 150 ppt at the age of 35 for
both women and men. The total TEQ intake matching 150 ppt (10  x TCDD) at 35 years was
8.91  x 1Q~3 ng/kg-day for males and 9.44 x 10~3 ng/kg-day for females.
                                         F-l

-------
      For the additive DLC intake method, where DLC-TEQ intakes (ng/kg-day) are added to

modeled TCDD intakes, a simple intake-scaling approach was used. The assumed ratio of TEQ

LASC to TCDD LASC was applied to the TCDD intake estimate. For the Needham scenario, a

total-TEQ LASC of 80.6 ppt is 1.88 times the TCDD LASC of 40.5 ppt. With the assumption

that TCDD comprises 10% of the total background TEQ, the ratio of DLC-TEQ:TCDD is 9:1 for

background exposures. Scaling the male TCDD background intake of 3.5 x 10~3 ng/kg-day by

this factor gives a DLC-TEQ intake of 3.15 x 10 3 ng/kg-day. The corresponding female

DLC-TEQ intake is 3.51 x 1(T3 ng/kg-day (3.9 x 10~4 x 1.88).



F.I. 1.2. Input for Continuous Exposure to Measurement

% MODEL PARAMETERS
output @clear
prepare gclear  T CBSNGKGLIADJ CBNGKG

% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON  =0.       %  CONTINUOUS EXPOSURE BEGINS AT  BIRTH (AGE 0 HOURS)
EXP_TIME_OFF =  306600.  % AGE AT MEASUREMENT  (HOURS)
DAY_CYCLE    =  24.      %  LENGTH OF DAY  (HOURS/DAY)
BCK_TIME_ON  =  0.  %  BACKGROUND EXPOSURE BEGINS AT BIRTH  (AGE 0  HOURS)
BCK_TIME_OFF =  306600.  % AGE AT END OF BACKGROUND EXPOSURE  (HOURS)
TIMELIMIT    =  306600.  % AGE AT MEASUREMENT  (HOURS)
MSTOTBCKGR   =0.       %  /KG-DAYBACKGROUND EXPOSURE INCLUDED IN MSTOT

% CONTINUOUS EXPOSURE DOSE (NG/KG-DAY)
MSTOT = 3.5E-4   %  TCDD-ONLY,  MALES (15 ppt at  35  years)
      % 3.9E-4   %  TCDD-ONLY,  FEMALES  (15 ppt at 35 years)
      % 8.91E-3  %  TOTAL TEQ,  MALES (150 ppt at  35 years)
      % 9.44E-3  %  TOTAL TEQ,  FEMALES  (150 ppt  at  35 years)

% HUMAN VARIABLE PARAMETERS
MALE   =  1.
FEMALE =  0.
YO     =  0. %  0  YEARS OLD  AT BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
CBSNGKGLIADJ
                                       F-2

-------
F. 1.1.3. Needham Background Scenario Results
       Table F-l. Estimated background intakes for Needham scenario
Age at
measurement
(years)
35
Measured
TCDD
LASC (ppt)
15
Assumed
TEQ LASC
(ppt)
150a
Continuous
intake matching
TCDD LASC
(ng/kg-day)
3.5E-04
(males)
3.9E-04
(females)
Continuous
intake matching
TEQ LASCa
(ng/kg-day)
8.91E-03
(males)
9.44E-03
(females)
Additive
DLC-TEQ
intakeb
(ng/kg-day)
3.15E-03
(males)
3.51E-03
(females)
       Intakes rounded to the nearest 10~5 ng/kg-day
       "For use in modeled-TEQ method
       bFor use in additive DLC-intake method
F.1.2. Eskenazi Background Scenario
F. 1.2.1.  Summary of Modeling Approach
      Eskenazi et al. (2004) reported TCDD levels for the Seveso Women's Cohort from
pooled samples from individuals living in zone non-ABR (unexposed regions) in 1976,
representing background exposure levels to TCDD and total TEQ. Table 3 in that study reports
mean TCDD and TEQ for three different age groups.  As an alternative background intake for
endpoints measured in children compared with the Needham background, the 0-12 age group
(girls) was used to determine background exposure using the Emond model.  The two pooled
sample results were averaged to give an average background TCDD LASC of 40.5 ppt. It was
assumed that both males and females had this average concentration.  The Emond model was run
until the intake resulted in an average LASC of 40.5 when averaged between ages 0 and 12. The
corresponding male and female oral intakes were 4.22 x 10 3 ng/kg-day and 4.29 x 10 3
ng/kg-day, respectively. The background intake was then rounded to the nearest 10~5 ng/kg-day.
      For direct modeling of total TEQ LASC, background TEQ LASC was estimated from
Eskenazi et al. (2004).  The average total TEQ levels for the 0-12 year-old group, as reported by
Eskenazi et al. (2004), was 116.6 ppt, with 76.1 ppt attributed to DLCs.  The Emond model was
run until the TCDD-equivalent intake resulted in an average LASC of 116.6 when averaged
between ages 0 and 12. The estimated male and female TEQ intakes were 0.01803 and
                                         F-3

-------
0.01807 ng/kg-day, respectively. These estimates were used for direct modeling of TEQ

expressed as LASC to obtain corresponding TEQ intakes.

      For the additive DLC method, where DLC-TEQ intakes (ng/kg-day) are added to

modeled TCDD intakes, a simple intake-scaling approach was used. The assumed or measured

ratio of TEQ LASC to TCDD LASC was applied to the TCDD intake estimate.  For the Eskenazi

scenario, measured DLC-TEQ LASC of 76.1 ppt is 1.88 times the TCDD LASC of 40.5 ppt.

Scaling the male TCDD background intake of 4.22 x  10 3 ng/kg-day by this factor gives a DLC-

TEQ intake of 7.93 x 10~3 ng/kg-day. The corresponding female DLC-TEQ intake is 8.07 x 10~3

ng/kg-day (4.29 x  10~3 x 1.88).



F.I.2.2. Input for Continuous Exposure to Measurement

% MODEL PARAMETERS
output @clear
prepare gclear  T CBSNGKGLIADJ CBNGKG

% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON  =0.       % CONTINUOUS EXPOSURE BEGINS  AT BIRTH (AGE 0 HOURS)
EXP_TIME_OFF =  105120.  % UPPER AGE RANGE IN SAMPLE  (HOURS)
DAY_CYCLE    =  24.      % LENGTH OF DAY  (HOURS/DAY)
BCK_TIME_ON  =0.       % BACKGROUND EXPOSURE BEGINS  AT BIRTH (AGE 0 HOURS)
BCK_TIME_OFF =  105120.  % AGE AT END OF BACKGROUND EXPOSURE (HOURS)
TIMELIMIT    =  105120.  % UPPER AGE RANGE IN SAMPLE  (HOURS)
MSTOTBCKGR   =0.       % /KG-DAYBACKGROUND  EXPOSURE  INCLUDED IN MSTOT

% CONTINUOUS EXPOSURE DOSE (NG/KG-DAY)
MSTOT = 4.22E-3  %  TCDD-ONLY,  MALES  (10-year avg  =40.5 ppt)
      % 4.29E-3  %  TCDD-ONLY,  FEMALES (10-year  avg =40.5  ppt)
      % 1.32E-2  %  TOTAL TEQ,  MALES  (10-year avg  = 93.7 ppt)
      % 1.33E-2  %  TOTAL TEQ,  FEMALES (10-year  avg =93.7  ppt)

% HUMAN VARIABLE PARAMETERS
MALE   =  1.
FEMALE =  0.
YO     =  0.  % 0  YEARS OLD AT BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
mean(_cbsngkgliadj)
                                       F-4

-------
F. 1.2.3. Eskenazi et Background Scenario Results
       Table F-2. Estimated background intakes for Eskenazi background scenario
Age range at
measurement
(years)
0-12
Average
measured
TCDD LASC
(ppt)
40.5
Average
measured
TEQ LASC
(ppt)
116.6
Continuous
intake matching
measured
TCDD LASC
(ng/kg-day)
4.22E-03
(males)
4.29E-03
(females)
Continuous
intake matching
measured
TEQ LASCa
(ng/kg-day)
1.32E-02
(males)
1.33E-02
(females)
Additive
DLC-TEQ
intakeb
(ng/kg-day)
7.93E-03
(males)
8.07E-03
(females)
       Intakes rounded to the nearest 10" ng/kg-day
       "For use in modeled-TEQ method
       bFor use in additive DLC-intake method
                                           F-5

-------
F.2. KINETIC MODELING OF EPIDEMIOLOGIC STUDIES CONSIDERED FOR RfD
F.2.1. Baccarelli et al. (2008)
F.2.1.1. Input for Exposure During Pregnancy
% EXPOSURE  PARAMETERS
CINT = 1.
EXP_TIME_ON =  0.
EXP_TIME_OFF =  401190.
DAY_CYCLE    =  24.
BCK_TIME_ON =  401190.
BCK_TIME_OFF =  401190.
CONCEPTION_T =  262800.
TIMELIMIT    =  269184.
TRANSTIME_ON =  264312.
MSTOTBCKGR   =  0.
        % CONTINUOUS  EXPOSURE BEGINS AT BIRTH  (AGE  0  HOURS)
        % LENGTH OF CRITICAL WINDOW (HOURS)
        % LENGTH OF DAY (HOURS/DAY)
        % AGE AT BEGINNING OF BACKGROUND EXPOSURE  (HOURS)
        % AGE AT END  OF BACKGROUND EXPOSURE  (HOURS)
        % AGE AT CONCEPTION (HOURS)
        % AGE AT END  OF PREGNANCY  (HOURS)
        % AGE AT MOTHER-FETUS EXCHANGE  (HOURS)
        % /KG-DAYBACKGROUND EXPOSURE INCLUDED  IN MSTOT
% CONTINUOUS  EXPOSURE DOSE (NG/KG-DAY)
MSTOT = 0.021  %  MATCHING MATERNAL LASC OF 235 NG/KG
F.2.1.2. Baccarelli et al. (2008) Results
      Table F-3. Estimated continuous TCDD intake corresponding to maternal
      serum concentration
Variable
Infant b-TSH
Maternal lipid adjusted serum
Intake
Value
5 uU/mL
235 ng/kg
0.020 ng/kg-day
Notes
Adverse response level
From Figure 2A in Baccarelli et al. (2008)
From Emond model; pregnancy at 30 years
TSH = thyroid stimulating hormone.
F.2.2. Mocarelli et al. (2008)

F.2.2.1. Input for Exposure from Event to LASC Measurement
% MODEL PARAMETERS
output @clear
prepare @clear  T  CBSNGKGLIADJ CBNGKG

% EXPOSURE  PARAMETERS
MAXT = 0.5.
CINT = 1.
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
MSTOTBCKGR
54312.  % AGE AT  EXPOSURE (HOURS)
54335.  % AGE AT  END  OF EXPOSURE (HOURS)
24.     % LENGTH  OF DAY (HOURS/DAY)
0.      % BACKGROUND  EXPOSURE BEGINS AT BIRTH  (AGE  0  HOURS)
613200. % AGE AT  END  OF BACKGROUND EXPOSURE  (HOURS)
58692.  % AGE AT  LASC MEASUREMENT (HOURS)
0.00035 % NEEDHAM BACKGROUND EXPOSURE DOSE  (NG/KG-DAY)
                                       F-6

-------
% EVENT EXPOSURE DOSE  (NG/KG-DAY)
MSTOT = 8.2   % 1ST  QUARTILE
      % 22.5  % 2ND  QUARTILE
      % 78.4  % 3RD  QUARTILE
      % 231.9 % 4TH  QUARTILE

% HUMAN VARIABLE PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0. % AGE AT BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
CBSNGKGLIADJ_oneday=mean(_cbsngkgliadj(find(_t==58524):length(t))
F.2.2.2. Input for Exposure from Event to End of Critical Window

% MODEL PARAMETERS
output @clear
prepare @clear T CBSNGKGLIADJ CBNGKG

% EXPOSURE PARAMETERS
MAXT =0.5.
CINT = 1.
EXP_TIME_ON  = 54312.   %  AGE AT EXPOSURE (HOURS)
EXP_TIME_OFF = 54335.   %  AGE AT END OF EXPOSURE  (HOURS)
DAY_CYCLE    =24.      %  LENGTH OF DAY (HOURS/DAY)
BCK_TIME_ON  =0.       %  BACKGROUND EXPOSURE BEGINS AT BIRTH  (AGE  0  HOURS)
BCK_TIME_OFF = 613200.  %  AGE AT END OF BACKGROUND EXPOSURE  (HOURS)
TIMELIMIT    = 87600.   %  LENGTH OF CRITICAL WINDOW  (HOURS)
MSTOTBCKGR   = 0.00035  %  NEEDHAM BACKGROUND EXPOSURE DOSE  (NG/KG-DAY)

% EVENT EXPOSURE DOSE  (NG/KG-DAY)
MSTOT =8.2   % 1ST  QUARTILE
      % 22.5  % 2ND  QUARTILE
      % 78.4  % 3RD  QUARTILE
      % 231.9 % 4TH  QUARTILE

% HUMAN VARIABLE PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0. % AGE AT BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj(find(_t==EXP_TIME_ON):length(_t)));
meanCBSNGKGLIADJ
maxCBSNGKGLIADJ=max(_cbsngkgliadj);
maxCBSNGKGLIADJ
F.2.2.3. Input for Continuous Exposure over Critical Window
% MODEL PARAMETERS
output @clear
prepare @clear  T CBSNGKGLIADJ CBNGKG

                                      F-7

-------
% EXPOSURE  PARAMETERS
MAXT = 0.5.
CINT = 1.
EXP_TIME_ON  =  0.
EXP_TIME_OFF  =  87601.
DAY_CYCLE     =  24.
BCK_TIME_ON  =  0.
BCK_TIME_OFF  =  613200.
TIMELIMIT     =  87600.
MSTOTBCKGR    =  0.
                  %  CONTINUOUS EXPOSURE BEGINS AT  BIRTH (AGE 0 HOURS)
                  %  LENGTH OF CRITICAL WINDOW  (HOURS)
                  %  LENGTH OF DAY (HOURS/DAY)
                  %  BACKGROUND EXPOSURE BEGINS AT  BIRTH (AGE 0 HOURS)
                  %  AGE AT END OF BACKGROUND EXPOSURE  (HOURS)
                  %  LENGTH OF CRITICAL WINDOW  (HOURS)
                  %  /KG-DAYBACKGROUND EXPOSURE INCLUDED IN MSTOT
% CONTINUOUS  EXPOSURE DOSE (NG/KG-DAY)
MSTOT = 7.97E-3
      % 2.08E-2
            1ST QUARTILE - MATCHING MEAN
            2ND QUARTILE - MATCHING MEAN
% 7.21E-2 % 3RD QUARTILE - MATCHING MEAN
% 2.12E-1 % 4TH QUARTILE - MATCHING MEAN
% 3.21E-2 % 1ST QUARTILE - MATCHING MAX
% 1.41E-1 % 2ND QUARTILE - MATCHING MAX
% 8.73E-1 % 3RD QUARTILE - MATCHING MAX
% 3.89E+0 % 4TH QUARTILE - MATCHING MAX
% HUMAN VARIABLE  PARAMETERS
MALE   =  1.
FEMALE =  0.
YO     =  0.  %  0 YEARS OLD AT BEGINNING OF  SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj);
maxCBSNGKGLIADJ=max(  cbsngkgliadj);
F.2.2.4. Mocaretti (2008) Results
      Table F-4. Matching peak and average after pulse to 10-year childhood
      intake for Mocarelli et al. (2008)
TCDD only
Subject
modeled
Quartile
Measured
LASC
(ng/kg)
Event
dose
(ng/kg)
Average
LASC
after
pulse
dose
(ng/kg)
TCDD Needham scenario; background LASC = 15
Male
Male
Male
Male
1st
^nd
3rd
4th
68
142
345
733
8.2
22.5
78.4
231.9
57.7
116.8
276.7
579.4
Continuous
intake
matching
average
LASC
(ng/kg-day)
Peak
LASC
after
pulse
dose
(ng/kg)
Continuous
intake
matching
peak LASC
(ng/kg-day)
Average of
continuous
intake rates
(ng/kg-day)
3pt(3.5E-04ng/kg-day)a
7.97E-03
2.08E-02
7.21E-02
2.12E-01
249.0
668.7
2288.7
6658.9
3.21E-02
1.41E-01
8.73E-01
3.89E+00
2.01E-02
8.08E-02
4.73E-01
2.05E+00
"See Table F-l.
                                       F-8

-------
F.2.3. Alaluusua et al. (2004)

F.2.3.1. Input for Exposure from Event to LASC Measurement
% MODEL PARAMETERS
output @clear
prepare gclear T CBSNGKGLIADJ CBNGKG

% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
MSTOTBCKGR
  21900.
  21923.
  24.
  0.
  613200.
  26280.
% AGE AT EXPOSURE  (HOURS)
% AGE AT END OF EXPOSURE  (HOURS)
% LENGTH OF DAY  (HOURS/DAY)
% BACKGROUND EXPOSURE BEGINS AT  BIRTH (AGE 0 HOURS)
% AGE AT END OF BACKGROUND  EXPOSURE (HOURS)
% AGE AT LASC MEASUREMENT  (HOURS)
= 0.00035 % NEEDHAM BACKGROUND EXPOSURE DOSE (NG/KG-DAY)
  EVENT EXPOSURE  DOSE  (NG/KG-DAY)
MSTOT
= 10.9 1
% 10.4 1
% 105.9 1
% 102.3 1
% 3419.2 S
% 4266.1 1
i 1ST TERTILE
I 1ST TERTILE
•; 2ND TERTILE
i 2ND TERTILE
i 3RD TERTILE
i 3RD TERTILE
- MALE
- FEMALE
- MALE
- FEMALE
- MALE
- FEMALE
% HUMAN VARIABLE  PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0. % AGE AT  BEGINNING OF SIMULATION
% POST-PROCESSING
start @nocallback
CBSNGKGLIADJ_oneday=mean(_cbsngkgliadj(findi
                                 t==26112):length(t))
F.2.3.2. Input for Exposure from Event to the End of the Assumed Critical Exposure Window
% MODEL PARAMETERS
output @clear
prepare @clear  T  CBSNGKGLIADJ CBNGKG

% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
MSTOTBCKGR
  21900.  % AGE AT EXPOSURE  (HOURS)
  21923.  % AGE AT END  OF  EXPOSURE (HOURS)
  24.     % LENGTH OF DAY  (HOURS/DAY)
  0.      % BACKGROUND  EXPOSURE BEGINS AT BIRTH  (AGE 0 HOURS)
  613200. % AGE AT END  OF  BACKGROUND EXPOSURE  (HOURS)
  43800.  % LENGTH OF CRITICAL WINDOW (HOURS)
  0.00035 % NEEDHAM BACKGROUND EXPOSURE,  MALES  (NG/KG-DAY)
  0.00039 % NEEDHAM BACKGROUND EXPOSURE,  FEMALES  (NG/KG-DAY)
% EVENT EXPOSURE  DOSE  (NG/KG-DAY)
                                       F-9

-------
MSTOT =10.9
      % 10.4
      % 105.9
      % 102.3
      % 3419.2
% 1ST TERTILE
% 1ST TERTILE
% 2ND TERTILE
% 2ND TERTILE
% 3RD TERTILE
      % 4266.1 % 3RD TERTILE  -
MALE
FEMALE
MALE
FEMALE
MALE
FEMALE
% HUMAN VARIABLE PARAMETERS
MALE   = 1.    % 0 FOR  FEMALE  SIMULATION
FEMALE =0.    % 1 FOR  FEMALE  SIMULATION
YO     = 0. % AGE AT BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj(find(_t==EXP_TIME_ON):length(_t)));
meanCBSNGKGLIADJ
maxCBSNGKGLIADJ=max(_cbsngkgliadj);
maxCBSNGKGLIADJ
F.2.3.3. Input for Continuous Exposure over Assumed Critical Exposure Window
% MODEL PARAMETERS
output @clear
prepare @clear  T CBSNGKGLIADJ  CBNGKG
% EXPOSURE PARAMETERS
MAXT
CINT
EXP
EXP
DAY
BCK
BCK
= 0
= 1
TIME
TIME
.
.


5

ON =
OFF =
CYCLE
TIME
TIME


ON =
OFF =
TIMELIMIT
MSTOTBCKGR


0.












% CONTINUOUS EXPOSURE BEGINS AT BIRTH (AGE 0
43801.
24,
0.

% LENGTH
% LENGTH


HOURS)
OF CRITICAL WINDOW (HOURS)
OF DAY
(HOURS /DAY)
% BACKGROUND EXPOSURE BEGINS AT BIRTH (AGE 0
613200
43800.
0.
. % AGE AT
% LENGTH
END OF
BACKGROUND
EXPOSURE (HOURS)
HOURS)

OF CRITICAL WINDOW (HOURS)
% /KG-DAYBACKGROUND EXPOSURE INCLUDED IN MSTOT
% CONTINUOUS EXPOSURE
MSTOT =











%
%
%
%
%
%
%
%
%
%
%
1
1
1
1
4
4
6
6
1
1
1
1
.62E-2
.51E-2
.53E-1
.44E-1
.94E+0
.68E+0
.95E-2
.15E-2
.72E+0
.58E+0
.14E+2
.08E+2
%
%
%
%
%
%
%
%
%
%
%
%
1ST
1ST
2ND
2ND
3RD
3RD
1ST
1ST
2ND
2ND
3RD
3RD
DOSE (NG/KG-DAY)
TERTILE -
TERTILE -
TERTILE -
TERTILE -
TERTILE -
TERTILE -
TERTILE -
TERTILE -
TERTILE -
TERTILE -
TERTILE -
TERTILE -
MALE
FEMALE
MALE
FEMALE
MALE
FEMALE
MALE
FEMALE
MALE
FEMALE
MALE
FEMALE
-
-
-
-
-
-
-
-
-
-
-
-
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MEAN
MEAN
MEAN
MEAN
MEAN
MEAN
MAX
MAX
MAX
MAX
MAX
MAX












% HUMAN VARIABLE PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0. % 0 YEARS OLD AT  BEGINNING OF SIMULATION
                                      F-10

-------
% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj
maxCBSNGKGLIADJ=max(_cbsngkgliadj) ;
F.2.3.4. Alaluusua et al (2004) Results
      Table F-5. Matching peak and average after pulse to chronic intake for
      Alaluusua et al. (2004)





Subject
modeled






Tertile
TCDD Only



Measured
LASC
(ng/kg)



Event
dose
(ng/kg)
Average
LASC
after
pulse
dose
(ng/kg)
Continuous
intake
matching
average
LASC
(ng/kg-day)

Peak
LASC
after pulse
dose
(ng/kg)

Continuous
intake
matching
peak LASC
(ng/kg-day)


Average of
continuous
intake rates
(ng/kg-day)
Average of
male and
female
continuous
intake rates
(ng/kg-day)
TEQa
Average of
male and
female
continuous
intake rates
(ng/kg-day)
Needham background
Male
Female
Male
Female
Male
Female
1st

2nd

3rd

72.1

375.4

4266.1

10.9
10.4
105.9
102.3
3419.2
4266.1
61.8
62.1
316.3
318.1
3559.0
3581.9
1.62E-02
1.51E-02
1.53E-01
1.44E-01
4.94E+00
4.68E+00
286.7
271.2
2626.9
2536.8
79877.5
78251.9
6.95E-02
6.15E-02
1.72E+00
1.58E+00
1.14E+02
1.08E+02
4.28E-02
3.83E-02
9.34E-01
8.60E-01
5.95E+01
5.64E+01
4.06E-02

8.97E-01

5.79E+01

4.39E-02

9.01E-01

5.79E+01

"TCDD male/female average + DLC background intake (3.3 x 10 ng/kg-day).
F.2.4. Eskanazi et al. (2002)

F.2.4.1.  Input for Exposure from Event to LASC Measurement
% MODEL  PARAMETERS
output  @clear
prepare  gclear T CBSNGKGLIADJ CBNGKG

% EXPOSURE  PARAMETERS
MAXT =0.5
CINT =  1.
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
MSTOTBCKGR
58692.
58715.
24.
0.
613200.
63072.
0.00039
% AGE AT EXPOSURE  (HOURS)
% AGE AT END OF  EXPOSURE (HOURS)
  LENGTH OF DAY  (HOURS/DAY)
  BACKGROUND EXPOSURE BEGINS AT BIRTH  (AGE 0 HOURS)
% AGE AT END OF  BACKGROUND EXPOSURE  (HOURS)
% AGE AT LASC MEASUREMENT (HOURS)
% NEEDHAM BACKGROUND EXPOSURE DOSE  (NG/KG-DAY)
% EVENT EXPOSURE DOSE (NG/KG-DAY)
MSTOT =5.4     % 28-DAY   EC GROUP
      % 2684.8  % Over 1000 ppt GROUP
% HUMAN VARIABLE PARAMETERS
                                       F-ll

-------
MALE   = 0.
FEMALE = 1.
YO     = 0. % AGE AT  BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
CBSNGKGLIADJ_oneday=mean(_cbsngkgliadj(find(_t==62904):length(t))
F.2.4.2. Input for Exposure from Event to the End of the Assumed Critical Exposure Window

% MODEL PARAMETERS
output @clear
prepare @clear T CBSNGKGLIADJ CBNGKG

% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON  = 58692.
EXP_TIME_OFF = 58715.
DAY_CYCLE    =24.
BCK_TIME_ON  = 0.
BCK_TIME_OFF = 613200.
TIMELIMIT    = 113880.
MSTOTBCKGR   = 0.00039
        % AGE AT EXPOSURE  (HOURS)
        % AGE AT END OF EXPOSURE  (HOURS)
        % LENGTH OF DAY  (HOURS/DAY)
        % BACKGROUND EXPOSURE BEGINS  AT BIRTH (AGE 0 HOURS)
        % AGE AT END OF BACKGROUND  EXPOSURE (HOURS)
        % LENGTH OF CRITICAL WINDOW (HOURS)
        % NEEDHAM BACKGROUND EXPOSURE DOSE (NG/KG-DAY)
% EVENT EXPOSURE DOSE  (NG/KG-DAY)
MSTOT =5.4    % 28-DAY    EC  GROUP
      % 2684.8 % Over  1000 ppt  GROUP

% HUMAN VARIABLE PARAMETERS
MALE   = 0.
FEMALE = 1.
YO     = 0. % AGE AT BEGINNING  OF SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj(find(_t==EXP_TIME_ON):length(_t)));
meanCBSNGKGLIADJ
maxCBSNGKGLIADJ=max(_cbsngkgliadj) ;
maxCBSNGKGLIADJ
F.2.4.3. Input for Continuous Exposure over Assumed Critical Exposure Window
% MODEL PARAMETERS
output @clear
prepare @clear  T CBSNGKGLIADJ CBNGKG

% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK TIME OFF
0.      % CONTINUOUS EXPOSURE  BEGINS  AT BIRTH (AGE 0 HOURS)
113881. % LENGTH OF CRITICAL WINDOW (HOURS)
24.     % LENGTH OF DAY  (HOURS/DAY)
0.      % BACKGROUND EXPOSURE  BEGINS  AT BIRTH (AGE 0 HOURS)
613200. % AGE AT END OF  BACKGROUND  EXPOSURE (HOURS)
                                      F-12

-------
TIMELIMIT     = 113880. % LENGTH OF CRITICAL WINDOW (HOURS)
MSTOTBCKGR   =0.       % /KG-DAYBACKGROUND EXPOSURE INCLUDED IN MSTOT

% CONTINUOUS  EXPOSURE DOSE  (NG/KG-DAY)
MSTOT =  3.64E-3 %  28-DAY    EC  EXPOSURE GROUP  -  MATCHING MEAN
      %  1.51E+0 %  Over 1000  ppt EXPOSURE GROUP  -  MATCHING MEAN
      %  1.68E-2 %  28-DAY    EC  EXPOSURE GROUP  -  MATCHING MAX
      %  6.06E+1 %  Over 1000  ppt EXPOSURE GROUP  -  MATCHING MAX

% HUMAN  VARIABLE PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0.  % 0 YEARS OLD AT BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj) ;
maxCBSNGKGLIADJ=max( cbsngkgliadj);
F.2.4.4.  Eskenazi et al (2002) Results
      Table F-6.  Matching peak and average after pulse to chronic intake for
      Eskenazi et al. (2002)
Subject
modeled
Exposure
group
TCDD Only
Measured
LASC
(ng/kg)
Event
dose
(ng/kg)
Average
LASC
after
pulse
dose
(ng/kg)
Continuous
intake
matching
average
LASC
(ng/kg-day)
Peak
LASC
after pulse
dose
(ng/kg)
Continuous
intake
matching
peak LASC
(ng/kg-day)
Average of
continuous
intake rates
(ng/kg-day)
TEQa
Average of
continuous
intake
rates
(ng/kg-day)
Needham background
Female
Female
28-day EC
Over 1,000
ppt
50
4,060
5.4
2684.8
37.3
2548.8
3.64E-03
1.51E+00
166.9
74597.2
1.68E-02
6.06E+01
1.02E-02
3.11E+01
1.37E-02
3.11E+01
"TCDD average + DLC background intake (3.5 x 10 3 ng/kg-day).
EC = estrous cycle.
                                       F-13

-------
F.3.  KINETIC MODELING OF EPIDEMIOLOGIC STUDIES FOR SENSITIVITY
     ANALYSIS
F.3.1. Alaluusua et al. (2004)
F.3.1.1.  Summary of Modeling Approach
      For the sensitivity analysis, modeling for Alaluusua et al. (2004) (detailed in
Section 4.2.3.3) was repeated using alternative male and female Eskenazi scenario background
intakes for children aged 0-12 as described in Section F.I.2. EPA used the Emond human
PBPK model to estimate continuous daily oral TCDD intakes for each exposure tertile from
corresponding measured LASC values estimated by calculating the geometric mean of the tertile
ranges provided by Alaluusua et al. (2004).  Serum levels were measured within one year of the
incident; in the absence of further specific information about measurement lag, a lag time of 6
months between the event and the measurement was assumed. This value was then used to
model the associated peak and mean LASC from time of the event (average age 2.5 years) to the
end of the critical window (5 years). Continuous daily intakes matching the peak and mean
LASC were determined by modeling exposure from birth to the end of the critical exposure
window. Male and female estimates were modeled separately and then averaged to give a single
continuous intake estimate for each exposure tertile. Total TEQ intake was estimated using the
additive method for both the Needham and Eskenazi scenarios as described previously (see
Sections F. 1.1 and F.I.2).
       Table F-7. Model inputs derived from study details for Alaluusua et al.
       (2004)
Average age at event
(years)
2.5
Time lag between exposure
and LASC measurement
(years)
0.5
Time lag between
exposure and effect
(years)
2.5
Critical exposure
window
(years)
5
F.3.1.2. Input for Exposure from Event to LASC Measurement
% MODEL PARAMETERS
output @clear
prepare @clear T CBSNGKGLIADJ CBNGKG
                                        F-14

-------
% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP TIME ON =21900. % AGE AT EXPOSURE (HOURS)
EXP TIME OFF =21923. % AGE AT END OF EXPOSURE (HOURS)
DAY CYCLE =24. % LENGTH OF DAY (HOURS/DAY)






BCK TIME ON = 0. % BACKGROUND EXPOSURE BEGINS AT BIRTH (AGE 0 HOURS)
BCK TIME OFF = 613200. % AGE AT END OF BACKGROUND EXPOSURE
TIMELIMIT =26280. % AGE AT LASC MEASUREMENT (HOURS)
MSTOTBCKGR = 0.00422 % ESKENAZI BACKGROUND EXPOSURE DOSE
% EVENT EXPOSURE DOSE (NG/KG-DAY)
MSTOT =8.2 % 1ST TERTILE - MALE
% 7.5 % 1ST TERTILE - FEMALE
% 103.1 % 2ND TERTILE - MALE
% 99.4 % 2ND TERTILE - FEMALE
% 3416.5 % 3RD TERTILE - MALE
% 3343.3 % 3RD TERTILE - FEMALE
(HOURS)

(NG/KG-DAY)







% HUMAN VARIABLE PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0. % AGE AT BEGINNING  OF  SIMULATION
% POST-PROCESSING
start @nocallback
CBSNGKGLIADJ_oneday=mean(_cbsngkgliadj(findi
t==26112):length(t)))
F.3.1.3. Input for Exposure from Event to the End of the Assumed Critical Exposure Window
% MODEL PARAMETERS
output @clear
prepare @clear T CBSNGKGLIADJ  CBNGKG
% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP TIME ON =
EXP TIME OFF =
DAY CYCLE
BCK TIME ON =
BCK TIME OFF =
TIMELIMIT
MSTOTBCKGR
%


21900. % AGE AT EXPOSURE (HOURS)
21923. % AGE AT END OF EXPOSURE (HOURS)
24. % LENGTH OF DAY (HOURS/DAY)
0. % BACKGROUND EXPOSURE BEGINS AT





BIRTH (AGE 0 HOURS)
613200. % AGE AT END OF BACKGROUND EXPOSURE (HOURS)
43800. % LENGTH OF CRITICAL EXPOSURE WINDOW (HOURS)
0.00422 % ESKENAZI BACKGROUND EXPOSURE,
0.00429 % ESKENAZI BACKGROUND EXPOSURE,
MALES (NG/KG-DAY)
FEMALES (NG/KG-DAY)
% EVENT EXPOSURE DOSE (NG/KG-DAY)
MSTOT =8.2
% 7.5
% 103.1
% 99.4
% 3416.5
% 3343.3
% 1ST TERTILE - MALE
% 1ST TERTILE - FEMALE
% 2ND TERTILE - MALE
% 2ND TERTILE - FEMALE
% 3RD TERTILE - MALE
% 3RD TERTILE - FEMALE






% HUMAN VARIABLE PARAMETERS
MALE   = 1.
                                      F-15

-------
FEMALE = 0.
YO     = 0. % AGE AT  BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj(find(_t==EXP_TIME_ON):length(_t)));
meanCBSNGKGLIADJ
maxCBSNGKGLIADJ=max(_cbsngkgliadj) ;
maxCBSNGKGLIADJ
F.3.1.4. Input for Continuous Exposure over Assumed Critical Exposure Window
% MODEL PARAMETERS
output @clear
prepare @clear  T CBSNGKGLIADJ  CBNGKG
% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON  = 0.
EXP_TIME_OFF = 43801.
DAY_CYCLE    = 24.
BCK_TIME_ON  = 0.
BCK_TIME_OFF = 613200.
TIMELIMIT    = 43800.
MSTOTBCKGR   = 0.
   CONTINUOUS EXPOSURE BEGINS AT  BIRTH  (AGE 0 HOURS)
   LENGTH OF ASSUMED CRITICAL EXPOSURE  WINDOW (HOURS)
   LENGTH OF DAY  (HOURS/DAY)
   BACKGROUND EXPOSURE BEGINS AT  BIRTH  (AGE 0 HOURS)
   AGE AT END OF BACKGROUND  EXPOSURE  (HOURS)
   LENGTH OF CRITICAL EXPOSURE WINDOW (HOURS)
   /KG-DAYBACKGROUND EXPOSURE INCLUDED  IN MSTOT
% CONTINUOUS EXPOSURE
MSTOT = 1.81E-2 %  1ST
      % 1.69E-2 %  1ST
      % 1.56E-1 %  2ND
      % 1.46E-1 %  2ND
      % 4.94E+0 %  3RD
      % 4.68E+0 %  3RD
      % 4.70E-2 %  1ST
      % 4.04E-2 %  1ST
      % 1.58E+0 %  2ND
      % 1.45E+0 %  2ND
      % 1.13E+2 %  3RD
      % 1.07E+2 %  3RD
DOSE  (NG/KG-DAY)
TERTILE - MALE
TERTILE - FEMALE
TERTILE - MALE
TERTILE - FEMALE
TERTILE - MALE
TERTILE - FEMALE
TERTILE - MALE
TERTILE - FEMALE
TERTILE - MALE
TERTILE - FEMALE
TERTILE - MALE
TERTILE - FEMALE
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MATCHING
MEAN
MEAN
MEAN
MEAN
MEAN
MEAN
PEAK
PEAK
PEAK
PEAK
PEAK
PEAK
% HUMAN VARIABLE PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0. % 0 YEARS OLD AT  BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj);
maxCBSNGKGLIADJ=max( cbsngkgliadj);
                                      F-16

-------
F.3.1.5. Alaluusua et al (2004) Results
       Table F-8. Matching peak and average after pulse to chronic intake for
       Alaluusua et al. (2004) using alternate background value
Subject
modeled
Tertile
TCDD Only
Measured
LASC
(ng/kg)
Event
dose
(ng/kg)
Average
LASC
after
pulse
dose
(ng/kg)
Continuous
intake
matching
average
LASC
(ng/kg-day)
Peak
LASC
after pulse
dose
(ng/kg)
Continuous
intake
matching
peak LASC
(ng/kg-day)
Average of
continuous
intake rates
(ng/kg-day)
Average of
male and
female
continuous
intake rates
(ng/kg-day)
TEQa
Average of
male and
female
continuous
intake rates
(ng/kg-day)
Eskenazi background
Male
Female
Male
Female
Male
Female
1st
^ncl
3rd
72.1
375.4
4266.1
8.2
7.5
103.1
99.4
3416.5
3343.3
67.5
68.0
319.4
321.2
3560.0
3582.9
1.81E-02
1.69E-02
1.56E-01
1.46E-01
4.94E+00
4.68E+00
218.4
203.0
2479.1
2390.4
79502.9
77847.7
4.70E-02
4.04E-02
1.58E+00
1.45E+00
1.13E+02
1.07E+02
3.25E-02
2.87E-02
8.68E-01
7.97E-01
5.92E+01
5.61E+01
3.06E-02
8.32E-01
5.76E+01
3.86E-02
8.40E-01
5.76E+01
"TCDD male/female average + DLC male/female average background intake (8.0 x 10 3 ng/kg-day).

F.3.2. Baccarelli et al. (2008)
F.3.2.1. Summary of Modeling Approach
       For the sensitivity analysis, total TEQ intakes were estimated.  For Baccarelli et al.
(2008), total TEQ exposure was obtained from the study author's Figure 2D by digitizing the
figure and finding the TEQ concentration on the regression line associated with a b-TSH
of 5 uU/mL (489 ppt). Modeling was then repeated as described in Section F.3.1.1 to determine
the continuous daily intake associated with this concentration.

F.3.2.2. Baccarelli et al (2008) Results
       Table F-9. Estimated continuous intake corresponding to maternal serum
       concentration for TEQ
Variable
Infant b-TSH
Maternal lipid adjusted serum TEQ
Intake
Value
5 uU/mL
489 ng/kg
0.059 ng/kg-day
Notes
BMR
From Figure 2D in For Baccarelli et al. (2008)
From Emond model; pregnancy at 30 years
TSH = thyroid stimulating hormone; BMR = benchmark response.
                                          F-17

-------
F.3.3. Eskenazi et al. (2002)
F.3.3.1.  Summary of Modeling Approach
      For the sensitivity analysis, modeling for Eskenazi et al. (2002) (detailed in
Section 4.2.3.4) was repeated using the Eskenazi scenario female background intake (see
Section F. 1.2). Modeling was carried out for the mid and high exposure tertiles as described in
Section F.3.1.1 using this alternative background value. The measured LASC of the lowest
exposure tertile was lower than the estimated background exposure; thus, for this tertile, the
Emond human PBPK model was used to find the chronic intake over the critical exposure
window (13 years) which matched the measured concentration.
      As part of the sensitivity analysis, the total TEQ intakes were estimated. For the mid and
high tertiles, this was done by adding the Eskenazi scenario female background DLC intake to
the calculated TCDD intake as discussed in Section F.3.1.1.  Total TEQ intake  was estimated for
the lowest tertile assuming that TEQ intake is equal to ten times the modeled TCDD intake.
       Table F-10. Model inputs derived from study details for Eskenazi et al.
       (2002)
Average age at event
(years)
6.7
Time lag between exposure
and LASC measurement
(years)
0.5
Time lag between
exposure and effect
(years)
6.7
Critical exposure
window
(years)
13
F.3.3.2. Input for Exposure from Event to LASC Measurement
% MODEL PARAMETERS
output @clear
prepare @clear T CBSNGKGLIADJ CBNGKG
% EXPOSURE PARAMETERS
MAXT =0.5
CINT =  1.
EXP_TIME_ON  = 58692.   %
EXP_TIME_OFF = 58715.   %
DAY_CYCLE     = 24.      %
BCK_TIME_ON  =0.       %
BCK_TIME_OFF = 613200.  %
TIMELIMIT     = 63072.   %
MSTOTBCKGR   = 0.00422  %
AGE AT  EXPOSURE  (HOURS)
AGE AT  END OF EXPOSURE (HOURS)
LENGTH  OF DAY (HOURS/DAY)
BACKGROUND EXPOSURE  BEGINS AT BIRTH (AGE 0 HOURS)
AGE AT  END OF BACKGROUND EXPOSURE  (HOURS)
AGE AT  LASC MEASUREMENT (HOURS)
ESKENAZI  BACKGROUND  EXPOSURE, FEMALES (NG/KG-DAY)
% EVENT  EXPOSURE DOSE (NG/KG-DAY)
MSTOT  =  2679.4  % Over 1000 ppt  GROUP

% HUMAN  VARIABLE PARAMETERS
                                        F-18

-------
MALE   = 0.
FEMALE = 1.
YO     = 0. % AGE AT  BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
CBSNGKGLIADJ_oneday=mean(_cbsngkgliadj(find(_t==62904):length(t))
F.3.3.3. Input for Exposure from Event to the End of the Assumed Critical Exposure Window

% MODEL PARAMETERS
output @clear
prepare @clear T CBSNGKGLIADJ CBNGKG

% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
MSTOTBCKGR
58692.
58715.
24.
0.
613200.
113880.
            AGE AT EXPOSURE  (HOURS)
            AGE AT END OF  EXPOSURE (HOURS)
            LENGTH OF DAY  (HOURS/DAY)
            BACKGROUND EXPOSURE  BEGINS AT BIRTH (AGE 0 HOURS)
            AGE AT END OF  BACKGROUND  EXPOSURE (HOURS)
            LENGTH OF CRITICAL WINDOW (HOURS)
= 0.00429 % ESKENAZI BACKGROUND  EXPOSURE, FEMALES (NG/KG-DAY)
% EVENT EXPOSURE DOSE  (NG/KG-DAY)
MSTOT = 2679.4  % Over  1000  ppt GROUP

% HUMAN VARIABLE PARAMETERS
MALE   = 0.
FEMALE = 1.
YO     = 0. % AGE AT BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj(find(_t==EXP_TIME_ON):length(_t)));
meanCBSNGKGLIADJ
maxCBSNGKGLIADJ=max(_cbsngkgliadj);
maxCBSNGKGLIADJ
F.3.3.4. Input for Continuous Exposure over Assumed Critical Exposure Window
output @clear
prepare @clear  T CBSNGKGLIADJ CBNGKG
% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON  =  0.
EXP_TIME_OFF =  113881.
DAYJCYCLE    =24.
BCK_TIME_ON  =  0.
BCK_TIME_OFF =  613200.
TIMELIMIT    =  113880.
MSTOTBCKGR   =  0.
        % CONTINUOUS EXPOSURE  BEGINS  AT BIRTH (AGE 0 HOURS)
        % LENGTH OF CRITICAL WINDOW (HOURS)
        % LENGTH OF DAY  (HOURS/DAY)
        % BACKGROUND EXPOSURE  BEGINS  AT BIRTH (AGE 0 HOURS)
        % AGE AT END OF  BACKGROUND  EXPOSURE (HOURS)
        % LENGTH OF CRITICAL WINDOW (HOURS)
        % /KG-DAYBACKGROUND EXPOSURE  INCLUDED IN MSTOT
                                      F-19

-------
% CONTINUOUS EXPOSURE  DOSE (NG/KG-DAY)
MSTOT =  3.08E-3 % 28-DAY   EC EXPOSURE  GROUP
      %  1.52E+0 % Over 1000 ppt EXPOSURE GROUP
      %  6.00E+1 % Over 1000 ppt EXPOSURE GROUP
MATCHING  MEAN
MATCHING  MAX
% HUMAN  VARIABLE PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0.  % 0 YEARS  OLD AT BEGINNING  OF SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj) ;
maxCBSNGKGLIADJ=max(_cbsngkgliadj)  ;

F.3.3.5. Eskenazi et al (2002) Results
      Table F-ll. Matching peak and average after pulse to chronic intake for
      Eskenazi et al. (2002) using alternate background value









Subject
modeled









Exposure
group
TCDD Only







Measured
LASC
(ng/kg)







Event
dose
(ng/kg)




Average
LASC
after
pulse
dose
(ng/kg)




Continuous
intake
matching
average
LASC
(ng/kg-day)




Peak
LASC
after
pulse
dose
(ng/kg)
Continuous
intake
matching
peak LASC/
measured
concentration
(if LASC
below
background)
(ng/kg-day)






Average of
continuous
intake rates
(ng/kg-day)
TEQa






Average of
continuous
intake rates
(ng/kg-day)
Eskenazi background
Female


Female
28-day EC
Over 1000

ppt
50

4060

Below background

2679.4


2552.8


1.52E+00


73933.1

3.08E-03

6.00E+01

3.08E-03

3.08E+01

1.12E-02

3.08E+01

"TCDD average + DLC background intake (8.07 x 10 3 ng/kg-day).


F.3.4. Eskenazi et al. (2005)
F.3.4.1.  Summary of Modeling Approach
      Eskenazi et al. (2005) investigated the association of TCDD exposure and age at
menopause in women who were premenopausal in 1976 and living near Seveso, Italy.  Study
authors divided TCDD exposures into quintiles for analysis (reported in Table 3 inEskenazi et
al., [2005]). Because the dose-response trend is not clear, it was difficult to determine a NOAEL
and LOAEL for this study, and all quintiles were modeled. Measured LASC values for the
                                        F-20

-------
second, third, and fourth quintiles were estimated by calculating the geometric means of the
quintile ranges rounded to the nearest tenth. No range was specified for the first quintile
(defined as <20.4 ppt) and fifth quintile (defined as >300 ppt). Instead, for the first quintile,
measured LASC was estimated by dividing the upper bound of the exposure range by 2 to give
an estimate of 10.2 ppt. For the fifth quintile, the lower bound of the exposure range was used as
the measured LASC estimate.
       The mean age at time of the incident was not reported by Eskenazi et al.(2005). Thus, the
age at incident was approximated by subtracting the lag between event and interview (21 years)
from the mean age at menopause (56.6, Table 1 in the study report) to get an approximate mean
age at incident of 35.6 years old. A critical susceptibility window for this endpoint could not be
determined.  Because women are susceptible to ovarian function effects until menopause, an
assumed critical exposure window of 50 years was assigned for the sensitivity analysis. Serum
levels were measured within one year of the incident, and an LASC measurement lag time of 0.5
years was assumed.  Modeling was carried out as detailed in Section F.3.1.1 for the second, third,
fourth, and fifth quintiles using the Needham background  scenario intake estimated (see
Section F. 1.1). The measured LASC of the first quintile was lower than the estimated Needham
background scenario exposure; thus, for this quintile, the Emond human PBPK model  was used
to find the intake over the assumed critical exposure window which matched the measured
LASC value.
       As part of the sensitivity analysis, total TEQ intakes were estimated for the second, third,
fourth, and fifth quintiles by adding the Needham scenario background DLC intake to the
modeled TCDD intake as discussed in Section F.3.1.1.  Total TEQ intake for the first quintile
was estimated assuming that total TEQ intake is equal to ten times the modeled TCDD intake.
       Table F-12.  Model inputs derived from study details for Eskenazi et al.
       (2005)
Average age at event
(years)
35.6
Time lag between exposure
and LASC measurement
(years)
0.5
Time lag between
exposure and effect
(years)
13.6
Assumed critical
exposure window
(years)
50
                                         F-21

-------
F.3.4.2. Input for Exposure from Event to LASC Measurement

% MODEL PARAMETERS
output @clear
prepare gclear T CBSNGKGLIADJ CBNGKG

% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON  = 311856.
EXP_TIME_OFF = 311879.
DAY_CYCLE    = 24.
BCK_TIME_ON  = 0.
BCK_TIME_OFF = 613200.
TIMELIMIT    = 316236.
MSTOTBCKGR
          % AGE AT EXPOSURE  (HOURS)
          % AGE AT END OF  EXPOSURE  (HOURS)
          % LENGTH OF DAY  (HOURS/DAY)
          % BACKGROUND EXPOSURE  BEGINS  AT BIRTH (AGE 0 HOURS)
          % AGE AT END OF  BACKGROUND  EXPOSURE (HOURS)
          % AGE AT LASC MEASUREMENT  (HOURS)
= 0.00039 % NEEDHAM BACKGROUND  EXPOSURE DOSE (NG/KG-DAY)
% EVENT EXPOSURE DOSE  (NG/KG-DAY)
MSTOT =2.1  % 2ND QUINTILE
      % 5.5  % 3RD QUINTILE
      % 13.8 % 4TH QUINTILE
      % 23.4 % 5TH QUINTILE

% HUMAN VARIABLE PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0. % AGE AT BEGINNING OF  SIMULATION
% POST-PROCESSING
start @nocallback
CBSNGKGLIADJ oneday=meani
             cbsngkgliadj(find(  t==316068):length(  t)))
F.3.4.3. Input for Exposure from Event to the End of the Assumed Critical Exposure Window
% MODEL PARAMETERS
output @clear
prepare @clear T CBSNGKGLIADJ CBNGKG

% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
MSTOTBCKGR
  311856. % AGE AT EXPOSURE  (HOURS)
  311879. % AGE AT END OF  EXPOSURE  (HOURS)
  24.     % LENGTH OF DAY  (HOURS/DAY)
  0.      % BACKGROUND EXPOSURE  BEGINS  AT BIRTH (AGE 0 HOURS)
  613200. % AGE AT END OF  BACKGROUND  EXPOSURE (HOURS)
  438000. % LENGTH OF ASSUMED  CRITICAL  EXPOSURE WINDOW (HOURS)
  0.00039 % NEEDHAM BACKGROUND EXPOSURE DOSE (NG/KG-DAY)
% EVENT EXPOSURE DOSE  (NG/KG-DAY)
MSTOT =2.1  % 2ND QUINTILE
      % 5.5  % 3RD QUINTILE
      % 13.8 % 4TH QUINTILE
      % 23.4 % 5TH QUINTILE

% HUMAN VARIABLE PARAMETERS
                                      F-22

-------
MALE   = 1.
FEMALE = 0.
YO     = 0. % AGE AT  BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj(find(_t==EXP_TIME_ON):length(_t)));
meanCBSNGKGLIADJ
maxCBSNGKGLIADJ=max(_cbsngkgliadj);
maxCBSNGKGLIADJ
F.3.4.4. Input for Continuous Exposure over Assumed Critical Exposure Window
% MODEL PARAMETERS
output @clear
prepare @clear  T CBSNGKGLIADJ  CBNGKG
% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON  = 0.
EXP_TIME_OFF = 438001.
DAY_CYCLE    = 24.
BCK_TIME_ON  = 0.
BCK_TIME_OFF = 613200.
TIMELIMIT    = 438000.
MSTOTBCKGR   = 0.
% CONTINUOUS EXPOSURE BEGINS AT BIRTH  (AGE  0  HOURS)
% LENGTH OF ASSUMED CRITICAL EXPOSURE WINDOW  (HOURS)
% LENGTH OF DAY  (HOURS/DAY)
% BACKGROUND EXPOSURE BEGINS AT BIRTH  (AGE  0  HOURS)
% AGE AT END OF BACKGROUND EXPOSURE  (HOURS)
% LENGTH OF ASSUMED CRITICAL EXPOSURE WINDOW  (HOURS)
% /KG-DAYBACKGROUND EXPOSURE INCLUDED IN MSTOT
% CONTINUOUS EXPOSURE  DOSE  (NG/KG-DAY)
MSTOT = 1.04E-3 % 2ND  QUINTILE  -
                   3RD QUINTILE
                   4TH QUINTILE
                   5TH QUINTILE
                   2ND QUINTILE
                   3RD QUINTILE
                   4TH QUINTILE
                   5TH QUINTILE
          MATCHING
          MATCHING
          MATCHING
          MATCHING
          MATCHING
          MATCHING
          MATCHING
          MATCHING
MEAN
MEAN
MEAN
MEAN
MAX
MAX
MAX
MAX
% HUMAN VARIABLE PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0. % 0 YEARS OLD AT  BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj);
maxCBSNGKGLIADJ=max(_cbsngkgliadj) ;
                                      F-23

-------
F.3.4.5. Eskenazi et al (2005) Results
       Table F-13. Matching peak and average after pulse to chronic intake for
       Eskenazi et al. (2005)







Subject
modeled
Female
Female
Female
Female
Female








Quintile
1st
^nd
3rd
4th
5th
TCDD only





Measured
LASC
(ng/kg)
10.2
26.4
43.1
80.0
118.0





Event
dose
(ng/kg)


Average
LASC
after
pulse
dose
(ng/kg)


Continuous
intake
matching
average
LASC
(ng/kg-day)


Peak
LASC
after
pulse
dose
(ng/kg)
LASC below background
2.1
5.5
13.8
23.4
25.9
37.7
62.1
85.9
1.04E-03
1.73E-03
3.44E-03
5.47E-03
89.4
209.4
506.1
848.3
Continuous
intake matching
peak LASC/
measured
concentration (if
LASC below
background)
(ng/kg-day)
1.57E-04
3.42E-03
1.29E-02
5.16E-02
1.15E-01




Average of
continuous
intake rates
(ng/kg-day)
1.57E-04
2.23E-03
7.31E-03
2.75E-02
6.02E-02
TEQa




Average of
continuous
intake rates
(ng/kg-day)
1.57E-03b
5.74E-03
1.08E-02
3.10E-02
6.37E-02
"TCDD average + DLC background intake (Needham = 3.51 x 10~3 ng/kg-day).
"Values below background multiplied by 10, assuming total TEQ = 10 x TCDD.
F.3.5. Mocarelli et al. (2000)
F.3.5.1. Summary of Modeling Approach
       Mocarelli et al. (2000) examined sex ratio of offspring born to parents exposed to dioxin
in Seveso, Italy.  Sex and age at exposure were also tested as factors possibly affecting sex ratio.
Because no difference in sex ratio was observed in groups in which only the mothers were
exposed to TCDD, only male exposures were modeled. Because the authors conducted this
statistical test using a dichotomous exposure variable (exposed vs. unexposed or <15 ppt), and
because there is no clear dose-response trend in sex ratios of offspring and father's TCDD
concentrations, a NOAEL and LOAEL were difficult to establish  for this study.  All quintiles
(reported in Table 2 in the study report) of fathers' exposure were modeled using the Emond
human PBPK model. Measured LASC values for all quintiles were estimated by calculating the
geometric mean of the quintile ranges reported in Table 2 in the study.
       Average ages at conception for various year ranges were provided in the study in Table 5.
From these ages, a population-weighted average age at conception of 31.0 and average age at the
time of exposure in 1976 of 20.5 were calculated. No critical susceptibility window could be
determined for this study; however, an assumed critical exposure window of 31.0 years was
                                          F-24

-------
assumed to match the average age at time of conception. Modeling was carried out as detailed in
Section F.3.1.1 using the Needham scenario background intake (see Section F. 1.1) with the
exception that a 5-year response surface was used to find continuous intakes matching the
modeled peak and mean LASC values, as detailed in Section F.3.5.1.
      As part of the sensitivity analysis, total TEQ intakes were estimated for all tertiles by
adding the Needham scenario background DLC intake to the modeled TCDD intake as described
in Section F.3.1.1.
      Table F-14. Model inputs derived from study details for Mocarelli et al.
      (2000)
Average age at event
(years)
20.5
Time lag between exposure and
LASC measurement
(years)
0.5
Time lag between exposure and
effect
(years)
20
Assumed critical
exposure window
(years)
31.0
F.3.5.2.  Input for Exposure from Event to LASC Measurement

% MODEL PARAMETERS
output  @clear
prepare @clear T CBSNGKGLIADJ CBNGKG

% EXPOSURE PARAMETERS
MAXT =0.5
CINT =  1.
EXP_TIME_ON  = 179580.
EXP_TIME_OFF = 179603.
DAY_CYCLE     = 24.
BCK_TIME_ON  = 0.
BCK_TIME_OFF = 613200.
TIMELIMIT     = 183960.
MSTOTBCKGR
           %  AGE AT EXPOSURE  (HOURS)
           %  AGE AT END OF  EXPOSURE  (HOURS)
           %  LENGTH OF DAY  (HOURS/DAY)
           %  BACKGROUND EXPOSURE BEGINS AT  BIRTH (AGE 0 HOURS)
           %  AGE AT END OF  BACKGROUND EXPOSURE (HOURS)
           %  AGE AT LASC MEASUREMENT  (HOURS)
= 0.00035  %  NEEDHAM BACKGROUND EXPOSURE DOSE  (NG/KG-DAY)
% EVENT  EXPOSURE DOSE  (NG/KG-DAY)
MSTOT =1.2     % 1ST QUINTILE
      %  4.2     % 2ND QUINTILE
      %  11.0   % 3RD QUINTILE
      %  30.2   % 4TH QUINTILE
      %  1420.0 % 5TH QUINTILE

% HUMAN  VARIABLE PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0.  % AGE AT BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback

                                       F-25

-------
CBSNGKGLIADJ_oneday=mean(_cbsngkgliadj(find(_t==183792):length(_
                                                  t) ) )
F.3.5.3. Input for Exposure from Event to the End of the Assumed Critical Exposure Window
% MODEL PARAMETERS
output @clear
prepare @clear T CBSNGKGLIADJ  CBNGKG

% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
MSTOTBCKGR
179580. % AGE AT EXPOSURE  (HOURS)
179603. % AGE AT END OF EXPOSURE  (HOURS)
24.     % LENGTH OF DAY  (HOURS/DAY)
0.       % BACKGROUND EXPOSURE BEGINS AT  BIRTH (AGE 0 HOURS)
613200. % AGE AT END OF BACKGROUND  EXPOSURE (HOURS)
271560. % LENGTH OF ASSUMED CRITICAL EXPOSURE WINDOW (HOURS)
0.00035 % NEEDHAM BACKGROUND EXPOSURE  DOSE  (NG/KG-DAY)
  EVENT EXPOSURE DOSE  (NG/KG-DAY)
MSTOT =
  1ST QUINTILE
  2ND QUINTILE
  3RD QUINTILE
  4TH QUINTILE
  5TH QUINTILE
% HUMAN VARIABLE PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0. % AGE AT BEGINNING  OF  SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj(find(_t==EXP_TIME_ON):length(_t)));
meanCBSNGKGLIADJ
maxCBSNGKGLIADJ=max(_cbsngkgliadj) ;
maxCBSNGKGLIADJ
                                      F-26

-------
F.3.5.4. Mocaretti et al (2000) Results
       Table F-15. Matching peak and average after pulse to 5-year average
       response surface for Mocarelli et al. (2000)






Subject
modeled
Male
Male
Male
Male
Male







Quintile
1st
-^nd
ord
J
4th
5th
TCDD only




Measured
LASC
(ng/kg)
21.7
44
84.8
176.5
2723.7




Event
dose
(ng/kg)
1.2
4.2
11.0
30.2
1420.0

Average
LASC
after
pulse
dose
(ng/kg)
19.0
33.0
46.9
112.4
1485.2
5-Year
response
surface
matching
average
LASC
(ng/kg-day)
2.82E-04
6.56E-04
1.58E-03
4.69E-03
2.66E-01

Peak
LASC
after
pulse
dose
(ng/kg)
52.4
160.0
397.3
1072.0
48470.7

5-Year
response
surface
matching
peak LASC
(ng/kg-day)
1.35E-03
7.93E-03
3.41E-02
1.62E-01
2.63E+01

Average of
5-Year
response
surface
values
(ng/kg-day)
8.17E-04
4.30E-03
1.78E-02
8.31E-02
1.33E+01
TEQ

Average of
5-Year
response
surface
values
(ng/kg-day)
3.97E-03
7.45E-03
2.10E-02
8.63E-02
1.33E+01
F.3.6. Mocarelli et al. (2008)
F.3.6.1. Summary of Modeling Approach
       For the sensitivity analysis, modeling for Mocarelli et al. (2008) (detailed in
Section 4.2.3.2) was repeated for the 1st quartile (LOAEL group), only, using the male TCDD
background intake of 4.22 x  10~3 ng/kg-day estimated for the Eskenazi scenario (see Table F-2)
for children aged 0-12. Modeling was carried out as described in Section F.3.1.1 using this
alternative background value.
       As part of the sensitivity analysis, total TEQ intakes also were modeled for the 1st quartile
using the Needham and Eskenazi scenario background TEQ intakes (see Tables F-2 and F-2).
This approach models the exposure directly, by matching the total TEQ (as LASC ppt, TCDD
included) at the time of TCDD measurement (i.e., serum sampling for boys 6.7 years old) with
the corresponding intake using the Emond model. For the Needham scenario, background TEQ
LASC at the time of measurement was estimated by running the Emond model from birth to
age 6.7 with a constant exposure of 8.9 x 10~3 ng/kg-day. The resulting total TEQ background
LASC of 80.6 ppt was multiplied by 0.9 to obtain the corresponding DLC-TEQ LASC
(72.5 ppt), which was added to the measured TCDD LASC of 68 as an estimate of total TEQ
LASC at time of measurement.
                                        F-27

-------
      For the Eskenazi scenario, the background TCDD and TEQ LASC values are given as
age-group averages, rather than time-point values.  The averages were assumed to be the
background levels at time of measurement.  The total TEQ LASC is 93.7 ppt (see Table F-2).
The DLC-TEQ contribution to background exposure is 53.2 ppt, which is added to the measured
TCDD LASC of 68 as an estimate of total TEQ LASC at time of measurement.
      An additional TCDD-only analysis was run for a Hill coefficient (HILL) value of 1 and
an elimination constant (KELV) of 0.005, which was optimized.
      Table F-16. Model inputs derived from study details for Mocarelli et al.
      (2008)
Average age at event
(years)
6.2
Time lag between exposure
and LASC measurement
(years)
0.5
Time lag between
exposure and effect
(years)
3.8
Critical exposure
window
(years)
10
F.3.6.2. Input for Exposure from Event to LASC Measurement
% MODEL PARAMETERS
output @clear
prepare @clear  T  CBSNGKGLIADJ CBNGKG
% EXPOSURE  PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
MSTOTBCKGR
= 54312.
= 54335.
= 24.
= 0.
= 613200.
= 58692.
= 0.00422
% 0.0132
% 0.0089
% AGE AT EXPOSURE  (HOURS)
% AGE AT END OF  EXPOSURE (HOURS)
% LENGTH OF DAY  (HOURS/DAY)
% BACKGROUND EXPOSURE  BEGINS AT BIRTH  (AGE  0  HOURS)
% AGE AT END OF  BACKGROUND EXPOSURE  (HOURS)
% AGE AT LASC MEASUREMENT  (HOURS)
% ESKENAZI BACKGROUND  TCDD INTAKE  (NG/KG-DAY)
% ESKENAZI BACKGROUND  TEQ  INTAKE  (NG/KG-DAY)
% NEEDHAM BACKGROUND TEQ INTAKE (NG/KG-DAY)
% EVENT EXPOSURE DOSE (NG/KG-DAY)
MSTOT = 3.36  %  TCDD only, ESKENAZI SCENARIO,  1ST QUARTILE
      % 2.9   %  TOTAL TEQ, ESKENAZI SCENARIO,  1ST QUARTILE
      % 11.8  %  TOTAL TEQ, NEEDHAM SCENARIO,  1ST QUARTILE

% HUMAN VARIABLE PARAMETERS
MALE   =  1.
FEMALE =  0.
YO     =  0. % AGE AT BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
                                       F-28

-------
CBSNGKGLIADJ_oneday=mean(_cbsngkgliadj(find(_t==58524):length(_t
                                                                  ) )
F.3.6.3. Input for Exposure from Event to the End of the Assumed Critical Exposure Window
% MODEL PARAMETERS
output @clear
prepare @clear T CBSNGKGLIADJ CBNGKG
% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
MSTOTBCKGR
               54312.
               54335.
               24.
               0.
               613200.
               87600.
               0.00422
               0.0132
               0.0089
/KG-DAY
% EVENT EXPOSURE
MSTOT =3.36
        2.9
        11.8
AGE AT EXPOSURE  (HOURS)
AGE AT END OF EXPOSURE  (HOURS)
LENGTH OF DAY  (HOURS/DAY)
BACKGROUND EXPOSURE BEGINS  AT BIRTH (AGE 0 HOURS)
AGE AT END OF BACKGROUND  EXPOSURE (HOURS)
LENGTH OF CRITICAL EXPOSURE WINDOW (HOURS)
ESKENAZI BACKGROUND TCDD  INTAKE (NG/KG-DAY)
ESKENAZI BACKGROUND TEQ INTAKE (NG/KG-DAY)
NEEDHAM BACKGROUND TEQ INTAKE (NG/KG-DAY)
                 DOSE  (NG/KG-DAY)
                TCDD only,  ESKENAZI  SCENARIO
                TOTAL  TEQ,  ESKENAZI  SCENARIO
                TOTAL  TEQ,  NEEDHAM SCENARIO
% HUMAN VARIABLE  PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0. % AGE AT  BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj(find(_t==EXP_TIME_ON):length(_t)));
meanCBSNGKGLIADJ
maxCBSNGKGLIADJ=max(_cbsngkgliadj);
maxCBSNGKGLIADJ
F.3.6.4. Input for Continuous Exposure over Assumed Critical Exposure Window
% MODEL PARAMETERS
output @clear
prepare @clear  T CBSNGKGLIADJ CBNGKG
% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON  =  0.
EXP_TIME_OFF =  87601.
DAYJCYCLE    =24.
BCK_TIME_ON  =  0.
BCK_TIME_OFF =  613200.
TIMELIMIT    =  87600.
MSTOTBCKGR   =  0.
                        %  CONTINUOUS  EXPOSURE BEGINS AT BIRTH  (AGE 0 HOURS)
                        %  LENGTH  OF ASSUMED CRITICAL EXPOSURE WINDOW  (HOURS)
                        %  LENGTH  OF DAY (HOURS/DAY)
                        %  BACKGROUND  EXPOSURE BEGINS AT BIRTH  (AGE 0 HOURS)
                        %  AGE AT  END  OF BACKGROUND EXPOSURE  (HOURS)
                        %  LENGTH  OF CRITICAL EXPOSURE WINDOW  (HOURS)
                        %  /KG-DAYBACKGROUND EXPOSURE INCLUDED IN MSTOT
                                      F-29

-------
% CONTINUOUS EXPOSURE DOSE  (NG/KG-DAY)
MSTOT  =  1.03E-2 % TCDD,  ESKENAZI  SCENARIO
       %  1.34E-2 % TCDD,  ESKENAZI  SCENARIO
         2.45E-2 % TEQ,  ESKENAZI  SCENARIO -
         2.02E-2 % TEQ,  ESKENAZI  SCENARIO -
       %  2.56E-2 % TEQ,  NEEDHAM SCENARIO -
       %  6.66E-2 % TEQ,  NEEDHAM SCENARIO -
- MATCHING MEAN
- MATCHING PEAK
• MATCHING MEAN
• MATCHING PEAK
MATCHING MEAN
MATCHING PEAK
% HUMAN  VARIABLE PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0.  % 0 YEARS  OLD AT BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj);
maxCBSNGKGLIADJ=max(_cbsngkgliadj) ;
F.3.6.5. Mocarelli et al (2008) Results
       Table F-17. Matching peak and average after pulse to critical-window intake
       for Mocarelli et al. (2008) using alternate background value
Subject
modeled
Quartile
Measured
LASC
(ng/kg)
Event
dose
(ng/kg)
Average
LASC
after pulse
dose
(ng/kg)
Continuous
intake matching
average LASC
(ng/kg-day)
Peak
LASC
after pulse
dose
(ng/kg)
Continuous
intake
matching
peak LASC
(ng/kg-day)
Average of
continuous
intake rates
(ng/kg-day)
TCDD Eskenazi scenario; background LASC = 40.5 (4.22E-03 ng/kg-day)
Male
1st
68
3.36
69.7
1.03E-02
137.6
1.34E-02
1.18E-02
TEQ Eskenazi scenario; background LASC = 93.7 (1.32E-03 ng/kg-day)
Male
1st
121.23
2.9
131.2
2.45E-02
181.7
2.02E-02
C
TEQ Needham scenario; background LASC = 150 ppt (8.9E-03 ng/kg-day)
Male
1st
140.5b
11.8
135.4
2.56E-02
403.7
6.66E-02
4.61E-02
TCDD alternate Hill coefficient scenario"1; background intake = 1.9E-04 ng/kg-day
Male
1st
86
4.11
64.2
3.73E-03
254.8
7.61E-03
5.67E-03
a68 ppt TCDD + 72.5 ppt DLC-TEQ.
bWindow-average > Peak; overall average not meaningful.
C68 ppt TCDD + 53.2 ppt DLC-TEQ.
dHILL = 1, KELV= 0.005, Needham background scenario (15 ppt at 35 years).
F.3.7. Mocarelli et al. (2011)
F.3.7.1.  Summary of Modeling Approach
      Mocarelli et al. (2011) examined sperm effects in boys who experienced perinatal TCDD
exposure during the Seveso event in 1976.  Study authors used a model based on lst-order
kinetics to extrapolate the measured LASC concentrations to the concentration at conception.
                                        F-30

-------
For consistency with all other exposure estimates, EPA did not use the study authors' exposure
estimates and instead used the Emond human PBPK model to estimate concentrations at
conception. The median measured LASC for mothers who breastfed was provided in the study
(reported in Table 2 of the study) and was selected as a LOAEL. Measured LASC of the
comparison group was assumed by the study authors to be equal to the value reported in
Eskenazi et al. (2004) (average  of 10.4 ppt) for the 20-40 age group.
       The average age of the women in the study was 24.8 years at the time of the incident, as
reported in the study text in the Materials and Methods section. The average age of the women
at conception in the exposed group was reported to be 28.2 years.  Two mean ages-at-conception
were evaluated by EPA: 30 and 45 years old. Serum levels were measured within one year of
the incident, and an LASC measurement lag time of 0.5 years was assumed. Modeling was
carried out for the exposure group that breastfed as detailed in Section F.3.1.1 using a scenario-
specific background intake modeled for an assumption of 10.4 ppt TCDD at age 30; the
background intake was assumed to be the same at age 45. Continuous daily TCDD intakes were
modeled to delivery (age at conception + 9 months) for both alternative ages-at-conception.
       As part of the sensitivity analysis, total TEQ intakes were estimated for the exposure
group that breastfed by assuming that total TEQ intake is equal to ten times the modeled TCDD
background intake.  The resulting background DLC-TEQ intake of 2.61 x 10 3 ng/kg-day
(2.9 x io~4 x 9) Was added to the modeled TCDD intakes to obtain the total TEQ intake
estimates.
       Table F-18. Model inputs derived from study details for Mocarelli et al.
       (2011)
Average age at event
(years)
24.8
Time lag between exposure
and LASC measurement
(years)
0.5
Time lag between
exposure and effect
(years)
5.2, 20.2
Target population
exposure window"
(years)
30.75, 45.75
"Age at delivery
F.3.7.2. Input for Exposure from Event to LASC Measurement
% EMOND HUMAN NON-GESTATION MODEL
% MODEL PARAMETERS
output @clear
                                        F-31

-------
prepare gclear T CBSNGKGLIADJ CBNGKG
% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON  = 217248.
EXP_TIME_OFF = 217249.
BCK_TIME_ON  =0.       %
BCK_TIME_OFF = 613200.  %
TIMELIMIT    = 221628.  %
MSTOTBCKGR   = 0.00029  %
% AGE AT EXPOSURE  (HOURS)
% AGE AT END OF EXPOSURE  (HOURS)
% BEGIN BACKGROUND EXPOSURE  (HOURS)
% END BACKGROUND EXPOSURE  (HOURS)
% AGE AT LASC MEASUREMENT  (HOURS)  (25.3 years)
  STUDY-SPECIFIC BACKGROUND  EXPOSURE (NG/KG-DAY)
% EVENT EXPOSURE DOSE  (NG/KG-DAY)
MSTOT = 6.37 % BREASTFEEDING GROUP (46.8 ppt TCDD measured)

% HUMAN VARIABLE PARAMETERS
MALE   = 0.
FEMALE = 1.
YO     = 0. % AGE AT BEGINNING OF SIMULATION
% POST-PROCESSING
start @nocallback
CBSNGKGLIADJ oneday=meani
   cbsngkgliadj(find( t==58524):length(  t)))
F.3.7.3. Input for Exposure from Event to the Study-Average Age at Delivery

% EMOND HUMAN GESTATION MODEL
% MODEL PARAMETERS
output @clear
prepare @clear T CBSNGKGLIADJ CBNGKG

% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
BCK_TIME_ON  = 0.
BCK_TIME_OFF = 613200.
EXP_TIME_ON  = 217248.
EXP_TIME_OFF = 217249.
CONCEPT      = 247032
TIMELIMIT    = 253602.
MSTOTBCKGR   = 0.00029
% BEGIN BACKGROUND EXPOSURE  (HOURS)
% END BACKGROUND EXPOSURE  (HOURS)
% AGE AT EXPOSURE  (HOURS)  (24.8  years)
% AGE AT END OF EXPOSURE  (HOURS)
% AGE AT CONCEPTION  (HOURS)  (28.2  years)
% AGE AT DELIVERY  (HOURS)  (28.95 years)
% MODELED BACKGROUND EXPOSURE  DOSE (NG/KG-DAY)
% EVENT EXPOSURE DOSE  (NG/KG-DAY)
MSTOT = 6.37 % BREASTFEEDING GROUP
MALE   = 0.
FEMALE = 1.
YO     = 0.
              AGE AT  BEGINNING OF SIMULATION
% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj(find(_t==EXP_TIME_ON):length(_t)));
meanCBSNGKGLIADJ
maxCBSNGKGLIADJ=max(_cbsngkgliadj);
maxCBSNGKGLIADJ
                                      F-32

-------
F.3.7.4. Input for Continuous Exposure until Age at Delivery for General Population
% EMOND HUMAN  GESTATION MODEL
% MODEL PARAMETERS
output @clear
prepare gclear T CBSNGKGLIADJ CBNGKG

% EXPOSURE  PARAMETERS
CINT  = 1
MAXT  =0.5
CONCEPT
EXP_TIME_ON
EXP_TIME_OFF

TIMELIMIT
MSTOTBCKGR
262801. % AGE  30  AT CONCEPTION  (HOURS)
394201. % AGE  45  AT CONCEPTION  (HOURS)
0.      % CONTINUOUS EXPOSURE BEGINS AT  BIRTH (AGE 0 HOURS)
262801. % AGE  30.75 AT DELIVERY  (HOURS)
394201. % AGE  45.75 AT DELIVERY  (HOURS)
EXP_TIME_OFF
0.      % BACKGROUND EXPOSURE INCLUDED IN MSTOT
% CONTINUOUS  EXPOSURE DOSE  (NG/KG-DAY)
MSTOT = 2.90E-4  % COMPARISON GROUP     - 10.4 PPT AT AGE  30  (BACKGROUND)
      % 1.50E-3  % BREASTFEEDING  GROUP  - 38.3 PPT AT AGE  30.75  AT DELIVERY
      % 1.04E-3  % BREASTFEEDING  GROUP  - 38.3 PPT AT AGE  45.75  AT DELIVERY

% HUMAN VARIABLE PARAMETERS
MALE   =  0.
FEMALE =  1.
YO     =  0.  % 0  YEARS OLD AT BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj);
maxCBSNGKGLIADJ=max(_cbsngkgliadj) ;
F.3.7.5. Mocarelli et al (2011) Results
      Table F-19.  Matching concentration at conception for the study population
      to chronic intake for the general population for Mocarelli et al. (2011)
Subject
modeled
Female
Female
Female
Female
Exposure
group
Comparison
Breastfed
General
population
age at
conception
30
45
30
45
TCDD only
Measured
LASC
(ng/kg)
10.4
46.8
Event
dose
(ng/kg)
Terminal
LASC at
conception
(ng/kg)
LASC at background
6.357
38.3
Continuous
intake matching
average LASC
(ng/kg-day)
2.90E-04
1.50E-03
1.04E-03
TEQa
Continuous intake
matching average
LASC (ng/kg-day)
2.90E-03
4.11E-03
3.65E-03
"TCDD average + DLC background intake (2.61 x 10 3 ng/kg-day).
                                       F-33

-------
F.3.8. Warner et al. (2004)
F.3.8.1. Summary of Modeling Approach
       Warner et al. (2004) studied age at onset of menarche in girls who were premenarcheal in
1976 at the time of first exposure.  Study authors divided exposure groups into quartiles, and
reported the exposures as ranges of measured TCDD LASC.  EPA determined that the highest
exposure group (4  quartile) was a NOAEL, so only the fourth quartile was evaluated for the
sensitivity analysis. For the fourth quartile, the lower bound of the exposure range was used as
the measured LASC estimate for estimating TCDD intakes.
       The average age of the subjects on July 10, 1976 was reported to be 6.9 years in the text
in the Results section. The critical susceptibility window for this endpoint could not be
determined; however, an assumed critical exposure window of 12.8 was established for modeling
purposes based on the age at menarche (12.8 ±1.6 years) reported by Warner et al. (2004).
Serum levels were measured within one year of the incident, therefore an LASC measurement
lag time of 0.5 years was assumed. Modeling was carried out as detailed in Section F.3.3.1.
Intakes were modeled with the Needham and Eskenazi background intakes as defined previously
(see Section F. 1.1 and F. 1.2).  Total TEQ was estimated by adding the background DLC intake
for the corresponding scenario to the calculated TCDD intakes as described in Section F.3.1.1.

       Table F-20. Model inputs derived from study details for Warner et al. (2004)
Average age at event
(years)
6.9
Time lag between exposure
and LASC measurement
(years)
0.5
Time lag between
exposure and effect
(years)
5.9
Assumed critical
exposure window
(years)
12.8
F.3.8.2. Input for Exposure from Event to LASC Measurement
% MODEL PARAMETERS
output @clear
prepare gclear T  CBSNGKGLIADJ  CBNGKG
% EXPOSURE PARAMETERS
MAXT  =0.5
CINT  = 1.
EXP_TIME_ON  = 60444.  % AGE AT EXPOSURE  (HOURS)
EXP_TIME_OFF = 60467.  % AGE AT END OF EXPOSURE  (HOURS)
DAY_CYCLE    =24.      % LENGTH OF DAY  (HOURS/DAY)
BCK_TIME_ON  =0.       % BACKGROUND EXPOSURE BEGINS AT BIRTH  (AGE 0 HOURS)
BCK_TIME_OFF = 613200.  % AGE AT END OF BACKGROUND EXPOSURE  (HOURS)
TIMELIMIT    = 64824.  % AGE AT LASC MEASUREMENT  (HOURS)
                                        F-34

-------
MSTOTBCKGR   = 0.00039  %  NEEDHAM BACKGROUND EXPOSURE DOSE  (NG/KG-DAY)
             % 0.00429  %  ESKENAZI BACKGROUND EXPOSURE DOSE  (NG/KG-DAY)

% EVENT EXPOSURE DOSE  (NG/KG-DAY)
MSTOT =
1ST QUARTILE
1ST QUARTILE
2ND QUARTILE
3RD QUARTILE
4TH QUARTILE
2ND QUARTILE
3RD QUARTILE
4TH QUARTILE
                             LOW -  NEEDHAM BACKGROUND
                             HIGH - NEEDHAM BACKGROUND
                             -  NEEDHAM BACKGROUND
                             -  NEEDHAM BACKGROUND
                             -  NEEDHAM BACKGROUND
                             -  ESKENAZI BACKGROUND
                             -  ESKENAZI BACKGROUND
                             -  ESKENAZI BACKGROUND
% HUMAN VARIABLE PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0. % AGE AT BEGINNING  OF SIMULATION
% POST-PROCESSING
start @nocallback
CBSNGKGLIADJ oneday=mean(  cbsngkgliadj(find(
                               t==64656):length(  t)))
F.3.8.3. Input for Exposure from Event to the End of the Assumed Critical Exposure Window
% MODEL PARAMETERS
output @clear
prepare @clear T CBSNGKGLIADJ CBNGKG
% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
MSTOTBCKGR
60444.  % AGE AT EXPOSURE  (HOURS)
60467.  % AGE AT END OF EXPOSURE  (HOURS)
24.     % LENGTH OF DAY  (HOURS/DAY)
0.       % BACKGROUND EXPOSURE  BEGINS AT  BIRTH (AGE 0 HOURS)
613200. % AGE AT END OF BACKGROUND  EXPOSURE (HOURS)
112128. % LENGTH OF ASSUMED CRITICAL EXPOSURE WINDOW (HOURS)
0.00039 % NEEDHAM BACKGROUND EXPOSURE  DOSE (NG/KG-DAY)
0.00429 % ESKENAZI BACKGROUND  EXPOSURE DOSE (NG/KG-DAY)
% EVENT EXPOSURE DOSE  (NG/KG-DAY)
MSTOT = 64.8 % 4TH QUARTILE  -  NEEDHAM BACKGROUND
      % 59.3 % 4TH QUARTILE  -  ESKENAZI BACKGROUND

% HUMAN VARIABLE PARAMETERS
MALE   = 0.
FEMALE = 1.
YO     = 0. % AGE AT BEGINNING OF  SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj(find(_t==EXP_TIME_ON):length(_t)));
meanCBSNGKGLIADJ
maxCBSNGKGLIADJ=max(_cbsngkgliadj);
maxCBSNGKGLIADJ
                                      F-35

-------
F.3.8.4. Input for Continuous Exposure over Assumed Critical Exposure Window
% MODEL PARAMETERS
output @clear
prepare gclear  T CBSNGKGLIADJ CBNGKG
% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON  =  0.
EXP_TIME_OFF =  112129.
DAY_CYCLE    =24.
BCK_TIME_ON  =  0.
BCK_TIME_OFF =  613200.
TIMELIMIT    =  112128.
MSTOTBCKGR   =  0.
% CONTINUOUS EXPOSURE BEGINS AT  BIRTH  (AGE 0 HOURS)
% LENGTH OF ASSUMED CRITICAL EXPOSURE  WINDOW (HOURS)
% LENGTH OF DAY  (HOURS/DAY)
% BACKGROUND EXPOSURE BEGINS AT  BIRTH  (AGE 0 HOURS)
% AGE AT END OF BACKGROUND EXPOSURE  (HOURS)
% LENGTH OF ASSUMED CRITICAL EXPOSURE  WINDOW (HOURS)
% /KG-DAYBACKGROUND EXPOSURE INCLUDED  IN MSTOT
% CONTINUOUS EXPOSURE  DOSE  (NG/KG-DAY)
MSTOT = 3.94E-2.  %  4TH QUARTILE - NEEDHAM  BACKGROUND
      % 4.24E-2.  %  4TH QUARTILE - ESKENAZI BACKGROUND
      % 6.04E-1   %  4TH QUARTILE - NEEDHAM  BACKGROUND
      % 5.17E-1   %  4TH QUARTILE - ESKENAZI BACKGROUND
                                  MATCHING MEAN
                                  MATCHING MEAN
                                  MATCHING PEAK
                                  MATCHING PEAK
% HUMAN VARIABLE PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0. % 0 YEARS OLD AT  BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj);
maxCBSNGKGLIADJ=max( cbsngkgliadj);
                                      F-36

-------
F.3.8.5. Warner et al (2004) Results
       Table F-21. Matching peak and average after pulse to chronic intake for
       Warner et al. (2004)









Subject
modeled










Quartile
TCDD only







Measured
LASC
(ng/kg)







Event
dose
(ng/kg)




Average
LASC
after
pulse
dose
(ng/kg)




Continuous
intake
matching
average
LASC
(ng/kg-day)




Peak
LASC
after
pulse
dose
(ng/kg)
Continuous
intake
matching peak
LASC/
measured
concentration
(if LASC
below
background)
(ng/kg-day)






Average of
continuous
intake rates
(ng/kg-day)
TEQa






Average of
continuous
intake rates
(ng/kg-day)
Needham background
Female
4*
300.0
64.8
207.2
3.94E-02
1896.6
6.04E-01
3.22E-01
3.25E-01
Eskenazi background
Female
4*
300.0
59.3
218.2
4.24E-02
1708.9
5.17E-01
2.80E-01
2.88E-01
"TCDD average + DLC background intake (Needham = 3.5 x 10 3 ng/kg-day; Eskenazi = 8.1 x 10 3 ng/kg-day).
F.3.9. Warner et al. (2007)
F.3.9.1. Summary of Modeling Approach
       Warner et al. (2007) examined ovarian function in women residents of Seveso, Italy
in 1996-1998, approximately 21 years after the incident. For analysis of ovulation status, the
study authors divided the exposure range into quartile groups (reported in Table 3 in the study
report).  EPA determined that the highest exposure group (4th quartile) was a NOAEL, so only
the fourth quartile was evaluated for the sensitivity analysis. For the fourth quartile, the lower
bound of the exposure group was used as the measured LASC estimate for estimating TCDD
intakes.
       Warner et al., (2007) reported the average age of women at the time of the interviews
(1996-1998) to be 31.3 years old in the text in the Results section. Because interviews took
place on average 21 years after the incident, average age at the time of the incident was estimated
to be 10 years old.  Serum values were collected within a year of the incident, and an LASC
measurement lag time of 0.5 years was assumed.  A critical susceptibility window for this
endpoint could not be determined. Because women are susceptible to ovarian function effects
until menopause, an assumed critical exposure window of 50 years was assigned as a
                                          F-37

-------
conservative estimate for the sensitivity analysis. Modeling was carried out as detailed in
Section F.3.1.1 using the Needham scenario background intake (see Section F. 1.1).
      As part of the sensitivity analysis, the intake when including DLCs was estimated by
adding the background DLC-TEQ intake to the modeled TCDD intake as described in
Section F. 1.1 using the Needham scenario female additive background DLC intake factor.
      Table F-22.  Model inputs derived from study details for Warner et al. (2007)
Average age at event
(years)
10
Time lag between exposure
and LASC measurement
(years)
0.5
Time lag between
exposure and effect
(years)
21
Assumed critical
exposure window
(years)
50
F.3.9.2.  Input for Exposure from Event to LASC Measurement

% MODEL  PARAMETERS
output  @clear
prepare  @clear T CBSNGKGLIADJ  CBNGKG

% EXPOSURE  PARAMETERS
MAXT =0.5
CINT =  1.
EXP_TIME_ON  = 87600.
EXP_TIME_OFF  = 87623.
DAY_CYCLE     =24.
BCK_TIME_ON  = 0.
BCK_TIME_OFF  = 613200.
TIMELIMIT     = 91980.
MSTOTBCKGR
                        % AGE AT  EXPOSURE  (HOURS)
                        % AGE AT  END OF EXPOSURE  (HOURS)
                        % LENGTH  OF DAY (HOURS/DAY)
                        % BACKGROUND EXPOSURE BEGINS  AT BIRTH  (AGE  0  HOURS)
                        % AGE AT  END OF BACKGROUND  EXPOSURE  (HOURS)
                        % AGE AT  LASC MEASUREMENT  (HOURS)
              = 0.00039 % NEEDHAM BACKGROUND EXPOSURE DOSE (NG/KG-DAY)

% EVENT  EXPOSURE DOSE  (NG/KG-DAY)
MSTOT =0.1    % 1ST QUARTILE
      %  3.7    % 2ND QUARTILE
      %  127.8 % 3RD QUARTILE
      %  212.0 % 4TH QUARTILE

% HUMAN  VARIABLE PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0.  % AGE AT BEGINNING  OF SIMULATION

% POST-PROCESSING
start @nocallback
CBSNGKGLIADJ oneday=mean(  cbsngkgliadj(find( t==91812):length( t)))
                                       F-38

-------
F.3.9.3. Input for Exposure from Event to the End of the Assumed Critical Exposure Window
% MODEL PARAMETERS
output @clear
prepare gclear T  CBSNGKGLIADJ CBNGKG

% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON
EXP_TIME_OFF
DAY_CYCLE
BCK_TIME_ON
BCK_TIME_OFF
TIMELIMIT
MSTOTBCKGR
87600.  % AGE AT EXPOSURE  (HOURS)
87623.  % AGE AT END OF  EXPOSURE  (HOURS)
24.     % LENGTH OF DAY  (HOURS/DAY)
0.      % BACKGROUND EXPOSURE  BEGINS AT BIRTH (AGE 0 HOURS)
613200. % AGE AT END OF  BACKGROUND  EXPOSURE (HOURS)
438000. % LENGTH OF ASSUMED  CRITICAL EXPOSURE WINDOW  (HOURS)
0.00039 % NEEDHAM BACKGROUND EXPOSURE DOSE (NG/KG-DAY)
% EVENT EXPOSURE  DOSE  (NG/KG-DAY)
MSTOT = 212.0 % 4TH  QUARTILE

% HUMAN VARIABLE  PARAMETERS
MALE   = 1.
FEMALE = 0.
YO     = 0. % AGE AT BEGINNING OF SIMULATION

% POST-PROCESSING
start @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj(find(_t==EXP_TIME_ON):length(_t)));
meanCBSNGKGLIADJ
maxCBSNGKGLIADJ=max(_cbsngkgliadj);
maxCBSNGKGLIADJ
F.3.9.4. Input for Continuous Exposure over Assumed Critical Exposure Window
% MODEL PARAMETERS
output @clear
prepare @clear  T  CBSNGKGLIADJ CBNGKG
% EXPOSURE PARAMETERS
MAXT =0.5
CINT = 1.
EXP_TIME_ON  =  0.
EXP_TIME_OFF =  438001.
DAYJCYCLE    =24.
BCK_TIME_ON  =  0.
BCK_TIME_OFF =  613200.
TIMELIMIT    =  438000.
MSTOTBCKGR   =  0.
        % CONTINUOUS EXPOSURE  BEGINS AT BIRTH (AGE 0 HOURS)
        % LENGTH OF ASSUMED  CRITICAL EXPOSURE WINDOW  (HOURS)
        % LENGTH OF DAY  (HOURS/DAY)
        % BACKGROUND EXPOSURE  BEGINS AT BIRTH (AGE 0 HOURS)
        % AGE AT END OF  BACKGROUND EXPOSURE (HOURS)
        % LENGTH OF ASSUMED  CRITICAL EXPOSURE WINDOW  (HOURS)
        % /KG-DAYBACKGROUND  EXPOSURE INCLUDED IN MSTOT
% CONTINUOUS EXPOSURE  DOSE (NG/KG-DAY)
MSTOT = 3.00E-3  %  4TH  QUARTILE - MATCHING MEAN
      % 2.04E-1  %  4TH  QUARTILE - MATCHING PEAK

% HUMAN VARIABLE PARAMETERS
MALE   = 0.
FEMALE = 1.
                                      F-39

-------
YO
= 0. %  0  YEARS OLD AT  BEGINNING OF  SIMULATION
% POST-PROCESSING
start  @nocallback
meanCBSNGKGLIADJ=mean(_cbsngkgliadj);
maxCBSNGKGLIADJ=max( cbsngkgliadj);
F.3.9.5. Warner et al (2007) Results
       Table F-23. Matching peak and average after pulse to chronic intake for
       Warner et al. (2007)
Subject
modeled
Quartile
TCDD only
Measured
LASC
(ng/kg)
Event
dose
(ng/kg)
Average
LASC
after
pulse
dose
(ng/kg)
Continuous
intake
matching
average
LASC
(ng/kg-day)
Peak
LASC
after
pulse
dose
(ng/kg)
Continuous
intake
matching
peak LASC
(ng/kg-day)
Average of
continuous
intake rates
(ng/kg-day)
TEQa
Average of
continuous
intake rates
(ng/kg-day)
Needham background
Female
4*
212.0
39.4
56.3
3.00E-03
1229.7
2.04E-01
1.04E-01
1.07E-01
"TCDD average + DLC background intake (Needham =3.5 x 10 3 ng/kg-day).
                                        F-40

-------
F.4.  REFERENCES

Alaluusua, S: Calderara, P; Gerthoux, PM: Lukinmaa, PL; Kovero, O: Needham, L: Patterson Jr,
       DG: Tuomisto, J: Mocarelli, P. (2004). Developmental dental aberrations after the dioxin
       accident in Seveso. Environ Health Perspect 112: 1313-1318.
Baccarelli, A; Giacomini, SM: Corbetta, C: Landi, MT: Bonzini, M: Consonni, D: Grillo, P;
       Patterson, DG: Pesatori, AC: Bertazzi, PA. (2008). Neonatal thyroid function in Seveso
       25 years after maternal exposure to dioxin. PLoS Med 5: e!61.
Eskenazi, B: Warner, M: Mocarelli, P: Samuels, S: Needham, LL: Patterson, DG, Jr: Lippman,
       S: Vercellini, P: Gerthoux. PM: Brambilla, P: Olive. D.  (2002). Serum dioxin
       concentrations and menstrual cycle characteristics. Am J Epidemiol 156: 383-392.
Eskenazi, B: Mocarelli, P: Warner, M: Needham, L: Patterson, DG, Jr: Samuels,  S: Turner, W;
       Gerthoux, PM: Brambilla, P. (2004). Relationship of serum TCDD concentrations and
       age at exposure of female residents of Seveso, Italy.  Environ Health Perspect 112: 22-27.
       http://dx.doi.org/10.1289/ehp.6573.
Eskenazi, B: Warner, M: Marks, AR: Samuels, S: Gerthoux, PM: Vercellini, P: Olive, PL;
       Needham, L: Patterson, D, Jr: Mocarelli, P. (2005). Serum dioxin concentrations and age
       at menopause. Environ Health Perspect 113: 858-862.
Mocarelli, P: Needham, LL: Marocchi, A; Patterson, DG, Jr: Brambilla, P: Gerthoux, PM:
       Meazza, L: Carreri, V. (1991).  Serum concentrations of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin and test results from selected residents of Seveso, Italy. J Toxicol Environ  Health
       A 32: 357-366. http://dx.doi.org/10.1080/15287399109531490.
Mocarelli, P: Gerthoux, PM: Ferrari, E: Patterson Jr, DG: Kieszak, SM: Brambilla, P: Vincoli,
       N; Signorini, S: Tramacere, P: Carreri, V: Sampson, EJ: Turner, WE. (2000). Paternal
       concentrations of dioxin and sex ratio of offspring. Lancet 355: 1858-1863.
       http://dx.doi.org/10.1016/S0140-6736(00)02290-X.
Mocarelli, P: Gerthoux, PM: Patterson, DG, Jr: Milani, S: Limonata,  G: Bertona, M: Signorini,
       S: Tramacere, P: Colombo, L: Crespi, C: Brambilla, P: Sarto, C: Carreri, V: Sampson,
       EJ: Turner, WE; Needham, LL. (2008). Dioxin exposure, from infancy through puberty,
       produces endocrine disruption and affects human semen quality. Environ Health Perspect
       116: 70-77. http://dx.doi.org/10.1289/ehp.10399.
Mocarelli, P, Gerthoux, P. M., Needham, L. L., Patterson, D. G., Jr Limonta, G., Falbo, R.,
       Signorini, S., Bertona, M., Crespi, C., Sarto, C., Scott, P. K., Turner, W. E., Brambilla, P.
       (2011). Perinatal exposure to low doses of dioxin can permanently impair human  semen
       quality. Environ Health Perspect 119: 713-718.
Needham, LL: Gerthoux, PM: Patterson, DG: Brambilla, P: Turner, WE: Beretta, C: Pirkle, JL:
       Colombo, L: Sampson, EJ:  Tramacere, PL; Signorini, S: Meazza, L: Carreri, V: Jackson,
       RJ; Mocarelli, P. (1998). Serum dioxin levies in Seveso, Italy, population in 1976.
       Teratog Carcinog Mutagen 17: 225-240.
Warner, M: Samuels, S: Mocarelli, P: Gerthoux, PM: Needham, L: Patterson, DG, Jr: Eskenazi,
       R (2004). Serum dioxin concentrations and age at menarche. Environ Health Perspect
       112: 1289-1292. http://dx.doi.org/10.1289/ehp.7004.
Warner, M: Eskenazi, B: Olive, PL; Samuels, S: Quick-Miles, S: Vercellini, P: Gerthoux, PM:
       Needham, L: Patterson, DG, Jr: Mocarelli, P. (2007). Serum dioxin concentrations and
       quality of ovarian function in women of seveso. Environ Health Perspect  115: 336-340.
       http://dx.doi.org/10.1289/ehp.9667.

                                         F-41

-------
                                         EPA/600/R-10/038F
                                          www.epa.gov/iris
               APPENDIX G
Noncancer Benchmark Dose Modeling
                   January 2012
          National Center for Environmental Assessment
             Office of Research and Development
             U.S. Environmental Protection Agency
                    Cincinnati, OH

-------
          CONTENTS—APPENDIX G: Noncancer Benchmark Dose Modeling
APPENDIX G.   NONCANCER BENCHMARK DOSE MODELING	G-l
     G.I. BENCHMARK DOSE SOFTWARE (BMDS) INPUT TABLES	G-l
         G.I.I.  Amin et al. (2000)	G-l
         G.1.2.  Bell et al. (2007)	G-l
         G.I.3.  Cantonietal. (1981)	G-2
         G.1.4.  Crofton et al. (2005)	G-2
         G.I.5.  DeCaprio et  al. (1986)	G-3
         G.I.6.  Franc et al. (2001)	G-4
         G.I.7.  Hojoetal. (2002)	G-4
         G.I.8.  Kattainen et  al. (2001)	G-5
         G.I.9.  Keller et al. (2008a; 2008b; 2007)	G-5
         G.I.10. Kocibaetal. (1978)	G-6
         G.I.11. Kuchiiwa et  al. (2002)	G-6
         G.I.12. Latchoumycandane and Mathur (2002)	G-7
         G.I.13. Lietal. (1997)	G-7
         G.I.14. Lietal. (2006)	G-8
         G.I.15. Markowski et al. (2001)	G-8
         G.I.16. Miettinen et  al. (2006)	G-9
         G.I.17. National Toxicology Program (1982)	G-9
         G.I.18. National Toxicology Program (2006)	G-10
         G.I.19. Ohsakoetal. (2001)	G-ll
         G.1.20. Sewalletal.  (1995)	G-ll
         G.1.21. Shi et al. (2007)	G-12
         G.1.22. Smialowicz et al. (2008)	G-12
         G.1.23. Smith etal. (1976)	G-13
         G.1.24. Sparschu etal. (1971)	G-13
         G.1.25. Toth etal.  (1979)	G-14
         G.1.26. van Birgelen et al. (1995)	G-14
         G.1.27. White etal. (1986)	G-15
     G.2. ALTERNATE DOSE: WHOLE BLOOD BMDS RESULTS	G-15
         G.2.1.  Amin etal. (2000): 0.25% Saccharin Consumed, Female	G-15
                G.2.1.1.  Summary Table of BMDS Modeling Results	G-15
                G.2.1.2.  Output for Selected Model: Linear	G-15
                G.2.1.3.  Figure for Selected Model: Linear	G-19
                G.2.1.4.  Output for Additional Model Presented: Power, Unrestricted ...G-19
                G.2.1.5.  Figure for Additional Model Presented: Power, Unrestricted.... G-23
         G.2.2.  Amin et al. (2000): 0.25% Saccharin Preference Ratio, Female	G-23
                G.2.2.1.  Summary Table of BMDS Modeling Results	G-23
                G.2.2.2.  Output for Selected Model: Linear	G-24
                G.2.2.3.  Figure for Selected Model: Linear	G-27
         G.2.3.  Amin et al. (2000): 0.50% Saccharin Consumed, Female	G-27
                G.2.3.1.  Summary Table of BMDS Modeling Results	G-27
                G.2.3.2.  Output for Selected Model: Linear	G-28

                                         G-ii

-------
                       CONTENTS (continued)
        G.2.3.3.  Figure for Selected Model: Linear	G-31
        G.2.3.4.  Output for Additional Model Presented: Power, Unrestricted... G-31
        G.2.3.5.  Figure for Additional Model Presented: Power, Unrestricted.... G-3 5
G.2.4.   Amin et al. (2000): 0.50% Saccharin Preference Ratio, Female	G-35
        G.2.4.1.  Summary Table of BMDS Modeling Results	G-35
        G.2.4.2.  Output for Selected Model: Linear	G-36
        G.2.4.3.  Figure for Selected Model: Linear	G-39
        G.2.4.4.  Output for Additional Model Presented: Power, Unrestricted... G-39
        G.2.4.5.  Figure for Additional Model Presented: Power, Unrestricted.... G-43
G.2.5.   Bell et al. (2007): Balano-Preputial Separation, Postnatal Day
        (PND)49	G-44
        G.2.5.1.  Summary Table of BMDS Modeling Results	G-44
        G.2.5.2.  Output for Selected Model: Log-Logistic	G-44
        G.2.5.3.  Figure for Selected Model: Log-Logistic	G-47
        G.2.5.4.  Output for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-47
        G.2.5.5.  Figure for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-50
G.2.6.   Cantoni et al. (1981): Urinary Coproporhyrins, 3 Months	G-51
        G.2.6.1.  Summary Table of BMDS Modeling Results	G-51
        G.2.6.2.  Output for Selected Model: Exponential (M4)	G-51
        G.2.6.3.  Figure for Selected Model: Exponential (M4)	G-55
        G.2.6.4.  Output for Additional Model Presented: Power, Unrestricted... G-55
        G.2.6.5.  Figure for Additional Model Presented: Power, Unrestricted.... G-59
G.2.7.   Cantoni et al. (1981): Urinary Porphyrins	G-60
        G.2.7.1.  Summary Table of BMDS Modeling Results	G-60
        G.2.7.2.  Output for Selected Model: Exponential (M2)	G-60
        G.2.7.3.  Figure for Selected Model: Exponential (M2)	G-63
G.2.8.   Crofton et al. (2005): Serum, T4	G-64
        G.2.8.1.  Summary Table of BMDS Modeling Results	G-64
        G.2.8.2.  Output for Selected Model: Exponential (M4)	G-64
        G.2.8.3.  Figure for Selected Model: Exponential (M4)	G-68
G.2.9.   Franc et al. (2001): Sprague-Dawley (S-D) Rats, Relative Liver
        Weight	G-69
        G.2.9.1.  Summary Table of BMDS Modeling Results	G-69
        G.2.9.2.  Output for Selected Model: Power	G-69
        G.2.9.3.  Figure for Selected Model: Power	G-72
G.2.10.  Franc et al. (2001): Long-Evans (L-E) Rats, Relative Liver Weight	G-73
        G.2.10.1.  Summary Table of BMDS Modeling Results	G-73
        G.2.10.2.  Output for Selected Model: Hill	G-73
        G.2.10.3.  Figure for Selected Model: Hill	G-77
        G.2.10.4.  Output for Additional Model Presented: Hill, Unrestricted	G-77
        G.2.10.5.  Figure for Additional Model Presented: Hill, Unrestricted	G-81

                                 G-iii

-------
                      CONTENTS (continued)
G.2.11. Franc et al. (2001): S-D Rats, Relative Thymus Weight	G-82
       G.2.11.1.  Summary Table of BMDS Modeling Results	G-82
       G.2.11.2.  Output for Selected Model: Exponential (M4)	G-82
       G.2.11.3.  Figure for Selected Model: Exponential (M4)	G-86
       G.2.11.4.  Output for Additional Model Presented: Polynomial, 3-
                 degree	G-86
       G.2.11.5.  Figure for Additional Model Presented: Polynomial, 3-
                 degree	G-90
G.2.12. Franc et al. (2001): Long-Evans (L-E) Rats, Relative Thymus Weight	G-91
       G.2.12.1.  Summary Table of BMDS Modeling Results	G-91
       G.2.12.2.  Output for Selected Model: Exponential (M4)	G-91
       G.2.12.3.  Figure for Selected Model: Exponential (M4)	G-95
G.2.13. Franc et al. (2001): Han/Wistar (H/W) Rats, Relative Thymus Weight	G-96
       G.2.13.1.  Summary Table of BMDS Modeling Results	G-96
       G.2.13.2.  Output for Selected Model: Exponential (M2)	G-96
       G.2.13.3.  Figure for Selected Model: Exponential (M2)	G-100
G.2.14. Hojoetal. (2002): DRL Reinforce per Minute	G-101
       G.2.14.1.  Summary Table of BMDS Modeling Results	G-101
       G.2.14.2.  Output for Selected Model: Exponential (M4)	G-101
       G.2.14.3.  Figure for Selected Model: Exponential (M4)	G-105
G.2.15. Hojoetal. (2002): DRL Response per Minute	G-106
       G.2.15.1.  Summary Table of BMDS Modeling Results	G-106
       G.2.15.2.  Output for Selected Model: Exponential (M4)	G-106
       G.2.15.3.  Figure for Selected Model: Exponential (M4)	G-l 10
G.2.16. Kattainen et al. (2001): 3rd Molar Eruption, Female	G-lll
       G.2.16.1.  Summary Table of BMDS Modeling Results	G-lll
       G.2.16.2.  Output for Selected Model: Log-Logistic	G-lll
       G.2.16.3.  Figure for Selected Model: Log-Logistic	G-l 14
       G.2.16.4.  Output for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-l 14
       G.2.16.5.  Figure for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-l 17
G.2.17. Kattainen et al. (2001): 3rd Molar Length, Female	G-l 18
       G.2.17.1.  Summary Table of BMDS Modeling Results	G-118
       G.2.17.2.  Output for Selected Model: Hill	G-118
       G.2.17.3.  Figure for Selected Model: Hill	G-122
       G.2.17.4.  Output for Additional Model Presented: Hill, Unrestricted	G-122
       G.2.17.5.  Figure for Additional Model Presented: Hill, Unrestricted	G-126
G.2.18. Keller et al. (2007):  Missing Mandibular Molars, CBA J	G-127
       G.2.18.1.  Summary Table of BMDS Modeling Results	G-127
       G.2.18.2.  Output for Selected Model: Multistage, 1-Degree	G-127
       G.2.18.3.  Figure for Selected Model: Multistage, 1-Degree	G-130
G.2.19. Kocibaetal. (1978): Urinary Coproporphyrin, Females	G-130

                                G-iv

-------
                      CONTENTS (continued)
        G.2.19.1. Summary Table of BMDS Modeling Results	G-130
        G.2.19.2. Output for Selected Model: Exponential (M4)	G-131
        G.2.19.3. Figure for Selected Model: Exponential (M4)	G-134
G.2.20.  Kociba et al. (1978): Uroporphyrin per Creatinine, Female	G-134
        G.2.20.1. Summary Table of BMDS Modeling Results	G-134
        G.2.20.2. Output for Selected Model: Linear	G-135
        G.2.20.3. Figure for Selected Model: Linear	G-138
G.2.21.  Kuchiiwa et al. (2002): Immunoreactive Neurons in Dorsalis, Males	G-138
        G.2.21.1. Summary Table of BMDS Modeling Results	G-138
        G.2.21.2. Output for Selected Model: Linear	G-139
        G.2.21.3. Figure for Selected Model: Linear	G-142
G.2.22.  Kuchiiwa et al. (2002): Immunoreactive Neurons in Medianus, Males... G-142
        G.2.22.1. Summary Table of BMDS Modeling Results	G-142
        G.2.22.2. Output for Selected Model: Linear	G-142
        G.2.22.3. Figure for Selected Model: Linear	G-146
G.2.23.  Kuchiiwa et al. (2002): Immunoreactive Neurons in B9, Males	G-146
        G.2.23.1. Summary Table of BMDS Modeling Results	G-146
        G.2.23.2. Output for Selected Model: Linear	G-146
        G.2.23.3. Figure for Selected Model: Linear	G-150
G.2.24.  Kuchiiwa et al. (2002): Immunoreactive Neurons in Magnus, Males	G-150
        G.2.24.1. Summary Table of BMDS Modeling Results	G-150
        G.2.24.2. Output for Selected Model: Linear	G-150
        G.2.24.3. Figure for Selected Model: Linear	G-154
G.2.25.  Latchoumycandane and Mathur (2002): Sperm Production	G-155
        G.2.25.1. Summary Table of BMDS Modeling Results	G-155
        G.2.25.2. Output for Selected Model: Hill	G-155
        G.2.25.3. Figure for Selected Model: Hill	G-159
        G.2.25.4. Output for Additional Model Presented: Hill, Unrestricted	G-159
        G.2.25.5. Figure for Additional Model Presented: Hill, Unrestricted	G-163
G.2.26.  Li et al. (1997): Follicle-Stimulating Hormone (FSH)	G-164
        G.2.26.1. Summary Table of BMDS Modeling Results	G-164
        G.2.26.2. Output for Selected Model: Power	G-164
        G.2.26.3. Figure for Selected Model: Power	G-168
        G.2.26.4. Output for Additional Model Presented: Power, Unrestricted. G-168
        G.2.26.5. Figure for Additional Model Presented: Power, Unrestricted.. G-172
G.2.21.  Lietal. (2006): Estradiol, 3-Day	G-173
        G.2.21.1. Summary Table of BMDS Modeling Results	G-173
        G.2.27.2. Output for Selected Model: Linear	G-173
        G.2.27.3. Figure for Selected Model: Linear	G-176
G.2.28.  Li et al. (2006): Progesterone, 3-Day	G-177
        G.2.28.1. Summary Table of BMDS Modeling Results	G-177
        G.2.28.2. Output for Selected Model: Hill	G-177
        G.2.28.3. Figure for Selected Model: Hill	G-181

                                 G-v

-------
                       CONTENTS (continued)
G.2.29. Markowski et al. (2001): FR10 Run Opportunities	G-182
       G.2.29.1. Summary Table of BMDS Modeling Results	G-182
       G.2.29.2. Output for Selected Model: Exponential (M2)	G-182
       G.2.29.3. Figure for Selected Model: Exponential (M2)	G-186
G.2.30. Markowski et al. (2001): FR2 Revolutions	G-187
       G.2.30.1. Summary Table of BMDS Modeling Results	G-187
       G.2.30.2. Output for Selected Model: Hill	G-187
       G.2.30.3. Figure for Selected Model: Hill	G-191
       G.2.30.4. Output for Additional Model Presented: Power, Unrestricted. G-191
       G.2.30.5. Figure for Additional Model Presented: Power, Unrestricted.. G-195
G.2.31. Markowski et al. (2001): FR5 Run Opportunities	G-196
       G.2.31.1. Summary Table of BMDS Modeling Results	G-196
       G.2.31.2. Output for Selected Model: Hill	G-196
       G.2.31.3. Figure for Selected Model: Hill	G-200
       G.2.31.4. Output for Additional Model Presented: Power, Unrestricted. G-200
       G.2.31.5. Figure for Additional Model Presented: Power, Unrestricted.. G-204
G.2.32. Miettinen et al. (2006): Cariogenic Lesions, Pups	G-205
       G.2.32.1. Summary Table of BMDS Modeling Results	G-205
       G.2.32.2. Output for Selected Model: Log-Logistic	G-205
       G.2.32.3. Figure for Selected Model: Log-Logistic	G-208
       G.2.32.4. Output for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-208
       G.2.32.5. Figure for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-211
G.2.33. Murray et al. (1979): Fertility in F2 Generation	G-212
       G.2.33.1. Summary Table of BMDS Modeling Results	G-212
       G.2.33.2. Output for Selected Model: Multistage, 2-Degree	G-212
       G.2.33.3. Figure for Selected Model: Multistage, 2-Degree	G-215
G.2.34. National Toxicology Program (1982): Toxic Hepatitis, Male Mice	G-215
       G.2.34.1. Summary Table of BMDS Modeling Results	G-215
       G.2.34.2. Output for Selected Model: Multistage, 3-Degree	G-216
       G.2.34.3. Figure for Selected Model: Multistage, 3-Degree	G-218
G.2.35. National Toxicology Program (2006): Alveolar Metaplasia	G-219
       G.2.35.1. Summary Table of BMDS Modeling Results	G-219
       G.2.35.2. Output for Selected Model: Log-Logistic	G-219
       G.2.35.3. Figure for Selected Model: Log-Logistic	G-221
G.2.36. National Toxicology Program (2006): Eosinophilic Focus,  Liver	G-222
       G.2.36.1. Summary Table of BMDS Modeling Results	G-222
       G.2.36.2. Output for Selected Model: Probit	G-222
       G.2.36.3. Figure for Selected Model: Probit	G-224
G.2.37. National Toxicology Program (2006): Fatty Change Diffuse, Liver	G-225
       G.2.37.1. Summary Table of BMDS Modeling Results	G-225
       G.2.37.2. Output for Selected Model: Weibull	G-225

                                G-vi

-------
                       CONTENTS (continued)
        G.2.37.3. Figure for Selected Model:  Weibull	G-227
G.2.38.  National Toxicology Program (2006): Gingival Hyperplasia,
        Squamous, 2 Years	G-228
        G.2.38.1. Summary Table of BMDS Modeling Results	G-228
        G.2.38.2. Output for Selected Model: Log-Logistic	G-228
        G.2.38.3. Figure for Selected Model:  Log-Logistic	G-231
        G.2.38.4. Output for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-231
        G.2.38.5. Figure for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-234
G.2.39.  National Toxicology Program (2006): Hepatocyte Hypertrophy,
        2 Years	G-235
        G.2.39.1. Summary Table of BMDS Modeling Results	G-235
        G.2.39.2. Output for Selected Model: Multistage, 5-Degree	G-235
        G.2.39.3. Figure for Selected Model:  Multistage, 5-Degree	G-238
G.2.40.  National Toxicology Program (2006): Necrosis, Liver	G-239
        G.2.40.1. Summary Table of BMDS Modeling Results	G-239
        G.2.40.2. Output for Selected Model: Log-Probit, Unrestricted	G-239
        G.2.40.3. Figure for Selected Model:  Log-Probit, Unrestricted	G-241
G.2.41.  National Toxicology Program (2006): Oval Cell Hyperplasia	G-242
        G.2.41.1. Summary Table of BMDS Modeling Results	G-242
        G.2.41.2. Output for Selected Model: Probit	G-242
        G.2.41.3. Figure for Selected Model:  Probit	G-244
        G.2.41.4. Output for Additional Model Presented: Weibull	G-245
        G.2.41.5. Figure for Additional Model Presented: Weibull	G-247
G.2.42.  National Toxicology Program (2006): Pigmentation, Liver	G-247
        G.2.42.1. Summary Table of BMDS Modeling Results	G-247
        G.2.42.2. Output for Selected Model: Log-Probit	G-248
        G.2.42.3. Figure for Selected Model:  Log-Probit	G-250
G.2.43.  National Toxicology Program (2006): Toxic Hepatopathy	G-250
        G.2.43.1. Summary Table of BMDS Modeling Results	G-250
        G.2.43.2. Output for Selected Model: Multistage, 5-Degree	G-251
        G.2.43.3. Figure for Selected Model:  Multistage, 5-Degree	G-253
G.2.44.  Ohsako et al. (2001): Ano-Genital Length, PND  120	G-254
        G.2.44.1. Summary Table of BMDS Modeling Results	G-254
        G.2.44.2. Output for Selected Model: Hill	G-254
        G.2.44.3. Figure for Selected Model:  Hill	G-258
        G.2.44.4. Output for Additional Model Presented: Hill, Unrestricted	G-258
        G.2.44.5. Figure for Additional Model Presented: Hill, Unrestricted	G-262
G.2.45.  Sewall et al. (1995): T4 In Serum	G-263
        G.2.45.1. Summary Table of BMDS Modeling Results	G-263
        G.2.45.2. Output for Selected Model: Hill	G-263
        G.2.45.3. Figure for Selected Model:  Hill	G-267

                                G-vii

-------
                      CONTENTS (continued)
        G.2.45.4. Output for Additional Model Presented: Hill, Unrestricted	G-267
        G.2.45.5. Figure for Additional Model Presented: Hill, Unrestricted	G-271
G.2.46.  Shi et al. (2007): Estradiol 17B, PE9	G-272
        G.2.46.1. Summary Table of BMDS Modeling Results	G-272
        G.2.46.2. Output for Selected Model: Exponential (M4)	G-272
        G.2.46.3. Figure for Selected Model: Exponential (M4)	G-276
G.2.47.  Smialowicz et al. (2008): PFC per 106 Cells	G-277
        G.2.47.1. Summary Table of BMDS Modeling Results	G-277
        G.2.47.2. Output for Selected Model: Power, Unrestricted	G-277
        G.2.47.3. Figure for Selected Model: Power, Unrestricted	G-281
G.2.48.  Smialowicz et al. (2008): PFC per Spleen	G-282
        G.2.48.1. Summary Table of BMDS Modeling Results	G-282
        G.2.48.2. Output for Selected Model: Power, Unrestricted	G-282
        G.2.48.3. Figure for Selected Model: Power, Unrestricted	G-285
G.2.49.  Smith etal. (1976): Cleft Palate in Pups	G-286
        G.2.49.1. Summary Table of BMDS Modeling Results	G-286
        G.2.49.2. Output for Selected Model: Log-Logistic	G-286
        G.2.49.3. Figure for Selected Model: Log-Logistic	G-288
G.2.50.  Sparschu et al. (1976): Fetal Body Weight, Male	G-289
        G.2.50.1. Summary Table of BMDS Modeling Results	G-289
        G.2.50.2. Output for Selected Model: exponential (M5)	G-289
        G.2.50.3. Figure for Selected Model: Exponential (M5)	G-292
G.2.51.  Sparschu et al. (1971): Fetal Body Weight, Female	G-293
        G.2.51.1. Summary Table of BMDS Modeling Results	G-293
        G.2.51.2. Output for Selected Model: Exponential (M2)	G-293
        G.2.51.3. Figure for Selected Model: Exponential (M2)	G-296
G.2.52.  Toth et al. (1979): Amyloidosis	G-297
        G.2.52.1. Summary Table of BMDS Modeling Results	G-297
        G.2.52.2. Output for Selected Model: Log-Logistic	G-297
        G.2.52.3. Figure for Selected Model: Log-Logistic	G-300
        G.2.52.4. Output for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-300
        G.2.52.5. Figure for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-303
G.2.53.  Toth etal. (1979): Skin Lesions	G-304
        G.2.53.1. Summary Table of BMDS Modeling Results	G-304
        G.2.53.2. Output for Selected Model: Log-Logistic	G-304
        G.2.53.3. Figure for Selected Model: Log-Logistic	G-307
        G.2.53.4. Output for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-307
        G.2.53.5. Figure for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-310
G.2.54.  van Birgelen et al. (1995): Hepatic Retinol	G-311

                                G-viii

-------
                           CONTENTS (continued)
            G.2.54.1. Summary Table of BMDS Modeling Results	G-311
            G.2.54.2. Output for Selected Model: Exponential (M4)	G-311
            G.2.54.3. Figure for Selected Model: Exponential (M4)	G-315
            G.2.54.4. Output for Additional Model Presented: Power, Unrestricted. G-315
            G.2.54.5. Figure for Additional Model Presented: Power, Unrestricted.. G-319
    G.2.55. van Birgelen et al. (1995): Hepatic Retinol Palmitate	G-320
            G.2.55.1. Summary Table of BMDS Modeling Results	G-320
            G.2.55.2. Output for Selected Model: Exponential (M4)	G-320
            G.2.55.3. Figure for Selected Model: Exponential (M4)	G-324
            G.2.55.4. Output for Additional Model Presented: Power, Unrestricted. G-324
            G.2.55.5. Figure for Additional Model Presented: Power, Unrestricted.. G-328
    G.2.56. White et al. (1986): CH50	G-329
            G.2.56.1. Summary Table of BMDS Modeling Results	G-329
            G.2.56.2. Output for Selected Model: Hill	G-329
            G.2.56.3. Figure for Selected Model: Hill	G-333
            G.2.56.4. Output for Additional Model Presented: Hill, Unrestricted	G-333
            G.2.56.5. Figure for Additional Model Presented: Hill, Unrestricted	G-337
G.3. ADMINISTERED DOSE: BMDS RESULTS	G-337
    G.3.1.   Aminetal. (2000): 0.25% Saccharin Consumed, Female	G-337
            G.3.1.1.   Summary Table of BMDS Modeling Results	G-337
            G.3.1.2.   Output for Selected Model: Linear	G-338
            G.3.1.3.   Figure for Selected Model: Linear	G-341
            G.3.1.4.   Output for Additional Model Presented: Power, Unrestricted. G-341
            G.3.1.5.   Figure for Additional Model Presented: Power, Unrestricted.. G-345
    G.3.2.   Amin et al. (2000): 0.25% Saccharin Preference Ratio, Female	G-345
            G.3.2.1.   Summary Table of BMDS Modeling Results	G-345
            G.3.2.2.   Output for Selected Model: Linear	G-346
            G.3.2.3.   Figure for Selected Model: Linear	G-349
    G.3.3.   Amin et al. (2000): 0.50% Saccharin Consumed, Female	G-349
            G.3.3.1.   Summary Table of BMDS Modeling Results	G-349
            G.3.3.2.   Output for Selected Model: Linear	G-350
            G.3.3.3.   Figure for Selected Model: Linear	G-353
            G.3.3.4.   Output for Additional Model Presented: Power, Unrestricted. G-353
            G.3.3.5.   Figure for Additional Model Presented: Power, Unrestricted.. G-357
    G.3.4.   Amin et al. (2000): 0.50% Saccharin Preference Ratio, Female	G-357
            G.3.4.1.   Summary Table of BMDS Modeling Results	G-357
            G.3.4.2.   Output for Selected Model: Linear	G-358
            G.3.4.3.   Figure for Selected Model: Linear	G-361
            G.3.4.4.   Output for Additional Model Presented: Power, Unrestricted. G-361
            G.3.4.5.   Figure for Additional Model Presented: Power, Unrestricted.. G-365
    G.3.5.   Belletal. (2007): Balano-Preputial Separation, PND 49	G-366
            G.3.5.1.   Summary Table of BMDS Modeling Results	G-366
            G.3.5.2.   Output for Selected Model: Log-Logistic	G-366

                                    G-ix

-------
                      CONTENTS (continued)
       G.3.5.3.   Figure for Selected Model: Log-Logistic	G-369
       G.3.5.4.   Output for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-369
       G.3.5.5.   Figure for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-372
G.3.6.  Cantoni et al. (1981): Urinary Coproporhyrins, 3 Months	G-373
       G.3.6.1.   Summary Table of BMDS Modeling Results	G-373
       G.3.6.2.   Output for Selected Model: Exponential (M4)	G-373
       G.3.6.3.   Figure for Selected Model: Exponential (M4)	G-377
       G.3.6.4.   Output for Additional Model Presented: Power, Unrestricted. G-377
       G.3.6.5.   Figure for Additional Model Presented: Power, Unrestricted.. G-381
G.3.7.  Cantoni et al. (1981): Urinary Porphyrins	G-382
       G.3.7.1.   Summary Table of BMDS Modeling Results	G-382
       G.3.7.2.   Output for Selected Model: Exponential (M2)	G-382
       G.3.7.3.   Figure for Selected Model: Exponential (M2)	G-385
G.3.8.  Crofton et al. (2005): Serum, T4	G-386
       G.3.8.1.   Summary Table of BMDS Modeling Results	G-386
       G.3.8.2.   Output for Selected Model: Exponential (M4)	G-386
       G.3.8.3.   Figure for Selected Model: Exponential (M4)	G-390
G.3.9.  Franc et al. (2001): S-D Rats, Relative Liver Weight	G-391
       G.3.9.1.   Summary Table of BMDS Modeling Results	G-391
       G.3.9.2.   Output for Selected Model: Power	G-391
       G.3.9.3.   Figure for Selected Model: Power	G-395
       G.3.9.4.   Output for Additional Model Presented: Power, Unrestricted. G-395
       G.3.9.5.   Figure for Additional Model Presented: Power, Unrestricted.. G-399
G.3.10. Franc et al. (2001): Long-Evans (L-E) Rats, Relative Liver Weight	G-400
       G.3.10.1.  Summary Table of BMDS Modeling Results	G-400
       G.3.10.2.  Output for Selected Model: Hill	G-400
       G.3.10.3.  Figure for Selected Model: Hill	G-404
       G.3.10.4.  Output for Additional Model Presented: Hill, Unrestricted	G-404
       G.3.10.5.  Figure for Additional Model Presented: Hill, Unrestricted	G-408
G.3.11. Franc et al. (2001): S-D Rats, Relative Thymus Weight	G-409
       G.3.11.1.  Summary Table of BMDS Modeling Results	G-409
       G.3.11.2.  Output for Selected Model: Exponential (M4)	G-409
G.3.12. Figure for Selected Model: Exponential (M4)	G-413
G.3.13. Output for Additional  Model Presented: Polynomial, 3-Degree	G-413
       G.3.13.1.  Figure for Additional Model Presented: Polynomial,
                 3-Degree	G-417
G.3.14. Franc et al. (2001): Long-Evans (L-E) Rats, Relative Thymus Weight... G-418
       G.3.14.1.  Summary Table of BMDS Modeling Results	G-418
       G.3.14.2.  Output for Selected Model: Exponential (M4)	G-418
       G.3.14.3.  Figure for Selected Model: Exponential (M4)	G-422
G.3.15. Franc et al. (2001): Han/Wistar (H/W) Rats, Relative Thymus Weight... G-423

                                G-x

-------
                      CONTENTS (continued)
       G.3.15.1. Summary Table of BMDS Modeling Results	G-423
       G.3.15.2. Output for Selected Model: Exponential (M2)	G-423
       G.3.15.3. Figure for Selected Model: Exponential (M2)	G-427
       G.3.15.4. Output for Additional Model Presented: Exponential (M4) .... G-427
       G.3.15.5. Figure for Additional Model Presented: Exponential (M4)	G-431
G.3.16. Hojoetal. (2002): DRL Reinforce per Minute	G-432
       G.3.16.1. Summary Table of BMDS Modeling Results	G-432
       G.3.16.2. Output for Selected Model: Linear	G-432
       G.3.16.3. Figure for Selected Model: Linear	G-435
       G.3.16.4. Output for Additional Model Presented: Exponential (M4) .... G-436
       G.3.16.5. Figure for Additional Model Presented: Exponential (M4)	G-439
G.3.17. Hojoetal. (2002): DRL Response per Minute	G-440
       G.3.17.1. Summary Table of BMDS Modeling Results	G-440
       G.3.17.2. Output for Selected Model: Exponential (M4)	G-440
       G.3.17.3. Figure for Selected Model: Exponential (M4)	G-444
G.3.18. Kattainen et al. (2001): 3rd Molar Eruption, Female	G-445
       G.3.18.1. Summary Table of BMDS Modeling Results	G-445
       G.3.18.2. Output for Selected Model: Log-Logistic	G-445
       G.3.18.3. Figure for Selected Model: Log-Logistic	G-448
       G.3.18.4. Output for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-448
       G.3.18.5. Figure for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-451
G.3.19. Kattainen et al. (2001): 3rd Molar Length, Female	G-452
       G.3.19.1. Summary Table of BMDS Modeling Results	G-452
       G.3.19.2. Output for Selected Model: Hill	G-452
       G.3.19.3. Figure for Selected Model: Hill	G-456
       G.3.19.4. Output for Additional Model Presented: Hill, Unrestricted	G-456
       G.3.19.5. Figure for Additional Model Presented: Hill, Unrestricted	G-460
G.3.20. Keller etal. (2007): Missing Mandibular Molars, CBAJ	G-461
       G.3.20.1. Summary Table of BMDS Modeling Results	G-461
       G.3.20.2. Output for Selected Model: Multistage, 1-Degree	G-461
       G.3.20.3. Figure for Selected Model: Multistage, 1-Degree	G-464
G.3.21. Kociba et al. (1978): Urinary Coproporphyrin, Females	G-465
       G.3.21.1. Summary Table of BMDS Modeling Results	G-465
       G.3.21.2. Output for Selected Model: Exponential (M4)	G-465
       G.3.21.3. Figure for Selected Model: Exponential (M4)	G-469
G.3.22. Kociba et al. (1978): Uroporphyrin per Creatinine, Female	G-470
       G.3.22.1. Summary Table of BMDS Modeling Results	G-470
       G.3.22.2. Output for Selected Model: Linear	G-470
       G.3.22.3. Figure for Selected Model: Linear	G-473
G.3.23. Kuchiiwa et al. (2002): Immunoreactive Neurons in Dorsalis, Males	G-474
       G.3.23.1. Summary Table of BMDS Modeling Results	G-474

                                G-xi

-------
                      CONTENTS (continued)
        G.3.23.2. Output for Selected Model: Linear	G-474
        G.3.23.3. Figure for Selected Model: Linear	G-477
G.3.24.  Kuchiiwa et al. (2002): Immunoreactive Neurons in Medianus, Males... G-478
        G.3.24.1. Summary Table of BMDS Modeling Results	G-478
        G.3.24.2. Output for Selected Model: Linear	G-478
        G.3.24.3. Figure for Selected Model: Linear	G-481
G.3.25.  Kuchiiwa et al. (2002): Immunoreactive Neurons in B9, Males	G-482
        G.3.25.1. Summary Table of BMDS Modeling Results	G-482
        G.3.25.2. Output for Selected Model: Linear	G-482
        G.3.25.3. Figure for Selected Model: Linear	G-485
G.3.26.  Kuchiiwa et al. (2002): Immunoreactive Neurons in Magnus, Males	G-486
        G.3.26.1. Summary Table of BMDS Modeling Results	G-486
        G.3.26.2. Output for Selected Model: Linear	G-486
        G.3.26.3. Figure for Selected Model: Linear	G-489
G.3.27.  Latchoumycandane and Mathur (2002): Sperm Production	G-490
        G.3.27.1. Summary Table of BMDS Modeling Results	G-490
        G.3.27.2. Output for Selected Model: Hill	G-490
        G.3.27.3. Figure for Selected Model: Hill	G-494
        G.3.27.4. Output for Additional Model Presented: Hill, Unrestricted	G-494
        G.3.27.5. Figure for Additional Model Presented: Hill, Unrestricted	G-498
G.3.28.  Lietal. (1997): FSH	G-499
        G.3.28.1. Summary Table of BMDS Modeling Results	G-499
        G.3.28.2. Output for Selected Model: Power	G-499
        G.3.28.3. Figure for Selected Model: Power	G-503
        G.3.28.4. Output for Additional Model Presented: Power, Unrestricted. G-503
        G.3.28.5. Figure for Additional Model Presented: Power, Unrestricted.. G-507
G.3.29.  Li et al. (2006): Estradiol, 3-Day	G-508
        G.3.29.1. Summary Table of BMDS Modeling Results	G-508
        G.3.29.2. Output for Selected Model: Linear	G-508
        G.3.29.3. Figure for Selected Model: Linear	G-511
G.3.30.  Li et al. (2006): Progesterone, 3-Day	G-512
        G.3.30.1. Summary Table of BMDS Modeling Results	G-512
        G.3.30.2. Output for Selected Model: Exponential (M4)	G-512
        G.3.30.3. Figure for Selected Model: Exponential (M4)	G-516
        G.3.30.4. Output for Additional Model Presented: Hill, Unrestricted	G-516
        G.3.30.5. Figure for Additional Model Presented: Hill, Unrestricted	G-520
G.3.31.  Markowski et al. (2001): FR10 Run Opportunities	G-521
        G.3.31.1. Summary Table of BMDS Modeling Results	G-521
        G.3.31.2. Output for Selected Model: Exponential (M2)	G-521
        G.3.31.3. Figure for Selected Model: Exponential (M2)	G-525
G.3.32.  Markowski et al. (2001): FR2 Revolutions	G-526
        G.3.32.1. Summary Table of BMDS Modeling Results	G-526
        G.3.32.2. Output for Selected Model: Hill	G-526

                                G-xii

-------
                       CONTENTS (continued)
        G.3.32.3. Figure for Selected Model: Hill	G-530
        G.3.32.4. Output for Additional Model Presented: Power, Unrestricted. G-530
        G.3.32.5. Figure for Additional Model Presented: Power, Unrestricted.. G-534
G.3.33.  Markowski et al. (2001): FR5 Run Opportunities	G-535
        G.3.33.1. Summary Table of BMDS Modeling Results	G-535
        G.3.33.2. Output for Selected Model: Hill	G-535
        G.3.33.3. Figure for Selected Model: Hill	G-539
        G.3.33.4. Output for Additional Model Presented: Power, Unrestricted. G-539
        G.3.33.5. Figure for Additional Model Presented: Power, Unrestricted.. G-543
G.3.34.  Miettinen et al. (2006): Cariogenic Lesions, Pups	G-544
        G.3.34.1. Summary Table of BMDS Modeling Results	G-544
        G.3.34.2. Output for Selected Model: Log-Logistic	G-544
        G.3.34.3. Figure for Selected Model: Log-Logistic	G-547
        G.3.34.4. Output for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-547
        G.3.34.5. Figure for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-550
G.3.35.  Murray et al. (1979): Fertility in F2 Generation	G-551
        G.3.35.1. Summary Table of BMDS Modeling Results	G-551
        G.3.35.2. Output for Selected Model: Multistage, 2-Degree	G-551
        G.3.35.3. Figure for Selected Model: Multistage, 2-Degree	G-554
G.3.36.  National Toxicology Program (1982): Toxic Hepatitis, Male Mice	G-554
        G.3.36.1. Summary Table of BMDS Modeling Results	G-554
        G.3.36.2. Output for Selected Model: Multistage, 3-Degree	G-555
        G.3.36.3. Figure for Selected Model: Multistage, 3-Degree	G-557
G.3.37.  National Toxicology Program (2006): Alveolar Metaplasia	G-558
        G.3.37.1. Summary Table of BMDS Modeling Results	G-558
        G.3.37.2. Output for Selected Model: Log-Logistic	G-558
        G.3.37.3. Figure for Selected Model: Log-Logistic	G-561
        G.3.37.4. Output for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-561
        G.3.37.5. Figure for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-564
G.3.38.  National Toxicology Program (2006): Eosinophilic Focus, Liver	G-565
        G.3.38.1. Summary Table of BMDS Modeling Results	G-565
        G.3.38.2. Output for Selected Model: Probit	G-565
        G.3.38.3. Figure for Selected Model: Probit	G-567
G.3.39.  National Toxicology Program (2006): Fatty Change Diffuse, Liver	G-568
        G.3.39.1. Summary Table of BMDS Modeling Results	G-568
        G.3.39.2. Output for Selected Model: Weibull	G-568
        G.3.39.3. Figure for Selected Model: Weibull	G-570
G.3.40.  National Toxicology Program (2006): Gingival Hyperplasia,
        Squamous, 2 Years	G-571

                                G-xiii

-------
                       CONTENTS (continued)
        G.3.40.1. Summary Table of BMDS Modeling Results	G-571
        G.3.40.2. Output for Selected Model: Log-Logistic	G-571
        G.3.40.3. Figure for Selected Model: Log-Logistic	G-574
        G.3.40.4. Output for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-574
        G.3.40.5. Figure for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-577
G.3.41.  National Toxicology Program (2006): Hepatocyte Hypertrophy,
        2 Years	G-578
        G.3.41.1. Summary Table of BMDS Modeling Results	G-578
        G.3.41.2. Output for Selected Model: Multistage, 5-Degree	G-578
        G.3.41.3. Figure for Selected Model: Multistage, 5-Degree	G-581
G.3.42.  National Toxicology Program (2006): Necrosis, Liver	G-582
        G.3.42.1. Summary Table of BMDS Modeling Results	G-582
        G.3.42.2. Output for Selected Model: Log-Probit, Unrestricted	G-582
        G.3.42.3. Figure for Selected Model: Log-Probit, Unrestricted	G-585
G.3.43.  National Toxicology Program (2006): Oval Cell  Hyperplasia	G-586
        G.3.43.1. Summary Table of BMDS Modeling Results	G-586
        G.3.43.2. Output for Selected Model: Probit	G-586
        G.3.43.3. Figure for Selected Model: Probit	G-588
        G.3.43.4. Output for Additional Model Presented: Weibull	G-589
        G.3.43.5. Figure for Additional Model Presented: Weibull	G-591
G.3.44.  National Toxicology Program (2006): Pigmentation, Liver	G-591
        G.3.44.1. Summary Table of BMDS Modeling Results	G-591
        G.3.44.2. Output for Selected Model: Log-Probit	G-592
        G.3.44.3. Figure for Selected Model: Log-Probit	G-594
G.3.45.  National Toxicology Program (2006): Toxic Hepatopathy	G-594
        G.3.45.1. Summary Table of BMDS Modeling Results	G-594
        G.3.45.2. Output for Selected Model: Multistage, 5-Degree	G-595
        G.3.45.3. Figure for Selected Model: Multistage, 5-Degree	G-597
G.3.46.  Ohsako et al. (2001): Ano-Genital Length, PND  120	G-598
        G.3.46.1. Summary Table of BMDS Modeling Results	G-598
        G.3.46.2. Output for Selected Model: Hill	G-598
        G.3.46.3. Figure for Selected Model: Hill	G-602
        G.3.46.4. Output for Additional Model Presented: Hill, Unrestricted	G-602
        G.3.46.5. Figure for Additional Model Presented: Hill, Unrestricted	G-606
G.3.47.  Sewalletal. (1995): T4 In Serum	G-607
        G.3.47.1. Summary Table of BMDS Modeling Results	G-607
        G.3.47.2. Output for Selected Model: Hill	G-607
        G.3.47.3. Figure for Selected Model: Hill	G-611
        G.3.47.4. Output for Additional Model Presented: Hill, Unrestricted	G-611
        G.3.47.5. Figure for Additional Model Presented: Hill, Unrestricted	G-615
G.3.48.  Shi et al. (2007): Estradiol 17B, PE9	G-616

                                G-xiv

-------
                      CONTENTS (continued)
       G.3.48.1.  Summary Table of BMDS Modeling Results	G-616
       G.3.48.2.  Output for Selected Model: Exponential (M4)	G-616
       G.3.48.3.  Figure for Selected Model: Exponential (M4)	G-620
G.3.49. Smialowicz et al. (2008): PFC per 106 Cells	G-621
       G.3.49.1.  Summary Table of BMDS Modeling Results	G-621
       G.3.49.2.  Output for Selected Model: Power, Unrestricted	G-621
       G.3.49.3.  Figure for Selected Model: Power, Unrestricted	G-625
       G.3.49.4.  Output for Additional Model Presented: Power	G-625
       G.3.49.5.  Figure for Additional Model Presented: Power	G-629
G.3.50. Smialowicz et al. (2008): PFC per Spleen	G-630
       G.3.50.1.  Summary Table of BMDS Modeling Results	G-630
       G.3.50.2.  Output for Selected Model: Power, Unrestricted	G-630
       G.3.50.3.  Figure for Selected Model: Power, Unrestricted	G-634
       G.3.50.4.  Output for Additional Model Presented: Power	G-634
       G.3.50.5.  Figure for Additional Model Presented: Power	G-638
G.3.51. Smith etal. (1976): Cleft Palate in Pups	G-639
       G.3.51.1.  Summary Table of BMDS Modeling Results	G-639
       G.3.51.2.  Output for Selected Model: Log-Logistic	G-639
       G.3.51.3.  Figure for Selected Model: Log-Logistic	G-641
G.3.52. Sparschu et al. (1971): Fetal Body Weight, Male	G-642
       G.3.52.1.  Summary Table of BMDS Modeling Results	G-642
       G.3.52.2.  Output for Selected Model: Exponential (M5)	G-642
       G.3.52.3.  Figure for Selected Model: Exponential (M5)	G-645
G.3.53. Sparschu et al. (1971): Fetal Body Weight, Female	G-646
       G.3.53.1.  Summary Table of BMDS Modeling Results	G-646
       G.3.53.2.  Output for Selected Model: Exponential (M2)	G-646
       G.3.53.3.  Figure for Selected Model: Exponential (M2)	G-649
G.3.54. Toth et al. (1979): Amyloidosis	G-650
       G.3.54.1.  Summary Table of BMDS Modeling Results	G-650
       G.3.54.2.  Output for Selected Model: Log-Logistic	G-650
       G.3.54.3.  Figure for Selected Model: Log-Logistic	G-653
       G.3.54.4.  Output for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-653
       G.3.54.5.  Figure for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-656
G.3.55. Toth etal. (1979): Skin Lesions	G-657
       G.3.55.1.  Summary Table of BMDS Modeling Results	G-657
       G.3.55.2.  Output for Selected Model: Logistic	G-657
       G.3.55.3.  Figure for Selected Model: Logistic	G-659
       G.3.55.4.  Output for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-660
       G.3.55.5.  Figure for Additional Model Presented: Log-Logistic,
                 Unrestricted	G-662

                                G-xv

-------
                           CONTENTS (continued)
    G.3.56. van Birgelen et al. (1995): Hepatic Retinol	G-663
            G.3.56.1. Summary Table of BMDS Modeling Results	G-663
            G.3.56.2. Output for Selected Model: Exponential (M4)	G-663
            G.3.56.3. Figure for Selected Model: Exponential (M4)	G-667
            G.3.56.4. Output for Additional Model Presented: Power, Unrestricted. G-667
            G.3.56.5. Figure for Additional Model Presented: Power, Unrestricted.. G-671
    G.3.57. van Birgelen et al. (1995): Hepatic Retinol Palmitate	G-672
            G.3.57.1. Summary Table of BMDS Modeling Results	G-672
            G.3.57.2. Output for Selected Model: Linear	G-672
            G.3.57.3. Figure for Selected Model: Linear	G-675
            G.3.57.4. Output for Additional Model Presented: Power, Unrestricted. G-676
            G.3.57.5. Figure for Additional Model Presented: Power, Unrestricted.. G-679
    G.3.58. White et al. (1986): CH50	G-680
            G.3.58.1. Summary Table of BMDS Modeling Results	G-680
            G.3.58.2. Output for Selected Model: Hill	G-680
            G.3.58.3. Figure for Selected Model: Hill	G-684
            G.3.58.4. Output for Additional Model Presented: Hill, Unrestricted	G-684
            G.3.58.5. Figure for Additional Model Presented: Hill, Unrestricted	G-688
G.4. REFERENCES	G-688
                                    G-xvi

-------
        APPENDIX G.    NONCANCER BENCHMARK DOSE MODELING


G.I.  BENCHMARK DOSE SOFTWARE (BMDS) INPUT TABLES

G.I.I.  Amin et al. (2000)
Endpoinf
Saccharin consumed, female rats (0.25%)
(mL saccharin solution/100 gbody weight)0
Saccharin consumed, female rats (0.50%)
(mL saccharin solution/100 gbody weight)0
Saccharin preference ratio, female rats (0.25%) (ratio of
saccharin solution consumed to total fluid consumed)"1
Saccharin preference ratio, female rats (0.50%) (ratio of
saccharin solution consumed to total fluid consumed)"1
Administered dose (ng/kg-day)
0
25a
100
Internal dose (ng/kg blood)b
0
(» = 10)
31.67 ±6.53
22.40 ±5.05
82. 14 ±4.22
72.73 ±7.79
3.38
(» = 10)
24.60 ±3.79
11.38 ±2.42
58.12 ± 10.71
44.48 ± 10.39
10.57
(it = 10)
10.70 ± 1.68
4.54 ±1.05
54.87 ±6. 17
33.77 ±7.79
a Lowest-observed-adverse-effect level (LOAEL) identified.
b From the Emond physiologically based pharmacokinetic (PBPK) model described in Section 3.3.
0 Values are the mean ± standard error (SE). Data obtained from Figure 2 in Amin et al. (2000).
d Values are the ratio ± SE. Data obtained from Figure 3 in Amin et al. (2000).
G.1.2.  Bell et al. (2007)
Endpoint
Proportion of male rat pups that had not
undergone balano-preputial separation on
PND490
Administered dose (ng/kg-day)
0
2.4a
8
46
Internal dose (ng/kg blood)b
0
(n = 30)
1/30 (3%)
2.20
(n = 30)
5/30 (17%)
5.14
(« = 30)
6/30 (20%)
18.41
(n = 30)
15/30 (50%)
a LOAEL identified.
b From the Emond PBPK model described in Section 3.3.
0 Data obtained from Figure 2 in Bell et al. (2007).

PND = postnatal day.
                                             G-l

-------
G.1.3.   Cantoni et al. (1981)
Endpoint
Urinary coproporphyrins in female
rats (ug coproporphyrin methyl
ester/24 hr) at 3 months'
Urinary porphyrins in rats
(nmol/24 hr) after 45 weeks'
Administered dose (ng/kg-day)
0
1.43a
14.3
143
Internal dose (ng/kg blood)b
0
(»=4)
0.74 ±0.17
2.27 ±0.49
1.85
(» = 4)
1.81±0.42d
5.55±0.85d
8.84
(»=3)
2.73±0.75e
7.62±1.79d
50.05
(»=3)
3.00±1.30e
196.89 ± 63. 14e
aLOAEL identified.
b From the Emond PBPK model described in Section 3.3.
0 Values are the mean ± SE. Data for urinary coproporphyrins and urinary porphyrins obtained from Figure 1 and
Table 1, respectively, in Cantoni et al. (1981).
d Statistically significant as compared to control (p < 0.05).
e Statistically significant as compared to control (p < 0.01).
G.1.4.   Crofton et al. (2005)
Endpoint
Serum T4 in
female rats (%
control)d
Administered dose (ng/kg-day)
0
0.1
3
10
30a
100b
300
1,000
3,000
10,000
Internal dose (ng/kg blood)0
0
(n = 14)
100.00 ±
15.44
0.02
(«=6)
96.27 ±
14.98
0.49
(n = 12)
98.57 ±
18.11
1.38
(«=6)
99.76 ±
19.04
3.46
(«=6)
93.32±
12.11
9.26
(« = 6)
70.94 ±
12.74
23.07
(« = 6)
62.52 ±
14.75
65.65
(«=6)
52.68 ±
22.73
180.90
(« = 6)
54.66 ±
19.71
583.48
(« = 4)
49.15 ±
11.15
aNo-observed-adverse-effect level (NOAEL) identified.
bLOAEL identified.
0 From the Emond PBPK model described in Section 3.3.
d Values are the mean ± SD. Data were obtained from a Crofton et al. (2005) supplemental file, available at
 http://ehp.niehs.nih.gov/docs/2005/8195/supplemental.pdf.
                                                 G-2

-------
G.1.5.  DeCaprio et al. (1986)
Endpoint
Absolute kidney weight (g),
males'1
Absolute thymus weight (g),
males'1
Body weight (g), males6
Relative brain weight, males'1
Relative liver weight, males'1
Relative thymus weight, males'1
Endpoint
Body weight (g), females6
Relative liver weight, females'1
Administered dose (ng/kg-day)
0
0.12
0.61a
4.9b
26
Internal dose (ng/kg blood) c
NM
(» = 10)
5.49 ±0.17
0.56 ±0.050
713 ±15
0.54 ±0.015
4.54 ±0.23
0.078 ± 0.006
NM
(it = 10)
5.14±0.12
0.45 ± 0.022
682 ± 16
0.56 ±0.016
4.1±0.14
0.066 ± 0.003
NM
(« = 11)
4.71 ±0.12
0.44 ±0.034
651 ±19
0.6 ±0.016
5. 36 ±0.61
0.068 ± 0.004
NM
(n = 10)
4.3±0.15f
0.35 ± 0.167s
603 ± 20f
0.65±0.016f
5.63±0.29f
0.06±0.003f
NM
(«=4)
-
-
433±38h
-
-
-
Administered dose (ng/kg-day)
0
0.12
0.68
4.86
31
Internal dose (ng/kg blood)0
0
(» = 8)
602 ± 12
4.3 ±0.26
NM
(n = 10)
583 ± 22
4.49 ±0.35
NM
(» = 9)
570 ± 22
4.27 ±0.16
NM
(n = 10)
531±14f
5.54±0.43f
NM
(«=4)
351±49h
-
a NO AEL identified.
bLOAEL identified.
0 Internal dose not calculated using the Emond PBPK (guinea pigs).
d Organ weight data in guinea pigs obtained from Table 2 of DeCaprio et al. (1986). Values are the mean ± SE.
Relative organs weights were calculated as organ weight (g)/body weight (g) x 100.
e Body weight data in guinea pigs obtained from Table 1 of DeCaprio et al. (1986). Values are the mean ± SE.
Statistically significant as compared to control (p < 0.05).
8 Statistically significant as compared to control (p < 0.01).
h Statistically significant as compared to control (p< 0.001).

NM = not modeled.
                                                  G-3

-------
G.1.6.   Franc et al. (2001)
Endpoint
S-D rats, relative liver weight"1
L-E rats, relative liver weight"1
S-D rats, relative thymus weight"1
L-E rats, relative thymus weight"1
H/W rats, relative thymus weight"1
Administered dose (ng/kg-day)
0
10a
30b
100
Internal dose (ng/kg blood) c
0
(» = 8)
100.0 ±5.0
100.0 ±3. 5
100.2 ±29.4
103.4 ±19.3
101.2 ±12.7
6.59
(»=8)
108.1±6.0e
106.3 ±6.3
91.2 ±17.0
95.4 ±24.9
97.5 ±11.7.0
14.48
(» =8)
116.8 ±9.2e
116.8 ±3.2e
51.4±15.4e
38.7±17.0e
71.0±8.5e
36.43
(» = 8)
155.3 ±10.9e
122.2 ±7.0e
22.8 ± 10.6e
35.0±27.6e
49.3 ± 15.4e
a NO AEL identified.
bLOAEL identified.
0 From the Emond PBPK model described in Section 3.3.
d
 Values are the mean ± SE.  Data obtained from Figure 5 in Franc et al. (2001).
e Statistically significant as compared to control (p < 0.05).

H/W = Han/Wistar; L-E = Long-Evans; S-D = Sprague-Dawley.
G.1.7.   Hojo et al. (2002)
Endpoint
DRL reinforcements/min, rat litters0
DRL responses/min, rat litters0
Administered dose (ng/kg-day)
0
20a
60
180
Internal dose (ng/kg blood)b
0
(» = 5)
-0.814 ±0.45
18.44 ±7.99
1.62
(»=5)
-0.364 ± 0.82
-0.99 ±10.96
4.17
(»=6)
0.374 ±0.54
-4.52 ±7.19
10.70
(» = 5)
-0.163 ±0.44
-0.41 ±15.23
aLOAEL identified.
b From the Emond PBPK model described in Section 3.3.
0 DRL = differential reinforcement of low rate. Values are the mean ± SD.  Data obtained from Table 5 in
 Hojo et al. (2002).
                                                G-4

-------
G.1.8.   Kattainen et al. (20011
Endpoint
3rd molar mesio-distal
length in female rat
offspring (molar
development) (mm)0
Proportion of female rat
offspring without 3rd molar
eruption on PND 35e
Administered dose (ng/kg-day)
0
30a
100
300
1,000
Internal dose (ng/kg blood)b
0
(it = 16)
1.86 ±0.017
1/16 (10%)
2.23
(» = I?)
1.58±0.045d
3/17 (20%)
6.25
(» = 15)
1.6±0.069d
4/15 (30%)
16.08
(» = 12)
1.5±0.064d
6/12 (50%)d
46.86
(» = 19)
1.35±0.118d
13/19 (70%)d
aLOAEL identified.
b From the Emond PBPK model described in Section 3.3.
0 Values are the mean ± SE.  Data were obtained from Figure 3 in Kattainen et al. (2001).
d Statistically significant as compared to control (p < 0.05).
e Data were obtained from Figure 2 in Kattainen et al. (2001).
G.1.9.   Keller et al. (2008a; 2008b; 2007)
Endpoint
Frequency of missing 3rd mandibular molars in CB A J
mice0
Administered dose (ng/kg-day)
0
10a
100
1,000
Internal dose (ng/kg blood)b
0
0/29 (0%)
0.54
2/23 (10%)
4.29
6/29 (20%)
34.06
30/30 (100%)
aLOAEL identified.
b From the Emond PBPK model described in Section 3.3.
c Data obtained from Table 1 in Keller et al. (2007).
                                                G-5

-------
G.1.10. Kociba et al. (1978)
Endpoint
Urinary coproporphyrin (ug/48 h),
female ratsd
ug uroporphyrin per mg creatinine,
female ratsd
Administered dose (ng/kg-day)
0
la
10b
100
Internal dose (ng/kg blood) c
0
(» = 5)
9.8 ±1.3
0.157 ±0.05
1.55
(» = 5)
8.6 ±2
0.143 ±0.037
7.15
(» = 5)
16.4 ± 4.7 e
0.181 ±0.053
38.56
(» =5)
17.4 ±4e
0.296 ± 0.074 e
a NO AEL identified.
bLOAEL identified.
0 From the Emond PBPK model described in Section 3.3.
d Values are the mean ± SD. Data obtained from Table 2 in Kociba et al. (1978).
e Statistically significant as compared to control (p < 0.05).
G.l.ll. Kuchiiwa et al. (2002)
Endpoint
Immunoreactive neurons in dorsalis, males0
Immunoreactive neurons in medianus, males0
Immunoreactive neurons inB9, males0
Immunoreactive neurons in magnus, males0
Administered Dose (ng/kg-day)
0
0.7a
70
Internal Dose (ng/kg blood)b
0
(»=6)
237.1 ±29.0
91.1 ±12.2
152.1 ± 16.0
43.61 ±3.40
0.26
(» = 6)
136.6 ±22.4d
33.3±4.55d
46.8±12.1d
19.82 ±10.20d
9.12
(»=6)
86.0 ± 13.2d'e
23.1±8.10d'e
19.6 ± 15.2d'e
11.10 ±3.884e
a LOAEL identified.
b From the Emond PRPK model described in Section 3.3.
0 Values are the mean ± SD.  Data obtained from Figure 2 in Kuchiiwa et al. (2002).
d Statistically significant as compared to control (p < 0.01).
e Dose dropped from Benchmark Dose (BMD) modeling
                                               G-6

-------
G.I.12. Latchoumycandane and Mathur (2002)
Endpoint
Daily sperm production (x 106) in adult
male rats (mg)°
Administered dose (ng/kg-day)
0
la
10
100
Internal dose (ng/kg blood)b
0
(» = 6)
22. 19 ±2.67
0.78
(» = 6)
15.67 ±2.65d
4.65
(» = 6)
13.65 ±2.19d
27.27
(»=6)
13.1±3.16d
aLOAEL identified.
b From the Emond PBPK model described in Section 3.3.
! Values are the mean ± SD. Data obtained from Table 1 in Latchoumycandane and Mathur (2002).
1 Statistically significant as compared to control (p < 0.05).
G.1.13. Li et al.
Endpoint
Serum FSH
(ng/mL) in female
ratsd
Administered dose (ng/kg-day)
0
3a
10b
30
100
300
1,000
3,000
10,000
30,000
Internal dose (ng/kg blood)0
0
(it = 10)
23.86
±9.38
0.27
(» = 10)
22.16
±15.34
0.80
(» = 10)
85.23
±29.83
2.1
(it = 10)
73.30 ±
15.34
5.87
(it = 10)
126.14 ±
50.28
15
(if = 10)
132.10 ±
36.65
43.33
(if = 10)
1 16.76 ±
16.19
119.94
(it = 10)
304.26 ±
48.58
385.96
(it = 10)
346.88 ±
47.73
1,171.90
(if = 10)
455.11 ±
90.34
a NO AEL identified.
bLOAEL identified.
0 From the Emond PBPK model described in Section 3.3.
d
 Values are the mean ± SE. Data obtained from Figure 3 in Li et al. (1997).
FSH = follicle stimulatin hormone.
                                               G-7

-------
G.1.14. Li et al. (2006)
Endpoint
Serum estradiol/Cpg-mL)"1 in female
mice (l~3d)c
Serum progesterone (ng-mL)"1 in
female mice (l~3d)°
Administered dose (ng/kg-day)
0
2a
50
100
Internal dose (ng/kg blood)b
0
(» = 10)
10.17 ±3. 85
61.74 ±3. 51
0.16
(» = 10)
19.91 ±6.31
30.56 ±12.80d
2.84
(if = 10)
24.72 ± 4.60
16.93 ±10.53
5.12
(it = 10)
18.09 ±5.57
11.36 ±13.83
aLOAEL identified.
b From the Emond PBPK model described in Section 3.3.
! Values are the mean ± SE.  Data obtained from Figures 3 (estradiol) and 4 (progesterone) in Li et al. (2006).
1 Statistically significant as compared to control (p < 0.01).
G.1.15. Markowski et al. (20011
Endpoint
FR10 earned run opportunities, adult
female offspring0
FR2 total revolutions, adult female
offspring0
FR5 earned run opportunities, adult
female offspring0
Administered dose (ng/kg-day)
0
20a
60
180
Internal dose (ng/kg blood)b
0
(it = 7)
13.29 ±8.65
119.29 ±69.9
26. 14 ±12.28
1.56
(it = 4)
11.25 ±5.56
108.5 ±61
23.5 ±7.04
4.03
(» = 6)
5.75 ±3.53
56.5 ±31.21
12.8 ±6.17
10.32
(it = 7)
7 ±6.01
68.14 ±33.23
13.14±7.14
aLOAEL identified.
b From the Emond PBPK model described in Section 3.3.
0 Values are the mean ± SD. Data obtained from Table 3 in Markowski et al. (2001)
                                                G-8

-------
G.I. 16. Miettinen et al. (2006)
Endpoint
Cariogenic lesions in rat pups0
Administered dose (ng/kg-day)
0
30a
100
300
1,000
Internal dose (ng/kg blood)b
0
(11 =42)
25/42 (60%)
2.22
(11 = 29)
23/29 (79%)d
6.23
(» = 15)
19/25 (76%)
16.01
(»=24)
20/24 (83%)d
46.64
(» = 32)
29/32 (91%)d
aLOAEL identified.
b From the Emond PBPK model described in Section 3.3.
c Data obtained from Table 2 in Miettinen et al. (2006).
d Statistically significant as compared to control (p < 0.05).
G.I.17. National Toxicology Program (1982)
Endpoint
Numbers of male mice with toxic
hepatitis0
Administered dose (ng/kg-day)
0
1.43"
7.14
71.4
Internal dose (ng/kg blood)b
0
(» = 73)
1/73 (1.4%)
0.77
(»=49)
5/49 (10%)
2.27
(» = 49)
3/49(6.1%)
11.24
(n = 50)
44/50 (88%)
aLOAEL identified.
b From the Emond PBPK model described in Section;
0 Data obtained from Table 11 in NTP (1982).
                                              G-9

-------
G.I.18. National Toxicology Program (2006)
Endpoint0
Gingival squamous hyperplasia
Liver, hepatocyte hypertrophy
Heart, cardiomyopathy
Liver, eosinophilic focus, multiple
Liver, fatty change, diffuse
Liver, necrosis
Liver, pigmentation
Liver, toxic hepatopathy
Oval cell hyperplasia
Lung, alveolar to bronchiolar
epithelial metaplasia (Alveolar
epithelium, metaplasia, bronchiolar)
Administered dose (ng/kg-day)
0
2.14a
7.14
15.7
32.9
71.4
Internal dose (ng/kg blood)b
0
(» = 10)
1/53
(2%)
0/53
(0%)
10/53
(19%)
3/53
(6%)
0/53
(0%)
1/53
(2%)
4/53
(8%)
0/53
(0%)
0/53
(0%)
2/53
(4%)
2.56
(» = 10)
7/54
(13%)d
19/54
(40%)e
12/54
(22%)
8/54
(15%)
2/54
(4%)
4/54
(7%)
9/54
(17%)
2/54
(4%)
4/54
(10%)d
19/54e
(35%)
5.69
(11 = 10)
14/53
(26%)e
19/53
(40%)c
22/53e
(42%)
14/53
(26%)
12/53e
(23%)
4/53
(8%)
34/53e
(64%)
8/53
(15%)
3/53
(10%)
33/53e
(62%)
9.79
(it = 10)
13/53
(25%)e
42/53
(80%)e
25/52e
(48%)
17/53
(32%)
17/53e
(32%)
8/53d
(15%)
48/53e
(91%)
30/53
(57%)
20/53
(40%)e
35/52e
(67%)
16.57
(it = 10)
15/53
(28%)e
41/53
(80%)e
32/53e
(60%)
22/53
(42%)
30/53e
(57%)
10/53e
(19%)
52/53e
(98%)
45/50
(85%)
38/53
(70%)d
45/53e
(85%)
29.70
(it = 10)
16/53
(30%)e
52/53
(100%)e
36/52e
(69%)
42/53
(79%)
48/53e
(91%)
17/53e
(32%)
53/53e
(100%)
53/53
(100%)
53/53
(100%)e
46/52e
(89%)
aLOAEL identified.
b From the Emond PBPK model described in Section 3.3.
0 Data are for female rats in 2-year gavage study.  Data for all endpoints obtained from Table A5b in NTP (2006).
d Statistically significant as compared to control (p < 0.05).
e Statistically significant as compared to control (p < 0.01).
                                               G-10

-------
G.1.19. Ohsako et al. (2001)
Endpoint
Anogenital distance (mm) in male
rat offspring, PND120d
Administered dose (ng/kg-day)
0
12.5a
50b
200
800
Internal dose (ng/kg blood)0
0
(» = 12)
28.91 ±0.90
1.04
(it = 10)
27.94 ±0.79
3.47
(» = 10)
25.17 ±1.02e
11.36
(» = 10)
26.01 ±0.90f
38.42
(» = 12)
23.80±0.45e
a NO AEL for selected endpoint.
b LO AEL for selected endpoint.
0 From the Emond PBPK model described in Section 3.3.
d Values are the mean ± SE. Data obtained from Figure 7 in Ohsako et al. (2001).
e Statistically significant as compared to control (p < 0.01).
f Statistically significant as compared to control (p < 0.05).
G.1.20. Sewall et al. (1995)
Endpoint
Serum levels of T4 (nmol/L),
saline non noninitiatedd
Administered dose (ng/kg-day)
0
3.5
10.7a
35b
125
Internal dose (ng/kg blood)0
0
(» = 9)
30.70 ±1.55
3.29
(»=9)
27.88 ±2.39
7.11
(» = 9)
25.90 ±2.27
16.63
(» = 9)
23.56 ±1.79e
44.66
(»=9)
18.40 ± 1.37e
a NO AEL for selected endpoint.
b LOAEL for selected endpoint.
0 From the Emond PBPK model described in Section 3.3.
d Values are the mean ± SE. Data obtained from Figure 1 in Sewall et al. (1995).
e Statistically significant as compared to control (p < 0.05).
                                                G-ll

-------
G.1.21. Shi et al. (2007)
Endpoint
Serum estradiol — 17(3 at
proestrus 9 in female rats at
9 mo. of age (pg/mL)d
Administered dose (ng/kg-day)
0
0.143a
0.714b
7.14
28.6
Internal dose (ng/kg blood)0
0
(it = 10)
102.86 ±13. 10
0.34
(» = 10)
86.19 ±6.19
1.07
(» = 10)
63.33 ±9.29e
5.23
(if = 10)
48.1±5.95e
13.91
(it = 10)
38.57 ±7.14e
a NO AEL identified.
bLOAEL identified.
0 From the Emond PBPK model described in Section 3.3.
d Values are the mean ± SE. Data obtained from Figure 4 in Shi et al. (2007).
e Statistically significant as compared to control (p < 0.05).
G.I.22. Smialowicz et al. (2008)
Endpoint
PFC per 106 cells in female
mice0
PFC x 104 per spleen in female
mice0
Administered dose (ng/kg-day)
0
1.07"
10.7
107
321
Internal dose (ng/kg blood)b
0
(it = 15)
1,491 ±716
27.8 ± 13.4
0.44
(it = 14)
1,129 ±171d
21±13.6d
2.46
(if = 15)
945±516d
17.6±9.4d
13.40
(it = 15)
677 ± 465d
12.6±8.7d
31.65
(if = 8)
161±117d
3.0±3.1d
aLOAEL identified.
b From the Emond PBPK model described in Section 3.3.
0 Values are the mean ± SD. Data obtained from Table 4 in Smialowicz et al. (2008)
d Statistically significant as compared to control (p < 0.05).
                                               G-12

-------
G.1.23. Smith et al. (1976)
Endpoint
Cleft palate in pupsd
Administered Dose (ng/kg-day)
0
1
10
100a
l,000b
3,000
Internal Dose (ng/kg blood)0
0
0/34 (0%)
0.12
2/41 (4.9%)
1.01
0/19 (0%)
7.11
1/17 (5.9%)
50.59
4/19 (21%)e
138.07
10/14 (71%)e
a NO AEL identified
b LOAEL identified
0 From the Emond PBPK model described in Section 3.3.
d
 Values are the incidence and number of litter groups. Data obtained from Table 3 in Smith et al. (1976)
' Statistically significant as compared to control (p < 0.01).
G.1.24. Sparschu et al. (19711
Endpoint
Body weight of male
fetusesd

Body weight of
female fetusesd
Administered Dose (ng/kg-day)
0
30a
125b
500
2,000
Internal Dose (ng/kg blood)0
0
(» = 117)
4.03 ±0.37
(n = 129)
3. 89 ±0.39
5.09
(n = 55)
4.14 ±0.26
(n = 60)
3.98 ±0.35
16.28
(n = 66)
3.85 ± 0.35 e
(n = 58)
3.71±0.37e
52.87
(» = 39)
3.86 ± 0.61 e
(» = 54)
3.78±0.54e
188.26
(»=3)
2.72 ± 0.25 e
(» =4)
2.69±0.19e
                                         a NO AEL identified
                                         b LOAEL identified
                         0 From the Emond PBPK model described in Section 3.3.
               1 Values are the mean ± SD. Data obtained from Table 4 in Sparschu et al. (19711
                        e Statistically significant as compared to control (p < 0.05).
                                               G-13

-------
G.1.25. Toth et al. (1979)
Endpoint
Number with amyloidosis plus skin
lesions in mice0
Number with skin lesions in mice0
Administered dose (ng/kg-day)
0
la
100
1,000
Internal dose (ng/kg blood)b
0
(» =38)
0/38 (0%)
0/38 (0%)
0.57
(» = 44)
5/44(11%)
5/44(11%)
14.21
(11 =44)
10/44 (23%)
13/44 (30%)
91.21
(»=43)
17/43 (40%)
25/43 (58%)
aLOAEL identified.
b From the Emond PBPK model described in Section 3.3.
0 Data obtained from Table 2 in Toth et al. (1979).
G.1.26. van Birgelen et al.
Endpoint
Hepatic retinol (mg/g liver) in
female rats0
Hepatic retinol palmitate
(mg/g liver) in female rats0
Plasma FT4 (pmol/L) in female
rats0
Plasma TT4 (nmol/L) in female
rats0
Administered dose (ng/kg-day)
0
14a
26
47
320
1,024
Internal dose (ng/kg blood)b
0
(»=8)
14.9 ±3.1
472 ± 96
23.4 ± 1.1
40.9 ±2.4
7.20
(» = 8)
8.4±1.2d
94 ± 24d
24.5 ±2.0
41.4 ±1.9
11.76
(»=8)
8.2±0.8d
107 ± 27d
22.4 ± 1.0
41.4 ±2.3
18.09
(»=8)
5.1±0.3d
74 ± 14d
19.3 ±3.3
32.3±2.6d
86.41
(» = 8)
2.2 ± 0.3d
22±8d
16.3±1.5d
33.6±2.2d
250.16
(»=8)
0.6±0.2d
3±ld
10.3±1.7d
25.5±2.7d
aLOAEL identified.
b From the Emond PBPK model described in Section 3.3.
0 Values are the mean ± SE. Data obtained from Table 3 in van Birgelen et al. (1995)
d Statistically significant as compared to control (p < 0.05).

FT4 = free thyroxine; TT4 = total thyroxine.
                                               G-14

-------
G.1.27. White et al. (1986)
Endpoint
CH50 (U/mL) in
female mice0
Administered dose (ng/kg-day)
0
10a
50
100
500
1,000
2,000
Internal dose (ng/kg blood)b
0
(» = 8)
91 ±5
1.09
(»=8)
54±3d
4.08
(» = 8)
63±4d
7.14
(»=8)
56±9d
26.81
(»=8)
41±6d
48.72
(» = 8)
32±6d
90.56
(»=8)
17±6d
aLOAEL identified.
b From the Emond PBPK model described in Section 3.3.
0 Values are the mean ± SE. Data obtained from Table 1 in White et al. (1986).
d Statistically significant as compared to control (p < 0.05).
G.2.  ALTERNATE DOSE: WHOLE BLOOD BMDS RESULTS

G.2.1.  Amin et al. (2000): 0.25% Saccharin Consumed, Female

G.2.1.1.  Summary Table of BMDS Modeling Results
Model3
Linear0
Polynomial, 2 -degree
Power
Power, unrestrictedd
Degrees of
freedom
1
1
1
0
X2/7-value
0.551
0.551
0.551
N/A
AIC
179.214
179.214
179.214
180.858
BMD (ng/kg)
9.147E+00
9.147E+00
9.147E+00
8.367E+00
BMDLb
(ng/kg)
6.094E+00
6.094E+00
6.094E+00
3.419E+00
Notes


power bound hit
(power =1)
unrestricted
(power = 0.736)
a Nonconstant variance model selected (p = 0.0005).
b BMDL = Benchmark Dose Level.
0 Best-fitting model, BMDS output presented in this appendix.
d Alternate model, BMDS output also presented in this appendix.
G.2.1.2.   Output for Selected Model: Linear

Amin et al. (2000): 0.25% Saccharin Consumed, Female
         Polynomial Model.  (Version:  2.13;  Date:  04/08/2008)
         Input Data File: C:\l\Blood\l_Amin_2000_25_SC_Linear_l.(d)
         Gnuplot Plotting File:   C:\l\Blood\l_Amin_2000_25_SC_Linear_l.plt
                                               Mon  Feb 08  10:44:22  2010
                                          G-15

-------
   The form of the response function is:

   Y[dose]  = beta 0 + beta l*dose + beta 2*doseA2 + ...
   Dependent variable = Mean
   Independent variable = Dose
   Signs of the polynomial coefficients are not restricted
   The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  * rho)

   Total number of dose groups = 3
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =      5.29482
                            rho =            0
                         beta_0 =      31.5112
                         beta 1 =     -1.97726
           Asymptotic Correlation Matrix of Parameter Estimates

                 lalpha          rho       beta_0       beta_l

    lalpha            1        -0.99       -0.029        0.044

       rho        -0.99            1        0.026        -0.04

    beta_0       -0.029        0.026            1        -0.94

    beta 1        0.044        -0.04        -0.94            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha         -2.54215
0.692726
            rho          2.40985
3.4717
         beta_0          31.2644
39.4823
         beta_l          -1.9414
-1.08672
Parameter Estimates



       Std.  Err.

         1.65048

        0.541771

          4.1929

        0.436071
   95.0% Wald

Lower Conf.  Limit

       -5.77702

        1.34799

        23.0464

       -2.79609
     Table of Data and Estimated Values of Interest
                                     G-16

-------
 Dose
Res .
                 Obs Mean
                              Est Mean
                              Obs Std Dev  Est Std Dev
                                     Scaled
    0    10
3.378    10
10.57    10
       31.7
       24.6
       10.7
31.3
24.7
10.8
20.6
  12
5.33
17.8
13.4
4.91
 0.0727
-0.0264
-0.0362
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -92.841935
              -85.255316
              -85.429148
              -85.606998
              -98.136607
          # Param's
                4
                6
                5
                4
                2
            AIC
         193.683870
         182.510632
         180.858295
         179.213995
         200.273213
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
-2*log(Likelihood Ratio)  Test df
            25.7626
            15.1732
           0.347663
             0.3557
                     p-value

                    <.0001
                 0.0005072
                    0.5554
                    0.5509
                                     G-17

-------
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is greater than  .1.  The model chosen seems
to adequately describe the data
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD =        9.14709
            BMDL =
                          6.09414
                                     G-18

-------
G.2.1.3.  Figure for Selected Model: Linear
                             Linear Model with 0.95 Confidence Level
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 OJ
                                                                       10
                                           dose
   10:4402/082010
G.2.1.4.  Output for Additional Model Presented: Power, Unrestricted

Amin et al. (2000): 0.25% Saccharin Consumed, Female
         Power Model.  (Version:  2.15;   Date: 04/07/2008)

         Input Data File: C:\l\Blood\l_Amin_2000_25_SC_Pwr_U_l.(d)

         Gnuplot Plotting File:   C:\l\Blood\l_Amin_2000_25_SC_Pwr_U_l.plt

                                            Mon  Feb  08  10:44:22 2010
   The form of the response  function is:



   Y[dose]  = control + slope  *  doseApower
   Dependent  variable = Mean
                                       G-19

-------
Independent variable = Dose
The power is not restricted
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  * rho)

Total number of dose groups = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
               Default Initial Parameter Values
                      lalpha =      5.29482
                         rho =            0
                     control =      31.6727
                       slope =      -2.2195
                       power =     0.952715
        Asymptotic Correlation Matrix of Parameter Estimates

lalpha
rho
control
slope
power
lalpha
1
-0.99
0.34
-0.17
-0.061
rho
-0.99
1
-0.42
0.19
0.068
control
0
-0

-0
-0
.34
.42
1
.72
.56
slope
-0.
0.
-0.

0.
17
19
72
1
97
power
-0.061
0.068
-0.56
0.97
1
                              Parameter Estimates
Confidence Interval
Variable
Estimate
Std. Err.
Upper Conf. Limit

1.

3.

43

3.

1.

60693

74094

.5886

61299

42318
lalpha

rho

control

slope

power

-2.48291

2.38455

32.99

-3.91099

0.735877

2.08669

0.692047

5.40754

3.83883

0.350669

                                                      95.0% Wald

                                                   Lower Conf. Limit

                                                          -6.57274

                                                           1.02817

                                                           22.3914

                                                           -11.435

                                                         0.0485775
  Table of Data and Estimated Values of Interest
                                  G-20

-------
 Dose
Res .
                 Obs Mean
                              Est Mean
                              Obs Std Dev  Est Std Dev
                                     Scaled
    0    10
3.378    10
10.57    10
       31.7
       24.6
       10.7
  33
23.4
10.8
20.6
  12
5.33
18.7
12.4
4.94
-0.223
 0.302
 -0.08
 Warning: Likelihood for fitted model larger than  the  Likelihood  for  model
A3.
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -92.841935
              -85.255316
              -85.429148
              -85.429148
              -98.136607
          # Param's
                4
                6
                5
                5
                2
            AIC
         193.683870
         182.510632
         180.858295
         180.858295
         200.273213
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among  Dose  levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs.  fitted)
 (Note:  When rho=0 the results of Test 3 and Test  2  will  be  the  same.

                     Tests of Interest
   Test

   Test 1
   Test 2
-2*log(Likelihood Ratio)  Test df
            25.7626
            15.1732
                     p-value

                    <.0001
                 0.0005072
                                     G-21

-------
   Test 3             0.347663          1          0.5554
   Test 4         -8.2423e-013          0              NA

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 8.36678


            BMDL = 3.41906
                                     G-22

-------
G.2.1.5.  Figure for Additional Model Presented: Power, Unrestricted
                                 Power Model with 0.95 Confidence Level
  c
  o
  Q.
  (/)
  0)
 o:

  c
  (0
  0)
50




45




40




35




30




25




20




15




10
                   Power
                                 BMDL
                                                             BMD
                                         4            6

                                                dose
                                                                       10
   10:4402/082010
G.2.2.  Amin et al. (2000): 0.25% Saccharin Preference Ratio, Female

G.2.2.1.  Summary Table ofBMDS Modeling Results
Model3
Linear1"
Polynomial, 2 -degree
Power
Degrees of
freedom
1
1
1
X2 /7-value
0.002
0.002
0.002
AIC
227.807
227.807
227.807
BMD (ng/kg)
1.162E+01
1.162E+01
1.162E+01
BMDL (ng/kg)
5.572E+00
5.572E+00
5.572E+00
Notes


power bound hit
(power =1)
a Nonconstant variance model selected (p = 0.0135).

b Best-fitting model, BMDS output presented in this appendix.
                                            G-23

-------
G.2.2.2.   Output for Selected Model: Linear
Amin et al. (2000): 0.25% Saccharin Preference Ratio, Female
        Polynomial Model.  (Version: 2.13;   Date:  04/08/2008)
        Input Data File: C:\l\Blood\2_Amin_2000_25_SP_Linear_l.(d)
        Gnuplot Plotting File:  C:\l\Blood\2_Amin_2000_25_SP_Linear_l.plt
                                          Mon  Feb 08  10:44:49  2010
   The form of the response function is:

   Y[dose] = beta 0 + beta l*dose + beta  2*doseA2  +  ...
   Dependent variable = Mean
   Independent variable = Dose
   Signs of the polynomial coefficients  are  not  restricted
   The variance is to be modeled as Var(i) = exp(lalpha  + log(mean(i))  * rho)

   Total number of dose groups = 3
   Total number of records with missing  values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set to:  le-008
   Parameter Convergence has been set to:  le-008
                  Default Initial Parameter Values
                         lalpha =       6.34368
                            rho =             0
                         beta_0 =       75.4888
                         beta 1 =     -2.24733
           Asymptotic Correlation Matrix  of  Parameter  Estimates

                 lalpha          rho       beta_0        beta_l

    lalpha            1           -1          0.22         -0.31

       rho           -1             1         -0.22          0.31

    beta_0         0.22        -0.22             1         -0.77

    beta_l        -0.31         0.31         -0.77             1



                                 Parameter Estimates
                                     G-24

-------
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha          3.00523
21.0608
            rho         0.797764
5.13199
         beta_0          75.1087
88.3249
         beta_l         -2.16469
-0.188553
                Std. Err.

                   9.2122

                  2.21138

                  6.74312

                  1.00825
   95.0% Wald

Lower Conf. Limit

       -15.0503

       -3.53646

        61.8924

       -4.14082
     Table of Data and Estimated Values of Interest
Dose
Res .
0
3.378
10.57
N
10
10
10
Obs
82
58
54
Mean
.1
.1
.9
Est
75
67
52
Mean
.1
.8
.2
Obs St
13.
33.
19.
d Dev
3
9
5
Est S
25
24
21
td Dev
.2
.2
.8
Seal
0
0
ed
.884
1.27
.383
 Model Descriptions for likelihoods calculated
 Model Al:        Yij
           Var{e(ij) }

 Model A2:        Yij
           Var{e(ij) }
Mu(i)  + e(i j '
SigmaA2

Mu(i)  + e(i j '
Sigma(i)A2
 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
Model
Al
A2
A3
fitted
R
Log (likelihood)
-108.574798
-104.269377
-105.147952
-109.903705
-112.382522
# Param' s
4
6
5
4
2
AIC
225.149597
220.538754
220.295903
227.807410
228.765045
                   Explanation of Tests

                                     G-25

-------
 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.
                     Tests of Interest
           -2*log(Likelihood Ratio)  Test df
p-value
16.2263
8.61084
1.75715
9.51151
4
2
1
1
0.00273
0.0135
0.185
0.002042
Test

Test 1
Test 2
Test 3
Test 4
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD =        11.6241
            BMDL =
                          5.57215
                                     G-26

-------
G.2.2.3.  Figure for Selected Model: Linear
                                  Linear Model with 0.95 Confidence Level
  c
  o
  Q.
  (/)
  0)
 o:

  c
  (0
  OJ
         90
         80
         70
60
         50
         40
         30
                     Linear
                                    BMDL
BlvH)
                                                    6

                                                 dose
                                                                   10
 12
   10:4402/082010
G.2.3.  Amin et al. (2000): 0.50% Saccharin Consumed, Female


G.2.3.1.  Summary Table ofBMDS Modeling Results
Model3
Linear1"
Polynomial, 2 -degree
Power
Power, unrestricted0
Degrees of
freedom
1
1
1
0
X2 /7-value
0.060
0.060
0.060
N/A
AIC
158.591
158.591
158.591
157.060
BMD (ng/kg)
1.016E-K)!
1.016E+01
1.016E+01
6.567E+00
BMDL
(ng/kg)
6.567E-KJO
6.567E+00
6.567E+00
1.155E+00
Notes


power bound hit
(power =1)
unrestricted
(power =0.396)
a Nonconstant variance model selected (p = <0.0001).

b Best-fitting model, BMDS output presented in this appendix.

c Alternate model, BMDS output also presented in this appendix.
                                             G-27

-------
G.2.3.2.   Output for Selected Model: Linear
Amin et al. (2000): 0.50% Saccharin Consumed, Female
        Polynomial Model.  (Version: 2.13;  Date:  04/08/2008)
        Input Data File: C:\l\Blood\3_Amin_2000_50_SC_Linear_l.(d)
        Gnuplot Plotting File:  C:\l\Blood\3_Amin_2000_50_SC_Linear_l.plt
                                          Mon Feb  08  10:45:20  2010
   The form of the response function is:

   Y[dose] = beta 0 + beta l*dose + beta 2*doseA2 +  ...
   Dependent variable = Mean
   Independent variable = Dose
   Signs of the polynomial coefficients are not restricted
   The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  *  rho)

   Total number of dose groups = 3
   Total number of records with missing values =  0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =      4.68512
                            rho =            0
                         beta_0 =      20.0631
                         beta 1 =     -1.57142
           Asymptotic Correlation Matrix of Parameter Estimates

                 lalpha          rho       beta_0       beta_l

    lalpha            1        -0.96        0.019      -0.0016

       rho        -0.96            1       -0.031         0.015

    beta_0        0.019       -0.031             1         -0.96

    beta_l      -0.0016        0.015        -0.96             1



                                 Parameter Estimates
                                     G-28

-------
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
0.943084
            rho
2.90435
         beta_0
24.8462
         beta_l
-0.70108
 Estimate

-0.982115

  2.11808

  18.6171

 -1.33226
Std.  Err.

 0.982262

 0.401166

   3.1782

 0.322037
    95.0% Wald

 Lower Conf.  Limit

        -2.90731

         1.33181

         12.3879

        -1.96344
     Table of Data and Estimated Values of Interest
Dose
Res .
0
3.378
10.57
N
10
10
10
Obs Mean
22.4
11.4
4.54
Est Mean
18.6
14.1
4.54
Obs Std Dev
16
7.66
3.33
Est Std Dev
13.5
10.1
3.04
Scaled
0.873
-0.856
-0.00339
 Model Descriptions for likelihoods calculated
 Model Al:        Yij
           Var{e(ij) }

 Model A2:        Yij
           Var{e(ij) }
 Mu(i)  + e(i j '
 SigmaA2

 Mu(i)  + e(i j '
 Sigma(i)A2
 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
Log(likelihood)
  -83.696404
  -73.511830
  -73.530233
  -75.295363
  -90.294746
 # Param's
       4
       6
       5
       4
       2
   AIC
175.392808
159.023660
157.060467
158.590726
184.589492
                   Explanation of Tests

                                     G-29

-------
 Test 1:

 Test 2:
 Test 3:
 Test 4:
 (Note:
   Test
 Do responses and/or variances differ among Dose levels?
 (A2 vs. R)
 Are Variances Homogeneous?  (Al vs A2)
 Are variances adequately modeled?  (A2 vs. A3)
 Does the Model for the Mean Fit?  (A3 vs. fitted)
When rho=0 the results of Test 3 and Test 2 will be the same.
                     Tests of Interest
  -2*log(Likelihood Ratio)  Test df
p-value
Test 1
Test 2
Test 3
Test 4
33.5658
20.3691
0.0368066
3.53026
4
2
1
1
<.0001
<.0001
0.8479
0.06026
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD =        10.1633
            BMDL =
                          6.56742
                                     G-30

-------
 c
 o
 Q.
 o:

 c
 (0
 OJ
     20
     15
        10
G.2.3.3.  Figure for Selected Model: Linear



                             Linear Model with 0.95 Confidence Level


                  I inpar 	
       35





       30





       25
                                               BMDL
           0           2



10:4502/082010
4          6

      dose
                                                                      BMD
                                                                       10
G.2.3.4.  Output for Additional Model Presented: Power, Unrestricted


Amin et al. (2000): 0.50% Saccharin Consumed, Female
         Power Model. (Version:  2.15;   Date: 04/07/2008)

         Input Data File: C:\l\Blood\3_Amin_2000_50_SC_Pwr_U_l.(d)

         Gnuplot Plotting File:   C:\l\Blood\3_Amin_2000_50_SC_Pwr_U_l.plt

                                            Mon Feb  08  10:45:20 2010
   The form  of the response  function is:



   Y[dose] = control + slope  *  doseApower





   Dependent variable = Mean
                                       G-31

-------
Independent variable = Dose
The power is not restricted
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  * rho)

Total number of dose groups = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
               Default Initial Parameter Values
                      lalpha =      4.68512
                         rho =            0
                     control =      22.3564
                       slope =     -6.53901
                       power =     0.425213
        Asymptotic Correlation Matrix of Parameter Estimates
lalpha
lalpha
rho
control
slope
power

-0.
0.
-0.
-0.
1
96
34
31
15
rho
-0.

-0.
0.
0.
96
1
47
36
15
control
0
-0

-0
-0
.34
.47
1
.81
.52
slope
-0.
0.
-0.

0.
31
36
81
1
92
power
-0.
0.
-0.
0.

15
15
52
92
1
                              Parameter Estimates
Confidence Interval
Variable
Estimate
Std. Err.
Upper Conf. Limit

1.

2.

31

0.

0.

83541

99953

.4181

824743

726173
lalpha

rho

control

slope

power

-0.708629

1.96142

22.6293

-7.10123

0.395571

1.298

0.529653

4.48416

4.04394

0.168677

                                                      95.0% Wald

                                                   Lower Conf. Limit

                                                          -3.25267

                                                          0.923323

                                                           13.8405

                                                          -15.0272

                                                         0.0649698
  Table of Data and Estimated Values of Interest
                                  G-32

-------
 Dose
Res .
                 Obs Mean
                              Est Mean
                              Obs Std Dev  Est Std Dev
                                     Scaled
    0    10
3.378    10
10.57    10
       22.4
       11.4
       4.54
22.6
11.1
4.58
  16
7.66
3.33
  15
7.46
3.12
-0.0577
  0.105
-0.0475
Degrees of freedom for Test A3 vs fitted <= 0
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) +
           Var{e(ij)} = SigmaA2
 Model A2:
           Var{e(ij)} = Sigma(i)A2
 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e (i) } = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -83.696404
              -73.511830
              -73.530233
              -73.530233
              -90.294746
# Param' s
4
6
5
5
2
AIC
175.392808
159.023660
157.060467
157.060467
184.589492
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
-2*log(Likelihood Ratio)  Test df
            33.5658
            20.3691
          0.0368066
                     p-value

                    <.0001
                    <.0001
                    0.8479
                                     G-33

-------
   Test 4                    0          0              NA

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 6.56719


            BMDL = 1.15476
                                     G-34

-------
G.2.3.5.  Figure for Additional Model Presented: Power, Unrestricted
                                Power Model with 0.95 Confidence Level
  03
  i/)
  I
        35
        30
        25
20
         15
         10
                                                                              10
   10:4502/082010
G.2.4.  Amin et al. (2000): 0.50% Saccharin Preference Ratio, Female

G.2.4.1.  Summary Table ofBMDS Modeling Results
Model3
Linear1"
Polynomial, 2 -degree
Power
Power, unrestricted0
Degrees of
freedom
1
1
1
0
X2 /7-value
0.135
0.135
0.135
N/A
AIC
234.250
234.250
234.250
234.020
BMD (ng/kg)
8.144E-KJO
8.144E+00
8.144E+00
2.598E+00
BMDL
(ng/kg)
5.105E-K)0
5.105E+00
5.105E+00
1.057E-14
Notes


power bound hit
(power =1)
unrestricted
(power =0.282)
a Constant variance model selected (p = 0.5593).
b Best-fitting model, BMDS output presented in this appendix.
c Alternate model, BMDS output also presented in this appendix.
                                             G-35

-------
G.2.4.2.   Output for Selected Model: Linear
Amin et al. (2000): 0.50% Saccharin Preference Ratio, Female
        Polynomial Model.  (Version: 2.13;  Date:  04/08/2008)
        Input Data File: C:\l\Blood\4_Amin_2000_50_SP_LinearCV_l.(d)
        Gnuplot Plotting File:  C:\l\Blood\4_Amin_2000_50_SP_LinearCV_l.plt
                                          Mon  Feb 08  10:45:50  2010
   The form of the response function is:

   Y[dose] = beta 0 + beta l*dose + beta 2*doseA2  +  ...
   Dependent variable = Mean
   Independent variable = Dose
   rho is set to 0
   Signs of the polynomial coefficients are not  restricted
   A constant variance model is fit

   Total number of dose groups = 3
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      764.602
                            rho =             0    Specified
                         beta_0 =      65.8627
                         beta 1 =     -3.34297
           Asymptotic Correlation Matrix of Parameter  Estimates

            ( *** The model parameter(s)  -rho
                 have been estimated at a boundary point,  or  have  been
specified by the user,
                 and do not appear in the correlation  matrix  )

                  alpha       beta_0       beta_l

     alpha            1     2.6e-008     2.1e-009

    beta_0     2.6e-008            1        -0.73

    beta 1     2.1e-009        -0.73             1
                                     G-36

-------
                                 Parameter Estimates
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          741.255
1116.38
         beta_0          65.8627
80.0239
         beta_l         -3.34297
-1.13183
                 Std. Err.

                   191.391

                   7.22524

                   1.12815
                95.0% Wald

             Lower Conf. Limit

                     366.135

                     51.7015

                    -5.55412
     Table of Data and Estimated Values of Interest
Dose
Res .
0
3.378
10.57
N
10
10
10
Obs
72
44
33
Mean
.7
.5
.8
Est
65
54
30
Mean
.9
.6
.5
Obs S
24
32
24
td Dev
.6
.9
.6
Est S
27
27
27
td Dev
.2
.2
.2
Seal
0
0
ed
.797
1.17
.375
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
Log(likelihood)
 -113.009921
 -112.428886
 -113.009921
 -114.125184
 -117.976057
# Param's
      4
      6
      4
      3
      2
   AIC
234.019841
236.857773
234.019841
234.250368
239.952114
                                     G-37

-------
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.
                     Tests of Interest
           -2*log(Likelihood Ratio)  Test df
        p-value
11.0943
1.16207
1.16207
2.23053
4
2
2
1
0.02552
0.5593
0.5593
0.1353
Test

Test 1
Test 2
Test 3
Test 4
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
The modeled variance appears
The model chosen seems
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD =        8.14425
            BMDL =
                          5.10523
                                     G-38

-------
G.2.4.3.  Figure for Selected Model: Linear
                             Linear Model with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 HI
90


80


70


60


50


40


30


20


10
                  Linear
                                       BMDL
BMD
                                     4          6
                                           dose
                                                                10
   10:4502/082010
G.2.4.4.  Output for Additional Model Presented: Power, Unrestricted
Amin et al. (2000): 0.50% Saccharin Preference Ratio, Female
         Power Model.  (Version:  2.15;  Date:  04/07/2008)
         Input Data File: C:\l\Blood\4_Amin_2000_50_SP_PwrCV_U_l.(d)
         Gnuplot Plotting File:   C:\l\Blood\4_Amin_2000_50_SP_PwrCV_U_l.plt
                                            Mon  Feb 08 10:45:50 2010
   The form of the response  function is:

   Y[dose]  = control + slope  *  doseApower


   Dependent variable = Mean
                                       G-39

-------
   Independent variable = Dose
   rho is set to 0
   The power is not restricted
   A constant variance model is fit

   Total number of dose groups = 3
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      764.602
                            rho =
                        control =
                          slope =
                          power =
            0
      72.7273
     -20.0402
     0.281985
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
control
slope
power


-1
-1
-2
alpha
1
.2e-009
.2e-009
.2e-010
control
-1.2e-009
1
-0.51
-0.22
slope
-1.2e-009
-0.51
1
0.92
power
-2.2e-010
-0.22
0.92
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          688.142
1036.38
        control          72.7273
88.986
          slope         -20.0402
9.47219
          power         0.281985
0.920661
Parameter Estimates



       Std.  Err.

         177.677

         8.29543

         15.0576

        0.325861
        95.0% Wald

     Lower Conf.  Limit

               339.9

             56.4686

            -49.5526

            -0.35669
                                     G-40

-------
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est  Std  Dev    Scaled
    0    10
3.378    10
10.57    10
       72.7
       44.5
       33.8
72.7
44.5
33.8
24.6
32.9
24.6
26.2
26.2
26.2
4.67e-009
1.52e-008
1.77e-008
 Warning: Likelihood for fitted model larger than the Likelihood  for model
A3.
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -113.009921
             -112.428886
             -113.009921
             -113.009921
             -117.976057
# Param's
      4
      6
      4
      4
      2
   AIC
234.019841
236.857773
234.019841
234.019841
239.952114
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose  levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs.  fitted)
 (Note:  When rho=0 the results of Test 3 and Test  2 will  be  the same.

                     Tests of Interest
   Test
-2*log(Likelihood Ratio)  Test df
           p-value
                                     G-41

-------
   Test 1              11.0943          4         0.02552
   Test 2              1.16207          2          0.5593
   Test 3              1.16207          2          0.5593
   Test 4        -2.84217e-014          0              NA

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.  The modeled variance appears
 to be appropriate here

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 2.59831


            BMDL = 1.05661e-014
                                     G-42

-------
G.2.4.5.   Figure for Additional Model Presented: Power, Unrestricted
 OJ
 t/)

 o
 Q.
 o:

 c
 (0
 HI
          10
   10:4502/082010
                                 Power Model with 0.95 Confidence Level
                                                                               10
                                                dose
                                           G-43

-------
G.2.5.  Bell et al. (2007): Balano-Preputial Separation, Postnatal Day (PND) 49
G.2.5.1.  Summary Table ofBMDS Modeling Results
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage, 3 -degree
Probit
Weibull
Gamma, unrestricted
Log-logistic,
unrestricted13
Log-probit,
unrestricted
Weibull, unrestricted
Degrees of
freedom
2
2
2
2
2
2
2
1
1
1
1
X2 p-value
0.684
0.342
0.777
0.269
0.684
0.367
0.684
0.566
0.501
0.456
0.551
AIC
112.136
113.915
111.908
114.254
112.136
113.713
112.136
113.746
113.871
113.977
113.771
BMD (ng/kg)
2.867E+00
6.159E+00
2.246E+00
5.322E+00
2.867E+00
5.715E+00
2.867E+00
1.862E+00
1.998E+00
2.038E+00
1.914E+00
BMDL
(ng/kg)
1.943E+00
4.746E+00
1.394E-KJO
3.512E+00
1.943E+00
4.422E+00
1.943E+00
1.829E-01
2.795E-01
3.250E-01
2.346E-01
Notes
power bound hit (power =1)
negative intercept
(intercept = -2.246)
slope bound hit (slope = 1)
slope bound hit (slope =1)
final B = 0

power bound hit (power =1)
unrestricted (power = 0.741)
unrestricted (slope = 0.93)
unrestricted (slope = 0.54)
unrestricted (power = 0.795)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.
G.2.5.2.  Output for Selected Model: Log-Logistic
Bell et al. (2007): Balano-Preputial Separation, PND 49
         Logistic Model.  (Version:  2.12; Date: 05/16/2008)
         Input Data File: C:\l\Blood\5_Bell_2007_BPS_LogLogistic_l.(d)
         Gnuplot Plotting File:   C:\l\Blood\5_Bell_2007_BPS_LogLogistic_l.plt
                                            Mon Feb  08  10:46:18 2010
   The  form of the probability  function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
   Dependent variable = DichEff
   Independent variable = Dose
   Slope  parameter is restricted as slope >= 1

   Total  number of observations  = 4
   Total  number of records with  missing values =  0
   Maximum number of iterations  = 250
   Relative Function Convergence has been set to:  le-008
   Parameter Convergence has  been set to: le-008
                                       G-44

-------
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =    0.0333333
                      intercept =     -2.99896
                          slope =            1
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -slope
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

             background    intercept

background            1        -0.49

 intercept        -0.49            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background         0.038005
*
      intercept         -3.00658
*
          slope                1
                                 Parameter Estimates
                      Std.  Err.
                 95.0% Wald

              Lower Conf.  Limit
* - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.7817
  Reduced model
0.0001309

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -53.7077         4
      -53.954         2      0.492596      2
     -63.9797
      111.908
1
20.544
3
                            P-value
                                  Goodness  of  Fit

                                     G-45

-------

Dose
0.0000
2.2040
5.1378
18.4110

Est. Prob.
0.0380
0.1326
0.2329
0.4965

Expected
1.140
3.977
6.988
14.895

Observed
1.000
5.000
6.000
15.000

Size
30
30
30
30
Scaled
Residual
-0.134
0.551
-0.427
0.038
 ChiA2 = 0.50      d.f. = 2        P-value = 0.7769







   Benchmark Dose Computation




Specified effect =            0.1




Risk Type        =      Extra risk




Confidence level =           0.95




             BMD =        2.24647




            BMDL =        1.39385
                                     G-46

-------
G.2.5.3.  Figure for Selected Model: Log-Logistic
                            Log-Logistic Model with 0.95 Confidence Level
 •
 I
 C
 o
 13
 ro
         0.7
         0.6
         0.5
0.4
0.3
         0.2
         0.1
          0  -
   10:4602/082010
                                                                15
                                            dose
G.2.5.4.  Output for Additional Model Presented: Log-Logistic, Unrestricted
Bell et al. (2007): Balano-Preputial Separation, PND 49
         Logistic Model.  (Version: 2.12; Date:  05/16/2008)
         Input Data File:  C:\l\Blood\5_Bell_2007_BPS_LogLogistic_U_l.(d)
         Gnuplot Plotting  File:
C:\l\Blood\5_Bell_2007_BPS_LogLogistic_U_l.plt
                                             Mon  Feb 08 10:46:18  2010
   The  form of the probability function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
                                       G-47

-------
   Dependent variable = DichEff
   Independent variable = Dose
   Slope parameter is not restricted

   Total number of observations = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =    0.0333333
                      intercept =     -2.68464
                          slope =     0.858398
           Asymptotic Correlation Matrix of Parameter Estimates

             background    intercept        slope

background            1        -0.48         0.35

 intercept        -0.48            1        -0.94

     slope         0.35        -0.94            1
Confidence Interval
       Variable
Upper Conf. Limit
     background
*
      intercept
*
          slope
               Parameter Estimates

                                       95.0% Wald

      Estimate        Std.  Err.     Lower Conf. Limit

     0.0353402            *                *

      -2.84051            *                *

      0.929645            *                *
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.4997
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -53.7077         4
     -53.9354         3      0.455534      1
                                     G-48
P-value

-------
  Reduced model
0.0001309

           AIC:
    -63.9797
     113.871
              20.544
                                  Goodness  of  Fit

Dose
0.0000
2.2040
5.1378
18.4110

Est. Prob.
0.0353
0.1400
0.2389
0.4858

Expected
1.060
4.201
7.166
14.573

Observed
1.000
5.000
6.000
15.000

Size
30
30
30
30
Scaled
Residual
-0.060
0.420
-0.499
0.156
 ChiA2 = 0.45
d.f.  =1
P-value = 0.5005
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        1.99765

            BMDL =       0.279534
                                     G-49

-------
 •

 I
 C
 o
 •*=
 o
 (0
G.2.5.5.  Figure for Additional Model Presented: Log-Logistic, Unrestricted



                                Log-Logistic Model with 0.95 Confidence Level


               : '  '  '       Tc

          0.7




          0.6




          0.5
0.4
0.3
          0.2
          0.1
            0   -

              EMDL
   10:4602/082010
                                                                        15
                                                  dose
                                             G-50

-------
G.2.6.  Cantoni et al. (1981): Urinary Coproporhyrins, 3 Months
G.2.6.1.  Summary Table ofBMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Power
Power, unrestricted0
Hill, unrestricted
Degrees of
freedom
2
2
1
1
1
2
2
2
1
0
X2 p-value
0.003
0.003
0.486
0.486
0.788
0.005
0.005
0.005
0.610
N/A
AIC
32.882
32.882
23.459
23.459
23.047
31.595
31.595
31.595
23.235
24.974
BMD (ng/kg)
3.209E+01
3.209E+01
5.339E-01
5.339E-01
4.333E-01
1.464E+01
1.464E+01
1.464E+01
2.766E-02
2.602E-01
BMDL
(ng/kg)
1.567E+01
1.567E+01
1.803E-01
1.803E-01
error
2.753E+00
2.753E+00
2.753E+00
2.031E-05
error
Notes

power hit bound (d=\)

power hit bound (d=\)
n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted
(power =0.304)
unrestricted (n = 0.739)
a Nonconstant variance model selected (p = 0.0039).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.2.6.2.  Output for Selected Model: Exponential (M4)
Cantoni et al. (1981): Urinary Coproporhyrins, 3 Months
         Exponential Model.  (Version: 1.61;   Date:  7/24/2009)
         Input Data File:  C:\l\Blood\6_Cantoni_1981_UriCopro_Exp_l.(d)
         Gnuplot Plotting  File:
                                            Mon Feb 08 10:46:46  2010
 Figurel-UrinaryCoproporphyrin Smonths
   The  form of the response function by Model:
       Model 2:
       Model 3:
       Model 4:
       Model 5:
Y[dose]
Y[dose]
Y[dose]
Y[dose]
a
a
a
a
expfsign  *  b  *  dose}
exp{sign  *  (b * dose)Ad}
[c-(c-l)  *  exp{-b * dose}]
[c-(c-l)  *  exp{-(b * dose)Ad}]
    Note:  Y[dose] is the  median response  for exposure = dose;
           sign = +1 for increasing trend  in  data;
           sign = -1 for decreasing trend.

       Model 2 is nested within Models 3 and  4.
       Model 3 is nested within Model 5.
       Model 4 is nested within Model 5.
   Dependent variable = Mean
   Independent variable =  Dose
                                       G-51

-------
Data are assumed to be distributed: normally
Variance Model: exp(lnalpha +rho *ln(Y[dose]))
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  * rho)

Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to:  le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 4

                 Inalpha             -1.50063
                     rho              2.60979
                       a             0.704303
                       b            0.0604961
                       c              4.47268
                       d                    1
                  Parameter Estimates

                Variable          Model 4

                 Inalpha            -1.75302
                     rho              2.6322
                       a            0.761218
                       b            0.241561
                       c             4.15597
                       d                   1


         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0      4
  1.847      4
  8.839      4
  50.05      4
               Estimated Values of Interest

   Dose      Est Mean      Est Std     Scaled Residual
0.7414
1.807
2.734
3
0.3475
0.8341
1.506
2.6
      0        0.7612       0.2907          -0.1366
  1.847         1.626       0.7892           0.4588
  8.839          2.88        1.674          -0.1743
  50.05         3.164        1.895          -0.1725


                                  G-52

-------
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i))  *  rho)

     Model  R:        Yij = Mu + e(i)
               Var{e(ij)} = SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
4
-12.90166
-6.203643
-6.487204
-15.73713
-6.729737
5
8
6
2
5
35.80333
28.40729
24.97441
35.47427
23.45947
   Additive constant for all log-likelihoods =      -14.7.   This  constant
added to the
   above values gives the log-likelihood including the  term  that  does  not
   depend on the model parameters.
R)
                              Explanation of Tests

Test 1:  Does response and/or variances differ among Dose  levels?  (A2  vs.

Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)

Test 6a: Does Model 4 fit the data?  (A3 vs  4)
     Test

     Test 1
     Test 2
     Test 3
    Test 6a
                         Tests of Interest

                 -2*log(Likelihood Ratio)
                                                  D. F.
p-value
19.07
13.4
0.5671
0.4851
6
3
2
1
0.004052
0.003854
0.7531
0.4861
     The p-value for Test 1 is less than  .05.  There appears  to be  a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

                                     G-53

-------
  The p-value for Test 2 is less than .1.  A non-homogeneous
  variance model appears to be appropriate.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is greater than  .1.  Model 4 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =     0.533855

              BMDL =     0.180293
                                  G-54

-------
G.2.6.3.  Figure for Selected Model: Exponential (M4)
                          Exponential Model 4 with 0.95 Confidence Level
 OJ
 c/)

 o
 Q.
 (/)
 
-------
   Dependent variable = Mean
   Independent variable = Dose
   The power is not restricted
   The variance is to be modeled as Var(i)
          = exp(lalpha + log(mean(i))  * rho)
   Total number of dose groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =      0.90039
                            rho =            0
                        control =     0.741372
                          slope =      0.93685
                          power =     0.224904
           Asymptotic Correlation Matrix of Parameter Estimates

lalpha
rho
control
slope
power
lalpha
1
-0.62
-0.53
-0.036
0.024
rho
-0.62
1
0.43
-0.2
-0.16
control
-0.53
0.43
1
-0.28
0.086
slope
-0.036
-0.2
-0.28
1
-0.77
power
0.024
-0.16
0.086
-0.77
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha         -1.78125
-0.570373
            rho          2.64332
4.10338
        control          0.75678
1.03113
          slope         0.845767
1.48247
          power         0.304211
0.568909
Parameter Estimates



       Std.  Err.

        0.617807

        0.744946

        0.139979

        0.324854

        0.135053
   95.0% Wald

Lower Conf. Limit

       -2.99213

        1.18325

       0.482426

       0.209065

      0.0395119
     Table of Data and Estimated Values of Interest

                                     G-56

-------
 Dose
Res .
                 Obs Mean
                              Est Mean
                              Obs Std Dev  Est Std Dev
                                      Scaled
    0
1.847
8.839
50.05
      0.741
       1.81
       2.73
          3
0.757
 1.78
  2.4
 3.54
0.348
0.834
 1.51
  2.6
0.284
0.877
  1.3
 2.18
-0.109
0.0705
 0.515
-0.493
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -12.901663
               -6.203643
               -6.487204
               -6.617347
              -15.737135
           # Param's
                 5
                 8
                 6
                 5
                 2
             AIC
           35.803325
           28.407287
           24.974409
           23.234694
           35.474269
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
-2*log(Likelihood Ratio)  Test df
             19.067
             13.396
           0.567122
          6
          3
          2
         p-value

      0.004052
      0.003854
        0.7531
                                     G-57

-------
   Test 4
0.260285
0.6099
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is greater than  .1.  The model chosen seems
to adequately describe the data
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 0.0276599


            BMDL = 2.03143e-005
                                     G-58

-------
G.2.6.5.   Figure for Additional Model Presented: Power, Unrestricted



                               Power Model with 0.95 Confidence Level
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 OJ
 5



 4



 3



 2



 1



 0



-1
                Power
        BI/IDL|BMD
                           10
                                 20            30

                                       dose
40
50
   10:4602/082010
                                           G-59

-------
G.2.7.  Cantoni et al. (1981): Urinary Porphyrins
G.2.7.1.  Summary Table ofBMDS Modeling Results
Model3
Exponential (M2)b
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Power
Degrees of
freedom
2
2
1
0
0
2
1
1
X2 p-value
0.001
0.001
0.0001
N/A
N/A
0.001
0.0001
0.0001
AIC
55.465
55.465
59.187
61.084
62.199
57.187
10.000
59.084
BMD (ng/kg)
3.760E+00
3.760E+00
2.484E-01
2.878E-01
6.233E+00
2.484E-01
error
2.878E-01
BMDL
(ng/kg)
2.762E+00
2.762E+00
1.448E-01
1.461E-01
3.341E+00
1.448E-01
error
1.461E-01
Notes

power hit bound (d=l)






a Nonconstant variance model selected (p = O.0001).
b Best-fitting model, BMDS output presented in this appendix.
G.2.7.2.  Output for Selected Model: Exponential (M2)
Cantoni et al. (1981): Urinary Porphyrins
         Exponential Model.  (Version:  1.61;   Date: 7/24/2009)
         Input Data File: C:\l\Blood\7_Cantoni_1981_UriPor_Exp_l.(d)
         Gnuplot Plotting File:
                                            Mon Feb 08 10:47:24  2010
 Table  1,  dose converted to ng per  kg  per day
   The  form of the response function  by Model:
      Model 2:      Y[dose] = a  *  exp{sign * b * dose}
      Model 3:      Y[dose] = a  *  exp{sign * (b * dose)Ad}
      Model 4:      Y[dose] = a  *  [c-(c-l)  * exp{-b * dose}]
      Model 5:      Y[dose] = a  *  [c-(c-l)  * exp{-(b * dose)Ad}]

    Note:  Y[dose]  is the median response for exposure = dose;
           sign = +1 for increasing  trend in data;
           sign = -1 for decreasing  trend.

      Model 2  is nested within  Models 3 and 4.
      Model 3  is nested within  Model  5.
      Model 4  is nested within  Model  5.
   Dependent variable = Mean
   Independent variable = Dose
   Data  are  assumed to be distributed:  normally
   Variance  Model:  exp(lnalpha +rho  *ln(Y[dose]))
   The variance is  to be modeled  as  Var(i)  = exp(lalpha +  log(mean(i))  * rho)

   Total number of  dose groups =  4
                                       G-60

-------
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 2

                 Inalpha             -3.57509
                     rho              2.23456
                       a              3.36453
                       b            0.0819801
                       c                    0
                       d                    1
                  Parameter Estimates

                Variable          Model 2

                 Inalpha          -1.85879
                     rho           1.82273
                       a           3.57896
                       b         0.0803347
                       c                 0
                       d                 1


         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0      4
  1.847      4
  8.839      3
  50.05      3
               Estimated Values of Interest

   Dose      Est Mean      Est Std     Scaled Residual
2.27
5.55
7.62
196.9
0.49
0.85
1.79
63.14
      0         3.579        1.262           -2.074
  1.847         4.152        1.445            1.936
  8.839          7.28         2.41           0.2441
  50.05         199.5        49.25         -0.09069
Other models for which likelihoods are calculated:

  Model Al:        Yij = Mu(i) + e(ij)

                                  G-61

-------
               Var{e (ij) }  = SigmaA2
     Model A2:         Yij
               Var{e(ij) }

     Model A3:         Yij
               Var{e(ij) }

     Model  R:         Yij
               Var{e(ij) }
                         Mu(i)  + e(i j)
                         Sigma(i)A2

                         Mu(i)  + e (i j)
                         exp(lalpha + log(mean(i)) * rho)

                         Mu + e(i)
                         SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
Al
A2
A3
R
2
-51.42175
-15.31211
-15.66963
-68.75058
-23.73254
                                                                AIC
5
8
6
2
4
112
46.
43.
141
55.
.8435
62422
33925
.5012
46509
   Additive constant for all log-likelihoods =     -12.87.  This constant
added to the
   above values gives the log-likelihood including the term that does not
   depend on the model parameters.
R)
Test 1:

Test 2:
Test 3:
Test 4:
                     Explanation of Tests

Does response and/or variances differ among Dose levels?  (A2 vs.

Are Variances Homogeneous?  (A2 vs. Al)
Are variances adequately modeled?  (A2 vs. A3)
Does Model 2 fit the data?  (A3 vs. 2)
     Test

     Test 1
     Test 2
     Test 3
     Test 4
                Tests of Interest

        -2*log(Likelihood Ratio)

                        106.9
                        72.22
                        0.715
                        16.13
D.  F.

  6
  3
  2
  2
                                                             p-value
                                                              < 0.0001
                                                              < 0.0001
                                                                0.6994
                                                              0.000315
     The p-value for Test 1 is less than .05.  There appears to be a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

     The p-value for Test 2 is less than .1.  A non-homogeneous
     variance model appears to be appropriate.
     The p-value for Test 3 is greater than  .1.
     variance appears to be appropriate here.
                                              The modeled
                                     G-62

-------
     The p-value for Test 4 is  less  than .1.  Model  2  may not adequately

     describe  the data; you may want to consider another model.
   Benchmark  Dose Computations:


     Specified Effect = 1.000000



            Risk Type = Estimated  standard deviations  from control


     Confidence Level = 0.950000


                   BMD =      3.75968



                  BMDL =      2.76247



G.2.7.3.  Figure for Selected Model: Exponential (M2)



                           Exponential Model 2 with 0.95 Confidence Level
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 0)
        350
        300
        250
200
150
        100
         50
                        Exponential
              BMDL BMD
                           10
                              20
30
40
50
                                           dose
   10:4702/082010
                                       G-63

-------
G.2.8.  Crofton et al. (2005): Serum, T4
G.2.8.1.  Summary Table ofBMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Hill
Linear
Polynomial, 8 -degree
Power
Power, unrestricted
Degrees of
freedom
8
8
7
6
6
8
8
8
7
X2 p-value
0.0001
0.0001
0.942
0.912
0.972
0.0001
0.0001
0.0001
0.018
AIC
516.356
516.356
476.449
478.234
477.450
522.460
522.460
522.460
491.101
BMD (ng/kg)
1.144E+02
1.144E+02
5.190E-H)0
5.757E+00
5.724E+00
2.406E+02
2.406E+02
2.406E+02
2.449E+00
BMDL
(ng/kg)
6.239E+01
6.239E+01
3.029E+00
3.094E+00
3.024E+00
1.761E+02
1.761E+02
1.761E+02
3.307E-01
Notes

power hit bound (d = 1)





power bound hit
(power =1)
unrestricted
(power = 0.243)
a Constant variance model selected (p = 0.7647).
b Best-fitting model, BMDS output presented in this appendix.
G.2.8.2.  Output for Selected Model: Exponential (M4)
Crofton et al. (2005): Serum, T4
         Exponential Model.  (Version:  1.61;  Date: 7/24/2009)
         Input Data File: C:\l\Blood\8_Crofton_2005_T4_ExpCV_l.(d)
         Gnuplot Plotting File:
                                            Mon Feb 08  10:48:04 2010
   The  form of the response  function by Model:
      Model 2:      Y[dose] = a  *  exp{sign * b * dose}
      Model 3:      Y[dose] = a  *  exp{sign *  (b * dose)Ad}
      Model 4:      Y[dose] = a  *  [c-(c-l) * exp{-b * dose}]
      Model 5:      Y[dose] = a  *  [c-(c-l) * exp{-(b *  dose)Ad}]

    Note:  Y[dose]  is the median response for exposure  =  dose;
           sign = +1 for increasing  trend in data;
           sign = -1 for decreasing  trend.

      Model 2 is nested within  Models 3 and 4.
      Model 3 is nested within  Model 5.
      Model 4 is nested within  Model 5.
   Dependent variable = Mean
   Independent variable = Dose
   Data  are assumed to be distributed:  normally
   Variance Model:  exp(lnalpha  +rho  *ln(Y[dose])

                                       G-64

-------
rho is set to 0.
A constant variance model is fit.

Total number of dose groups = 10
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values
               Variable

                 Inalpha
                     rho(S)
                       a
                       b
                       c
                       d
Model 4

     5.47437
           0
     104.999
  0.00641895
    0.445764
           1
   (S) = Specified
                  Parameter Estimates
                Variable

                 Inalpha
                     rho
                       a
                       b
                       c
                       d
 Model 4

    5.50623
          0
    100.332
   0.076678
   0.523626
          1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
0
0.0202
0.4882
1.384
3.455
9.257
23.07
65.65
180.9
583.5
14
6
12
6
6
6
6
6
6
4
100
96.27
98.57
99.76
93.32
70.94
62.52
52.68
54.66
49.15
15.44
14.98
18.11
19.04
12.11
12.74
14.75
22.73
19.71
11.15
               Estimated Values of Interest
                                  G-65

-------
      Dose
                Est Mean
                              Est Std
                       Scaled Residual
         0
    0.0202
    0.4882
     1.384
     3.455
     9.257
     23.07
     65.65
     180.9
     583.5
100.3
100.3
98.58
95.52
89.21
76.04
60.69
52.85
52.54
52.54
15.69
15.69
15.69
15.69
15.69
15.69
15.69
15.69
15.69
15.69
 -0.07952
  -0.6231
-0.000744
   0.6614
   0.6422
  -0.7962
   0.2854
 -0.02621
   0.3319
  -0.4323
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) *  rho)

     Model  R:        Yij = Mu + e(i)
               Var{e (ij)  } = SigmaA2
                     Model

                        Al
                        A2
                        A3
                         R
                         4
             Likelihoods of Interest

             Log(likelihood)      DF
              -233.0774
              -230.2028
              -233.0774
              -268.4038
              -234.2243
                                AIC
11
20
11
2
4
488.1549
500.4056
488.1549
540.8076
476.4486
   Additive constant for all log-likelihoods =     -66.16.   This  constant
added to the
   above values gives the log-likelihood including the term  that  does  not
   depend on the model parameters.
                                 Explanation of Tests

   Test 1:  Does response and/or variances differ among Dose  levels?  (A2  vs.
R)
   Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
   Test 3:  Are variances adequately modeled?  (A2 vs. A3)

   Test 6a: Does Model 4 fit the data?  (A3 vs  4)
                                     G-66

-------
76.4
5.749
5.749
2.294
18
9
9
7
< 0.0001
0.7647
0.7647
0.9418
                         Tests of Interest

  Test          -2*log(Likelihood Ratio)       D. F.         p-value

  Test 1
  Test 2
  Test 3
 Test 6a
  The p-value for Test 1 is less than .05.  There appears to be a
  difference between response and/or variances among the dose
  levels, it seems appropriate to model the data.

  The p-value for Test 2 is greater than  .1.  A homogeneous
  variance model appears to be appropriate here.

  The p-value for Test 3 is greater than  .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is greater than .1.  Model 4 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =      5.18983

              BMDL =      3.02894
                                  G-67

-------
G.2.8.3.  Figure for Selected Model: Exponential (M4)



                             Exponential Model 4 with 0.95 Confidence Level
 c
 o
 Q.
 o:

 c
 (0
 OJ
         120
         100
80
          60
          40
          20
                          Exponential
            BMDLBMD
100        200
                                                300

                                               dose
400        500
                                                                       600
   10:4802/082010
                                          G-68

-------
G.2.9.  Franc et al. (2001): Sprague-Dawley (S-D) Rats, Relative Liver Weight
G.2.9.1.  Summary Table ofBMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Powerb
Degrees of
freedom
2
1
1
0
0
2
1
1
/2p-value
0.968
0.880
0.580
N/A
N/A
0.858
0.935
0.839
AIC
234.369
236.327
236.610
238.346
238.346
234.610
236.311
236.346
BMD (ng/kg)
7.800E+00
9.201E+00
6.365E+00
9.474E+00
9.479E+00
6.365E+00
8.946E+00
9.474E+00
BMDL (ng/kg)
6.040E+00
6.051E+00
4.512E+00
4.425E+00
3.004E+00
4.512E+00
4.598E+00
4.587E+00
Notes








a Constant variance model selected (p = 0.107).
b Best-fitting model, BMDS output presented in this appendix.
G.2.9.2.  Output for Selected Model: Power
Franc et al. (2001): S-D Rats, Relative Liver Weight
         Power Model.  (Version:  2.15;   Date:  04/07/2008)
         Input Data File: C:\l\Blood\88_Franc_2001_SD_RelLivWt_PowerCV_l.(d)
         Gnuplot Plotting File:
C:\l\Blood\8 8_Franc_2 0 0!_SD_RelLivWt_PowerCV_l.pit
                                            Thu Apr 15 11:46:32  2010
 Figure  5,  SD rats,  relative liver  weight


   The  form of the response function  is:

   Y[dose]  = control + slope * doseApower
   Dependent  variable = Mean
   Independent variable = Dose
   rho  is  set to 0
   The  power  is restricted to be  greater than or equal to  1
   A  constant variance model is fit

   Total number of dose groups =  4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence  has been set to: le-008
   Parameter  Convergence has been set  to: le-008
                   Default Initial  Parameter Values
                           alpha  =       527.447
                             rho  =             0   Specified
                                       G-69

-------
                        control =
                          slope =
                          power =
          100
     0.947018
      1.13144
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
control
slope
power


-6.
5.
-4.
alpha
1
3e-009
4e-009
7e-009
control
-6.3e-009
1
-0.74
0.71
slope
5.4e-009
-0.74
1
-1
power
-4.7e-009
0.71
-1
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          462.113
688.544
        control          100.494
114.824
          slope         0.593276
3.17131
          power          1.25841
2.43011
Parameter Estimates



       Std.  Err.

         115.528

         7.31114

         1.31535

        0.597816
   95.0% Wald

Lower Conf.  Limit

        235.682

        86.1645

       -1.98476

       0.086712
     Table of Data and Estimated Values of Interest
Dose
Res .
0
6.587
14.48
36.43
N
8
8
8
8
Obs Mean
100
108
117
155
Est Mean
100
107
118
155
Obs S

16
25
30
td Dev
14
.9
.9
.9
Est S
21
21
21
21
td Dev
.5
.5
.5
.5
Seal
-0
0
-0
0.
ed
.065
.158
.109
0157
 Model Descriptions for likelihoods calculated
                                     G-70

-------
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
                       Log(likelihood)
                        -114.152281
                        -111.103649
                        -114.152281
                        -114.172940
                        -125.052064
 # Param's
       5
       8
       5
       4
       2
     AIC
  238.304562
  238.207299
  238.304562
  236.345880
  254.104127
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
           -2*log(Likelihood Ratio)  Test df
                       27.8968
                       6.09726
                       6.09726
                     0.0413179
6
3
3
1
 p-value

<.0001
 0.107
 0.107
0.8389
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than  .1.
model appears to be appropriate here
                                            A homogeneous variance
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
                                            The modeled variance appears
                                            The model chosen seems
                                     G-71

-------
                Benchmark  Dose Computation


Specified  effect =            0.1


Risk Type         =      Relative risk


Confidence level =           0.95


              BMD = 9.47408




             BMDL = 4.5873




G.2.9.3.  Figure for Selected Model: Power



                              Power Model with 0.95 Confidence Level



        180  ^
        160
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 OJ
        140
120  -
        100
         80  -
   11:4604/152010
                                                                         35
                                        G-72

-------
G.2.10. Franc et al. (2001): Long-Evans (L-E) Rats, Relative Liver Weight
G.2.10.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hillb
Linear
Polynomial, 3 -degree
Power
Hill, unrestricted0
Power, unrestricted
Degrees of
freedom
2
2
1
1
1
2
1
2
0
1
/2 p-value
0.441
0.441
0.785
0.785
0.829
0.499
0.0001
0.499
N/A
0.965
AIC
208.974
208.974
209.408
209.408
209.381
208.725
10.000
208.725
211.337
209.336
BMD (ng/kg)
1.708E+01
1.708E+01
7.997E+00
7.997E+00
7.725E+00
1.570E+01
8.604E+00
1.570E+01
7.217E+00
7.193E+00
BMDL
(ng/kg)
1.098E+01
1.098E+01
2.601E+00
2.601E+00
1.225E+00
9.619E+00
error
9.619E+00
1.147E+00
error
Notes

power hit bound (d=\)

power hit bound (d=\)
n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted (n = 0.545)
unrestricted
(power = 0.524)
a Nonconstant variance model selected (p = 0.0632).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.2.10.2. Output for Selected Model: Hill
Franc et al. (2001): L-E Rats, Relative Liver Weight
         Hill Model.  (Version:  2.14;  Date:  06/26/2008)
         Input Data File:  C:\l\Blood\89_Franc_2001_LE_RelLivWt_Hill_l.(d)
         Gnuplot Plotting  File:
C:\l\Blood\89_Franc_2001_LE_RelLivWt_Hill_l.pit
                                             Thu Apr 15 11:48:44  2010
 Figure  5,  L-E rats, relative liver weight


   The  form of the response  function is:

   Y[dose]  = intercept  +  v*doseAn/(kAn + doseAn)
   Dependent variable = Mean
   Independent variable =  Dose
   Power parameter restricted to be greater  than 1
   The  variance is to be modeled as Var(i) = exp(lalpha
rho * In(mean(i)))
   Total  number of dose  groups = 4
   Total  number of records  with missing values  = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set  to:  le-008
   Parameter Convergence has  been set to:  le-008
                                       G-73

-------
                  Default Initial Parameter Values
                         lalpha =      5.41581
                            rho =            0
                      intercept =          100
                              v =       22.225
                              n =     0.443155
                              k =       18.746
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -n
                 have been estimated at a boundary point,  or have been
specified by the user,
                 and do not appear in the correlation matrix )

lalpha
rho
intercept
V
k
lalpha
1
-1
-0.21
0.33
0.18
rho
-1
1
0.21
-0.33
-0.18
intercept
-0.21
0.21
1
0.028
0.35
V
0.33
-0.33
0.028
1
0.91

0
-0
0
0

k
.18
.18
.35
.91
1
                                 Parameter Estimates
Confidence Interval
Variable
Estimate
Std.
Err.
Upper Conf. Limit

16.6449

11.9842
lalpha

rho

intercept
106.616

83.8004


75.9042

V

n
k

-17

4.

99

36


20

.2754

77884

.5348

.3963

1
.5223

17

3.

3.

24


28

.3066

67625

61286

.1862

NA
.2566

                                                         95.0% Wald

                                                      Lower Conf.  Limit

                                                             -51.1957

                                                             -2.42648

                                                              92.4538

                                                             -11.0079


                                                             -34.8596
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
     Table of Data and Estimated Values of Interest

                                     G-74

-------
 Dose
Res .
                 Obs Mean
                              Est Mean
                              Obs Std Dev  Est Std Dev
                                     Scaled
    0
6.584
14.47
36.41
        100
        106
        117
        122
99.5
 108
 115
 123
  10
17.9
8.97
19.9
10.5
12.9
14.8
17.4
  0.125
 -0.455
  0.426
-0.0954
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) +
           Var{e(ij)} = SigmaA2
 Model A2:
           Var{e(ij)} = Sigma(i)A2
 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e (i) } = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -100.516456
              -96.870820
              -99.666984
              -99.690373
             -105.717087
          # Param's
                5
                8
                6
                5
                2
            AIC
         211.032912
         209.741641
         211.333969
         209.380746
         215.434174
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
-2*log(Likelihood Ratio)  Test df
            17.6925
            7.29127
            5.59233
         6
         3
         2
        p-value

     0.007048
      0.06317
      0.06104
                                     G-75

-------
   Test 4            0.0467774          1          0.8288

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is less than .1.  You may want to consider a
different variance model

The p-value for Test 4 is greater than  .1.  The model chosen seems
to adequately describe the data
        Benchmark Dose Computation

Specified effect =           0.1

Risk Type        =     Relative risk

Confidence level =           0.95

             BMD =        7.72492

            BMDL =       1.22451
                                     G-76

-------
G.2.10.3. Figure for Selected Model: Hill
                               Hill Model with 0.95 Confidence Level
 c
 o
 Q.
 o:
 c
 (0
 0)
        140
        130
        120
        110
        100
         90
   11:4804/152010
                                                                        35
G.2.10.4. Output for Additional Model Presented: Hill, Unrestricted
Franc et al. (2001): L-E Rats, Relative Liver Weight
         Hill  Model.  (Version:  2.14;  Date:  06/26/2008)
         Input Data File:  C:\l\Blood\89_Franc_2001_LE_RelLivWt_Hill_U_l.(d)
         Gnuplot Plotting  File:
C:\l\Blood\8 9_Franc_2 0 0!_LE_RelLivWt_Hill_U_l.pit
                                            Thu  Apr 15 11:48:50  2010
 Figure  5,  L-E rats, relative  liver weight


   The form of the response  function is:

   Y[dose]  = intercept + v*doseAn/(kAn  + doseAn)
                                       G-77

-------
   Dependent variable = Mean
   Independent variable = Dose
   Power parameter is not restricted
   The variance is to be modeled as Var(i)
                      = exp(lalpha
               rho * In(mean(i))
   Total number of dose groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
lalpha =
rho =
intercept =
v =
n =
k =
5.41581
0
100
22.225
0.443155
18.746
           Asymptotic Correlation Matrix of Parameter Estimates
                 lalpha
            rho    intercept
    lalpha
-0.15
       rho
0.15
 intercept
0.013
-1
                  -0.14
             -1
           0.22
           0.14
-0.22
0.022
-0.14
              0.14
             0.022
0.24
             -0.24
              0.11
              -0.9
                   0.24
                  -0.15
          -0.24
           0.15
 0.11
0.013
 -0.9
             -0.92
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha         -19.2405
16.4505
            rho          5.19575
12.7781
            Parameter Estimates



                   Std.  Err.

                       18.21

                     3.86861
             95.0% Wald

          Lower Conf.  Limit

                 -54.9315

                 -2.38657
                                     G-78

-------
      intercept
106.43

27308.5

1.97744

958637
      99.5348

      440.285

     0.544741

      7266.27
           3.51796

           13708.5

          0.730981

            485402
                  92.6398

                 -26427.9

                -0.887956

                  -944104
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est  Std Dev
                                                 Scaled
    0
6.584
14.47
36.41
100
106
117
122
99.5
 109
 114
 123
  10
17.9
8.97
19.9
Degrees of freedom for Test A3 vs fitted <= 0
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
10.3
  13
14.6
17.8
Model
Al
A2
A3
fitted
R
Log (likelihood)
-100.516456
-96.870820
-99.666984
-99.668321
-105.717087
# Param's
5
8
6
6
2
AIC
211.032912
209.741641
211.333969
211.336641
215.434174
  0.128
 -0.589
  0.558
-0.0957
                   Explanation of Tests
                                     G-79

-------
 Test 1:

 Test 2:
 Test 3:
 Test 4:
 (Note:
   Test
 Do responses and/or variances differ among Dose levels?
 (A2 vs. R)
 Are Variances Homogeneous?  (Al vs A2)
 Are variances adequately modeled?  (A2 vs. A3)
 Does the Model for the Mean Fit?  (A3 vs. fitted)
When rho=0 the results of Test 3 and Test 2 will be the same.
                     Tests of Interest
  -2*log(Likelihood Ratio)  Test df
p-value
Test 1
Test 2
Test 3
Test 4
17.6925
7.29127
5.59233
0.00267242
6
3
2
0
0.007048
0.06317
0.06104
NA
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1.
model appears to be appropriate

The p-value for Test 3 is less than .1.
different variance model
                                A non-homogeneous variance
                                You may want to consider a
NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
        Benchmark Dose Computation

Specified effect =           0.1

Risk Type        =     Relative risk

Confidence level =           0.95

             BMD =        7.21718

            BMDL =       1.14742
                                     G-80

-------
G.2.10.5. Figure for Additional Model Presented: Hill, Unrestricted
                                 Hill Model with 0.95 Confidence Level
 CD
 CO
 c
 o
 Q.
 CO
 CD
 CD
 CD
         140
         130
120
         110
         100
          90
                    Hill
              BMDL
                 0
                     BMD
   11:4804/152010
                         10       15       20       25      30       35

                                      dose
                                             G-81

-------
G.2.11. Franc et al. (2001): S-D Rats, Relative Thymus Weight
G.2.11.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree0
Power
Power, unrestricted
Degrees of
freedom
2
1
1
0
0
2
2
2
1
X2p-value
0.814
0.016
0.720
N/A
N/A
0.404
0.404
0.404
0.483
AIC
285.107
292.452
286.825
288.696
288.696
286.508
286.508
286.508
287.189
BMD (ng/kg)
2.478E+00
3.173E+01
1.878E+00
3.296E+00
3.625E+00
4.783E+00
4.783E+00
4.783E+00
6.795E-01
BMDL
(ng/kg)
1.535E+00
1.007E+00
9.221E-01
9.365E-01
6.199E-01
3.893E+00
3.893E+00
3.893E+00
3.271E-03
Notes







power bound hit
(power =1)
unrestricted
(power = 0.515)
a Nonconstant variance model selected (p = 0.0320).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.2.11.2. Output for Selected Model: Exponential (M4)
Franc et al. (2001): S-D Rats, Relative Thymus Weight
         Exponential Model.  (Version: 1.61;   Date:  7/24/2009)
         Input Data File:  C:\l\Blood\91_Franc_2001_SD_RelThyWt_Exp_l.(d)
         Gnuplot Plotting  File:
                                            Thu  Apr 15 11:51:19  2010
 Figure  5,  SD rats, relative  thymus weight
   The  form of the response  function by Model:
      Model 2:     Y[dose] = a * exp{sign  *  b  *  dose}
      Model 3:     Y[dose] = a * exp{sign  *  (b  * dose)Ad}
      Model 4:     Y[dose] = a * [c-(c-l)  *  exp{-b * dose}]
      Model 5:     Y[dose] = a * [c-(c-l)  *  exp{-(b * dose)Ad}]

    Note:  Y[dose] is the median response for exposure = dose;
           sign = +1 for increasing trend in  data;
           sign = -1 for decreasing trend.

      Model 2 is nested within Models 3 and  4.
      Model 3 is nested within Model 5.
      Model 4 is nested within Model 5.
   Dependent variable = Mean
   Independent variable =  Dose
   Data  are assumed to be  distributed: normally

                                       G-82

-------
Variance Model: exp(lnalpha +rho *ln(Y[dose]))
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))

Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to:  le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
                                                      rho)
               Initial Parameter Values
               Variable

                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                 Model 4
                      3.35464
                      1.08199
                          105
                    0.0569979
                     0.108531
                            1
                  Parameter Estimates
                Variable

                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                  Model 4

                      2.4312
                     1.28672
                     110.959
                   0.0663498
                    0.146486
                           1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
  6.587
  14.48
  36.43
100
91.17
51.41
22.79
83.2
47.97
43.48
29.98
   Dose

      0
  6.587
  14.48
  36.43
               Estimated Values of Interest

             Est Mean      Est Std     Scaled Residual
  111
77.43
52.49
 24.7
69.78
55.36
43.11
26.54
-0.4442
 0.7019
-0.0709
-0.2031
                                  G-83

-------
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)

     Model  R:        Yij = Mu + e(i)
               Var{e (ij)  } = SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
4
-141.9834
-137.5818
-138.3482
-146.9973
-138.4123
5
8
6
2
5
293.9669
291.1637
288.6964
297.9946
286.8245
   Additive constant for all log-likelihoods =     -29.41.  This  constant
added to the
   above values gives the log-likelihood including the term that  does  not
   depend on the model parameters.
R)
                              Explanation of Tests

Test 1:  Does response and/or variances differ among Dose  levels?  (A2  vs.

Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)

Test 6a: Does Model 4 fit the data?  (A3 vs  4)
     Test

     Test 1
     Test 2
     Test 3
    Test 6a
                         Tests of Interest

                -2*log(Likelihood Ratio)
                                                  D. F.
p-value
18.83
8.803
1.533
0.1282
6
3
2
1
0.004459
0.03203
0.4647
0.7203
     The p-value for Test 1 is less than  .05.  There appears to be  a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.
                                     G-84

-------
  The p-value for Test 2 is less than .1.  A non-homogeneous
  variance model appears to be appropriate.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is greater than  .1.  Model 4 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 0.100000

         Risk Type = Relative deviation

  Confidence Level = 0.950000

               BMD =      1.87814

              BMDL =     0.922136
                                  G-85

-------
G.2.11.3. Figure for Selected Model: Exponential (M4)

                         Exponential_beta Model 4 with 0.95 Confidence Level
        150
 c
 o
 Q.
 o:
 c
 (0
 OJ
        100
         50
                        Exponential
             3MDL
BMD
                               10
                     15       20
                         dose
25
30
   11:51 04/152010
35
G.2.11.4. Output for Additional Model Presented: Polynomial, 3-degree
Franc et al. (2001): S-D Rats, Relative Thymus Weight
         Polynomial Model.  (Version:  2.13;  Date:  04/08/2008)
         Input  Data File: C:\l\Blood\91_Franc_2001_SD_RelThyWt_Poly_l.(d)
         Gnuplot Plotting File:
C:\l\Blood\91_Franc_2001_SD_RelThyWt_Poly_l.plt
                                            Thu Apr 15  11:51:20 2010
 Figure 5,  SD  rats,  relative thymus  weight


   The form of the response function is:

   Y[dose]  = beta 0  + beta l*dose  +  beta 2*doseA2 +
                                       G-86

-------
   Dependent variable = Mean
   Independent variable = Dose
   The polynomial coefficients are restricted to be negative
   The variance is to be modeled as Var(i)  = exp(lalpha + log(mean(i))  * rho)

   Total number of dose groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =       8.0075
                            rho =
                         beta_0 =
                         beta_l =
                         beta 2 =
                         beta 3 =
        0
      100
        0
-0.475283
        0
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -beta_2    -beta_3
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

lalpha
rho
beta 0
beta 1
lalpha
1
-0.99
0.018
0.0095
rho
-0.99
1
-0.022
-0.0024
beta 0
0.018
-0.022
1
-0.87
beta 1
0.0095
-0.0024
-0.87
1
                                 Parameter Estimates
Confidence Interval
Variable
Estimate
Std. Err.
Upper Conf. Limit

6

2


.18885

.01593

lalpha

rho

beta 0
2.8315

1.19884

94.5944
1.71297

0.416889

14.6685
123.344

-



0.978362


beta 1

beta 2
beta 3
-1.97776

0
0
0.509904

NA
NA
                                                         95.0% Wald

                                                      Lower Conf. Limit

                                                            -0.525852

                                                             0.381756

                                                              65.8446

                                                             -2.97715
                                     G-87

-------
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
    0
6.587
14.48
36.43
 100
91.2
51.4
22.8
94.6
81.6
  66
22.5
83.2
  48
43.5
  30
  63
57.6
50.7
26.7
 0.243
 0.471
-0.811
0.0269
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) +
           Var{e(ij)} = SigmaA2
 Model A2:
           Var{e(ij)} = Sigma(i)A2
 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e (i) } = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
     Log(likelihood)
      -141.983433
      -137.581833
      -138.348184
      -139.254163
      -146.997301
          # Param's
                5
                8
                6
                4
                2
            AIC
         293.966865
         291.163667
         288.696368
         286.508326
         297.994602
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.
                                     G-88

-------
   Test
         Tests of Interest

2*log(Likelihood Ratio)   Test df
p-value
Test 1
Test 2
Test 3
Test 4
18.8309
8.8032
1.5327
1.81196
6
3
2
2
0.004459
0.03203
0.4647
0.4041
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is greater than  .1.  The model chosen seems
to adequately describe the data
             Benchmark Dose Computation

Specified effect =           0.1

Risk Type        =     Relative risk

Confidence level =          0.95

             BMD =        4.78292
            BMDL =
                           3.8932
                                     G-89

-------
G.2.11.5. Figure for Additional Model Presented: Polynomial, 3-degree



                              Polynomial Model with 0.95 Confidence Level
 o
 Q.
 CD
 CD
         150
         100
          50
                        Polynomial
                  BMDL
BMD
                                 10      15       20      25       30       35

                                              dose
   11:51 04/152010
                                           G-90

-------
G.2.12. Franc et al. (2001): L-E Rats, Relative Thymus Weight
G.2.12.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Power
Power, unrestricted
Degrees of
freedom
2
2
1
0
0
2
2
2
1
)C2p-value
0.440
0.440
0.227
N/A
N/A
0.304
0.304
0.304
0.168
AIC
301.449
301.449
303.266
303.805
303.805
302.186
302.186
302.186
303.710
BMD (ng/kg)
2.726E+00
2.726E+00
2.084E+00
7.859E+00
7.480E+00
5.045E+00
5.045E+00
5.045E+00
1.374E+00
BMDL
(ng/kg)
1.212E+00
1.212E+00
5.926E-01
9.801E-01
7.512E-01
3.349E+00
3.349E+00
3.349E+00
9.032E-09
Notes

power hit bound (d = 1)





power bound hit
(power =1)
unrestricted
(power =0.601)
a Constant variance model selected (p = 0.5063).
b Best-fitting model, BMDS output presented in this appendix.
G.2.12.2. Output for Selected Model: Exponential (M4)
Franc et al. (2001): L-E Rats, Relative Thymus Weight
         Exponential Model.  (Version:  1.61;   Date: 7/24/2009)
         Input Data File: C:\l\Blood\92_Franc_2001_LE_RelThyWt_ExpCV_l.(d)
         Gnuplot Plotting File:
                                            Thu Apr 15  11:53:37  2010
 Figure  5,  L-E rats, relative thymus  weight
   The  form of the response function by Model:
      Model 2:      Y[dose] = a  *  exp{sign * b * dose}
      Model 3:      Y[dose] = a  *  exp{sign *  (b * dose)Ad}
      Model 4:      Y[dose] = a  *  [c-(c-l)  * exp{-b * dose}]
      Model 5:      Y[dose] = a  *  [c-(c-l)  * exp{-(b * dose)Ad}]

    Note:  Y[dose]  is the median response for exposure =  dose;
           sign = +1 for increasing  trend in data;
           sign = -1 for decreasing  trend.

      Model 2 is nested within  Models 3 and 4.
      Model 3 is nested within  Model 5.
      Model 4 is nested within  Model 5.
   Dependent variable = Mean
   Independent variable = Dose
   Data  are assumed to be distributed:  normally
   Variance Model:  exp(lnalpha  +rho  *ln(Y[dose])
   rho is  set to 0.
                                       G-91

-------
A constant variance model is fit.

Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 4

                 Inalpha               8.1814
                     rho(S)                 0
                       a                  105
                       b            0.0506168
                       c             0.166582
                       d                    1

   (S) = Specified
                  Parameter Estimates

                Variable          Model 4

                 Inalpha             8.22706
                     rho                   0
                       a             105.977
                       b           0.0660042
                       c            0.221786
                       d                   1


         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
  6.584
  14.47
  36.41
               Estimated Values of Interest

   Dose      Est Mean      Est Std     Scaled Residual
100
95.41
38.69
34.98
54.72
70.46
47.97
77.96
      0           106        61.16          -0.2764
  6.584         76.91        61.16           0.8555
  14.47         55.24        61.16           -0.765
  36.41         30.96        61.16            0.186


                                  G-92

-------
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)

     Model  R:        Yij = Mu + e(i)
               Var{e(ij)} = SigmaA2
                     Model
             Likelihoods of Interest

             Log(likelihood)       DF
                                                                AIC
Al
A2
A3
R
4
-146.9024
-145.7361
-146.9024
-150.6049
-147.6329
5
8
5
2
4
303.8049
307.4723
303.8049
305.2098
303.2658
   Additive constant for all log-likelihoods =     -29.41.  This  constant
added to the
   above values gives the log-likelihood including the term that  does  not
   depend on the model parameters.
                                 Explanation of Tests

   Test 1:  Does response and/or variances differ among Dose levels?  (A2  vs.
R)
   Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
   Test 3:  Are variances adequately modeled?  (A2 vs. A3)

   Test 6a: Does Model 4 fit the data?  (A3 vs  4)
     Test

     Test 1
     Test 2
     Test 3
    Test 6a
         Tests  of Interest

-2*log(Likelihood Ratio)
                                                  D. F.
9.738
2.333
2.333
1.461
6
3
3
1
p-value
                                                0.1362
                                                0.5063
                                                0.5063
                                                0.2268
     The p-value for Test 1 is greater than  .05.  There may not be  a
     diffence between responses and/or variances among the dose levels
     Modelling the data with a dose/response curve may not be  appropriate.

                                     G-93

-------
  The p-value for Test 2 is greater than .1.  A homogeneous
  variance model appears to be appropriate here.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is greater than .1.  Model 4 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 0.100000

         Risk Type = Relative deviation

  Confidence Level = 0.950000

               BMD =      2.08379

              BMDL =     0.592601
                                  G-94

-------
G.2.12.3. Figure for Selected Model: Exponential (M4)



                           Exponential_beta Model 4 with 0.95 Confidence Level
 0)
 to
 c
 o
 CL
 (/)
 0)
 or

 c
 TO
 0)
         150
         100
50
              MDL
                          Exponential
           BMD
   11:5304/152010
                                  10
15       20

    dose
25       30
                                                                   35
                                           G-95

-------
G.2.13. Franc et al. (2001): Han/Wistar (HAV) Rats, Relative Thymus Weight
G.2.13.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)b
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Power
Power, unrestricted
Degrees of
freedom
2
1
1
0
0
2
2
2
1
)C2p-value
0.698
0.407
0.396
N/A
N/A
0.645
0.645
0.645
0.363
AIC
261.646
263.616
263.646
264.927
264.927
261.804
261.804
261.804
263.755
BMD (ng/kg)
5.094E+00
5.944E+00
5.063E+00
9.945E+00
9.638E+00
6.874E+00
6.874E+00
6.874E+00
5.487E+00
BMDL
(ng/kg)
3.132E+00
3.140E+00
1.864E+00
2.127E+00
1.853E+00
5.006E+00
5.006E+00
5.006E+00
2.573E-01
Notes







power bound hit
(power =1)
unrestricted
(power =0.881)
a Constant variance model selected (p = 0.4331).
b Best-fitting model, BMDS output presented in this appendix.
G.2.13.2. Output for Selected Model: Exponential (M2)
Franc et al. (2001): HAV Rats, Relative Thymus Weight
         Exponential Model.  (Version:  1.61;   Date: 7/24/2009)
         Input Data File: C:\l\Blood\93_Franc_2001_HW_RelThyWt_ExpCV_l.(d)
         Gnuplot Plotting File:
                                            Thu Apr 15 11:55:55  2010
 Figure  5,  H/W rats,  relative thymus  weight
   The  form of the response function  by Model:
      Model 2:      Y[dose] = a *  exp{sign * b * dose}
      Model 3:      Y[dose] = a *  exp{sign * (b * dose)Ad}
      Model 4:      Y[dose] = a *  [c-(c-l)  * exp{-b * dose}]
      Model 5:      Y[dose] = a *  [c-(c-l)  * exp{-(b * dose)Ad}]

    Note:  Y[dose]  is the median response for exposure = dose;
           sign = +1 for increasing  trend in data;
           sign = -1 for decreasing  trend.

      Model 2  is nested within Models 3 and 4.
      Model 3  is nested within Model  5.
      Model 4  is nested within Model  5.
   Dependent variable = Mean
   Independent variable = Dose
   Data  are  assumed to be distributed:  normally
   Variance  Model:  exp(lnalpha +rho  *ln(Y[dose])

                                       G-96

-------
rho is set to 0.
A constant variance model is fit.

Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values
               Variable

                 Inalpha
                     rho(S)
                       a
                       b
                       c
                       d
                 Model 2
                      6.96647
                            0
                      56.9433
                    0.0204806
                            0
                            1
   (S) = Specified
                  Parameter Estimates
                Variable

                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                  Model 2

                   6.98895
                         0
                   103.047
                 0.0206828
                         0
                         1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
  6.588
  14.48
  36.44
100
97.53
71.02
49.29
35.98
32.98
23.99
43.48
   Dose

      0
  6.588
  14.48
  36.44
               Estimated Values of Interest

             Est Mean      Est Std     Scaled Residual
  103
89.92
76.38
48.49
32.93
32.93
32.93
32.93
-0.2617
 0.6532
-0.4596
0.06871
                                  G-97

-------
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)

     Model  R:        Yij = Mu + e(i)
               Var{e (ij)  } = SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
2
-127.4636
-126.0925
-127.4636
-132.935
-127.8231
5
8
5
2
3
264.9271
268.185
264.9271
269.87
261.6463
   Additive constant for all log-likelihoods =     -29.41.  This  constant
added to the
   above values gives the log-likelihood including the term that  does not
   depend on the model parameters.
R)
Test 1:

Test 2:
Test 3:
Test 4:
                     Explanation of Tests

Does response and/or variances differ among Dose  levels?  (A2  vs.

Are Variances Homogeneous?  (A2 vs. Al)
Are variances adequately modeled?  (A2 vs. A3)
Does Model 2 fit the data?  (A3 vs. 2)
     Test

     Test 1
     Test 2
     Test 3
     Test 4
                         Tests of Interest

                -2*log(Likelihood Ratio)
                                                  D. F.
13.69
2.742
2.742
0.7192
6
3
3
2
                                                    p-value
                                                                0.03336
                                                                 0.4331
                                                                 0.4331
                                                                  0.698
     The p-value for Test 1 is less than  .05.  There appears to be  a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

                                     G-98

-------
  The p-value for Test 2 is greater than .1.  A homogeneous
  variance model appears to be appropriate here.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 4 is greater than .1.  Model 2 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 0.100000

         Risk Type = Relative deviation

  Confidence Level = 0.950000

               BMD =      5.09411

              BMDL =      3.13214
                                  G-99

-------
G.2.13.3. Figure for Selected Model: Exponential (M2)


                           Exponential_beta Model 2 with 0.95 Confidence Level
 01
 to
 c
 o
 Q.
 (/)
 0)
 or
 £=
 TO
 0)
         140
         120
         100
80
60
          40
          20
                          Exponential
                 BMDL
                BMD
                 0
   11:5504/152010
                        10
15       20

    dose
25       30
35
                                          G-100

-------
G.2.14. Hojo et al. (2002): DRL Reinforce per Minute
G.2.14.1. Summary Table of BMDS Modeling Results
Model3
Hill
Linear
Polynomial, 3 -degree
Power
Power, unrestricted
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Degrees of
freedom
1
2
2
2
1
2
2
1
0
X2 p-value
0.101
0.009
0.009
0.009
0.025
0.007
0.007
0.054
N/A
AIC
4.465
9.124
9.124
9.124
6.780
9.612
9.612
5.488
6.465
BMD (ng/kg)
1.667E+00
1.352E+01
1.352E+01
1.352E+01
2.428E-01
1.623E+01
1.623E+01
1.316E-HM)
1.728E+00
BMDL
(ng/kg)
6.209E-08
6.020E+00
6.020E+00
6.020E+00
1.070E-14
8.673E+00
8.673E+00
2.367E-03
9.452E-03
Notes
n upper bound hit (n = 18)


power bound hit
(power =1)
unrestricted
(power =0.103)

power hit bound (d = 1)


a Constant variance model selected (p = 0.4321).
b Best-fitting model, BMDS output presented in this appendix.
G.2.14.2. Output for Selected Model: Exponential (M4)
Hojo et al. (2002): DRL Reinforce Per Minute
         Exponential Model.  (Version:  1.61;   Date: 7/24/2009)
         Input Data File: C:\l\Blood\21_Hojo_2002_DRLrein_ExpCV_l.(d)
         Gnuplot Plotting File:
                                            Mon Feb 08 10:49:08  2010
 Table  5,  values adjusted by a  constant to allow exponential  model
   The  form of the response function  by Model:
      Model 2:      Y[dose] = a  *  exp{sign * b * dose}
      Model 3:      Y[dose] = a  *  exp{sign * (b * dose)Ad}
      Model 4:      Y[dose] = a  *  [c-(c-l)  * exp{-b * dose}]
      Model 5:      Y[dose] = a  *  [c-(c-l)  * exp{-(b * dose)Ad}]

    Note:  Y[dose]  is the median response for exposure = dose;
           sign = +1 for increasing  trend in data;
           sign = -1 for decreasing  trend.

      Model 2  is nested within  Models 3 and 4.
      Model 3  is nested within  Model  5.
      Model 4  is nested within  Model  5.
   Dependent  variable = Mean
   Independent variable = Dose
   Data  are assumed to be distributed:  normally
   Variance Model:  exp(lnalpha +rho  *ln(Y[dose])

                                      G-101

-------
rho is set to 0.
A constant variance model is fit.

Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values
               Variable

                 Inalpha
                     rho(S)
                       a
                       b
                       c
                       d
                   Model 4
                       -1.29672
                              0
                         0.0817
                        0.15642
                        16.3733
                              1
   (S) = Specified
                  Parameter Estimates
                Variable

                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                    Model 4

                      -1.11961
                             0
                     0.0547452
                      0.708154
                        18.214
                             1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
  1.625
  4.169
   10.7
0.086
0.536
1.274
0.737
0.448
0.821
0.54
0.443
   Dose

      0
  1.625
  4.169
   10.7
               Estimated Values of Interest

             Est Mean      Est Std     Scaled Residual
0.05475
 0.6989
 0.9479
 0.9966
0.5713
0.5713
0.5713
0.5713
 0.1223
-0.6375
  1.398
 -1.016
                                  G-102

-------
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)

     Model  R:        Yij = Mu + e(i)
               Var{e (ij)  } = SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
4
3.11555
4.489557
3.11555
-2.435087
1.255891
5
8
5
2
4
3.7689
7.020886
3.7689
8.870174
5.488219
   Additive constant for all log-likelihoods =      -19.3.  This  constant
added to the
   above values gives the log-likelihood including the term that  does  not
   depend on the model parameters.
R)
                              Explanation of Tests

Test 1:  Does response and/or variances differ among Dose levels?  (A2  vs.

Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)

Test 6a: Does Model 4 fit the data?  (A3 vs  4)
     Test

     Test 1
     Test 2
     Test 3
    Test 6a
                         Tests of Interest

                -2*log(Likelihood Ratio)
                                                  D. F.
13.85
2.748
2.748
3.719
6
3
3
1
p-value
                                                                0.03137
                                                                 0.4321
                                                                 0.4321
                                                                0.05379
     The p-value for Test 1 is less than  .05.  There appears to be  a
     difference between response and/or variances among the dose

                                     G-103

-------
  levels, it seems appropriate to model the data.

  The p-value for Test 2 is greater than .1.  A homogeneous
  variance model appears to be appropriate here.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is less than .1.   Model 4 may not adequately
  describe the data; you may want to consider another model.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =      1.31616

              BMDL =   0.00236664
                                  G-104

-------
G.2.14.3. Figure for Selected Model: Exponential (M4)



                              Exponential Model 4 with 0.95 Confidence Level
 01
 to
 £=
 O
 CL
 or
 CO
 01
         1.5
         0.5
         -0.5
            BMDL
   10:4902/082010
                          Exponential
BMD
                                               dose
                                                                            10
                                           G-105

-------
G.2.15. Hojo et al. (2002): DRL Response per Minute
G.2.15.1. Summary Table of BMDS Modeling Results
Model3
Hill
Linear
Polynomial, 3 -degree
Power
Power, unrestricted
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Degrees of
freedom
0
2
2
2
2
2
2
1
0
X2 p-value
N/A
0.006
0.006
0.006
0.741
0.570
0.570
0.477
N/A
AIC
126.353
132.243
132.243
132.243
122.455
122.980
122.980
124.360
126.353
BMD (ng/kg)
1.373E+00
1.064E+01
1.064E+01
1.064E+01
1.070E+03
5.027E-01
5.027E-01
3.813E-01
8.430E-01
BMDL
(ng/kg)
1.070E-14
5.340E+00
5.340E+00
5.340E+00
error
error
error
1.553E-02
2.221E-02
Notes



power bound hit
(power =1)
unrestricted (power = 0)

power hit bound (d=\)


a Constant variance model selected (p = 0.3004).
b Best-fitting model, BMDS output presented in this appendix.
G.2.15.2. Output for Selected Model: Exponential (M4)
Hojo et al. (2002): DRL Response Per Minute
         Exponential Model.  (Version:  1.61;   Date: 7/24/2009)
         Input Data File: C:\l\Blood\23_Hojo_2002_DRLresp_ExpCV_l.(d)
         Gnuplot Plotting File:
                                            Mon Feb 08 10:50:10  2010
 Table  5,  values adjusted by a constant  to allow exponential model
   The  form of the response function  by Model:
      Model 2:      Y[dose] = a * exp{sign *  b * dose}
      Model 3:      Y[dose] = a * exp{sign *  (b  * dose)Ad}
      Model 4:      Y[dose] = a *  [c-(c-l)  *  exp{-b * dose}]
      Model 5:      Y[dose] = a *  [c-(c-l)  *  exp{-(b * dose)Ad}]

    Note:  Y[dose]  is the median response for exposure = dose;
           sign = +1 for increasing  trend in  data;
           sign = -1 for decreasing  trend.

      Model 2  is nested within Models 3 and  4.
      Model 3  is nested within Model  5.
      Model 4  is nested within Model  5.
   Dependent  variable = Mean
   Independent variable = Dose
   Data  are assumed to be distributed:  normally
   Variance Model:  exp(lnalpha +rho  *ln(Y[dose])
   rho is  set to 0.
                                      G-106

-------
A constant variance model is fit.

Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 4

                 Inalpha              4.51689
                     rho(S)                 0
                       a              24.6362
                       b             0.379327
                       c            0.0184785
                       d                    1

   (S) = Specified
                  Parameter Estimates

                Variable          Model 4

                 Inalpha             4.54096
                     rho                   0
                       a             23.4674
                       b             1.61185
                       c            0.101317
                       d                   1


         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0      5
  1.625      5
  4.169      6
   10.7      5
               Estimated Values of Interest

   Dose      Est Mean      Est Std     Scaled Residual
23.46
4.013
0.478
4.594
7.986
10.96
7.194
15.23
      0         23.47        9.684        -0.001008
  1.625         3.915        9.684          0.02265
  4.169         2.403        9.684          -0.4869
   10.7         2.378        9.684           0.5118


                                  G-107

-------
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i))  *  rho)

     Model  R:        Yij = Mu + e(i)
               Var{e(ij)} = SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
4
-57.92733
-56.09669
-57.92733
-64.49611
-58.1801
5
8
5
2
4
125.8547
128.1934
125.8547
132.9922
124.3602
   Additive constant for all log-likelihoods =      -19.3.  This  constant
added to the
   above values gives the log-likelihood including the term that  does  not
   depend on the model parameters.
R)
                              Explanation of Tests

Test 1:  Does response and/or variances differ among Dose  levels?  (A2  vs.

Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)

Test 6a: Does Model 4 fit the data?  (A3 vs  4)
     Test

     Test 1
     Test 2
     Test 3
    Test 6a
                         Tests of Interest

                 -2*log(Likelihood Ratio)
                                                  D. F.
16.8
3.661
3.661
0.5056
6
3
3
1
p-value
                                                                0.01005
                                                                 0.3004
                                                                 0.3004
                                                                 0.4771
     The p-value for Test 1 is less than  .05.  There appears  to be  a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

                                     G-108

-------
  The p-value for Test 2 is greater than .1.  A homogeneous
  variance model appears to be appropriate here.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is greater than .1.  Model 4 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =     0.381347

              BMDL =    0.0155267
                                 G-109

-------
G.2.15.3. Figure for Selected Model: Exponential (M4)



                            Exponential Model 4 with 0.95 Confidence Level
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 0)
        30
        20
10
       -10
          BMDL   BMD
                        Exponential
   10:5002/082010
                                                                            10
                                              dose
                                          G-110

-------
G.2.16. Kattainen et al. (2001): 3rd Molar Eruption, Female
G.2.16.1. Summary Table of BMDS Modeling Results
Model
Logistic
Log-logistic"
Log-probit
Probit
Multistage, 4 -degree
Log-logistic,
unrestricted13
Log-probit, unrestricted
Degrees of
freedom
3
3
3
3
3
2
2
X2 p-value
0.360
0.982
0.522
0.379
0.781
0.949
0.941
AIC
88.508
85.227
87.424
88.352
86.155
87.162
87.181
BMD (ng/kg)
9.223E+00
2.399E-KJO
7.346E+00
8.802E+00
4.042E+00
1.931E+00
2.075E+00
BMDL
(ng/kg)
6.671E+00
1.328E+00
4.561E+00
6.549E+00
2.626E+00
1.840E-01
2.395E-01
Notes

slope bound hit
(slope = 1)
slope bound hit
(slope =1)

final B = 0
unrestricted
(slope = 0.91)
unrestricted
(slope = 0.549)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.
G.2.16.2. Output for Selected Model: Log-Logistic
Kattainen et al. (2001): 3rd Molar Eruption, Female
         Logistic Model.  (Version:  2.12; Date: 05/16/2008)
         Input Data File: C:\l\Blood\24_Katt_2001_Erup_LogLogistic_BMRl.(d)
         Gnuplot Plotting File:
C:\l\Blood\24_Katt_2001_Erup_LogLogistic_BMRl.pit
                                            Mon Feb  08  10:50:39 2010
 Figure  2
   The  form of the probability  function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
   Dependent variable = DichEff
   Independent variable = Dose
   Slope  parameter is restricted as slope >= 1

   Total  number of observations  = 5
   Total  number of records with  missing values =  0
   Maximum number of iterations  = 250
   Relative Function Convergence has been set to:  le-008
   Parameter Convergence has  been set to: le-008
                                      G-lll

-------
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =       0.0625
                      intercept =     -3.07535
                          slope =            1
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -slope
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

             background    intercept

background            1        -0.53

 intercept        -0.53            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background        0.0699339

      intercept         -3.07219

          slope                1
                                 Parameter Estimates
                                   Std. Err.
                                                         95.0% Wald

                                                      Lower Conf. Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.9823
  Reduced model
0.0004142

           AIC:
                   Analysis of Deviance Table

             Log(likelihood)   # Param's  Deviance  Test d.f.
                  -40.5286         5
                  -40.6137         2      0.170195      3
                                                                    P-value
                  -50.7341
                   85.2274
                                        1
20.411
Dose
         Est.Prob.
                                  Goodness  of  Fit

                            Expected    Observed     Size
                                                                 Scaled
                                                                Residual
                                     G-112

-------
    0.0000     0.0699         1.119     1.000          16       -0.117
    2.2297     0.1570         2.669     3.000          17        0.221
    6.2523     0.2788         4.182     4.000          15       -0.105
   16.0824     0.4670         5.604     6.000          12        0.229
   46.8576     0.7066        13.426    13.000          19       -0.215

 ChiA2 = 0.17      d.f. = 3        P-value = 0.9820
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        2.39879

            BMDL =        1.32815
                                     G-113

-------
G.2.16.3. Figure for Selected Model: Log-Logistic
                            Log-Logistic Model with 0.95 Confidence Level
 T3
 £
 o
 o
 "
 o
 (0
         0.8
         0.6
         0.4
         0.2
                          Log-Logistic
             BMDL
BMD
                            10
20
                                  30
40
                                            dose
   10:5002/082010
G.2.16.4. Output for Additional Model Presented: Log-Logistic, Unrestricted
Kattainen et al. (2001): 3rd Molar Eruption, Female
         Logistic Model.  (Version:  2.12; Date:  05/16/2008)
         Input Data File:  C:\l\Blood\24_Katt_2001_Erup_LogLogistic_U_BMRl.(d)
         Gnuplot Plotting  File:
C:\l\Blood\24_Katt_2001_Erup_LogLogistic_U_BMRl.pit
                                            Mon  Feb 08 10:50:40  2010
 Figure  2
   The  form of the probability function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
                                       G-114

-------
   Dependent variable = DichEff
   Independent variable = Dose
   Slope parameter is not restricted

   Total number of observations = 5
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =       0.0625
                      intercept =      -2.7659
                          slope =     0.901885
           Asymptotic Correlation Matrix of Parameter Estimates

             background    intercept        slope

background            1        -0.52         0.38

 intercept        -0.52            1        -0.94

     slope         0.38        -0.94            1
Confidence Interval
       Variable
Upper Conf. Limit
     background
*
      intercept
*
          slope
               Parameter Estimates

                                       95.0% Wald

      Estimate        Std.  Err.      Lower Conf. Limit

     0.0630045            *                *

      -2.79616            *                *

      0.910333            *                *
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.9488
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -40.5286         5
     -40.5811         3      0.105049      2
                                     G-115
P-value

-------
  Reduced model
0.0004142

           AIC:
    -50.7341
     87.1622
              20.411
Goodness
of Fit
Scaled

0
2
6
16
46
Dose
.0000
.2297
.2523
.0824
.8576
Est
0.
0.
0.
0.
0.
. Prob.
0630
1683
2922
4692
6903
Expected
1.
2.
4.
5.
13.
008
862
383
631
116
Observed
1.
3.
4.
6.
13.
000
000
000
000
000
Size
16
17
15
12
19
Residual
-0
0
-0
0
-0
.008
.090
.217
.214
.058
 ChiA2 = 0.10
d.f. = 2
P-value = 0.9491
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        1.93079

            BMDL =        0.18403
                                     G-116

-------
G.2.16.5. Figure for Additional Model Presented: Log-Logistic, Unrestricted
 T3

 £
 O
 c
 O
 •*=
 O
 (0
                                Log-Logistic Model with 0.95 Confidence Level
          0.8
          0.6
0.4
          0.2
                             Log-Logistic
             Bli/IDL
           BMD
                                10
20            30

    dose
                                                                 40
   10:5002/082010
                                            G-117

-------
G.2.17. Kattainen et al. (2001): 3rd Molar Length, Female
G.2.17.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hillb
Linear
Polynomial,
4-degree
Power
Hill, unrestricted0
Power, unrestricted
Degrees of
freedom
3
3
2
2
2
3
o
J
3
1
2
X2 p-value
0.0001
0.0001
0.002
0.002
0.022
0.0001
O.0001
O.0001
O.0001
0.263
AIC
-124.866
-124.866
-147.120
-147.120
-152.239
-124.024
-124.024
-124.024
-130.856
-157.201
BMD (ng/kg)
1.669E+01
1.669E+01
4.237E-01
4.237E-01
3.132E-01
1.982E+01
1.982E+01
1.982E+01
1.215E-02
1.964E-03
BMDL (ng/kg)
9.933E+00
9.933E+00
2.530E-01
2.530E-01
1.679E-01
1.277E+01
1.277E+01
1.277E+01
error
8.002E-06
Notes

power hit bound
(rf=l)

power hit bound
(d=l)
n lower bound hit
(» = D


power bound hit
(power =1)
unrestricted
(n = 13.042)
unrestricted
(power =0.195)
a Nonconstant variance model selected (p = O.0001).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.2.17.2. Output for Selected Model: Hill
Kattainen et al. (2001): 3rd Molar Length, Female
         Hill Model.  (Version: 2.14;   Date:  06/26/2008)
         Input Data  File:  C:\l\Blood\25_Katt_2001_Length_Hill_l.(d)
         Gnuplot Plotting  File:  C:\l\Blood\25_Katt_2001_Length_Hill_l.plt
                                             Mon Feb  08  10:51:09 2010
 Figure 3 female only


   The  form of the  response function  is:

   Y[dose]  = intercept  + v*doseAn/(kAn  +  doseAn)
    Dependent variable  = Mean
    Independent variable = Dose
    Power parameter  restricted to be  greater than 1
    The  variance is  to  be modeled as  Var(i)  = exp(lalpha  + rho * In(mean(i)))

    Total number of  dose groups = 5
                                       G-118

-------
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =     -2.37155
                            rho =            0
                      intercept =
                              v =
                              n =
                              k =
  1.85591
-0.507874
 0.845932
  2.03129
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -n
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
lalpha
lalpha
rho
intercept
V
k

-0.
-0.
0.
-0.
1
98
16
84
38
rho
-0.98
1
0.2
-0.79
0.4
intercept
-0.16
0.2
1
-0.3
-0.11
V
0.84
-0.79
-0.3
1
-0.52
k
-0.38
0.4
-0.11
-0.52
1
                                 Parameter Estimates
Confidence Interval
Variable
Upper Conf. Limit
lalpha
6.06262
rho
-9.11612
intercept
1.88609
V
-0.332105
n
k
3.13675

Estimate

3.31084

-14.2657

1.85483

-0.453667

1
1.91219


Std. Err.

1.404

2.62739

0.0159477

0.0620227

NA
0.624785

                                                         95.0% Wald

                                                      Lower Conf. Limit

                                                             0.559057

                                                             -19.4153

                                                              1.82357

                                                            -0.575229


                                                             0.687636
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
                                     G-119

-------
     has no standard error.
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev    Scaled
    0
 2.23
6.252
16.08
46.86
16
17
15
12
19
1.86
1.58
 1.6
 1.5
1.35
1.85
1.61
1.51
1.45
1.42
0.0661
 0.185
 0.265
 0.221
 0.515
0.0639
 0.175
  0.28
 0.371
 0.431
0.0674
-0.789
  1.22
  0.51
-0.716
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
              Log(likelihood)
                 56.758717
                 85.856450
                 84.934314
                 81.119648
                 45.373551
                       # Param's
                             6
                            10
                             7
                             5
                             2
                         AIC
                     -101.517434
                     -151.712901
                     -155.868628
                     -152.239295
                      -86.747101
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest

                                     G-120

-------
           -2*log(Likelihood Ratio)  Test df
p-value
80.9658
58.1955
1.84427
7.62933
8
4
3
2
<.0001
<.0001
0.6053
0.02205
Test

Test 1
Test 2
Test 3
Test 4
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =       0.313211

            BMDL =      0.167922
                                     G-121

-------
G.2.17.3. Figure for Selected Model: Hill
                               Hill Model with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 HI
                   Hill
         1.9


         1.8


         1.7


         1.6


         1.5


         1.4
 ^_

         1.3


         1.2


         1.1
           BI/IDLJBMD

                0            10           20            30           40
                                            dose
   10:51 02/082010



G.2.17.4. Output for Additional Model Presented: Hill, Unrestricted

Kattainen et al. (2001): 3rd Molar Length, Female



         Hill Model.  (Version:  2.14;   Date:  06/26/2008)
         Input Data File:  C:\l\Blood\25_Katt_2001_Length_Hill_U_l.(d)
         Gnuplot Plotting  File:   C:\l\Blood\25_Katt_2001_Length_Hill_U_l.plt
                                             Mon Feb  08  10:51:09 2010


 Figure  3  female only
   The  form of the response  function is:

   Y[dose]  = intercept +  v*doseAn/(kAn  +  doseAn)


   Dependent variable = Mean
                                       G-122

-------
   Independent variable = Dose
   Power parameter is not restricted
   The variance is to be modeled as Var(i) = exp(lalpha  + rho * In(mean(i))

   Total number of dose groups = 5
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =     -2.37155
                            rho =            0
                      intercept =      1.85591
                              v =    -0.507874
                              n =     0.845932
                              k =      2.03129
           Asymptotic Correlation Matrix of Parameter Estimates
                 lalpha
               rho    intercept
    lalpha
3.3e-017

       rho
-5.1e-017

 intercept
1.4e-017
-6.2e-017
             -0.98
-0.16
                   0.84
0.22
             -0.77
            -0.16
                           0.22
 0.84
1.4e-016
                         -0.77    -2.2e-016
-0.35
               1.4e-016    -2.2e-016
            -0.35
                         6e-017    -2.6e-016
  6e-017
         -2.6e-016
               3.3e-017    -5.1e-017     1.4e-017    -6.2e-017
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha          4.25154
7.37044
            rho         -15.7639
-10.0776
               Parameter Estimates



                      Std.  Err.

                         1.5913

                        2.90127
                         95.0% Wald

                      Lower Conf.  Limit

                              1.13265

                             -21.4503
                                     G-123

-------
      intercept
1.88729
              v
-0.266393
              n
9.10027e+013
              k
5.05155e+011
  1.85591

-0.357293

  13.0417

0.0136512
   0.0160104

   0.0463784

4.64308e+013

2.57737e+011
      1.82453

    -0.448193

-9.10027e+013

-5.05155e+011
     Table of Data and Estimated Values of Interest
Dose
Res .
0
2.23
6.252
16.08
46.86
N
16
17
15
12
19
Ob
1
1


1
s

.
1
1

Mean
86
58
.6
.5
35
Est
1.
1
1
1
1
Mean
86
.5
.5
.5
.5
Obs
0.
0
0
0
0
Std Dev
0661
.185
.265
.221
.515
Est
0
0
0
0
0
Std Dev
.064
.345
.345
.345
.345
Scaled
2.09e-009
0.937
1.09
0.0534
-1.9
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
Model
Al
A2
A3
fitted
R
Log (likelihood)
56.758717
85.856450
84.934314
71.427978
45.373551
# Param's
6
10
7
6
2
AIC
-101.517434
-151.712901
-155.868628
-130.855955
-86.747101
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?

                                     G-124

-------
           (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.
Test

Test 1
Test 2
Test 3
Test 4
                     Tests of Interest
           -2*log(Likelihood Ratio)  Test df
                       80.9658
                       58.1955
                       1.84427
                       27.0127
 p-value

<.0001
<.0001
0.6053
<.0001
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =       0.012148


BMDL computation failed.
                                     G-125

-------
G.2.17.5. Figure for Additional Model Presented: Hill, Unrestricted
 OJ
 c/)
 c
 o
 Q.
 o:

 c
 (0
 0)
1.9




1.8




1.7




1.6




1.5




1.4




1.3




1.2




1.1
                      Hill
                   MD
                                                Hill Model
                                10
                                    20
30
40
                                                  dose
   10:51 02/082010
                                            G-126

-------
G.2.18. Keller et al. (2007): Missing Mandibular Molars, CBA J
G.2.18.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probit
Multistage, l-degreea
Multistage, 2 -degree
Multistage, 3 -degree
Probit
Weibull
Degrees of
freedom
1
2
1
1
3
1
1
2
1
X2 p-value
0.105
0.335
0.105
0.105
0.255
0.122
0.150
0.342
0.108
AIC
52.510
49.984
52.524
52.524
50.425
51.391
50.853
49.904
52.219
BMD (ng/kg)
3.342E+00
3.069E+00
4.009E+00
3.845E+00
1.091E-HM)
1.916E+00
1.713E+00
2.927E+00
2.744E+00
BMDL (ng/kg)
8.986E-01
2.212E+00
2.411E+00
2.421E+00
7.624E-01
9.654E-01
9.584E-01
2.053E+00
9.350E-01
Notes









a Best-fitting model, BMDS output presented in this appendix.


G.2.18.2. Output for Selected Model: Multistage, 1-Degree
Keller et al. (2007): Missing Mandibular Molars, CBA J
        Multistage Model. (Version: 3.0;   Date:  05/16/2008)
        Input  Data File: C:\l\Blood\26_Keller_2007_Molars_Multil_l.(d)
        Gnuplot Plotting File:  C:\l\Blood\26_Keller_2007_Molars_Multil_l.plt
                                            Mon Feb 08 10:51:47 2010
 Table  1 using  mandibular molars only
   The  form of  the probability function  is:

   P[response]  = background +  (1-background)*[1-EXP(
                  -betal*doseAl)]

   The  parameter betas are restricted to be  positive
   Dependent  variable = DichEff
   Independent  variable = Dose

 Total number of observations = 4
 Total number of records with missing values
 Total number of parameters in model =  2
 Total number of specified parameters = 0
 Degree of  polynomial = 1
= 0
 Maximum number of iterations = 250
 Relative  Function Convergence has been  set  to:  le-008
 Parameter Convergence has been set to:  le-008
                                      G-127

-------
                  Default Initial Parameter Values
                     Background =            0
                        Beta(l) = 3.03988e+018
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -Background
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
                Beta(l)
   Beta(l)
                                 Parameter Estimates
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     Background                0
*
        Beta(l)         0.096571
                      Std. Err.
                      95.0% Wald

                   Lower Conf. Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.1533
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -21.5798         4
     -24.2126         1       5.26564      3
      -71.326

      50.4251
     1
99.4926
3
                                 P-value
<.0001
                                  Goodness  of  Fit

Dose
0.0000
0.5374
4.2881
34.0560

Est. Prob.
0.0000
0.0506
0.3391
0.9627

Expected
0.000
1.163
9.833
28.881

Observed
0.000
2.000
6.000
30.000

Size
29
23
29
30
Scaled
Residual
0.000
0.796
-1.504
1.078
 ChiA2 =4.06
 d.f. = 3
P-value = 0.2554
                                     G-128

-------
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        1.09102

            BMDL =       0.762404

            BMDU =        1.56496

Taken together,  (0.762404, 1.56496) is a 90     % two-sided confidence
interval for the BMD
                                     G-129

-------
G.2.18.3. Figure for Selected Model: Multistage, 1-Degree

                             Multistage Model with 0.95 Confidence Level
 I
         0.8
         0.6
         0.4
         0.2
                                                                     30
35
   10:51 02/082010
G.2.19. Kociba et al. (1978): Urinary Coproporphyrin, Females

G.2.19.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Power
Power, unrestricted
Degrees of
freedom
2
2
1
0
0
2
2
2
1
X2 p-value
O.0001
O.0001
0.006
N/A
N/A
O.001
O.001
<0.001
0.001
AIC
82.975
82.975
73.823
69.047
69.047
82.233
82.233
82.233
78.691
BMD (ng/kg)
2.378E+01
2.378E+01
1.566E+00
6.225E+00
5.473E+00
1.790E+01
1.790E+01
1.790E+01
1.148E+00
BMDL
(ng/kg)
1.340E+01
1.340E+01
7.180E-01
1.586E+00
error
3.862E+00
3.862E+00
3.862E+00
8.984E-09
Notes

power hit bound (d = 1)





power bound hit
(power =1)
unrestricted
(power =0.4 16)
a Nonconstant variance model selected (p = 0.0298).
b Best-fitting model, BMDS output presented in this appendix.
                                            G-130

-------
G.2.19.2. Output for Selected Model: Exponential (M4)
Kociba et al. (1978): Urinary Coproporphyrin, Females
        Exponential Model.  (Version:  1.61;   Date:  7/24/2009)
        Input Data File: C:\l\Blood\29_Kociba_1978_Copro_Exp_l.(d)
        Gnuplot Plotting File:
                                          Mon  Feb  08  10:52:47  2010
 Table2-UrinaryCoproporphyrin
   The form of the response  function by Model:
      Model 2:     Y[dose] = a * exp{sign  *  b  *  dose}
      Model 3:     Y[dose] = a * exp{sign  *  (b  * dose)Ad}
      Model 4:     Y[dose] = a *  [c-(c-l)  *  exp{-b  *  dose}]
      Model 5:     Y[dose] = a *  [c-(c-l)  *  exp{-(b *  dose)Ad}]

    Note: Y[dose] is the median response for exposure  = dose;
          sign = +1 for increasing trend in  data;
          sign = -1 for decreasing trend.

      Model 2 is nested within Models  3 and  4.
      Model 3 is nested within Model 5.
      Model 4 is nested within Model 5.
   Dependent variable = Mean
   Independent variable = Dose
   Data are assumed to be distributed:  normally
   Variance Model: exp(lnalpha +rho  *ln(Y[dose]))
   The variance is to be modeled as  Var(i)  =  exp(lalpha  + log(mean(i))  * rho)

   Total number of dose groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set to:  le-008
   Parameter Convergence has been set  to:  le-008

   MLE solution provided: Exact
                  Initial Parameter Values

                  Variable          Model 4

                    Inalpha              -5.58269
                        rho              2.98472
                          a                  8.17
                          b             0.0692478
                          c              2.23623
                          d                     1
                                     G-131

-------
   Dose
                  Parameter Estimates

                Variable          Model  4
Inalpha
rho
a
b
c
d
-4.90852
2.80743
8.91071
0.15304
1.97526
1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std  Dev
0
1.547
7.155
38.56
5
5
5
5
9.8
8.6
16.4
17.4
1.3
2
4.7
4
               Estimated Values of Interest

             Est Mean      Est Std      Scaled  Residual
0
1.547
7.155
38.56
8.911
10.74
14.69
17.58
1.852
2.407
3.736
4.805
1.074
-1.991
1.021
-0.08246
Other models for which likelihoods are  calculated:

  Model Al:        Yij = Mu(i) + e(ij)
            Var{e (ij)  } = SigmaA2

  Model A2:        Yij = Mu(i) + e(ij)
            Var{e(ij)} = Sigma(i)A2

  Model A3:        Yij = Mu(i) + e(ij)
            Var{e(ij)} = exp(lalpha + log(mean(i))  *  rho)

  Model  R:        Yij = Mu + e(i)
            Var{e (ij)  } = SigmaA2
                  Model
Likelihoods of Interest

Log(likelihood)      DF
                                  G-132
                                                              AIC
Al
A2
A3
R
-31.69739
-27.21541
-28.16434
-41.73188
5
8
6
2
73.39478
70.43081
68.32868
87.46376

-------
                                 -31.91136            5      73.82272
   Additive constant for all log-likelihoods =     -18.38.  This constant
added to the
   above values gives the log-likelihood including the term that does not
   depend on the model parameters.
                                 Explanation of Tests

   Test 1:  Does response and/or variances differ among Dose levels?  (A2 vs.
R)
   Test 2:  Are Variances Homogeneous? (A2 vs. Al)
   Test 3:  Are variances adequately modeled?  (A2 vs. A3)

   Test 6a: Does Model 4 fit the data? (A3 vs 4)


                            Tests of Interest

     Test          -2*log(Likelihood Ratio)       D. F.         p-value
29.03
8.964
1.898
7.494
6
3
2
1
< 0.0001
0.02977
0.3872
0.00619
     Test 1
     Test 2
     Test 3
    Test 6a
     The p-value for Test 1 is less than .05.  There appears to be a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

     The p-value for Test 2 is less than .1.  A non-homogeneous
     variance model appears to be appropriate.

     The p-value for Test 3 is greater than  .1.  The modeled
     variance appears to be appropriate here.

     The p-value for Test 6a is less than .1.  Model 4 may not adequately
     describe the data; you may want to consider another model.
   Benchmark Dose Computations:

     Specified Effect = 1.000000

            Risk Type = Estimated standard deviations from control

     Confidence Level = 0.950000

                  BMD =      1.56562

                 BMDL =     0.718033
                                     G-133

-------
G.2.19.3. Figure for Selected Model: Exponential (M4)


                           Exponential Model 4 with 0.95 Confidence Level
  CD
  to
  c
  o
  Q.
  0)
                       Exponential
              0       5


   10:5202/082010
10
15
  20

dose
25
30
35
40
G.2.20. Kociba et al. (1978): Uroporphyrin per Creatinine, Female

G.2.20.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear1"
Polynomial, 3 -degree
Power
Degrees of
freedom
2
2
1
0
0
2
2
1
/2/7-value
0.755
0.755
0.499
N/A
N/A
0.793
0.793
0.497
AIC
-93.828
-93.828
-91.935
-90.190
-90.190
-93.928
-93.928
-91.928
BMD (ng/kg)
1.641E+01
1.641E+01
1.216E+01
7.542E+00
7.607E+00
1.306E-H)!
1.306E+01
1.326E+01
BMDL (ng/kg)
1.259E+01
1.259E+01
3.958E+00
4.128E+00
3.966E+00
9.287E+00
9.287E+00
9.287E+00
Notes

power hit bound (d = 1)






a Constant variance model selected (p = 0.4919).

b Best-fitting model, BMDS output presented in this appendix.
                                            G-134

-------
G.2.20.2. Output for Selected Model: Linear
Kociba et al. (1978): Uroporphyrin per Creatinine, Female
        Polynomial Model.  (Version: 2.13;  Date:  04/08/2008)
        Input Data File: C:\l\Blood\28_Kociba_1978_Uropor_LinearCV_l.(d)
        Gnuplot Plotting File:
C:\l\Blood\28_Kociba_1978_Uropor_LinearCV_l.plt
                                          Mon  Feb 08  10:52:17  2010
 Table 2


   The form of the response function is:

   Y[dose] = beta 0 + beta l*dose + beta 2*doseA2  +  ...
   Dependent variable = Mean
   Independent variable = Dose
   rho is set to 0
   Signs of the polynomial coefficients are not  restricted
   A constant variance model is fit

   Total number of dose groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set to: le-008
   Parameter Convergence has been set to:  le-008
                  Default Initial Parameter Values
                          alpha =     0.0030385
                            rho =             0    Specified
                         beta_0 =      0.149139
                         beta 1 =    0.00381789
           Asymptotic Correlation Matrix of  Parameter  Estimates

            ( *** The model parameter(s)  -rho
                 have been estimated at a boundary point,  or  have  been
specified by the user,
                 and do not appear in  the correlation  matrix  )

                  alpha       beta 0       beta  1

     alpha            1     1.9e-009    -2.6e-009

    beta_0     1.9e-009            1         -0.6

    beta_l    -2.6e-009         -0.6             1


                                     G-135

-------
Confidence Interval
       Variable
Upper Conf. Limit
          alpha
0.00402961
         beta_0
0.176517
         beta_l
0.00521295
           Parameter Estimates

                                   95.0% Wald

  Estimate        Std.  Err.     Lower Conf. Limit

0.00248773      0.000786688         0.000945846

  0.149139        0.0139684            0.121761

0.00381789      0.000711776          0.00242284
     Table of Data and Estimated Values of Interest
Dose
Res .
-
1.
7.
38
0
547
155
.56
N
5
5
5
5
Obs Mean
0
0
0
0
.157
.143
.181
.296
Est Mean
0
0
0
0
.149
.155
.176
.296
Obs

0
0
0
Std Dev Est
0.05
.037
.053
.074
0.
0.
0.
0.
Std Dev Scaled
0499
0499
0499
0499
0.352
-0.54
0.204
-0.0161
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
 Log(likelihood)
    50.195349
    51.400051
    50.195349
    49.963863
    41.049755
# Param's
      5
      8
      5
      3
      2
   AIC
-90.390697
-86.800103
-90.390697
-93.927727
-78.099510
                                     G-136

-------
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.
                     Tests of Interest
           -2*log(Likelihood Ratio)  Test df
        p-value
20.7006
2.40941
2.40941
0.46297
6
3
3
2
0.002076
0.4919
0.4919
0.7934
Test

Test 1
Test 2
Test 3
Test 4
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
The modeled variance appears
The model chosen seems
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD =         13.064
            BMDL =
                          9.28715
                                     G-137

-------
  OJ
  c/)
  c
  o
  Q.
  (/)
  0)
 o:

  c
  (0
  0)
G.2.20.3. Figure for Selected Model: Linear



                                    Linear Model with 0.95 Confidence Level


                         I inasir  	
            0.4





           0.35
            0.3
0.25
            0.2
           0.15
            0.1
              Linear
                               BMDL
                                BMD
                                     10
                                  15
  20

dose
25
30
35
40
   10:5202/082010
G.2.21. Kuchiiwa et al. (2002): Immunoreactive Neurons in Dorsalis, Males

G.2.21.1. Summary Table of BMDS Modeling Results
Model3
Linear1"
Degrees of
Freedom
0
X2 p-value
N/AC
AIC
93.91
BMD
(ng/kg-day)
6.044E-02
BMDL
(ng/kg-day)
4.270E-02
Notes

a Constant variance model selected (p = 0.530).

b Best-fitting model, BMDS output presented in this appendix.

c p-value could not be calculated because there were no available degrees of freedom.
                                             G-138

-------
G.2.21.2. Output for Selected Model: Linear
        Polynomial Model. (Version: 2.13;  Date: 04/08/2008)
        Input Data File:
C:\USEPA\BMDS21\l\79_Kuchiiwa_2002_dors_blood_dd_LinearCV_l.(d)
        Gnuplot Plotting File:
C:\USEPA\BMDS21\l\79_Kuchiiwa_2002_dors_blood_dd_LinearCV_l.plt
                                          Tue Aug 16 13:54:37 2011
 number_labeled_cells_dorsalis_TWAblooddose


   The form of the response function is:

   Y[dose]  = beta 0 + beta l*dose + beta 2*doseA2 + ...
   Dependent variable = Mean
   Independent variable = Dose
   rho is set to 0
   Signs of the polynomial coefficients are not restricted
   A constant variance model is fit

   Total number of dose groups = 2
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      670.324
                            rho =            0   Specified
                         beta_0 =      237.097
                         beta 1 =     -391.046
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

                  alpha       beta_0       beta_l

     alpha            1    -4.2e-008     2.3e-008

    beta_0    -4.2e-008            1        -0.71

    beta 1     2.3e-008        -0.71            1
                                     G-139

-------
Confidence Interval
       Variable
Upper Conf. Limit
          alpha
1005.57
         beta_0
256.008
         beta_l
-287.021
              Parameter Estimates

                                      95. 0% Wald

     Estimate        Std. Err.     Lower Conf. Limit

      558.603          228.049             111.636

      237.097          9.64886             218.186

     -391.046          53.0749            -495.071
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
    0
0.2571
237
 137
237
 137
 29
22.4
23.6
 23.6
Degrees of freedom for Test A3 vs fitted <= 0
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
1.03e-007
 2.15e-008
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
    Log(likelihood)
      -43.952634
      -43.755407
      -43.952634
      -43.952634
      -54.206960
         # Param's
               3
               4
               3
               3
               2
           AIC
         93.905267
         95.510815
         93.905267
         93.905267
        112.413921
                                     G-140

-------
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.
   Test

   Test 1
   Test 2
   Test 3
   Test 4
                     Tests of Interest
-2*log(Likelihood Ratio)   Test df
            20.9031
           0.394453
           0.394453
        i.81073e-013
2
1
1
0
 p-value

<.0001
  0.53
  0.53
    NA
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.  The modeled variance appears
 to be appropriate here

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD =      0.0604398
            BMDL =
                        0.0427028
                                     G-141

-------
G.2.21.3.  Figure for Selected Model: Linear

                                Linear Model with 0.95 Confidence Level
  c
  o
  Q.
 o:
 c
 (0
 OJ
280


260


240


220


200


180


160


140


120


100
                     Linear
                      BMDL
BMD
                            0.05
0.1          0.15
      dose
                                                        0.2
                                             0.25
   13:5408/162011
G.2.22. Kuchiiwa et al. (2002): Immunoreactive Neurons in Medianus, Males
G.2.22.1. Summary Table of BMDS Modeling Results
Model3
Linear1"
Degrees of
Freedom
0
X2 p-value
N/AC
AIC
65.97
BMD
(ng/kg-day)
4.928E-02
BMDL
(ng/kg-day)
3.227E-02
Notes

a Modeled variance model selected (p = 0.025).
b Best-fitting model, BMDS output presented in this appendix.
c/rvalue could not be calculated because there were no available degrees of freedom.
G.2.22.2.  Output for Selected Model: Linear


         Polynomial Model.   (Version: 2.13;   Date:  04/08/2008)
         Input Data File:
C:\USEPA\BMDS21\l\80_Kuchiiwa_2002_med_blood_dd_Linear_l.(d)
         Gnuplot Plotting File:
C:\USEPA\BMDS21\1\8 0_Kuchiiwa_2 0 02_med_blood_dd_Linear_l.pit
                                         G-142

-------
                                          Tue Aug 16 13:55:40 2011
 number_labeled_cells_medianus_TWAblooddose


   The form of the response function is:

   Y[dose]  = beta 0 + beta l*dose + beta 2*doseA2 + ...
   Dependent variable = Mean
   Independent variable = Dose
   Signs of the polynomial coefficients are not restricted
   The variance is to be modeled as Var(i)  = exp(lalpha + log(mean(i))  * rho)

   Total number of dose groups = 2
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to:  le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =      4.43247
                            rho =            0
                         beta_0 =      91.1157
                         beta 1 =     -225.014
           Asymptotic Correlation Matrix of Parameter Estimates

                 lalpha          rho       beta 0       beta 1

    lalpha            1        -0.99     2.7e-009    -1.9e-009

       rho        -0.99            1      -3e-009     2.2e-009

    beta_0     2.7e-009      -3e-009            1        -0.94

    beta 1    -1.9e-009     2.2e-009        -0.94            1
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
2.44349
            rho
3.53497
         beta_0
99.9878
         Parameter Estimates

                                 95.0% Wald

Estimate        Std.  Err.     Lower Conf. Limit

-3.97249          3.27352            -10.3885

  1.9468         0.810306            0.358628

 91.1157          4.52665             82.2436


            G-143

-------
         beta_l         -225.014           18.8038             -261.869
-188.16
     Table of Data and Estimated Values of  Interest
Dose
Res .
0
0.2571
N Obs Mean

6 91.1
6 33.3
Est Mean

91.1
33.3
Obs Std Dev

12.1
4.55
Est Std Dev

11.1
4.16
Scaled

4.41e-009
-4.19e-009
Degrees of freedom for Test A2 vs A3 <=  0

 Warning: Likelihood for fitted model larger  than  the  Likelihood for model
A3.
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest

            Model      Log(likelihood)    #  Param's      AIC
             Al          -31.500916             3       69.001832
             A2          -28.985335             4       65.970670
             A3          -28.985335             4       65.970670
         fitted          -28.985335             4       65.970670
              R          -46.859574             2       97.719148
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ  among  Dose  levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3  vs.  fitted)

                                     G-144

-------
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest

   Test    -2*log(Likelihood Ratio)  Test df        p-value
   Test 1
   Test 2
   Test 3
   Test 4
      35.7485
      5.03116
 2.47269e-012
-2.47269e-012
2
1
0
0
<.0001
0.0249
    NA
    NA
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

NA - Degrees of freedom for Test 3 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD =      0.0492768

            BMDL =       0.032269
                                     G-145

-------
G.2.22.3. Figure for Selected Model: Linear
                                Linear Model with 0.95 Confidence Level
  c
  o
  Q.
 o:
 c
 (0
 0)
110


100


 90


 80


 70


 60


 50


 40


 30
                     Linear
                    BMDL
                    BMD
                            0.05
                               0.1          0.15
                                     dose
0.2
0.25
   13:5508/162011
G.2.23. Kuchiiwa et al. (2002): Immunoreactive Neurons in B9, Males
G.2.23.1. Summary Table of BMDS Modeling Results
Model3
Linear1"
Degrees of
Freedom
0
X2 p-value
N/AC
AIC
86.12
BMD
(ng/kg-day)
4.172E-02
BMDL
(ng/kg-day)
3.015E-02
Notes

a Constant variance model selected (p = 0.504).
b Best-fitting model, BMDS output presented in this appendix.
c/>-value could not be calculated because there were no available degrees of freedom.
G.2.23.2. Output for Selected Model: Linear


         Polynomial Model.  (Version: 2.13;   Date:  04/08/2008)
         Input Data File:
C:\USEPA\BMDS21\l\81_Kuchiiwa_2002_b9_blood_dd_LinearCV_l.(d)
         Gnuplot Plotting File:
C:\USEPA\BMDS21\1\81  Kuchiiwa  2002 b9  blood dd  LinearCV  l.plt
                                         G-146

-------
                                          Tue Aug 16 13:57:44 2011
 number_labeled_cells_b9_TWAblooddose


   The form of the response function is:

   Y[dose]  = beta 0 + beta l*dose + beta 2*doseA2 + ...
   Dependent variable = Mean
   Independent variable = Dose
   rho is set to 0
   Signs of the polynomial coefficients are not restricted
   A constant variance model is fit

   Total number of dose groups = 2
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to:  le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      350.225
                            rho =            0   Specified
                         beta_0 =      152.086
                         beta 1 =     -409.531
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
                  alpha

     alpha            1

    beta_0     2.2e-007

    beta 1    -2.5e-007
  beta_0       beta_l

2.2e-007    -2.5e-007

       1        -0.71

   -0.71            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          291.854
525.381
     Parameter Estimates



            Std.  Err.

              119.149
   95.0% Wald

Lower Conf.  Limit

        58.3265
                                     G-147

-------
         beta 0
165.756
         beta 1
      152.086

      409.531
           6.9744

          38.3637
                     138.416

                    -484.722
-334.339
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
    0
0.2571
152
46.8
152
46.8
     16
    21.1
Degrees of freedom for Test A3 vs fitted <= 0
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
17.1
 17.1
-5.3e-007
 3.27e-007
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
    Log(likelihood)
      -40.057520
      -39.834453
      -40.057520
      -40.057520
      -54.163617
# Param's
      3
      4
      3
      3
      2
                        AIC
                      86.115041
                      87.668907
                      86.115041
                      86.115041
                     112.327234
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                                     G-148

-------
   Test

   Test 1
   Test 2
   Test 3
   Test 4
          Tests  of Interest

-2*log(Likelihood Ratio)   Test df
            28.6583
           0.446134
           0.446134
       1.87583e-012
2
1
1
0
 p-value

<.0001
0.5042
0.5042
    NA
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.  The modeled variance appears
 to be appropriate here

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD =      0.0417154

            BMDL =      0.0301486
                                     G-149

-------
G.2.23.3.  Figure for Selected Model: Linear
                               Linear Model with 0.95 Confidence Level
 c
 o
 Q.
o:
 c
 (0
 OJ
180


160


140


120


100


 80


 60


 40


 20
  13:5708/162011
                           0.05
                                           0.15
0.2
                                             dose
G.2.24. Kuchiiwa et al. (2002): Immunoreactive Neurons in Magnus, Males
G.2.24.1. Summary Table of BMDS Modeling Results
0.25
Model3
Linear1"
Degrees of
Freedom
0
X2 p-value
N/AC
AIC
60.36
BMD
(ng/kg-day)
3.354E-02
BMDL
(ng/kg-day)
2.048E-02
Notes

a Modeled variance model selected (p = 0.013).
b Best-fitting model, BMDS output presented in this appendix.
c p-value could not be calculated because there were no available degrees of freedom.
G.2.24.2.  Output for Selected Model: Linear
         Polynomial Model.   (Version: 2.13;   Date:  04/08/2008)
         Input  Data File:
C:\USEPA\BMDS21\l\82_Kuchiiwa_2002_mag_blood_dd_Linear_l.(d)
         Gnuplot Plotting File:
C:\USEPA\BMDS21\l\82_Kuchiiwa_2002_mag_blood_dd_Linear_l.plt

                                         G-150

-------
                                          Tue Aug 16 13:56:37 2011
 number_labeled_cells_magnus_TWAblooddose


   The form of the response function is:

   Y[dose]  = beta 0 + beta l*dose + beta 2*doseA2 +
   Dependent variable = Mean
   Independent variable = Dose
   Signs of the polynomial coefficients are not restricted
   The variance is to be modeled as Var(i)  = exp(lalpha + log(mean(i))  * rho)

   Total number of dose groups = 2
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to:  le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =      4.05645
                            rho =            0
                         beta_0 =      43.6123
                         beta 1 =     -92.5263
           Asymptotic Correlation Matrix of Parameter Estimates

                 lalpha          rho       beta 0       beta 1

    lalpha            1        -0.99     4.1e-009    -5.6e-008

       rho        -0.99            1    -4.6e-009     5.3e-008

    beta_0     4.1e-009    -4.6e-009            1        -0.32

    beta 1    -5.6e-008     5.3e-008        -0.32            1
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
19.6944
            rho
-0.757015
         beta_0
46.0952
         Parameter Estimates

                                 95.0% Wald

Estimate        Std.  Err.     Lower Conf. Limit

 12.7854          3.52508             5.87638

-2.78668          1.03556            -4.81635

 43.6123          1.26679             41.1294


            G-151

-------
         beta 1
      -92.5263
           15.5809
                -123.064
-61.9882
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev    Scaled
    0
0.2571
43.6
 19.E
43.6
 19.E
3.4
10.2
Degrees of freedom for Test A2 vs A3 <= 0

Degrees of freedom for Test A3 vs fitted <= 0



 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
3.1
9.31
1.13e-008
 1.88e-008
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
     Log(likelihood)
       -29.244768
       -26.179929
       -26.179929
       -26.179929
       -37.469939
          # Param's
                3
                4
                4
                4
                2
           AIC
         64.489536
         60.359859
         60.359859
         60.359859
         78.939878
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.
                                     G-152

-------
   Test

   Test 1
   Test 2
   Test 3
   Test 4
          Tests of Interest

-2*log(Likelihood Ratio)   Test df
              22.58
            6.12968
       7.10543e-015
                  0
2
1
0
0
  p-value

 <.0001
0.01329
     NA
     NA
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

NA - Degrees of freedom for Test 3 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD =      0.0335363

            BMDL =       0.020483
                                     G-153

-------
G.2.24.3. Figure for Selected Model: Linear



                                 Linear Model with 0.95 Confidence Level
 c
 o
 Q.
 o:

 c
 (0
 OJ
50




45




40




35




30




25




20




15




10
                    Linear
                BMDL
                BMD
                           0.05
                                 0.1
0.15
0.2
0.25
                                                dose
   13:5608/162011
                                           G-154

-------
G.2.25. Latchoumycandane and Mathur (2002): Sperm Production
G.2.25.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hillb
Linear
Polynomial, 3 -degree
Power
Hill, unrestricted0
Power, unrestricted
Degrees of
freedom
2
2
1
0
1
2
2
2
0
1
X2 p-value
0.0001
0.0001
0.700
N/A
0.962
0.0001
0.0001
0.0001
N/A
0.501
AIC
93.831
93.831
75.261
77.263
75.115
94.250
94.250
94.250
77.113
75.566
BMD (ng/kg)
1.739E+01
1.739E+01
1.912E-01
2.925E-01
1.171E-01
1.995E+01
1.995E+01
1.995E+01
9.955E-02
6.921E-06
BMDL
(ng/kg)
9.432E+00
9.432E+00
7.976E-02
7.970E-02
1.324E-02
1.212E+01
1.212E+01
1.212E+01
1.228E-09
6.921E-06
Notes

power hit bound (d = 1)


n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted (n = 0.916)
unrestricted
(power = 0.087)
a Constant variance model selected (p = 0.8506).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.2.25.2. Output for Selected Model: Hill
Latchoumycandane and Mathur (2002): Sperm Production
         Hill Model.  (Version:  2.14;   Date: 06/26/2008)
         Input Data File:  C:\l\Blood\30_Latch_2002_Sperm_HillCV_l.(d)
         Gnuplot Plotting  File:   C:\l\Blood\30_Latch_2002_Sperm_HillCV_l.plt
                                            Mon  Feb  08  10:53:26 2010
  (xlOA6)  Table 1 without Vitamin E


   The  form of the response  function is:

   Y[dose]  = intercept + v*doseAn/(kAn + doseAn)
   Dependent variable = Mean
   Independent variable =  Dose
   rho  is  set to 0
   Power parameter restricted to be greater than  1
   A  constant variance model is  fit

   Total number of dose groups = 4
   Total number of records with  missing values =  0
   Maximum number of iterations  = 250
   Relative Function Convergence has been set to:  le-008
   Parameter Convergence has been set to: le-008
                                      G-155

-------
                  Default Initial Parameter Values
                          alpha =      7.23328
                            rho =
                      intercept =
                              v =
                              n =
                              k =
       0
   22.19
   -9.09
 1.93059
0.546864
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho    -n
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
alpha
alpha 1
intercept -2.2e-009
v -3.7e-008
k -5.9e-009


Confidence Interval
Variable
Upper Conf. Limit
alpha
9.43907
intercept
24.154
v
-6.60026
n
k
0.752259
intercept v
-2.2e-009 -3.7e-008
1 -0.76
-0.76 1
-0.23 -0.24
Parameter Estimates


Estimate Std. Err.

6.0283 1.74022

22.1894 1.00236

-9.16715 1.30966

1 NA
0.320198 0.220443

k
-5.9e-009
-0.23
-0.24
1

95.0% Wald

Lower Conf. Limit

2.61753

20.2248

-11.734


-0.111862

NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
                                    G-156

-------
    0
0.7845
4.651
27.27
       22.2
        15.7
       13.7
       13.1
22.2
 15.7
13.6
13.1
2.67
 2.65
2.19
3.16
2.46
 2.46
2.46
2.46
0.000631
 -0.00931
  0.0372
 -0.0285
 Model Descriptions for likelihoods calculated


 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -33.556444
              -33.158811
              -33.556444
              -33.557588
              -47.392394
          # Param's
                5
                8
                5
                4
                2
            AIC
          77.112888
          82.317623
          77.112888
          75.115176
          98.784788
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.
   Test

   Test 1
   Test 2
   Test 3
   Test 4
                     Tests of Interest
-2*log(Likelihood Ratio)  Test df
            28.4672
           0.795266
           0.795266
         0.00228746
         6
         3
         3
         1
        p-value

       <.0001
       0.8506
       0.8506
       0.9619
The p-value for Test 1 is less than .05.  There appears to be a

                                    G-157

-------
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
The modeled variance appears
The model chosen seems
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =       0.117131

            BMDL =     0.0132353
                                     G-158

-------
G.2.25.3. Figure for Selected Model: Hill

                             Hill Model with 0.95 Confidence Level
26
24
22
Response
->• K)
00 O
£=
o> 16
5
14
12
10
Bl
;
r '

-


-
~r
7IDL

Hill

-


\

-
-P-
\
v_

-
-
-
BMD
                                  10
15
20
                                         dose
   10:5302/082010
25
G.2.25.4. Output for Additional Model Presented: Hill, Unrestricted
Latchoumycandane and Mathur (2002): Sperm Production
        Hill  Model.  (Version: 2.14;   Date:  06/26/2008)
        Input Data File: C:\l\Blood\30_Latch_2002_Sperm_HillCV_U_l.(d)
        Gnuplot Plotting File:   C:\l\Blood\30_Latch_2002_Sperm_HillCV_U_l.plt
                                            Mon Feb 08  10:53:26 2010
  (xlOA6) Table  1  without Vitamin  E


   The  form  of  the response function  is:

   Y[dose] = intercept + v*doseAn/(kAn + doseAn)
   Dependent  variable = Mean
   Independent  variable = Dose
   rho is  set to 0
                                      G-159

-------
   Power parameter is not restricted
   A constant variance model is fit

   Total number of dose groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to:  le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      7.23328
                            rho =
                      intercept =
                              v =
                              n =
                              k =
       0
   22.19
   -9.09
 1.93059
0.546864
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point,  or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
intercept
V
n
k


-9
1
1
1
alpha
1
.8e-009
.6e-007
.6e-007
.2e-007
intercept
-9.8e-009
1
-0.5
-0.015
-0.13
V
1.6e-007
-0.5
1
0.76
0.56
n
1.6e-007
-0.015
0.76
1
0.86
k
1.2e-007
-0.13
0.56
0.86
1
                                 Parameter Estimates
Confidence Interval
Variable
Upper Conf. Limit
alpha
9.43818
intercept
24.1545
V
-5.25394
n
4.17544
k
1.16518

Estimate

6.02773

22.19

-9.23667

0.916265

0.301742


Std

1

1

2

1

0.


. Err.

.74006

.00231

.03204

.66287

440535

                                                         95.0% Wald

                                                      Lower Conf.  Limit

                                                              2.61728

                                                              20.2255

                                                             -13.2194

                                                             -2.34291

                                                            -0.561692
                                    G-160

-------
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev    Scaled
6
6
6
6
22.2
15.7
13.7
13.1
 Dose
Res .
    0
0.7845
4.651
27.27
Degrees of freedom for Test A3 vs fitted <= 0
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e (i) } = SigmaA2
22.2
15.7
13.6
13.1
2.67
2.65
2.19
3.16
2.46
2.46
2.46
2.46
3.4e-008
-1.51e-007
2.62e-007
-5.45e-007
                       Likelihoods of Interest
   Model
    Al
    A2
    A3
fitted
     R
                       Log (likelihood)
                         -33.556444
                         -33.158811
                         -33.556444
                         -33.556444
                         -47.392394
# Param's
      5
      8
      5
      5
      2
  AIC
77.112888
82.317623
77.112888
77.112888
98.784788
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest

                                     G-161

-------
   Test    -2*log(Likelihood Ratio)  Test df        p-value

   Test 1              28.4672          6          <.0001
   Test 2             0.795266          3          0.8506
   Test 3             0.795266          3          0.8506
   Test 4         6.96332e-013          0              NA

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.  The modeled variance appears
 to be appropriate here

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =      0.0995543

            BMDL =  1.22818e-009
                                     G-162

-------
G.2.25.5. Figure for Additional Model Presented: Hill, Unrestricted



                                  Hill Model with 0.95 Confidence Level


        26  H' ' ' '          ^~^
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 0)
24




22




20




18




16




14




12




10
           Hill
          BMDL
        3MD
                                        10
                                            15
20
25
                                               dose
   10:5302/082010
                                           G-163

-------
G.2.26. Li et al. (1997): Follicle-Stimulating Hormone (FSH)
G.2.26.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear
Polynomial,
8-degree
Powerb
Hill, unrestricted
Power, unrestricted0
Degrees of
freedom
8
8
7
6
7
8
9
8
6
7
X2 p-value
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
O.OOOl
0.001
0.002
AIC
1,095.292
1,095.292
1,059.480
1,066.195
1,056.459
1,077.695
1,155.670
1,077.695
1,039.481
1,037.474
BMD (ng/kg)
5.222E+02
5.222E+02
3.432E+01
1.019E+02
5.423E+00
2.003E+02
error
2.003E-H)2
2.204E-01
1.963E-01
BMDL
(ng/kg)
4.121E+02
4.121E+02
9.930E+00
8.583E-01
error
1.357E+02
1.916E+02
1.357E402
error
2.484E-02
Notes

power hit bound (d=\)


n lower bound hit (n = 1)


power bound hit
(power = 1)
unrestricted (n = 0.32)
unrestricted
(power =0.305)
a Nonconstant variance model selected (p = O.OOOl).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.2.26.2. Output for Selected Model: Power
Li et al. (1997): FSH
         Power Model.  (Version:  2.15;  Date:  04/07/2008)
         Input Data File:  C:\l\Blood\72_Li_1997_FSH_Pwr_l.(d)
         Gnuplot Plotting  File:   C:\l\Blood\72_Li_1997_FSH_Pwr_l.plt
                                            Mon  Feb  08 13:36:35 2010
 Figure  3:  FSH in female  S-D  rats 24hr after dosing,  22 day old rats


   The  form of the response function is:

   Y[dose]  = control + slope  *  doseApower
   Dependent variable = Mean
   Independent variable =  Dose
   The  power is restricted to  be greater than  or  equal to 1
   The  variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  *  rho)

   Total  number of dose groups = 10
   Total  number of records with missing values  =  0
   Maximum number of iterations = 250
   Relative Function Convergence has been set  to:  le-008
   Parameter Convergence has been set to: le-008
                                       G-164

-------
                  Default Initial Parameter Values
                         lalpha =       9.8191
                            rho =            0
                        control =      22.1591
                          slope =       52.284
                          power =     0.294106
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -power
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

lalpha
rho
control
slope
lalpha
1
-0.99
-0.29
-0.033
rho
-0.99
1
0.2
0.033
control
-0.29
0.2
1
-0.36
slope
-0.033
0.033
-0.36
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha          3.50054
5.9015
            rho          1.27087
1.74492
        control          87.4348
112.786
          slope         0.492306
0.672567
          power                1
Parameter Estimates



       Std.  Err.

           1.225

        0.241869

         12.9347

       0.0919718

              NA
   95.0% Wald

Lower Conf.  Limit

        1.09958

       0.796814

        62.0833

       0.312044
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
                                     G-165

-------
0
0.266
0.7988
2.097
5.867
15
43.33
119.9
386
1172
10
10
10
10
10
10
10
10
10
10
23.9
22.2
85.2
73.3
126
132
117
304
347
455
87.4
87.6
87.8
88.5
90.3
94.8
109
146
277
664
29.6
48.5
94.3
48.5
159
116
51.2
154
151
286
98.6
98.7
98.9
99.4
101
104
113
137
205
358
-2.04
-2.1
-0.0832
-0.483
1.12
1.14
0.223
3.65
1.07
-1.85
Model Descriptions for likelihoods calculated
Model Al:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = SigmaA2

Model A2:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = Sigma(i)A2

Model A3:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
    Model A3 uses any fixed variance parameters  that
    were specified by the user

Model  R:         Yi = Mu + e(i)
           Var{e(i)} = SigmaA2
                      Likelihoods of Interest
           Model
            Al
            A2
            A3
        fitted
             R
            Log(likelihood)
             -535.687163
             -496.367061
             -502.709623
             -534.847518
             -574.835246
# Param's
11
20
12
4
2
AIC
1093.374327
1032.734122
1029.419246
1077.695035
1153.670492
                  Explanation of Tests

Test 1:  Do responses and/or variances differ among  Dose  levels?
         (A2 vs. R)
Test 2:  Are Variances Homogeneous?  (Al vs A2)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)
Test 4:  Does the Model for the Mean Fit?  (A3 vs.  fitted)
(Note:  When rho=0 the results of Test 3 and Test  2  will  be  the  same.

                    Tests of Interest
  Test

  Test 1
  Test 2
-2*log(Likelihood Ratio)  Test df
            156.936
            78.6402
18
 9
 p-value

<.0001
<.0001
                                    G-166

-------
   Test 3              12.6851          8          0.1232
   Test 4              64.2758          8          <.0001

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 200.314


            BMDL = 135.673
                                     G-167

-------
 o
 Q.
 o:
 c
 (0
 OJ
        400
        300
        200
        100
                  Power
G.2.26.3. Figure for Selected Model: Power

                              Power Model with 0.95 Confidence Level

        700


        600


        500
                  BMDL
                         BMD
                        200
                                  400
 600
dose
800
1000
1200
   13:3602/082010
G.2.26.4. Output for Additional Model Presented: Power, Unrestricted
Lietal. (1997): FSH
        Power  Model.  (Version: 2.15;   Date:  04/07/2008)
        Input  Data File: C:\l\Blood\72_Li_1997_FSH_Pwr_U_l.(d)
        Gnuplot  Plotting File:  C:\l\Blood\72_Li_1997_FSH_Pwr_U_l.plt
                                            Mon Feb 08 13:36:46  2010
 Figure 3:  FSH  in female S-D rats  24hr  after dosing, 22 day  old rats


   The form of  the response function  is:

   Y[dose]  = control + slope * doseApower
   Dependent  variable = Mean
   Independent  variable = Dose
                                      G-168

-------
The power is not restricted
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  * rho)

Total number of dose groups = 10
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
               Default Initial Parameter Values
                      lalpha =       9.8191
                         rho =            0
                     control =      22.1591
                       slope =       52.284
                       power =     0.294106
        Asymptotic Correlation Matrix of Parameter Estimates

lalpha
rho
control
slope
power
lalpha
1
-0.99
-0.69
-0.06
0.26
rho
-0.99
1
0.65
0.0089
-0.23
control
-0.69
0.65
1
-0.23
0.029
slope
-0.06
0.0089
-0.23
1
-0.85
power
0.26
-0.23
0.029
-0.85
1
                              Parameter Estimates
Confidence Interval
Variable
Estimate
Std. Err.
Upper Conf. Limit

5.87265

1.613

29.299

71.0853

0.37088
lalpha

rho

control

slope

power

3.67487

1.17882

15.8201

52.528

0.304867

1.12134

0.221526

6.87715

9.46821

0.0336805

                                                      95.0% Wald

                                                   Lower Conf. Limit

                                                           1.47708

                                                          0.744632

                                                           2.34113

                                                           33.9706

                                                          0.238855
  Table of Data and Estimated Values of Interest
                                  G-169

-------
Dose
Res .
0
0.266
0.7988
2.097
5.867
15
43.33
119.9
386
1172
N
10
10
10
10
10
10
10
10
10
10
Obs Mean
23.9
22.2
85.2
73.3
126
132
117
304
347
455
                             Est Mean
                                        Obs  Std  Dev  Est  Std Dev
Scaled
15.8
50.9
64.9
81.7
106
136
182
242
339
469
29.6
48.5
94.3
48.5
159
116
51.2
154
151
286
32
63.7
73.5
84.1
98.1
114
135
160
195
236
0.795
-1.43
0.876
-0.314
0.652
-0.102
-1.52
1.24
0.134
-0.182
Model Descriptions for likelihoods calculated
Model Al:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = SigmaA2

Model A2:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = Sigma(i)A2

Model A3:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = exp(lalpha +  rho*ln(Mu(i)))
    Model A3 uses any fixed variance parameters  that
    were specified by the user

Model  R:         Yi = Mu + e(i)
           Var{e(i)} = SigmaA2
                      Likelihoods of  Interest
           Model
            Al
            A2
            A3
        fitted
             R
Log (likelihood)
-535.687163
-496.367061
-502.709623
-513.737215
-574.835246
# Param' s
11
20
12
5
2
AIC
1093.374327
1032.734122
1029.419246
1037.474431
1153.670492
                  Explanation of Tests

Test 1:  Do responses and/or variances differ  among  Dose  levels?
         (A2 vs. R)
Test 2:  Are Variances Homogeneous?  (Al vs A2)
Test 3:  Are variances adequately modeled?  (A2  vs. A3)
Test 4:  Does the Model for the Mean  Fit?  (A3  vs.  fitted)
(Note:  When rho=0 the results of Test 3 and Test  2  will  be  the same.

                    Tests of Interest
                                    G-170

-------
156.936
78.6402
12.6851
22.0552
18
9
8
7
<.0001
<.0001
0.1232
0.002485
   Test    -2*log(Likelihood Ratio)  Test df        p-value

   Test 1
   Test 2
   Test 3
   Test 4

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 0.196278


            BMDL = 0.0248364
                                     G-171

-------
 OJ
 c/)
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 0)
G.2.26.5. Figure for Additional Model Presented: Power, Unrestricted



                                Power Model with 0.95 Confidence Level


         700  \-      Power




         600




         500
400
300
         200
         100
            BMDLBMD
                           200
                             400
 600

dose
800
1000
1200
   13:3602/082010
                                          G-172

-------
G.2.27. Li et al. (2006): Estradiol, 3-Day
G.2.27.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear1"
Polynomial, 3 -degree
Power
Hill, unrestricted
Power, unrestricted
Degrees of
freedom
2
2
1
0
0
2
2
2
0
1
X2 p-value
0.156
0.156
0.341
N/A
N/A
0.162
0.162
0.162
N/A
0.328
AIC
269.027
269.027
268.212
270.212
270.212
268.952
268.952
268.952
270.265
268.265
BMD (ng/kg)
1.416E+01
1.416E+01
error
error
error
1.606E+01
1.606E+01
1.606E+01
9.273E+12
9.455E+10
BMDL
(ng/kg)
5.544E+00
5.544E+00
error
error
error
5.379E-KJO
5.379E+00
5.379E+00
9.273E+12
error
Notes

power hit bound (d=\)





power bound hit
(power =1)
unrestricted (n = 0.03)
unrestricted
(power =0.015)
a Constant variance model selected (p = 0.4372).
b Best-fitting model, BMDS output presented in this appendix.
G.2.27.2. Output for Selected Model: Linear
Li et al. (2006): Estradiol, 3-Day
         Polynomial Model.  (Version:  2.13;  Date: 04/08/2008)
         Input Data File: C:\l\Blood\31_Li_2006_Estra_LinearCV_l.(d)
         Gnuplot Plotting File:   C:\l\Blood\31_Li_2006_Estra_LinearCV_l.plt
                                            Mon Feb  08  10:54:00 2010
 Figure  3,  3-day estradiol


   The  form of the response  function is:

   Y[dose]  = beta 0 + beta l*dose  + beta 2*doseA2 +
   Dependent variable = Mean
   Independent variable = Dose
   rho  is  set to 0
   Signs of the polynomial coefficients are not restricted
   A  constant variance model  is  fit

   Total number of dose groups = 4
   Total number of records with  missing values = 0
   Maximum number of iterations  = 250
   Relative Function Convergence has  been set to: le-008
   Parameter Convergence has  been set to: le-008
                                      G-173

-------
                  Default Initial Parameter Values
                          alpha =      267.211
                            rho =            0   Specified
                         beta_0 =      16.1705
                         beta 1 =       1.0106
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
beta 0
beta 1
alpha
1
2.1e-012
5e-014
beta 0
2.1e-012
1
-0.69
beta 1
5e-014
-0.69
1
Confidence Interval
       Variable
Upper Conf. Limit
          alpha
378.888
         beta_0
23.147
         beta_l
3.39156
         Parameter Estimates

                                 95.0% Wald

Estimate        Std.  Err.      Lower Conf.  Limit

 263.435          58.9057              147.981

 16.1705          3.55949              9.19407

  1.0106           1.2148             -1.37037
     Table of Data and Estimated Values of Interest
Dose
Res .
0
0.1588
2.839
5.124
N
10
10
10
10
Obs
10.
19
24.
18.
Mean
2
.9
7
1
Est
16
1

21
Mean
.2
6.3
19
.3
Obs S
12

14
17
td Dev
.2
20
.6
.6
Est St
16.
16
16.
16.
d Dev
2
.2
2
2
Scaled
-1.
0.
1.
-0.6

17
697
11
35
 Model Descriptions for likelihoods calculated
                                     G-174

-------
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses  any fixed variance parameters that
     were specified by the user

 Model  R:          Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -129.653527
             -128.294657
             -129.653527
             -131.476097
             -131.819169
 # Param's
       5
       8
       5
       3
       2
     AIC
  269.307054
  272.589314
  269.307054
  268.952193
  267.638338
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
-2*log(Likelihood Ratio)  Test df
            7.04902
            2.71774
            2.71774
            3.64514
6
3
3
2
 p-value

0.3163
0.4372
0.4372
0.1616
The p-value for Test 1 is greater than  .05.  There may not be a
diffence between responses and/or variances among the dose levels
Modelling the data with a dose/response curve may not be appropriate
The p-value for Test 2 is greater than  .1.
model appears to be appropriate here
                                 A homogeneous variance
The p-value for Test 3 is greater than  .1.
 to be appropriate here

The p-value for Test 4 is greater than  .1.
to adequately describe the data
                                 The modeled variance appears
                                 The model chosen seems
                                     G-175

-------
              Benchmark Dose  Computation



Specified  effect =              1



Risk Type         =     Estimated standard  deviations from  the control mean



Confidence level =           0.95



              BMD =         16.0605
             BMDL =
                            5.37895
G.2.27.3. Figure for Selected Model: Linear



                             Linear Model with 0.95 Confidence Level
 c
 o
 Q.
 o:

 c
 (0
 OJ
       35
       30
       25
20
15
        10
                  Linear
                              BMDL
                                                                    BMD
                                            8

                                           dose
                                            10
12
14      16
   10:5402/082010
                                       G-176

-------
G.2.28. Li et al. (2006): Progesterone, 3-Day
G.2.28.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hillb
Linear
Polynomial,
3 -degree
Power
Power, unrestricted
Degrees of
freedom
2
2
1
0
1
2
2
2
1
X2 p-value
0.001
0.001
0.384
N/A
0.386
0.001
0.001
0.001
0.404
AIC
329.928
328.101
315.734
317.734
315.728
330.729
330.729
330.729
315.673
BMD (ng/kg)
2.619E+00
1.340E-01
1.074E-02
4.301E-02
9.461E-04
3.891E+00
3.891E+00
3.891E+00
2.812E-59
BMDL
(ng/kg)
error
error
6.633E-03
4.272E-03
8.006E-11
2.626E+00
2.626E+00
2.626E+00
2.812E-59
Notes

power hit bound (d=\)


n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted
(power =0.01)
a Nonconstant variance model selected (p = 0.0013).
b Best-fitting model, BMDS output presented in this appendix.
G.2.28.2. Output for Selected Model: Hill
Li et al. (2006): Progesterone, 3-Day
         Hill Model.  (Version:  2.14;   Date: 06/26/2008)
         Input Data File: C:\l\Blood\32_Li_2006_Progest_Hill_l.(d)
         Gnuplot Plotting File:   C:\l\Blood\32_Li_2006_Progest_Hill_l.plt
                                            Wed Feb  10  10:57:14 2010
 Figure  4,  3-day progesterone


   The  form of the response  function is:

   Y[dose]  = intercept + v*doseAn/(kAn + doseAn)
   Dependent variable = Mean
   Independent variable = Dose
   Power  parameter restricted to  be greater than  1
   The  variance is to be modeled  as Var(i)  = exp(lalpha   + rho * In(mean(i)))

   Total  number of dose groups  =  4
   Total  number of records with missing values =  0
   Maximum number of iterations = 250
   Relative Function Convergence  has been set to: le-008
   Parameter Convergence has been set to: le-008
                                      G-177

-------
                  Default Initial Parameter Values
                         lalpha =      7.08699
                            rho =            0
                      intercept =
                              v =
                              n =
                              k =
 61.7404
-50.3835
 1.47286
0.128302
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -n
                 have been estimated at a boundary point,  or have been
specified by the user,
                 and do not appear in the correlation matrix )

lalpha
rho
intercept
V
k
lalpha
1
-0.99
-0.093
0.82
0.22
rho
-0.99
1
0.12
-0.79
-0.2
intercept
-0.093
0.12
1
-0.43
0.014
V
0.82
-0.79
-0.43
1
0.035
k
0.22
-0.2
0.014
0.035
1
                                 Parameter Estimates
Confidence Interval
Variable
Upper Conf. Limit
lalpha
20.6775
rho
-0.585772
intercept
68.2898
V
-26.9922
n
k
0.0432411

Estimate

14.0902

-2.27438

61.7488

-42.1007

1
0.00282851


Std. Err.

3.36095

0.861553

3.3373

7.70852

NA
0.020619

NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
     Table of Data and Estimated Values of Interest
                                                         95.0% Wald

                                                      Lower Conf.  Limit

                                                              7.50284

                                                               -3.963

                                                              55.2078

                                                             -57.2091


                                                            -0.037584
                                    G-178

-------
 Dose
Res .
                 Obs Mean
                              Est Mean
                                Obs Std Dev  Est Std Dev
                                                  Scaled
    0
0.1588
2.839
5.124
10
 10
10
10
61.7
 30.6
16.9
11.4
61.7
 20.4
19.7
19.7
11.1
 40.5
33.3
43.7
10.6
 37.2
38.7
38.8
-0.00251
    0.865
  -0.225
  -0.678
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
              Log(likelihood)
               -159.632675
               -151.812765
               -152.488175
               -152.863841
               -165.698875
                       # Param's
                             5
                             8
                             6
                             5
                             2
                         AIC
                      329.265349
                      319.625529
                      316.976349
                      315.727683
                      335.397750
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
  -2*log(Likelihood Ratio)  Test df
              27.7722
              15.6398
              1.35082
             0.751333
                      6
                      3
                      2
                      1
                     p-value

                 0.0001037
                  0.001344
                    0.5089
                    0.3861
                                     G-179

-------
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is greater than  .1.  The model chosen seems
to adequately describe the data
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =    0.000946102

            BMDL =  8.00639e-011
                                     G-180

-------
G.2.28.3. Figure for Selected Model: Hill


                                  Hill Model with 0.95 Confidence Level
 01
 to
 £=
 O
 Q.
 or
 £=
 TO
 O)
        60
        40
20
        -20
                   Hill
          BMDLBMD
               0
   10:5702/102010
                                  2            3

                                       dose
                                            G-181

-------
G.2.29. Markowski et al. (2001): FR10 Run Opportunities
G.2.29.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)b
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Power
Power, unrestricted
Degrees of
freedom
2
2
1
0
0
2
2
2
1
X2 p-value
0.304
0.304
0.371
N/A
N/A
0.226
0.226
0.226
0.239
AIC
117.150
117.150
117.570
118.918
118.918
117.744
117.744
117.744
118.158
BMD (ng/kg)
8.570E+00
8.570E+00
3.452E+00
2.315E+00
1.801E+00
1.106E+01
1.106E+01
1.106E+01
5.768E+00
BMDL
(ng/kg)
2.887E-K)0
2.887E+00
1.299E-02
1.391E-02
1.274E-09
5.741E+00
5.741E+00
5.741E+00
1.032E-14
Notes

power hit bound (d = 1)





power bound hit
(power =1)
unrestricted
(power = 0.276)
a Constant variance model selected (p = 0.1719).
b Best-fitting model, BMDS output presented in this appendix.
G.2.29.2. Output for Selected Model: Exponential (M2)
Markowski et al. (2001): FR10 Run Opportunities
         Exponential Model.  (Version:  1.61;   Date: 7/24/2009)
         Input Data File: C:\l\Blood\33_Mark_2001_FR10opp_ExpCV_l.(d)
         Gnuplot Plotting File:
                                            Mon Feb 08 10:55:13  2010
 Table  3
   The  form of the response function  by Model:
      Model 2:      Y[dose] = a  *  exp{sign * b * dose}
      Model 3:      Y[dose] = a  *  exp{sign * (b * dose)Ad}
      Model 4:      Y[dose] = a  *  [c-(c-l)  * exp{-b * dose}]
      Model 5:      Y[dose] = a  *  [c-(c-l)  * exp{-(b * dose)Ad}]

    Note:  Y[dose]  is the median response for exposure = dose;
           sign = +1 for increasing  trend in data;
           sign = -1 for decreasing  trend.

      Model 2  is nested within  Models 3 and 4.
      Model 3  is nested within  Model  5.
      Model 4  is nested within  Model  5.
   Dependent variable = Mean
   Independent variable = Dose
   Data  are  assumed to be distributed:  normally
   Variance  Model:  exp(lnalpha +rho  *ln(Y[dose])

                                      G-182

-------
rho is set to 0.
A constant variance model is fit.

Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values
               Variable

                 Inalpha
                     rho(S)
                       a
                       b
                       c
                       d
                 Model 2
                       3.5321
                            0
                      6.77975
                    0.0581937
                            0
                            1
   (S) = Specified
                  Parameter Estimates
                Variable

                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                  Model 2

                   3.63127
                         0
                   12.2901
                 0.0808832
                         0
                         1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
  1.557
   4.03
  10.32
13.29
11.25
5.75
7
8.65
5.56
3.53
6.01
   Dose

      0
  1.557
   4.03
  10.32
               Estimated Values of Interest

             Est Mean      Est Std     Scaled Residual
12.29
10.84
8.871
5.335
  145
  145
6.145
6.145
0.4305
0.1347
-1.244
 0.717
                                  G-183

-------
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)

     Model  R:        Yij = Mu + e(i)
               Var{e (ij)  } = SigmaA2
         Model

            Al
            A2
            A3
             R
             2
                                Likelihoods of Interest

                                Log(likelihood)      DF
                                  -54.38526
                                  -51.88568
                                  -54.38526
                                  -57.45429
                                  -55.57522
                                                             AIC
5
8
5
2
3
118.7705
119.7714
118.7705
118.9086
117.1504
   Additive constant for all log-likelihoods =     -22.05.  This  constant
added to the
   above values gives the log-likelihood including the term that  does  not
   depend on the model parameters.
R)
Test 1:

Test 2:
Test 3:
Test 4:
                     Explanation of Tests

Does response and/or variances differ among Dose  levels?  (A2  vs.

Are Variances Homogeneous?  (A2 vs. Al)
Are variances adequately modeled?  (A2 vs. A3)
Does Model 2 fit the data?  (A3 vs. 2)
     Test

     Test 1
     Test 2
     Test 3
     Test 4
                Tests of Interest

        -2*log(Likelihood Ratio)

                        11.14
                        4.999
                        4.999
                         2.38
D.  F.

  6
  3
  3
  2
                                                             p-value
                                                                0.08423
                                                                 0.1719
                                                                 0.1719
                                                                 0.3042
     The p-value for Test 1 is greater than  .05.  There may not be  a
     diffence between responses and/or variances among the dose levels
     Modelling the data with a dose/response curve may not be appropriate.
                                     G-184

-------
  The p-value for Test 2 is greater than .1.  A homogeneous
  variance model appears to be appropriate here.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 4 is greater than .1.  Model 2 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =      8.56961

              BMDL =      2.88708
                                 G-185

-------
G.2.29.3. Figure for Selected Model: Exponential (M2)



                            Exponential Model 2 with 0.95 Confidence Level
 o
 Q.
 ro
 CD
        20
        15
        10
                                                                            10
   10:5502/082010
                                           G-186

-------
G.2.30. Markowski et al. (2001): FR2 Revolutions
G.2.30.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hillb
Linear
Polynomial, 3 -degree
Power
Power, unrestricted0
Degrees of
freedom
2
2
1
0
1
2
2
2
1
X2 p-value
0.236
0.236
0.263
N/A
0.654
0.180
0.180
0.180
0.161
AIC
217.219
217.219
217.583
218.532
216.532
217.764
217.764
217.764
218.294
BMD (ng/kg)
8.486E+00
8.486E+00
3.413E+00
2.415E+00
1.840E-H)0
1.058E+01
1.058E+01
1.058E+01
5.739E+00
BMDL
(ng/kg)
3.232E+00
3.232E+00
1.766E-02
9.313E-01
5.992E-01
5.602E+00
5.602E+00
5.602E+00
1.032E-14
Notes

power hit bound (d=\)


n upper bound hit
(it = 18)


power bound hit
(power =1)
unrestricted
(power =0.3 18)
a Constant variance model selected (p = 0.1092).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.2.30.2. Output for Selected Model: Hill
Markowski et al. (2001): FR2 Revolutions
         Hill Model.  (Version:  2.14;  Date:  06/26/2008)
         Input Data File:  C:\l\Blood\34_Mark_2001_FR2rev_HillCV_l.(d)
         Gnuplot Plotting  File:   C:\l\Blood\34_Mark_2001_FR2rev_HillCV_l.plt
                                            Mon Feb 08 10:55:47  2010
 Table  3


   The  form of the response  function is:

   Y[dose]  = intercept  +  v*doseAn/(kAn + doseAn)
   Dependent variable = Mean
   Independent variable =  Dose
   rho  is  set to 0
   Power parameter restricted to be greater  than 1
   A  constant variance model is fit

   Total number of dose groups = 4
   Total number of records  with missing values  = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set  to:  le-008
   Parameter Convergence has been set to:  le-008
                                       G-187

-------
                  Default Initial Parameter Values
                          alpha =      2598.74
                            rho =
                      intercept =
                              v =
                              n =
                              k =
      0
 119.29
 -62.79
2.13752
2.53662
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho    -n
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
                                                      3.5e-008

                                                         -0.52

                                                          0.37

                                                             1
alpha
alpha 1
intercept 1.2e-008
v le-009
k 3.5e-008

Confidence Interval
Variable
Upper Conf. Limit
alpha
3419.46
intercept
153.909
v
-13.5831
n
k
2.26502
intercept v
1.2e-008 le-009
1 -0.81
-0.81 1
-0.52 0.37
Parameter Estimates

Estimate Std. Err.

2183.85 630.425

119.29 17.6629

-56.5223 21.9082

18 NA
1.68653 0.295154

                                                         95.0% Wald

                                                      Lower Conf. Limit

                                                              948.245

                                                              84.6713

                                                             -99.4615


                                                              1.10804
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
                                     G-188

-------
    0
1.557
 4.03
10.32
        119
        109
       56.5
       68.1
 119
 108
62.8
62.8
69.9
  61
31.2
33.2
46.7
46.7
46.7
46.7
-2.41e-007
 2.29e-007
    -0.329
     0.304
 Model Descriptions for likelihoods calculated


 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e (i) } = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -104.165520
             -101.140174
             -104.165520
             -104.266162
             -107.599268
          # Param's
                5
                8
                5
                4
                2
            AIC
         218.331040
         218.280349
         218.331040
         216.532324
         219.198536
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.
   Test

   Test 1
   Test 2
   Test 3
   Test 4
                     Tests of Interest
-2*log(Likelihood Ratio)  Test df
            12.9182
            6.05069
            6.05069
           0.201284
         6
         3
         3
         1
        p-value

      0.04435
       0.1092
       0.1092
       0.6537
The p-value for Test 1 is less than  .05.  There appears to be a

                                     G-189

-------
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
The modeled variance appears
The model chosen seems
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =        1. 83952

            BMDL =      0.599228
                                     G-190

-------
G.2.30.3. Figure for Selected Model: Hill

                               Hill Model with 0.95 Confidence Level
        200
        150
 o
 Q.
 o:
 c
 (0
 OJ
100
         50
                   Hill
          0  - BMDL
                  BMD
                                                                        10
                                           dose
   10:5502/082010
G.2.30.4. Output for Additional Model Presented: Power, Unrestricted
Markowski et al. (2001): FR2 Revolutions
         Power Model.  (Version:  2.15;   Date: 04/07/2008)
         Input Data File: C:\l\Blood\34_Mark_2001_FR2rev_PowerCV_U_l.(d)
         Gnuplot Plotting File:
C:\l\Blood\34_Mark_2001_FR2rev_PowerCV_U_l.plt
                                            Mon Feb  08  10:55:49 2010
 Table  3


   The  form of the response  function is:

   Y[dose]  = control + slope  *  doseApower
   Dependent  variable = Mean
                                      G-191

-------
   Independent variable = Dose
   rho is set to 0
   The power is not restricted
   A constant variance model is fit

   Total number of dose groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      2598.74
                            rho =
                        control =
                          slope =
                          power =
            0
       119.29
     -10.3599
     0.824761
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
control
slope
power
alpha
1
-3e-010
6.9e-010
9.9e-010
control
-3e-010
1
-0.63
-0.28
slope
6.9e-010
-0.63
1
0.87
power
9.9e-010
-0.28
0.87
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          2350.22
3679.95
        control          120.082
155.514
          slope         -27.8164
19.7023
          power         0.317923
1.00556
Parameter Estimates



       Std.  Err.

         678.449

         18.0782

         24.2447

        0.350841
        95.0% Wald

     Lower Conf.  Limit

             1020.48

             84.6491

            -75.3352

           -0.369713
                                     G-192

-------
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
    0
1.557
 4.03
10.32
        119
        109
       56.5
       68.1
 120
88.1
76.8
61.7
69.9
  61
31.2
33.2
48.5
48.5
48.5
48.5
-0.0432
  0.843
  -1.02
  0.353
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -104.165520
             -101.140174
             -104.165520
             -105.147159
             -107.599268
          # Param's
                5
                8
                5
                4
                2
            AIC
         218.331040
         218.280349
         218.331040
         218.294317
         219.198536
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
-2*log(Likelihood Ratio)  Test df
            12.9182
            6.05069
                     p-value

                   0.04435
                    0.1092
                                     G-193

-------
   Test 3
   Test 4
6.05069
1.96328
0.1092
0.1612
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
                     The modeled variance appears
                     The model chosen seems
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 5.73906


            BMDL = 1.03181e-014
                                     G-194

-------
G.2.30.5. Figure for Additional Model Presented: Power, Unrestricted



                               Power Model with 0.95 Confidence Level
 o
 Q.
 ro
 CD
        200
         150
         100
         50
          0
           BMDL
   10:5502/082010
                                                                           10
                                          G-195

-------
G.2.31. Markowski et al. (2001): FR5 Run Opportunities
G.2.31.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hillb
Linear
Polynomial, 3 -degree
Power
Power, unrestricted0
Degrees of
freedom
2
2
1
0
1
2
2
2
1
X2 p-value
0.205
0.205
0.254
N/A
0.939
0.122
0.122
0.122
0.134
AIC
133.193
133.193
133.328
134.032
132.032
134.229
134.229
134.229
134.268
BMD (ng/kg)
5.078E+00
5.078E+00
2.160E+00
2.124E+00
1.723E+00
7.234E+00
7.234E+00
7.234E+00
2.666E+00
BMDL
(ng/kg)
2.439E+00
2.439E+00
6.854E-01
9.667E-01
9.085E-01
4.430E+00
4.430E+00
4.430E+00
1.032E-14
Notes

power hit bound (d = 1)


n upper bound hit
(» = 18)


power bound hit
(power =1)
unrestricted
(power =0.392)
a Constant variance model selected (p = 0.2262).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.2.31.2. Output for Selected Model: Hill
Markowski et al. (2001): FR5 Run Opportunities
         Hill Model.  (Version:  2.14;  Date:  06/26/2008)
         Input Data File:  C:\l\Blood\35_Mark_2001_FR5opp_HillCV_l.(d)
         Gnuplot Plotting  File:   C:\l\Blood\35_Mark_2001_FR5opp_HillCV_l.plt
                                            Mon  Feb 08 10:56:24 2010
 Table  3


   The  form of the response  function is:

   Y[dose]  = intercept +  v*doseAn/(kAn + doseAn)
   Dependent variable = Mean
   Independent variable =  Dose
   rho  is  set to 0
   Power parameter restricted to be greater  than 1
   A  constant variance model is fit

   Total number of dose groups = 4
   Total number of records  with missing values  = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to:  le-008
   Parameter Convergence has been set to: le-008
                                       G-196

-------
                  Default Initial Parameter Values
                          alpha =      77.4849
                            rho =
                      intercept =
                              v =
                              n =
                              k =
      0
  26.14
 -13.34
2.77257
2.48811
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho    -n
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
                                                      6.2e-008

                                                         -0.51

                                                          0.36

                                                             1
alpha
alpha 1
intercept -3.2e-009
v 1.9e-008
k 6.2e-008

Confidence Interval
Variable
Upper Conf. Limit
alpha
101.129
intercept
32.0935
v
-5.77257
n
k
2.08973
intercept v
-3.2e-009 1.9e-008
1 -0.81
-0.81 1
-0.51 0.36
Parameter Estimates

Estimate Std. Err.

64.5863 18.6445

26.14 3.03753

-13.1569 3.7676

18 NA
1.68073 0.208677

                                                         95.0% Wald

                                                      Lower Conf. Limit

                                                              28.0438

                                                              20.1865

                                                             -20.5413


                                                              1.27173
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
                                    G-197

-------
    0
1.557
 4.03
10.32
       26.1
       23.5
       12.8
       13.1
26.1
23.5
  13
  13
12.3
7.04
6.17
7.14
,04
,04
,04
,04
 -1.9e-008
-1.94e-007
   -0.0558
    0.0517
 Model Descriptions for likelihoods calculated
 Model Al:        Yij
           Var{e(ij)}

 Model A2:        Yij
           Var{e(ij)}
             Mu(i)  + e (i j'
             SigmaA2

             Mu(i)  + e (i j'
             Sigma(i)A2
 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -62.013133
              -59.839035
              -62.013133
              -62.016025
              -67.530040
          # Param's
                5
                8
                5
                4
                2
            AIC
         134.026266
         135.678070
         134.026266
         132.032049
         139.060081
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.
   Test

   Test 1
   Test 2
   Test 3
   Test 4
                     Tests of Interest
-2*log(Likelihood Ratio)  Test df
             15.382
             4.3482
             4.3482
         0.00578335
         6
         3
         3
         1
        p-value

      0.01748
       0.2262
       0.2262
       0.9394
The p-value for Test 1 is less than  .05.  There appears to be a

                                     G-198

-------
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
The modeled variance appears
The model chosen seems
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =        1.72335

            BMDL =      0.908491
                                     G-199

-------
G.2.31.3. Figure for Selected Model: Hill



                              Hill Model with 0.95 Confidence Level
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 0)
       40
        35
        30
       25
20
        15
        10
                 Hill
               BMDL
                 BMD
              0           2



   10:5602/082010
4           6

     dose
                                                                 10
G.2.31.4. Output for Additional Model Presented: Power, Unrestricted

Markowski et al. (2001): FR5 Run Opportunities
         Power Model.  (Version:  2.15;   Date: 04/07/2008)

         Input Data File: C:\l\Blood\35_Mark_2001_FR5opp_PwrCV_U_l.(d)

         Gnuplot Plotting File:   C:\l\Blood\35_Mark_2001_FR5opp_PwrCV_U_l.plt

                                            Mon  Feb  08  10:56:24 2010
 Table  3





   The  form of the response  function is:



   Y[dose]  = control + slope  *  doseApower





   Dependent variable = Mean
                                      G-200

-------
   Independent variable = Dose
   rho is set to 0
   The power is not restricted
   A constant variance model is fit

   Total number of dose groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      77.4849
                            rho =
                        control =
                          slope =
                          power =
            0
        26.14
      -2.3827
     0.844532
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
control
slope
power


-9
1
9
alpha
1
.3e-009
.4e-008
.3e-009
control
-9.3e-009
1
-0.64
-0.34
slope
1.4e-008
-0.64
1
0.9
power
9.3e-009
-0.34
0.9
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          70.8926
111.003
        control          26.3582
32.4909
          slope         -5.73309
2.16433
          power         0.391903
0.944342
Parameter Estimates



       Std.  Err.

         20.4649

         3.12902

         4.02937

        0.281862
        95.0% Wald

     Lower Conf.  Limit

             30.7821

             20.2254

            -13.6305

           -0.160536
                                     G-201

-------
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev    Scaled
    0
1.557
 4.03
10.32
       26.1
       23.5
       12.8
       13.1
26.4
19.5
16.5
  12
12.3
7.04
6.17
7.14
,42
,42
,42
,42
-0.0686
  0.941
  -1.06
  0.343
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -62.013133
              -59.839035
              -62.013133
              -63.134001
              -67.530040
          # Param's
                5
                8
                5
                4
                2
            AIC
         134.026266
         135.678070
         134.026266
         134.268002
         139.060081
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
-2*log(Likelihood Ratio)  Test df
             15.382
             4.3482
                     p-value

                   0.01748
                    0.2262
                                     G-202

-------
   Test 3
   Test 4
 4.3482
2.24174
0.2262
0.1343
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
                     The modeled variance appears
                     The model chosen seems
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 2.66625


            BMDL = 1.03181e-014
                                     G-203

-------
G.2.31.5. Figure for Additional Model Presented: Power, Unrestricted


                               Power Model with 0.95 Confidence Level
 c
 (0
 OJ
        40
        35
        30
       25

 o
 Q.

 a:      20
        15
        10
                  Power
          BMDL
BMD
        4            6

              dose
                                                                              10
   10:5602/082010
                                          G-204

-------
G.2.32. Miettinen et al. (2006): Cariogenic Lesions, Pups
G.2.32.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage, 4 -degree
Probit
Weibull
Gamma, unrestricted
Log-logistic,
unrestricted13
Log-probit, unrestricted
Weibull, unrestricted
Degrees of
freedom
3
3
3
3
3
3
3
2
2
2
2
X2 p-value
0.410
0.371
0.602
0.300
0.410
0.350
0.410
0.798
0.728
0.732
0.766
AIC
162.280
162.518
161.292
163.040
162.280
162.656
162.280
161.801
161.983
161.972
161.884
BMD (ng/kg)
3.401E+00
4.108E+00
1.428E-K)0
6.321E+00
3.401E+00
4.548E+00
3.401E+00
3.374E-03
4.942E-02
6.495E-02
1.792E-02
BMDL (ng/kg)
1.889E+00
2.450E+00
5.175E-01
3.127E+00
1.889E+00
2.889E+00
1.889E+00
8.884E-242
error
error
error
Notes
power bound hit
(power =1)

slope bound hit
(slope = 1)
slope bound hit
(slope =1)
final B = 0

power bound hit
(power =1)
unrestricted
(power =0.215)
unrestricted
(slope = 0.465)
unrestricted
(slope = 0.289)
unrestricted
(power =0.324)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.


G.2.32.2. Output for Selected Model: Log-Logistic
Miettinen et al. (2006): Cariogenic Lesions, Pups
         Logistic Model.  (Version: 2.12; Date:  05/16/2008)
         Input Data File:  C:\l\Blood\36_Miet_2006_Cariogenic_LogLogistic_l.(d)
         Gnuplot Plotting  File:
C:\l\Blood\36_Miet_2006_Cariogenic_LogLogistic_l.pit
                                            Mon  Feb 08 10:56:59  2010
 Table  2  converting the percentage into the  number of animals,  and control is
Control II  from the study.  Dose is in ng per kg and is from Table  1
   The  form of the probability function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
   Dependent variable =  DichEff
   Independent variable  =  Dose
                                       G-205

-------
   Slope parameter is restricted as slope >= 1

   Total number of observations = 5
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =     0.595238
                      intercept =       -2.494
                          slope =            1
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -slope
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

             background    intercept

background            1        -0.66

 intercept        -0.66            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background         0.644165

      intercept         -2.55354

          slope                1
                                 Parameter Estimates
                      Std.  Err.
                                                         95.0% Wald

                                                      Lower Conf. Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.5853
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -77.6769         5
      -78.646         2       1.93832      3
                                     G-206
                                                                    P-value

-------
  Reduced model
0.0259

           AIC:
    -83.2067
     161.292
             11.0597
Goodness
of Fit
Scaled

0
2
6
16
46
Dose
.0000
.2195
.2259
.0142
.6355
Est
0.
0.
0.
0.
0.
. Prob.
6442
6966
7603
8416
9231
Expected
27
20
19
20
29
.055
.200
.007
.198
.540
Observed
25.
23.
19.
20.
29.
000
000
000
000
000
Size
42
29
25
24
32
Residual
-0
1
-0
-0
-0
.662
.131
.003
.111
.358
 ChiA2 = 1.86
d.f. = 3
P-value = 0.6024
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        1.42805

            BMDL =       0.517495
                                     G-207

-------
G.2.32.3. Figure for Selected Model: Log-Logistic

                            Log-Logistic Model with 0.95 Confidence Level

                          i ^
          1
         0.9
 T3
 £
 O
 c
 O
 •*=
 O
 (0
         0.8  -
         0.7  -
Log-Logistic
         0.6  -
         0.5
         0.4 BflvlDL  BMP

               0            10           20           30           40
                                            dose
   10:5602/082010



G.2.32.4. Output for Additional Model Presented: Log-Logistic, Unrestricted

Miettinen et al. (2006): Cariogenic Lesions, Pups



         Logistic Model.  (Version:  2.12;  Date:  05/16/2008)
         Input  Data File:
C:\l\Blood\36_Miet_2006_Cariogenic_LogLogistic_U_l.(d)
         Gnuplot  Plotting File:
C:\l\Blood\36_Miet_2006_Cariogenic_LogLogistic_U_l.plt
                                            Mon  Feb  08 10:56:59 2010


 Table  2  converting the percentage into the number  of animals, and  control is
Control  II  from  the study. Dose  is in ng per kg  and is from Table 1
   The  form of the probability  function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]

                                      G-208

-------
   Dependent variable = DichEff
   Independent variable = Dose
   Slope parameter is not restricted

   Total number of observations = 5
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =     0.595238
                      intercept =    -0.739403
                          slope =     0.442847
           Asymptotic Correlation Matrix of Parameter Estimates

             background    intercept        slope

background            1        -0.51         0.24

 intercept        -0.51            1        -0.89

     slope         0.24        -0.89            1
Confidence Interval
       Variable
Upper Conf. Limit
     background
*
      intercept
*
          slope
          Parameter Estimates

                                  95.0% Wald

 Estimate        Std.  Err.      Lower Conf. Limit

 0.597745            *                *

-0.798024            *                *

 0.465259            *                *
* - Indicates that this value is not calculated.
                        Analysis of Deviance Table

       Model      Log(likelihood)   # Param's  Deviance  Test d.f.
     Full model        -77.6769         5
                                     G-209
                                             P-value

-------
   Fitted model
0.7301
  Reduced model
0.0259

           AIC:
    -77.9915

    -83.2067


     161.983
     3      0.629204      2

     1       11.0597      4
                                  Goodness  of  Fit
Dose
0.0000
2.2195
6.2259
16.0142
46.6355
Est. Prob.
0.5977
0.7566
0.8042
0.8474
0.8910
Expected
25.105
21.940
20.105
20.338
28.512
Observed
25.000
23.000
19.000
20.000
29.000
Size
42
29
25
24
32
Scaled
Residual
-0.033
0.458
-0.557
-0.192
0.277
 ChiA2 =0.63
d.f. = 2
P-value = 0.7281
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =       0.049422

           Benchmark dose computation failed.
                            Lower limit includes zero.
                                     G-210

-------
 T3

 £
 O
 c
 O
 •*=
 O
 (0
G.2.32.5. Figure for Additional Model Presented: Log-Logistic, Unrestricted



                                            Log-Logistic Model



            1                 ' "" ' """"*'"
          0.9
          0.8
          0.7
          0.6
          0.5
          0.4
Log-Logistic
                                10
                 20            30

                    dose
40
   10:5702/082010
                                            G-211

-------
G.2.33. Murray et al. (1979): Fertility in F2 Generation
G.2.33.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Multistage, 1 -degree
Multistage,
2-degreea
Probit
Weibull
Log-probit,
unrestricted
Degrees of
freedom
0
1
0
1
1
1
0
0
X2 p-value
N/A
0.051
N/A
0.031
0.079
0.048
N/A
N/A
AIC
61.729
61.318
61.729
63.154
60.464
61.544
61.729
61.729
BMD (ng/kg)
4.481E+00
2.420E+00
4.971E+00
1.598E+00
2.733E+00
2.250E+00
5.042E+00
4.244E+00
BMDL
(ng/kg)
1.590E+00
1.722E+00
1.565E+00
8.747E-01
1.366E+00
1.590E+00
1.604E+00
1.506E+00
Notes







unrestricted
(slope = 3. 182)
a Best-fitting model, BMDS output presented in this appendix.


G.2.33.2. Output for Selected Model: Multistage, 2-Degree
Murray et al. (1979): Fertility in F2 Generation
        Multistage Model.  (Version: 3.0;   Date:  05/16/2008)
        Input  Data File: C:\l\Blood\Murray_1979_fert_index_f2_Multi2_l.(d)
        Gnuplot Plotting File:
C:\l\Blood\Murray_1979_fert_index_f2_Multi2_l.pit
                                            Wed  Feb 10 16:06:28 2010
 Table  1 but  expressed as number of dams  who  do not produce offspring
   The  form of  the probability function  is:

   P[response]  = background +  (1-background)*[1-EXP(
                  -betal*doseAl-beta2*doseA2)]

   The  parameter betas are restricted  to be  positive
   Dependent  variable = DichEff
   Independent  variable = Dose

 Total number of observations = 3
 Total number of records with missing values
 Total number of parameters in model =  3
 Total number of specified parameters = 0
 Degree of  polynomial = 2
= 0
 Maximum number of iterations = 250
 Relative  Function Convergence has been  set  to:  le-008

                                      G-212

-------
 Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                     Background =    0.0567204
                        Beta(l) =            0
                        Beta(2) =    0.0155037
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -Beta(l)
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

             Background      Beta(2)

Background            1        -0.45

   Beta(2)        -0.45            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     Background        0.0780188
*
        Beta(l)                0
*
        Beta(2)        0.0141051
                                 Parameter Estimates
                      Std. Err.
                 95.0% Wald

              Lower Conf.  Limit
* - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.02805
  Reduced model
0.0002798

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -25.8194         3
     -28.2318         2       4.82474      1
     -34.0009
      60.4636
1
16.363
2
                            P-value
                                  Goodness  of  Fit

     Dose     Est. Prob.    Expected    Observed     Size

                                     G-213
                                               Scaled
                                              Residual

-------
0.0000
1.1242
5.8831
0.0780
0.0943
0.4341
2.497
1.886
8.683
4.000
0.000
9.000
32
20
20
0.991
-1.443
0.143
 ChiA2 = 3.08      d.f. = 1        P-value = 0.0790


   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        2.73307

            BMDL =        1.36619

            BMDU =        4.10938

Taken together,  (1.36619, 4.10938) is a 90     % two-sided confidence
interval for the BMD
                                     G-214

-------
G.2.33.3. Figure for Selected Model: Multistage, 2-Degree


                                Multistage Model with 0.95 Confidence Level


               ''''      ' Mu'ltis
          0.7




          0.6




          0.5
 T3

 £
 O
 c
 O
 •*=
 O
 (0
0.4
0.3   -  -T-
          0.2
          0.1
   16:0602/102010
G.2.34. National Toxicology Program (1982): Toxic Hepatitis, Male Mice

G.2.34.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probit
Multistage, 3-degreea
Probit
Weibull
Degrees of
freedom
1
2
1
1
1
2
1
X2/7-value
0.027
0.092
0.026
0.027
0.036
0.082
0.025
AIC
113.103
110.352
113.089
113.111
112.045
110.512
113.044
BMD (ng/kg)
3.823E+00
3.108E+00
3.797E+00
3.565E+00
2.782E+00
2.763E+00
3.967E+00
BMDL (ng/kg)
2.005E+00
2.465E+00
2.141E+00
2.294E+00
1.343E-K)0
2.241E+00
1.704E+00
Notes







' Best-fitting model, BMDS output presented in this appendix.
                                           G-215

-------
G.2.34.2. Output for Selected Model: Multistage, 3-Degree
National Toxicology Program (1982): Toxic Hepatitis, Male Mice
        Multistage Model.  (Version:  3.0;   Date:  05/16/2008)
        Input Data File: C:\l\Blood\37_NTP_1982_ToxHep_Multi3_l.(d)
        Gnuplot Plotting File:  C:\l\Blood\37_NTP_1982_ToxHep_Multi3_l.plt
                                           Mon  Feb  08  10:57:32 2010
   The form of the probability  function  is:

   P[response] = background +  (1-background)*[1-EXP(
                 -betal*doseAl-beta2*doseA2-beta3*doseA3)

   The parameter betas are restricted  to be  positive
   Dependent variable = DichEff
   Independent variable = Dose

 Total number of observations =  4
 Total number of records with missing  values  =  0
 Total number of parameters in model = 4
 Total number of specified parameters  = 0
 Degree of polynomial = 3
 Maximum number of iterations = 250
 Relative Function Convergence has been  set  to:  le-008
 Parameter Convergence has been set  to:  le-008
                  Default Initial  Parameter  Values
                     Background =     0.0471757
                        Beta(l) =    0.00749116
                        Beta(2) =             0
                        Beta(3) =    0.00139828
           Asymptotic Correlation Matrix  of  Parameter Estimates

            ( *** The model parameter(s)   -Beta(2)
                 have been estimated  at a boundary point,  or have been
specified by the user,
                 and do not appear  in the correlation matrix )

             Background      Beta(l)       Beta(3)

Background            1        -0.77          0.69


                                     G-216

-------
   Beta(l)

   Beta(3)
-0.77

 0.69
    1

-0.95
-0.95

    1
                                 Parameter Estimates
Confidence Interval
       Variable
Upper Conf. Limit
     Background
*
        Beta(l)
*
        Beta(2)
*
        Beta(3)
      Estimate

     0.0267933

     0.0283198

             0

     0.0012342
         Std. Err.
             95.0% Wald

          Lower Conf. Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.04777
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -51.0633         4
     -53.0224         3       3.91812      1
     -121.743

      112.045
         1
    141.358
3
                                     P-value
<.0001
                                  Goodness  of  Fit

Dose
0.0000
0.7665
2.2711
11.2437

Est. Prob.
0.0268
0.0482
0.1005
0.8775

Expected
1.956
2.363
4.925
43.877

Observed
1.000
5.000
3.000
44.000

Size
73
49
49
50
Scaled
Residual
-0.693
1.759
-0.915
0.053
 ChiA2 =4.41
 d.f. =1
    P-value = 0.0357
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        2.78201
                                     G-217

-------
             BMDL  =



             BMDU  =
                    1.34308



                     4.5214
Taken  together,  (1.34308,  4.5214  )  is  a 90

interval for the  BMD
                                           %  two-sided confidence
 O


 I


 O

 13
 (0
G.2.34.3. Figure for Selected Model: Multistage, 3-Degree



                              Multistage Model with 0.95 Confidence Level


           1  p.           Multistage
         0.8
         0.6
0.4
         0.2
                   BMDL,
                       BMD
                             4          6

                                    dose
                                                                       10
   10:5702/082010
                                        G-218

-------
G.2.35. National Toxicology Program (2006): Alveolar Metaplasia
G.2.35.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage, 5 -degree
Probit
Weibull
Gamma, unrestricted
Log-probit,
unrestricted
Weibull, unrestricted
Degrees of
freedom
4
4
3
4
4
4
4
3
o
J
o
J
X2 p-value
0.010
<0.001
0.723
0.024
0.010
0.001
0.010
0.426
0.696
0.522
AIC
320.093
343.283
312.558
318.680
320.093
347.071
320.093
314.011
312.677
313.492
BMD (ng/kg)
9.886E-01
2.389E+00
6.497E-01
1.566E+00
9.886E-01
2.542E+00
9.886E-01
1.642E-01
6.818E-01
2.644E-01
BMDL
(ng/kg)
8.393E-01
2.052E+00
3.751E-01
1.318E+00
8.393E-01
2.219E+00
8.393E-01
1.874E-02
2.740E-01
6.947E-02
Notes
power bound hit
(power =1)


slope bound hit
(slope =1)
final B = 0

power bound hit
(power =1)
unrestricted
(power =0.503)
unrestricted
(slope = 0.677)
unrestricted
(power = 0.661)
a Best-fitting model, BMDS output presented in this appendix.


G.2.35.2. Output for Selected Model: Log-Logistic
National Toxicology Program (2006): Alveolar Metaplasia
         Logistic Model.  (Version:  2.12;  Date:  05/16/2008)
         Input Data File: C:\l\Blood\40_NTP_2006_AlvMeta_LogLogistic_l.(d)
         Gnuplot Plotting File:
C:\l\Blood\40_NTP_2006_AlvMeta_LogLogistic_l.pit
                                            Mon Feb 08 10:58:58  2010
   The  form of the probability function  is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
   Dependent  variable = DichEff
   Independent variable = Dose
   Slope parameter is restricted as  slope >= 1

   Total number of observations =  6
   Total number of records with missing  values = 0
   Maximum number of iterations =  250
   Relative Function Convergence has been set to: le-008

                                      G-219

-------
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =    0.0377358
                      intercept =     -1.69494
                          slope =      1.12282
           Asymptotic Correlation Matrix of Parameter Estimates

             background    intercept        slope

background            1        -0.21          0.1

 intercept        -0.21            1        -0.93

     slope          0.1        -0.93            1
Confidence Interval
       Variable
Upper Conf. Limit
     background
*
      intercept
*
          slope
               Parameter Estimates

                                       95.0% Wald

      Estimate        Std.  Err.     Lower Conf. Limit

     0.0373462            *                *

      -1.70923            *                *

       1.13164            *                *
* - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.7227
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -152.615         6
     -153.279         3       1.32728      3
     -216.802

      312.558
1
128.374
                            P-value
<.0001
                                  Goodness  of  Fit

     Dose     Est._Prob.    Expected    Observed     Size


                                     G-220
                                               Scaled
                                              Residual

-------
0.0000
2.5565
5.6937
9.7882
16.5688
29.6953
0.0373
0.3682
0.5807
0.7162
0.8197
0.8976
1.979
19.881
30.776
37.243
43.446
46.674
2.000
19.000
33.000
35.000
45.000
46.000
53
54
53
52
53
52
0.015
-0.249
0.619
-0.690
0.555
-0.308
 ChiA2  =  1.33
            d.f. =3
P-value  = 0.7232
   Benchmark Dose Computation



Specified effect =              0.1



Risk Type        =       Extra  risk



Confidence level =            0.95



              BMD =         0.64971



             BMDL =        0.375051






G.2.35.3. Figure for Selected Model: Log-Logistic
                           Log-Logistic Model with 0.95 Confidence Level
 I
 "o
 ro
        0.8
        0.6
0.4
        0.2
          0  -

           EMDL
   10:5802/082010
                                                                         30
                                       G-221

-------
G.2.36. National Toxicology Program (2006): Eosinophilic Focus, Liver
G.2.36.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probit
Multistage, 5 -degree
Probit3
Weibull
Log-probit,
unrestricted
Degrees of
freedom
3
4
3
4
3
4
3
3
X2 p-value
0.293
0.405
0.152
0.192
0.752
0.459
0.324
0.116
AIC
331.902
330.400
333.515
332.312
329.328
329.945
331.628
334.150
BMD (ng/kg)
3.573E+00
5.949E+00
4.139E+00
4.889E+00
3.393E+00
5.583E-H)0
3.770E+00
4.146E+00
BMDL
(ng/kg)
2.225E+00
5.137E+00
2.077E+00
3.980E+00
2.466E+00
4.864E+00
2.249E+00
2.152E+00
Notes



slope bound hit
(slope =1)



unrestricted
(slope = 0.895)
a Best-fitting model, BMDS output presented in this appendix.


G.2.36.2. Output for Selected Model: Probit
National Toxicology Program (2006): Eosinophilic Focus, Liver
         Probit  Model.  (Version: 3.1;  Date:  05/16/2008)
         Input Data File: C:\l\Blood\45_NTP_2006_LivEosFoc_Probit_l.(d)
         Gnuplot Plotting File:  C:\l\Blood\45_NTP_2006_LivEosFoc_Probit_l.plt
                                            Mon Feb 08 11:00:54 2010
   The  form of  the probability function  is:

   P[response]  = CumNorm(Intercept+Slope*Dose) ,

   where  CumNorm(.)  is the cumulative normal  distribution function
   Dependent  variable = DichEff
   Independent  variable = Dose
   Slope parameter is not restricted

   Total number of observations = 6
   Total number of records with missing values  = 0
   Maximum  number of iterations = 250
   Relative Function Convergence has been  set  to:  le-008
   Parameter  Convergence has been set to:  le-008
                   Default Initial
                      background =
(and Specified)  Parameter Values
           0   Specified

  G-222

-------
                      intercept =
                          slope =
                    -1.28017
                   0.0712441
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -background
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
              intercept

 intercept            1

     slope        -0.77
             slope

             -0.77

                 1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
      intercept         -1.23453
-0.989279
          slope        0.0688678
0.085005
               Parameter Estimates



                      Std.  Err.

                       0.125132

                     0.00823346
                      95.0% Wald

                   Lower Conf. Limit

                          -1.47979

                         0.0527305
       Model
     Full model
   Fitted model
0.4331
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
      -161.07         6
     -162.972         2       3.80461      4
     -202.816

      329.945
     1
83.4925
                                 P-value
<.0001
                                  Goodness  of  Fit

0
2
5
9
16
29
Dose
.0000
.5565
.6937
.7882
.5688
.6953
Est
0.
0.
0.
0.
0.
0.
. Prob.
1085
1449
1998
2876
4628
7912
Exp
5
7
10
15
24
41
ected
.751
.826
.588
.242
.526
.932
Ob
3.
8.
14.
17.
22.
42.
served
000
000
000
000
000
000
Size
53
54
53
53
53
53
Scaled
Residual
-1.215
0.067
1.172
0.533
-0.696
0.023
 ChiA2 =3.62
 d.f.  =4
P-value = 0.4593
                                     G-223

-------
   Benchmark Dose  Computation


Specified effect =             0.1


Risk Type        =      Extra  risk


Confidence level =           0.95


              BMD =        5.58309


             BMDL =        4.86394




G.2.36.3. Figure for Selected Model: Probit



                               Probit Model with 0.95 Confidence Level
 •

 I
 o
 13
 (0
         0.8
         0.6
0.4
         0.2
                     Probit
                     BMDL  BMD
                                    10
                                     15

                                    dose
20
25
30
   11:0002/082010
                                        G-224

-------
G.2.37. National Toxicology Program (2006): Fatty Change Diffuse, Liver
G.2.37.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probit
Multistage, 5 -degree
Probit
Weibull3
Degrees of
freedom
4
4
4
4
3
4
4
X2 p-value
0.659
0.056
0.359
0.367
0.581
0.075
0.724
AIC
252.348
262.132
254.413
254.428
254.045
260.915
251.989
BMD (ng/kg)
4.028E+00
5.890E+00
4.254E+00
4.204E+00
3.524E+00
5.567E+00
3.917E+00
BMDL (ng/kg)
2.923E+00
5.042E+00
3.228E+00
3.277E+00
2.234E+00
4.784E+00
2.856E-H)0
Notes







a Best-fitting model, BMDS output presented in this appendix.


G.2.37.2. Output for Selected Model: Weibutt
National Toxicology Program (2006): Fatty Change Diffuse, Liver
        Weibull  Model using Weibull Model  (Version:  2.12;   Date: 05/16/2008)
        Input  Data File:  C:\l\Blood\47_NTP_2006_LivFatDiff_Weibull_l.(d)
        Gnuplot  Plotting File:
C:\l\Blood\47_NTP_2006_LivFatDiff_Weibull_l.plt
                                           Mon  Feb  08 11:01:56 2010
 NTP_liver_fatty_change_diffuse


   The  form  of  the probability function is:

   P[response]  =  background +  (1-background)*[1-EXP(-slope*doseApower)]
   Dependent  variable = DichEff
   Independent  variable = Dose
   Power parameter is restricted as power >=1

   Total number of observations = 6
   Total number of records with missing values  =  0
   Maximum  number of iterations = 250
   Relative Function Convergence has been set to:  le-008
   Parameter  Convergence has been set to: le-008
                   Default Initial  (and Specified)  Parameter Values
                      Background =   0.00925926
                           Slope =   0.00721355
                           Power =      1. 69678
                                      G-225

-------
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -Background
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

Slope
Power
Slope
1
-0.98
Power
-0.98
1
                                 Parameter Estimates
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     Background                0
          Slope        0.0135075
0.0260603
          Power          1.50444
1.83564
                      Std.  Err.

                             NA
                     0.00640459

                       0.168981
                      95.0% Wald

                   Lower Conf.  Limit


                        0.00095478

                           1.17324
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
       Model
     Full model
   Fitted model
0.7349
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -122.992         6
     -123.995         2       2.00444      4
     -204.846

      251.989
     1
163.708
                                 P-value
<.0001
                                  Goodness  of  Fit

0
2
5
9
16
29
Dose
.0000
.5565
.6937
.7882
.5688
.6953
Est
0.
0.
0.
0.
0.
0.
. Prob.
0000
0539
1688
3415
6024
8913
Exp
0
2
8
18
31
47
ected
.000
.912
.949
.102
.929
.238
Ob
0.
2.
12.
17.
30.
48.
served
000
000
000
000
000
000
Size
53
54
53
53
53
53
Scaled
Residual
0.000
-0.550
1.119
-0.319
-0.542
0.336
 ChiA2 =2.06
 d.f.  = 4
P-value = 0.7243
                                     G-226

-------
   Benchmark Dose Computation


Specified effect =             0.1


Risk Type        =       Extra risk


Confidence level =            0.95


              BMD =         3.91723


             BMDL =        2.85566




G.2.37.3. Figure for Selected Model: Weibull



                              Weibull Model with 0.95 Confidence Level
 •

 I
 o
 13
 (0
          1
         0.8
         0.6
0.4
         0.2
                     Weibul
                 BMDL  BMD
                                    10
                                     15

                                    dose
20
25
30
   11:01 02/082010
                                       G-227

-------
G.2.38. National Toxicology Program (2006): Gingival Hyperplasia, Squamous, 2 Years
G.2.38.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage, 5 -degree
Probit
Weibull
Gamma, unrestricted
Log-logistic,
unrestricted13
Log-probit, unrestricted
Weibull, unrestricted
Degrees of
freedom
4
4
4
4
4
4
4
3
3
3
3
X2 p-value
0.036
0.016
0.055
0.005
0.036
0.018
0.036
0.633
0.655
0.668
0.644
AIC
314.985
318.602
313.351
321.426
314.985
318.240
314.985
307.618
307.507
307.444
307.562
BMD (ng/kg)
7.743E+00
1.392E+01
5.850E+00
1.535E+01
7.743E+00
1.318E+01
7.743E+00
5.309E-01
7.049E-01
8.357E-01
6.143E-01
BMDL
(ng/kg)
5.166E+00
1.056E+01
3.730E-K)0
1.038E+01
5.166E+00
9.924E+00
5.166E+00
9.859E-07
1.260E-05
4.796E-05
3.872E-06
Notes
power bound hit
(power =1)

slope bound hit
(slope = 1)
slope bound hit
(slope = 1)
final B = 0

power bound hit
(power =1)
unrestricted
(power = 0.282)
unrestricted
(slope = 0.374)
unrestricted
(slope = 0.22)
unrestricted
(power = 0.325)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.


G.2.38.2. Output for Selected Model: Log-Logistic
National Toxicology Program (2006): Gingival Hyperplasia, Squamous, 2 Years
         Logistic Model.  (Version: 2.12; Date:  05/16/2008)
         Input Data File:  C:\l\Blood\42_NTP_2006_GingHypSq_LogLogistic_l.(d)
         Gnuplot Plotting  File:
C:\l\Blood\42_NTP_2006_GingHypSq_LogLogistic_l.pit
                                            Mon  Feb 08 10:59:57  2010
  [insert  study notes]
   The  form of the probability function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
   Dependent variable =  DichEff
   Independent variable  =  Dose
   Slope  parameter is restricted as slope  >=  1
                                       G-228

-------
   Total number of observations = 6
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =    0.0188679
                      intercept =     -3.75308
                          slope =            1
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -slope
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

             background    intercept

background            1        -0.79

 intercept        -0.79            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background        0.0671812
*
      intercept         -3.96371
*
          slope                1
                                 Parameter Estimates
                      Std.  Err.
   95.0% Wald

Lower Conf.  Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.05077
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
      -149.95         6
     -154.675         2       9.45085      4
              P-value
                                     G-229

-------
  Reduced model
0.0001186

           AIC:
    -162.631
     313.351
             25.3627
Goodness
of Fit
Scaled

0
2
5
9
16
29
Dose
.0000
.5565
.6937
.7882
.5688
.6953
Est
0.
0.
0.
0.
0.
0.
. Prob.
0672
1104
1582
2134
2905
4036
Expected
3
5
8
11
15
21
.561
.960
.385
.311
.394
.389
Observed
1.
7.
14.
13.
15.
16.
000
000
000
000
000
000
Size
53
54
53
53
53
53
Residual
-1
0
2
0
-0
-1
.405
.452
.113
.566
.119
.509
 ChiA2 = 9.26
d.f.  =4
P-value = 0.0550
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        5. 85026

            BMDL =         3.7296
                                     G-230

-------
G.2.38.3. Figure for Selected Model: Log-Logistic

                            Log-Logistic Model with 0.95 Confidence Level
 T3
 £
 O
 c
 O
 •*=
 O
 (0
         0.4
         0.3
0.2
         0.1
          0  -
   10:5902/082010
                                                                           30
G.2.38.4. Output for Additional Model Presented: Log-Logistic, Unrestricted
National Toxicology Program (2006): Gingival Hyperplasia, Squamous, 2 Years
         Logistic Model.  (Version:  2.12; Date:  05/16/2008)
         Input Data File:  C:\l\Blood\42_NTP_2006_GingHypSq_LogLogistic_U_l.(d)
         Gnuplot Plotting  File:
C:\l\Blood\42_NTP_2006_GingHypSq_LogLogistic_U_l.plt
                                            Mon  Feb 08 10:59:57 2010
  [insert  study notes]
   The  form of the probability function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
                                       G-231

-------
   Dependent variable = DichEff
   Independent variable = Dose
   Slope parameter is not restricted

   Total number of observations = 6
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =    0.0188679
                      intercept =         -2.2
                          slope =     0.424326
           Asymptotic Correlation Matrix of Parameter Estimates

             background    intercept        slope

background            1        -0.27         0.11

 intercept        -0.27            1        -0.93

     slope         0.11        -0.93            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background        0.0185138
*
      intercept         -2.06653
*
          slope         0.373721
                                 Parameter Estimates
                      Std.  Err.
   95.0% Wald

Lower Conf.  Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.6578
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
      -149.95         6
     -150.753         3       1.60697      3
              P-value
                                     G-232

-------
  Reduced model
0.0001186

           AIC:
    -162.631
     307.507
             25.3627
                                  Goodness  of  Fit

0
2
5
9
16
29
Dose
.0000
.5565
.6937
.7882
.5688
.6953
Est
0.
0.
0.
0.
0.
0.
. Prob.
0185
1681
2101
2433
2792
3230
Exp
0
9
11
12
14
17
ected
.981
.078
.136
.893
.795
.117
Ob
1.
7.
14.
13.
15.
16.
served
000
000
000
000
000
000
Size
53
54
53
53
53
53
Scaled
Residual
0.019
-0.756
0.966
0.034
0.063
-0.328
 ChiA2 =1.62
d.f.  =3
P-value = 0.6554
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =       0.704898

            BMDL =   1.26034e-005
                                     G-233

-------
G.2.38.5. Figure for Additional Model Presented: Log-Logistic, Unrestricted



                               Log-Logistic Model with 0.95 Confidence Level
 T3

 £
 O
 c
 O
 •*=
 O
 (0
          0.4
          0.3
0.2
          0.1
                             Log-Logistic
            BMDL
         BMD
                                       10
                                         15

                                       dose
20         25
30
   10:5902/082010
                                           G-234

-------
G.2.39. National Toxicology Program (2006): Hepatocyte Hypertrophy, 2 Years
G.2.39.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probit
Multistage, 5-degreea
Probit
Weibull
Gamma, unrestricted
Log-probit,
unrestricted
Weibull, unrestricted
Degrees of
freedom
5
4
4
5
4
4
5
4
4
4
X2 p-value
0.034
0.001
0.006
0.006
0.018
0.001
0.034
0.027
0.008
0.024
AIC
273.875
297.895
279.210
277.800
275.693
299.731
273.875
275.270
278.360
275.439
BMD (ng/kg)
9.091E-01
2.475E+00
1.137E+00
1.530E+00
9.272E-01
2.453E+00
9.091E-01
error
1.191E+00
7.345E-01
BMDL
(ng/kg)
7.868E-01
2.122E+00
6.491E-01
1.321E+00
7.906E-01
2.137E+00
7.868E-01
error
7.038E-01
3.588E-01
Notes
power bound hit
(power =1)





power bound hit
(power =1)
unrestricted
(power =0.844)
unrestricted
(slope = 0.864)
unrestricted
(power =0.92)
a Best-fitting model, BMDS output presented in this appendix.


G.2.39.2. Output for Selected Model: Multistage, 5-Degree
National Toxicology Program (2006): Hepatocyte Hypertrophy, 2 Years
        Multistage Model. (Version: 3.0;   Date:  05/16/2008)
        Input  Data File: C:\l\Blood\43_NTP_2006_HepHyper_Multi5_l.(d)
        Gnuplot Plotting File:  C:\l\Blood\43_NTP_2006_HepHyper_Multi5_l.plt
                                            Mon Feb 08 11:00:25 2010
  [insert  study notes]
   The  form of  the probability function  is:

   P[response]  = background +  (1-background)*[1-EXP(
                  -betal*doseAl-beta2*doseA2-beta3*doseA3-beta4*doseA4-
beta5*doseA5)]

   The  parameter betas are restricted to be  positive
   Dependent  variable = DichEff
   Independent  variable = Dose

 Total number of observations = 6
 Total number of records with missing values
 Total number of parameters in model =  6
 Total number of specified parameters = 0

                                      G-235
= o

-------
 Degree of polynomial = 5
 Maximum number of iterations = 250
 Relative Function Convergence has been set to: le-008
 Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                     Background =     0.112745
                        Beta(l) =    0.0950808
                        Beta(2) =            0
                        Beta(3) =            0
                        Beta (4) =            0
                        Beta (5) = 4.39515e-008
           Asymptotic Correlation Matrix of Parameter Estimates

            ( *** The model parameter(s)  -Background    -Beta (2)     -Beta (3)
-Beta(4)
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

                Beta(l)      Beta(5)

   Beta(l)            1         -0.5

   Beta(5)         -0.5            1
                                 Parameter Estimates

                                                          95.0% Wald
Confidence Interval
       Variable         Estimate        Std. Err.     Lower Conf.  Limit
Upper Conf. Limit
     Background                0            *                 *
*
        Beta(l)         0.113632            *                 *
*
        Beta(2)                0            *                 *
*
        Beta(3)                0            *                 *
*
        Beta(4)                0            *                 *
*
        Beta(5)     1.71322e-008            *                 *
*


* - Indicates that this value is not calculated.
                                     G-236

-------
       Model
     Full model
   Fitted model
0.01955
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -129.986         6
     -135.847         2       11.7216      4
      -219.97

      275.693
     1
179.968
                                 P-value
<.0001
                                  Goodness  of  Fit

0
2
5
9
16
29
Dose
.0000
.5565
.6937
.7882
.5688
.6953
Est
0.
0.
0.
0.
0.
0.
. Prob.
0000
2521
4764
6717
8510
9769
Expe
0.
13.
25.
35.
45.
51.
cted
000
614
251
599
106
778
Ob
0.
19.
19.
42.
41.
52.
served
000
000
000
000
000
000
Size
53
54
53
53
53
53
Scaled
Residual
0.000
1.688
-1.719
1.872
-1.584
0.203
 ChiA2 = 11.86
 d.f. =4
P-value = 0.0184
   Benchmark Dose Computation
Specified effect =

Risk Type

Confidence level =

             BMD =

            BMDL =

            BMDU =
            0.1

      Extra risk

           0.95

        0.92721

       0.790637

        1.14523
Taken together, (0.790637, 1.14523) is a 90
interval for the BMD
                              % two-sided confidence
                                     G-237

-------
G.2.39.3. Figure for Selected Model: Multistage, 5-Degree



                                 Multistage Model with 0.95 Confidence Level


               "          Miiltis

            1
          0.8
 T3

 £
 O
 c
 O
 •*=
 O
 (0
0.6
0.4
          0.2
                                                                                      30
   11:0002/082010
                                            G-238

-------
G.2.40. National Toxicology Program (2006): Necrosis, Liver
G.2.40.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probit
Multistage, 5 -degree
Probit
Weibull
Gamma, unrestricted
Log-logistic, unrestricted
Log-probit,
unrestricted"
Weibull, unrestricted
Degrees of
freedom
4
4
4
4
4
4
4
3
3
3
3
x2
p-value
0.939
0.601
0.943
0.572
0.939
0.666
0.939
0.883
0.860
0.805
0.879
AIC
234.400
236.742
234.382
236.863
234.400
236.293
234.400
236.290
236.377
236.598
236.302
BMD (ng/kg)
8.655E+00
1.484E+01
7.928E+00
1.333E+01
8.655E+00
1.393E+01
8.655E+00
7.726E+00
7.733E+00
7.501E+00
7.763E+00
BMDL
(ng/kg)
6.340E+00
1.240E+01
5.605E+00
1.024E+01
6.340E+00
1.154E+01
6.340E+00
3.453E+00
3.536E+00
3.504E+00
3.508E+00
Notes
power bound hit
(power =1)

slope bound hit
(slope = 1)
slope bound hit
(slope = 1)
final B = 0

power bound hit
(power =1)
unrestricted
(power = 0.87)
unrestricted
(slope = 0.974)
unrestricted
(slope = 0.517)
unrestricted
(power =0.895)
a Best-fitting model, BMDS output presented in this appendix.


G.2.40.2. Output for Selected Model: Log-Probit, Unrestricted
National Toxicology Program (2006): Necrosis, Liver
         Probit Model.  (Version:  3.1;   Date:  05/16/2008)
         Input Data File: C:\l\Blood\50_NTP_2006_LivNec_LogProbit_U_l.(d)
         Gnuplot Plotting File:
C:\l\Blood\50_NTP_2006_LivNec_LogProbit_U_l.plt
                                            Mon Feb 08  11:29:30 2010
 NTP liver  necrosis
   The  form of the probability  function is:

   P[response]  = Background
                + (1-Background)  *  CumNorm(Intercept+Slope*Log(Dose) ) ,

   where  CumNorm(.)  is the cumulative normal distribution  function
   Dependent  variable = DichEff
   Independent variable = Dose
   Slope  parameter is not restricted
                                      G-239

-------
   Total number of observations = 6
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial (and Specified) Parameter Values
                     background =    0.0188679
                      intercept =     -2.16223
                          slope =     0.457376
           Asymptotic Correlation Matrix of Parameter Estimates

             background    intercept        slope

background            1        -0.65         0.55

 intercept        -0.65            1        -0.97

     slope         0.55        -0.97            1
Confidence Interval
       Variable
Upper Conf. Limit
     background
0.065499
      intercept
-1.23311
          slope
0.879823
               Parameter Estimates

                                       95.0% Wald

      Estimate        Std.  Err.      Lower Conf. Limit

     0.0221151        0.0221351           -0.0212689

      -2.32352         0.556343             -3.41393

      0.517104         0.185064             0.154385
       Model
     Full model
   Fitted model
0.808
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -114.813         6
     -115.299         3      0.972184      3
      -127.98

      236.598
1
26.3331
                            P-value
<.0001
                                  Goodness  of  Fit

                                     G-240

-------

0
2
5
9
16
29
Dose
.0000
.5565
.6937
.7882
.5688
.6953
Est
0.
0.
0.
0.
0.
0.
. Prob.
0221
0544
0976
1457
2096
3002
Exp
1
2
5
7
11
15
ected
.172
.938
.174
.720
.106
.908
Ob
1.
4.
4.
8.
10.
17.
served
000
000
000
000
000
000
Size
53
54
53
53
53
53
Scaled
Residual
-0.161
0.637
-0.543
0.109
-0.373
0.327
 ChiA2  = 0.99
            d.f.  =3
P-value  = 0.8048
   Benchmark Dose  Computation




Specified effect =             0.1




Risk Type        =      Extra risk




Confidence level =            0.95




              BMD =         7.50077




             BMDL =          3.5039







G.2.40.3. Figure for Selected Model: Log-Probit, Unrestricted





                            LogProbit Model with 0.95 Confidence Level
 I
 c
 g

 t3
 ro
        0.5
        0.4
        0.3
0.2
        0.1
                      LogProbit
                 BMDL
                     BMD
                                  10
  11:2902/082010
                                   15


                                  dose






                                G-241
               20
25
30

-------
G.2.41. National Toxicology Program (2006): Oval Cell Hyperplasia
G.2.41.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probit
Multistage, 5 -degree
Probit"
Weibullb
Degrees of
freedom
o
J
4
o
3
o
5
o
3
4
o
J
X2 p-value
0.074
0.171
0.042
0.072
0.207
0.227
0.077
AIC
199.468
196.803
201.659
200.121
195.962
195.448
198.375
BMD (ng/kg)
6.739E+00
6.064E+00
6.936E+00
7.090E+00
4.785E+00
5.673E+00
5.718E+00
BMDL (ng/kg)
5.074E+00
5.145E+00
5.604E+00
5.931E+00
3.105E+00
4.793E+00
4.088E+00
Notes







a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.


G.2.41.2. Output for Selected Model: Probit

National Toxicology Program (2006): Oval Cell Hyperplasia
         Probit Model.  (Version:  3.1;   Date: 05/16/2008)
         Input Data File: C:\l\Blood\53_NTP_2006_OvalHyper_Probit_l.(d)
         Gnuplot Plotting File:   C:\l\Blood\53_NTP_2006_OvalHyper_Probit_l.plt
                                            Mon Feb 08  13:25:23 2010
   The  form of the probability  function is:

   P[response]  = CumNorm(Intercept+Slope*Dose) ,

   where  CumNorm(.)  is the cumulative normal distribution  function
   Dependent variable = DichEff
   Independent variable = Dose
   Slope  parameter is not restricted

   Total  number of observations  =  6
   Total  number of records with  missing values = 0
   Maximum number of iterations  =  250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been  set to: le-008
                   Default Initial  (and Specified) Parameter Values

                                      G-242

-------
                     background =
                      intercept =
                          slope =
                           0   Specified
                    -2.29925
                    0.169545
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -background
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
              intercept

 intercept            1

     slope        -0.87
             slope

             -0.87

                 1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
      intercept         -2.18988
-1.78217
          slope         0.172453
0.208211
               Parameter Estimates



                      Std.  Err.

                       0.208021

                      0.0182446
                      95.0% Wald

                   Lower Conf.  Limit

                           -2.5976

                          0.136694
       Model
     Full model
   Fitted model
0.1668
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -92.4898         6
     -95.7242         2       6.46873      4
     -210.191

      195.448
     1
235.402
                                 P-value
<.0001
                                  Goodness  of  Fit

0
2
5
9
16
29
Dose
.0000
.5565
.6937
.7882
.5688
.6953
Est
0.
0.
0.
0.
0.
0.
. Prob.
0143
0401
1135
3079
7478
9983
Expe
0.
2.
6.
16.
39.
52.
cted
756
168
017
317
631
911
Ob
0.
4.
3.
20.
38.
53.
served
000
000
000
000
000
000
Size
53
54
53
53
53
53
Scaled
Residual
-0.876
1.270
-1.306
1.096
-0.516
0.299
 ChiA2 =5.64
 d.f.  = 4
P-value = 0.2274
                                     G-243

-------
   Benchmark Dose  Computation



Specified effect =             0.1



Risk Type        =      Extra  risk



Confidence level =           0.95



              BMD =        5.67298



             BMDL =        4.79341





G.2.41.3. Figure for Selected Model: Probit
 T3
 0)
 C
 o
 •*=
 o
 (0
         0.8
         0.6
0.4
         0.2
                               Probit Model with 0.95 Confidence Level
                                                                             30
   13:2502/082010
                                        G-244

-------
G.2.41.4. Output for Additional Model Presented: Weibull
National Toxicology Program (2006): Oval Cell Hyperplasia
        Weibull Model using Weibull Model  (Version:  2.12;   Date:  05/16/2008)
        Input Data File: C:\l\Blood\53_NTP_2006_OvalHyper_Weibull_l.(d)
        Gnuplot Plotting File:
C:\l\Blood\53_NTP_2006_OvalHyper_Weibull_l.plt
                                           Mon  Feb  08 13:25:23 2010
   The form of the probability  function  is:

   P[response] = background +  (1-background)*[1-EXP(-slope*doseApower)]
   Dependent variable = DichEff
   Independent variable = Dose
   Power parameter is restricted as power  >=1

   Total number of observations =  6
   Total number of records with missing  values  =  0
   Maximum number of iterations =  250
   Relative Function Convergence has been  set to:  le-008
   Parameter Convergence has been  set  to:  le-008
                  Default Initial  (and  Specified)  Parameter Values
                     Background =    0.00925926
                          Slope =    0.00296825
                          Power =       2 . 17092
           Asymptotic Correlation Matrix  of  Parameter Estimates

             Background         Slope         Power

Background            1         -0.72           0.7

     Slope        -0.72             1         -0.99

     Power          0.7         -0.99             1



                                 Parameter Estimates

                                                          95.0% Wald
Confidence Interval
       Variable         Estimate        Std.  Err.      Lower Conf.  Limit
Upper Conf. Limit

                                     G-245

-------
     Background
0.0598245
          Slope
0.00559745
          Power
3.28628
0.0164137
0.00162074
2.39427
0.0221488
0.00202897
0.455116
-0.0269971
-0.00235596
1.50226
       Model
     Full model
   Fitted model
0.06031
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -92.4898         6
     -96.1875         3        7.3953      3
                                 P-value
Reduced model
AIC:
Dose
0.
2.
5.
9.
16.
29.
0000
5565
6937
7882
5688
6953
-210.191
198.375
Est. Prob.
0
0
0
0
0
0
.0164
.0314
.1138
.3285
.7440
.9957
1
Goodness
Expected
0.
1.
6.
17.
39.
52.
870
695
034
411
431
774
235.402
of Fit
Observed
0.
4.
3.
20.
38.
53.
000
000
000
000
000
000
Size
53
54
53
53
53
53
5 <.0
Scaled
Residual
-0
1
-1
0
-0
0
.940
.799
.312
.757
.450
.476
 ChiA2 = 6.85
 d.f.  = 3
P-value = 0.0770
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        5.71754

            BMDL =       4.08823
                                     G-246

-------
G.2.41.5.  Figure for Additional Model Presented: Weibull



                                Weibull Model with 0.95 Confidence Level


              "     '  Weibull' '

           1
          0.8
 T3

 £
 O
 c
 O
 •*=
 O
 (0
0.6
0.4
          0.2
                                                                                  30
   13:2502/082010
G.2.42. National Toxicology Program (2006): Pigmentation, Liver


G.2.42.1.  Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probita
Multistage, 5 -degree
Probit
Weibull
Degrees of
freedom
3
4
3
3
3
4
O
J
/2/7-value
0.552
0.247
0.984
0.962
0.058
0.004
0.219
AIC
196.971
197.066
195.530
195.526
199.955
200.504
199.007
BMD (ng/kg)
2.172E+00
1.853E+00
2.566E+00
2.463E+00
1.822E+00
1.710E+00
1.756E+00
BMDL (ng/kg)
1.493E+00
1.521E+00
1.937E+00
1.890E-H)0
9.916E-01
1.430E+00
1.190E+00
Notes




final B = 0


' Best-fitting model, BMDS output presented in this appendix.
                                          G-247

-------
G.2.42.2. Output for Selected Model: Log-Probit
National Toxicology Program (2006): Pigmentation, Liver
        Probit Model.  (Version:  3.1;   Date:  05/16/2008)
        Input Data File: C:\l\Blood\54_NTP_2006_Pigment_LogProbit_l.(d)
        Gnuplot Plotting File:
C:\l\Blood\54_NTP_2006_Pigment_LogProbit_l.plt
                                          Mon  Feb  08  13:25:55  2010
   The form of the probability  function  is:

   P[response] = Background
               +  (1-Background)  * CumNorm(Intercept+Slope*Log(Dose) ) ,

   where CumNorm(.) is the cumulative normal  distribution function
   Dependent variable = DichEff
   Independent variable = Dose
   Slope parameter is restricted as  slope  >=  1

   Total number of observations =  6
   Total number of records with missing  values  =  0
   Maximum number of iterations =  250
   Relative Function Convergence has been  set to:  le-008
   Parameter Convergence has been  set  to:  le-008
   User has chosen the log transformed model
                  Default Initial  (and  Specified)  Parameter Values
                     background =     0.0754717
                      intercept =     -2.48683
                          slope =       1.53221
           Asymptotic Correlation Matrix  of  Parameter  Estimates

             background    intercept         slope

background            1        -0.42          0.33

 intercept        -0.42             1         -0.96

     slope         0.33        -0.96             1
                                     G-248

-------
Confidence Interval
       Variable
Upper Conf. Limit
     background
0.138962
      intercept
-1.97787
          slope
2.31569
               Parameter Estimates

                                       95.0% Wald

      Estimate        Std.  Err.     Lower Conf. Limit

     0.0725473        0.0338856          0.00613263

      -2.93268         0.487158             -3.8875

       1.83184         0.246868             1.34798
       Model
     Full model
   Fitted model
0.9617
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -94.6177         6
     -94.7632         3      0.291072      3
     -210.717

      195.526
     1
232.198
                                 P-value
<.0001
                                  Goodness  of  Fit

0
2
5
9
16
29
Dose
.0000
.5565
.6937
.7882
.5688
.6953
Est
0.
0.
0.
0.
0.
0.
. Prob.
0725
1769
6291
9013
9874
9995
Exp
3
9
33
47
52
52
ected
.845
.553
.342
.771
.334
.974
Ob
4.
9.
34.
48.
52.
53.
served
000
000
000
000
000
000
Size
53
54
53
53
53
53
Scaled
Residual
0.082
-0.197
0.187
0.105
-0.412
0.160
 ChiA2 = 0.29
 d.f.  = 3
P-value = 0.9624
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        2.46293

            BMDL =        1.88981
                                     G-249

-------
G.2.42.3.  Figure for Selected Model: Log-Probit



                               LogProbit Model with 0.95 Confidence Level
 T3

 £
 O
 c
 O
 •*=
 O
 (0
           1
         0.8
0.6
         0.4
         0.2
                        LogProbit
                BMDL
             BMD
                                      10
                                        15

                                      dose
20
25
30
   13:2502/082010
G.2.43. National Toxicology Program (2006): Toxic Hepatopathy


G.2.43.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probit
Multistage, 5-degreea
Probit
Weibull
Degrees of
freedom
4
4
3
3
4
4
4
X2/7-value
0.754
0.159
0.391
0.394
0.693
0.231
0.716
AIC
185.763
191.136
189.577
189.580
185.924
189.820
185.785
BMD (ng/kg)
4.302E+00
4.833E+00
4.697E+00
4.972E+00
3.980E-H)0
4.621E+00
4.089E+00
BMDL (ng/kg)
3.463E+00
4.068E+00
3.818E+00
3.780E+00
3.059E-H)0
3.860E+00
3.215E+00
Notes




final 15 = 0


' Best-fitting model, BMDS output presented in this appendix.
                                          G-250

-------
G.2.43.2. Output for Selected Model: Multistage, 5-Degree
National Toxicology Program (2006): Toxic Hepatopathy
        Multistage Model.  (Version:  3.0;   Date:  05/16/2008)
        Input Data File: C:\l\Blood\55_NTP_2006_ToxHepa_Multi5_l.(d)
        Gnuplot Plotting File:  C:\l\Blood\55_NTP_2006_ToxHepa_Multi5_l.plt
                                           Mon  Feb  08  13:26:28  2010
   The form of the probability  function  is:

   P[response] = background +  (1-background)*[1-EXP(
                 -betal*doseAl-beta2*doseA2-beta3*doseA3-beta4*doseA4-
beta5*doseA5)]

   The parameter betas are restricted  to be  positive
   Dependent variable = DichEff
   Independent variable = Dose

 Total number of observations =  6
 Total number of records with missing values  =  0
 Total number of parameters in model =  6
 Total number of specified parameters = 0
 Degree of polynomial = 5
 Maximum number of iterations = 250
 Relative Function Convergence has been  set  to:  le-008
 Parameter Convergence has been set  to:  le-008
                  Default Initial  Parameter Values
                     Background =             0
                        Beta(l) =             0
                        Beta(2) =             0
                        Beta(3) =             0
                        Beta(4) =             0
                        Beta(5) =  4.36963e+012
           Asymptotic Correlation Matrix  of  Parameter Estimates

            ( *** The model parameter(s)   -Background    -Beta(l)     -Beta(4)
-Beta(5)
                 have been estimated  at a boundary point,  or have been
specified by the user,
                 and do not appear  in the correlation matrix )


                                     G-251

-------
   Beta (2)

   Beta(3)
Beta(2)

      1

  -0.95
Beta(3)

  -0.95

      1
Confidence Interval
       Variable
Upper Conf. Limit
     Background
*
        Beta(l)
*
        Beta(2)
*
        Beta (3)
*
        Beta (4)
*
        Beta(5)
        Estimate

               0

               0

      0.00639021

     6.5404e-005

               0

               0
                                 Parameter Estimates
           Std. Err.
         95.0% Wald

      Lower Conf. Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.6792
  Reduced model

           AIC:
        Analysis of Deviance Table

  Log(likelihood)   # Param's  Deviance  Test d.f.
       -89.8076         6
       -90.9619         2       2.30853      4
       -218.207

        185.924
           1
256.799
                                       P-value
<.0001
                                  Goodness  of  Fit

0
2
5
9
16
29
Dose
.0000
.5565
.6937
.7882
.5688
.6953
Est
0.
0.
0.
0.
0.
0.
. Prob.
0000
0420
1969
4901
8715
9994
Expe
0.
2.
10.
25.
46.
52.
cted
000
265
434
976
189
966
Ob
0.
2.
8.
30.
45.
53.
served
000
000
000
000
000
000
Size
53
54
53
53
53
53
Scaled
Residual
0.000
-0.180
-0.841
1.106
-0.488
0.185
 ChiA2 = 2.23
   d.f. =4
      P-value = 0.6928
   Benchmark Dose Computation
                                     G-252

-------
Specified effect =



Risk Type



Confidence level =



              BMD =



             BMDL =



             BMDU =
                       0.1



                 Extra risk



                      0.95



                   3.98025



                   3.05855



                   4.89735
Taken together,  (3.05855,  4.89735)  is  a 90

interval  for the BMD
                                          %  two-sided  confidence
G.2.43.3. Figure for Selected Model: Multistage, 5-Degree



                             Multistage Model with 0.95 Confidence Level
 •

 I
 C
 o
 •*=
 o
 (0
          1
         0.8
         0.6
0.4
         0.2
                        Multistage
                 BMDL  BMD
                                    10
                                     15

                                   dose
20
25
30
   13:2602/082010
                                       G-253

-------
G.2.44. Ohsako et al. (2001): Ano-Genital Length, PND 120
G.2.44.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hillb
Linear
Polynomial, 4 -degree
Power
Hill, unrestricted0
Power, unrestricted
Degrees of
freedom
3
3
2
1
2
3
3
3
1
2
X2 p-value
0.027
0.027
0.106
0.049
0.154
0.025
0.025
0.025
0.056
0.153
AIC
171.073
171.073
168.392
169.789
167.647
171.258
171.258
171.258
169.555
167.654
BMD (ng/kg)
2.592E+01
2.592E+01
2.248E+00
2.193E+00
2.879E+00
2.700E+01
2.700E+01
2.700E+01
3.494E+00
4.151E+00
BMDL
(ng/kg)
1.750E+01
1.750E+01
8.445E-01
9.382E-01
8.028E-01
1.881E+01
1.881E+01
1.881E+01
3.046E-01
2.395E-01
Notes

power hit bound (d = 1)


n lower bound hit
(« = 1)


power bound hit
(power =1)
unrestricted (n = 0.591)
unrestricted
(power =0.291)
a Constant variance model selected (p = 0.165).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.2.44.2. Output for Selected Model: Hill
Ohsako et al. (2001): Ano-Genital Length, PND 120
         Hill Model.  (Version:  2.14;  Date:  06/26/2008)
         Input Data File:  C:\l\Blood\56_Ohsako_2001_Anogen_HillCV_l.(d)
         Gnuplot Plotting  File:   C:\l\Blood\56_Ohsako_2001_Anogen_HillCV_l.plt
                                            Mon  Feb 08 13:27:02 2010
 Figure  7


   The  form of the response  function is:

   Y[dose]  = intercept + v*doseAn/(kAn + doseAn)
   Dependent variable = Mean
   Independent variable =  Dose
   rho  is  set to 0
   Power parameter restricted to be greater than  1
   A  constant variance model  is  fit

   Total number of dose groups = 5
   Total number of records  with  missing values  =  0
   Maximum number of iterations  = 250
   Relative Function Convergence has been set to:  le-008
   Parameter Convergence has  been set to: le-008

                                       G-254

-------
                  Default Initial Parameter Values
                          alpha =      7.27386
                            rho =
                      intercept =
                              v =
                              n =
                              k =
      0
 28.905
-5.1065
1.57046
 2.4317
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho    -n
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
                                                      7.2e-008

                                                         -0.52

                                                         -0.23

                                                             1
alpha
alpha 1
intercept 4.4e-008
v -9.8e-008
k 7.2e-008

Confidence Interval
Variable
Upper Conf. Limit
alpha
9.7422
intercept
30.4431
v
-2.9672
n
k
6.70661
intercept v
4.4e-008 -9.8e-008
1 -0.57
-0.57 1
-0.52 -0.23
Parameter Estimates

Estimate Std. Err.

7.07394 1.36138

28.9732 0.74996

-5.02686 1.05086

1 NA
2.56203 2.11462

                                                         95.0% Wald

                                                      Lower Conf. Limit

                                                              4.40568

                                                              27.5034

                                                             -7.08651


                                                             -1.58255
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
                                    G-255

-------
    0
 1.04
3.471
11.36
38.42
12
10
10
10
12
28.9
27.9
25.2
  26
23.8
  29
27.5
26.1
24.9
24.3
3.13
 2.5
3.21
2.85
1.56
2.66
2.66
2.66
2.66
2.66
-0.0889
  0.495
  -1.09
   1.35
 -0.602
 Model Descriptions for likelihoods calculated
 Model Al:         Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:         Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:         Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model  A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:          Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
              Log(likelihood)
                -77.952340
                -74.703868
                -77.952340
                -79.823277
                -89.824703
                       # Param's
                             6
                            10
                             6
                             4
                             2
                         AIC
                      167.904680
                      169.407736
                      167.904680
                      167.646555
                      183.649405
                   Explanation of Tests

 Test 1:   Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:   Are Variances Homogeneous?  (Al vs A2)
 Test 3:   Are variances adequately modeled?  (A2 vs. A3)
 Test 4:   Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
  -2*log(Likelihood Ratio)   Test df
              30.2417
              6.49694
              6.49694
              3.74187
                                  p-value

                              0.0001916
                                  0.165
                                  0.165
                                  0.154
                                     G-256

-------
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
The modeled variance appears
The model chosen seems
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =        2.87863

            BMDL =      0.802782
                                     G-257

-------
G.2.44.3. Figure for Selected Model: Hill

                              Hill Model with 0.95 Confidence Level
31
30
29

-------
   Dependent variable = Mean
   Independent variable = Dose
   rho is set to 0
   Power parameter is not restricted
   A constant variance model is fit

   Total number of dose groups = 5
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      7.27386
                            rho =
                      intercept =
                              v =
                              n =
                              k =
            0
       28.905
      -5.1065
      1.57046
       2.4317
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
intercept
V
n
k
alpha
1
-3. le-008
7.5e-009
1.7e-008
-8.8e-009
intercept
-3. le-008
1
0.001
0.0016
-0.13
V
7.5e-009
0.001
1
0.98
-0.99
n
1.7e-008
0.0016
0.98
1
-0.97

k
-8.8e-009
-0.
-0.
-0.

13
99
97
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          7.06192
9.72564
      intercept          28.9618
30.4404
              v         -6.82284
14.9532
Parameter Estimates



       Std.  Err.

         1.35907

        0.754441

         11.1104
        95.0% Wald

     Lower Conf.  Limit

              4.3982

             27.4831

            -28.5989
                                     G-259

-------
2.62979

101.553
               0.591421

                7.47064
                           1.04

                         48.002
                              -1.44695

                              -86.6115
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev
                                                           Scaled
    0
 1.04
3.471
11.36
38.42
12
10
10
10
12
28.9
27.9
25.2
  26
23.8
  29
27.3
26.3
25.1
  24
3.13
 2.5
3.21
2.85
1.56
2.66
2.66
2.66
2.66
2.66
-0.074
  0.71
 -1.36
  1.04
-0.284
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model      Log(likelihood)
             Al          -77.952340
             A2          -74.703868
             A3          -77.952340
         fitted          -79.777354
              R          -89.824703
# Param's
6
10
6
5
2
AIC
167.904680
169.407736
167.904680
169.554709
183.649405
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
                                     G-260

-------
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
-2*log(Likelihood Ratio)   Test df
            30.2417
            6.49694
            6.49694
            3.65003
    p-value

0.0001916
    0.165
    0.165
  0.05607
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =        3.49389

            BMDL =      0.304602
                                     G-261

-------
G.2.44.5. Figure for Additional Model Presented: Hill, Unrestricted



                                  Hill Model with 0.95 Confidence Level
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 0)
31



30




29



28




27



26




25



24




23



22
                   Hill
           BMDL
BMD
                                 10
                                 15
                            20

                          dose
25
30
35
40
   13:2702/082010
                                           G-262

-------
G.2.45. Sewall et al. (1995): T4 In Serum
G.2.45.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hillb
Linear
Polynomial, 4 -degree
Power
Hill, unrestricted0
Power, unrestricted
Degrees of
freedom
3
3
2
2
2
3
3
3
1
2
X2 p-value
0.722
0.722
0.854
0.854
0.898
0.576
0.576
0.576
0.864
0.985
AIC
204.495
204.495
205.483
205.483
205.382
205.150
205.150
205.150
207.196
205.197
BMD (ng/kg)
1.869E+01
1.869E+01
1.106E+01
1.106E+01
1.031E+01
2.238E+01
2.238E+01
2.238E+01
9.706E+00
9.726E+00
BMDL
(ng/kg)
1.243E+01
1.243E+01
4.650E+00
4.650E+00
3.603E+00
1.619E+01
1.619E+01
1.619E+01
1.973E+00
1.914E+00
Notes

power hit bound (d = 1)

power hit bound (d = 1)
n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted (n = 0.569)
unrestricted
(power =0.538)
a Constant variance model selected (p = 0.4078).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.2.45.2. Output for Selected Model: Hill
Sewall et al. (1995): T4 In Serum
         Hill Model.  (Version:  2.14;  Date:  06/26/2008)
         Input Data File:  C:\l\Blood\58_Sewall_1995_T4_HillCV_l.(d)
         Gnuplot Plotting  File:   C:\l\Blood\58_Sewall_1995_T4_HillCV_l.plt
                                            Mon  Feb 08 13:28:15  2010
 Figure  1,  Saline noninitiated


   The  form of the response  function is:

   Y[dose]  = intercept +  v*doseAn/(kAn  + doseAn)
   Dependent variable = Mean
   Independent variable =  Dose
   rho  is  set to 0
   Power parameter restricted to be greater  than 1
   A  constant variance model is fit

   Total number of dose groups = 5
   Total number of records  with missing values  = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set  to:  le-008
   Parameter Convergence has been set to:  le-008
                                       G-263

-------
                  Default Initial Parameter Values
                          alpha =      33.0913
                                                 Specified
      rho =
intercept =
        v =
        n =
        k =
       0
 30.6979
-12.2937
0.950815
 12.5808
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho    -n
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
                                                      1.5e-008

                                                         -0.65

                                                         -0.89

                                                             1
alpha
alpha 1
intercept -1.2e-009
v -1.8e-008
k 1.5e-008

Confidence Interval
Variable
Upper Conf. Limit
alpha
41.7679
intercept
33.7031
v
-3.10154
n
k
76.7035
intercept v
-1.2e-009 -1.8e-008
1 0.3
0.3 1
-0.65 -0.89
Parameter Estimates

Estimate Std. Err.

29.5556 6.23087

30.3957 1.68747

-18.2488 7.72836

1 NA
24.2883 26.743

                                                         95.0% Wald

                                                      Lower Conf. Limit

                                                              17.3433

                                                              27.0883

                                                             -33.3961


                                                              -28.127
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
                                    G-264

-------
    0
3.291
7.107
16.63
44.66
       30.7
       27.9
       25.9
       23.6
       18.4
30.4
28.2
26.3
  23
18.6
4.66
7.17
6.81
5.38
4.12
5.44
5.44
5.44
5.44
5.44
  0.167
 -0.188
 -0.204
  0.319
-0.0942
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -98.583448
              -96.590204
              -98.583448
              -98.691143
             -109.013252
          # Param's
                6
               10
                6
                4
                2
            AIC
         209.166896
         213.180407
         209.166896
         205.382286
         222.026503
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
-2*log(Likelihood Ratio)  Test df
            24.8461
            3.98649
            3.98649
            0.21539
                     p-value

                  0.001651
                    0.4078
                    0.4078
                    0.8979
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
                                     G-265

-------
It seems appropriate to model the data

The p-value for Test 2 is greater than .1.
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
A homogeneous variance
The modeled variance appears
The model chosen seems
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =         10.306

            BMDL =       3.60269
                                     G-266

-------
G.2.45.3. Figure for Selected Model: Hill

                                Hill Model with 0.95 Confidence Level
                !	i-iiii"'-	
             35
             30
        c
        o
        Q.
       ce
             25
             20
             15
                   BMDL
BMD
                  0     5    10    15     20    25    30     35    40    45
                                          dose
         13:2802/082010
G.2.45.4. Output for Additional Model Presented: Hill, Unrestricted
Sewall et al. (1995): T4 In Serum
        Hill  Model.  (Version:  2.14;   Date: 06/26/2008)
        Input Data File: C:\l\Blood\58_Sewall_1995_T4_HillCV_U_l.(d)
        Gnuplot Plotting File:   C:\l\Blood\58_Sewall_1995_T4_HillCV_U_l.plt
                                            Mon  Feb  08  13:28:15 2010
 Figure 1,  Saline noninitiated


   The form of  the response  function is:

   Y[dose]  = intercept + v*doseAn/(kAn + doseAn)
   Dependent  variable = Mean
   Independent  variable = Dose
   rho is  set to 0
   Power parameter is not restricted
   A constant variance model  is  fit

   Total number of dose groups = 5
                                      G-267

-------
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to:  le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      33.0913
                                                 Specified
      rho =
intercept =
        v =
        n =
        k =
       0
 30.6979
-12.2937
0.950815
 12.5808
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point,  or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
intercept
V
n
k
alpha
1
-3.9e-005
0.00022
0.00021
-0.00022
intercept
-3.9e-005
1
-0.17
-0.31
0.18
V
0.00022
-0.17
1
0.97
-1
n
0.00021
-0.31
0.97
1
-0.98
k
-0.00022
0.18
-1
-0.98
1
                                 Parameter Estimates
Confidence Interval
Variable
Upper Conf. Limit
alpha
41.5957
intercept
34.2336
V
7642.29
n
2.42564
k
338374

Estimate

29.4337

30.7096

-143.244

0.569063

2856.29


Std. Err.

6.20518

1.79801

3972.28

0.947248

171186

                                                         95.0% Wald

                                                      Lower Conf.  Limit

                                                              17.2718

                                                              27.1855

                                                             -7928.78

                                                             -1.28751

                                                              -332662
     Table of Data and Estimated Values of Interest

                                    G-268

-------
 Dose
Res .
    0
3.291
7.107
16.63
44.66
                 Obs Mean
                              Est Mean
                              Obs Std Dev  Est Std Dev
                                Scaled
       30.7
       27.9
       25.9
       23.6
       18.4
30.
27.
26.
23.
18.
7
7
1
4
4
4.
7.
6.
5.
4.
66
17
81
38
12
5.
5.
5.
5.
5.
43
43
43
43
43
-0.
0
-
0
-0.
00646
.0842
0.134
.0657
00948
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -98.583448
              -96.590204
              -98.583448
              -98.598183
             -109.013252
     #  Param's
           6
          10
           6
           5
           2
       AIC
    209.166896
    213.180407
    209.166896
    207.196367
    222.026503
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
-2*log(Likelihood Ratio)  Test df
            24.8461
            3.98649
    4

G-269
   p-value

0.001651
  0.4078

-------
   Test 3
   Test 4
  3.98649
0.0294713
0.4078
0.8637
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
                       The modeled variance appears
                       The model chosen seems
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =        9.70574

            BMDL =       1.97319
                                     G-270

-------
G.2.45.5. Figure for Additional Model Presented: Hill, Unrestricted


                                  Hill Model with 0.95 Confidence Level


            :'	i-iiii	
        35
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 0)
        30
25
        20
        15
             BMDL
   13:2802/082010
                      BMD
                              10      15     20      25      30     35      40     45

                                                dose
                                           G-271

-------
G.2.46. Shi et al. (2007): Estradiol 17B, PE9
G.2.46.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Hill
Linear
Polynomial, 4 -degree
Power
Hill, unrestricted
Power, unrestricted
Degrees of
freedom
3
3
2
2
2
3
3
3
1
2
X2 p-value
0.010
0.010
0.690
0.690
0.975
0.003
0.003
0.003
0.897
0.506
AIC
391.638
391.638
382.969
382.969
382.278
394.308
394.308
394.308
384.243
383.590
BMD (ng/kg)
6.976E+00
6.976E+00
8.068E-01
8.068E-01
7.239E-01
9.841E+00
9.841E+00
9.841E+00
7.086E-01
6.280E-01
BMDL
(ng/kg)
3.761E+00
3.761E+00
3.544E-01
3.544E-01
error
6.687E+00
6.687E+00
6.687E+00
error
3.304E-02
Notes

power hit bound (d=\)

power hit bound (d=\)
n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted (n = 0.875)
unrestricted
(power = 0.222)
a Nonconstant variance model selected (p = 0.0521).
b Best-fitting model, BMDS output presented in this appendix.
G.2.46.2. Output for Selected Model: Exponential (M4)
Shi et al. (2007): Estradiol 17B, PE9
         Exponential Model.  (Version:  1.61;  Date: 7/24/2009)
         Input Data File: C:\l\Blood\59_Shi_2007_Estradiol_Exp_l.(d)
         Gnuplot Plotting File:
                                            Mon Feb 08  13:28:52  2010
 Figure  4  PE9 only
   The  form of the response  function by Model:
      Model 2:      Y[dose] = a  *  exp{sign * b * dose}
      Model 3:      Y[dose] = a  *  exp{sign *  (b * dose)Ad}
      Model 4:      Y[dose] = a  *  [c-(c-l) * exp{-b * dose}]
      Model 5:      Y[dose] = a  *  [c-(c-l) * exp{-(b *  dose)Ad}]

    Note:  Y[dose]  is the median response for exposure  =  dose;
           sign = +1 for increasing  trend in data;
           sign = -1 for decreasing  trend.

      Model 2 is nested within  Models 3 and 4.
      Model 3 is nested within  Model 5.
      Model 4 is nested within  Model 5.
   Dependent variable = Mean
   Independent variable = Dose
   Data  are assumed to be distributed:  normally
   Variance Model:  exp(lnalpha  +rho  *ln(Y[dose])

                                      G-272

-------
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  * rho)

Total number of dose groups = 5
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 4
                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                          2.65881
                         0.913414
                              108
                         0.277637
                         0.340136
                                1
                  Parameter Estimates

                Variable          Model 4
                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                         1.66773
                         1.15314
                         103.146
                         1.00685
                        0.418742
                               1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
  0.3418
  1.075
   5.23
  13.91
10
 10
10
10
10
102.9
86.19
63.33
48.1
38.57
41.41
19.58
29.36
18.82
22.59
   Dose

      0
 0.3418
  1.075
   5.23
  13.91
               Estimated Values of Interest

             Est Mean      Est Std     Scaled Residual
    103.1
    85.69
    63.51
     43.5
    43.19
33.35
29.96
25.21
20.27
20.19
-0.02738
 0.05296
-0.02238
  0.7167
 -0.7237
                                  G-273

-------
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)

     Model  R:        Yij = Mu + e(i)
               Var{e(ij)} = SigmaA2
                     Model
             Likelihoods of Interest

             Log(likelihood)       DF
                                                                AIC
Al
A2
A3
R
4
-188.3615
-183.667
-186.1132
-203.3606
-186.4844
6
10
7
2
5
388.7231
387.3339
386.2263
410.7211
382.9687
   Additive constant for all log-likelihoods =     -45.95.  This  constant
added to the
   above values gives the log-likelihood including the term that  does  not
   depend on the model parameters.
                                 Explanation of Tests

   Test 1:  Does response and/or variances differ among Dose  levels?  (A2  vs.
R)
   Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
   Test 3:  Are variances adequately modeled?  (A2 vs. A3)

   Test 6a: Does Model 4 fit the data?  (A3 vs  4)
     Test

     Test 1
     Test 2
     Test 3
    Test 6a
         Tests  of Interest

-2*log(Likelihood Ratio)
                                                  D. F.
p-value
39.39
9.389
4.892
0.7424
8
4
3
2
< 0.0001
0.05208
0.1798
0.6899
     The p-value for Test 1 is less than  .05.  There appears  to be  a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

                                     G-274

-------
  The p-value for Test 2 is less than .1.  A non-homogeneous
  variance model appears to be appropriate.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is greater than  .1.  Model 4 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =     0.806817

              BMDL =     0.354366
                                 G-275

-------
G.2.46.3. Figure for Selected Model: Exponential (M4)




                           Exponential Model 4 with 0.95 Confidence Level
c


I

ce

c
ro
01
        140
        120
        100
         80
         60
         40
         20
                        Exponential
             3MDL
                  BMD
                                         6        8


                                            dose
10       12
                                                                          14
   13:2802/082010
                                            G-276

-------
G.2.47. Smialowicz et al. (2008): PFC per 106 Cells
G.2.47.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear
Polynomial, 4 -degree
Power
Hill, unrestricted
Power, unrestricted1"
Degrees of
freedom
3
3
2
2
2
3
3
3
1
2
X2 p-value
0.101
0.101
0.044
0.044
0.063
0.048
0.048
0.048
0.213
0.481
AIC
901.897
901.897
903.897
903.897
903.192
903.585
903.585
903.585
901.219
899.130
BMD (ng/kg)
8.343E+00
8.343E+00
8.325E+00
8.325E+00
3.669E+00
1.373E+01
1.374E+01
1.373E+01
1.928E+00
1.902E-K)0
BMDL
(ng/kg)
5.064E+00
5.064E+00
1.465E+00
1.465E+00
6.970E-01
1.053E+01
1.053E+01
1.053E+01
2.208E-01
2.158E-01
Notes

power hit bound (d=\)

power hit bound (d=\)
n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted (n = 0.35)
unrestricted
(power = 0.333)
a Constant variance model selected (p = O.0001).
b Best-fitting model, BMDS output presented in this appendix.
G.2.47.2. Output for Selected Model: Power, Unrestricted
Smialowicz et al. (2008): PFC per 106 Cells
         Power Model.  (Version:  2.15;   Date:  04/07/2008)
         Input Data File: C:\l\Blood\60_Smial_2008_PFCcells_PwrCV_U_l.(d)
         Gnuplot Plotting File:
C:\l\Blood\60_Smial_2008_PFCcells_PwrCV_U_l.pit
                                            Mon Feb 08 13:29:38  2010
 Anti  Response to SRBCs, PFC per  10to6 cells,  Table 4


   The form of the response function  is:

   Y[dose]  = control + slope * doseApower
   Dependent  variable = Mean
   Independent variable = Dose
   rho  is  set to 0
   The  power  is not restricted
   A  constant variance model is  fit

   Total number of dose groups = 5
   Total number of records with  missing values = 0
   Maximum number of iterations  = 250
   Relative Function Convergence has been set to: le-008
   Parameter  Convergence has been set  to:  le-008
                                      G-277

-------
                  Default Initial Parameter Values
                          alpha =       232385
                            rho =
                        control =
                          slope =
                          power =
            0
         1491
     -491.716
     0.288021
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
control
slope
power


-3
1
-1
alpha
1
.4e-009
.8e-009
.2e-010
control
-3.4e-009
1
-0.82
-0.65
slope
1.8e-009
-0.82
1
0.94
power
-1.2e-010
-0.65
0.94
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha           219793
294222
        control          1470.48
1712.99
          slope         -378.406
-70.6872
          power         0.333124
0.555581
Parameter Estimates



       Std.  Err.

         37974.5

          123.73

         157.002

        0.113501
        95.0% Wald

     Lower Conf.  Limit

              145365

             1227.98

            -686.125

            0.110666
     Table of Data and Estimated Values of Interest
Dose
Res .
-
0
2

0
.438
.464
13.4
15
14
15
15
N Obs Mean Est Mean
1.49e+003
1.13e+003
945
677
1.47e+003
1.18e+003
959
572
Obs Std Dev
716
171
516
465
Est Std Dev
469
469
469
469
Scaled
0.169
-0.431
-0.12
0.867
                                     G-278

-------
31.65
        161
274
117
469
-0.684
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -444.832859
             -425.402825
             -444.832859
             -445.564823
             -463.753685
         # Param's
               6
              10
               6
               4
               2
           AIC
        901.665718
        870.805651
        901.665718
        899.129647
        931.507371
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
-2*log(Likelihood Ratio)  Test df
            76.7017
            38.8601
            38.8601
            1.46393
                    p-value

                   <.0001
                   <.0001
                   <.0001
                    0.481
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1.
non-homogeneous variance model
                              Consider running a
                                     G-279

-------
The p-value for Test 3 is less than .1.  You may want to consider a
different variance model

The p-value for Test 4 is greater than .1.  The model chosen seems
to adequately describe the data
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 1. 90249


            BMDL = 0.215843
                                     G-280

-------
G.2.47.3. Figure for Selected Model: Power, Unrestricted


                                 Power Model with 0.95 Confidence Level
 01
 to
 £=
 O
 CL
 (/)
 0)
 a:
 £=
 TO
          1500
1000
           500
                     Power
              BVIDL
             BMD
                  0
                            10
15

dose
20
25
30
   13:2902/082010
                                           G-281

-------
G.2.48. Smialowicz et al. (2008): PFC per Spleen
G.2.48.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear
Polynomial, 4 -degree
Power
Hill, unrestricted
Power, unrestricted1"
Degrees of
freedom
3
2
3
1
2
3
3
3
1
2
X2 p-value
0.124
0.069
0.124
0.021
0.116
0.126
0.126
0.126
0.103
0.270
AIC
377.565
379.138
377.565
381.138
378.108
377.522
377.522
377.522
378.463
376.420
BMD (ng/kg)
1.334E+01
1.536E+01
1.334E+01
1.536E+01
1.568E+01
2.055E+01
2.055E+01
2.055E+01
1.202E+01
1.187E-K)!
BMDL
(ng/kg)
8.593E+00
8.895E+00
8.593E+00
8.895E+00
error
1.624E+01
1.624E+01
1.624E+01
error
3.762E-K)0
Notes




n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted (n = 0.544)
unrestricted
(power = 0.531)
a Nonconstant variance model selected (p = 0.0011).
b Best-fitting model, BMDS output presented in this appendix.
G.2.48.2. Output for Selected Model: Power, Unrestricted
Smialowicz et al. (2008): PFC per Spleen
         Power Model.  (Version:  2.15;   Date:  04/07/2008)
         Input Data File: C:\l\Blood\61_Smial_2008_PFCspleen_Pwr_U_l.(d)
         Gnuplot Plotting File:
C:\l\Blood\61_Smial_2008_PFCspleen_Pwr_U_l.pit
                                            Mon Feb 08 13:30:16  2010
 Anti  Response to SRBCs - PFC x  10  to the 4 per spleen, Table  4


   The form of the response function  is:

   Y[dose]  = control + slope * doseApower
   Dependent  variable = Mean
   Independent variable = Dose
   The  power  is not restricted
   The  variance is to be modeled  as  Var(i)
= exp(lalpha  +  log(mean(i)
rho)
   Total  number of dose groups =  5
   Total  number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence  has been set to: le-008
   Parameter Convergence has been set  to:  le-008
                                      G-282

-------
                  Default Initial Parameter Values
                         lalpha =      4.76607
                            rho =            0
                        control =         27.8
                          slope =     -9.21898
                          power =     0.286443
           Asymptotic Correlation Matrix of Parameter Estimates
lalpha
lalpha
rho
control
slope
power

-0
0
-0
-0
1
.98
.25
.28
.22
rho
-0.98
1
-0.3
0.28
0.22
control
0.25
-0.3
1
-0.83
-0.74
slope
-0
0
-0

0
.28
.28
.83
1
.99
power
-0
0
-0
0

.22
.22
.74
.99
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha         0.746922
2.74721
            rho          1.36826
2.06567
        control          25.3816
31.1967
          slope          -3.5662
1.38385
          power         0.531216
0.875637
Parameter Estimates



       Std.  Err.

         1.02058

        0.355827

         2.96691

         2.52558

        0.175728
   95.0% Wald

Lower Conf.  Limit

       -1.25337

        0.67085

        19.5666

       -8.51626

       0.186796
     Table of Data and Estimated Values of Interest
Dose
Res .
0
0.438
2.464
13.4
31.65
N
15
14
15
15
8
Obs
27

17
12

Mean
.8
21
.6
.6
3
Est
25
23
19
11
3.
Mean
.4
.1
.6
.2
03
Obs St
13.
13.
9.
8.
3.
d Dev
4
6
4
7
1
Est S
13
12
11
7
3
td Dev
.3
.4
.1
.6
.1
Seal
0
-0
-0
0
-0.
ed
.706
.626
.704
.702
0313
                                     G-283

-------
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -190.565019
             -181.476284
             -181.900030
             -183.210137
             -204.636496
# Param's
      6
     10
      7
      5
      2
   AIC
393.130038
382.952569
377.800059
376.420274
413.272993
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
-2*log(Likelihood Ratio)  Test df
            46.3204
            18.1775
            0.84749
            2.62021
           p-value

          <.0001
        0.001139
          0.8381
          0.2698
The p-value for Test 1 is less than  .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than  .1.
model appears to be appropriate
                              A non-homogeneous variance
The p-value for Test 3 is greater than  .1.
 to be appropriate here
                                 The modeled variance appears
                                     G-284

-------
The p-value  for  Test 4 is greater  than  .1.   The model chosen  seems

to adequately  describe the data
                Benchmark Dose Computation



Specified effect  =             1



Risk Type         =     Estimated  standard deviations from  the control mean



Confidence  level  =          0.95



              BMD  = 11.8748





             BMDL  = 3.76161
G.2.48.3. Figure for Selected Model: Power, Unrestricted



                             Power Model with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 HI
       35
       30
       25
20
15
       10
                Power
                BMDL
   13:3002/082010
                             BMD
                                10
                                   15

                                   dose
20
25
30
                                      G-285

-------
G.2.49. Smith et al. (1976): Cleft Palate in Pups
G.2.49.1. Summary Table of BMDS Modeling Results
Model3
Gamma
Logistic
Log-logistic a
Log-probit
Multistage, 5th degree
Probit
Weibull
Gamma, unrestricted
Log-logistic, unrestricted
Log-probit, unrestricted
Weibull, unrestricted
Degrees of
freedom
3
4
3
3
3
4
3
3
3
3
3
X2 p-value
0.4216
0.5620
0.4218
0.4667
0.4490
0.6133
0.4340
0.4216
0.4218
0.4134
0.4339
AIC
69.75
68.48
69.79
69.96
69.41
67.98
69.64
69.75
69.79
69.89
69.64
BMD
(ng/kg-day)
3.242E+01
4.592E+01
3.525E+01
3.854E+01
2.504E+01
4.096E+01
3.104E+01
3.242E+01
3.525E+01
3.806E+01
3.104E+01
BMDL
(ng/kg-day)
1.123E+01
3.437E+01
1.064E+01
1.903E+01
1.165E+01
3.113E+01
1.136E+01
8.310E+00
1.064E+01
1.086E+01
9.231E+00
Notes











a Best-fitting model, BMDS output presented in this appendix.
G.2.49.2. Output for Selected Model: Log-Logistic
        Logistic Model.  (Version:  2.12; Date: 05/16/2008)
        Input Data  File:
C:\USEPA\BMDS21\la\76_Smith_1976_cleft_palate_b_LogLogistic_l.(d)
        Gnuplot Plotting  File:
C:\USEPA\BMDS21\la\76_Smith_1976_cleft_palate_b_LogLogistic_l.plt
                                           Fri Sep 02  08:12:55  2011
 Table 3 cleft palate
   The form of the probability function is:

   P[response] = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
   Dependent variable  =  DichEff
   Independent variable  =  Dose
   Slope parameter is  restricted as slope >= 1

   Total number of observations = 6
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has  been set to: le-008
   User has chosen the  log  transformed model
                                     G-286

-------
                  Default Initial Parameter Values
                     background =            0
                      intercept =     -4.88569
                          slope =            1
           Asymptotic Correlation Matrix of Parameter Estimates

             background    intercept        slope

background            1        -0.22         0.21

 intercept        -0.22            1        -0.99

     slope         0.21        -0.99            1
Confidence Interval
       Variable
Upper Conf. Limit
     background
*
      intercept
*
          slope
               Parameter Estimates

                                       95.0% Wald

      Estimate        Std.  Err.      Lower Conf. Limit

     0.0259253            *                *

      -10.1275            *                *

       2.22613            *                *
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.2733
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -29.9486         6
     -31.8949         3       3.89258      3
     -52.2767

      69.7899
1
44.6562
                            P-value
<.0001
                                  Goodness  of  Fit

0
0
1
7
50
138
Dose
.0000
.1242
.0125
.1100
.5906
.0663
Est
0.
0.
0.
0.
0.
0.
. Prob.
0259
0259
0260
0290
2197
7067
Exp
0
1
0
0
4
9
ected
.881
.063
.493
.493
.175
.894
Ob
0.
2.
0.
1.
4.
10.
served
000
000
000
000
000
000
Size
34
41
19
17
19
14
Scaled
Residual
-0.951
0.921
-0.712
0.733
-0.097
0.062
                                     G-287

-------
 ChiA2  =2.81
d.f.  =  3
P-value = 0.4218
   Benchmark Dose Computation

Specified effect =             0.1

Risk  Type         =       Extra risk

Confidence level =            0.95

              BMD =         35.2466

             BMDL =         10.6443


G.2.49.3. Figure for Selected Model: Log-Logistic
        0.6
        0.4
        0.2
              BMDL
                          Log-Logistic Model with 0.95 Confidence Level
                        Log-Logistic
                      20
                            BMD
                              40
                                      60      80
                                        dose
                                                     100
                                                             120
                                                                     140
  08:12 09/02 2011
                                        G-288

-------
G.2.50. Sparschu et al. (1976): Fetal Body Weight, Male
G.2.50.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5) b
Hill
Linear
Polynomial, 3 -degree
Power
Hill, unrestricted
Power, unrestricted
Degrees of
freedom
3
3
2
1
1
3
0
3
1
2
x2
p-value
0.0002
0.0002
0.0001
<0.0001
<0001
0.0001
NA
0.0001
<0001
<0001
AIC
-247.04
-247.04
-246.68
-246.18
-246.76
-246.33
-151.65
-246.33
-246.76
-244.93
BMD (ng/kg-
day)
6.844E+01
6.844E+01
6.436E+01
5.736E-K)!
5.421E+01
7.217E+01
6.931E+01
7.217E+01
5.421E+01
7.132E+01
BMDL (ng/kg-
day)
4.399E+01
4.399E+01
3.808E+01
1.685E-K)!
error
4.697E+01
2.162E+01
4.697E+01
error
4.420E+01
Notes










a Modeled variance model presented (p < 0.0001); variance not appropriately captured (/?-test 3 = 0.008).
b Best-fitting model, BMDS output presented in this appendix.
G.2.50.2. Output for Selected Model: exponential (MS)


         Exponential Model.  (Version: 1.61;   Date:  7/24/2009)
         Input Data File:
C:\USEPA\BMDS21\la\74_Sparschu_1971_pup_bw_male_b_Exp_l.(d)
         Gnuplot Plotting  File:
                                            Thu  Sep  01 14:59:46 2011
 Table 4 males
    The form of the response  function by Model:
       Model 2:
       Model 3:
       Model 4:
       Model 5:
Y[dose]
Y[dose]
Y[dose]
Y[dose]
a
a
a
a
exp{sign * b  *  dose}
exp{sign *  (b * dose)Ad}
[c-(c-l) * exp{-b * dose}]
[c-(c-l) * exp{-(b *  dose)Ad}]
     Note:  Y[dose] is the  median response  for  exposure = dose;
           sign = +1 for increasing trend  in data;
           sign = -1 for decreasing trend.

       Model 2 is nested within Models 3 and 4.
       Model 3 is nested within Model 5.
       Model 4 is nested within Model 5.
    Dependent variable = Mean
    Independent variable =  Dose
    Data are assumed to be  distributed: normally
    Variance Model: exp(lnalpha +rho *ln(Y[dose]))
    The variance is to be modeled as Var(i) =  exp(lalpha + log(mean(i))

    Total number of dose groups = 5
                                                         rho)
                                       G-289

-------
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 5

                 Inalpha             -4.28192
                     rho              1.66816
                       a                4.347
                       b            0.0041752
                       c             0.312859
                       d                    1
                  Parameter Estimates

                Variable          Model 5

                 Inalpha           16.8213
                     rho          -13.5946
                       a           4.04383
                       b         0.0163183
                       c           0.86046
                       d           1.40496


         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
0
5.09
16.28
52.87
188.3
117
55
66
39
3
4.03
4.14
3.85
3.86
2.72
0.37
0.26
0.35
0.61
0.25
               Estimated Values of Interest

   Dose      Est Mean      Est Std     Scaled Residual
0
5.09
16.28
52.87
188.3
4.044
4.027
3.963
3.73
3.484
0.3374
0.3471
0.3873
0.5844
0.929
-0.4433
2.415
-2.363
1.39
-1.424
Other models for which likelihoods are calculated:

                                  G-290

-------
     Model Al:        Yij
               Var{e(ij) }

     Model A2:        Yij
               Var{e(ij) }

     Model A3:        Yij
               Var{e(ij) }

     Model  R:        Yij
               Var{e(ij) }
                         Mu(i) + e (ij)
                         SigmaA2

                         Mu(i) + e (ij)
                         Sigma(i)A2

                         Mu(i) + e (i j)
                         exp(lalpha + log(mean(i)) * rho)

                         Mu + e(i)
                         SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
5
126.4055
145.7666
137.4206
101.5293
129.0908
6
10
7
2
6
-240.8109
-271.5331
-260.8413
-199.0587
-246.1816
   Additive constant for all log-likelihoods =     -257.3.  This  constant
added to the
   above values gives the log-likelihood including the term that  does not
   depend on the model parameters.
R)
                              Explanation of Tests

Test 1:  Does response and/or variances differ among Dose levels?  (A2 vs.

Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)

Test 7a: Does Model 5 fit the data?  (A3 vs  5)
     Test

     Test 1
     Test 2
     Test 3
    Test 7a
                         Tests of Interest

                -2*log(Likelihood Ratio)
                                                  D. F.
p-value
88.47
38.72
16.69
16.66
8
4
3
1
< 0.0001
< 0.0001
0.0008177
< 0.0001
     The p-value for Test 1 is less than  .05.  There appears to be a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

     The p-value for Test 2 is less than  .1.  A non-homogeneous
     variance model appears to be appropriate.

                                     G-291

-------
     The p-value  for Test 3 is less  than  .1.   You may want to

     consider  a different variance model.



     The p-value  for Test 7a is less  than  .1.   Model 5 may not  adequately

     describe  the data;  you may want  to consider another model.
   Benchmark  Dose  Computations:



     Specified  Effect = 1.000000



            Risk Type = Estimated standard deviations from control



     Confidence Level = 0.950000



                   BMD =      57.3555



                 BMDL =      16.8535
G.2.50.3. Figure for Selected Model: Exponential (MS)



                         Exponential_beta Model 5 with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 HI
         3.5
        2.5
                BMDL
                        Exponential
   14:5909/01 2011
BMD
                              50
            100

          dose
150
                                      G-292

-------
 G.2.51. Sparschu et al. (1971): Fetal Body Weight, Female
 G.2.51.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2) b
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Power
Hill, unrestricted
Power, unrestricted
Degrees of
freedom
3
2
2
1
1
3
3
2
1
2
X2 p-value
0.0340
0.0025
0.0146
0.0037
0.0037
0.0315
0.0315
0.0025
0.0037
0.0136
AIC
-229.963
-224.657
-228.182
-226.196
-226.226
-229.794
-229.794
-224.657
-226.226
-228.035
BMD (ng/kg-
day)
.027E+02
.713E+02
.044E+02
.037E+02
.044E+02
.035E+02
.035E+02
.746E+02
.044E+02
.054E+02
BMDL (ng/kg-
day)
6.523E+01
5.467E+01
6.131E+01
6.028E+01
6.055E+01
6.725E+01
6.725E+01
5.742E+01
6.055E+01
6.491E+01
Notes










a Modeled variance model presented (p = 0.001)); variance not appropriately captured (/?-test 3 = 0.005).
b Best-fitting model, BMDS output presented in this appendix.
 G.2.51.2. Output for Selected Model: Exponential (M2)


         Exponential Model.  (Version: 1.61;   Date:  7/24/2009)
         Input Data File:
 C:\USEPA\BMDS21\la\75_Sparschu_1971_pup_bw_fm_b_Exp_l.(d)
         Gnuplot Plotting  File:
                                            Thu  Sep  01 15:03:28 2011
  Table 4 females
    The form of the response function by Model:
       Model 2:
       Model 3:
       Model 4:
       Model 5:
Y[dose]
Y[dose]
Y[dose]
Y[dose]
a
a
a
a
expfsign * b  *  dose}
exp{sign *  (b * dose)Ad}
[c-(c-l) * exp{-b * dose}]
[c-(c-l) * exp{-(b *  dose)Ad}]
     Note:  Y[dose] is  the  median response  for  exposure = dose;
           sign = +1 for increasing trend  in data;
           sign = -1 for decreasing trend.

       Model 2 is nested within Models 3 and 4.
       Model 3 is nested within Model 5.
       Model 4 is nested within Model 5.
    Dependent variable  =  Mean
    Independent variable  =  Dose
    Data are assumed to be  distributed: normally
    Variance Model: exp(lnalpha +rho *ln(Y[dose]))
    The variance is to  be modeled as Var(i) =  exp(lalpha + log(mean(i))  * rho)

    Total number of dose  groups = 5
                                       G-293

-------
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values
               Variable

                 Inalpha
                     rho
                       a
                       b
                       c
                       d
  Model 2

      -7.22746
       4.02075
       3.74918
    0.00140938
             0
             1
                  Parameter Estimates

                Variable          Model 2
                 Inalpha
                     rho
                       a
                       b
                       c
                       d
    11.1109
   -9.58142
    3.90142
0.000999148
          0
          1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
0
5.09
16.28
52.87
188.3
129
60
58
54
4
3.89
3.98
3.71
3.78
2.69
0.39
0.35
0.37
0.54
0.19
   Dose
               Estimated Values of Interest

             Est Mean      Est Std     Scaled Residual
0
5.09
16.28
52.87
188.3
3.901
3.882
3.838
3.701
3.232
0.3805
0.3899
0.4113
0.49
0.9369
-0.3408
1.955
-2.379
1.189
-1.158
Other models for which likelihoods are calculated:

                                  G-294

-------
     Model Al:        Yij
               Var{e(ij) }

     Model A2:        Yij
               Var{e(ij) }

     Model A3:        Yij
               Var{e(ij) }

     Model  R:        Yij
               Var{e(ij) }
                         Mu(i) + e (ij)
                         SigmaA2

                         Mu(i) + e (ij)
                         Sigma(i)A2

                         Mu(i) + e (i j)
                         exp(lalpha + log(mean(i)) * rho)

                         Mu + e(i)
                         SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
2
123.0729
132.131
123.3163
100.5646
118.9813
6
10
7
2
4
-234.1458
-244.262
-232.6326
-197.1292
-229.9626
   Additive constant for all log-likelihoods =     -280.3.  This  constant
added to the
   above values gives the log-likelihood including the term that  does not
   depend on the model parameters.
R)
Test 1:

Test 2:
Test 3:
Test 4:
                     Explanation of Tests

Does response and/or variances differ among Dose levels?  (A2  vs.

Are Variances Homogeneous?  (A2 vs. Al)
Are variances adequately modeled?  (A2 vs. A3)
Does Model 2 fit the data?  (A3 vs. 2)
     Test

     Test 1
     Test 2
     Test 3
     Test 4
                         Tests of Interest

                -2*log(Likelihood Ratio)
                                                  D. F.
                                                    p-value
63.13
18.12
17.63
8.67
8
4
3
3
< 0.0001
0.001171
0.0005244
0.03402
     The p-value for Test 1 is less than  .05.  There appears to be a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

     The p-value for Test 2 is less than  .1.  A non-homogeneous
     variance model appears to be appropriate.
                                     G-295

-------
     The p-value  for Test 3 is less  than .1.   You may want to

     consider  a different variance model.



     The p-value  for Test 4 is less  than .1.   Model 2 may not  adequately

     describe  the data;  you may want  to  consider another model.
   Benchmark  Dose  Computations:



     Specified  Effect = 1.000000



            Risk Type = Estimated standard deviations from  control



     Confidence Level = 0.950000



                   BMD =      102.699



                 BMDL =      65.2254
G.2.51.3. Figure for Selected Model: Exponential (M2)


                         Exponential_beta Model 2 with 0.95 Confidence Level
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 OJ
         3.5
        2.5
   15:0309/01 2011
                        Exponential
                               BMDL
    BMD
                              50
  100

dose
150
                                      G-296

-------
G.2.52. Toth et al. (1979): Amyloidosis
G.2.52.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage, 3 -degree
Probit
Weibull
Gamma, unrestricted
Log-logistic,
unrestricted13
Log-probit, unrestricted
Weibull, unrestricted
Degrees of
freedom
2
2
2
2
2
2
2
2
2
2
2
X2 p-value
0.040
0.019
0.053
0.009
0.040
0.021
0.040
0.959
0.903
0.870
0.933
AIC
149.120
151.340
148.269
152.855
149.120
151.115
149.120
140.119
140.240
140.315
140.174
BMD (ng/kg)
1.965E+01
3.701E+01
1.503E+01
3.782E+01
1.965E+01
3.467E+01
1.965E+01
4.349E-01
4.843E-01
4.960E-01
4.641E-01
BMDL
(ng/kg)
1.283E+01
2.858E+01
8.747E+00
2.502E+01
1.283E+01
2.657E+01
1.283E+01
2.891E-03
5.312E-03
7.292E-03
4.069E-03
Notes
power bound hit
(power =1)

slope bound hit
(slope = 1)
slope bound hit (slope =
1)
final B = 0

power bound hit
(power =1)
unrestricted
(power = 0.254)
unrestricted
(slope = 0.326)
unrestricted
(slope = 0.186)
unrestricted
(power = 0.289)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.


G.2.52.2. Output for Selected Model: Log-Logistic
Toth et al. (1979): Amyloidosis
         Logistic Model.  (Version: 2.12;  Date:  05/16/2008)
         Input Data File:  C:\l\Blood\62_Toth_1979_Amylyr_LogLogistic_l.(d)
         Gnuplot Plotting  File:
C:\l\Blood\62_Toth_1979_Amylyr_LogLogistic_l.pit
                                             Mon Feb 08  13:30:54 2010
 Table  2
   The  form of the probability function  is:

   P[response] = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
    Dependent variable  =  DichEff
    Independent variable  = Dose
    Slope parameter is  restricted as slope  >= 1
                                       G-297

-------
   Total number of observations = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =            0
                      intercept =     -4.54593
                          slope =            1
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -slope
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

             background    intercept

background            1        -0.49

 intercept        -0.49            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background        0.0699918
*
      intercept         -4.90704
*
          slope                1
                                 Parameter Estimates
                      Std.  Err.
                     95.0% Wald

                  Lower  Conf.  Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.01628
  Reduced model
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
      -68.017         4
     -72.1346         2       8.23525      2
     -82.0119
    1

G-298
27.99
3
                                P-value
<.0001

-------
           AIC:         148.269
                                  Goodness  of  Fit

Dose
0.0000
0.5732
14.2123
91.2070

Est. Prob.
0.0700
0.0739
0.1584
0.4446

Expected
2.660
3.252
6.971
19.117

Observed
0.000
5.000
10.000
17.000

Size
38
44
44
43
Scaled
Residual
-1.691
1.007
1.251
-0.650
 ChiA2 = 5.86      d.f. = 2        P-value = 0.0534







   Benchmark Dose Computation




Specified effect =            0.1




Risk Type        =      Extra risk




Confidence level =           0.95




             BMD =        15.0264




            BMDL =        8.74665
                                     G-299

-------
G.2.52.3. Figure for Selected Model: Log-Logistic

                            Log-Logistic Model with 0.95 Confidence Level
 O

 I
 O
 13
 (0
         0.6
         0.5
         0.4
0.3
         0.2
         0.1
                          Log-Logistic
                 BMDL
                 BMD
                            20
                                40
60
80
                                            dose
   13:3002/082010
G.2.52.4. Output for Additional Model Presented: Log-Logistic, Unrestricted
Toth et al. (1979): Amyloidosis
         Logistic Model.  (Version:  2.12; Date:  05/16/2008)
         Input Data File:  C:\l\Blood\62_Toth_1979_Amylyr_LogLogistic_U_l.(d)
         Gnuplot Plotting  File:
C:\l\Blood\62_Toth_1979_Amylyr_LogLogistic_U_l.pit
                                            Mon  Feb 08 13:30:54  2010
 Table 2
   The  form of the probability function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
                                       G-300

-------
   Dependent variable = DichEff
   Independent variable = Dose
   Slope parameter is not restricted

   Total number of observations = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =            0
                      intercept =     -1.92722
                          slope =     0.314472
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -background
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

              intercept        slope

 intercept            1        -0.84

     slope        -0.84            1
                                 Parameter Estimates

                                                         95.0% Wald
Confidence Interval
       Variable         Estimate        Std. Err.     Lower Conf. Limit
Upper Conf. Limit
     background                0            *                *
*
      intercept         -1.96073            *                *
*
          slope         0.326156            *                *
*

* - Indicates that this value is not calculated.
                        Analysis of Deviance Table

       Model      Log(likelihood)   # Param's  Deviance  Test d.f.   P-value

                                    G-301

-------
     Full model
   Fitted model
0.902
  Reduced model

           AIC:
-68.017
68.1201
82.0119
140.24
4
2
1

                            0.206341

                               27.99
                                  Goodness  of  Fit
                          2

                          3
<.0001

Dose
0.0000
0.5732
14.2123
91.2070

Est. Prob.
0.0000
0.1051
0.2507
0.3802

Expected
0.000
4.623
11.029
16.348

Observed
0.000
5.000
10.000
17.000

Size
38
44
44
43
Scaled
Residual
0.000
0.186
-0.358
0.205
 ChiA2 = 0.20
d.f.  =2
P-value = 0.9028
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =       0.484272

            BMDL =     0.00531211
                                     G-302

-------
G.2.52.5. Figure for Additional Model Presented: Log-Logistic,  Unrestricted



                                Log-Logistic Model with 0.95 Confidence Level
 T3

 £
 O
 c
 O
 •*=
 O
 (0
          0.6
          0.5
          0.4
          0.3
 il        0.2
          0.1
             BljflDL||BMD
                             Log-Logistic
                                20
40             60

   dose
80
   13:3002/082010
                                            G-303

-------
G.2.53. Toth et al. (1979): Skin Lesions
G.2.53.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage, 3 -degree
Probit
Weibull
Gamma, unrestricted
Log-logistic,
unrestricted13
Log-probit, unrestricted
Weibull, unrestricted
Degrees of
freedom
2
2
2
2
2
2
2
2
2
2
2
X2 p-value
0.032
0.005
0.078
0.003
0.032
0.006
0.032
0.945
0.744
0.670
0.866
AIC
156.346
161.421
153.963
161.788
156.346
160.991
156.346
147.148
147.631
147.844
147.324
BMD (ng/kg)
1.037E+01
2.487E+01
6.413E+00
1.887E+01
1.037E+01
2.309E+01
1.037E+01
error
5.969E-01
5.939E-01
5.539E-01
BMDL
(ng/kg)
7.470E+00
1.982E+01
4.025E-K)0
1.280E+01
7.470E+00
1.858E+01
7.470E+00
error
6.773E-02
8.147E-02
5.181E-02
Notes
power bound hit
(power =1)

slope bound hit
(slope = 1)
slope bound hit
(slope = 1)
final B = 0

power bound hit
(power =1)
unrestricted
(power =0.341)
unrestricted
(slope = 0.48)
unrestricted
(slope = 0.279)
unrestricted
(power = 0.405)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.


G.2.53.2. Output for Selected Model: Log-Logistic
Toth et al. (1979): Skin Lesions
         Logistic Model.  (Version: 2.12;  Date:  05/16/2008)
         Input Data File:  C:\l\Blood\63_Toth_1979_SkinLes_LogLogistic_l.(d)
         Gnuplot Plotting  File:
C:\l\Blood\63_Toth_1979_SkinLes_LogLogistic_l.plt
                                             Wed Feb 10 14:47:53 2010
 Table  2
   The  form of the probability function  is:

   P[response] = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
    Dependent variable  =  DichEff
    Independent variable  = Dose
    Slope parameter is  restricted as slope  >= 1
                                       G-304

-------
   Total number of observations = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =            0
                      intercept =     -3.94312
                          slope =            1
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -slope
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

             background    intercept

background            1        -0.43

 intercept        -0.43            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background        0.0564562
*
      intercept         -4.05558
*
          slope                1
                                 Parameter Estimates
                      Std.  Err.
                 95.0% Wald

              Lower Conf.  Limit
* - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.03132
  Reduced model
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -71.5177         4
     -74.9813         2       6.92722      2
     -95.8498
1
48.6642
3
                            P-value
<.0001
                                     G-305

-------
           AIC:         153.963
                                  Goodness  of  Fit

Dose
0.0000
0.5732
14.2123
91.2070

Est. Prob.
0.0565
0.0657
0.2429
0.6343

Expected
2.145
2.892
10.687
27.275

Observed
0.000
5.000
13.000
25.000

Size
38
44
44
43
Scaled
Residual
-1.508
1.282
0.813
-0.720
 ChiA2 = 5.10      d.f. = 2        P-value = 0.0782







   Benchmark Dose Computation




Specified effect =            0.1




Risk Type        =      Extra risk




Confidence level =           0.95




             BMD =         6.4132




            BMDL =         4.0249
                                     G-306

-------
G.2.53.3. Figure for Selected Model: Log-Logistic

                         Log-Logistic Model with 0.95 Confidence Level
        0.7
        0.6
        0.5
 I
 C
 g
 13
 ro
                       Log-Logistic
        0.4


        0.3


        0.2


        0.1


         0
           jBMDL  BMP
              0           20          40          60          80
                                       dose
  14:4702/102010



G.2.53.4. Output for Additional Model Presented: Log-Logistic, Unrestricted

Toth et al. (1979): Skin Lesions



        Logistic Model.  (Version:  2.12; Date:  05/16/2008)
        Input Data File: C:\l\Blood\63_Toth_1979_SkinLes_LogLogistic_U_l.(d)
        Gnuplot Plotting File:
C:\l\Blood\63_Toth_1979_SkinLes_LogLogistic_U_l.plt
                                            Wed Feb 10 14:47:54  2010


 Table  2
   The form of the probability function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
   Dependent  variable = DichEff
   Independent variable =  Dose
   Slope parameter is not  restricted
                                       G-307

-------
   Total number of observations = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =            0
                      intercept =     -1.87608
                          slope =     0.458888
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -background
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
              intercept

 intercept            1

     slope        -0.86
             slope

             -0.86

                 1
                                 Parameter Estimates
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background                0
*
      intercept         -1.94946
*
          slope           0.4802
                      Std.  Err.
                     95.0% Wald

                  Lower  Conf.  Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.7426
  Reduced model
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -71.5177         4
     -71.8153         2       0.59526      2
     -95.8498
    1

G-308
48.6642
3
                                P-value
<.0001

-------
           AIC:         147.631
                                  Goodness  of  Fit

Dose
0.0000
0.5732
14.2123
91.2070

Est. Prob.
0.0000
0.0983
0.3374
0.5542

Expected
0.000
4.323
14.845
23.832

Observed
0.000
5.000
13.000
25.000

Size
38
44
44
43
Scaled
Residual
0.000
0.343
-0.588
0.358
 ChiA2 = 0.59      d.f. = 2        P-value = 0.7438







   Benchmark Dose Computation




Specified effect =            0.1




Risk Type        =      Extra risk




Confidence level =           0.95




             BMD =       0.596932




            BMDL =        0.06773
                                     G-309

-------
G.2.53.5. Figure for Additional Model Presented: Log-Logistic, Unrestricted



                                Log-Logistic Model with 0.95 Confidence Level
 T3

 £
 O
 c
 O
 •*=
 O
 (0
0.7




0.6




0.5




0.4




0.3




0.2




0.1
             BMDLBMD
                             Log-Logistic
                                20
                                     40
60
80
                                                 dose
   14:4702/102010
                                            G-310

-------
G.2.54. van Birgelen et al. (1995): Hepatic Retinol
G.2.54.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Hill
Linear
Polynomial, 5 -degree
Power
Hill, unrestricted
Power, unrestricted0
Degrees of
freedom
4
4
3
3
3
4
4
4
2
3
X2 p-value
0.0001
0.0001
<0.001
0.001
0.239
0.0001
0.0001
0.0001
0.241
0.011
AIC
159.735
3,222.700
141.454
141.454
124.865
176.828
176.828
176.828
125.495
131.771
BMD (ng/kg)
7.790E+00
5.542E+01
2.488E-K)!
2.488E+01
5.316E+00
1.877E+02
1.877E+02
1.877E+02
3.595E+00
3.802E-01
BMDL
(ng/kg)
4.150E+00
error
3.363E+00
3.363E+00
error
1.437E+02
1.437E+02
1.437E+02
error
1.393E-02
Notes

power hit bound (d=\)

power hit bound (d=\)
n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted (n = 0.763)
unrestricted
(power = 0.14)
a Nonconstant variance model selected (p = O.0001).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.2.54.2. Output for Selected Model: Exponential (M4)
van Birgelen et al. (1995): Hepatic Retinol
         Exponential Model.  (Version:  1.61;   Date: 7/24/2009)
         Input Data File:  C:\l\Blood\65_VanB_1995a_HepRet_Exp_l.(d)
         Gnuplot Plotting  File:
                                             Mon Feb 08  13:32:00 2010
 Tbl3,  hepatic retinol
   The  form of the response function by  Model:
       Model 2:
       Model 3:
       Model 4:
       Model 5:
Y[dose]
Y[dose]
Y[dose]
Y[dose]
a
a
a
a
expfsign  *  b  *  dose}
exp{sign  *  (b * dose)Ad}
[c-(c-l)  *  exp{-b * dose}]
[c-(c-l)  *  exp{-(b * dose)Ad}]
    Note:  Y[dose] is  the  median response  for exposure =  dose;
           sign = +1 for increasing trend  in data;
           sign = -1 for decreasing trend.

       Model 2 is nested within Models  3 and 4.
       Model 3 is nested within Model 5.
       Model 4 is nested within Model 5.
    Dependent variable  =  Mean
    Independent variable  = Dose
                                       G-311

-------
Data are assumed to be distributed: normally
Variance Model: exp(lnalpha +rho *ln(Y[dose]))
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  * rho)

Total number of dose groups = 6
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to:  le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 4

                 Inalpha             -1.16065
                     rho              1.53688
                       a               15.645
                       b            0.0254351
                       c            0.0365247
                       d                    1
                  Parameter Estimates

                Variable          Model 4

                 Inalpha            -0.92683
                     rho             1.77262
                       a             11.5049
                       b           0.0286598
                       c           0.0653043
                       d                   1


         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
  7.204
  11.76
  18.09
  86.41
  250.2
               Estimated Values of Interest

   Dose      Est Mean      Est Std     Scaled Residual
14
8
8
5
2
0
.9
.4
.2
.1
.2
.6
8
3
2
0.
0.
0.
.768
.394
.263
8485
8485
5657
      0          11.5        5.483            1.751
  7.204         9.499        4.627          -0.6719
  11.76         8.428        4.161          -0.1552

                                  G-312

-------
     18.09         7.154        3.599           -1.615
     86.41         1.655       0.9832            1.568
     250.2        0.7596       0.4931          -0.9155
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) *  rho)

     Model  R:        Yij = Mu + e(i)
               Var{e (ij)  } = SigmaA2
                                Likelihoods of Interest

                     Model      Log(likelihood)      DF         AIC
Al
A2
A3
R
4
-87.1567
-47.28742
-55.32422
-109.967
-65.72714
7
12
8
2
5
188.3134
118.5748
126.6484
223.934
141.4543
   Additive constant for all log-likelihoods =     -44.11.  This  constant
added to the
   above values gives the log-likelihood including the term that  does  not
   depend on the model parameters.
                                 Explanation of Tests

   Test 1:  Does response and/or variances differ among Dose  levels?  (A2  vs.
R)
   Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
   Test 3:  Are variances adequately modeled?  (A2 vs. A3)

   Test 6a: Does Model 4 fit the data?  (A3 vs  4)


                            Tests of Interest

     Test          -2*log(Likelihood Ratio)       D. F.         p-value

     Test 1
     Test 2
     Test 3
    Test 6a


                                     G-313
125.4
79.74
16.07
20.81
10
5
4
3
< 0.0001
< 0.0001
0.002922
0.0001155

-------
  The p-value for Test 1 is less than .05.  There appears to be a
  difference between response and/or variances among the dose
  levels,  it seems appropriate to model the data.

  The p-value for Test 2 is less than .1.   A non-homogeneous
  variance model appears to be appropriate.

  The p-value for Test 3 is less than .1.   You may want to
  consider a different variance model.

  The p-value for Test 6a is less than .1.  Model 4 may not adequately
  describe the data; you may want to consider another model.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =      24.8811

              BMDL =      3.36281
                                 G-314

-------
G.2.54.3. Figure for Selected Model: Exponential (M4)

                          Exponential Model 4 with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 OJ
       20
        15
10
                       Exponential
          EMDL
                         50
                              100          150
                                   dose
200
250
   13:3202/082010
G.2.54.4. Output for Additional Model Presented: Power, Unrestricted
van Birgelen et al. (1995): Hepatic Retinol
         Power Model.  (Version:  2.15;   Date: 04/07/2008)
         Input Data File: C:\l\Blood\65_VanB_1995a_HepRet_Pwr_U_l.(d)
         Gnuplot Plotting File:   C:\l\Blood\65_VanB_1995a_HepRet_Pwr_U_l.plt
                                            Mon Feb  08  13:32:03 2010
 Tbl3, hepatic retinol


   The form of the response  function is:

   Y[dose]  = control + slope  *  doseApower
   Dependent  variable = Mean
   Independent variable = Dose
   The power  is not restricted
                                      G-315

-------
   The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  * rho)

   Total number of dose groups = 6
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =      2.76506
                            rho =            0
                        control =         14.9
                          slope =     -3.98831
                          power =     0.231232
           Asymptotic Correlation Matrix of Parameter Estimates

lalpha
rho
control
slope
power
lalpha
1
-0.8
-0.042
0.038
0.063
rho
-0.8
1
-0.089
0.0044
-0.1
control
-0.042
-0.089
1
-0.95
-0.81
slope
0.038
0.0044
-0.95
1
0.95
power
0.063
-0.1
-0.81
0.95
1
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
-0.212609
            rho
2.07625
        control
21.302
          slope
-3.50543
          power
0.192707
 Estimate

-0.986251

  1.67858

  16.9266

 -7.51118

 0.139871
Parameter Estimates



       Std.  Err.

        0.394722

        0.202896

         2.23237

         2.04379

       0.0269576
   95.0% Wald

Lower Conf.  Limit

       -1.75989

        1.28091

        12.5513

       -11.5169

      0.0870351
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
                                     G-316

-------
    0
7.204
11.76
18.09
86.41
250.2
       14.9
        8.4
        8.2
        5.1
        2.2
        0.6
 16.9
 7.03
 6.32
 5.67
 2.91
0.666
 8.77
 3.39
 2.26
0.849
0.849
0.566
 6.56
 3.14
 2.87
 2.62
  1.5
0.434
-0.874
  1.24
  1.85
-0.611
 -1.34
-0.427
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -87.156698
              -47.287416
              -55.324218
              -60.885746
             -109.967018
           # Param's
                 7
                12
                 8
                 5
                 2
             AIC
          188.313395
          118.574833
          126.648436
          131.771493
          223.934036
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
-2*log(Likelihood Ratio)  Test df
            125.359
            79.7386
            16.0736
            11.1231
         10
          5
          4
          3
         p-value

        <.0001
        <.0001
      0.002922
       0.01108
                                     G-317

-------
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is less than .1.  You may want to consider a
different variance model

The p-value for Test 4 is less than .1.  You may want to try a different
model
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 0.380208


            BMDL = 0.013927
                                     G-318

-------
G.2.54.5. Figure for Additional Model Presented: Power, Unrestricted
 (D
 W
 o
 Q.
 ro
 cu
                             Power Model with 0.95 Confidence Level
                          50
100         150

     dose
200
250
   13:3202/082010
                                            G-319

-------
G.2.55. van Birgelen et al. (1995): Hepatic Retinol Palmitate
G.2.55.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Hill
Linear
Polynomial, 5 -degree
Power
Hill, unrestricted
Power, unrestricted0
Degrees of
freedom
4
4
3
3
3
4
0
4
3
3
X2 p-value
0.0001
0.0001
<0.0001
0.0001
0.009
0.0001
N/A
0.0001
0.0001
0.239
AIC
460.282
460.282
446.995
446.995
416.233
486.375
584.170
486.375
527.310
408.982
BMD (ng/kg)
error
error
1.415E-KJ2
1.415E+02
3.657E+00
3.487E+02
error
3.487E+02
6.875E-14
5.262E-02
BMDL
(ng/kg)
error
error
3.647E+01
3.647E+01
error
2.412E+02
5.617E+02
2.412E+02
6.875E-14
5.889E-05
Notes

power hit bound (d=l)

power hit bound (d=l)
n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted (n = 0.613)
unrestricted
(power = 0.064)
a Nonconstant variance model selected (p = O.0001).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.2.55.2. Output for Selected Model: Exponential (M4)
van Birgelen et al. (1995): Hepatic Retinol Palmitate
         Exponential Model.  (Version:  1.61;   Date: 7/24/2009)
         Input Data File:  C:\l\Blood\66_VanB_1995a_HepRetPalm_Exp_l.(d)
         Gnuplot Plotting  File:
                                             Mon Feb 08  13:32:41 2010
 Tbl3,  hepatic retinol  palmitate
   The  form of the response function by  Model:
      Model 2:     Y[dose]  = a * exp{sign * b * dose}
      Model 3:     Y[dose]  = a * exp{sign * (b * dose)Ad}
      Model 4:     Y[dose]  = a * [c-(c-l)  * exp{-b * dose}]
      Model 5:     Y[dose]  = a * [c-(c-l)  * exp{-(b *  dose)Ad}]

    Note:  Y[dose] is  the  median response for exposure  =  dose;
           sign = +1 for increasing trend in data;
           sign = -1 for decreasing trend.

      Model 2 is nested within Models  3  and 4.
      Model 3 is nested within Model 5.
      Model 4 is nested within Model 5.
    Dependent variable  =  Mean
    Independent variable  = Dose
                                       G-320

-------
Data are assumed to be distributed: normally
Variance Model: exp(lnalpha +rho *ln(Y[dose]))
The variance is to be modeled as Var(i)  = exp(lalpha + log(mean(i))  * rho)

Total number of dose groups = 6
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to:  le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 4

                 Inalpha             0.284674
                     rho              1.77158
                       a                495.6
                       b            0.0337826
                       c           0.00576502
                       d                    1
                  Parameter Estimates

                Variable          Model 4

                 Inalpha           -0.241601
                     rho             2.03456
                       a             223.848
                       b           0.0300737
                       c           0.0129253
                       d                   1

  NC = No Convergence


         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
472
94
107
74
22
3
271.5
67.88
76.37
39.6
22.63
2.828
      0
  7.204
  11.76
  18.09
  86.41
  250.2
               Estimated Values of Interest

   Dose      Est Mean      Est Std     Scaled Residual

      0         223.8        217.8            3.222

                                  G-321

-------
7.204
11.76
18.09
86.41
250.2
180.8
158
131.1
19.33
3.013
175.3
152.9
126.4
18.03
2.721
-1.401
-0.9443
-1.278
0.4197
-0.01317
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i))  *  rho)

     Model  R:        Yij = Mu + e(i)
               Var{e (ij)  } = SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
4
-250.5548
-196.7557
-197.3832
-276.7896
-218.4977
7
12
8
2
5
515.1096
417.5115
410.7663
557.5793
446.9954
   Additive constant for all log-likelihoods =     -44.11.   This  constant
added to the
   above values gives the log-likelihood including the  term  that  does  not
   depend on the model parameters.
R)
                              Explanation of Tests

Test 1:  Does response and/or variances differ among  Dose  levels?  (A2  vs.

Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)

Test 6a: Does Model 4 fit the data?  (A3 vs  4)
     Test

     Test 1
     Test 2
     Test 3
                         Tests of Interest

                 -2*log(Likelihood Ratio)

                                 160.1
                                 107.6
                                 1.255

                                  G-322
D.  F.

 10
  5
  4
p-value
 < 0.0001
 < 0.0001
    0.869

-------
 Test 6a                         42.23           3            < 0.0001
  The p-value for Test 1 is less than .05.  There appears to be a
  difference between response and/or variances among the dose
  levels,  it seems appropriate to model the data.

  The p-value for Test 2 is less than .1.   A non-homogeneous
  variance model appears to be appropriate.

  The p-value for Test 3 is greater than  .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is less than .1.  Model 4 may not adequately
  describe the data; you may want to consider another model.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =      141.528

              BMDL =      36.4721
                                 G-323

-------
G.2.55.3. Figure for Selected Model: Exponential (M4)



                           Exponential Model 4 with 0.95 Confidence Level
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 0)
        700
        600
        500
400
300
        200
        100
                        Exponential
                                                 BMD
                           50
                              100         150

                                   dose
200
250
   13:3202/082010
G.2.55.4. Output for Additional Model Presented: Power, Unrestricted


van Birgelen et al. (1995): Hepatic Retinol Palmitate
         Power Model.  (Version:  2.15;  Date:  04/07/2008)

         Input Data File:  C:\l\Blood\66_VanB_1995a_HepRetPalm_Pwr_U_l.(d)

         Gnuplot Plotting  File:

C:\l\Blood\66_VanB_1995a_HepRetPalm_Pwr_U_l.pit

                                            Mon  Feb 08 13:32:47  2010
 Tbl3, hepatic retinol palmitate





   The form of the response  function is:



   Y[dose]  = control + slope *  doseApower
                                       G-324

-------
   Dependent variable = Mean
   Independent variable = Dose
   The power is not restricted
   The variance is to be modeled as Var(i)
          = exp(lalpha + log(mean(i))  * rho)
   Total number of dose groups = 6
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =      9.57332
                            rho =            0
                        control =          472
                          slope =     -320.514
                          power =    0.0711173
           Asymptotic Correlation Matrix of Parameter Estimates

lalpha
rho
control
slope
power
lalpha
1
-0.95
0.3
-0.31
-0.3
rho
-0.95
1
-0.41
0.39
0.29
control
0.3
-0.41
1
-0.98
-0.82
slope
-0.31
0.39
-0.98
1
0.9
power
-0.3
0.29
-0.82
0.9
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha        0.0640168
1.74855
            rho          1.81132
2.19835
        control           464.29
635.925
          slope         -324.216
-160.887
          power        0.0639088
0.0913048
Parameter Estimates



       Std.  Err.

        0.859472

        0.197468

         87.5705

         83.3327

       0.0139778
   95.0% Wald

Lower Conf.  Limit

       -1.62052

        1.42429

        292.655

       -487.545

      0.0365129
     Table of Data and Estimated Values of Interest

                                     G-325

-------
 Dose
Res .
                 Obs Mean
                              Est Mean
                      Obs Std Dev  Est Std Dev
                                     Scaled
    0
7.204
11.76
18.09
86.41
250.2
472
 94
107
 74
 22
  3
 464
96.5
84.8
74.2
33.2
2.86
 272
67.9
76.4
39.6
22.6
2.83
 269
64.7
57.6
  51
24.6
2.68
  0.0812
  -0.108
    1.09
-0.00941
   -1.28
   0.145
 Model Descriptions for likelihoods calculated
 Model Al:        Yij
           Var{e(ij) }

 Model A2:        Yij
           Var{e(ij) }
     Mu(i)  + e(i j '
     SigmaA2

     Mu(i)  + e(i j '
     Sigma(i)A2
 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model      Log(likelihood)
             Al         -250.554817
             A2         -196.755746
             A3         -197.383174
         fitted         -199.490808
              R         -276.789644
# Param' s
7
12
8
5
2
AIC
515.109634
417.511491
410.766347
408.981615
557.579287
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.
                     Tests of Interest

   Test    -2*log(Likelihood Ratio)  Test df

   Test 1              160.068         10

                                     G-326
                                 p-value

                                <.0001

-------
   Test 2
   Test 3
   Test 4
107.598
1.25486
4.21527
<.0001
 0.869
0.2391
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is greater than  .1.  The model chosen seems
to adequately describe the data
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 0.0526247


            BMDL = 5.88883e-005
                                     G-327

-------
G.2.55.5. Figure for Additional Model Presented: Power, Unrestricted



                                Power Model with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
         700
         600
         500
         400
300
         200
         100
                   Power
            BMDLBMD
                             50
100          150

      dose
                                                          200
250
   13:3202/082010
                                         G-328

-------
G.2.56. White et al. (1986): CH50
G.2.56.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hillb
Linear
Polynomial, 6 -degree
Power
Hill, unrestricted0
Power, unrestricted
Degrees of
freedom
5
5
4
4
4
5
5
5
3
4
X2 p-value
0.002
0.002
0.001
0.001
0.002
0.001
0.001
0.001
0.071
0.148
AIC
389.664
389.664
390.632
390.632
389.601
394.446
394.446
394.446
381.520
379.265
BMD (ng/kg)
1.957E+01
1.957E+01
1.411E+01
1.411E+01
8.632E+00
3.497E+01
3.497E+01
3.497E+01
1.481E-01
1.211E-01
BMDL
(ng/kg)
1.261E+01
1.261E+01
5.177E+00
5.177E+00
1.498E-K)0
2.568E+01
2.568E+01
2.568E+01
4.351E-03
1.225E-03
Notes

power hit bound (d = 1)

power hit bound (d = 1)
n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted (n = 0.246)
unrestricted
(power = 0.227)
a Nonconstant variance model selected (p = 0.0871).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.2.56.2. Output for Selected Model: Hill
White et al. (1986): CH50
         Hill Model.  (Version:  2.14;  Date:  06/26/2008)
         Input Data File:  C:\l\Blood\71_White_1986_CH50_Hill_l.(d)
         Gnuplot Plotting  File:   C:\l\Blood\71_White_1986_CH50_Hill_l.plt
                                            Mon Feb 08 13:35:56  2010
  [insert  study notes]


   The  form of the response  function is:

   Y[dose]  = intercept  +  v*doseAn/(kAn + doseAn)
   Dependent variable = Mean
   Independent variable =  Dose
   Power parameter restricted to be greater  than 1
   The  variance is to be modeled as Var(i) = exp(lalpha  + rho  *  In(mean(i)))

   Total number of dose groups = 7
   Total number of records  with missing values  = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set to:  le-008
   Parameter Convergence has been set to:  le-008
                                       G-329

-------
                  Default Initial Parameter Values
                         lalpha =
                            rho =
                      intercept =
                              v =
                              n =
                              k =
 5.60999
       0
      91
     -74
0.118036
   1.094
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -n
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
lalpha
lalpha
rho
intercept
V
k

-0
0
0
-0
1
.99
.27
.23
.32
-0

-0
-0
0
rho
.99
1
.28
.24
.33
intercept
0
-0

0
-0
.27
.28
1
.39
.78
0
-0
0

-0
V
.23
.24
.39
1
.85

-0
0
-0
-0

k
.32
.33
.78
.85
1
                                 Parameter Estimates
Confidence Interval
Variable
Upper Conf. Limit
lalpha
7.83989
rho
1.15888
intercept
87.0562
V
-37.2264
n
k
62.6223

Estimate

4.581

0.31293

74.6365

-66.2096

1
20.8286


Std. Err.

1.66273

0.431616

6.33673

14.7876

NA
21.3237

NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
     Table of Data and Estimated Values of Interest
                                                         95.0% Wald

                                                      Lower Conf. Limit

                                                              1.32211

                                                            -0.533022

                                                              62.2167

                                                             -95.1928


                                                              -20.965
                                    G-330

-------
 Dose
Res .
    0
1.094
4.085
 7.14
26.81
48.72
90.56
                 Obs Mean
                              Est Mean
                     Obs Std Dev  Est Std Dev
Scaled
91
54
63
56
41
32
17
74
71
63
57
37
28
20
.6
.3
.8
.7
.4
.3
.8
14
8.
11
25



.1
49
.3
.5
17
17
17
19.
19.
18.
18.
17.
16.
15.
4
3
9
6
4
7
9
2
-2
-0.
-0.
0.
0.
-0.
.39
.54
117
263
589
636
678
 Model Descriptions for likelihoods calculated
 Model Al:        Yij
           Var{e(ij) }

 Model A2:        Yij
           Var{e(ij) }
    Mu (i)  + e (ij )
    SigmaA2

    Mu(i)  + e (ij )
    Sigma(i)A2
 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
Log (likelihood)
-181.340979
-175.820265
-181.238690
-189.800288
-212.367055
# Param' s
8
14
9
5
2
AIC
378.681959
379.640529
380.477380
389.600575
428.734109
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose  levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs.  fitted)
 (Note:   When rho=0 the results of Test 3 and Test  2 will  be  the same.
                     Tests of Interest

   Test    -2*log(Likelihood Ratio)  Test df

   Test 1              73.0936         12

                                     G-331
                                p-value

                               <.0001

-------
   Test 2
   Test 3
   Test 4
11.0414
10.8369
17.1232
6
5
4
  0.0871
 0.05471
0.001829
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1,
model appears to be appropriate

The p-value for Test 3 is less than .1,
different variance model
                  A non-homogeneous variance
                  You may want to consider a
The p-value for Test 4 is less than .1.  You may want to try a different
model
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =        8.63239

            BMDL =       1.49823
                                     G-332

-------
G.2.56.3. Figure for Selected Model: Hill

                               Hill Model with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 OJ
        100
         80
         60
         40
         20
                   Hill
             MDL
BMD
   13:3502/082010
                      10     20     30     40     50     60     70     80     90
                                           dose
G.2.56.4. Output for Additional Model Presented: Hill, Unrestricted
White etal. (1986): CH50
        Hill  Model.  (Version:  2.14;   Date: 06/26/2008)
        Input Data File: C:\l\Blood\71_White_1986_CH50_Hill_U_l.(d)
        Gnuplot Plotting File:   C:\l\Blood\71_White_1986_CH50_Hill_U_l.plt
                                            Mon  Feb  08  13:35:57 2010
 [insert  study notes]


   The  form of the response  function is:

   Y[dose]  = intercept + v*doseAn/(kAn + doseAn)
   Dependent  variable = Mean
   Independent variable = Dose
                                      G-333

-------
   Power parameter is not restricted
   The variance is to be modeled as Var(i) = exp(lalpha

   Total number of dose groups = 7
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                                        rho * In(mean(i))
                  Default Initial Parameter Values
                         lalpha =
                            rho =
                      intercept =
                              v =
                              n =
                              k =
                    5.60999
                          0
                         91
                        -74
                   0.118036
                      1.094
           Asymptotic Correlation Matrix of Parameter Estimates
                 lalpha
              rho    intercept
    lalpha
-0.014
               -1
       rho
0.011
 intercept
0.015
-0.93
-0.35
  -1
0.16
                   0.19
                   -0.4
                 -0.014
-0.16
            -0.19
              0.4
            0.011
              0.16
             -0.16
              0.15
             -0.58
             0.015
 0.19
-0.19
 0.15
-0.02
-0.93
 -0.4
  0.4
-0.58
             -0.02
-0.35
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
10.6349
            rho
0.815757
      intercept
100.595
              Parameter Estimates

                                      95.0% Wald

     Estimate        Std.  Err.      Lower Conf.  Limit

      6.54093          2.08879              2.44698

    -0.245847         0.541645             -1.30745

      89.6302          5.59428              78.6656
                                     G-334

-------
798.315

0.361333

5.88059e+006
    -628.486

    0.246409

      493877
           727.973

          0.058636

      2.74838e+006
                 -2055.29

                 0.131484

            -4.89284e+006
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev
                                                Scaled
    0
1.094
4.085
 7.14
26.81
48.72
90.56
91
54
63
56
41
32
17
89.6
65.2
56.3
51.7
38.3
30.9
22.3
14.1
8.49
11.3
25.5
  17
  17
  17
15.1
15.8
  16
16.2
16.8
17.3
  18
 0.256
 -2.01
  1.17
 0.746
 0.453
 0.175
-0.831
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model      Log(likelihood)
             Al         -181.340979
             A2         -175.820265
             A3         -181.238690
         fitted         -184.759769
              R         -212.367055
# Param's
8
14
9
6
2
AIC
378.681959
379.640529
380.477380
381.519538
428.734109
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?

                                     G-335

-------
           (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.
   Test

   Test 1
   Test 2
   Test 3
   Test 4
                     Tests of Interest
-2*log(Likelihood Ratio)   Test df
            73.0936
            11.0414
            10.8369
            7.04216
12
 6
 5
 3
  p-value

 <.0001
 0.0871
0.05471
0.07057
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1.
model appears to be appropriate

The p-value for Test 3 is less than .1.
different variance model
                              A non-homogeneous variance
                              You may want to consider a
The p-value for Test 4 is less than .1.  You may want to try a different
model
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =       0.148074

            BMDL =    0.00435112
                                     G-336

-------
G.2.56.5. Figure for Additional Model Presented: Hill, Unrestricted



                                   Hill Model with 0.95 Confidence Level
  o
  Q.
 o:

 c
 (0
 OJ
         100
          80
          60
          40
          20
                     Hill
            BMDLBMD
                        10
20
30
40     50

   dose
60
70
   13:3502/082010
80
90
G.3.  ADMINISTERED DOSE: BMDS RESULTS


G.3.1.  Amin et al. (2000): 0.25% Saccharin Consumed, Female


G.3.1.1.  Summary Table of BMDS Modeling Results
Model3
Linear1"
Polynomial,
2-degree
Power
Power, unrestricted0
Degrees of
freedom
1
1
1
0
X2 /7-value
0.358
0.358
0.358
N/A
AIC
179.702
179.702
179.702
180.858
BMD
(ng/kg-day)
8.816E+01
8.816E+01
8.816E+01
7.530E+01
BMDL
(ng/kg-day)
5.890E+01
5.890E+01
5.890E+01
2.537E+01
Notes


power bound hit
(power =1)
unrestricted
(power = 0.605)
a Nonconstant variance model selected (p = 0.0005).

b Best-fitting model, BMDS output presented in this appendix.

c Alternate model, BMDS output also presented in this appendix.
                                           G-337

-------
G.3.1.2.   Output for Selected Model: Linear
Amin et al. (2000): 0.25% Saccharin Consumed, Female
        Polynomial Model.  (Version: 2.13;  Date:  04/08/2008)
        Input Data File: C:\l\l_Amin_2000_25_SC_Linear_l.(d)
        Gnuplot Plotting File:  C:\l\l_Amin_2000_25_SC_Linear_l.plt
                                          Tue  Feb  16  17:22:16  2010
   The form of the response function is:

   Y[dose] = beta 0 + beta l*dose + beta 2*doseA2 +  ...
   Dependent variable = Mean
   Independent variable = Dose
   Signs of the polynomial coefficients are not restricted
   The variance is to be modeled as Var(i) = exp(lalpha  + log(mean(i))  *  rho)

   Total number of dose groups = 3
   Total number of records with missing values =  0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =      5.29482
                            rho =             0
                         beta_0 =      30.8266
                         beta 1 =    -0.204134
           Asymptotic Correlation Matrix of Parameter  Estimates

                 lalpha          rho       beta_0       beta_l

    lalpha            1        -0.99       -0.016          0.03

       rho        -0.99            1        0.013       -0.026

    beta_0       -0.016        0.013             1         -0.94

    beta_l         0.03       -0.026        -0.94             1



                                 Parameter Estimates



                                     G-338

-------
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
0.698746
            rho
3.48995
         beta_0
38.3069
         beta_l
-0.109803
 Estimate

 -2.55843

  2.42056

  30.3968

-0.196699
Std.  Err.

  1.66185

 0.545617

  4.03582

0.0443352
    95.0% Wald

 Lower Conf.  Limit

         -5.8156

         1.35117

         22.4868

       -0.283594
     Table of Data and Estimated Values of Interest
Dose
Res .
0
25
100
N
10
10
10
Obs
31
24
10
Mean
.7
.6
.7
Est
30
25
10
Mean
.4
.5
.7
Obs Std Dev
20.6
12
5.33
Est Std Dev
17.3
14
4.92
Scaled
0.233
-0.2
-0.0204
 Model Descriptions for likelihoods calculated
 Model Al:        Yij
           Var{e(ij) }

 Model A2:        Yij
           Var{e(ij) }
 Mu(i)  + e(i j '
 SigmaA2

 Mu(i)  + e(i j '
 Sigma(i)A2
 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
Log(likelihood)
  -92.841935
  -85.255316
  -85.429148
  -85.851107
  -98.136607
 # Param's
       4
       6
       5
       4
       2
   AIC
193.683870
182.510632
180.858295
179.702213
200.273213
                   Explanation of Tests

                                     G-339

-------
 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.
                     Tests of Interest
           -2*log(Likelihood Ratio)  Test df
        p-value
25.7626
15.1732
0.347663
0.843918
4
2
1
1
<.0001
0.0005072
0.5554
0.3583
Test

Test 1
Test 2
Test 3
Test 4
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
The modeled variance appears
The model chosen seems
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD =        88.1623
            BMDL =
                          58.9029
                                     G-340

-------
G.3.1.3.  Figure for Selected Model: Linear

                             Linear Model with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 OJ
        10
                                                                         100
   17:2202/162010
G.3.1.4.  Output for Additional Model Presented: Power, Unrestricted
Amin et al. (2000): 0.25% Saccharin Consumed, Female
        Power  Model.  (Version:  2.15;   Date: 04/07/2008)
        Input  Data File: C:\l\l_Amin_2000_25_SC_Pwr_U_l.(d)
        Gnuplot Plotting File:   C:\l\l_Amin_2000_25_SC_Pwr_U_l.plt
                                            Tue Feb  16  17:22:17 2010
   The form of  the response function is:

   Y[dose]  = control + slope * doseApower
   Dependent  variable = Mean
   Independent  variable = Dose
                                      G-341

-------
   The power is not restricted
   The variance is to be modeled as Var(i)  = exp(lalpha + log(mean(i))  * rho)

   Total number of dose groups = 3
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =      5.29482
                            rho =            0
                        control =      31.6727
                          slope =    -0.567889
                          power =     0.783745
           Asymptotic Correlation Matrix of Parameter Estimates

lalpha
rho
control
slope
power
lalpha
1
-0.99
0.34
-0.14
-0.061
rho
-0.99
1
-0.42
0.15
0.068
control
0
-0

-0
-0
.34
.42
1
.67
.56
slope
-0.
0.
-0.

0.
14
15
67
1
99
power
-0.061
0.068
-0.56
0.99
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha         -2.48291
1.60693
            rho          2.38455
3.74094
        control            32.99
43.5886
          slope         -1.36469
2.5799
          power         0.605364
1.17077
Parameter Estimates



       Std.  Err.

         2.08669

        0.692047

         5.40754

         2.01258

        0.288476
   95.0% Wald

Lower Conf.  Limit

       -6.57274

        1.02817

        22.3914

       -5.30927

      0.0399625
     Table of Data and Estimated Values of Interest
                                     G-342

-------
 Dose
Res .
                 Obs Mean
                              Est Mean
                              Obs Std Dev  Est Std Dev
                                     Scaled
    0    10
   25    10
  100    10
       31.7
       24.6
       10.7
  33
23.4
10.8
20.6
  12
5.33
18.7
12.4
4.94
-0.223
 0.302
 -0.08
 Warning: Likelihood for fitted model larger than  the  Likelihood  for  model
A3.
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -92.841935
              -85.255316
              -85.429148
              -85.429148
              -98.136607
          # Param's
                4
                6
                5
                5
                2
            AIC
         193.683870
         182.510632
         180.858295
         180.858295
         200.273213
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among  Dose  levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs.  fitted)
 (Note:  When rho=0 the results of Test 3 and Test  2  will  be  the  same.

                     Tests of Interest
   Test

   Test 1
   Test 2
-2*log(Likelihood Ratio)  Test df
            25.7626
            15.1732
                     p-value

                    <.0001
                 0.0005072
                                     G-343

-------
   Test 3             0.347663          1          0.5554
   Test 4         -8.2423e-013          0              NA

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 75.2994


            BMDL = 25.3717
                                     G-344

-------
G.3.1.5.   Figure for Additional Model Presented: Power, Unrestricted



                                Power Model with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
50




45




40




35




30




25




20




15




10




 5
                  Power
                           BMDL
                                                          BMD
                            20
                                 40            60

                                       dose
80
100
   17:2202/162010
G.3.2.  Amin et al. (2000): 0.25% Saccharin Preference Ratio, Female


G.3.2.1.  Summary Table ofBMDS Modeling Results
Model3
Linear1"
Polynomial, 2 -degree
Power
Degrees of
freedom
1
1
1
X2/7-value
0.002
0.002
0.002
AIC
228.094
228.094
228.094
BMD
(ng/kg-day)
1.264E+02
1.264E+02
1.264E+02
BMDL
(ng/kg-day)
6.128E-K)!
6.128E+01
6.128E+01
Notes


power bound hit
(power =1)
' Nonconstant variance model selected (p = 0.0135).

' Best-fitting model, BMDS output presented in this appendix.
                                          G-345

-------
G.3.2.2.   Output for Selected Model: Linear
Amin et al. (2000): 0.25% Saccharin Preference Ratio, Female
        Polynomial Model.  (Version: 2.13;   Date:  04/08/2008)
        Input Data File: C:\l\2_Amin_2000_25_SP_Linear_l.(d)
        Gnuplot Plotting File:  C:\l\2_Amin_2000_25_SP_Linear_l.plt
                                           Tue  Feb 16  17:22:44  2010
   The form of the response function is:

   Y[dose] = beta 0 + beta l*dose + beta  2*doseA2  +  ...
   Dependent variable = Mean
   Independent variable = Dose
   Signs of the polynomial coefficients are  not  restricted
   The variance is to be modeled as Var(i) = exp(lalpha  +  log(mean(i))  *  rho)

   Total number of dose groups = 3
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set to:  le-008
   Parameter Convergence has been set to:  le-008
                  Default Initial Parameter Values
                         lalpha =       6.34368
                            rho =             0
                         beta_0 =       74.2008
                         beta 1 =    -0.219781
           Asymptotic Correlation Matrix  of  Parameter  Estimates

                 lalpha          rho       beta_0        beta_l

    lalpha            1           -1           0.2         -0.28

       rho           -1             1         -0.19          0.28

    beta_0          0.2        -0.19             1         -0.76

    beta_l        -0.28         0.28         -0.76             1



                                 Parameter Estimates



                                     G-346

-------
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
18.4443
            rho
5.78472
         beta_0
86.7211
         beta_l
-0.00907442
 Estimate

 0.338774

  1.43998

  73.6633

-0.207175
Std.  Err.

  9.23768

  2.21674

   6.6623

 0.101074
    95.0% Wald

 Lower Conf. Limit

        -17.7667

        -2.90476

         60.6054

       -0.405276
     Table of Data and Estimated Values of Interest
Dose
Res .
0
25
100
N
10
10
10
Obs
82
58
54
Mean
.1
.1
.9
Est
73
68
52
Mean
.7
.5
.9
Obs St
13.
33.
19.
d Dev
3
9
5
Est St
26.
24.
20.
d Dev
2
8
6
Scale
1
-1
0.
d
.02
.32
295
 Model Descriptions for likelihoods calculated
 Model Al:        Yij
           Var{e(ij) }

 Model A2:        Yij
           Var{e(ij) }
 Mu(i)  + e(i j '
 SigmaA2

 Mu(i)  + e(i j '
 Sigma(i)A2
 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
Log(likelihood)
 -108.574798
 -104.269377
 -105.147952
 -110.046917
 -112.382522
 # Param's
       4
       6
       5
       4
       2
   AIC
225.149597
220.538754
220.295903
228.093834
228.765045
                   Explanation of Tests

                                     G-347

-------
 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2 )
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.)
                     Tests of Interest
-2*log (Likelihood Ratio)   Test df
p-value
16.2263
8.61084
1.75715
9.79793
4
2
1
1
0.00273
0.0135
0.185
0.001747
   Test

   Test 1
   Test 2
   Test 3
   Test 4
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD =        126.365
            BMDL =
                          61.2812
                                     G-348

-------
G.3.2.3.  Figure for Selected Model: Linear



                                  Linear Model with 0.95 Confidence Level
  o
  Q.
 o:

  c
  (0
  OJ
         90
         80
         70
60
         50
         40
         30
                     Linear
                                    BMDL
                                     BlvH)
                          20
                             40
60

 dose
80
100
120
   17:2202/162010
G.3.3.  Amin et al. (2000): 0.50% Saccharin Consumed, Female


G.3.3.1.  Summary Table ofBMDS Modeling Results
Model3
Linear1"
Polynomial, 2 -degree
Power
Power, unrestricted0
Degrees of
freedom
1
1
1
0
/2/7-value
0.031
0.031
0.031
N/A
AIC
159.737
159.737
159.737
157.060
BMD
(ng/kg-day)
9.874E-K)!
9.874E+01
9.874E+01
5.610E+01
BMDL
(ng/kg-day)
6.417E-K)!
6.417E+01
6.417E+01
6.781E+00
Notes


power bound hit
(power =1)
unrestricted
(power = 0.325)
a Nonconstant variance model selected (p = <0.0001).

b Best-fitting model, BMDS output presented in this appendix.

c Alternate model, BMDS output also presented in this appendix.
                                            G-349

-------
G.3.3.2.   Output for Selected Model: Linear
Amin et al. (2000): 0.50% Saccharin Consumed, Female
        Polynomial Model.  (Version: 2.13;  Date:  04/08/2008)
        Input Data File: C:\l\3_Amin_2000_50_SC_Linear_l.(d)
        Gnuplot Plotting File:  C:\l\3_Amin_2000_50_SC_Linear_l.plt
                                          Tue Feb  16  17:23:14  2010
   The form of the response function is:

   Y[dose] = beta 0 + beta l*dose + beta 2*doseA2 +  ...
   Dependent variable = Mean
   Independent variable = Dose
   Signs of the polynomial coefficients are not restricted
   The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  *  rho)

   Total number of dose groups = 3
   Total number of records with missing values =  0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =      4.68512
                            rho =            0
                         beta_0 =      19.3484
                         beta 1 =    -0.158141
           Asymptotic Correlation Matrix of Parameter Estimates

                 lalpha          rho       beta_0       beta_l

    lalpha            1        -0.97        0.018      -0.0021

       rho        -0.97            1       -0.027         0.014

    beta_0        0.018       -0.027             1         -0.95

    beta_l      -0.0021        0.014        -0.95             1



                                 Parameter Estimates



                                     G-350

-------
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
0.948397
            rho
2.9301
         beta_0
24.1962
         beta_l
-0.0707631
 Estimate

-0.997428

  2.13634

  18.1144

-0.135736
Std.  Err.

 0.992786

 0.404989

  3.10302

0.0331501
    95.0% Wald

 Lower Conf.  Limit

        -2.94325

         1.34257

         12.0326

       -0.200709
     Table of Data and Estimated Values of Interest
Dose
Res .
0
25
100
N
10
10
10
Obs Mean
22.4
11.4
4.54
Est Mean
18.1
14.7
4.54
Obs Std Dev
16
7.66
3.33
Est Std Dev
13.4
10.7
3.06
Scaled
1
-0.983
-0.00393
 Model Descriptions for likelihoods calculated
 Model Al:        Yij
           Var{e(ij) }

 Model A2:        Yij
           Var{e(ij) }
 Mu(i)  + e(i j '
 SigmaA2

 Mu(i)  + e(i j '
 Sigma(i)A2
 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
Log(likelihood)
  -83.696404
  -73.511830
  -73.530233
  -75.868688
  -90.294746
 # Param's
       4
       6
       5
       4
       2
   AIC
175.392808
159.023660
157.060467
159.737377
184.589492
                   Explanation of Tests

                                     G-351

-------
 Test 1:

 Test 2:
 Test 3:
 Test 4:
 (Note:
   Test
 Do responses and/or variances differ among Dose levels?
 (A2 vs. R)
 Are Variances Homogeneous?  (Al vs A2)
 Are variances adequately modeled?  (A2 vs. A3)
 Does the Model for the Mean Fit?  (A3 vs. fitted)
When rho=0 the results of Test 3 and Test 2 will be the same.
                     Tests of Interest
  -2*log(Likelihood Ratio)  Test df
p-value
Test 1
Test 2
Test 3
Test 4
33.5658
20.3691
0.0368066
4.67691
4
2
1
1
<.0001
<.0001
0.8479
0.03057
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD =        98.7409
            BMDL =
                           64.169
                                     G-352

-------
G.3.3.3.  Figure for Selected Model: Linear

                             Linear Model with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 OJ
       35
       30
       25
20
15
        10
                  Linear
                                                BMDL
                                                                  BMD
                         20
                              40          60
                                   dose
80
100
   17:2302/162010
G.3.3.4.  Output for Additional Model Presented: Power, Unrestricted
Amin et al. (2000): 0.50% Saccharin Consumed, Female
         Power Model. (Version:  2.15;   Date: 04/07/2008)
         Input Data File: C:\l\3_Amin_2000_50_SC_Pwr_U_l.(d)
         Gnuplot Plotting File:   C:\l\3_Amin_2000_50_SC_Pwr_U_l.plt
                                            Tue Feb  16  17:23:15 2010
   The  form of the response  function is:

   Y[dose]  = control + slope  *  doseApower
   Dependent  variable = Mean
   Independent  variable = Dose
                                      G-353

-------
   The power is not restricted
   The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  * rho)

   Total number of dose groups = 3
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =      4.68512
                            rho =            0
                        control =      22.3564
                          slope =     -3.55874
                          power =     0.349799
           Asymptotic Correlation Matrix of Parameter Estimates

lalpha
rho
control
slope
power
lalpha
1
-0.96
0.34
-0.26
-0.15
rho
-0.96
1
-0.47
0.3
0.15
control
0.34
-0.47
1
-0.73
-0.52
slope
-0.26
0.3
-0.73
1
0.96
power
-0.15
0.15
-0.52
0.96
1
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
1.83541
            rho
2.99953
        control
31.4181
          slope
2.26526
          power
0.597381
 Estimate

-0.708629

  1.96142

  22.6293

 -4.03215

 0.325414
Parameter Estimates



       Std.  Err.

           1.298

        0.529653

         4.48416

         3.21302

        0.138761
   95.0% Wald

Lower Conf.  Limit

       -3.25267

       0.923323

        13.8405

       -10.3296

       0.053447
     Table of Data and Estimated Values of Interest
                                     G-354

-------
 Dose
Res .
                 Obs Mean
                              Est Mean
                              Obs Std Dev  Est Std Dev
                                     Scaled
    0    10
   25    10
  100    10
       22.4
       11.4
       4.54
22.6
11.1
4.58
  16
7.66
3.33
  15
7.46
3.12
-0.0577
  0.105
-0.0475
 Warning: Likelihood for fitted model larger than the  Likelihood  for  model
A3.
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -83.696404
              -73.511830
              -73.530233
              -73.530233
              -90.294746
# Param's
4
6
5
5
2
AIC
175.392808
159.023660
157.060467
157.060467
184.589492
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among  Dose  levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs.  fitted)
 (Note:  When rho=0 the results of Test 3 and Test  2  will  be  the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
-2*log(Likelihood Ratio)  Test df
            33.5658
            20.3691
                     p-value

                    <.0001
                    <.0001
                                     G-355

-------
   Test 3            0.0368066          1          0.8479
   Test 4        -2.84217e-014          0              NA

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 56.0967


            BMDL = 6.78112
                                     G-356

-------
G.3.3.5.  Figure for Additional Model Presented: Power, Unrestricted



                                 Power Model with 0.95 Confidence Level


                   Pnwpr  	
         35





         30





         25
  o
  Q.
 o:

  c
  (0
  OJ
20
15
         10
   17:2302/162010
G.3.4.  Amin et al. (2000): 0.50% Saccharin Preference Ratio, Female


G.3.4.1.  Summary Table ofBMDS Modeling Results
                                                                                    100
Model3
Linear1"
Polynomial,
2-degree
Power
Power, unrestricted0
Degrees of
freedom
1
1
1
0
/2/7-value
0.088
0.088
0.088
N/A
AIC
234.936
234.936
234.936
234.020
BMD
(ng/kg-day)
8.278E+01
8.278E+01
8.278E+01
1.817E+01
BMDL
(ng/kg-day)
5.100E+01
5.100E+01
5.100E+01
l.OOOE-13
Notes


power bound hit
(power =1)
unrestricted
(power = 0.232)
a Constant variance model selected (p = 0.5593).

b Best-fitting model, BMDS output presented in this appendix.

c Alternate model, BMDS output also presented in this appendix.
                                            G-357

-------
G.3.4.2.   Output for Selected Model: Linear
Amin et al. (2000): 0.50% Saccharin Preference Ratio, Female
        Polynomial Model.  (Version: 2.13;  Date:  04/08/2008)
        Input Data File: C:\l\4_Amin_2000_50_SP_LinearCV_l.(d)
        Gnuplot Plotting File:  C:\l\4_Amin_2000_50_SP_LinearCV_l.plt
                                          Tue  Feb 16  17:23:43  2010
   The form of the response function is:

   Y[dose] = beta 0 + beta l*dose + beta 2*doseA2  +  ...
   Dependent variable = Mean
   Independent variable = Dose
   rho is set to 0
   Signs of the polynomial coefficients are not  restricted
   A constant variance model is fit

   Total number of dose groups = 3
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set to:  le-008
   Parameter Convergence has been set to:  le-008
                  Default Initial Parameter Values
                          alpha =      764.602
                            rho =             0    Specified
                         beta_0 =      64.1858
                         beta 1 =    -0.332668
           Asymptotic Correlation Matrix of  Parameter  Estimates

            ( *** The model parameter(s)  -rho
                 have been estimated at a boundary  point,  or  have  been
specified by the user,
                 and do not appear in  the correlation  matrix  )

                              beta_0       beta_l

                              2e-008     1.4e-009

                                   1         -0.7

                                -0.7             1
                                     G-358

alpha
beta 0
beta 1
alpha
1
2e-008
1.4e-009

-------
                                 Parameter Estimates
Confidence Interval
       Variable
Upper Conf. Limit
          alpha
1142.19
         beta_0
77.9876
         beta_l
-0.100752
 Estimate

  758.396

  64.1858

-0.332668
Std.  Err.

  195.817

  7.04184

 0.118327
    95.0% Wald

 Lower Conf. Limit

         374.602

         50.3841

       -0.564584
     Table of Data and Estimated Values of Interest
Dose
Res .
0
25
100
N
10
10
10
Obs
72
44
33
Mean
.7
.5
.8
Est
64
55
30
Mean
.2
.9
.9
Obs S
24
32
24
td Dev
.6
.9
.6
Est S
27
27
27
td Dev
.5
.5
.5
Seal
0
0
ed
.981
1.31
.327
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
Log(likelihood)
 -113.009921
 -112.428886
 -113.009921
 -114.468091
 -117.976057
 # Param's
       4
       6
       4
       3
       2
   AIC
234.019841
236.857773
234.019841
234.936183
239.952114
                                     G-359

-------
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.
                     Tests of Interest
           -2*log(Likelihood Ratio)  Test df
p-value
11.0943
1.16207
1.16207
2.91634
4
2
2
1
0.02552
0.5593
0.5593
0.08769
Test

Test 1
Test 2
Test 3
Test 4
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD =        82.7823
            BMDL =
                          50.9971
                                     G-360

-------
G.3.4.3.  Figure for Selected Model: Linear

                             Linear Model with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 OJ
90


80


70


60


50


40


30


20


10
                  Linear
                                        BMDL
  BMD
                          20
                              40          60
                                   dose
80
100
   17:2302/162010
G.3.4.4.  Output for Additional Model Presented: Power, Unrestricted
Amin et al. (2000): 0.50% Saccharin Preference Ratio, Female
         Power Model.  (Version:  2.15;  Date:  04/07/2008)
         Input Data File: C:\l\4_Amin_2000_50_SP_PwrCV_U_l.(d)
         Gnuplot Plotting File:   C:\l\4_Amin_2000_50_SP_PwrCV_U_l.plt
                                            Tue  Feb 16 17:23:44 2010
   The  form of the response  function is:

   Y[dose]  = control + slope  *  doseApower
   Dependent  variable = Mean
   Independent variable = Dose
                                       G-361

-------
   rho is set to 0
   The power is not restricted
   A constant variance model is fit

   Total number of dose groups = 3
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      764.602
                            rho =
                        control =
                          slope =
                          power =
            0
      72.7273
      -13.387
     0.231973
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
control
slope
power


-1
5
2
alpha
1
.3e-008
.9e-009
.5e-009
control
-1.3e-008
1
-0.4
-0.22
slope
5.9e-009
-0.4
1
0.97
power
2.5e-009
-0.22
0.97
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          688.142
1036.38
        control          72.7273
88.986
          slope          -13.387
17.9639
          power         0.231973
0.757376
Parameter Estimates



       Std.  Err.

         177.677

         8.29543

         15.9957

        0.268067
        95.0% Wald

     Lower Conf.  Limit

               339.9

             56.4686

             -44.738

           -0.293429
     Table of Data and Estimated Values of Interest

                                     G-362

-------
 Dose
Res .
                 Obs Mean
                              Est Mean
                                Obs Std Dev  Est Std Dev
                                                  Scaled
    0
   25
  100
10
10
10
72.7
44.5
33.8
72.7
44.5
33.8
24.6
32.9
24.6
26.2
26.2
26.2
 5.16e-008
-1.27e-008
   -2e-008
Degrees of freedom for Test A3 vs fitted <= 0
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e (i) } = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
              Log(likelihood)
               -113.009921
               -112.428886
               -113.009921
               -113.009921
               -117.976057
# Param's
4
6
4
4
2
AIC
234.019841
236.857773
234.019841
234.019841
239.952114
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
  -2*log(Likelihood Ratio)  Test df
              11.0943
              1.16207
                                  p-value

                                0.02552
                                 0.5593
                                     G-363

-------
   Test 3              1.16207          2          0.5593
   Test 4                    0          0              NA

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.  The modeled variance appears
 to be appropriate here

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 18.1732


            BMDL = le-013
                                     G-364

-------
G.3.4.5.   Figure for Additional Model Presented: Power, Unrestricted



                                 Power Model with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
          10
   17:2302/162010
                                          40
60
80
100
                                               dose
                                          G-365

-------
G.3.5.  Bell et al. (2007): Balano-Preputial Separation, PND 49
G.3.5.1.  Summary Table ofBMDS Modeling Results
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage, 3 -degree
Probit
Weibull
Gamma, unrestricted
Log-logistic,
unrestricted13
Log-probit, unrestricted
Weibull, unrestricted
Degrees of
freedom
2
2
2
2
2
2
2
1
1
1
1
X2 p-value
0.369
0.237
0.456
0.178
0.369
0.248
0.369
0.566
0.484
0.439
0.534
AIC
113.514
114.853
112.952
115.488
113.514
114.723
113.514
113.746
113.908
114.021
113.802
BMD
(ng/kg-day)
7.332E+00
1.501E+01
5.209E-H)0
1.428E+01
7.332E+00
1.399E+01
7.332E+00
1.894E+00
2.127E+00
2.179E+00
2.007E+00
BMDL
(ng/kg-day)
4.687E+00
1.137E+01
2.870E-H)0
9.138E+00
4.687E+00
1.061E+01
4.687E+00
7.609E-02
1.363E-01
1.671E-01
1.075E-01
Notes
power bound hit
(power =1)

slope bound hit
(slope = 1)
slope bound hit
(slope =1)
final B = 0

power bound hit
(power =1)
unrestricted
(power =0.506)
unrestricted
(slope = 0.67)
unrestricted
(slope = 0.389)
unrestricted
(power =0.574)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.


G.3.5.2.  Output for Selected Model: Log-Logistic
Bell et al. (2007): Balano-Preputial Separation, PND 49
         Logistic Model.  (Version: 2.12;  Date:  05/16/2008)
         Input Data File:  C:\l\5_Bell_2007_BPS_LogLogistic_l.(d)
         Gnuplot Plotting  File:   C:\l\5_Bell_2007_BPS_LogLogistic_l.plt
                                             Tue  Feb 16 17:24:10  2010
   The  form of the probability function  is:

   P[response] = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
   Dependent variable =  DichEff
   Independent variable  =  Dose
   Slope parameter is restricted as slope  >= 1

   Total number of observations = 4
                                       G-366

-------
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =    0.0333333
                      intercept =     -3.75371
                          slope =            1
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -slope
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

             background    intercept

background            1        -0.58

 intercept        -0.58            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background        0.0635251

      intercept         -3.84765

          slope                1
                                 Parameter Estimates
                      Std.  Err.
                                                         95. 0% Wald

                                                      Lower Conf. Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.4638
  Reduced model
0.0001309
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -53.7077         4
      -54.476         2       1.53661      2
                                                                    P-value
     -63.9797
                                        1
20.544
3
                                     G-367

-------
           AIC:         112.952
                                  Goodness  of  Fit

Dose
0.0000
2.4000
8.0000
46.0000

Est. Prob.
0.0635
0.1091
0.2000
0.5273

Expected
1.906
3.274
6.001
15.819

Observed
1.000
5.000
6.000
15.000

Size
30
30
30
30
Scaled
Residual
-0.678
1.011
-0.000
-0.300
 ChiA2 = 1.57      d.f. = 2        P-value = 0.4559







   Benchmark Dose Computation




Specified effect =            0.1




Risk Type        =      Extra risk




Confidence level =           0.95




             BMD =        5.20918




            BMDL =        2.86991
                                     G-368

-------
G.3.5.3.  Figure for Selected Model: Log-Logistic

                        Log-Logistic Model with 0.95 Confidence Level

           ; ' '  '      "Tc
        0.7


        0.6


        0.5
 I
 C
 o
 '•5
 ro
        0.4
0.3
        0.2
        0.1
         0  -
  17:2402/162010
                                   20
                                       30
40
                                      dose
G.3.5.4.  Output for Additional Model Presented: Log-Logistic, Unrestricted
Bell et al. (2007): Balano-Preputial Separation, PND 49
         Logistic Model.  (Version: 2.12; Date:  05/16/2008)
         Input Data File:  C:\l\5_Bell_2007_BPS_LogLogistic_U_l.(d)
         Gnuplot Plotting  File:   C:\l\5_Bell_2007_BPS_LogLogistic_U_l.plt
                                             Tue  Feb 16 17:24:10  2010
   The  form of the probability function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose) ) ]
   Dependent variable = DichEff
   Independent variable =  Dose
   Slope  parameter is not  restricted

   Total  number of observations = 4
                                       G-369

-------
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =    0.0333333
                      intercept =     -2.54947
                          slope =     0.615936
           Asymptotic Correlation Matrix of Parameter Estimates

             background    intercept        slope

background            1        -0.49         0.35

 intercept        -0.49            1        -0.93

     slope         0.35        -0.93            1
                                 Parameter Estimates
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background        0.0354714
*
      intercept         -2.70296
*
          slope         0.670238
                      Std.  Err.
                 95.0% Wald

              Lower Conf.  Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.4827
  Reduced model
0.0001309

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -53.7077         4
     -53.9541         3      0.492844      1
     -63.9797
      113.908
1
20.544
3
                            P-value
                                     G-370

-------
                                  Goodness  of  Fit

Dose
0.0000
2.4000
8.0000
46.0000

Est. Prob.
0.0355
0.1392
0.2405
0.4848

Expected
1.064
4.176
7.216
14.544

Observed
1.000
5.000
6.000
15.000

Size
30
30
30
30
Scaled
Residual
-0.063
0.435
-0.520
0.167
 ChiA2 = 0.49      d.f. = 1        P-value = 0.4836







   Benchmark Dose Computation




Specified effect =            0.1




Risk Type        =      Extra risk




Confidence level =           0.95




             BMD =        2 . 12667




            BMDL =        0. 13633
                                     G-371

-------
G.3.5.5.  Figure for Additional Model Presented: Log-Logistic, Unrestricted



                               Log-Logistic Model with 0.95 Confidence Level


               ; '           "Tc
          0.7




          0.6




          0.5
 T3

 £
 O
 c
 O
 •*=
 O
 (0
          0.4
0.3
          0.2
          0.1
           0   -

            BMDL
   17:2402/162010
                                              20
                                                   30
40
                                                 dose
                                           G-372

-------
G.3.6.  Cantoni et al. (1981): Urinary Coproporhyrins, 3 Months
G.3.6.1.  Summary Table ofBMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Power
Power, unrestricted0
Hill, unrestricted
Degrees of
freedom
2
2
1
1
1
2
2
2
1
0
X2 p-value
0.002
0.002
0.341
0.341
0.535
0.002
0.002
0.002
0.665
N/A
AIC
33.792
33.792
23.881
23.881
23.359
33.301
33.301
33.301
23.162
24.974
BMD
(ng/kg-day)
1.101E+02
1.101E+02
3.741E-01
3.741E-01
3.273E-01
7.734E+01
7.734E+01
7.734E+01
4.637E-03
7.264E-02
BMDL
(ng/kg-day)
5.318E+01
5.318E+01
1.253E-01
1.253E-01
error
1.975E+01
1.975E+01
1.975E+01
8.796E-08
1.656E-04
Notes

power hit bound (d = 1)

power hit bound (d = 1)
n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted
(power = 0.22)
unrestricted (n = 0.48)
a Nonconstant variance model selected (p = 0.0039).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.3.6.2.  Output for Selected Model: Exponential (M4)
Cantoni et al. (1981): Urinary Coproporhyrins, 3 Months
         Exponential Model.  (Version: 1.61;   Date: 7/24/2009)
         Input Data File:  C:\l\6_Cantoni_1981_UriCopro_Exp_l.(d)
         Gnuplot Plotting  File:
                                             Tue Feb 16 17:24:39  2010
 Figurel-UrinaryCoproporphyrin Smonths
   The  form of the response function by Model:
       Model 2:
       Model 3:
       Model 4:
       Model 5:
Y[dose]
Y[dose]
Y[dose]
Y[dose]
a
a
a
a
expfsign  *  b  *  dose}
exp{sign  *  (b * dose)Ad}
[c-(c-l)  *  exp{-b * dose}]
[c-(c-l)  *  exp{-(b * dose)Ad}]
    Note:  Y[dose] is the  median response  for exposure = dose;
           sign = +1 for increasing trend  in data;
           sign = -1 for decreasing trend.

       Model 2 is nested within Models  3 and 4.
       Model 3 is nested within Model 5.
       Model 4 is nested within Model 5.
   Dependent variable  =  Mean
   Independent variable  = Dose
                                       G-373

-------
Data are assumed to be distributed: normally
Variance Model: exp(lnalpha +rho *ln(Y[dose]))
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i)) * rho)

Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to:  le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 4

                 Inalpha             -1.50063
                     rho              2.60979
                       a             0.704303
                       b            0.0205927
                       c              4.47268
                       d                    1
                  Parameter Estimates

                Variable          Model 4

                 Inalpha            -1.74154
                     rho             2.66803
                       a            0.755982
                       b              0.3715
                       c             3.93845
                       d                   1


         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0      4
   1.43      4
   14.3      4
    143      4
               Estimated Values of Interest

   Dose      Est Mean      Est Std     Scaled Residual
0.7414
1.807
2.734
3
0.3475
0.8341
1.506
2.6
      0         0.756       0.2882          -0.1014
   1.43         1.671       0.8307           0.3265
   14.3         2.966        1.786          -0.2607
    143         2.977        1.794          0.02532


                                  G-374

-------
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)

     Model  R:        Yij = Mu + e(i)
               Var{e(ij)} = SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
4
-12.90166
-6.203643
-6.487204
-15.73713
-6.940389
5
8
6
2
5
35.80333
28.40729
24.97441
35.47427
23.88078
   Additive constant for all log-likelihoods =      -14.7.  This  constant
added to the
   above values gives the log-likelihood including the term that  does  not
   depend on the model parameters.
R)
                              Explanation of Tests

Test 1:  Does response and/or variances differ among Dose  levels?  (A2  vs.

Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)

Test 6a: Does Model 4 fit the data?  (A3 vs  4)
     Test

     Test 1
     Test 2
     Test 3
    Test 6a
                         Tests of Interest

                -2*log(Likelihood Ratio)
                                                  D. F.
p-value
19.07
13.4
0.5671
0.9064
6
3
2
1
0.004052
0.003854
0.7531
0.3411
     The p-value for Test 1 is less than  .05.  There appears to be  a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

                                     G-375

-------
  The p-value for Test 2 is less than .1.  A non-homogeneous
  variance model appears to be appropriate.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is greater than  .1.  Model 4 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =     0.374114

              BMDL =     0.125287
                                 G-376

-------
G.3.6.3.  Figure for Selected Model: Exponential (M4)

                    Exponential_beta Model 4 with 0.95 Confidence Level
 c
 o
 Q_
 in
 
-------
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
               Default Initial Parameter Values
                      lalpha =      0.90039
                         rho =            0
                     control =     0.741372
                       slope =      1.00533
                       power =     0.163111
        Asymptotic Correlation Matrix of Parameter Estimates
lalpha
lalpha 1 -0
rho -0.62
control -0.53 0
slope -0.038 -0
power 0.027 -0
Confidence Interval
Variable Estimate
Upper Conf. Limit
lalpha -1.78404
-0.57478
rho 2.6428
4.10197
control 0.757242
1.03157
slope 0.927009
1.56581
power 0.220276
0.409334
Table of Data and Estimated
Dose N Obs Mean Est
Res .
rho control
.62 -0.53
1 0.43
.43 1
.24 -0.3
.16 0.09
Parameter Estimates
Std. Err.
0.61698
0.74449
0.139966
0.325923
0.0964599
Values of Interest
Mean Obs Std Dev
slope power
-0.038 0.027
-0.24 -0.16
-0.3 0.09
1 -0.72
-0.72 1
95. 0% Wald
Lower Conf. Limit
-2.9933
1.18363
0.482915
0.288212
0.031218
Est Std Dev Scaled
                                  G-378

-------
    0
 1.43
 14.3
  143
      0.741
       1.81
       2.73
          3
0.757
 1.76
 2.42
 3.52
0.348
0.834
 1.51
  2.6
0.284
0.865
 1.32
 2.16
-0.112
 0.108
 0.471
-0.483
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -12.901663
               -6.203643
               -6.487204
               -6.580755
              -15.737135
           # Param's
                 5
                 8
                 6
                 5
                 2
             AIC
           35.803325
           28.407287
           24.974409
           23.161510
           35.474269
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
-2*log(Likelihood Ratio)  Test df
             19.067
             13.396
           0.567122
           0.187101
          6
          3
          2
          1
         p-value

      0.004052
      0.003854
        0.7531
        0.6653
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
                                     G-379

-------
The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is greater than .1.  The model chosen seems
to adequately describe the data
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 0.00463746


            BMDL = 8.79634e-008
                                     G-380

-------
G.3.6.5.   Figure for Additional Model Presented: Power, Unrestricted



                               Power Model with 0.95 Confidence Level
 0)
 to

 o
 CL
 (/)
 0)
 or
 6




 5




 4




 3




 2




 1




 0




-1
                Power
        BMDLBMD
             0
   17:2402/162010
               20
40
60       80

    dose
100
120
140
                                           G-381

-------
G.3.7.  Cantoni et al. (1981): Urinary Porphyrins
G.3.7.1.  Summary Table ofBMDS Modeling Results
Model3
Exponential (M2)b
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Power
Power, unrestricted
Degrees of
freedom
2
2
1
1
0
2
1
2
1
X2 p-value
<0.0001
0.0001
0.0001
0.0001
N/A
0.0001
0.0001
0.0001
0.0001
AIC
58.753
58.753
63.138
63.138
62.356
62.487
10.000
62.487
59.914
BMD
(ng/kg-day)
1.223E-K)!
1.223E+01
2.227E-01
2.227E-01
9.363E+00
7.732E-01
error
7.732E-01
1.025E-01
BMDL
(ng/kg-day)
9.037E+00
9.037E+00
1.137E-01
1.137E-01
4.664E+00
2.816E-01
error
2.816E-01
2.389E-02
Notes

power hit bound (d=\)

power hit bound (d=\)



power bound hit
(power =1)
unrestricted
(power = 0.746)
a Nonconstant variance model selected (p = O.0001).
b Best-fitting model, BMDS output presented in this appendix.
G.3.7.2.  Output for Selected Model: Exponential (M2)
Cantoni et al. (1981): Urinary Porphyrins
         Exponential Model.  (Version:  1.61;   Date: 7/24/2009)
         Input Data File: C:\l\7_Cantoni_1981_UriPor_Exp_l.(d)
         Gnuplot Plotting File:
                                            Tue Feb 16  17:25:14  2010
 Table  1,  dose converted to ng per  kg per day
   The  form of the response function  by Model:
      Model 2:      Y[dose] = a  *  exp{sign * b * dose}
      Model 3:      Y[dose] = a  *  exp{sign * (b * dose)Ad}
      Model 4:      Y[dose] = a  *  [c-(c-l)  * exp{-b * dose}]
      Model 5:      Y[dose] = a  *  [c-(c-l)  * exp{-(b * dose)Ad}]

    Note:  Y[dose]  is the median response for exposure = dose;
           sign = +1 for increasing  trend in data;
           sign = -1 for decreasing  trend.

      Model 2  is nested within  Models 3 and 4.
      Model 3  is nested within  Model  5.
      Model 4  is nested within  Model  5.
   Dependent variable = Mean
   Independent variable = Dose
   Data  are assumed to be distributed:  normally
   Variance Model:  exp(lnalpha  +rho  *ln(Y[dose]))
   The variance is  to be modeled  as  Var(i)  = exp(lalpha  +  log(mean(i))

                                      G-382
rho)

-------
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 2

                 Inalpha             -3.57509
                     rho              2.23456
                       a              3.83141
                       b            0.0277822
                       c                    0
                       d                    1
                  Parameter Estimates

                Variable          Model 2

                 Inalpha          -1.55886
                     rho           1.77962
                       a           4.17268
                       b         0.0270415
                       c                 0
                       d                 1


         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
2.27
5.55
7.62
196.9
0.49
0.85
1.79
63.14
      0      4
   1.43      4
   14.3      3
    143      3
               Estimated Values of Interest

   Dose      Est Mean      Est Std     Scaled Residual
      0         4.173        1.635           -2.327
   1.43         4.337        1.692            1.433
   14.3         6.143        2.307            1.109
    143         199.4        51.04         -0.08645
Other models for which likelihoods are calculated:

                                  G-383

-------
     Model Al:        Yij
               Var{e(ij) }

     Model A2:        Yij
               Var{e(ij) }

     Model A3:        Yij
               Var{e(ij) }

     Model  R:        Yij
               Var{e(ij) }
                         Mu(i) + e (ij)
                         SigmaA2

                         Mu(i) + e (ij)
                         Sigma(i)A2

                         Mu(i) + e (i j)
                         exp(lalpha + log(mean(i)) * rho)

                         Mu + e(i)
                         SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
2
-51.42175
-15.31211
-15.66963
-68.75058
-25.37651
5
8
6
2
4
112.8435
46.62422
43.33925
141.5012
58.75302
   Additive constant for all log-likelihoods =     -12.87.  This  constant
added to the
   above values gives the log-likelihood including the term that  does not
   depend on the model parameters.
R)
Test 1:

Test 2:
Test 3:
Test 4:
                     Explanation of Tests

Does response and/or variances differ among Dose levels?  (A2  vs.

Are Variances Homogeneous?  (A2 vs. Al)
Are variances adequately modeled?  (A2 vs. A3)
Does Model 2 fit the data?  (A3 vs. 2)
     Test

     Test 1
     Test 2
     Test 3
     Test 4
                         Tests of Interest

                -2*log(Likelihood Ratio)
                                                  D. F.
                                                    p-value
106.9
72.22
0.715
19.41
6
3
2
2
< 0.0001
< 0.0001
0.6994
< 0.0001
     The p-value for Test 1 is less than  .05.  There appears to be a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

     The p-value for Test 2 is less than  .1.  A non-homogeneous
     variance model appears to be appropriate.
                                     G-384

-------
     The p-value  for Test 3 is greater than .1.  The modeled
     variance  appears to be appropriate here.

     The p-value  for Test 4 is less  than .1.   Model 2 may  not adequately
     describe  the data; you may want to consider another model.
   Benchmark  Dose Computations:

     Specified  Effect = 1.000000

            Risk Type = Estimated  standard deviations  from control

     Confidence Level = 0.950000

                   BMD =      12.2272

                  BMDL =      9.03732
G.3.7.3.  Figure for Selected Model: Exponential (M2)

                        Exponential_beta Model 2 with 0.95 Confidence Level
 o
 Q.
 ro
 CD
        350
        300
        250
200
        150
        100
         50
                       Exponential
              BMDL  BMD
                      20
                      40
60      80
   dose
100
120
140
  17:2502/162010
                                      G-385

-------
G.3.8.  Crofton et al. (2005): Serum, T4
G.3.8.1.  Summary Table ofBMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Hill
Linear
Polynomial, 8 -degree
Power
Power, unrestricted
Degrees of
freedom
8
8
7
7
6
8
8
8
7
X2 p-value
0.0001
0.0001
0.957
0.957
0.973
0.0001
0.0001
0.0001
0.030
AIC
518.241
518.241
476.204
476.204
477.434
523.518
523.518
523.518
489.670
BMD
(ng/kg-day)
2.136E+03
2.136E+03
5.633E-K)!
5.633E+01
5.564E+01
4.246E+03
4.246E+03
4.246E+03
2.179E+01
BMDL
(ng/kg-day)
1.157E+03
1.157E+03
3.006E-K)!
3.006E+01
2.590E+01
3.086E+03
3.086E+03
3.086E+03
2.271E+00
Notes

power hit bound (d = 1)

power hit bound (d = 1)



power bound hit
(power =1)
unrestricted
(power =0.2 17)
a Constant variance model selected (p = 0.7647).
b Best-fitting model, BMDS output presented in this appendix.
G.3.8.2.  Output for Selected Model: Exponential (M4)
Crofton et al. (2005): Serum, T4
         Exponential Model.  (Version:  1.61;  Date: 7/24/2009)
         Input Data File: C:\l\8_Crofton_2005_T4_ExpCV_l.(d)
         Gnuplot Plotting File:
                                            Tue Feb 16  17:26:01 2010
   The  form of the response  function by Model:
      Model 2:     Y[dose] = a  *  exp{sign * b * dose}
      Model 3:     Y[dose] = a  *  exp{sign *  (b * dose)Ad}
      Model 4:     Y[dose] = a  *  [c-(c-l) * exp{-b * dose}]
      Model 5:     Y[dose] = a  *  [c-(c-l) * exp{-(b *  dose)Ad}]

    Note:  Y[dose]  is the median response for exposure  =  dose;
           sign = +1 for increasing  trend in data;
           sign = -1 for decreasing  trend.

      Model 2 is nested within  Models 3 and 4.
      Model 3 is nested within  Model 5.
      Model 4 is nested within  Model 5.
   Dependent variable = Mean
   Independent variable = Dose
   Data  are assumed to be distributed:  normally
   Variance Model:  exp(lnalpha  +rho  *ln(Y[dose])

                                      G-386

-------
rho is set to 0.
A constant variance model is fit.

Total number of dose groups = 10
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values
               Variable

                 Inalpha
                     rho(S)
                       a
                       b
                       c
                       d
Model 4

     5.47437
           0
     104.999
 0.000371694
    0.445764
           1
   (S) = Specified
                  Parameter Estimates
                Variable

                 Inalpha
                     rho
                       a
                       b
                       c
                       d
 Model 4

    5.50283
          0
     99.776
 0.00728387
   0.533516
          1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
0
0.1
3
10
30
100
300
1000
3000
le+004
14
6
12
6
6
6
6
6
6

100
96.27
98.57
99.76
93.32
70.94
62.52
52.68
54.66
49.15
15.
14.
18.
19.
12.
12.
14.
22.
19.
11
44
98
11
04
11
74
75
73
71
.15
               Estimated Values of Interest
                                  G-387

-------
      Dose
                Est Mean
                              Est Std
          Scaled Residual
0
0.1
3
10
30
100
300
1000
3000
le+004
99.
99.
98.
96.
90.
75
58.
53.
53.
53.
78
74
77
51
64
.7
47
26
23
23
15
15
15
15
15
15
15
15
15
15
.66
.66
.66
.66
.66
.66
.66
.66
.66
.66
0.
-0
-0.
0
0
-
0
-0.
0
-0
05325
.5434
04357
.5085
.4195
0.744
.6334
09133
.2237
.5218
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i))  *  rho)

     Model  R:        Yij = Mu + e(i)
               Var{e(ij)} = SigmaA2
                     Model
Likelihoods of Interest

Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
4
-233.0774
-230.2028
-233.0774
-268.4038
-234.1019
11
20
11
2
4
488.1549
500.4056
488.1549
540.8076
476.2038
   Additive constant for all log-likelihoods =      -66.16.   This  constant
added to the
   above values gives the log-likelihood including  the  term that  does  not
   depend on the model parameters.
                                 Explanation of Tests

   Test 1:  Does response and/or variances differ among  Dose  levels?  (A2  vs.
R)
   Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
   Test 3:  Are variances adequately modeled?  (A2 vs. A3)

   Test 6a: Does Model 4 fit the data?  (A3 vs  4)
                                     G-388

-------
76.4
5.749
5.749
2.049
18
9
9
7
< 0.0001
0.7647
0.7647
0.9571
                         Tests of Interest

  Test          -2*log(Likelihood Ratio)        D. F.         p-value

  Test 1
  Test 2
  Test 3
 Test 6a
  The p-value for Test 1 is less than .05.  There appears to be a
  difference between response and/or variances among the dose
  levels,  it seems appropriate to model the data.

  The p-value for Test 2 is greater than  .1.  A homogeneous
  variance model appears to be appropriate here.

  The p-value for Test 3 is greater than  .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is greater than .1.  Model 4 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =      56.3321

              BMDL =      30.0635
                                 G-389

-------
G.3.8.3.  Figure for Selected Model: Exponential (M4)



                           Exponential_beta Model 4 with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
         120
         100
80
          60
          40
          20
                          Exponential
            BMDUBMD
                            2000        4000         6000         8000        10000

                                               dose
   17:2602/162010
                                         G-390

-------
G.3.9.  Franc et al. (2001): S-D Rats, Relative Liver Weight
G.3.9.1.  Summary Table ofBMDS Modeling Results
Model3
Hill
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Linear
Polynomial,
3 -degree
Powerb
Hill, unrestricted
Power, unrestricted0
Degrees of
freedom
1
2
2
1
1
2
2
2
0
1
X2 p-value
0.797
0.935
0.935
0.797
0.797
0.967
0.967
0.967
N/A
0.805
AIC
236.371
234.440
234.440
236.371
236.371
234.372
234.372
234.372
238.366
236.365
BMD
(ng/kg-day)
1.826E+01
2.262E+01
2.262E+01
1.827E+01
1.827E+01
1.861E+01
1.861E+01
1.861E-K)!
1.726E+01
1.725E+01
BMDL
(ng/kg-day)
5.463E+00
1.757E+01
1.757E+01
6.112E+00
6.112E+00
1.339E+01
1.339E+01
1.339E+01
2.022E+00
2.003E+00
Notes
n lower bound hit (n = 1)

power hit bound (d=\)

power hit bound (d=\)


power bound hit
(power = 1)
unrestricted (n = 0.965)
unrestricted
(power = 0.962)
a Constant variance model selected (p = 0.107).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.3.9.2.  Output for Selected Model: Power
Franc et al. (2001): S-D Rats, Relative Liver Weight
         Power Model.  (Version:  2.15;  Date:  04/07/2008)
         Input Data File:  C:\l\88_Franc_2001_SD_RelLivWt_PowerCV_l.(d)
         Gnuplot Plotting  File:   C:\l\88_Franc_2001_SD_RelLivWt_PowerCV_l.plt
                                             Fri  Apr 16 16:28:45  2010
 Figure  5,  SD rats, relative liver weight


   The  form of the response  function is:

   Y[dose]  = control +  slope * doseApower
   Dependent variable = Mean
   Independent variable =  Dose
   rho  is  set to 0
   The  power is restricted to be greater  than or equal to  1
   A  constant variance model is fit

   Total  number of dose groups = 4
   Total  number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to:  le-008
   Parameter Convergence has been set to:  le-008

                                       G-391

-------
                  Default Initial Parameter Values
                          alpha =      527.447
                            rho =
                        control =
                          slope =
                          power =
                          0
                        100
                    1.15946
                   0.839423
                 Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho    -power
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
                  alpha

     alpha            1

   control     1.3e-012

     slope    -6.2e-013
          control        slope

         1.3e-012    -6.2e-013

                1        -0.67

            -0.67            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          462.485
689.099
        control          101.047
111.053
          slope         0.542984
0.733788
          power                1
              Parameter Estimates



                     Std.  Err.

                       115.621

                       5.10511

                     0.0973507

                            NA
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
                         95.0% Wald

                      Lower Conf.  Limit

                              235.872

                              91.0415

                             0.352181
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev
                                                 Scaled
    0
   10
100
108
101
106
  14
16.9
21.5
21.5
-0.138
 0.208
                                     G-392

-------
   30
  100
        117
        155
117
155
25.9
30.9
21.5
21.5
 -0.0702
0.000298
 Model Descriptions for likelihoods calculated


 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses  any fixed variance parameters that
     were specified by the user

 Model  R:          Yi = Mu + e(i)
            Var{e(i)} = SigmaA2


                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -114.152281
             -111.103649
             -114.152281
             -114.185827
             -125.052064
         # Param's
               5
               8
               5
               3
               2
            AIC
         238.304562
         238.207299
         238.304562
         234.371654
         254.104127
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.
   Test

   Test 1
   Test 2
   Test 3
   Test 4
                     Tests of Interest
-2*log(Likelihood Ratio)  Test df
            27.8968
            6.09726
            6.09726
          0.0670927
        6
        3
        3
        2
        p-value

       <.0001
        0.107
        0.107
        0.967
The p-value for Test 1 is less than  .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than  .1.
model appears to be appropriate here
                                 A homogeneous variance
                                     G-393

-------
The p-value for Test 3 is greater than .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is greater than .1.  The model chosen seems
to adequately describe the data
               Benchmark Dose Computation

Specified effect =           0.1

Risk Type        =     Relative risk

Confidence level =          0.95

             BMD = 18.6096


            BMDL = 13.3879
                                     G-394

-------
 o
 Q.
 ro
 0)
G.3.9.3.  Figure for Selected Model: Power

                           Power Model with 0.95 Confidence Level

        180 ""
        160
        140
        120
        100
        80
Power
                 BMDL   BMD
                        20
                  40         60
                       dose
80
100
  16:2804/162010
G.3.9.4.  Output for Additional Model Presented: Power, Unrestricted
Franc et al. (2001): S-D Rats, Relative Liver Weight
         Power Model.  (Version:  2.15;   Date: 04/07/2008)
         Input Data File: C:\l\88_Franc_2001_SD_RelLivWt_PowerCV_U_l.(d)
         Gnuplot Plotting File:
C:\l\8 8_Franc_2 0 0!_SD_RelLivWt_PowerCV_U_l.pit
                                             Fri Apr  16 16:28:46 2010
 Figure  5,  SD rats, relative  liver weight


   The form of the response function is:

   Y[dose]  = control + slope  *  doseApower
   Dependent  variable = Mean
   Independent variable = Dose
   rho is  set to 0
   The power  is not restricted
                                      G-395

-------
   A constant variance model is fit

   Total number of dose groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      527.447
                            rho =
                        control =
                          slope =
                          power =
            0
          100
      1.15946
     0.839423
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
control
slope
power
alpha
1
le-009
-6.2e-010
4.7e-010
control
le-009
1
-0.74
0.71
slope
-6.2e-010
-0.74
1
-1
power
4.7e-010
0.71
-1
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          462.394
688.963
        control          100.636
114.927
          slope         0.650456
3.46718
          power         0.961853
1.87359
Parameter Estimates



       Std.  Err.

         115.598

         7.29156

         1.43713

        0.465182
        95.0% Wald

     Lower Conf.  Limit

             235.825

             86.3448

            -2.16627

           0.0501134
     Table of Data and Estimated Values of Interest
                                     G-396

-------
 Dose
Res .
                 Obs Mean
                              Est Mean
                              Obs Std Dev  Est Std Dev
                                    Scaled
    0
   10
   30
  100
        100
        108
        117
        155
101
107
118
155
  14
16.9
25.9
30.9
21.5
21.5
21.5
21.5
-0.0836
  0.192
 -0.128
 0.0192
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e (i) } = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -114.152281
             -111.103649
             -114.152281
             -114.182670
             -125.052064
         # Param's
               5
               8
               5
               4
               2
            AIC
         238.304562
         238.207299
         238.304562
         236.365340
         254.104127
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2 )
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.
   Test

   Test 1
   Test 2
   Test 3
   Test 4
                     Tests of Interest
-2*log(Likelihood Ratio)  Test df
            27.8968
            6.09726
            6.09726
           .0607785
        6
        3
        3
        1
        p-value

       <.0001
        0.107
        0.107
       0.8053
                                     G-397

-------
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
The modeled variance appears
The model chosen seems
               Benchmark Dose Computation

Specified effect =           0.1

Risk Type        =     Relative risk

Confidence level =          0.95

             BMD = 17.2469


            BMDL = 2.00336
                                     G-398

-------
G.3.9.5.  Figure for Additional Model Presented: Power, Unrestricted
                              Power Model with 0.95 Confidence Level
        180
        160
 8      14°

 o
 Q.
 ro
 (D
        120
        100
         80
                  Power
            BMDL
BMD
               0
   16:2804/162010
 20
40
60
80
100
                                            dose
                                           G-399

-------
G.3.10. Franc et al. (2001): L-E Rats, Relative Liver Weight
G.3.10.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hillb
Linear
Polynomial,
3 -degree
Power
Hill, unrestricted0
Power, unrestricted
Degrees of
freedom
2
2
1
1
1
2
1
2
0
1
X2 p-value
0.245
0.245
0.607
0.607
0.703
0.273
0.0001
0.273
N/A
0.940
AIC
210.148
210.148
209.599
209.599
209.480
209.933
10.000
209.933
211.341
209.340
BMD
(ng/kg-day)
5.143E+01
5.143E+01
1.476E+01
1.476E+01
1.321E+01
4.753E+01
1.505E+01
4.753E+01
1.163E+01
1.155E+01
BMDL
(ng/kg-day)
3.188E+01
3.188E+01
3.702E+00
3.702E+00
1.591E+00
2.788E+01
error
2.788E+01
9.756E-01
1.513E-02
Notes

power hit bound (d = 1)

power hit bound (d = 1)
n lower bound hit
(« = 1)


power bound hit
(power =1)
unrestricted (n = 0.418)
unrestricted
(power =0.394)
a Nonconstant variance model selected (p = 0.0632).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.3.10.2. Output for Selected Model: Hill
Franc et al. (2001): L-E Rats, Relative Liver Weight
         Hill Model.  (Version:  2.14;  Date:  06/26/2008)
         Input Data File:  C:\l\89_Franc_2001_LE_RelLivWt_Hill_l.(d)
         Gnuplot Plotting  File:   C:\l\89_Franc_2001_LE_RelLivWt_Hill_l.plt
                                             Fri Apr 16  16:29:20  2010
 Figure  5,  L-E rats,  relative liver weight


   The  form of the response function is:

   Y[dose]  = intercept  +  v*doseAn/(kAn +  doseAn)
   Dependent variable  =  Mean
   Independent variable  =  Dose
   Power parameter restricted to be greater  than 1
   The  variance is to  be modeled as Var(i) = exp(lalpha  +  rho * In(mean(i)))

   Total number of dose  groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set to:  le-008
   Parameter Convergence has been set to:  le-008

                                       G-400

-------
                  Default Initial Parameter Values
                         lalpha =      5.41581
                            rho =            0
                      intercept =          100
                              v =       22.225
                              n =     0.329526
                              k =      40.8403
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -n
                 have been estimated at a boundary point,  or have been
specified by the user,
                 and do not appear in the correlation matrix )

lalpha
rho
intercept
V
k
lalpha
1
-1
-0.18
0.38
0.2
rho
-1
1
0.17
-0.38
-0.2
intercept
-0.18
0.17
1
-0.13
0.39

0.
-0.
-0.

0.
V
38
38
13
1
77
k
0.2
-0.2
0.39
0.77
1
                                 Parameter Estimates
Confidence Interval
Variable
Upper Conf. Limit
lalpha
17.9973
rho
11.4729
intercept
106.853
V
53.6856
n
k
84.1966

Estimate

-15.3958

4.38043

99.5667

28.8965

1
25.1273


Std. Err.

17.0376

3.61867

3.7178

12.6477

NA
30.138

                                                         95.0% Wald

                                                      Lower Conf.  Limit

                                                             -48.7889

                                                             -2.71204

                                                                92.28

                                                              4.10739


                                                             -33.9421
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
                                    G-401

-------
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
    0
   10
   30
  100
        100
        106
        117
        122
99.6
 108
 115
 123
  10
17.9
8.97
19.9
10.8
12.8
14.9
  17
  0.114
 -0.329
  0.288
-0.0723
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -100.516456
              -96.870820
              -99.666984
              -99.739888
             -105.717087
          # Param's
                5
                8
                6
                5
                2
            AIC
         211.032912
         209.741641
         211.333969
         209.479776
         215.434174
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
-2*log(Likelihood Ratio)  Test df
            17.6925
            7.29127
                     p-value

                  0.007048
                   0.06317
                                     G-402

-------
   Test 3              5.59233          2         0.06104
   Test 4             0.145807          1          0.7026

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is less than .1.  You may want to consider a
different variance model

The p-value for Test 4 is greater than  .1.  The model chosen seems
to adequately describe the data
        Benchmark Dose Computation

Specified effect =           0.1

Risk Type        =     Relative risk

Confidence level =           0.95

             BMD =        13.2094

            BMDL =       1.59127
                                     G-403

-------
G.3.10.3. Figure for Selected Model: Hill

                               Hill Model with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 OJ
        140
        130
        120
        110
        100
         90
                           20
40          60
     dose
80
100
   16:2904/162010
G.3.10.4. Output for Additional Model Presented: Hill, Unrestricted
Franc et al. (2001): L-E Rats, Relative Liver Weight
        Hill  Model.  (Version:  2.14;   Date: 06/26/2008)
        Input Data File: C:\l\89_Franc_2001_LE_RelLivWt_Hill_U_l.(d)
        Gnuplot Plotting File:   C:\l\89_Franc_2001_LE_RelLivWt_Hill_U_l.plt
                                            Fri Apr 16 16:29:27 2010
 Figure  5,  L-E rats, relative  liver weight


   The form of the response  function is:

   Y[dose]  = intercept + v*doseAn/(kAn + doseAn)
   Dependent  variable = Mean
   Independent variable = Dose
                                      G-404

-------
   Power parameter is not restricted
   The variance is to be modeled as Var(i) = exp(lalpha

   Total number of dose groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                                         rho * In(mean(i))
                  Default Initial Parameter Values
lalpha =
rho =
intercept =
v =
n =
k =
5.41581
0
100
22.225
0.329526
40.8403
           Asymptotic Correlation Matrix of Parameter Estimates
                 lalpha
               rho    intercept
    lalpha
-0.13
                -1
       rho
0.13
 intercept
0.011
   -1
-0.21
                 -0.099
0.21
             0.099
            -0.21
             0.21
-0.099
 0.099
 0.023
            0.023
 0.23
-0.23
 0.14
              -0.84
                   0.23
                  -0.13
             -0.23
              0.13
             0.14
            0.011
 -0.84
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
16.5688
            rho
12.631
      intercept
106.453
               Parameter Estimates

                                       95.0% Wald

      Estimate        Std.  Err.      Lower Conf. Limit

      -18.8355          18.0637             -54.2397

        5.1098          3.83743             -2.41144

        99.526          3.53402              92.5994
                                     G-405

-------
9081.17

1.31479

3.02155e+006
                    286.422

                   0.418159

                    32981.9
            4487.2

          0.457476

      1.52481e+006
                 -8508.33

                -0.478477

            -2.95559e+006
 Dose
Res .
Table of Data and Estimated Values of Interest

       N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev
                                                                     Scaled
    0
   10
   30
  100
              100
              106
              117
              122
99.5
 109
 114
 123
  10
17.9
8.97
19.9
Degrees of freedom for Test A3 vs fitted <= 0
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
10.3
  13
14.6
17.7
Model
Al
A2
A3
fitted
R
Log (likelihood)
-100.516456
-96.870820
-99.666984
-99.670736
-105.717087
# Param' s
5
8
6
6
2
AIC
211.032912
209.741641
211.333969
211.341472
215.434174
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
   0.13
 -0.563
  0.529
-0.0942
                                     G-406

-------
 Test 2:
 Test 3:
 Test 4:
 (Note:
   Test
 Are Variances Homogeneous?  (Al vs A2)
 Are variances adequately modeled?  (A2 vs. A3)
 Does the Model for the Mean Fit?  (A3 vs. fitted)
When rho=0 the results of Test 3 and Test 2 will be the same.
                     Tests of Interest
  -2*log(Likelihood Ratio)  Test df
p-value
Test 1
Test 2
Test 3
Test 4
17.6925
7.29127
5.59233
0.00750301
6
3
2
0
0.007048
0.06317
0.06104
NA
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1,
model appears to be appropriate

The p-value for Test 3 is less than .1,
different variance model
                                A non-homogeneous variance
                                You may want to consider a
NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
        Benchmark Dose Computation

Specified effect =           0.1

Risk Type        =     Relative risk

Confidence level =           0.95

             BMD =        11.6342

            BMDL =      0.975601
                                     G-407

-------
G.3.10.5.  Figure for Additional Model Presented: Hill, Unrestricted



                                  Hill Model with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
         140
         130
         120
         110
         100
          90
                     Hill
             BvlDL
BMD
                             20
40           60

     dose
                                            80
100
   16:2904/162010
                                          G-408

-------
G.3.11. Franc et al. (2001): S-D Rats, Relative Thymus Weight
G.3.11.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree0
Power
Power, unrestricted
Degrees of
freedom
2
1
1
0
0
2
2
2
1
X2 p-value
0.551
0.0001
0.972
N/A
N/A
0.252
0.252
0.252
0.510
AIC
285.890
303.995
286.698
288.696
288.696
287.456
287.456
287.456
287.131
BMD
(ng/kg-day)
6.730E+00
3.858E+02
3.559E+00
3.796E+00
4.299E+00
1.330E+01
1.330E+01
1.330E+01
5.049E-01
BMDL
(ng/kg-day)
3.627E+00
6.615E-01
1.714E+00
1.714E+00
9.311E-01
1.062E+01
1.062E+01
1.062E+01
4.411E-04
Notes







power bound hit
(power =1)
unrestricted
(power =0.388)
a Nonconstant variance model selected (p = 0.0320).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.3.11.2. Output for Selected Model: Exponential (M4)
Franc et al. (2001): S-D Rats, Relative Thymus Weight
         Exponential Model.  (Version: 1.61;   Date:  7/24/2009)
         Input Data File:  C:\l\91_Franc_2001_SD_RelThyWt_Exp_l.(d)
         Gnuplot Plotting  File:
                                             Fri Apr  16 16:30:07  2010
 Figure  5,  SD rats, relative  thymus weight
   The  form of the response  function by Model:
      Model 2:
      Model 3:
      Model 4:
      Model 5:
Y[dose]
Y[dose]
Y[dose]
Y[dose]
a * exp{sign
a * exp{sign
a * [c-(c-1)
a * [c-(c-l)
b * dose}
 (b * dose)Ad}
exp{-b  *  dose}]
exp{-(b *  dose)Ad}
    Note:  Y[dose] is the median response for  exposure = dose;
           sign = +1 for increasing trend in data;
           sign = -1 for decreasing trend.

      Model 2 is nested within Models 3 and 4.
      Model 3 is nested within Model 5.
      Model 4 is nested within Model 5.
   Dependent variable = Mean
   Independent variable =  Dose
   Data  are assumed to be  distributed: normally
   Variance Model: exp(lnalpha +rho *ln(Y[dose])

                                      G-409

-------
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  * rho)

Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 4
                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                      3.35464
                      1.08199
                          105
                    0.0424361
                     0.206726
                            1
                  Parameter Estimates

                Variable          Model 4
                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                     2.54324
                     1.25901
                     108.904
                   0.0379343
                    0.208146
                           1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
     10
     30
    100
100
91.17
51.41
22.79
83.2
47.97
43.48
29.98
   Dose

      0
     10
     30
    100
               Estimated Values of Interest

             Est Mean      Est Std     Scaled Residual
108.9
81.68
 50.3
24.61
68.33
57.01
42.02
26.79
-0.3686
 0.4706
 0.0748
 -0.192
                                  G-410

-------
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)

     Model  R:        Yij = Mu + e(i)
               Var{e(ij)} = SigmaA2
                     Model
             Likelihoods of Interest

             Log(likelihood)       DF
                                                                AIC
Al
A2
A3
R
4
-141.9834
-137.5818
-138.3482
-146.9973
-138.3488
5
8
6
2
5
293.9669
291.1637
288.6964
297.9946
286.6976
   Additive constant for all log-likelihoods =     -29.41.  This  constant
added to the
   above values gives the log-likelihood including the term that  does not
   depend on the model parameters.
                                 Explanation of Tests

   Test 1:  Does response and/or variances differ among Dose levels?  (A2  vs.
R)
   Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
   Test 3:  Are variances adequately modeled?  (A2 vs. A3)

   Test 6a: Does Model 4 fit the data?  (A3 vs  4)
     Test

     Test 1
     Test 2
     Test 3
    Test 6a
         Tests  of Interest

-2*log(Likelihood Ratio)
                                                  D. F.
p-value
18.83
8.803
1.533
0.001216
6
3
2
1
0.004459
0.03203
0.4647
0.9722
     The p-value for Test 1 is less than  .05.  There appears to be  a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

     The p-value for Test 2 is less than  .1.  A non-homogeneous

                                     G-411

-------
  variance model appears to be appropriate.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is greater than  .1.  Model 4 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 0.100000

         Risk Type = Relative deviation

  Confidence Level = 0.950000

               BMD =      3.55883

              BMDL =      1.71399
                                  G-412

-------
G.3.12. Figure for Selected Model: Exponential (M4)
 CD
 W
 o
 Q.
 CO
 CD
        150
        100
        50
                       Exponential_beta Model 4 with 0.95 Confidence Level
                      Exponential
            MDLJ |BMD
                         20
40         60
    dose
80
100
  16:3004/162010
G.3.13. Output for Additional Model Presented: Polynomial, 3-Degree
Franc et al. (2001): S-D Rats, Relative Thymus Weight
         Polynomial Model.  (Version:  2.13;   Date: 04/08/2008)
         Input  Data File: C:\l\91_Franc_2001_SD_RelThyWt_Poly_l.(d)
         Gnuplot Plotting File:  C:\l\91_Franc_2001_SD_RelThyWt_Poly_l.plt
                                            Fri Apr 16  16:30:11 2010
 Figure  5,  3D  rats,  relative thymus  weight


   The form of the response function is:

   Y[dose]  = beta 0  + beta l*dose  +  beta  2*doseA2 +  ...
   Dependent  variable = Mean
   Independent  variable = Dose
   The polynomial coefficients are  restricted to be negative
                                      G-413

-------
   The variance is to be modeled as Var(i)  = exp(lalpha + log(mean(i))  * rho)

   Total number of dose groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =       8.0075
                            rho =
                         beta_0 =
                         beta_l =
                         beta 2 =
                             _
                         beta 3 =
         0
       100
 -0.352259
-0.0585481
         0
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -beta_2    -beta_3
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

lalpha
rho
beta 0
beta 1
lalpha
1
-0.99
0.031
-0.016
rho
-0.99
1
-0.034
0.022
beta 0
0.031
-0.034
1
-0.84
beta 1
-0.016
0.022
-0.84
1
                                 Parameter Estimates
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
6.33243
            rho
2.01271
         beta_0
116.774
         beta_l
-0.331634
         beta_2
         beta 3
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
Estimate
2.92328
1.18295
89.841
-0.675682
0
0
Std. Err.
1.7394
0.423359
13.7418
0.175538
NA
NA
                     95.0% Wald

                  Lower Conf. Limit

                        -0.485884

                         0.353177

                          62.9076

                         -1.01973
                                     G-414

-------
     has no standard error.
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
    0
   10
   30
  100
 100
91.2
51.4
22.8
83.1
69.6
22.3
83.2
  48
43.5
  30
61.7
58.9
  53
  27
 0.466
 0.388
-0.968
0.0543
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
     Log(likelihood)
      -141.983433
      -137.581833
      -138.348184
      -139.728204
      -146.997301
          # Param's
                5
                8
                6
                4
                2
            AIC
         293.966865
         291.163667
         288.696368
         287.456407
         297.994602
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
                                     G-415

-------
   Test    -2*log(Likelihood Ratio)  Test df        p-value
Test 1
Test 2
Test 3
Test 4
18.8309
8.8032
1.5327
2.76004
6
3
2
2
0.004459
0.03203
0.4647
0.2516
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is greater than  .1.  The model chosen seems
to adequately describe the data
             Benchmark Dose Computation

Specified effect =           0.1

Risk Type        =     Relative risk

Confidence level =          0.95

             BMD =        13.2963


            BMDL =        10.6163
                                     G-416

-------
G.3.13.1.  Figure for Additional Model Presented: Polynomial, 3-Degree


                              Polynomial Model with 0.95 Confidence Level
 o
 Q.
 CD
 CD
         150
         100
          50
                        Polynomial
                  BMDL
BMD
                            20
40          60

     dose
                                         80
100
   16:3004/162010
                                          G-417

-------
G.3.14. Franc et al. (2001): L-E Rats, Relative Thymus Weight
G.3.14.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Power
Power, unrestricted
Degrees of
freedom
2
2
1
0
0
2
2
2
1
X2 p-value
0.394
0.394
0.317
N/A
N/A
0.236
0.236
0.236
0.175
AIC
301.666
301.666
302.808
303.805
303.805
302.690
302.690
302.690
303.643
BMD
(ng/kg-day)
6.406E+00
6.406E+00
3.520E+00
1.280E+01
1.195E+01
1.429E+01
1.429E+01
1.429E+01
1.297E+00
BMDL
(ng/kg-day)
2.122E+00
2.122E+00
1.067E-K)0
1.450E+00
9.965E-01
9.087E+00
9.087E+00
9.087E+00
2.703E-08
Notes

power hit bound (d=\)





power bound hit
(power =1)
unrestricted
(power = 0.454)
a Constant variance model selected (p = 0.5063).
b Best-fitting model, BMDS output presented in this appendix.
G.3.14.2. Output for Selected Model: Exponential (M4)
Franc et al. (2001): L-E Rats, Relative Thymus Weight
         Exponential Model.  (Version:  1.61;   Date: 7/24/2009)
         Input Data File: C:\l\92_Franc_2001_LE_RelThyWt_ExpCV_l.(d)
         Gnuplot Plotting File:
                                            Fri Apr 16  16:30:58  2010
 Figure  5,  L-E rats, relative thymus  weight
   The  form of the response function by Model:
      Model 2:     Y[dose] = a  *  exp{sign * b * dose}
      Model 3:     Y[dose] = a  *  exp{sign * (b * dose)Ad}
      Model 4:     Y[dose] = a  *  [c-(c-l)  * exp{-b * dose}]
      Model 5:     Y[dose] = a  *  [c-(c-l)  * exp{-(b * dose)Ad}]

    Note:  Y[dose]  is the median response for exposure =  dose;
           sign = +1 for increasing  trend in data;
           sign = -1 for decreasing  trend.

      Model 2 is nested within  Models 3 and 4.
      Model 3 is nested within  Model 5.
      Model 4 is nested within  Model 5.
   Dependent variable = Mean
   Independent variable = Dose
   Data  are assumed to be distributed:  normally
   Variance Model:  exp(lnalpha  +rho  *ln(Y[dose])

                                      G-418

-------
rho is set to 0.
A constant variance model is fit.

Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values
               Variable

                 Inalpha
                     rho(S)
                       a
                       b
                       c
                       d
                 Model 4
                       8.1814
                            0
                          105
                    0.0413945
                       0.3173
                            1
   (S) = Specified
                  Parameter Estimates
                Variable

                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                  Model 4

                     8.21275
                           0
                      106.57
                   0.0425967
                     0.28189
                           1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
     10
     30
    100
100
95.41
38.69
34.98
54.72
70.46
47.97
77.96
   Dose

      0
     10
     30
    100
               Estimated Values of Interest

             Est Mean      Est Std     Scaled Residual
106.6
80.03
51.36
31.12
60.73
60.73
60.73
60.73
 -0.306
 0.7164
-0.5902
 0.1798
                                  G-419

-------
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)

     Model  R:        Yij = Mu + e(i)
               Var{e (ij)  } = SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
4
-146.9024
-145.7361
-146.9024
-150.6049
-147.404
5
8
5
2
4
303.8049
307.4723
303.8049
305.2098
302.8079
   Additive constant for all log-likelihoods =     -29.41.  This  constant
added to the
   above values gives the log-likelihood including the term that  does  not
   depend on the model parameters.
R)
                              Explanation of Tests

Test 1:  Does response and/or variances differ among Dose levels?  (A2  vs.

Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)

Test 6a: Does Model 4 fit the data?  (A3 vs  4)
     Test

     Test 1
     Test 2
     Test 3
    Test 6a
                         Tests of Interest

                -2*log(Likelihood Ratio)
                                                  D. F.
9.738
2.333
2.333
1.003
6
3
3
1
p-value
                                                                 0.1362
                                                                 0.5063
                                                                 0.5063
                                                                 0.3166
     The p-value for Test 1 is greater than  .05.  There may not be  a
     diffence between responses and/or variances among the dose levels

                                     G-420

-------
  Modelling the data with a dose/response curve may not be appropriate.

  The p-value for Test 2 is greater than .1.  A homogeneous
  variance model appears to be appropriate here.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is greater than .1.  Model 4 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 0.100000

         Risk Type = Relative deviation

  Confidence Level = 0.950000

               BMD =      3.52038

              BMDL =      1.06729
                                 G-421

-------
G.3.14.3. Figure for Selected Model: Exponential (M4)



                           Exponential_beta Model 4 with 0.95 Confidence Level
         150
         100
 o
 Q.
 o:

 c
 (0
 OJ
50
                          Exponential
            BMDL
         BMD
                             20
                                40          60

                                     dose
   16:3004/162010
80          100
                                         G-422

-------
G.3.15. Franc et al. (2001): H/W Rats, Relative Thymus Weight
G.3.15.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)b
Exponential (M3)
Exponential (M4)c
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Power
Power, unrestricted
Degrees of
freedom
2
2
1
0
0
2
2
2
1
X2 p-value
0.682
0.682
0.512
N/A
N/A
0.543
0.543
0.543
0.381
AIC
261.694
261.694
263.358
264.927
264.927
262.148
262.148
262.148
263.694
BMD
(ng/kg-day)
1.366E+01
1.366E+01
8.820E+00
1.776E+01
1.701E+01
1.919E+01
1.919E+01
1.919E+01
8.127E+00
BMDL
(ng/kg-day)
8.014E+00
8.014E+00
3.219E-H)0
3.500E+00
2.729E+00
1.373E+01
1.373E+01
1.373E+01
1.406E-01
Notes

power hit bound (d = 1)





power bound hit
(power =1)
unrestricted
(power = 0.665)
a Constant variance model selected (p = 0.4331).
b Alternate model, BMDS output also presented in this appendix.
0 Best-fitting model, BMDS output presented in this appendix.
G.3.15.2. Output for Selected Model: Exponential (M2)
Franc et al. (2001): H/W Rats, Relative Thymus Weight
         Exponential Model.  (Version: 1.61;   Date:  7/24/2009)
         Input Data File:  C:\l\93_Franc_2001_HW_RelThyWt_ExpCV_l.(d)
         Gnuplot Plotting  File:
                                             Fri  Apr 16 16:31:40  2010
 Figure  5,  H/W rats, relative thymus weight
   The  form of the response  function by Model:
      Model 2:     Y[dose] = a * exp{sign  * b  *  dose}
      Model 3:     Y[dose] = a * exp{sign  *  (b  * dose)Ad}
      Model 4:     Y[dose] = a * [c-(c-l)  * exp{-b * dose}]
      Model 5:     Y[dose] = a * [c-(c-l)  * exp{-(b * dose)Ad}]

    Note:  Y[dose] is the median response for exposure = dose;
           sign = +1 for increasing trend in data;
           sign = -1 for decreasing trend.

      Model 2 is nested within Models 3 and 4.
      Model 3 is nested within Model 5.
      Model 4 is nested within Model 5.
   Dependent variable = Mean
   Independent variable =  Dose
   Data  are assumed to be  distributed: normally

                                       G-423

-------
Variance Model: exp(lnalpha +rho *ln(Y[dose]))
rho is set to 0.
A constant variance model is fit.

Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to:  le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 2

                 Inalpha              6.96647
                     rho(S)                 0
                       a              59.5084
                       b           0.00715458
                       c                    0
                       d                    1

   (S) = Specified
                  Parameter Estimates

                Variable          Model 2

                 Inalpha           6.99043
                     rho                 0
                       a           99.7761
                       b        0.00771341
                       c                 0
                       d                 1


         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
100
97.53
71.02
49.29
35.98
32.98
23.99
43.48
      0
     10
     30
    100
               Estimated Values of Interest

   Dose      Est Mean      Est Std     Scaled Residual
      0         99.78        32.96          0.01921
     10         92.37        32.96           0.4426
     30         79.16        32.96          -0.6986

                                  G-424

-------
       100
                46.14
                    32.96
0.271
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)

     Model  R:        Yij = Mu + e(i)
               Var{e(ij)} = SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
2
-127.4636
-126.0925
-127.4636
-132.935
-127.8469
5
8
5
2
3
264.9271
268.185
264.9271
269.87
261.6939
   Additive constant for all log-likelihoods =     -29.41.  This  constant
added to the
   above values gives the log-likelihood including the term that  does not
   depend on the model parameters.
R)
Test 1:

Test 2:
Test 3:
Test 4:
                     Explanation of Tests

Does response and/or variances differ among Dose  levels?  (A2  vs.

Are Variances Homogeneous?  (A2 vs. Al)
Are variances adequately modeled?  (A2 vs. A3)
Does Model 2 fit the data?  (A3 vs. 2)
     Test

     Test 1
     Test 2
     Test 3
     Test 4
                         Tests of Interest

                -2*log(Likelihood Ratio)
                                                  D. F.
13.69
2.742
2.742
0.7668
6
3
3
2
                                                    p-value
                                                                0.03336
                                                                 0.4331
                                                                 0.4331
                                                                 0.6815
     The p-value for Test 1 is less than  .05.  There appears to be  a
     difference between response and/or variances among the dose

                                     G-425

-------
  levels, it seems appropriate to model the data.

  The p-value for Test 2 is greater than .1.  A homogeneous
  variance model appears to be appropriate here.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 4 is greater than .1.  Model 2 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 0.100000

         Risk Type = Relative deviation

  Confidence Level = 0.950000

               BMD =      13.6594

              BMDL =      8.01373
                                  G-426

-------
G.3.15.3. Figure for Selected Model: Exponential (M2)

                         Exponential_beta Model 2 with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 OJ
        140
        120
        100
         80
60
         40
         20
                        Exponential
                BMDL
               BMD
                           20
                             40
60
80
100
                                           dose
   16:31 04/162010
G.3.15.4. Output for Additional Model Presented: Exponential (M4)
Franc et al. (2001): H/W Rats, Relative Thymus Weight
        Exponential Model.  (Version:  1.61;   Date: 7/24/2009)
        Input  Data File: C:\l\93_Franc_2001_HW_RelThyWt_ExpCV_l.(d)
        Gnuplot Plotting File:
                                            Fri Apr 16  16:31:40 2010
 Figure  5,  H/W rats,  relative thymus  weight
   The  form of the response function  by Model:
      Model 2:      Y[dose] = a  *  exp{sign * b * dose}
      Model 3:      Y[dose] = a  *  exp{sign *  (b * dose)Ad}
      Model 4:      Y[dose] = a  *  [c-(c-l)  * exp{-b * dose}]
      Model 5:      Y[dose] = a  *  [c-(c-l)  * exp{-(b * dose)Ad}]

    Note:  Y[dose]  is the median response for exposure =  dose;

                                      G-427

-------
       sign = +1 for increasing trend in data;
       sign = -1 for decreasing trend.

   Model 2 is nested within Models 3 and 4.
   Model 3 is nested within Model 5.
   Model 4 is nested within Model 5.
Dependent variable = Mean
Independent variable = Dose
Data are assumed to be distributed: normally
Variance Model: exp(lnalpha +rho *ln(Y[dose]))
rho is set to 0.
A constant variance model is fit.

Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to:  le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 4
                 Inalpha
                     rho(S)
                       a
                       b
                       c
                       d
                6.96647
                      0
                    105
                0.03169
               0.447105
                      1
   (S) = Specified
                  Parameter Estimates

                Variable          Model 4
                 Inalpha
                     rho
                       a
                       b
                       c
                       d
               6.97993
                     0
               103.091
               0.02048
              0.394904
                     1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
     10
  100
97.53
  35.98
  32.98

G-428

-------
        30
       100
71.02
49.29
23.99
43.48
      Dose
                  Estimated Values of Interest

                Est Mean      Est Std     Scaled Residual
0
10
30
100
103.1
91.54
74.46
48.76
32.78
32.78
32.78
32.78
-0.2667
0.5166
-0.2961
0.04621
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i))  *  rho)

     Model  R:        Yij = Mu + e(i)
               Var{e (ij)  } = SigmaA2
                     Model
       Likelihoods of Interest

       Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
4
-127.4636
-126.0925
-127.4636
-132.935
-127.6789
5
8
5
2
4
264.9271
268.185
264.9271
269.87
263.3577
   Additive constant for all log-likelihoods =     -29.41.   This  constant
added to the
   above values gives the log-likelihood including the  term  that  does  not
   depend on the model parameters.
                                 Explanation of Tests

   Test 1:  Does response and/or variances differ among Dose  levels?  (A2  vs.
R)
   Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
   Test 3:  Are variances adequately modeled?  (A2 vs. A3)

   Test 6a: Does Model 4 fit the data?  (A3 vs  4)
                                     G-429

-------
                         Tests of Interest

  Test          -2*log(Likelihood Ratio)        D. F.         p-value
  Test 1                         13.69           6             0.03336
  Test 2                         2.742           3              0.4331
  Test 3                         2.742           3              0.4331
 Test 6a                        0.4306           1              0.5117
  The p-value for Test 1 is less than .05.  There appears to be a
  difference between response and/or variances among the dose
  levels,  it seems appropriate to model the data.

  The p-value for Test 2 is greater than  .1.  A homogeneous
  variance model appears to be appropriate here.

  The p-value for Test 3 is greater than  .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is greater than .1.  Model 4 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 0.100000

         Risk Type = Relative deviation

  Confidence Level = 0.950000

               BMD =      8.82023

              BMDL =      3.21928
                                 G-430

-------
G.3.15.5. Figure for Additional Model Presented: Exponential (M4)



                           Exponential_beta Model 4 with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
         140
         120
         100
          80
60
          40
          20
              BMDL
                          Exponential
             BMD
                             20
40           60

     dose
                                                          80
100
   16:31 04/162010
                                          G-431

-------
G.3.16. Hojo et al. (2002): DRL Reinforce per Minute
G.3.16.1. Summary Table of BMDS Modeling Results
Model3
Hill
Linear1"
Polynomial, 3 -degree
Power
Power, unrestricted
Exponential (M2)
Exponential (M3)
Exponential (M4)°
Exponential (M5)
Degrees of
freedom
0
2
2
2
1
2
2
1
0
X2 p-value
N/A
0.008
0.008
0.008
0.025
0.006
0.006
0.062
N/A
AIC
6.465
9.552
9.552
9.552
6.780
9.894
9.894
5.241
6.465
BMD
(ng/kg-day)
2.060E+01
2.677E+02
2.677E+02
2.677E+02
2.187E+00
3.043E+02
3.043E+02
1.734E+01
2.140E+01
BMDL
(ng/kg-day)
1.713E-05
1.100E-K)2
1.100E+02
1.100E+02
4.612E-08
1.505E+02
1.505E+02
3.827E-02
1.240E-05
Notes



power bound hit
(power =1)
unrestricted
(power = 0.089)

power hit bound (d=\)


a Constant variance model selected (p = 0.4321).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.3.16.2. Output for Selected Model: Linear
Hojo et al. (2002): DRL Reinforce Per Minute
         Polynomial Model.  (Version: 2.13;   Date:  04/08/2008)
         Input Data File:  C:\l\20_Hojo_2002_DRLrein_LinearCV_l.(d)
         Gnuplot Plotting  File:   C:\l\20_Hojo_2002_DRLrein_LinearCV_l.plt
                                            Tue  Feb 16 17:29:42  2010
 Table  5


   The  form of the response  function is:

   Y[dose]  = beta 0 + beta  l*dose + beta 2*doseA2 + ...
   Dependent variable = Mean
   Independent variable =  Dose
   rho  is  set to 0
   Signs  of the polynomial coefficients are  not  restricted
   A  constant variance model is fit

   Total  number of dose groups = 4
   Total  number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set  to:  le-008
   Parameter Convergence has been set to:  le-008
                                       G-432

-------
                  Default Initial Parameter Values
                          alpha =     0.337763
                            rho =            0   Specified
                         beta_0 =       -0.404
                         beta 1 =   0.00249615
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

                  alpha       beta 0       beta 1

     alpha            1    -1.4e-008     2.2e-008

    beta_0    -1.4e-008            1        -0.69

    beta 1     2.2e-008        -0.69            1
Confidence Interval
       Variable
Upper Conf. Limit
          alpha
0.69919
         beta_0
0.017352
         beta_l
0.00660807
           Parameter Estimates

                                   95.0% Wald

  Estimate        Std.  Err.      Lower Conf.  Limit

  0.435671         0.134451             0.172152

 -0.372098         0.198702            -0.761547

0.00246548       0.00211361          -0.00167711
     Table of Data and Estimated Values of Interest
Dose
Res .
0
20
60
180
N
5
5
6
5
Obs Mean
-0.814
-0.364
0.374
-0.163
Est Mean
-0.372
-0.323
-0.224
0.0717
Obs Std Dev
0.448
0.821
0.54
0.443
Est Std Dev
0.66
0.66
0.66
0.66
Scaled
-1.5
-0.14
2.22
-0.795
 Model Descriptions for likelihoods calculated
                                    G-433

-------
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses  any fixed variance parameters that
     were specified by the user

 Model  R:          Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
                3.115550
                4.489557
                3.115550
               -1.775882
               -2.435087
# Param' s
5
8
5
3
2
AIC
3.768900
7.020886
3.768900
9.551763
8.870174
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
-2*log(Likelihood Ratio)  Test df
            13.8493
            2.74801
            2.74801
            9.78286
6
3
3
2
   p-value

 0.03137
  0.4321
  0.4321
0.007511
The p-value for Test 1 is less than  .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than  .1.
model appears to be appropriate here
                                 A homogeneous variance
The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than  .1.  You may want to try a different
model
                                     G-434

-------
              Benchmark  Dose Computation



Specified effect =              1



Risk Type        =      Estimated  standard deviations from the  control mean



Confidence level =           0.95



              BMD =         267.718
             BMDL =
                            110.032
G.3.16.3. Figure for Selected Model: Linear



                               Linear Model with 0.95 Confidence Level
          1
         0.5
 o
 Q.
 o:

 c
 (0
 HI
        -0.5
          -1
        -1.5
   17:2902/162010
                    Linear
                                    BMDL
                                       BM1D
                           50
100
150
200
250
                                            dose
                                       G-435

-------
G.3.16.4. Output for Additional Model Presented: Exponential (M4)
Hojo et al. (2002): DRL Reinforce Per Minute
        Exponential Model.  (Version:  1.61;   Date:  7/24/2009)
        Input Data File: C:\l\21_Hojo_2002_DRLrein_ExpCV_l.(d)
        Gnuplot Plotting File:
                                           Tue  Feb  16  17:30:21 2010
 Table 5, values adjusted by a constant  to  allow exponential model
   The form of the response function by Model:
Model 2:
Model 3:
Model 4:
Model 5:
                   Y[dose]
                   Y[dose]
                   Y[dose]
                   Y[dose]
a * exp{sign
a * exp{sign
a * [c-(c-1)
a * [c-(c-l)
b * dose}
 (b * dose)Ad}
exp{-b * dose}]
exp{-(b * dose)Ad}
    Note: Y[dose] is the median response  for  exposure  = dose;
          sign = +1 for increasing trend  in data;
          sign = -1 for decreasing trend.

      Model 2 is nested within Models  3 and 4.
      Model 3 is nested within Model 5.
      Model 4 is nested within Model 5.
   Dependent variable = Mean
   Independent variable = Dose
   Data are assumed to be distributed:  normally
   Variance Model: exp(lnalpha +rho  *ln(Y[dose]))
   rho is set to 0.
   A constant variance model is  fit.

   Total number of dose groups =  4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence  has been  set to:  le-008
   Parameter Convergence has been set  to:  le-008

   MLE solution provided: Exact
                  Initial Parameter Values

                  Variable          Model 4
                    Inalpha
                        rho(S)
                          a
                          b
                          c
                          d
                                  -1.29672
                                          0
                                     0.0817
                                 0.00880867
                                    16.3733
                                          1
                                     G-436

-------
  (S)  = Specified
                  Parameter Estimates

                Variable          Model 4
                 Inalpha            -1.13136
                     rho                   0
                       a           0.0542868
                       b           0.0525016
                       c             18.5072
                       d                   1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0      5        0.086        0.448
     20      5        0.536        0.821
     60      6        1.274         0.54
    180      5        0.737        0.443
               Estimated Values of Interest

   Dose      Est Mean      Est Std     Scaled Residual
      0       0.05429        0.568           0.1249
     20        0.6721        0.568          -0.5359
     60         0.964        0.568             1.337
    180         1.005        0.568           -1.054
Other models for which likelihoods are calculated:

  Model Al:        Yij = Mu(i) + e(ij)
            Var{e(ij)} = SigmaA2

  Model A2:        Yij = Mu(i) + e(ij)
            Var{e(ij)} = Sigma(i)A2

  Model A3:        Yij = Mu(i) + e(ij)
            Var{e(ij)} = exp(lalpha + log(mean(i))  *  rho)

  Model  R:        Yij = Mu + e(i)
            Var{e(ij)} = SigmaA2
                             Likelihoods of Interest

                  Model      Log(likelihood)      DF         AIC

                     Al         3.11555             5         3.7689

                                  G-437

-------
                        A2        4.489557            8      7.020886
                        A3         3.11555            5        3.7689
                         R       -2.435087            2      8.870174
                         4        1.379312            4      5.241376
   Additive constant for all log-likelihoods =      -19.3.  This constant
added to the
   above values gives the log-likelihood including the term that does not
   depend on the model parameters.
                                 Explanation of Tests

   Test 1:  Does response and/or variances differ among Dose levels?  (A2 vs.
R)
   Test 2:  Are Variances Homogeneous? (A2 vs. Al)
   Test 3:  Are variances adequately modeled?  (A2 vs. A3)

   Test 6a: Does Model 4 fit the data? (A3 vs 4)


                            Tests of Interest

     Test          -2*log(Likelihood Ratio)       D. F.         p-value
     Test 1                         13.85           6             0.03137
     Test 2                         2.748           3              0.4321
     Test 3                         2.748           3              0.4321
    Test 6a                         3.472           1              0.0624
     The p-value for Test 1 is less than .05.  There appears to be a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

     The p-value for Test 2 is greater than  .1.  A homogeneous
     variance model appears to be appropriate here.

     The p-value for Test 3 is greater than  .1.  The modeled
     variance appears to be appropriate here.

     The p-value for Test 6a is less than .1.  Model 4 may not adequately
     describe the data; you may want to consider another model.
   Benchmark Dose Computations:

     Specified Effect = 1.000000

            Risk Type = Estimated standard deviations from control

     Confidence Level = 0.950000

                  BMD =      17.3391

                 BMDL =    0.0382689

                                     G-438

-------
G.3.16.5. Figure for Additional Model Presented: Exponential (M4)



                           Exponential_beta Model 4 with 0.95 Confidence Level
 01
 to

 o
 CL
 en
 0)
 or

 £=
 CO
 01
         1.5
0.5
         -0.5
            BMDL
                 0
   17:3002/162010
                          Exponential
              BMD
              20     40     60     80     100    120    140     160     180

                                      dose
                                           G-439

-------
G.3.17. Hojo et al. (2002): DRL Response per Minute
G.3.17.1. Summary Table of BMDS Modeling Results
Model3
Hill
Linear
Polynomial, 3 -degree
Power
Power, unrestricted
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Degrees of
freedom
0
2
2
2
2
2
2
1
0
X2 p-value
N/A
0.004
0.004
0.004
0.741
0.568
0.568
0.479
N/A
AIC
126.353
132.825
132.825
132.825
122.455
122.985
122.985
124.356
126.353
BMD
(ng/kg-day)
1.646E+01
2.067E+02
2.067E+02
2.067E+02
1.800E+04
6.184E+00
6.184E+00
4.775E+00
1.118E+01
BMDL
(ng/kg-day)
1.800E-13
9.757E+01
9.757E+01
9.757E+01
error
error
error
2.704E-01
2.127E-01
Notes



power bound hit
(power =1)
unrestricted (power = 0)

power hit bound (d=\)


a Constant variance model selected (p = 0.3004).
b Best-fitting model, BMDS output presented in this appendix.
G.3.17.2. Output for Selected Model: Exponential (M4)
Hojo et al. (2002): DRL Response Per Minute
         Exponential Model.  (Version:  1.61;   Date: 7/24/2009)
         Input Data File: C:\l\23_Hojo_2002_DRLresp_ExpCV_l.(d)
         Gnuplot Plotting File:
                                            Tue Feb 16 17:31:24  2010
 Table  5,  values adjusted by a constant  to allow exponential model
   The  form of the response function  by Model:
      Model 2:      Y[dose] = a * exp{sign *  b * dose}
      Model 3:      Y[dose] = a * exp{sign *  (b  * dose)Ad}
      Model 4:      Y[dose] = a *  [c-(c-l)  *  exp{-b * dose}]
      Model 5:      Y[dose] = a *  [c-(c-l)  *  exp{-(b * dose)Ad}]

    Note:  Y[dose]  is the median response for exposure = dose;
           sign = +1 for increasing  trend in  data;
           sign = -1 for decreasing  trend.

      Model 2  is nested within Models 3 and  4.
      Model 3  is nested within Model  5.
      Model 4  is nested within Model  5.
   Dependent  variable = Mean
   Independent variable = Dose
   Data  are assumed to be distributed:  normally
   Variance Model:  exp(lnalpha +rho  *ln(Y[dose])
   rho is  set to 0.
                                      G-440

-------
A constant variance model is fit.

Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 4

                 Inalpha              4.51689
                     rho(S)                 0
                       a              24.6362
                       b            0.0212679
                       c            0.0184785
                       d                    1

   (S) = Specified
                  Parameter Estimates

                Variable          Model 4

                 Inalpha             4.54075
                     rho                   0
                       a              23.465
                       b             0.12859
                       c            0.100615
                       d                   1


         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0      5
     20      5
     60      6
    180      5
               Estimated Values of Interest

   Dose      Est Mean      Est Std     Scaled Residual
23.46
4.013
0.478
4.594
7.986
10.96
7.194
15.23
      0         23.47        9.683       -0.0004677
     20         3.973        9.683         0.009182
     60          2.37        9.683          -0.4787
    180         2.361        9.683           0.5157


                                  G-441

-------
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) *  rho)

     Model  R:        Yij = Mu + e(i)
               Var{e(ij)} = SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
4
-57.92733
-56.09669
-57.92733
-64.49611
-58.17787
5
8
5
2
4
125.8547
128.1934
125.8547
132.9922
124.3557
   Additive constant for all log-likelihoods =      -19.3.  This  constant
added to the
   above values gives the log-likelihood including the term that  does  not
   depend on the model parameters.
R)
                              Explanation of Tests

Test 1:  Does response and/or variances differ among Dose  levels?  (A2  vs.

Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)

Test 6a: Does Model 4 fit the data?  (A3 vs  4)
     Test

     Test 1
     Test 2
     Test 3
    Test 6a
                         Tests of Interest

                 -2*log(Likelihood Ratio)
                                                  D. F.
16.8
3.661
3.661
0.5011
6
3
3
1
p-value
                                                                0.01005
                                                                 0.3004
                                                                 0.3004
                                                                  0.479
     The p-value for Test 1 is less than  .05.  There appears  to be  a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

                                     G-442

-------
  The p-value for Test 2 is greater than .1.  A homogeneous
  variance model appears to be appropriate here.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is greater than .1.  Model 4 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =      4.77493

              BMDL =     0.270447
                                 G-443

-------
G.3.17.3. Figure for Selected Model: Exponential (M4)



                          Exponential_beta Model 4 with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
        30
        20
        10
       -10
                        Exponential
          BMDL
              BMD
            0      20



17:31 02/162010

                             40     60     80     100    120    140     160     180

                                              dose
                                         G-444

-------
G.3.18. Kattainen et al. (2001): 3rd Molar Eruption, Female
G.3.18.1. Summary Table of BMDS Modeling Results
Model
Logistic
Log-logistic"
Log-probit
Probit
Multistage, 4 -degree
Log-logistic,
unrestricted13
Log-probit, unrestricted
Degrees of
freedom
3
3
3
3
o
J
2
2
x2
p-value
0.292
0.923
0.390
0.306
0.641
0.952
0.941
AIC
89.060
85.535
88.231
88.919
86.798
87.157
87.179
BMD
(ng/kg-day)
1.941E+02
4.763E+01
1.574E+02
1.858E+02
8.677E+01
2.599E+01
2.813E+01
BMDL
(ng/kg-day)
1.390E+02
2.481E-K)!
9.512E+01
1.370E+02
5.520E+01
1.730E+00
2.334E+00
Notes
negative intercept
(intercept = -1.508)
slope bound hit
(slope = 1)
slope bound hit
(slope =1)
negative intercept
(intercept =-0.927)
final B = 0
unrestricted
(slope = 0.794)
unrestricted
(slope = 0.478)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.


G.3.18.2. Output for Selected Model: Log-Logistic
Kattainen et al. (2001): 3rd Molar Eruption, Female
         Logistic Model.  (Version:  2.12; Date: 05/16/2008)
         Input Data File: C:\l\24_Katt_2001_Erup_LogLogistic_BMRl.(d)
         Gnuplot Plotting File:   C:\l\24_Katt_2001_Erup_LogLogistic_BMRl.plt
                                            Tue Feb  16  17:31:52 2010
 Figure  2
   The  form of the probability  function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
   Dependent variable = DichEff
   Independent variable = Dose
   Slope  parameter is restricted as slope >= 1

   Total  number of observations  = 5
   Total  number of records with  missing values =  0
   Maximum number of iterations  = 250
   Relative Function Convergence has been set to:  le-008
   Parameter Convergence has  been set to: le-008
                                      G-445

-------
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =       0.0625
                      intercept =       -6.063
                          slope =            1
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -slope
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

             background    intercept

background            1        -0.56

 intercept        -0.56            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background        0.0846785
*
      intercept         -6.06063
*
          slope                1
                                 Parameter Estimates
                      Std.  Err.
                 95.0% Wald

              Lower Conf.  Limit
* - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.9238
  Reduced model
0.0004142

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -40.5286         5
     -40.7674         2      0.477533      3
     -50.7341
      85.5347
1
20.411
                            P-value
                                  Goodness  of  Fit

     Dose     Est. Prob.    Expected    Observed     Size

                                     G-446
                                               Scaled
                                              Residual

-------
0.0000
30.0000
100.0000
300.0000
1000.0000
0.0847
0.1445
0.2578
0.4615
0.7254
1.355
2.457
3.867
5.538
13.782
1.000
3.000
4.000
6.000
13.000
16
17
15
12
19
-0.319
0.374
0.078
0.267
-0.402
 ChiA2 = 0.48
d.f.  =3
P-value = 0.9231
   Benchmark Dose Computation




Specified effect =            0.1




Risk Type        =      Extra risk




Confidence level =           0.95




             BMD =        47.6274




            BMDL =        24.8121
                                     G-447

-------
G.3.18.3. Figure for Selected Model: Log-Logistic

                            Log-Logistic Model with 0.95 Confidence Level
 T3
 £
 O
 c
 O
 •*=
 O
 (0
         0.8
         0.6
0.4
         0.2
          0  -
                          Log-Logistic
                           200
                              400          600
                                   dose
800
1000
   17:31 02/162010
G.3.18.4. Output for Additional Model Presented: Log-Logistic, Unrestricted
Kattainen et al. (2001): 3rd Molar Eruption, Female
         Logistic Model.  (Version: 2.12;  Date:  05/16/2008)
         Input Data File:  C:\l\24_Katt_2001_Erup_LogLogistic_U_BMRl.(d)
         Gnuplot Plotting  File:   C:\l\24_Katt_2001_Erup_LogLogistic_U_BMRl.plt
                                             Tue  Feb 16 17:31:53 2010
 Figure  2
   The  form of the probability function  is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
   Dependent variable =  DichEff
                                       G-448

-------
   Independent variable = Dose
   Slope parameter is not restricted

   Total number of observations = 5
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =       0.0625
                      intercept =     -4.71231
                          slope =     0.782659
           Asymptotic Correlation Matrix of Parameter Estimates

             background    intercept        slope

background            1        -0.48         0.39

 intercept        -0.48            1        -0.98

     slope         0.39        -0.98            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background        0.0633217
*
      intercept         -4.78282
*
          slope         0.793723
                                 Parameter Estimates
                      Std.  Err.
                 95.0% Wald

              Lower Conf.  Limit
* - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.9515
  Reduced model
0.0004142
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -40.5286         5
     -40.5783         3     0.0994416      2
     -50.7341
1
                                     G-449
20.411
                            P-value

-------
           AIC:         87.1566
                                  Goodness  of  Fit
Dose
0.0000
30.0000
100.0000
300.0000
1000.0000
Est. Prob.
0.0633
0.1670
0.2924
0.4721
0.6892
Expected
1.013
2.840
4.387
5.666
13.095
Observed
1.000
3.000
4.000
6.000
13.000
Size
16
17
15
12
19
Scaled
Residual
-0.013
0.104
-0.219
0.193
-0.047
 ChiA2 = 0.10      d.f. = 2        P-value = 0.9518







   Benchmark Dose Computation





Specified effect =            0.1





Risk Type        =      Extra risk





Confidence level =           0.95





             BMD =         25.986





            BMDL =        1.73001
                                     G-450

-------
G.3.18.5. Figure for Additional Model Presented: Log-Logistic, Unrestricted



                                Log-Logistic Model with 0.95 Confidence Level
 T3

 £
 O
 c
 O
 •*=
 O
 (0
          0.8
          0.6
0.4
          0.2
                             Log-Logistic
             BiyiDLj JBMD
                              200
                                 400          600

                                        dose
800          1000
   17:31 02/162010
                                            G-451

-------
G.3.19. Kattainen et al. (2001): 3rd Molar Length, Female
G.3.19.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hillb
Linear
Polynomial,
4-degree
Power
Hill, unrestricted0
Power, unrestricted
Degrees of
freedom
3
3
2
1
2
3
3
3
1
2
X2 p-value
0.0001
0.0001
0.0001
0.0001
0.013
0.0001
0.0001
0.0001
0.087
0.250
AIC
-122.954
-122.954
-80.747
-78.747
-151.152
-122.325
-122.325
-122.325
-154.939
-157.093
BMD
(ng/kg-day)
4.027E+02
4.027E+02
error
error
4.052E-K)0
4.659E+02
4.659E+02
4.659E+02
1.913E-02
9.098E-03
BMDL 1
(ng/kg-day) | Notes
2.366E+02 1
2 . 3 66E+02 power hit bound (d = 1 )
error 1
error
2.144E+00
2.963E+02
2.963E+02

n lower bound hit (n = 1)


2.963E+02 PO^r bound hit
(power =1)
1.928E-04 unrestricted (« = 0.197)
9.097E-03
unrestricted
(power =0.169)
a Nonconstant variance model selected (p = O.0001).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.3.19.2. Output for Selected Model: Hill
Kattainen et al. (2001): 3rd Molar Length, Female
         Hill Model.  (Version:  2.14;  Date:  06/26/2008)
         Input Data File:  C:\l\25_Katt_2001_Length_Hill_l.(d)
         Gnuplot Plotting  File:   C:\l\25_Katt_2001_Length_Hill_l.plt
                                            Tue  Feb 16 17:32:21  2010
 Figure  3 female only


   The  form of the response  function is:

   Y[dose]  = intercept +  v*doseAn/(kAn + doseAn)
   Dependent variable = Mean
   Independent variable =  Dose
   Power parameter restricted to be greater  than 1
   The  variance is to be modeled as Var(i) = exp(lalpha  + rho  *  In(mean(i)))

   Total number of dose groups = 5
   Total number of records  with missing values  = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set to:  le-008
   Parameter Convergence has been set to:  le-008
                                       G-452

-------
                  Default Initial Parameter Values
                         lalpha =     -2.37155
                            rho =            0
                      intercept =
                              v =
                              n =
                              k =
  1.85591
-0.507874
 0.826204
  27.3305
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -n
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
lalpha
lalpha
rho
intercept
V
k

-0.
-0.
0.
-0.
1
98
16
84
37
rho
-0.98
1
0.2
-0.79
0.39
intercept
-0.16
0.2
1
-0.31
-0.11

0.
-0.
-0.

-0.
V
84
79
31
1
48

-0
0
-0
-0

k
.37
.39
.11
.48
1
                                 Parameter Estimates
Confidence Interval
Variable
Upper Conf. Limit
lalpha
6.09824
rho
-9.19484
intercept
1.88597
V
-0.325818
n
k
39.4101

Estimate

3.34561

-14.3325

1.8548

-0.441166

1
24.0343


Std. Err.

1.40443

2.62129

0.0159017

0.058852

NA
7.84495

NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
                                                         95.0% Wald

                                                      Lower Conf. Limit

                                                             0.592981

                                                             -19.4701

                                                              1.82364

                                                            -0.556513


                                                              8.65852
     Table of Data and Estimated Values of Interest

                                    G-453

-------
 Dose
Res .
                 Obs Mean
                              Est Mean
                                Obs Std Dev  Est Std Dev
                                                  Scaled
    0
   30
  100
  300
 1000
16
17
15
12
19
1.86
1.58
 1.6
 1.5
1.35
1.85
1.61
 1.5
1.45
1.42
0.0661
 0.185
 0.265
 0.221
 0.515
0.0637
 0.176
 0.293
 0.378
 0.423
0.0692
-0.768
  1.28
 0.527
-0.783
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
              Log(likelihood)
                 56.758717
                 85.856450
                 84.934314
                 80.575940
                 45.373551
                       # Param's
                             6
                            10
                             7
                             5
                             2
                         AIC
                     -101.517434
                     -151.712901
                     -155.868628
                     -151.151880
                      -86.747101
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
  -2*log(Likelihood Ratio)  Test df
              80.9658
              58.1955
                      4

                  G-454
                     p-value

                    <.0001
                    <.0001

-------
   Test 3
   Test 4
1.84427
8.71675
0.6053
0.0128
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =        4 . 05231

            BMDL =       2.14357
                                     G-455

-------
G.3.19.3. Figure for Selected Model: Hill

                                Hill Model with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 OJ
1.9

1.8

1.7

1.6


1.5

1.4

1.3

1.2

1.1
                   Hill
           BMDLBMD
                           200
                              400          600
                                   dose
800
1000
   17:3202/162010
G.3.19.4. Output for Additional Model Presented: Hill, Unrestricted
Kattainen et al. (2001): 3rd Molar Length, Female
         Hill Model.  (Version:  2.14;  Date:  06/26/2008)
         Input Data File:  C:\l\25_Katt_2001_Length_Hill_U_l.(d)
         Gnuplot Plotting  File:   C:\l\25_Katt_2001_Length_Hill_U_l.plt
                                             Tue Feb 16  17:32:21 2010
 Figure  3  female only


   The form of the response function is:

   Y[dose]  = intercept  +  v*doseAn/(kAn  +  doseAn)
   Dependent variable = Mean
   Independent variable =  Dose
                                       G-456

-------
   Power parameter is not restricted
   The variance is to be modeled as Var(i) = exp(lalpha

   Total number of dose groups = 5
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                                   rho * In(mean(i))
                  Default Initial Parameter Values
                         lalpha =     -2.37155
                            rho =            0
                      intercept =      1.85591
                              v =    -0.507874
n = 0.826204
k = 27.3305
Asymptotic Correlation Matrix of Parameter
lalpha rho intercept
k
lalpha 1 -0.98 -0.18
-0.011
rho -0.98 1 0.22
0.011
intercept -0.18 0.22 1
0.0019
v 0.18 -0.18 -0.025
-0.96
n -0.28 0.29 -0.059
-0.71
k -0.011 0.011 0.0019


Estimates
v
0.18
-0.18

-0.025
1

0.51

-0.96
                                                                      -0.28
                                                                       0.29
                                                                     -0.059
                                                                       0.51
                                                                      -0.71
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
6.00607
            rho
-8.82777
      intercept
1.88704
         Parameter Estimates

                                 95.0% Wald

Estimate        Std.  Err.     Lower Conf. Limit

 3.21882           1.4221            0.431563

-14.0862          2.68292            -19.3446

 1.85564        0.0160224             1.82424
                                     G-457

-------
3.19148

0.29479

3.34593e+007
    -2.48572

    0.196925

1.92967e+006
     2.89658

   0.0499318

1.60869e+007
  -8.16291

 0.0990606

-2.96e+007
     Table of Data and Estimated Values of Interest
Dose
Res .
0
30
100
300
1000
N
16
17
15
12
19
Obs
1.
1.
1
1
1.
Mean
86
58
.6
.5
35
Est
1.
1
1.
1.
1
Mean
86
.6
54
48
.4
Obs
0.
0
0
0
0
Std Dev
0661
.185
.265
.221
.515
Est
0.

0
0
0
Std Dev
0643
0.18
.234
.316
.471
Seal
0.
-0
0
0
-0
ed
0164
.598
.857
.259
.466
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
Model
Al
A2
A3
fitted
R
Log (likelihood)
56.758717
85.856450
84.934314
83.469680
45.373551
# Param's
6
10
7
6
2
AIC
-101.517434
-151.712901
-155.868628
-154.939361
-86.747101
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)

                                     G-458

-------
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.
                     Tests of Interest
           -2*log(Likelihood Ratio)  Test df
p-value
80.9658
58.1955
1.84427
2.92927
8
4
3
1
<.0001
<.0001
0.6053
0.08699
Test

Test 1
Test 2
Test 3
Test 4
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =      0.0191282

            BMDL =     0.0001928
                                     G-459

-------
G.3.19.5. Figure for Additional Model Presented: Hill, Unrestricted



                                   Hill Model with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
1.9




1.8




1.7




1.6




1.5




1.4




1.3




1.2




1.1
                     Hill
                 UBI\
   BMDLBMD
                              200
                                  400          600

                                        dose
800
1000
   17:3202/162010
                                            G-460

-------
G.3.20. Keller et al. (2007): Missing Mandibular Molars, CBA J
G.3.20.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probit
Multistage, l-degreea
Multistage, 2 -degree
Multistage, 3 -degree
Probit
Weibull
Degrees of
freedom
1
2
1
1
3
1
1
2
1
X2 p-value
0.105
0.320
0.105
0.105
0.276
0.126
0.141
0.325
0.108
AIC
52.490
50.095
52.524
52.524
49.409
51.515
51.222
50.032
52.216
BMD
(ng/kg-day)
7.293E+01
7.168E+01
9.278E+01
8.849E+01
2.778E-K)!
4.619E+01
4.253E+01
6.848E+01
6.079E+01
BMDL
(ng/kg-day)
2.027E+01
5.142E+01
5.273E+01
5.297E+01
1.884E+01
2.214E+01
2.212E+01
4.775E+01
2.078E+01
Notes









a Best-fitting model, BMDS output presented in this appendix.


G.3.20.2. Output for Selected Model: Multistage, 1-Degree
Keller et al. (2007): Missing Mandibular Molars, CBA J
        Multistage Model. (Version: 3.0;   Date:  05/16/2008)
        Input  Data File: C:\l\26_Keller_2007_Molars_Multil_l.(d)
        Gnuplot Plotting File:  C:\l\26_Keller_2007_Molars_Multil_l.plt
                                            Tue  Feb 16 17:32:56 2010
 Table  1  using  mandibular molars only
   The  form of  the probability function  is:

   P[response]  = background +  (1-background)*[1-EXP(
                  -betal*doseAl)]

   The  parameter betas are restricted to be  positive
   Dependent  variable = DichEff
   Independent  variable = Dose

 Total number of observations = 4
 Total number of records with missing values
 Total number of parameters in model =  2
 Total number of specified parameters = 0
 Degree  of  polynomial = 1
= 0
 Maximum  number of iterations = 250
 Relative Function Convergence has been  set  to:  le-008
 Parameter Convergence has been set to:  le-008
                                      G-461

-------
                  Default Initial Parameter Values
                     Background =            0
                        Beta(l) = 1.02909e+017
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -Background
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
                Beta(l)
   Beta(l)
                                 Parameter Estimates
Confidence Interval
       Variable
Upper Conf. Limit
     Background
*
        Beta (1)
      Estimate

             0

    0.00379264
     Std.  Err.
         95.0% Wald

      Lower Conf.  Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.2358
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -21.5798         4
     -23.7044         1       4.24924      3
      -71.326

      49.4088
     1
99.4926
3
                                 P-value
<.0001
                                  Goodness  of  Fit

Dose
0.0000
10.0000
100.0000
1000.0000

Est. Prob.
0.0000
0.0372
0.3156
0.9775

Expected
0.000
0.856
9.153
29.324

Observed
0.000
2.000
6.000
30.000

Size
29
23
29
30
Scaled
Residual
0.000
1.260
-1.260
0.832
 ChiA2 = 3.87
 d.f. = 3
P-value = 0.2762
                                     G-462

-------
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        27.7803

            BMDL =        18.8447

            BMDU =        41.7256

Taken together,  (18.8447, 41.7256) is a 90     % two-sided confidence
interval for the BMD
                                     G-463

-------
G.3.20.3. Figure for Selected Model: Multistage, 1-Degree

                              Multistage Model with 0.95 Confidence Level
         0.8
         0.6
 1       0.4
         0.2
             EiMDL
   17:3202/162010
                            200
400          600
      dose
800
1000
                                             G-464

-------
G.3.21. Kociba et al. (1918): Urinary Coproporphyrin, Females
G.3.21.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Power
Power, unrestricted
Degrees of
freedom
2
2
1
0
0
2
2
2
1
X2 p-value
0.0001
0.0001
0.040
N/A
N/A
0.0001
0.0001
0.0001
0.001
AIC
84.006
84.006
70.556
69.092
69.047
83.713
83.713
83.713
78.260
BMD
(ng/kg-day)
7.054E+01
7.054E+01
1.625E+00
3.128E+00
6.677E+00
6.195E+01
6.195E+01
6.195E+01
7.808E-01
BMDL
(ng/kg-day)
4.341E+01
4.341E+01
7.300E-01
1.024E+00
error
3.112E+01
3.112E+01
3.112E+01
1.693E-08
Notes

power hit bound (d=l)





power bound hit
(power =1)
unrestricted
(power = 0.306)
a Nonconstant variance model selected (p = 0.0298).
b Best-fitting model, BMDS output presented in this appendix.
G.3.21.2. Output for Selected Model: Exponential (M4)
Kociba et al. (1978): Urinary Coproporphyrin, Females
         Exponential Model.  (Version:  1.61;   Date: 7/24/2009)
         Input Data File: C:\l\29_Kociba_1978_Copro_Exp_l.(d)
         Gnuplot Plotting File:
                                            Tue Feb 16  17:34:45  2010
 Table2-UrinaryCoproporphyrin
   The  form of the response function by Model:
      Model 2:     Y[dose] = a  *  exp{sign * b * dose}
      Model 3:     Y[dose] = a  *  exp{sign * (b * dose)Ad}
      Model 4:     Y[dose] = a  *  [c-(c-l)  * exp{-b * dose}]
      Model 5:     Y[dose] = a  *  [c-(c-l)  * exp{-(b * dose)Ad}]

    Note:  Y[dose]  is the median response for exposure =  dose;
           sign = +1 for increasing  trend in data;
           sign = -1 for decreasing  trend.

      Model 2 is nested within  Models 3 and 4.
      Model 3 is nested within  Model 5.
      Model 4 is nested within  Model 5.
   Dependent variable = Mean
   Independent variable = Dose
   Data  are assumed to be distributed:  normally
   Variance Model:  exp(lnalpha  +rho  *ln(Y[dose])

                                      G-465

-------
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  * rho)

Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 4

                 Inalpha             -5.58269
                     rho              2.98472
                       a                 8.17
                       b            0.0259469
                       c              2.23623
                       d                    1
                  Parameter Estimates

                Variable          Model 4

                 Inalpha            -4.94473
                     rho             2.76088
                       a             8.93039
                       b            0.136554
                       c              1.9753
                       d                   1


         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0      5
      1      5
     10      5
    100      5
               Estimated Values of Interest

   Dose      Est Mean      Est Std     Scaled Residual
9.8
8.6
16.4
17.4
1.3
2
4.7
4
      0          8.93        1.733            1.122
      1         10.04        2.038           -1.582
     10         15.42        3.683           0.5967
    100         17.64        4.436          -0.1211
                                  G-466

-------
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)

     Model  R:        Yij = Mu + e(i)
               Var{e(ij)} = SigmaA2
                     Model
             Likelihoods of Interest

             Log(likelihood)       DF
                                                                AIC
Al
A2
A3
R
4
-31.69739
-27.21541
-28.16434
-41.73188
-30.27804
5
8
6
2
5
73.39478
70.43081
68.32868
87.46376
70.55608
   Additive constant for all log-likelihoods =     -18.38.  This  constant
added to the
   above values gives the log-likelihood including the term that  does not
   depend on the model parameters.
                                 Explanation of Tests

   Test 1:  Does response and/or variances differ among Dose levels?  (A2  vs.
R)
   Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
   Test 3:  Are variances adequately modeled?  (A2 vs. A3)

   Test 6a: Does Model 4 fit the data?  (A3 vs  4)
     Test

     Test 1
     Test 2
     Test 3
    Test 6a
         Tests  of Interest

-2*log(Likelihood Ratio)
                                                  D. F.
p-value
29.03
8.964
1.898
4.227
6
3
2
1
< 0.0001
0.02977
0.3872
0.03978
     The p-value for Test 1 is less than  .05.  There appears to be  a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

     The p-value for Test 2 is less than  .1.  A non-homogeneous

                                     G-467

-------
  variance model appears to be appropriate.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is less than .1.   Model 4 may not adequately
  describe the data; you may want to consider another model.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =      1.62505

              BMDL =     0.729987
                                  G-468

-------
G.3.21.3. Figure for Selected Model: Exponential (M4)


                          Exponential_beta Model 4 with 0.95 Confidence Level
                        Exponential
 CD
 co
 c
 o
 Q.
 CO
 CD
 CO
 CD
                           20
40           60

     dose
80
100
   17:3402/162010
                                            G-469

-------
G.3.22. Kociba et al. (1918): Uroporphyrin per Creatinine, Female
G.3.22.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear1"
Polynomial, 3 -degree
Power
Power, unrestricted
Degrees of
freedom
2
2
1
0
0
2
2
2
1
X2 p-value
0.661
0.661
0.576
N/A
N/A
0.720
0.720
0.720
0.515
AIC
-93.561
-93.561
-92.078
-90.190
-90.190
-93.735
-93.735
-93.735
-91.967
BMD
(ng/kg-day)
4.357E+01
4.357E+01
1.719E+01
1.080E+01
1.099E+01
3.522E+01
3.522E+01
3.522E+01
2.274E+01
BMDL
(ng/kg-day)
3.328E+01
3.328E+01
5.516E+00
5.613E+00
5.088E+00
2.500E+01
2.500E+01
2.500E+01
3.334E+00
Notes

power hit bound (d=l)





power bound hit
(power =1)
unrestricted
(power =0.731)
a Constant variance model selected (p = 0.4919).
b Best-fitting model, BMDS output presented in this appendix.
G.3.22.2. Output for Selected Model: Linear
Kociba et al. (1978): Uroporphyrin per Creatinine, Female
         Polynomial Model.  (Version:  2.13;   Date: 04/08/2008)
         Input Data File: C:\l\28_Kociba_1978_Uropor_LinearCV_l.(d)
         Gnuplot Plotting File:   C:\l\28_Kociba_1978_Uropor_LinearCV_l.plt
                                            Tue Feb 16  17:34:12  2010
 Table  2


   The  form of the response  function is:

   Y[dose]  = beta 0 + beta l*dose  + beta  2*doseA2 +  ...
   Dependent variable = Mean
   Independent variable = Dose
   rho  is  set to 0
   Signs  of the polynomial coefficients are not restricted
   A  constant variance model is  fit

   Total  number of dose groups = 4
   Total  number of records with  missing values = 0
   Maximum number of iterations  = 250
   Relative Function Convergence has  been set to: le-008
   Parameter Convergence has been set to: le-008
                                      G-470

-------
                  Default Initial Parameter Values
                          alpha =    0.0030385
                            rho =            0   Specified
                         beta_0 =     0.154759
                         beta 1 =    0.0014231
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

                  alpha       beta_0       beta_l

     alpha            1    -2.2e-009     3.5e-009

    beta_0    -2.2e-009            1        -0.55

    beta 1     3.5e-009        -0.55            1
Confidence Interval
       Variable
Upper Conf. Limit
          alpha
0.00406867
         beta_0
0.181105
         beta_l
0.00194739
      Estimate

    0.00251184

      0.154759

     0.0014231
Parameter Estimates



       Std.  Err.

     0.000794315

       0.0134422

     0.000267497
   95.0% Wald

Lower Conf.  Limit

    0.000955015

       0.128413

    0.000898818
     Table of Data and Estimated Values of Interest
Dose
Res .
0
1
10
100
N
5
5
5
5
Obs Mean
0
0
0
0
.157
.143
.181
.296
Est Mean
0
0
0
0
.155
.156
.169
.297
Obs

0
0
0
Std Dev
0.05
.037
.053
.074
Est
0.
0.
0.
0.
Std Dev
0501
0501
0501
0501
Scaled

-0
0
-0.
0.1
.588
.536
0477
 Model Descriptions for likelihoods calculated
 Model Al:
Yij = Mu (i)  + e (ij ;
                                     G-471

-------
           Var{e(ij)} = SigmaA2

 Model A2:         Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:         Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:          Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model      Log(likelihood)
             Al           50.195349
             A2           51.400051
             A3           50.195349
         fitted           49.867385
              R           41.049755
# Param's
5
8
5
3
2
AIC
-90.390697
-86.800103
-90.390697
-93.734769
-78.099510
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
-2*log(Likelihood Ratio)  Test df
            20.7006
            2.40941
            2.40941
           0.655928
6
3
3
2
   p-value

0.002076
  0.4919
  0.4919
  0.7204
The p-value for Test 1 is less than  .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than
model appears to be appropriate here
                            .1.  A homogeneous variance
The p-value for Test 3 is greater than  .1.
 to be appropriate here

The p-value for Test 4 is greater than  .1.
to adequately describe the data
                                 The modeled variance appears
                                 The model chosen seems
                                     G-472

-------
              Benchmark Dose Computation



Specified effect =               1



Risk Type        =      Estimated  standard  deviations  from the control mean



Confidence level =           0.95



              BMD =         35.2176
             BMDL =
                            25.0024
                      Linear
G.3.22.3. Figure for Selected Model: Linear



                                Linear Model with 0.95 Confidence Level


                      I inasir  	
           0.4





          0.35





           0.3
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 OJ
          0.25
           0.2
          0.15
           0.1
                            BMDL
                             20
                                       BMD
40          60

     dose
                                                                 80
100
   17:3402/162010
                                        G-473

-------
G.3.23. Kuchiiwa et al. (2002): Immunoreactive Neurons in Dorsalis, Males
G.3.23.1. Summary Table of BMDS Modeling Results
Model3
Linear1"
Degrees of
Freedom
0
X2 p-value
N/AC
AIC
93.91
BMD
(ng/kg-day)
1.646E-01
BMDL
(ng/kg-day)
1.163E-01
Notes

a Constant variance model selected (p = 0.530).
b Best-fitting model, BMDS output presented in this appendix.
0 p-value could not be calculated because there were no available degrees of freedom.
G.3.23.2. Output for Selected Model: Linear


         Polynomial Model.  (Version:  2.13;   Date: 04/08/2008)
         Input Data File:
C:\USEPA\BMDS21\l\75_Kuchiiwa_2002_dors_admin_dd_LinearCV_l.(d)
         Gnuplot Plotting File:
C:\USEPA\BMDS21\l\75_Kuchiiwa_2002_dors_admin_dd_LinearCV_l.plt
                                            Tue Aug 16  13:41:50  2011
 number_labeled_cells_dorsalis


   The  form of the response  function is:

   Y[dose]  = beta 0 + beta l*dose  + beta  2*doseA2 +  ...
   Dependent variable = Mean
   Independent variable = Dose
   rho  is  set to 0
   Signs  of the polynomial coefficients are not restricted
   A  constant variance model is  fit

   Total  number of dose groups = 2
   Total  number of records with  missing values = 0
   Maximum number of iterations  = 250
   Relative Function Convergence has  been set to: le-008
   Parameter Convergence has been set to: le-008
                   Default Initial  Parameter Values
                           alpha  =       670.324
                             rho  =             0   Specified
                          beta_0  =       237.097
                          beta  1  =      -143.626
            Asymptotic Correlation  Matrix of Parameter  Estimates

            (  *** The model parameter(s)   -rho

                                      G-474

-------
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
                  alpha

     alpha            1

    beta_0     3.8e-008

    beta 1    -1.9e-008
           beta_0       beta_l

         3.8e-008    -1.9e-008

                1        -0.71

            -0.71            1
Confidence Interval
       Variable
Upper Conf. Limit
          alpha
1005.57
         beta_0
256.008
         beta_l
-105.419
              Parameter Estimates

                                      95.0% Wald

     Estimate        Std.  Err.     Lower Conf. Limit

      558.603          228.049             111.636

      237.097          9.64886             218.186

     -143.626          19.4936            -181.833
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
    0
  0.7
237
137
237
137
  29
22.4
Degrees of freedom for Test A3 vs fitted <= 0
 Model Descriptions for likelihoods calculated


 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user
23.6
23.6
-9.42e-008
 -2.9e-008
                                     G-475

-------
 Model  R:         Yi = Mu + e(i;
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
              Log(likelihood)
                -43.952634
                -43.755407
                -43.952634
                -43.952634
                -54.206960
# Param's
      3
      4
      3
      3
      2
   AIC
 93.905267
 95.510815
 93.905267
 93.905267
112.413921
 Test 1:

 Test 2:
 Test 3:
 Test 4:
 (Note:
   Test
          Explanation of Tests

 Do responses and/or variances differ among Dose levels?
 (A2 vs. R)
 Are Variances Homogeneous?  (Al vs A2)
 Are variances adequately modeled?  (A2 vs. A3)
 Does the Model for the Mean Fit?  (A3 vs. fitted)
When rho=0 the results of Test 3 and Test 2 will be the same.
                     Tests of Interest
  -2*log(Likelihood Ratio)  Test df
           p-value
Test 1
Test 2
Test 3
Test 4
20.9031
0.394453
0.394453
8.95284e-013
2
1
1
0
<.0001
0.53
0.53
NA
The p-value for Test 1 is less than  .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here


The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

NA - Degrees of freedom for Test 4 are  less than or equal to 0.  The Chi-
Square
     test for fit is not valid
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD =       0.164558

                                     G-476

-------
              BMDL  =
                             0.116266
G.3.23.3.  Figure for Selected Model: Linear



                                 Linear Model with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 HI
         280
         260
         140
         120
         100
   13:41 08/162011
                                                                                 0.7
                                          G-477

-------
G.3.24. Kuchiiwa et al. (2002): Immunoreactive Neurons in Medianus, Males
G.3.24.1. Summary Table of BMDS Modeling Results
Model3
Linear1"
Degrees of
Freedom
0
X2 p-value
N/AC
AIC
65.97
BMD
(ng/kg-day)
1.342E-01
BMDL
(ng/kg-day)
8.786E-02
Notes

a Modeled variance model selected (p = 0.025).
b Best-fitting model, BMDS output presented in this appendix.
0 p-value could not be calculated because there were no available degrees of freedom.
G.3.24.2. Output for Selected Model: Linear


         Polynomial Model.  (Version:  2.13;   Date: 04/08/2008)
         Input Data File:
C:\USEPA\BMDS21\l\76_Kuchiiwa_2002_med_admin_dd_Linear_l.(d)
         Gnuplot Plotting File:
C:\USEPA\BMDS21\l\76_Kuchiiwa_2002_med_admin_dd_Linear_l.pit
                                            Tue Aug 16  13:44:08  2011
 number_labeled_cells_medianus


   The  form of the response  function is:

   Y[dose]  = beta 0 + beta l*dose  +  beta  2*doseA2 +  ...
   Dependent variable = Mean
   Independent variable = Dose
   Signs  of the polynomial coefficients are not restricted
   The  variance is to be modeled  as  Var(i)  = exp(lalpha  + log(mean(i))  * rho)

   Total  number of dose groups =  2
   Total  number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence  has been set to: le-008
   Parameter Convergence has been set  to: le-008
                   Default Initial  Parameter Values
                          lalpha  =       4.43247
                             rho  =             0
                          beta_0  =       91.1157
                          beta  1  =      -82.6446
            Asymptotic Correlation  Matrix of Parameter Estimates
                                      G-478

-------

lalpha
rho
beta 0
beta 1
lalpha
1
-0.99
2.7e-009
-1.9e-009
rho
-0.99
1
-3e-009
2.2e-009
beta 0
2.7e-009
-3e-009
1
-0.94
beta 1
-1.9e-009
2.2e-009
-0.94
1
                                 Parameter Estimates
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha         -3.97249
2.44349
            rho           1.9468
3.53497
         beta_0          91.1157
99.9878
         beta_l         -82.6446
-69.1083
                      Std. Err.

                        3.27352

                       0.810306

                        4.52665

                        6.90638
                          95.0% Wald

                       Lower Conf. Limit

                              -10.3885

                              0.358628

                               82.2436

                              -96.1808
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est  Std Dev    Scaled
    0
  0.7
91.1
33.3
91.1
33.3
12.1
4.55
11.1
4.16
 4.41e-009
-4.19e-009
Degrees of freedom for Test A2 vs A3 <= 0
 Warning: Likelihood for fitted model larger than the Likelihood  for model
A3.
 Model Descriptions for likelihoods calculated
 Model Al:        Yij
           Var{e(ij) }

 Model A2:        Yij
           Var{e(ij) }

 Model A3:        Yij
           Var{e(ij) }
      Mu (i)  + e (ij )
      SigmaA2

      Mu (i)  + e (ij )
      Sigma(i)A2

      Mu (i)  + e (ij )
      exp(lalpha  + rho*ln(Mu(i))
                                     G-479

-------
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest

            Model      Log(likelihood)   # Param's      AIC
             Al          -31.500916            3      69.001832
             A2          -28.985335            4      65.970670
             A3          -28.985335            4      65.970670
         fitted          -28.985335            4      65.970670
              R          -46.859574            2      97.719148
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.)

                     Tests of Interest

   Test    -2*log(Likelihood Ratio)  Test df        p-value

   Test 1              35.7485          2          <.0001
   Test 2              5.03116          1          0.0249
   Test 3         2.47269e-012          0              NA
   Test 4        -2.47269e-012          0              NA

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

NA - Degrees of freedom for Test 3 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean


                                     G-480

-------
Confidence  level =



               BMD =





              BMDL =
                       0.95



                   0.134165





                  0.0878581
G.3.24.3. Figure for Selected Model: Linear
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 OJ
110




100




 90




 80




 70




 60




 50




 40




 30
                                 Linear Model with 0.95 Confidence Level
                      Linear
                    BMDL
                    BMD
                         0.1
                         0.2
0.3       0.4

    dose
0.5
0.6
0.7
   13:4408/162011
                                         G-481

-------
G.3.25. Kuchiiwa et al. (2002): Immunoreactive Neurons in B9, Males
G.3.25.1. Summary Table of BMDS Modeling Results
Model3
Linear1"
Degrees of
Freedom
0
X2 p-value
N/AC
AIC
86.12
BMD
(ng/kg-day)
1.136E-01
BMDL
(ng/kg-day)
8.208E-02
Notes

a Constant variance model selected (p = 0.504).
b Best-fitting model, BMDS output presented in this appendix.
0 p-value could not be calculated because there were no available degrees of freedom.
G.3.25.2. Output for Selected Model: Linear


         Polynomial Model.  (Version:  2.13;   Date: 04/08/2008)
         Input Data File:
C:\USEPA\BMDS21\l\77_Kuchiiwa_2002_b9_admin_dd_LinearCV_l.(d)
         Gnuplot Plotting File:
C:\USEPA\BMDS21\l\77_Kuchiiwa_2002_b9_admin_dd_LinearCV_l.pit
                                            Tue Aug 16  13:48:05  2011
 number_labeled_cells_b9


   The  form of the response  function is:

   Y[dose]  = beta 0 + beta l*dose  + beta  2*doseA2 +  ...
   Dependent variable = Mean
   Independent variable = Dose
   rho  is  set to 0
   Signs  of the polynomial coefficients are not restricted
   A  constant variance model is  fit

   Total  number of dose groups = 2
   Total  number of records with  missing values = 0
   Maximum number of iterations  = 250
   Relative Function Convergence has  been set to: le-008
   Parameter Convergence has been set to: le-008
                   Default Initial  Parameter Values
                           alpha  =       350.225
                             rho  =             0   Specified
                          beta_0  =       152.086
                          beta  1  =      -150.415
            Asymptotic Correlation  Matrix of Parameter  Estimates

                                      G-482

-------
           (  *** The model parameter(s)  -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
                  alpha

     alpha            1

    beta_0     9.2e-032

    beta 1    -2.9e-016
            beta_0       beta_l

            le-031    -2.9e-016

                 1        -0.71

             -0.71            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          291.854
525.381
         beta_0          152.086
165.756
         beta_l         -150.415
-122.798
               Parameter Estimates



                      Std.  Err.

                        119.149

                         6.9744

                        14.0904
                          95.0% Wald

                       Lower Conf. Limit

                               58.3265

                               138.416

                              -178.031
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
    0
  0.7
 152
46.8
 152
46.8
  16
21.1
Degrees of freedom for Test A3 vs fitted <= 0
 Model Descriptions for likelihoods calculated
17.1
17.1
        0
1.02e-015
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that

                                     G-483

-------
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -40.057520
              -39.834453
              -40.057520
              -40.057520
              -54.163617
 # Param's
       3
       4
       3
       3
       2
     AIC
   86.115041
   87.668907
   86.115041
   86.115041
  112.327234
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.
   Test

   Test 1
   Test 2
   Test 3
   Test 4
                     Tests of Interest
-2*log(Likelihood Ratio)   Test df
            28.6583
           0.446134
           0.446134
       1.37845e-012
2
1
1
0
 p-value

<.0001
0.5042
0.5042
    NA
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.  The modeled variance appears
 to be appropriate here

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

                                     G-484

-------
               BMD =         0.113578



              BMDL =        0.0820848
G.3.25.3. Figure for Selected Model: Linear
                                 Linear Model with 0.95 Confidence Level
 c
 o
 Q.
 o:

 c
 (0
 0)
180




160




140




120




100




 80




 60




 40




 20
                      Linear
                   BMDL
                   BMD
                         0.1
                         0.2
0.3       0.4

    dose
0.5
0.6
0.7
   13:4808/162011
                                          G-485

-------
G.3.26. Kuchiiwa et al. (2002): Immunoreactive Neurons in Magnus, Males
G.3.26.1. Summary Table of BMDS Modeling Results
Model3
Linear1"
Degrees of
freedom
0
X2 p-value
N/AC
AIC
60.36
BMD
(ng/kg-day)
9.131E-02
BMDL
(ng/kg-day)
5.577E-02
Notes

a Modeled variance model selected (p = 0.013).
b Best-fitting model, BMDS output presented in this appendix.
0 p-value could not be calculated because there were no available degrees of freedom.
G.3.26.2. Output for Selected Model: Linear


         Polynomial  Model.  (Version:  2.13;   Date: 04/08/2008)
         Input Data  File:
C:\USEPA\BMDS21\l\78_Kuchiiwa_2002_mag_admin_dd_Linear_l.(d)
         Gnuplot Plotting File:
C:\USEPA\BMDS21\l\78_Kuchiiwa_2002_mag_admin_dd_Linear_l.pit
                                            Tue Aug 16  13:46:34  2011
 number_labeled_cells_magnus


   The  form of the response function is:

   Y[dose]  = beta 0 + beta l*dose  +  beta  2*doseA2 +
   Dependent variable = Mean
   Independent variable = Dose
   Signs  of the polynomial coefficients  are not restricted
   The  variance is to be modeled  as  Var(i)  = exp(lalpha  + log(mean(i))  * rho)

   Total  number of dose groups =  2
   Total  number of records with missing  values = 0
   Maximum number of iterations = 250
   Relative Function Convergence  has been set to: le-008
   Parameter Convergence has been set  to: le-008
                   Default Initial  Parameter Values
                          lalpha  =       4.05645
                             rho  =             0
                          beta_0  =       43.6123
                          beta  1  =      -33.9836
            Asymptotic Correlation  Matrix of Parameter Estimates

                  lalpha          rho        beta 0       beta  1
                                      G-486

-------
lalpha
rho
beta 0
beta 1
1
-0.99
4.1e-009
-5.6e-008
-0.99
1
-4.6e-009
5.3e-008
4.1e-009
-4.6e-009
1
-0.32
-5.6e-008
5.3e-008
-0.32
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha          12.7854
19.6944
            rho         -2.78668
-0.757015
         beta_0          43.6123
46.0952
         beta_l         -33.9836
-22.7674
               Parameter Estimates



                      Std.  Err.

                        3.52508

                        1.03556

                        1.26679

                        5.72265
                          95.0% Wald

                       Lower Conf. Limit

                               5.87638

                              -4.81635

                               41.1294

                              -45.1998
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
    0
  0.7
43.6
19.8
43.6
19.8
 3.4
10.2
Degrees of freedom for Test A2 vs A3 <= 0

Degrees of freedom for Test A3 vs fitted <= 0



 Model Descriptions for likelihoods calculated
 Model Al:        Yij
           Var{e(ij)}

 Model A2:        Yij
           Var{e(ij)}
      Mu(i)  + e(i j '
      S i gma A 2

      Mu(i)  + e(i j'
      Sigma(i)A2
 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters that
     were specified by the user
 3.1
9.31
1.13e-008
1.88e-008
                                     G-487

-------
 Model  R:         Yi = Mu + e(i;
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest

            Model      Log(likelihood)   # Param's      AIC
             Al          -29.244768            3      64.489536
             A2          -26.179929            4      60.359859
             A3          -26.179929            4      60.359859
         fitted          -26.179929            4      60.359859
              R          -37.469939            2      78.939878
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.)

                     Tests of Interest

   Test    -2*log(Likelihood Ratio)  Test df        p-value

   Test 1                22.58          2          <.0001
   Test 2              6.12968          1         0.01329
   Test 3         7.10543e-015          0              NA
   Test 4                    0          0              NA

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

NA - Degrees of freedom for Test 3 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
             Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD =      0.0913086

                                     G-488

-------
              BMDL  =
                            0.0557686
G.3.26.3. Figure for Selected Model: Linear



                                Linear Model with 0.95 Confidence Level


        50



        45
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 OJ
        15
        10
                                                                        0.6
0.7
   13:4608/162011
                                          G-489

-------
G.3.27. Latchoumycandane and Mathur (2002): Sperm Production
G.3.27.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hillb
Linear
Polynomial, 3 -degree
Power
Hill, unrestricted0
Power, unrestricted
Degrees of
freedom
2
2
1
0
1
2
2
2
0
1
X2 p-value
0.0001
0.0001
0.699
N/A
0.859
0.0001
0.0001
0.0001
N/A
0.499
AIC
95.106
95.106
75.263
77.263
75.144
95.308
95.308
95.308
77.113
75.570
BMD
(ng/kg-day)
7.640E+01
7.640E+01
2.435E-01
3.697E-01
1.450E-01
8.275E+01
8.275E+01
8.275E+01
6.943E-02
2.706E-07
BMDL
(ng/kg-day)
3.992E+01
3.992E+01
1.016E-01
1.016E-01
1.559E-02
4.852E+01
4.852E+01
4.852E+01
2.060E-06
2.706E-07
Notes

power hit bound (d = 1)


n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted (n = 0.709)
unrestricted
(power = 0.067)
a Constant variance model selected (p = 0.8506).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.3.27.2. Output for Selected Model: Hill
Latchoumycandane and Mathur (2002): Sperm Production
         Hill Model.  (Version:  2.14;   Date: 06/26/2008)
         Input Data File:  C:\l\30_Latch_2002_Sperm_HillCV_l.(d)
         Gnuplot Plotting  File:   C:\l\30_Latch_2002_Sperm_HillCV_l.plt
                                            Tue  Feb  16 18:13:20 2010
  (xlOA6)  Table 1 without Vitamin E


   The  form of the response  function is:

   Y[dose]  = intercept + v*doseAn/(kAn + doseAn)
   Dependent variable = Mean
   Independent variable =  Dose
   rho  is  set to 0
   Power parameter restricted to be greater than  1
   A  constant variance model is  fit

   Total number of dose groups  = 4
   Total number of records with  missing values =  0
   Maximum number of iterations  = 250
   Relative Function Convergence has been set to:  le-008
   Parameter Convergence has been set to: le-008
                                      G-490

-------
                  Default Initial Parameter Values
                          alpha =      7.23328
                            rho =
                      intercept =
                              v =
                              n =
                              k =
       0
   22.19
   -9.09
 1.80484
0.697086
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho    -n
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
intercept
V
k
alpha
1
6.3e-010
3e-008
8.3e-009
intercept
6.3e-010
1
-0.78
-0.23
V
3e-008
-0.78
1
-0.17
k
8.3e-009
-0.23
-0.17
1
                                 Parameter Estimates
Confidence Interval
Variable
Estimate
Std. Err.
Upper Conf. Limit

9.45061
alpha

intercept
24.1547

-6.52343


0.907359

V

n
k

6.03567

22.1885

-9.00869

1
0.386669

1.74235

1.00316

1.26801

NA
0.265663

NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
                                                         95.0% Wald

                                                      Lower Conf. Limit

                                                              2.62073

                                                              20.2223

                                                             -11.4939


                                                            -0.134021
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
                                    G-491

-------
    0
    1
   10
  100
       22.2
       15.7
       13.7
       13.1
22.2
15.7
13.5
13.2
2.67
2.65
2.19
3.16
2.46
2.46
2.46
2.46
0.00151
-0.0218
  0.134
 -0.114
 Model Descriptions for likelihoods calculated


 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -33.556444
              -33.158811
              -33.556444
              -33.572245
              -47.392394
          # Param's
                5
                8
                5
                4
                2
            AIC
          77.112888
          82.317623
          77.112888
          75.144490
          98.784788
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.
   Test

   Test 1
   Test 2
   Test 3
   Test 4
                     Tests of Interest
-2*log(Likelihood Ratio)  Test df
            28.4672
           0.795266
           0.795266
           0.031602
         6
         3
         3
         1
        p-value

       <.0001
       0.8506
       0.8506
       0.8589
The p-value for Test 1 is less than .05.  There appears to be a

                                    G-492

-------
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
The modeled variance appears
The model chosen seems
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =       0.144988

            BMDL =     0.0155926
                                     G-493

-------
OJ
c/)
c
o
Q.
(/)
0)
o:

c
(0
0)
G.3.27.3. Figure for Selected Model: Hill



                               Hill Model with 0.95 Confidence Level


        26  t     Hill'




        24




        22




        20




        18




        16




        14




        12




        10


         BlflDLJEMD


              0           20           40          60          80          100

                                           dose
   18:1302/162010





G.3.27.4. Output for Additional Model Presented: Hill, Unrestricted


Latchoumycandane and Mathur (2002): Sperm Production






         Hill Model.  (Version:  2.14;   Date:  06/26/2008)

         Input Data File:  C:\l\30_Latch_2002_Sperm_HillCV_U_l.(d)

         Gnuplot Plotting  File:   C:\l\30_Latch_2002_Sperm_HillCV_U_l.plt

                                            Tue  Feb 16 18:13:21  2010




  (xlOA6) Table 1  without  Vitamin E
  The  form of the response  function is:



  Y[dose]  = intercept +  v*doseAn/(kAn  + doseAn)





  Dependent variable = Mean
                                      G-494

-------
   Independent variable = Dose
   rho is set to 0
   Power parameter is not restricted
   A constant variance model is fit

   Total number of dose groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      7.23328
                            rho =
                      intercept =
                              v =
                              n =
                              k =
            0
        22.19
        -9.09
      1.80484
     0.697086
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
intercept
V
n
k
alpha
1
-7.6e-009
8e-008
5e-008
1.9e-008
intercept
-7.6e-009
1
-0.5
-0.015
-0.13
V
8e-008
-0.5
1
0.75
0.55
n
5e-008
-0.015
0.75
1
0.86
k
1.9e-008
-0.13
0.55
0.86
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          6.02773
9.43818
      intercept            22.19
24.1545
              v         -9.23433
-5.27378
              n         0.709305
3.22451
Parameter Estimates



       Std.  Err.

         1.74006

         1.00231

         2.02073

         1.28329
        95.0% Wald

     Lower Conf.  Limit

             2.61728

             20.2255

            -13.1949

             -1.8059
                                     G-495

-------
                        0.290697
                       0.548737
                             -0.784807
1.3662
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
    0
    1
   10
  100
22.2
15.7
13.7
13.1
22.2
15.7
13.7
13.1
2.67
  65
  19
  16
2.46
2.46
2.46
2.46
 2.62e-008
 -1.5e-008
-4.56e-008
-3.52e-007
Degrees of freedom for Test A3 vs fitted <= 0
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e (i) } = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
     Log(likelihood)
       -33.556444
       -33.158811
       -33.556444
       -33.556444
       -47.392394
# Param's
5
8
5
5
2
AIC
77.112888
82.317623
77.112888
77.112888
98.784788
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.
                                     G-496

-------
                     Tests of Interest

   Test    -2*log(Likelihood Ratio)  Test df        p-value

   Test 1              28.4672          6          <.0001
   Test 2             0.795266          3          0.8506
   Test 3             0.795266          3          0.8506
   Test 4         2.84217e-014          0              NA

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.  The modeled variance appears
 to be appropriate here

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =      0.0694325

            BMDL =  2.06007e-006
                                     G-497

-------
G.3.27.5.  Figure for Additional Model Presented: Hill, Unrestricted



                                 Hill Model with 0.95 Confidence Level


        26  ^^———^—-
 o
 Q.
 o:

 c
 (0
 OJ
24




22




20




18




16




14




12




10
           Hill
          BMDLBMD
                            20
                                 40            60

                                       dose
80
100
   18:1302/162010
                                          G-498

-------
G.3.28. Li et al. (1997): FSH
G.3.28.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear
Polynomial,
8-degree
Power b
Hill, unrestricted
Power, unrestricted °
Degrees of
freedom
8
8
7
7
7
8
9
8
6
7
X2 p-value
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
O.OOOl
0.001
0.002
AIC
,095.240
,095.240
,061.243
,061.243
,059.547
,078.221
1,155.670
1,078.221
1,039.902
1,037.821
BMD
(ng/kg-day)
1.340E+04
1.340E+04
1.031E+03
1.031E+03
6.645E+02
5.287E+03
error
5.287E+03
2.809E+00
2.508E+00
BMDL
(ng/kg-day)
1.060E+04
1.060E+04
4.015E+02
4.015E+02
error
3.602E+03
error
3.602E+03
6.602E-01
2.525E-01
Notes

power hit bound (d=\)

power hit bound (d=\)
n lower bound hit (n = 1)


power bound hit
(power = 1)
unrestricted (n = 0.291)
unrestricted
(power = 0.279)
a Nonconstant variance model selected (p = O.OOOl).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.3.28.2. Output for Selected Model: Power
Li et al. (1997): FSH
         Power Model.  (Version:  2.15;  Date:  04/07/2008)
         Input Data File:  C:\l\72_Li_1997_FSH_Pwr_l.(d)
         Gnuplot Plotting  File:   C:\l\72_Li_1997_FSH_Pwr_l.plt
                                            Tue  Feb 16 20:07:31 2010
 Figure  3:  FSH in female  S-D  rats 24hr after dosing,  22 day old rats


   The  form of the response function is:

   Y[dose]  = control + slope  *  doseApower
   Dependent variable = Mean
   Independent variable =  Dose
   The  power is restricted to  be greater than  or  equal to 1
   The  variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  *  rho)

   Total  number of dose groups = 10
   Total  number of records with missing values  =  0
   Maximum number of iterations = 250
   Relative Function Convergence has been set  to:  le-008
   Parameter Convergence has been set to: le-008
                                       G-499

-------
                  Default Initial Parameter Values
                         lalpha =       9.8191
                            rho =            0
                        control =      22.1591
                          slope =      26.1213
                          power =     0.264963
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -power
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

lalpha
rho
control
slope
lalpha
1
-0.99
-0.29
-0.023
rho
-0.99
1
0.2
0.023
control
-0.29
0.2
1
-0.35
slope
-0.023
0.023
-0.35
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha           3.5473
5.9709
            rho          1.26137
1.74009
        control          88.9479
114.254
          slope        0.0188972
0.0257908
          power                1
Parameter Estimates



       Std.  Err.

         1.23656

        0.244246

         12.9114

      0.00351723

              NA
   95.0% Wald

Lower Conf.  Limit

        1.12369

       0.782659

        63.6419

      0.0120035
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
                                     G-500

-------
0
3
10
30
100
300
1000
3000
le+004
3e+004
10
10
10
10
10
10
10
10
10
10
23.9
22.2
85.2
73.3
126
132
117
304
347
455
88.9
89
89.1
89.5
90.8
94.6
108
146
278
656
29.6
48.5
94.3
48.5
159
116
51.2
154
151
286
99.9
99.9
100
100
101
104
113
136
205
352
-2.06
-2.12
-0.124
-0.511
1.1
1.14
0.25
3.68
1.06
-1.8
Model Descriptions for likelihoods calculated
Model Al:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = SigmaA2

Model A2:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = Sigma(i)A2

Model A3:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
    Model A3 uses any fixed variance parameters  that
    were specified by the user

Model  R:         Yi = Mu + e(i)
           Var{e(i)} = SigmaA2
                      Likelihoods of Interest
           Model
            Al
            A2
            A3
        fitted
             R
            Log(likelihood)
             -535.687163
             -496.367061
             -502.709623
             -535.110448
             -574.835246
# Param's
11
20
12
4
2
AIC
1093.374327
1032.734122
1029.419246
1078.220896
1153.670492
                  Explanation of Tests

Test 1:  Do responses and/or variances differ  among  Dose  levels?
         (A2 vs. R)
Test 2:  Are Variances Homogeneous?  (Al vs A2)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)
Test 4:  Does the Model for the Mean Fit?  (A3  vs.  fitted)
(Note:  When rho=0 the results of Test 3 and Test  2  will  be  the  same.

                    Tests of Interest
  Test

  Test 1
  Test 2
-2*log(Likelihood Ratio)  Test df
            156.936
            78.6402
18
 9
 p-value

<.0001
<.0001
                                    G-501

-------
   Test 3              12.6851          8          0.1232
   Test 4              64.8016          8          <.0001

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 5286.67


            BMDL = 3601.91
                                     G-502

-------
 o
 Q.
 o:
 c
 (0
 OJ
        400
        300
        200
        100
                  Power
G.3.28.3. Figure for Selected Model: Power

                              Power Model with 0.95 Confidence Level

        700


        600


        500
                  BMDL
                          BMD
                        5000      10000
                                           15000
                                           dose
20000     25000      30000
   20:0702/162010
G.3.28.4. Output for Additional Model Presented: Power, Unrestricted
Lietal. (1997): FSH
        Power  Model.  (Version: 2.15;   Date:  04/07/2008)
        Input  Data File: C:\l\72_Li_1997_FSH_Pwr_U_l.(d)
        Gnuplot Plotting File:  C:\l\72_Li_1997_FSH_Pwr_U_l.plt
                                            Tue Feb 16  20:07:33  2010
 Figure 3:  FSH  in female S-D rats  24hr after dosing, 22 day  old rats


   The form of  the response function  is:

   Y[dose]  = control + slope * doseApower
   Dependent  variable = Mean
   Independent  variable = Dose
                                      G-503

-------
   The power is not restricted
   The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  * rho)

   Total number of dose groups = 10
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =       9.8191
                            rho =            0
                        control =      22.1591
                          slope =      26.1213
                          power =     0.264963
           Asymptotic Correlation Matrix of Parameter Estimates
lalpha
lalpha
rho
control
slope
power

-0
-0
-0
0
1
.99
.69
.15
.28
-0

0
0
-0
rho
.99
1
.65
.11
.26
control
-0.69
0.65
1
-0.17
0.024
slope
-0.
0.
-0.

-0.
15
11
17
1
93
power
0.28
-0.26
0.024
-0.93
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha          3.72156
5.93861
            rho          1.17032
1.60788
        control          15.7412
29.4094
          slope           24.963
37.5651
          power         0.278637
0.339857
Parameter Estimates



       Std.  Err.

         1.13117

        0.223249

         6.97367

         6.42976

       0.0312355
   95.0% Wald

Lower Conf. Limit

         1.5045

       0.732758

        2.07307

        12.3609

       0.217417
     Table of Data and Estimated Values of Interest
                                     G-504

-------
Dose
Res .
0
3
10
30
100
300
1000
3000
le+004
3e+004
N
10
10
10
10
10
10
10
10
10
10
Obs Mean
23.9
22.2
85.2
73.3
126
132
117
304
347
455
                             Est Mean
                                        Obs  Std  Dev  Est  Std Dev
Scaled
15.7
49.6
63.2
80.1
106
138
187
248
341
457
29.6
48.5
94.3
48.5
159
116
51.2
154
151
286
32.3
63.2
72.7
83.6
98.4
115
137
162
195
232
0.796
-1.38
0.96
-0.259
0.654
-0.164
-1.62
1.1
0.0999
-0.0271
Model Descriptions for likelihoods  calculated
Model Al:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = SigmaA2

Model A2:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = Sigma(i)A2

Model A3:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = exp(lalpha +  rho*ln(Mu(i)))
    Model A3 uses any fixed variance parameters  that
    were specified by the user

Model  R:         Yi = Mu + e(i)
           Var{e(i)} = SigmaA2
                      Likelihoods of  Interest
           Model
            Al
            A2
            A3
        fitted
             R
Log (likelihood)
-535.687163
-496.367061
-502.709623
-513.910636
-574.835246
# Param' s
11
20
12
5
2
AIC
1093.374327
1032.734122
1029.419246
1037.821272
1153.670492
                  Explanation of Tests

Test 1:  Do responses and/or variances differ  among  Dose  levels?
         (A2 vs. R)
Test 2:  Are Variances Homogeneous?  (Al vs A2)
Test 3:  Are variances adequately modeled?  (A2  vs. A3)
Test 4:  Does the Model for the Mean  Fit?  (A3  vs.  fitted)
(Note:  When rho=0 the results of Test 3 and Test  2  will  be  the same.

                    Tests of Interest
                                    G-505

-------
156.936
78.6402
12.6851
22.402
18
9
8
7
<.0001
<.0001
0.1232
0.002165
   Test    -2*log(Likelihood Ratio)  Test df        p-value

   Test 1
   Test 2
   Test 3
   Test 4

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 2.50839


            BMDL = 0.252541
                                     G-506

-------
o
Q.
o:

c
(0
OJ
        400
         300
        200
        100
                   Power
G.3.28.5. Figure for Additional Model Presented: Power, Unrestricted



                                Power Model with 0.95 Confidence Level


         700




         600




         500
                0        5000      10000      15000     20000      25000      30000

                                              dose
  20:0702/162010
                                         G-507

-------
G.3.29. Li et al. (2006): Estradiol, 3-Day
G.3.29.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear1"
Polynomial,
3 -degree
Power
Hill, unrestricted
Power, unrestricted
Degrees of
freedom
2
2
1
0
0
2
2
2
0
1
X2 p-value
0.147
0.147
0.341
N/A
N/A
0.151
0.151
0.151
N/A
0.327
AIC
269.146
269.146
268.212
270.212
270.212
269.084
269.084
269.084
270.266
268.266
BMD
(ng/kg-day)
3.044E+02
3.044E+02
error
error
error
3.471E+02
3.471E+02
3.471E+02
1.059E+17
3.727E+14
BMDL
(ng/kg-day)
1.108E+02
1.108E+02
error
error
error
1.082E+02
1.082E+02
1.082E+02
1.059E+17
error
Notes

power hit bound (d = 1)





power bound hit
(power =1)
unrestricted (n = 0.025)
unrestricted
(power =0.0 12)
a Constant variance model selected (p = 0.4372).
b Best-fitting model, BMDS output presented in this appendix.
G.3.29.2. Output for Selected Model: Linear
Li et al. (2006): Estradiol, 3-Day
         Polynomial Model.  (Version:  2.13;  Date: 04/08/2008)
         Input Data File: C:\l\31_Li_2006_Estra_LinearCV_l.(d)
         Gnuplot Plotting File:   C:\l\31_Li_2006_Estra_LinearCV_l.plt
                                            Tue Feb  16  18:13:56 2010
 Figure  3,  3-day estradiol


   The  form of the response  function is:

   Y[dose]  = beta 0 + beta l*dose  + beta 2*doseA2 +
   Dependent variable = Mean
   Independent variable = Dose
   rho  is  set to 0
   Signs  of the polynomial coefficients are not restricted
   A  constant variance model  is  fit

   Total  number of dose groups = 4
   Total  number of records with  missing values = 0
   Maximum number of iterations  = 250
   Relative Function Convergence has  been set to: le-008
   Parameter Convergence has  been set to: le-008
                                      G-508

-------
                  Default Initial Parameter Values
                          alpha =      267.211
                            rho =            0   Specified
                         beta_0 =      16.4428
                         beta 1 =    0.0468351
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
                  alpha

     alpha            1

    beta_0    -2.6e-013

    beta 1    -4.5e-015
   beta_0       beta_l

-2.6e-013    -4.5e-015

        1        -0.68

    -0.68            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          264.303
380.137
         beta_0          16.4428
23.3111
         beta_l        0.0468351
0.16968
      Parameter Estimates



             Std.  Err.

                  59.1

               3.50431

              0.062677
   95.0% Wald

Lower Conf.  Limit

        148.469

        9.57445

     -0.0760095
     Table of Data and Estimated Values of Interest
Dose
Res .
0
2
50
100
N
10
10
10
10
Obs
10
19
24
18
Mean
.2
.9
.7
.1
Est
16
16
18
21
Mean
.4
.5
.8
.1
Obs S
12

14
17
td Dev
.2
20
.6
.6
Est St
16.
16.
16.
16.
d Dev
3
3
3
3
Seal

0

-0
ed
1.22
.656
1.16
.591
 Model Descriptions for likelihoods calculated
                                    G-509

-------
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -129.653527
             -128.294657
             -129.653527
             -131.541911
             -131.819169
 # Param's
       5
       8
       5
       3
       2
     AIC
  269.307054
  272.589314
  269.307054
  269.083823
  267.638338
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
-2*log(Likelihood Ratio)  Test df
            7.04902
            2.71774
            2.71774
            3.77677
6
3
3
2
 p-value

0.3163
0.4372
0.4372
0.1513
The p-value for Test 1 is greater than  .05.  There may not be a
diffence between responses and/or variances among the dose levels
Modelling the data with a dose/response curve may not be appropriate
The p-value for Test 2 is greater than  .1.
model appears to be appropriate here
                                 A homogeneous variance
The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is greater than  .1.  The model chosen seems
to adequately describe the data
                                     G-510

-------
              Benchmark  Dose Computation



Specified  effect =              1



Risk Type         =      Estimated standard deviations  from the control mean



Confidence level =           0.95



              BMD =          347.12
             BMDL =
                            108.173
G.3.29.3. Figure for Selected Model: Linear



                              Linear Model with 0.95 Confidence Level
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 0)
       35
        30
       25
20
15
        10
                  Linear
                             BMDL
                                                                     BMD
                      50       100      150       200

                                           dose
                                                  250
300
350
   18:1302/162010
                                       G-511

-------
G.3.30. Li et al. (2006): Progesterone, 3-Day
G.3.30.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)b
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Power
Power, unrestricted
Degrees of
freedom
2
2
1
0
1
2
2
2
1
X2 p-value
0.001
0.001
0.384
N/A
0.386
0.001
0.001
0.001
0.405
AIC
330.234
330.234
315.734
317.734
315.729
331.121
331.121
331.121
315.670
BMD
(ng/kg-day)
5.252E+01
5.252E+01
1.353E-01
5.225E-01
1.135E-02
7.765E+01
7.765E+01
7.765E+01
1.066E-63
BMDL
(ng/kg-day)
error
error
8.351E-02
7.503E-02
1.161E-05
5.264E+01
5.264E+01
5.264E+01
1.066E-63
Notes

power hit bound (d=\)


n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted
(power = 0.009)
a Nonconstant variance model selected (p = 0.0013).
b Best-fitting model, BMDS output presented in this appendix.
G.3.30.2. Output for Selected Model: Exponential (M4)
Li et al. (2006): Progesterone, 3-Day
         Exponential Model.  (Version:  1.61;  Date: 7/24/2009)
         Input Data File: C:\l\32_Li_2006_Progest_Exp_l.(d)
         Gnuplot Plotting File:
                                            Tue Feb 16  18:14:31  2010
 Figure  4,  3-day progesterone
   The  form of the response  function by Model:
      Model  2:
      Model  3:
      Model  4:
      Model  5:
Y[dose]
Y[dose]
Y[dose]
Y[dose]
a
a
a
a
exp{sign * b  *  dose}
exp{sign *  (b *  dose)Ad}
[c-(c-l) * exp{-b  *  dose}]
[c-(c-l) * exp{-(b *  dose)Ad}]
    Note:  Y[dose]  is the median  response for exposure =  dose;
           sign = +1 for increasing  trend in data;
           sign = -1 for decreasing  trend.

      Model  2 is nested within Models  3 and 4.
      Model  3 is nested within Model  5.
      Model  4 is nested within Model  5.
   Dependent variable = Mean
   Independent variable = Dose
   Data  are assumed to be distributed:  normally
   Variance Model:  exp(lnalpha  +rho  *ln(Y[dose]))
   The variance is  to be modeled  as  Var(i)  = exp(lalpha  +  log(mean(i))

                                      G-512
                                                        rho)

-------
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values
               Variable

                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                     Model 4
                          11.3313
                         -1.44835
                          64.8274
                        0.0456906
                         0.166844
                                1
                  Parameter Estimates
                Variable

                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                      Model 4

                          14.074
                        -2.27065
                         61.7474
                         2.13327
                        0.318566
                               1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
      2
     50
    100
10
10
10
10
61.74
30.56
16.93
11.36
11.1
40.48
33.3
43.75
   Dose

      0
      2
     50
    100
               Estimated Values of Interest

             Est Mean      Est Std     Scaled Residual
    61.75
    20.26
    19.67
    19.67
10.55
37.38
38.66
38.66
-0.002085
   0.8713
   -0.224
  -0.6801
Other models for which likelihoods are calculated:

                                  G-513

-------
     Model Al:        Yij
               Var{e(ij) }

     Model A2:        Yij
               Var{e(ij) }

     Model A3:        Yij
               Var{e(ij) }

     Model  R:        Yij
               Var{e(ij) }
                         Mu(i) + e (ij)
                         SigmaA2

                         Mu(i) + e (ij)
                         Sigma(i)A2

                         Mu(i) + e (i j)
                         exp(lalpha + log(mean(i)) * rho)

                         Mu + e(i)
                         SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
4
-159.6327
-151.8128
-152.4882
-165.6989
-152.8668
5
8
6
2
5
329.2653
319.6255
316.9763
335.3978
315.7335
   Additive constant for all log-likelihoods =     -36.76.  This  constant
added to the
   above values gives the log-likelihood including the term that  does not
   depend on the model parameters.
R)
                              Explanation of Tests

Test 1:  Does response and/or variances differ among Dose levels?  (A2  vs.

Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)

Test 6a: Does Model 4 fit the data?  (A3 vs  4)
     Test

     Test 1
     Test 2
     Test 3
    Test 6a
                         Tests of Interest

                -2*log(Likelihood Ratio)
                                                  D. F.
p-value
27.77
15.64
1.351
0.7572
6
3
2
1
0.0001037
0.001344
0.5089
0.3842
     The p-value for Test 1 is less than  .05.  There appears to be a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

     The p-value for Test 2 is less than  .1.  A non-homogeneous
     variance model appears to be appropriate.

                                     G-514

-------
  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is greater than .1.  Model 4 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =     0.135296

              BMDL =    0.0835054
                                 G-515

-------
G.3.30.3. Figure for Selected Model: Exponential (M4)

                      Exponential_beta Model 4 with 0.95 Confidence Level
 CD
 co
 c
 o
 Q.
 CO
 CD
 CO
 CD
       60
       40
       20
      -20
                     Exponential
         BMDLBMD
             0
   18:1402/162010
20
40         60
     dose
80
100
G.3.30.4. Output for Additional Model Presented: Hill, Unrestricted
Li et al. (2006): Progesterone, 3-Day
         Hill  Model.  (Version:  2.14;  Date:  06/26/2008)
         Input Data File:  C:\l\32_Li_2006_Progest_Hill_U_l.(d)
         Gnuplot Plotting  File:   C:\l\32_Li_2006_Progest_Hill_U_l.plt
                                            Tue  Feb 16 18:14:41  2010
 Figure  4,  3-day progesterone


   The form of the response  function is:

   Y[dose]  = intercept + v*doseAn/(kAn + doseAn)
   Dependent variable = Mean
   Independent variable =  Dose
   Power  parameter is not  restricted
   The variance is to be modeled as Var(i)
                      = exp(lalpha
                            rho  * In(mean(i)))
                                       G-516

-------
   Total number of dose groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
lalpha =
rho =
intercept =
v =
n =
k =
7.
61
-50
1.
1
08699
0
.7404
.3835
43997
.6159
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -k
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

lalpha
rho
intercept
V
n
lalpha
1
-0.99
-0.097
0.84
NA
rho
-0.

0.
-0.
NA
99
1
13
81

intercept
-0.097
0.13
1
-0.43
NA
V
0.84
-0.81
-0.43
1
NA

NA
NA
NA
NA

NA
NA - This parameter's variance has been estimated as zero or less.
THE MODEL HAS PROBABLY NOT CONVERGED!!!
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
NA
            rho
NA
      intercept
NA
         Parameter Estimates

                                 95.0% Wald

Estimate        Std.  Err.      Lower Conf.  Limit

 13.9863               NA                  NA

-2.25026               NA                  NA

 61.7404               NA                  NA
                                     G-517

-------
NA

NA
               -42.1239

                2.02774

                 le-013
                             NA

                             NA

                             NA
                                    NA

                                    NA
At least some variance estimates are negative.
THIS USUALLY MEANS THE MODEL HAS NOT CONVERGED!
Try again from another starting point.
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est  Std  Dev    Scaled
    0
    2
   50
  100
10
10
10
10
61.7
30.6
16.9
11.4
61.7
19.6
19.6
19.6
11.1
40.5
33.3
43.7
10.5
38.3
38.3
38.3
9.74e-008
    0.905
   -0.222
   -0.683
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
   Model
    Al
    A2
    A3
fitted
     R
                       Log (likelihood)
                        -159.632675
                        -151.812765
                        -152.488175
                        -152.873643
                        -165.698875
                       # Param's
                             5
                             8
                             6
                             5
                             2
                         AIC
                      329.265349
                      319.625529
                      316.976349
                      315.747285
                      335.397750
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose  levels?

                                     G-518

-------
           (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.
                     Tests of Interest
           -2*log(Likelihood Ratio)  Test df
           p-value
Test

Test 1
Test 2
Test 3
Test 4
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
27.7722
15.6398
1.35082
0.770936
6
3
2
1
0.0001037
0.001344
0.5089
0.3799
The p-value for Test 2 is less than .1,
model appears to be appropriate
A non-homogeneous variance
The p-value for Test 3 is greater than .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is greater than .1.  The model chosen seems
to adequately describe the data
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =   5.81703e-014

            BMDL =  5.81703e-014
                                     G-519

-------
G.3.30.5. Figure for Additional Model Presented: Hill,  Unrestricted



                                  Hill Model with 0.95 Confidence Level
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 0)
        60
        40
20
        -20
                   Hill
           BMDLJBMD
   18:1402/162010
                             20
                                  40            60

                                        dose
80
100
                                           G-520

-------
G.3.31. Markowski et al. (2001): FR10 Run Opportunities
G.3.31.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)b
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Power
Power, unrestricted
Degrees of
freedom
2
2
1
0
0
2
2
2
1
X2 p-value
0.248
0.248
0.412
N/A
N/A
0.190
0.190
0.190
0.238
AIC
117.557
117.557
117.445
118.918
118.918
118.089
118.089
118.089
118.164
BMD
(ng/kg-day)
1.653E-KJ2
1.653E+02
4.742E+01
3.178E+01
2.348E+01
2.081E+02
2.081E+02
2.081E+02
9.153E+01
BMDL
(ng/kg-day)
5.025E+01
5.025E+01
1.729E-01
3.967E-05
6.728E-06
1.051E+02
1.051E+02
1.051E+02
5.911E-07
Notes

power hit bound (d=\)





power bound hit
(power =1)
unrestricted
(power = 0.237)
a Constant variance model selected (p = 0.1719).
b Best-fitting model, BMDS output presented in this appendix.
G.3.31.2. Output for Selected Model: Exponential (M2)
Markowski et al. (2001): FR10 Run Opportunities
         Exponential Model.  (Version:  1.61;   Date: 7/24/2009)
         Input Data File: C:\l\33_Mark_2001_FR10opp_ExpCV_l.(d)
         Gnuplot Plotting File:
                                            Tue Feb 16 18:15:26  2010
 Table  3
   The  form of the response function  by Model:
      Model 2:      Y[dose] = a *  exp{sign * b * dose}
      Model 3:      Y[dose] = a *  exp{sign * (b * dose)Ad}
      Model 4:      Y[dose] = a *  [c-(c-l)  * exp{-b * dose}]
      Model 5:      Y[dose] = a *  [c-(c-l)  * exp{-(b * dose)Ad}]

    Note:  Y[dose]  is the median response for exposure = dose;
           sign = +1 for increasing  trend in data;
           sign = -1 for decreasing  trend.

      Model 2  is nested within Models 3 and 4.
      Model 3  is nested within Model  5.
      Model 4  is nested within Model  5.
   Dependent variable = Mean
   Independent variable = Dose
   Data  are  assumed to be distributed:  normally
   Variance  Model:  exp(lnalpha +rho  *ln(Y[dose])

                                      G-521

-------
rho is set to 0.
A constant variance model is fit.

Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values
               Variable

                 Inalpha
                     rho(S)
                       a
                       b
                       c
                       d
                 Model 2

                       3.5321
                            0
                      6.98169
                   0.00309891
                            0
                            1
   (S) = Specified
                  Parameter Estimates
                Variable

                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                  Model 2

                   3.64823
                         0
                   11.9443
                 0.0044262
                         0
                         1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
     20
     60
    180
13.29
11.25
5.75
7
8.65
5.56
3.53
6.01
   Dose

      0
     20
     60
    180
               Estimated Values of Interest

             Est Mean      Est Std     Scaled Residual
11.94
10.93
9.158
5.385
  197
  197
6.197
6.197
0.5745
0.1025
-1.347
0.6897
                                  G-522

-------
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)

     Model  R:        Yij = Mu + e(i)
               Var{e (ij)  } = SigmaA2
         Model

            Al
            A2
            A3
             R
             2
                                Likelihoods of Interest

                                Log(likelihood)      DF
                                  -54.38526
                                  -51.88568
                                  -54.38526
                                  -57.45429
                                  -55.77871
                                                             AIC
5
8
5
2
3
118.7705
119.7714
118.7705
118.9086
117.5574
   Additive constant for all log-likelihoods =     -22.05.  This  constant
added to the
   above values gives the log-likelihood including the term that  does  not
   depend on the model parameters.
R)
Test 1:

Test 2:
Test 3:
Test 4:
                     Explanation of Tests

Does response and/or variances differ among Dose  levels?  (A2  vs.

Are Variances Homogeneous?  (A2 vs. Al)
Are variances adequately modeled?  (A2 vs. A3)
Does Model 2 fit the data?  (A3 vs. 2)
     Test

     Test 1
     Test 2
     Test 3
     Test 4
                         Tests of Interest

                -2*log(Likelihood Ratio)

                                 11.14
                                 4.999
                                 4.999
                                 2.787
                                      D.  F.

                                         6
                                         3
                                         3
                                         2
p-value
  0.08423
   0.1719
   0.1719
   0.2482
     The p-value for Test 1 is greater than  .05.  There may not be  a
     diffence between responses and/or variances among the dose levels
     Modelling the data with a dose/response curve may not be appropriate.
                                     G-523

-------
  The p-value for Test 2 is greater than .1.  A homogeneous
  variance model appears to be appropriate here.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 4 is greater than .1.  Model 2 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =      165.284

              BMDL =      50.2488
                                 G-524

-------
G.3.31.3. Figure for Selected Model: Exponential (M2)




                        Exponential_beta Model 2 with 0.95 Confidence Level
 o

 &
 CD
 ro
 
-------
G.3.32. Markowski et al. (2001): FR2 Revolutions
G.3.32.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hillb
Linear
Polynomial,
3 -degree
Power
Power, unrestricted0
Degrees of
freedom
2
2
1
0
0
2
2
2
1
X2 p-value
0.192
0.192
0.298
N/A
N/A
0.150
0.150
0.150
0.160
AIC
217.636
217.636
217.415
218.532
218.532
218.129
218.129
218.129
218.302
BMD
(ng/kg-day)
1.627E+02
1.627E+02
4.668E+01
3.308E+01
2.364E-K)!
1.989E+02
1.989E+02
1.989E+02
9.101E+01
BMDL
(ng/kg-day)
5.807E+01
5.807E+01
1.965E-01
1.193E+01
7.336E-KJO
1.025E+02
1.025E+02
1.025E+02
1.800E-13
Notes

power hit bound (d = 1)


n upper bound hit
(n = 18)


power bound hit
(power =1)
unrestricted
(power = 0.272)
a Constant variance model selected (p = 0.1092).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.3.32.2. Output for Selected Model: Hill
Markowski et al. (2001): FR2 Revolutions
         Hill Model.  (Version:  2.14;  Date:  06/26/2008)
         Input Data File:  C:\l\34_Mark_2001_FR2rev_HillCV_l.(d)
         Gnuplot Plotting  File:   C:\l\34_Mark_2001_FR2rev_HillCV_l.plt
                                             Tue  Feb 16 18:16:03  2010
 Table  3


   The  form of the response  function is:

   Y[dose]  = intercept  +  v*doseAn/(kAn + doseAn)
   Dependent variable = Mean
   Independent variable =  Dose
   rho  is  set to 0
   Power parameter restricted to be greater  than 1
   A  constant variance model is fit

   Total number of dose groups = 4
   Total number of records with missing values  = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set  to:  le-008
   Parameter Convergence has been set to:  le-008

                                       G-526

-------
                  Default Initial Parameter Values
                          alpha =      2598.74
                            rho =
                      intercept =
                              v =
                              n =
                              k =
      0
 119.29
 -62.79
1.80602
  35.85
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
alpha intercept
alpha 1 -8.1e-009
intercept -8.1e-009 1
v 4.5e-008 -0.81
n -3e-005 -0.00013
k 3e-005 -0.0022
v
4.5e-008
-0.81
1
0.0002
0.0014
Parameter Estimates
Confidence Interval
Variable Estimate
Upper Conf. Limit
alpha 2183.85
3419.46
intercept 119.29
153.909
v -56.5223
-13.5831
n 18
17371.7
k 21.6708
1697.95
Table of Data and Estimated Values
Dose N Obs Mean Est Mean
Res .

Std. Err.

630.425

17.6629

21.9082

8854.08

855.263

of Interest
Obs Std Dev

                                                             n

                                                       -3e-005

                                                      -0.00013

                                                        0.0002

                                                             1

                                                            -1
                                   k

                              3e-005

                             -0.0022

                              0.0014

                                  -1

                                   1
                                                         95.0% Wald

                                                      Lower Conf. Limit

                                                              948.245

                                                              84.6713

                                                             -99.4615

                                                             -17335.7

                                                             -1654.61
                                    G-527

-------
    0
   20
   60
  180
        119
        109
       56.5
       68.1
 119
 108
62.8
62.8
69.9
  61
31.2
33.2
46.7
46.7
46.7
46.7
2.74e-008
8.42e-010
   -0.329
    0.304
Degrees of freedom for Test A3 vs fitted <= 0
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -104.165520
             -101.140174
             -104.165520
             -104.266162
             -107.599268
          # Param's
                5
                8
                5
                5
                2
            AIC
         218.331040
         218.280349
         218.331040
         218.532324
         219.198536
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
-2*log(Likelihood Ratio)  Test df
            12.9182
            6.05069
            6.05069
           0.201283
         6
         3
         3
         0
        p-value

      0.04435
       0.1092
       0.1092
           NA
                                     G-528

-------
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.  The modeled variance appears
 to be appropriate here

NA - Degrees of freedom for Test 4 are less than or equal to 0.  The Chi-
Square
     test for fit is not valid
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =        23.6366

            BMDL =       7.33648
                                     G-529

-------
G.3.32.3. Figure for Selected Model: Hill



                               Hill Model with 0.95 Confidence Level
        200
        150
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 0)
100
         50
                   Hill
             -BMDL

               BMD
                      20     40     60     80     100    120    140    160    180

                                            dose
   18:1602/162010
G.3.32.4. Output for Additional Model Presented: Power, Unrestricted

Markowski et al. (2001): FR2 Revolutions
         Power Model.  (Version:  2.15;  Date:  04/07/2008)

         Input Data File: C:\l\34_Mark_2001_FR2rev_PowerCV_U_l.(d)

         Gnuplot Plotting File:   C:\l\34_Mark_2001_FR2rev_PowerCV_U_l.plt

                                            Tue  Feb 16 18:16:04  2010
 Table  3





   The  form of the response  function is:



   Y[dose]  = control + slope  *  doseApower





   Dependent variable = Mean
                                       G-530

-------
   Independent variable = Dose
   rho is set to 0
   The power is not restricted
   A constant variance model is fit

   Total number of dose groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      2598.74
                            rho =
                        control =
                          slope =
                          power =
            0
       119.29
     -1.79436
     0.708231
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
control
slope
power


9
-1
-1
alpha
1
.7e-009
.9e-008
.6e-008
control
9.7e-009
1
-0.49
-0.28
slope
-1.9e-008
-0.49
1
0.96
power
-1.6e-008
-0.28
0.96
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha             2351
3681.17
        control          120.074
155.517
          slope         -14.1965
29.329
          power          0.27229
0.862913
Parameter Estimates



       Std.  Err.

         678.674

         18.0837

         22.2073

        0.301344
        95.0% Wald

     Lower Conf.  Limit

             1020.82

             84.6305

             -57.722

           -0.318334
                                     G-531

-------
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
    0
   20
   60
  180
        119
        109
       56.5
       68.1
 120
  88
76.8
61.7
69.9
  61
31.2
33.2
48.5
48.5
48.5
48.5
-0.0428
  0.846
  -1.02
  0.352
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -104.165520
             -101.140174
             -104.165520
             -105.151136
             -107.599268
          # Param's
                5
                8
                5
                4
                2
            AIC
         218.331040
         218.280349
         218.331040
         218.302271
         219.198536
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
-2*log(Likelihood Ratio)  Test df
            12.9182
            6.05069
                     p-value

                   0.04435
                    0.1092
                                     G-532

-------
   Test 3
   Test 4
6.05069
1.97123
0.1092
0.1603
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
                     The modeled variance appears
                     The model chosen seems
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 91.0145


            BMDL = 1.8e-013
                                     G-533

-------
G.3.32.5. Figure for Additional Model Presented: Power, Unrestricted



                                Power Model with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
         200
         150
         100
          50
                   Power
            BMDL
                       20     40     60     80     100    120    140    160    180

                                               dose
   18:1602/162010
                                         G-534

-------
G.3.33. Markowski et al. (2001): FR5 Run Opportunities
G.3.33.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hillb
Linear
Polynomial,
3 -degree
Power
Power, unrestricted0
Degrees of
freedom
2
2
1
0
1
2
2
2
1
X2 p-value
0.149
0.149
0.303
N/A
0.939
0.091
0.091
0.091
0.133
AIC
133.830
133.830
133.087
134.032
132.032
134.825
134.825
134.825
134.281
BMD
(ng/kg-day)
9.491E+01
9.491E+01
2.961E+01
2.871E+01
2.214E-K)!
1.349E+02
1.349E+02
1.349E+02
3.721E+01
BMDL
(ng/kg-day)
4.324E+01
4.324E+01
9.356E+00
1.226E+01
1.117E-K)!
8.118E+01
8.118E+01
8.118E+01
1.439E-07
Notes

power hit bound (d = 1)


n upper bound hit
(n = 18)


power bound hit
(power =1)
unrestricted
(power =0.336)
a Constant variance model selected (p = 0.2262).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.3.33.2. Output for Selected Model: Hill
Markowski et al. (2001): FR5 Run Opportunities
         Hill Model.  (Version:  2.14;  Date:  06/26/2008)
         Input Data File:  C:\l\35_Mark_2001_FR5opp_HillCV_l.(d)
         Gnuplot Plotting  File:   C:\l\35_Mark_2001_FR5opp_HillCV_l.plt
                                            Tue  Feb 16 18:16:39 2010
 Table  3


   The  form of the response  function is:

   Y[dose]  = intercept +  v*doseAn/(kAn + doseAn)
   Dependent variable = Mean
   Independent variable =  Dose
   rho  is  set to 0
   Power parameter restricted to be greater  than 1
   A  constant variance model is fit

   Total number of dose groups = 4
   Total number of records  with missing values  = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to:  le-008
   Parameter Convergence has been set to: le-008

                                       G-535

-------
                  Default Initial Parameter Values
                          alpha =      77.4849
                            rho =
                      intercept =
                              v =
                              n =
                              k =
      0
  26.14
 -13.34
2.36002
35.0654
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho    -n
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
                                                      3.6e-008

                                                         -0.51

                                                          0.36

                                                             1
alpha
alpha 1
intercept -3.6e-009
v 9.8e-009
k 3.6e-008

Confidence Interval
Variable
Upper Conf. Limit
alpha
101.129
intercept
32.0935
v
-5.77257
n
k
26.8517
intercept v
-3.6e-009 9.8e-009
1 -0.81
-0.81 1
-0.51 0.36
Parameter Estimates

Estimate Std. Err.

64.5863 18.6445

26.14 3.03753

-13.1569 3.7676

18 NA
21.5963 2.68136

                                                         95.0% Wald

                                                      Lower Conf. Limit

                                                              28.0438

                                                              20.1865

                                                             -20.5413


                                                              16.3409
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
                                    G-536

-------
    0
   20
   60
  180
       26.1
       23.5
       12.8
       13.1
26.1
23.5
  13
  13
12.3
7.04
6.17
7.14
,04
,04
,04
,04
 1.02e-008
-1.39e-007
   -0.0558
    0.0517
 Model Descriptions for likelihoods calculated


 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e (i) } = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -62.013133
              -59.839035
              -62.013133
              -62.016024
              -67.530040
          # Param's
                5
                8
                5
                4
                2
            AIC
         134.026266
         135.678070
         134.026266
         132.032049
         139.060081
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among  Dose  levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs.  fitted)
 (Note:   When rho=0 the results of Test 3 and Test  2  will  be  the same.
   Test

   Test 1
   Test 2
   Test 3
   Test 4
                     Tests of Interest
-2*log(Likelihood Ratio)  Test df
             15.382
             4.3482
             4.3482
          0.0057833
         6
         3
         3
         1
        p-value

      0.01748
       0.2262
       0.2262
       0.9394
The p-value for Test 1 is less than  .05.  There appears to be  a

                                     G-537

-------
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
The modeled variance appears
The model chosen seems
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =         22.144

            BMDL =        11.165
                                     G-538

-------
G.3.33.3. Figure for Selected Model: Hill

                              Hill Model with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 OJ
       40
        35
        30
       25
20
        15
        10
                 Hill
             BMDL
              BMD
                    20     40     60     80     100    120    140    160     180
                                           dose
   18:1602/162010
G.3.33.4. Output for Additional Model Presented: Power, Unrestricted
Markowski et al. (2001): FR5 Run Opportunities
         Power Model.  (Version:  2.15;   Date: 04/07/2008)
         Input Data File: C:\l\35_Mark_2001_FR5opp_PwrCV_U_l.(d)
         Gnuplot Plotting File:   C:\l\35_Mark_2001_FR5opp_PwrCV_U_l.plt
                                            Tue Feb  16  18:16:40 2010
 Table  3


   The  form of the response  function is:

   Y[dose]  = control + slope  *  doseApower
   Dependent  variable = Mean
   Independent variable = Dose
                                      G-539

-------
   rho is set to 0
   The power is not restricted
   A constant variance model is fit

   Total number of dose groups = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      77.4849
                            rho =
                        control =
                          slope =
                          power =
            0
        26.14
     -0.39517
     0.725538
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
alpha
alpha 1
control 7.4e-009
slope 4.3e-008
power 4.8e-008
control slope power
7.4e-009 4.3e-008 4.8e-008
1 -0.51 -0.34
-0.51 1 0.97
-0.34 0.97 1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          70.9323
111.065
        control          26.3567
32.492
          slope         -2.49841
3.71437
          power         0.336003
0.810375
Parameter Estimates



       Std.  Err.

         20.4764

         3.13032

         3.16984

        0.242031
        95.0% Wald

     Lower Conf.  Limit

             30.7993

             20.2213

            -8.71118

           -0.138368
     Table of Data and Estimated Values of Interest

                                     G-540

-------
 Dose
Res .
                 Obs Mean
                              Est Mean
                              Obs Std Dev  Est Std Dev
                                     Scaled
    0
   20
   60
  180
       26.1
       23.5
       12.8
       13.1
26.4
19.5
16.5
12.1
12.3
7.04
6.17
7.14
,42
,42
,42
,42
-0.0681
  0.945
  -1.07
  0.341
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -62.013133
              -59.839035
              -62.013133
              -63.140714
              -67.530040
          # Param's
                5
                8
                5
                4
                2
            AIC
         134.026266
         135.678070
         134.026266
         134.281428
         139.060081
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
-2*log(Likelihood Ratio)  Test df
             15.382
             4.3482
             4.3482
         6
         3
         3
        p-value

      0.01748
       0.2262
       0.2262
                                     G-541

-------
   Test 4
2.25516
0.1332
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
                     The modeled variance appears
                     The model chosen seems
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 37.2131


            BMDL = 1.43926e-007
                                     G-542

-------
G.3.33.5. Figure for Additional Model Presented: Power, Unrestricted


                               Power Model with 0.95 Confidence Level
 c
 (0
 OJ
        40
        35
        30
       25

 o
 Q.

 a:      20
        15
        10
                  Power
          BMDL
BMD
                      20      40     60     80     100     120     140    160    180

                                              dose
   18:1602/162010
                                         G-543

-------
G.3.34. Miettinen et al. (2006): Cariogenic Lesions, Pups
G.3.34.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage, 4 -degree
Probit
Weibull
Gamma, unrestricted
Log-logistic,
unrestricted13
Log-probit, unrestricted
Weibull, unrestricted
Degrees of
freedom
3
3
3
3
3
3
o
J
2
2
2
2
x2
p-value
0.345
0.315
0.506
0.257
0.345
0.299
0.345
0.797
0.723
0.726
0.761
AIC
162.699
162.909
161.767
163.393
162.699
163.031
162.699
161.805
161.998
161.987
161.897
BMD
(ng/kg-day)
7.505E+01
8.991E+01
3.130E-K)!
1.390E+02
7.505E+01
9.941E+01
7.505E+01
1.591E-02
3.713E-01
5.098E-01
1.174E-01
BMDL
(ng/kg-day)
4.086E+01
5.250E+01
1.054E-K)!
6.729E+01
4.086E+01
6.208E+01
4.086E+01
1.335E-240
error
error
error
Notes
power bound hit
(power =1)

slope bound hit
(slope = 1)
slope bound hit
(slope =1)
final B = 0

power bound hit
(power =1)
unrestricted
(power =0.184)
unrestricted
(slope = 0.403)
unrestricted
(slope = 0.25)
unrestricted
(power =0.281)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.


G.3.34.2. Output for Selected Model: Log-Logistic
Miettinen et al. (2006): Cariogenic Lesions, Pups
         Logistic Model.  (Version: 2.12; Date:  05/16/2008)
         Input Data File:  C:\l\36_Miet_2006_Cariogenic_LogLogistic_l.(d)
         Gnuplot Plotting  File:
C:\l\36_Miet_2006_Cariogenic_LogLogistic_l.pit
                                             Tue  Feb 16 18:17:16  2010
 Table  2  converting the percentage into the  number of animals,  and control is
Control II from the study.  Dose is in ng per kg and is from Table 1
   The  form of the probability function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose) ) ]
   Dependent variable =  DichEff
   Independent variable  =  Dose
                                       G-544

-------
   Slope parameter is restricted as slope >= 1

   Total number of observations = 5
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =     0.595238
                      intercept =     -5.52519
                          slope =            1
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -slope
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

             background    intercept

background            1        -0.64

 intercept        -0.64            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background         0.658158

      intercept         -5.64068

          slope                1
                                 Parameter Estimates
                      Std.  Err.
                                                         95.0% Wald

                                                      Lower Conf. Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.4911
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -77.6769         5
     -78.8837         2       2.41374      3
                                     G-545
                                                                    P-value

-------
  Reduced model
0.0259

           AIC:
    -83.2067
     161.767
             11.0597
Goodness
of Fit
Scaled

0
30
100
300
1000
Dose
.0000
.0000
.0000
.0000
.0000
Est
0.
0.
0.
0.
0.
. Prob.
6582
6911
7477
8345
9249
Expected
27
20
18
20
29
.643
.041
.693
.027
.596
Observed
25.
23.
19.
20.
29.
000
000
000
000
000
Size
42
29
25
24
32
Residual
-0
1
0
-0
-0
.860
.189
.141
.015
.400
 ChiA2 =2.33
d.f. = 3
P-value = 0.5062
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        31.2951

            BMDL =        10.5354
                                     G-546

-------
 T3
 £
 O
 c
 O
 •*=
 O
 (0
G.3.34.3. Figure for Selected Model: Log-Logistic

                            Log-Logistic Model with 0.95 Confidence Level
                          i ^
          1
         0.9
         0.8  -
         0.7  -
Log-Logistic
         0.6  -
         0.5
         0.4 BMDL  BMD
                          200
            400         600
                  dose
800
1000
   18:1702/162010
G.3.34.4. Output for Additional Model Presented: Log-Logistic, Unrestricted
Miettinen et al. (2006): Cariogenic Lesions, Pups
         Logistic Model.  (Version:  2.12; Date:  05/16/2008)
         Input Data File: C:\l\36_Miet_2006_Cariogenic_LogLogistic_U_l.(d)
         Gnuplot Plotting File:
C:\l\36_Miet_2006_Cariogenic_LogLogistic_U_l.pit
                                            Tue  Feb 16 18:17:18 2010
 Table 2  converting the percentage into the number of animals,  and control is
Control II  from the study.  Dose is in ng per  kg  and is from Table  1
   The  form of the probability function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose) ) ]
                                       G-547

-------
   Dependent variable = DichEff
   Independent variable = Dose
   Slope parameter is not restricted

   Total number of observations = 5
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =     0.595238
                      intercept =     -1.68849
                          slope =     0.382632
           Asymptotic Correlation Matrix of Parameter Estimates

             background    intercept        slope

background            1        -0.41         0.24

 intercept        -0.41            1        -0.96

     slope         0.24        -0.96            1
                                 Parameter Estimates
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background         0.597778
*
      intercept         -1.79836
*
          slope         0.402606
                      Std.  Err.
   95.0% Wald

Lower Conf.  Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.7247
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -77.6769         5
     -77.9988         3      0.643944      2
                                     G-548
              P-value

-------
  Reduced model
0.0259

           AIC:
    -83.2067
     161.998
             11.0597
Goodness
of Fit
Scaled

0
30
100
300
1000
Dose
.0000
.0000
.0000
.0000
.0000
Est
0.
0.
0.
0.
0.
. Prob.
5978
7564
8045
8480
8905
Expected
25
21
20
20
28
.107
.936
.112
.351
.495
Observed
25.
23.
19.
20.
29.
000
000
000
000
000
Size
42
29
25
24
32
Residual
-0
0
-0
-0
0
.034
.460
.561
.200
.286
 ChiA2 =0.65
d.f. = 2
P-value = 0.7227
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =       0.371315

           Benchmark dose computation failed.  Lower limit includes zero.
                                     G-549

-------
 T3

 £
 O
 c
 O
 •*=
 O
 (0
G.3.34.5. Figure for Additional Model Presented: Log-Logistic, Unrestricted



                                            Log-Logistic Model



            1                 ' "" ' """"*'"
          0.9
          0.8
          0.7
          0.6
          0.5
          0.4
Log-Logistic
                              200
              400           600

                    dose
800
1000
   18:1702/162010
                                            G-550

-------
G.3.35. Murray et al. (1979): Fertility in F2 Generation
G.3.35.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Multistage, 1 -degree
Multistage,
2-degreea
Probit
Weibull
Log-probit,
unrestricted
Degrees of
freedom
0
1
0
1
1
1
0
0
X2 p-value
N/A
0.072
N/A
0.053
0.094
0.070
N/A
N/A
AIC
61.729
60.497
61.729
61.644
59.935
60.613
61.729
61.729
BMD
(ng/kg-day)
7.016E+00
4.007E+00
7.902E+00
2.380E+00
4.548E+00
3.707E+00
8.115E+00
6.373E+00
BMDL
(ng/kg-day)
1.698E+00
2.836E+00
1.584E+00
1.320E+00
1.635E-K)0
2.615E+00
1.698E+00
1.503E+00
Notes







unrestricted
(slope = 2.306)
a Best-fitting model, BMDS output presented in this appendix.


G.3.35.2. Output for Selected Model: Multistage, 2-Degree
Murray et al. (1979): Fertility in F2 Generation
        Multistage Model. (Version: 3.0;   Date:  05/16/2008)
        Input  Data File: C:\l\Murray_1979_fert_index_f2_Multi2_l.(d)
        Gnuplot Plotting File:  C:\l\Murray_1979_fert_index_f2_Multi2_l.plt
                                            Tue  Feb 16 20:08:06 2010
 Table  1 but  expressed as number of dams who  do  not produce offspring
   The  form of  the probability function  is:

   P[response]  = background +  (1-background)*[1-EXP(
                  -betal*doseAl-beta2*doseA2)]

   The  parameter betas are restricted to be  positive
   Dependent  variable = DichEff
   Independent  variable = Dose

 Total number of observations = 3
 Total number of records with missing values
 Total number of parameters in model =  3
 Total number of specified parameters = 0
 Degree  of  polynomial = 2
= 0
 Maximum  number of iterations = 250
 Relative Function Convergence has been  set  to:  le-008
 Parameter Convergence has been set to:  le-008

                                      G-551

-------
                  Default Initial Parameter Values
                     Background =    0.0624181
                        Beta(l) =            0
                        Beta(2) =   0.00532688
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -Beta(l)
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

             Background      Beta(2)

Background            1        -0.44

   Beta(2)        -0.44            1
Confidence Interval
       Variable
Upper Conf. Limit
     Background
*
        Beta(l)
*
        Beta(2)
               Parameter Estimates

                                       95.0% Wald

      Estimate        Std. Err.     Lower Conf. Limit

     0.0772201            *                *

             0            *                *

    0.00509404            *                *
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.03821
  Reduced model
0.0002798

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -25.8194         3
     -27.9673         2       4.29584      1
                            P-value
     -34.0009
      59.9347
1
16.363
     Dose
              Est. Prob.
                Goodness  of  Fit

          Expected    Observed     Size
                         Scaled
                        Residual
                                     G-552

-------
    0.0000     0.0772         2.471     4.000          32        1.013
    1.0000     0.0819         1.638     0.000          20       -1.336
   10.0000     0.4455         8.911     9.000          20        0.040

 ChiA2 = 2.81      d.f. = 1        P-value = 0.0936
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        4.54787

            BMDL =        1.63487

            BMDU =        6.79105

Taken together,  (1.63487, 6.79105) is a 90     % two-sided confidence
interval for the BMD
                                     G-553

-------
G.3.35.3. Figure for Selected Model: Multistage, 2-Degree


                                Multistage Model with 0.95 Confidence Level


               ' ' '  '      ' Mul'tis
          0.7




          0.6




          0.5
 T3

 £
 O
 c
 O
 •*=
 O
 (0
0.4
0.3   -  -T-
          0.2
          0.1
   20:0802/162010
G.3.36. National Toxicology Program (1982): Toxic Hepatitis, Male Mice

G.3.36.1. Summary Table of BMDS Modeling Results
                                                                                   10
Model
Gamma
Logistic
Log-logistic
Log-probit
Multistage, 3-degreea
Probit
Weibull
Degrees of
freedom
1
2
1
1
1
2
1
/2/7-value
0.026
0.093
0.027
0.027
0.028
0.088
0.026
AIC
113.097
110.712
113.093
113.111
112.555
110.696
113.056
BMD
(ng/kg-day)
.552E+01
.769E+01
.499E+01
.360E+01
1.488E+01
.564E+01
.619E+01
BMDL
(ng/kg-day)
5.155E+00
1.383E+01
6.628E+00
7.237E+00
4.676E-K)0
1.261E+01
4.903E+00
Notes







' Best-fitting model, BMDS output presented in this appendix.
                                           G-554

-------
G.3.36.2. Output for Selected Model: Multistage, 3-Degree
National Toxicology Program (1982): Toxic Hepatitis, Male Mice
        Multistage Model.  (Version:  3.0;   Date:  05/16/2008)
        Input Data File: C:\l\37_NTP_1982_ToxHep_Multi3_l.(d)
        Gnuplot Plotting File:  C:\l\37_NTP_1982_ToxHep_Multi3_l.plt
                                           Tue  Feb  16 18:17:51  2010
   The form of the probability  function  is:

   P[response] = background +  (1-background)*[1-EXP(
                 -betal*doseAl-beta2*doseA2-beta3*doseA3)

   The parameter betas are restricted  to be  positive
   Dependent variable = DichEff
   Independent variable = Dose

 Total number of observations =  4
 Total number of records with missing  values  =  0
 Total number of parameters in model = 4
 Total number of specified parameters  = 0
 Degree of polynomial = 3
 Maximum number of iterations = 250
 Relative Function Convergence has been  set  to:  le-008
 Parameter Convergence has been set  to:  le-008
                  Default Initial  Parameter  Values
                     Background =     0.0525767
                        Beta(l) =    0.00243254
                        Beta(2) =             0
                        Beta(3) =  5.29052e-006
           Asymptotic Correlation Matrix  of  Parameter Estimates

            ( *** The model parameter(s)   -Beta(2)
                 have been estimated  at a boundary point,  or have been
specified by the user,
                 and do not appear  in the correlation matrix )

             Background      Beta(l)       Beta(3)

Background            1        -0.69          0.66


                                     G-555

-------
   Beta(l)

   Beta(3)
-0.69

 0.66
    1

-0.98
-0.98

    1
                                 Parameter Estimates
Confidence Interval
       Variable
Upper Conf. Limit
     Background
*
        Beta(l)
*
        Beta(2)
*
        Beta(3)
      Estimate

     0.0383474

    0.00605732

             0

  4.60855e-006
         Std. Err.
             95.0% Wald

          Lower Conf. Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.03534
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -51.0633         4
     -53.2776         3       4.42854      1
     -121.743

      112.555
         1
    141.358
3
                                     P-value
<.0001
                                  Goodness  of  Fit

Dose
0.0000
1.4000
7.1000
71.0000

Est. Prob.
0.0383
0.0465
0.0803
0.8798

Expected
2.799
2.278
3.937
43.990

Observed
1.000
5.000
3.000
44.000

Size
73
49
49
50
Scaled
Residual
-1.097
1.847
-0.492
0.004
 ChiA2 =4.86
 d.f. =1
    P-value = 0.0275
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        14.8848
                                     G-556

-------
             BMDL  =



             BMDU  =
                   4.67636



                   28.8293
Taken  together,  (4.67636, 28.8293) is  a  90

interval  for the  BMD
                                          %  two-sided  confidence
G.3.36.3. Figure for Selected Model: Multistage, 3-Degree



                              Multistage Model with 0.95 Confidence Level
 O


 I


 O

 13
 (0
         0.8
         0.6
0.4
         0.2
                                                                            70
   18:1702/162010
                                        G-557

-------
G.3.37. National Toxicology Program (2006): Alveolar Metaplasia
G.3.37.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage, 5 -degree
Probit
Weibull
Gamma, unrestricted
Log-logistic,
unrestricted13
Log-probit, unrestricted
Weibull, unrestricted
Degrees of
freedom
4
4
4
4
4
4
4
3
3
3
3
x2
p-value
0.001
0.001
0.409
0.001
0.001
0.001
0.001
0.407
0.739
0.727
0.586
AIC
340.127
358.346
312.970
340.296
340.127
362.181
340.127
314.135
312.487
312.543
313.176
BMD
(ng/kg-day)
2.240E+00
4.997E+00
6.644E-01
3.291E+00
2.240E+00
5.656E+00
2.240E+00
2.211E-02
3.062E-01
3.316E-01
9.000E-02
BMDL
(ng/kg-day)
1.791E+00
4.149E+00
5.041E-01
2.517E+00
1.791E+00
4.810E+00
1.791E+00
8.081E-04
7.972E-02
8.968E-02
1.341E-02
Notes
power bound hit
(power =1)

slope bound hit
(slope = 1)
slope bound hit
(slope =1)
final B = 0

power bound hit
(power =1)
unrestricted
(power = 0.297)
unrestricted
(slope = 0.785)
unrestricted
(slope = 0.471)
unrestricted
(power = 0.465)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.


G.3.37.2. Output for Selected Model: Log-Logistic
National Toxicology Program (2006): Alveolar Metaplasia
         Logistic Model.  (Version: 2.12;  Date:  05/16/2008)
         Input Data  File:  C:\l\40_NTP_2006_AlvMeta_LogLogistic_l.(d)
         Gnuplot Plotting  File:  C:\l\40_NTP_2006_AlvMeta_LogLogistic_l.plt
                                             Tue Feb 16  18:19:30 2010
   The  form of the probability function  is:

   P[response] = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose) ) ]
    Dependent variable  =  DichEff
    Independent variable  = Dose
    Slope parameter is  restricted as  slope >= 1
                                       G-558

-------
   Total number of observations = 6
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =    0.0377358
                      intercept =     -2.03745
                          slope =            1
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -slope
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

             background    intercept

background            1         -0.4

 intercept         -0.4            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background        0.0448753
*
      intercept         -1.78837
*
          slope                1
                                 Parameter Estimates
                      Std.  Err.
                 95.0% Wald

              Lower Conf.  Limit
* - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.4424
  Reduced model
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -152.615         6
     -154.485         2        3.7393      4
     -216.802
1
128.374
                            P-value
<.0001
                                     G-559

-------
           AIC:
                         312.97
                                  Goodness  of  Fit

0
2
7
15
32
71
Dose
.0000
.1400
.1400
.7000
.9000
.4000
Est
0.
0.
0.
0.
0.
0.
. Prob.
0449
2966
5647
7366
8531
9262
Expe
2.
16.
29.
38.
45.
48.
cted
378
017
928
301
214
162
Ob
2.
19.
33.
35.
45.
46.
served
000
000
000
000
000
000
Size
53
54
53
52
53
52
Scaled
Residual
-0.251
0.889
0.851
-1.039
-0.083
-1.147
 ChiA2 = 3.
d.f. = 4
P-value = 0.4088
   Benchmark Dose Computation




Specified effect =            0.1




Risk Type        =      Extra risk




Confidence level =           0.95




             BMD =       0. 664411




            BMDL =       0.504109
                                     G-560

-------
G.3.37.3. Figure for Selected Model: Log-Logistic

                            Log-Logistic Model with 0.95 Confidence Level

          1
 T3
 £
 O
 c
 O
 •*=
 O
 (0
0.4
         0.2
          0  -
   18:1902/162010
                                                                         70
G.3.37.4. Output for Additional Model Presented: Log-Logistic, Unrestricted
National Toxicology Program (2006): Alveolar Metaplasia
         Logistic Model.  (Version:  2.12; Date:  05/16/2008)
         Input Data File: C:\l\40_NTP_2006_AlvMeta_LogLogistic_U_l.(d)
         Gnuplot Plotting File:   C:\l\40_NTP_2006_AlvMeta_LogLogistic_U_l.plt
                                            Tue  Feb 16 18:19:31 2010
   The  form of the probability function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose) ) ]
   Dependent variable = DichEff
                                       G-561

-------
   Independent variable = Dose
   Slope parameter is not restricted

   Total number of observations = 6
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =    0.0377358
                      intercept =     -1.26694
                          slope =     0.784484
           Asymptotic Correlation Matrix of Parameter Estimates

             background    intercept        slope

background            1        -0.24         0.11

 intercept        -0.24            1         -0.9

     slope         0.11         -0.9            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background        0.0375286
*
      intercept         -1.26811
*
          slope         0.785033
                                 Parameter Estimates
                      Std.  Err.
                 95.0% Wald

              Lower Conf.  Limit
* - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.7395
  Reduced model
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -152.615         6
     -153.244         3        1.2566      3
     -216.802
1
128.374
                            P-value
<.0001
                                     G-562

-------
           AIC:
                        312.487
                                  Goodness  of  Fit

0
2
7
15
32
71
Dose
.0000
.1400
.1400
.7000
.9000
.4000
Est
0.
0.
0.
0.
0.
0.
. Prob.
0375
3631
5845
7205
8207
8934
Expe
1.
19.
30.
37.
43.
46.
cted
989
609
980
468
498
455
Ob
2.
19.
33.
35.
45.
46.
served
000
000
000
000
000
000
Size
53
54
53
52
53
52
Scaled
Residual
0.008
-0.172
0.563
-0.763
0.538
-0.204
 ChiA2 = 1.26
d.f. = 3
P-value = 0.7388
   Benchmark Dose Computation




Specified effect =            0.1




Risk Type        =      Extra risk




Confidence level =           0.95




             BMD =       0.306194




            BMDL =      0.0797223
                                     G-563

-------
 T3

 £
 O
 c
 O
 •*=
 O
 (0
0.6
0.4
          0.2
G.3.37.5. Figure for Additional Model Presented: Log-Logistic, Unrestricted



                               Log-Logistic Model with 0.95 Confidence Level



           1
          0.8
                   Log-Logistic
            BMDLBMD
                 0        10       20        30       40        50        60       70

                                                 dose
   18:1902/162010
                                           G-564

-------
G.3.38. National Toxicology Program (2006): Eosinophilic Focus, Liver
G.3.38.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probit
Multistage, 5 -degree
Probit3
Weibull
Log-probit,
unrestricted
Degrees of
freedom
4
4
3
4
3
4
4
3
X2 p-value
0.367
0.167
0.117
0.084
0.313
0.187
0.367
0.087
AIC
330.457
333.343
334.148
334.683
331.771
332.962
330.457
334.849
BMD
(ng/kg-day)
5.676E+00
1.258E+01
4.727E+00
1.078E+01
6.568E+00
1.196E-K)!
5.675E+00
4.750E+00
BMDL
(ng/kg-day)
4.532E+00
1.071E+01
2.867E+00
8.514E+00
4.666E+00
1.031E+01
4.532E+00
1.757E+00
Notes
power bound hit
(power =1)





power bound hit
(power =1)
unrestricted
(slope = 0.643)
a Best-fitting model, BMDS output presented in this appendix.


G.3.38.2. Output for Selected Model: Probit
National Toxicology Program (2006): Eosinophilic Focus, Liver
         Probit  Model.  (Version: 3.1;   Date:  05/16/2008)
         Input Data File: C:\l\45_NTP_2006_LivEosFoc_Probit_l.(d)
         Gnuplot Plotting File:  C:\l\45_NTP_2006_LivEosFoc_Probit_l.plt
                                            Tue Feb 16 18:25:56 2010
   The  form of the probability function  is:

   P[response]  = CumNorm(Intercept+Slope*Dose) ,

   where  CumNorm(.)  is the cumulative normal  distribution function
   Dependent  variable = DichEff
   Independent  variable = Dose
   Slope parameter is not restricted

   Total number of observations = 6
   Total number of records with missing  values  = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set to:  le-008
   Parameter  Convergence has been set  to:  le-008
                   Default Initial  (and  Specified)  Parameter Values

                                      G-565

-------
                     background =
                      intercept =
                          slope =
                           0   Specified
                    -1.11935
                   0.0279665
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -background
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
              intercept

 intercept            1

     slope        -0.69
             slope

             -0.69

                 1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
      intercept         -1.06148
-0.847497
          slope        0.0269279
0.0333525
               Parameter Estimates



                      Std.  Err.

                       0.109177

                     0.00327788
                      95.0% Wald

                   Lower Conf. Limit

                          -1.27546

                         0.0205034
       Model
     Full model
   Fitted model
0.1456
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
      -161.07         6
     -164.481         2        6.8221      4
     -202.816

      332.962
     1
83.4925
                                 P-value
<.0001
                                  Goodness  of  Fit

0
2
7
15
32
71
Dose
.0000
.1400
.1400
.7000
.9000
.4000
Est
0.
0.
0.
0.
0.
0.
. Prob.
1442
1577
1924
2615
4303
8054
Exp
7
8
10
13
22
42
ected
.645
.517
.195
.860
.807
.688
Ob
3.
8.
14.
17.
22.
42.
served
000
000
000
000
000
000
Size
53
54
53
53
53
53
Scaled
Residual
-1.816
-0.193
1.326
0.982
-0.224
-0.239
 ChiA2 = 6.16
 d.f.  = 4
P-value = 0.1873
                                     G-566

-------
   Benchmark Dose  Computation



Specified effect =             0.1



Risk Type        =      Extra  risk



Confidence level =           0.95



              BMD =        11.9584



             BMDL =        10.3075





G.3.38.3. Figure for Selected Model: Probit



                               Probit Model with 0.95 Confidence Level
 T3

 £
 O
 c
 O
 •*=
 O
 (0
         0.8
         0.6
0.4
         0.2
                     Probit
                    BMDL BMD
                        10
                        20
30       40

    dose
50
60
70
   18:2502/162010
                                        G-567

-------
G.3.39. National Toxicology Program (2006): Fatty Change Diffuse, Liver
G.3.39.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probit
Multistage, 5 -degree
Probit
Weibull3
Log-probit,
unrestricted
Degrees of
freedom
4
4
4
4
4
4
4
4
X2 p-value
0.668
0.005
0.292
0.118
0.808
0.005
0.679
0.282
AIC
252.294
269.825
255.082
257.548
251.545
269.430
252.218
255.258
BMD
(ng/kg-day)
4.224E+00
1.092E+01
4.697E+00
6.236E+00
4.021E+00
1.052E+01
4.252E-KJO
4.581E+00
BMDL
(ng/kg-day)
3.166E+00
9.292E+00
3.153E+00
5.204E+00
3.250E+00
9.068E+00
3.174E+00
3.193E+00
Notes



slope bound hit
(slope = 1)



unrestricted
(slope = 0.824)
a Best-fitting model, BMDS output presented in this appendix.


G.3.39.2. Output for Selected Model: Weibull
National Toxicology Program (2006): Fatty Change Diffuse, Liver
        Weibull  Model using Weibull Model  (Version:  2.12;   Date: 05/16/2008)
        Input  Data File:  C:\l\47_NTP_2006_LivFatDiff_Weibull_l.(d)
        Gnuplot  Plotting File:  C:\l\47_NTP_2006_LivFatDiff_Weibull_l.plt
                                            Tue  Feb 16 18:26:57 2010
 NTP liver  fatty change diffuse


   The  form of  the probability function is:

   P[response]  = background +  (1-background)*[1-EXP(-slope*doseApower)]
   Dependent  variable = DichEff
   Independent  variable = Dose
   Power parameter is restricted as power  >=1

   Total number of observations = 6
   Total number of records with missing values  =  0
   Maximum  number of iterations = 250
   Relative Function Convergence has been  set to:  le-008
   Parameter  Convergence has been set to:  le-008
                   Default Initial  (and Specified)  Parameter Values
                      Background =   0.00925926
                           Slope =   0.00962604
                           Power =      1.28042
                                      G-568

-------
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -Background
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

Slope
Power
Slope
1
-0.97
Power
-0.97
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     Background                0
          Slope        0.0223474
0.0409874
          Power          1.07133
1.31071
                                 Parameter Estimates
                      Std.  Err.

                             NA
                     0.00951041

                       0.122134
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
                 95.0% Wald

              Lower Conf.  Limit


                    0.0037073

                     0.831952
       Model
     Full model
   Fitted model
0.6928
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -122.992         6
     -124.109         2       2.23388      4
     -204.846

      252.218
1
163.708
                            P-value
<.0001
                                  Goodness  of  Fit

0
2
7
15
32
71
Dose
.0000
.1400
.1400
.7000
.9000
.4000
Est
0.
0.
0.
0.
0.
0.
. Prob.
0000
0492
1677
3475
6107
8851
Expe
0.
2.
8.
18.
32.
46.
cted
000
659
889
420
365
909
Ob
0.
2.
12.
17.
30.
48.
served
000
000
000
000
000
000
Size
53
54
53
53
53
53
Scaled
Residual
0.000
-0.414
1.144
-0.409
-0.666
0.470
                                     G-569

-------
 ChiA2 =2.31
            d.f.  = 4
P-value  =  0.6785
   Benchmark Dose Computation


Specified effect =             0.1


Risk Type        =       Extra risk


Confidence level =            0.95


              BMD =         4.25219


             BMDL =        3.17375




G.3.39.3. Figure for Selected Model: Weibull



                              Weibull Model with 0.95 Confidence Level
 •

 I
 o
 13
 (0
          1
         0.8
         0.6
0.4
         0.2
                     Weibull
              BMDL
           BMD
                        10
                       20
   30      40

       dose
50
60
70
   18:2602/162010
                                       G-570

-------
G.3.40. National Toxicology Program (2006): Gingival Hyperplasia, Squamous, 2 Years
G.3.40.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage, 5 -degree
Probit
Weibull
Gamma, unrestricted
Log-logistic,
unrestricted13
Log-probit, unrestricted
Weibull, unrestricted
Degrees of
freedom
4
4
4
4
4
4
4
3
3
3
3
x2
p-value
0.012
0.008
0.015
0.003
0.012
0.008
0.012
0.651
0.675
0.688
0.663
AIC
318.867
320.908
317.969
323.633
318.867
320.687
318.867
307.529
307.416
307.354
307.471
BMD
(ng/kg-day)
2.295E+01
3.594E+01
1.838E-K)!
4.313E+01
2.295E+01
3.436E+01
2.295E+01
2.480E-01
3.710E-01
4.688E-01
3.076E-01
BMDL
(ng/kg-day)
1.417E+01
2.564E+01
1.044E-K)!
2.794E+01
1.417E+01
2.425E+01
1.417E+01
5.096E-09
1.505E-07
8.851E-07
3.210E-08
Notes
power bound hit
(power =1)

slope bound hit
(slope = 1)
slope bound hit
(slope =1)
final B = 0

power bound hit
(power =1)
unrestricted
(power =0.199)
unrestricted
(slope = 0.265)
unrestricted
(slope = 0.156)
unrestricted
(power =0.23)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.


G.3.40.2. Output for Selected Model: Log-Logistic
National Toxicology Program (2006): Gingival Hyperplasia, Squamous, 2 Years
         Logistic Model.  (Version: 2.12; Date:  05/16/2008)
         Input Data File:  C:\l\42_NTP_2006_GingHypSq_LogLogistic_l.(d)
         Gnuplot Plotting  File:   C:\l\42_NTP_2006_GingHypSq_LogLogistic_l.plt
                                             Tue  Feb 16 18:20:29  2010
  [insert study notes]
   The  form of the probability function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
   Dependent variable =  DichEff
   Independent variable  =  Dose
   Slope parameter is restricted as slope  >=  1
                                       G-571

-------
   Total number of observations = 6
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =    0.0188679
                      intercept =      -4.5509
                          slope =            1
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -slope
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

             background    intercept

background            1        -0.71

 intercept        -0.71            1
                                 Parameter Estimates
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background         0.117717
*
      intercept         -5.10866
*
          slope                1
                      Std.  Err.
                 95.0% Wald

              Lower Conf.  Limit
* - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.007076
  Reduced model
0.0001186
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
      -149.95         6
     -156.985         2       14.0696      4
     -162.631
1
                                     G-572
25.3627
                            P-value

-------
           AIC:
                        317.969
Goodness
of Fit
Scaled

0
2
7
15
32
71
Dose
.0000
.1400
.1400
.7000
.9000
.4000
Est
0.
0.
0.
0.
0.
0.
. Prob.
1177
1290
1542
1942
2641
3837
Expected
6.
6.
8.
10.
13.
20.
239
965
174
292
995
335
Observed
1.
7.
14.
13.
15.
16.
000
000
000
000
000
000
Size
53
54
53
53
53
53
Residual
-2
0
2
0
0
-1
.233
.014
.216
.940
.313
.225
 ChiA2 = 12.38
d.f.  =4
P-value = 0.0147
   Benchmark Dose Computation




Specified effect =            0.1




Risk Type        =      Extra risk




Confidence level =           0.95




             BMD =        18.3832




            BMDL =        10.4359
                                     G-573

-------
G.3.40.3. Figure for Selected Model: Log-Logistic

                            Log-Logistic Model with 0.95 Confidence Level
 T3
 £
 O
 c
 O
 •*=
 O
 (0
         0.4
         0.3
0.2
         0.1
                          Log-Logistic
          0  -
                               BMD
                        10
                       20
30      40
    dose
50
60
70
   18:2002/162010
G.3.40.4. Output for Additional Model Presented: Log-Logistic, Unrestricted
National Toxicology Program (2006): Gingival Hyperplasia, Squamous, 2 Years
         Logistic Model.  (Version: 2.12; Date:  05/16/2008)
         Input Data File:  C:\l\42_NTP_2006_GingHypSq_LogLogistic_U_l.(d)
         Gnuplot Plotting  File:
C:\l\42_NTP_2006_GingHypSq_LogLogistic_U_l.plt
                                             Tue  Feb 16 18:20:29  2010
  [insert  study notes]
   The  form of the probability function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose) ) ]
                                       G-574

-------
   Dependent variable = DichEff
   Independent variable = Dose
   Slope parameter is not restricted

   Total number of observations = 6
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =    0.0188679
                      intercept =     -2.04571
                          slope =     0.299277
           Asymptotic Correlation Matrix of Parameter Estimates

             background    intercept        slope

background            1         -0.3         0.12

 intercept         -0.3            1        -0.91

     slope         0.12        -0.91            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background        0.0185126
*
      intercept         -1.93464
*
          slope         0.264795
                                 Parameter Estimates
                      Std.  Err.
   95.0% Wald

Lower Conf. Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.6785
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
      -149.95         6
     -150.708         3        1.5163      3
              P-value
                                     G-575

-------
  Reduced model
0.0001186

           AIC:
    -162.631
     307.416
             25.3627
                                  Goodness  of  Fit

0
2
7
15
32
71
Dose
.0000
.1400
.1400
.7000
.9000
.4000
Est
0.
0.
0.
0.
0.
0.
. Prob.
0185
1659
2105
2447
2806
3219
Expe
0.
8.
11.
12.
14.
17.
cted
981
959
155
972
873
059
Ob
1.
7.
14.
13.
15.
16.
served
000
000
000
000
000
000
Size
53
54
53
53
53
53
Scaled
Residual
0.019
-0.717
0.959
0.009
0.039
-0.311
 ChiA2 = 1.53
d.f.  =3
P-value = 0.6750
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =       0.370958

            BMDL =   1.50494e-007
                                     G-576

-------
G.3.40.5. Figure for Additional Model Presented: Log-Logistic, Unrestricted



                               Log-Logistic Model with 0.95 Confidence Level
 T3

 £
 O
 c
 O
 •*=
 O
 (0
          0.4
          0.3
0.2
          0.1
                             Log-Logistic
            BMDL
        BMD
                 0        10       20        30       40        50        60       70

                                                 dose
   18:2002/162010
                                           G-577

-------
G.3.41. National Toxicology Program (2006): Hepatocyte Hypertrophy, 2 Years
G.3.41.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probit
Multistage, 5-degreea
Probit
Weibull
Gamma, unrestricted
Log-logistic,
unrestricted
Log-probit, unrestricted
Weibull, unrestricted
Degrees of
freedom
4
4
5
4
4
4
4
4
4
4
4
x2
p-value
0.001
0.001
0.010
0.001
<0.001
0.001
0.001
0.029
0.005
0.006
0.019
AIC
290.365
310.492
278.082
297.168
290.365
313.841
290.365
275.042
280.068
279.204
275.967
BMD
(ng/kg-day)
1.647E+00
4.315E+00
6.978E-01
2.930E+00
1.647E-KJO
4.564E+00
1.647E+00
error
6.672E-01
7.167E-01
3.709E-01
BMDL
(ng/kg-day)
1.340E+00
3.650E+00
5.454E-01
2.267E+00
1.340E+00
3.923E+00
1.340E+00
error
2.939E-01
3.322E-01
1.315E-01
Notes
power bound hit
(power =1)

slope bound hit
(slope = 1)
slope bound hit
(slope = 1)
final 15 = 0

power bound hit
(power =1)
unrestricted
(power = 0.478)
unrestricted
(slope = 0.984)
unrestricted
(slope = 0.594)
unrestricted
(power = 0.64)
a Best-fitting model, BMDS output presented in this appendix.


G.3.41.2. Output for Selected Model: Multistage, 5-Degree
National Toxicology Program (2006): Hepatocyte Hypertrophy, 2 Years
        Multistage Model.  (Version:  3.0;   Date: 05/16/2008)
        Input Data File: C:\l\43_NTP_2006_HepHyper_Multi5_l.(d)
        Gnuplot Plotting File:  C:\l\43_NTP_2006_HepHyper_Multi5_l.plt
                                            Tue Feb 16 18:21:00  2010
  [insert  study notes]
   The  form of the probability  function is:

   P[response]  = background +  (1-background)*[1-EXP(
                  -betal*doseAl-beta2*doseA2-beta3*doseA3-beta4*doseA4-
beta5*doseA5)]

   The  parameter betas are restricted to be  positive
   Dependent  variable = DichEff
   Independent variable = Dose
                                      G-578

-------
 Total number of observations = 6
 Total number of records with missing values
 Total number of parameters in model = 6
 Total number of specified parameters = 0
 Degree of polynomial = 5
                      = 0
 Maximum number of iterations = 250
 Relative Function Convergence has been set to: le-008
 Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                     Background =     0.232262
                        Beta(l) =     0.045074
                        Beta(2) =            0
                        Beta(3) =            0
                        Beta (4) =            0
                        Beta (5) = 2.59945e-010
           Asymptotic Correlation Matrix of Parameter Estimates

            ( *** The model parameter(s)  -Beta (2)    -Beta (3)    -Beta(4)
-Beta(5)
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

             Background      Beta(l)

Background            1        -0.64

   Beta(l)        -0.64            1
                                 Parameter Estimates
Confidence Interval
       Variable
Upper Conf. Limit
     Background
*
        Beta(l)
*
        Beta(2)
*
        Beta(3)
*
        Beta (4)
*
        Beta(5)
 Estimate

0.0541647

0.0639585

        0

        0

        0

        0
Std.  Err.
   95.0% Wald

Lower Conf. Limit
                                     G-579

-------
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
2.6361629e-005
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -129.986         6
     -143.183         2       26.3932      4
      -219.97

      290.365
     1
179.968
                                 P-value
<.0001
                                  Goodness  of  Fit

0
2
7
15
32
71
Dose
.0000
.1400
.1400
.7000
.9000
.4000
Est
0.
0.
0.
0.
0.
0.
. Prob.
0542
1752
4009
6535
8847
9902
Expe
2.
9.
21.
34.
46.
52.
cted
871
458
248
635
887
479
Ob
0.
19.
19.
42.
41.
52.
served
000
000
000
000
000
000
Size
53
54
53
53
53
53
Scaled
Residual
-1.742
3.416
-0.630
2.126
-2.532
-0.667
 ChiA2 = 26.48
 d.f.  = 4
P-value = 0.0000
   Benchmark Dose Computation
Specified effect =

Risk Type

Confidence level =

             BMD =

            BMDL =

            BMDU =
            0.1

      Extra risk

           0.95

        1.64733

        1.34007

         2.0581
Taken together, (1.34007, 2.0581 )  is a 90
interval for the BMD
                             % two-sided confidence
                                     G-580

-------
G.3.41.3. Figure for Selected Model: Multistage, 5-Degree


                                Multistage Model with 0.95 Confidence Level


               ""          Multis

            1
          0.8
 T3

 £
 O
 c
 O
 •*=
 O
 (0
0.6
0.4
          0.2
   18:21 02/162010
                                            G-581

-------
G.3.42. National Toxicology Program (2006): Necrosis, Liver
G.3.42.1. Summary Table of BMDS Modeling Results
Model
Logistic
Log-logistic
Log-probit
Multistage, 5 -degree
Probit
Weibull
Gamma, unrestricted
Log-logistic,
unrestricted
Log-probit,
unrestricted"
Weibull, unrestricted
Degrees of
freedom
4
4
4
4
4
4
3
3
3
3
x2
p-value
0.397
0.810
0.290
0.763
0.445
0.763
0.869
0.833
0.768
0.856
AIC
238.314
235.265
239.107
235.581
237.888
235.581
236.344
236.483
236.742
236.393
BMD
(ng/kg-day)
3.484E+01
1.791E+01
3.205E+01
2.019E+01
3.266E+01
2.019E+01
1.114E+01
1.112E+01
1.061E+01
1.117E+01
BMDL
(ng/kg-day)
2.842E+01
1.194E+01
2.382E+01
1.419E+01
2.637E+01
1.419E+01
3.487E+00
3.581E+00
3.498E+00
3.554E+00
Notes
negative intercept
(intercept = -2.601)
slope bound hit (slope =1)
slope bound hit (slope =1)
final B = 0

power bound hit
(power =1)
unrestricted
(power = 0.599)
unrestricted
(slope = 0.695)
unrestricted
(slope = 0.367)
unrestricted
(power = 0.64)
a Best-fitting model, BMDS output presented in this appendix.


G.3.42.2. Output for Selected Model: Log-Probit, Unrestricted
National Toxicology Program (2006): Necrosis, Liver
         Probit Model. (Version:  3.1;   Date:  05/16/2008)
         Input Data File: C:\l\50_NTP_2006_LivNec_LogProbit_U_l.(d)
         Gnuplot Plotting File:   C:\l\50_NTP_2006_LivNec_LogProbit_U_l.plt
                                            Tue Feb 16 18:34:31  2010
 NTP liver  necrosis
   The  form of the probability function  is:

   P[response]  = Background
                + (1-Background)  *  CumNorm(Intercept+Slope*Log(Dose)) ,

   where  CumNorm(.)  is the cumulative  normal distribution  function
   Dependent  variable = DichEff
   Independent variable = Dose
   Slope  parameter is not restricted

   Total  number of observations =  6
   Total  number of records with missing  values
   Maximum number of iterations =  250

                                      G-582
= o

-------
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial (and Specified) Parameter Values
                     background =    0.0188679
                      intercept =     -1.98094
                          slope =     0.316942
           Asymptotic Correlation Matrix of Parameter Estimates

             background    intercept        slope

background            1        -0.69         0.59

 intercept        -0.69            1        -0.97

     slope         0.59        -0.97            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background        0.0228339
0.0680734
      intercept         -2.14844
-1.11503
          slope         0.367034
0.639577
               Parameter Estimates



                      Std.  Err.

                      0.0230818

                       0.527256

                       0.139055
                 95.0% Wald

              Lower Conf.  Limit

                   -0.0224057

                     -3.18184

                    0.0944904
       Model
     Full model
   Fitted model
0.7733
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -114.813         6
     -115.371         3        1.1157      3
      -127.98

      236.742
1
26.3331
                            P-value
<.0001
     Dose     Est._Prob.

    0.0000     0.0228
                Goodness  of  Fit

          Expected    Observed     Size

            1.210     1.000          53

                  G-583
                         Scaled
                        Residual
                        -0.193

-------
2.1400
7.1400
15.7000
32.9000
71.4000
0.0529
0.0979
0.1475
0.2116
0.2968
2.858
5.187
7.819
11.215
15.729
4.000
4.000
8.000
10.000
17.000
                                                       54
                                                       53
                                                       53
                                                       53
                                                       53
                                              0.694
                                             -0.549
                                              0.070
                                             -0.409
                                              0.382
 ChiA2 = 1.14
d.f. = 3
P-value = 0.7678
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        10.6107

            BMDL =        3.49791
                                     G-584

-------
G.3.42.3. Figure for Selected Model: Log-Probit,  Unrestricted



                                LogProbit Model with 0.95 Confidence Level
 T3

 £
 O
 c
 O
 •*=
 O
 (0
          0.5
          0.4
          0.3
0.2
          0.1
                         LogProbit
               BMDL
                  BMD
                          10       20       30        40       50       60        70

                                                 dose
   18:3402/162010
                                           G-585

-------
G.3.43. National Toxicology Program (2006): Oval Cell Hyperplasia
G.3.43.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probit
Multistage, 5 -degree
Probit3
Weibull"
Degrees of
freedom
3
4
3
3
2
4
3
X2 p-value
0.072
0.069
0.039
0.068
0.066
0.112
0.075
AIC
199.446
199.875
202.012
200.421
198.641
198.166
198.690
BMD
(ng/kg-day)
8.970E+00
9.792E+00
9.708E+00
9.968E+00
5.424E+00
9.103E+00
7.712E+00
BMDL
(ng/kg-day)
5.499E+00
8.245E+00
7.247E+00
7.758E+00
3.514E+00
7.701E-H)0
4.692E+00
Notes







a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.
G.3.43.2. Output for Selected Model: Probit
National Toxicology Program (2006): Oval Cell Hyperplasia
         Probit Model.  (Version:  3.1;   Date:  05/16/2008)
         Input Data File: C:\l\53_NTP_2006_OvalHyper_Probit_l.(d)
         Gnuplot Plotting File:   C:\l\53_NTP_2006_OvalHyper_Probit_l.plt
                                            Tue Feb 16  19:51:52  2010
   The  form of the probability  function is:

   P[response]  = CumNorm(Intercept+Slope*Dose) ,

   where  CumNorm(.)  is the cumulative normal distribution  function
   Dependent variable = DichEff
   Independent variable = Dose
   Slope  parameter is not restricted

   Total  number of observations  =  6
   Total  number of records with  missing values = 0
   Maximum number of iterations  =  250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been  set  to: le-008
                   Default Initial  (and Specified) Parameter  Values
                      background  =             0   Specified
                       intercept  =      -1.92612
                           slope  =     0.0670004
                                      G-586

-------
           Asymptotic Correlation Matrix of Parameter Estimates
           (  *** The model parameter(s)  -background
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
              intercept

 intercept            1

     slope         -0.8
             slope

              -0.8

                 1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
      intercept         -1.82129
-1.489
          slope        0.0767832
0.0931523
               Parameter Estimates



                      Std.  Err.

                        0.16954

                     0.00835175
                      95.0% Wald

                   Lower Conf.  Limit

                          -2.15359

                          0.060414
       Model
     Full model
   Fitted model
0.0566
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -92.4898         6
     -97.0832         2       9.18683      4
     -210.191

      198.166
     1
235.402
                                 P-value
<.0001
                                  Goodness  of  Fit

0
2
7
15
32
71
Dose
.0000
.1400
.1400
.7000
.9000
.4000
Est
0.
0.
0.
0.
0.
0.
. Prob.
0343
0488
1015
2690
7596
9999
Exp
1
2
5
14
40
52
ected
.817
.633
.379
.258
.256
.993
Ob
0.
4.
3.
20.
38.
53.
served
000
000
000
000
000
000
Size
53
54
53
53
53
53
Scaled
Residual
-1.372
0.864
-1.082
1.779
-0.725
0.082
 ChiA2 =7.50
 d.f.  = 4
P-value = 0.1119
   Benchmark Dose Computation
                                     G-587

-------
Specified effect  =



Risk Type



Confidence level  =



              BMD  =



             BMDL  =
                        0.1



                 Extra risk



                       0.95



                     9.1026



                     7.7011
G.3.43.3. Figure for Selected Model: Probit



                                Probit Model with 0.95 Confidence Level
 T3

 £
 O
 c
 O
 •*=
 O
 (0
           1
         0.8
0.6
0.4
         0.2
                     Probit
                  BMDL
               BMD
       0       10       20       30      40       50       60

                                     dose
                                                                             70
   19:51 02/162010
                                        G-588

-------
G.3.43.4. Output for Additional Model Presented: Weibull
National Toxicology Program (2006): Oval Cell Hyperplasia
        Weibull Model using Weibull Model  (Version:  2.12;   Date:  05/16/2008)
        Input Data File: C:\l\53_NTP_2006_OvalHyper_Weibull_l.(d)
        Gnuplot Plotting File:  C:\l\53_NTP_2006_OvalHyper_Weibull_l.plt
                                           Tue  Feb  16 19:51:53 2010
   The form of the probability  function  is:

   P[response] = background +  (1-background)*[1-EXP(-slope*doseApower)]
   Dependent variable = DichEff
   Independent variable = Dose
   Power parameter is restricted as power  >=1

   Total number of observations =  6
   Total number of records with missing  values  =  0
   Maximum number of iterations =  250
   Relative Function Convergence has been  set to:  le-008
   Parameter Convergence has been  set  to:  le-008
                  Default Initial  (and  Specified)  Parameter Values
                     Background =    0.00925926
                          Slope =     0.0044452
                          Power =       1.63009
           Asymptotic Correlation Matrix  of  Parameter Estimates

             Background         Slope         Power

Background            1         -0.63          0.61

     Slope        -0.63             1         -0.99

     Power         0.61         -0.99             1
Confidence Interval
       Variable
Upper Conf. Limit
         Parameter Estimates

                                 95.0% Wald

Estimate        Std. Err.     Lower Conf. Limit



             G-589

-------
     Background
0.0601492
          Slope
0.0088166
          Power
2.37011
0.021258
0.0028715
1.76359
0.0198428
0.00303327
0.309457
-0.0176332
-0.0030736
1.15706
       Model
     Full model
   Fitted model
0.0524
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -92.4898         6
     -96.3448         3       7.70998      3
                                 P-value
Reduced model
AIC:
Dose
0.
2.
7.
15.
32.
71.
0000
1400
1400
7000
9000
4000
-210.191
198.69
Est. Prob.
0
0
0
0
0
0
.0213
.0320
.1073
.3234
.7490
.9953
1
Goodness
Expected
1
1
5
17
39
52
.127
.725
.685
.138
.698
.750
235.402
of Fit
Observed
0.
4.
3.
20.
38.
53.
000
000
000
000
000
000
Size
53
54
53
53
53
53
5 <.0
Scaled
Residual
-1
1
-1
0
-0
0
.073
.760
.192
.840
.538
.501
 ChiA2 = 6.92
 d.f.  = 3
P-value = 0.0746
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        7.71171

            BMDL =       4.69152
                                     G-590

-------
G.3.43.5. Figure for Additional Model Presented: Weibull


                                Weibull Model with 0.95 Confidence Level


              ""      ' Weibuil' '

           1
          0.8
 T3

 £
 O
 c
 O
 •*=
 O
 (0
0.6
0.4
          0.2
   19:51 02/162010
G.3.44. National Toxicology Program (2006): Pigmentation, Liver

G.3.44.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probita
Multistage, 5 -degree
Probit
Weibull
Degrees of
freedom
3
4
3
3
3
4
O
J
/2/7-value
0.385
<0.001
0.978
0.980
0.210
O.001
0.290
AIC
197.655
203.517
195.600
195.450
199.850
210.309
198.489
BMD
(ng/kg-day)
1.547E+00
2.259E+00
2.212E+00
2.072E+00
9.396E-01
2.259E+00
1.280E+00
BMDL
(ng/kg-day)
8.055E-01
1.872E+00
1.452E+00
1.399E+00
7.079E-01
1.916E+00
7.518E-01
Notes




final B = 0


' Best-fitting model, BMDS output presented in this appendix.
                                          G-591

-------
G.3.44.2. Output for Selected Model: Log-Probit
National Toxicology Program (2006): Pigmentation, Liver
        Probit Model.  (Version:  3.1;   Date:  05/16/2008)
        Input Data File: C:\l\54_NTP_2006_Pigment_LogProbit_l.(d)
        Gnuplot Plotting File:   C:\l\54_NTP_2006_Pigment_LogProbit_l.plt
                                           Tue  Feb  16  19:52:19  2010
   The form of the probability  function  is:

   P[response] = Background
               +  (1-Background)  * CumNorm(Intercept+Slope*Log(Dose) ) ,

   where CumNorm(.) is the cumulative normal  distribution function
   Dependent variable = DichEff
   Independent variable = Dose
   Slope parameter is restricted as  slope  >=  1

   Total number of observations =  6
   Total number of records with missing values  =  0
   Maximum number of iterations =  250
   Relative Function Convergence has been  set to:  le-008
   Parameter Convergence has been  set  to:  le-008
   User has chosen the log transformed model
                  Default Initial  (and  Specified)  Parameter Values
                     background =     0.0754717
                      intercept =     -1.91144
                          slope =       1.07385
           Asymptotic Correlation Matrix  of  Parameter  Estimates

             background    intercept         slope

background            1        -0.45          0.35

 intercept        -0.45             1         -0.94

     slope         0.35        -0.94             1



                                 Parameter Estimates

                                     G-592

-------
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background        0.0735956
0.140878
      intercept         -2.19294
-1.40885
          slope          1.25068
1.58335
                      Std. Err.

                      0.0343284

                       0.400053

                       0.169731
                      95.0% Wald

                   Lower Conf. Limit

                        0.00631316

                          -2.97703

                          0.918012
       Model
     Full model
   Fitted model
0.9753
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -94.6177         6
     -94.7248         3      0.214232      3
     -210.717

       195.45
     1
                                 P-value
<.0001
                                  Goodness  of  Fit

0
2
7
15
32
71
Dose
.0000
.1400
.1400
.7000
.9000
.4000
Est
0.
0.
0.
0.
0.
0.
. Prob.
0736
1729
6338
9023
9863
9992
Exp
3
9
33
47
52
52
ected
.901
.338
.591
.822
.275
.959
Ob
4.
9.
34.
48.
52.
53.
served
000
000
000
000
000
000
Size
53
54
53
53
53
53
Scaled
Residual
0.052
-0.122
0.117
0.082
-0.325
0.202
 ChiA2 = 0.18
 d.f.  = 3
P-value = 0.9801
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =        2.07241

            BMDL =        1.39932
                                     G-593

-------
G.3.44.3.  Figure for Selected Model: Log-Probit



                               LogProbit Model with 0.95 Confidence Level
 T3

 £
 O
 c
 O
 •*=
 O
 (0
           1
         0.8
0.6
         0.4
         0.2
                        LogProbit
             [iMDL
          BMD
                         10
                         20
30       40

    dose
50
60
70
   19:5202/162010
G.3.45. National Toxicology Program (2006): Toxic Hepatopathy


G.3.45.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic
Log-probit
Multistage, 5-degreea
Probit
Weibull
Degrees of
freedom
4
4
3
3
4
4
4
/2/7-value
0.772
0.012
0.362
0.378
0.577
0.019
0.745
AIC
185.634
198.445
190.061
189.858
186.521
197.159
185.657
BMD
(ng/kg-day)
4.668E+00
7.070E+00
5.676E+00
6.061E+00
4.163E+00
6.784E+00
4.454E+00
BMDL
(ng/kg-day)
3.317E+00
5.925E+00
4.040E+00
4.079E+00
2.701E-H)0
5.712E+00
3.159E+00
Notes




final 15 = 0


' Best-fitting model, BMDS output presented in this appendix.
                                          G-594

-------
G.3.45.2. Output for Selected Model: Multistage, 5-Degree
National Toxicology Program (2006): Toxic Hepatopathy
        Multistage Model.  (Version:  3.0;   Date:  05/16/2008)
        Input Data File: C:\l\55_NTP_2006_ToxHepa_Multi5_l.(d)
        Gnuplot Plotting File:  C:\l\55_NTP_2006_ToxHepa_Multi5_l.plt
                                           Tue  Feb  16  19:52:49 2010
   The form of the probability  function  is:

   P[response] = background +  (1-background)*[1-EXP(
                 -betal*doseAl-beta2*doseA2-beta3*doseA3-beta4*doseA4-
beta5*doseA5)]

   The parameter betas are restricted  to be  positive
   Dependent variable = DichEff
   Independent variable = Dose

 Total number of observations =  6
 Total number of records with missing  values  =  0
 Total number of parameters in model = 6
 Total number of specified parameters  = 0
 Degree of polynomial = 5
 Maximum number of iterations = 250
 Relative Function Convergence has been  set  to:  le-008
 Parameter Convergence has been set  to:  le-008
                  Default Initial  Parameter Values
                     Background =             0
                        Beta(l) =             0
                        Beta(2) =             0
                        Beta(3) =             0
                        Beta(4) =             0
                        Beta(5) =  5.40983e+010
           Asymptotic Correlation Matrix  of  Parameter Estimates

            ( *** The model parameter(s)   -Background    -Beta(3)     -Beta(4)
-Beta(5)
                 have been estimated  at a boundary point,  or have been
specified by the user,
                 and do not appear  in the correlation matrix )


                                     G-595

-------
   Beta (1)

   Beta(2)
Beta(l)

      1

  -0.91
Beta(2)

  -0.91

      1
Confidence Interval
       Variable
Upper Conf. Limit
     Background
*
        Beta(l)
*
        Beta(2)
*
        Beta (3)
*
        Beta (4)
*
        Beta(5)
        Estimate

               0

        0.019656

      0.00135796

               0

               0

               0
                                 Parameter Estimates
           Std. Err.
         95.0% Wald

      Lower Conf. Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.5737
  Reduced model

           AIC:
        Analysis of Deviance Table

  Log(likelihood)   # Param's  Deviance  Test d.f.
       -89.8076         6
       -91.2606         2       2.90597      4
       -218.207

        186.521
           1
256.799
                                       P-value
<.0001
                                  Goodness  of  Fit

0
2
7
15
32
71
Dose
.0000
.1400
.1400
.7000
.9000
.4000
Est
0.
0.
0.
0.
0.
0.
. Prob.
0000
0471
1891
4745
8796
9998
Expe
0.
2.
10.
25.
46.
52.
cted
000
545
021
146
616
987
Ob
0.
2.
8.
30.
45.
53.
served
000
000
000
000
000
000
Size
53
54
53
53
53
53
Scaled
Residual
0.000
-0.350
-0.709
1.335
-0.682
0.113
 ChiA2 =2.89
   d.f. =4
      P-value = 0.5771
   Benchmark Dose Computation
                                     G-596

-------
Specified  effect =

Risk Type

Confidence level =

              BMD =

             BMDL =

             BMDU =
                       0.1

                 Extra risk

                      0.95

                   4.16294

                   2.70063

                   6.00186
Taken together,  (2.70063,  6.00186) is  a  90
interval  for the BMD
                                         %  two-sided confidence
G.3.45.3. Figure for Selected Model: Multistage, 5-Degree

                             Multistage Model with 0.95 Confidence Level
 O

 I
 O
 13
 (0
          1
         0.8
         0.6
0.4
         0.2
                                                                          70
   19:5202/162010
                                       G-597

-------
G.3.46. Ohsako et al. (2001): Ano-Genital Length, PND 120
G.3.46.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hillb
Linear
Polynomial,
4-degree
Power
Hill, unrestricted0
Power, unrestricted
Degrees of
freedom
3
3
2
1
2
3
3
3
1
2
X2 p-value
0.019
0.019
0.117
0.049
0.148
0.018
0.018
0.018
0.055
0.151
AIC
171.804
171.804
168.204
169.789
167.727
171.954
171.954
171.954
169.600
167.689
BMD
(ng/kg-day)
5.650E+02
5.650E+02
2.854E+01
2.948E+01
3.722E-K)!
5.852E+02
5.852E+02
5.852E+02
5.101E+01
6.200E+01
BMDL
(ng/kg-day)
3.785E+02
3.785E+02
1.054E+01
1.135E+01
9.752E-KJO
4.047E+02
4.047E+02
4.047E+02
3.066E+00
2.291E+00
Notes

power hit bound (d = 1)


n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted (n = 0.502)
unrestricted
(power = 0.252)
a Constant variance model selected (p = 0.165).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.3.46.2. Output for Selected Model: Hill
Ohsako et al. (2001): Ano-Genital Length, PND 120
         Hill Model.  (Version:  2.14;  Date:  06/26/2008)
         Input Data File:  C:\l\56_Ohsako_2001_Anogen_HillCV_l.(d)
         Gnuplot Plotting  File:   C:\l\56_Ohsako_2001_Anogen_HillCV_l.plt
                                            Tue  Feb 16 19:53:25 2010
 Figure  7


   The  form of the response  function is:

   Y[dose]  = intercept + v*doseAn/(kAn + doseAn)
   Dependent variable = Mean
   Independent variable =  Dose
   rho  is  set to 0
   Power parameter restricted to be greater than  1
   A  constant variance model  is  fit

   Total number of dose groups = 5
   Total number of records  with  missing values  =  0
   Maximum number of iterations  = 250
   Relative Function Convergence has been set to:  le-008
   Parameter Convergence has  been set to: le-008

                                       G-598

-------
                  Default Initial Parameter Values
                          alpha =      7.27386
                            rho =
                      intercept =
                              v =
                              n =
                              k =
      0
 28.905
-5.1065
1.40226
33.9669
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho    -n
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
alpha
alpha 1
intercept -2.2e-009
v -2.4e-008
k -7.2e-009


Confidence Interval
Variable
Upper Conf. Limit
alpha
9.75666
intercept
30.4423
v
-2.86965
n
k
77.7767
intercept v
-2.2e-009 -2.4e-008
1 -0.66
-0.66 1
-0.5 -0.11
Parameter Estimates


Estimate Std. Err.

7.08444 1.3634

28.9809 0.745637

-4.79692 0.983318

1 NA
29.8628 24.4463

k
-7.2e-009
-0.5
-0.11
1

95.0%

Lower Co





-


-

                                                              4.41223

                                                              27.5195

                                                             -6.72418


                                                             -18.0511
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
                                    G-599

-------
   0
12.5
  50
 200
 800
12
10
10
10
12
28.9
27.9
25.2
  26
23.8
  29
27.6
  26
24.8
24.4
3.13
 2.5
3.21
2.85
1.56
2.66
2.66
2.66
2.66
2.66
-0.0988
  0.442
 -0.963
   1.42
 -0.726
Model Descriptions for likelihoods calculated
Model Al:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = SigmaA2

Model A2:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = Sigma(i)A2

Model A3:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = SigmaA2
    Model A3 uses any fixed variance parameters that
    were specified by the user

Model  R:         Yi = Mu + e(i)
           Var{e(i)} = SigmaA2
                      Likelihoods of Interest
           Model
            Al
            A2
            A3
        fitted
             R
              Log(likelihood)
                -77.952340
                -74.703868
                -77.952340
                -79.863340
                -89.824703
                       # Param's
                             6
                            10
                             6
                             4
                             2
                         AIC
                      167.904680
                      169.407736
                      167.904680
                      167.726680
                      183.649405
                  Explanation of Tests

Test 1:  Do responses and/or variances differ among Dose levels?
         (A2 vs. R)
Test 2:  Are Variances Homogeneous?  (Al vs A2)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)
Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
(Note:  When rho=0 the results of Test 3 and Test 2 will be the same.

                    Tests of Interest
  Test

  Test 1
  Test 2
  Test 3
  Test 4
  -2*log(Likelihood Ratio)   Test df
              30.2417
              6.49694
              6.49694
                3.822
                                  p-value

                              0.0001916
                                  0.165
                                  0.165
                                 0.1479
                                    G-600

-------
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
The modeled variance appears
The model chosen seems
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =        37.2249

            BMDL =       9.75249
                                     G-601

-------
G.3.46.3. Figure for Selected Model: Hill

                              Hill Model with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 OJ
31

30

29

28

27

26

25

24

23

22
                 Hill
          EMDL
BMD
              0      100     200     300     400     500     600     700     800
                                           dose
   19:5302/162010
G.3.46.4. Output for Additional Model Presented: Hill, Unrestricted
Ohsako et al. (2001): Ano-Genital Length, PND 120
        Hill  Model.  (Version:  2.14;   Date: 06/26/2008)
        Input Data File: C:\l\56_Ohsako_2001_Anogen_HillCV_U_l.(d)
        Gnuplot Plotting File:   C:\l\56_Ohsako_2001_Anogen_HillCV_U_l.plt
                                            Tue Feb  16 19:53:26  2010
 Figure 7


   The form  of  the response  function is:

   Y[dose] = intercept + v*doseAn/(kAn + doseAn)
   Dependent  variable = Mean
   Independent  variable = Dose
                                      G-602

-------
   rho is set to 0
   Power parameter is not restricted
   A constant variance model is fit

   Total number of dose groups = 5
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      7.27386
                            rho =
                      intercept =
                              v =
                              n =
                              k =
            0
       28.905
      -5.1065
      1.40226
      33.9669
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
intercept
v
n
k


2
-1
-1
1
alpha
1
.le-009
.8e-008
.7e-008
.6e-008
intercept
2. le-009
1
0.012
0.0075
-0.13
v
-1.8e-008
0.012
1
0.98
-0.99
n
-1.7e-008
0.0075
0.98
1
-0.97

k
1.6e-008
-0.
-0.
-0.

13
99
97
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          7.06785
9.73381
      intercept          28.9608
30.4413
              v         -6.94236
17.07
              n         0.501942
2.29563
Parameter Estimates



       Std.  Err.

         1.36021

        0.755363

         12.2514

        0.915162
        95.0% Wald

     Lower Conf.  Limit

             4.40189

             27.4803

            -30.9547

            -1.29174
                                     G-603

-------
2232.84
                         131.957
                                  1071.9
                                           -1968.92
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
    0
 12.5
   50
  200
  800
12
10
10
10
12
28.9
27.9
25.2
  26
23.8
  29
27.3
26.3
25.1
  24
3.13
 2.5
3.21
2.85
1.56
2.66
2.66
2.66
2.66
2.66
-0.0727
   0.72
  -1.37
   1.04
 -0.287
 Model Descriptions for likelihoods calculated


 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2


                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
              Log(likelihood)
                -77.952340
                -74.703868
                -77.952340
                -79.800035
                -89.824703
# Param' s
6
10
6
5
2
AIC
167.904680
169.407736
167.904680
169.600070
183.649405
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.
                                     G-604

-------
   Test
                     Tests of Interest
-2*log(Likelihood Ratio)   Test df
p-value
Test 1
Test 2
Test 3
Test 4
30.2417
6.49694
6.49694
3.69539
8
4
4
1
0.0001916
0.165
0.165
0.05456
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here


The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =        51.0107

            BMDL =       3.06631
                                     G-605

-------
G.3.46.5. Figure for Additional Model Presented: Hill, Unrestricted



                                  Hill Model with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
31




30




29




28




27




26




25




24




23




22
                   Hill
           Bli/IDL
BMD
                       100
                       200
                    300
 400

dose
500
600
700
800
   19:5302/162010
                                           G-606

-------
G.3.47. Sewall et al. (1995): T4 In Serum
G.3.47.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M5)
Hillb
Linear
Polynomial,
4-degree
Power
Hill, unrestricted0
Power, unrestricted
Degrees of
freedom
3
3
2
2
3
3
3
1
2
X2 p-value
0.424
0.424
0.611
0.702
0.332
0.332
0.332
0.844
0.983
AIC
205.966
205.966
206.152
205.875
206.584
206.584
206.584
207.205
205.200
BMD
(ng/kg-day)
5.762E+01
5.762E+01
2.523E+01
2.071E-K)!
6.788E+01
6.788E+01
6.788E+01
1.657E+01
1.658E+01
BMDL
(ng/kg-day)
3.783E+01
3.783E+01
8.442E+00
5.164E-KJO
4.858E+01
4.858E+01
4.858E+01
1.903E+00
1.820E+00
Notes

power hit bound (d=\)
power hit bound (d=\)
n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted (n = 0.427)
unrestricted
(power = 0.403)
a Constant variance model selected (p = 0.4078).
b Best-fitting model, BMDS output presented in this appendix.
0 Alternate model, BMDS output also presented in this appendix.
G.3.47.2. Output for Selected Model: Hill
Sewall et al. (1995): T4 In Serum
         Hill Model.  (Version:  2.14;  Date:  06/26/2008)
         Input Data File:  C:\l\58_Sewall_1995_T4_HillCV_l.(d)
         Gnuplot Plotting  File:   C:\l\58_Sewall_1995_T4_HillCV_l.plt
                                            Tue  Feb 16 19:54:30  2010
 Figure  1,  Saline noninitiated


   The  form of the response  function is:

   Y[dose]  = intercept +  v*doseAn/(kAn + doseAn)
   Dependent variable = Mean
   Independent variable =  Dose
   rho  is  set to 0
   Power parameter restricted to be greater  than 1
   A  constant variance model is fit

   Total number of dose groups = 5
   Total number of records  with missing values  = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set  to:  le-008
   Parameter Convergence has been set to:  le-008
                                       G-607

-------
                  Default Initial Parameter Values
                          alpha =      33.0913
                                                 Specified
      rho =
intercept =
        v =
        n =
        k =
       0
 30.6979
-12.2937
0.695384
 24.6674
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho    -n
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
alpha
alpha 1
intercept 1.2e-008
v 4.1e-008
k -2.4e-008


Confidence Interval
Variable
Upper Conf. Limit
alpha
42.2274
intercept
33.1899
v
-5.69537
n
k
105.905
intercept v
1.2e-008 4.1e-008
1 0.14
0.14 1
-0.66 -0.76
Parameter Estimates


Estimate Std. Err.

29.8807 6.29941

29.9609 1.64749

-14.2338 4.35645

1 NA
33.2198 37.0852

k
-2.4e-008
-0.66
-0.76
1

95.0% Wald

Lower Conf. Limit

17.5341

26.7319

-22.7723


-39.4658

NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
                                    G-608

-------
   0
 3.5
10.7
  35
 125
       30.7
       27.9
       25.9
       23.6
       18.4
  30
28.6
26.5
22.7
18.7
4.66
7.17
6.81
5.38
4.12
5.47
5.47
5.47
5.47
5.47
 0.404
-0.399
-0.328
 0.493
-0.171
Model Descriptions for likelihoods calculated
Model Al:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = SigmaA2

Model A2:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = Sigma(i)A2

Model A3:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = SigmaA2
    Model A3 uses any fixed variance parameters that
    were specified by the user

Model  R:         Yi = Mu + e(i)
           Var{e(i)} = SigmaA2
                      Likelihoods of Interest
           Model
            Al
            A2
            A3
        fitted
             R
            Log(likelihood)
              -98.583448
              -96.590204
              -98.583448
              -98.937315
             -109.013252
          # Param's
                6
               10
                6
                4
                2
            AIC
         209.166896
         213.180407
         209.166896
         205.874631
         222.026503
                  Explanation of Tests

Test 1:  Do responses and/or variances differ among Dose levels?
         (A2 vs. R)
Test 2:  Are Variances Homogeneous?  (Al vs A2)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)
Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
(Note:  When rho=0 the results of Test 3 and Test 2 will be the same.

                    Tests of Interest
  Test

  Test 1
  Test 2
  Test 3
  Test 4
-2*log(Likelihood Ratio)  Test df
            24.8461
            3.98649
            3.98649
           0.707735
                     p-value

                  0.001651
                    0.4078
                    0.4078
                     0.702
                                    G-609

-------
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
The modeled variance appears
The model chosen seems
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =        20.7117

            BMDL =       5.16405
                                     G-610

-------
G.3.47.3. Figure for Selected Model: Hill

                              Hill Model with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 OJ
       35
       30
       25
       20
       15
                 Hill
            BMDL
BMD
                       20
         40
60
dose
80
100
120
   19:5402/162010
G.3.47.4. Output for Additional Model Presented: Hill, Unrestricted
Sewall et al. (1995): T4 In Serum
        Hill  Model.  (Version:  2.14;   Date: 06/26/2008)
        Input Data File: C:\l\58_Sewall_1995_T4_HillCV_U_l.(d)
        Gnuplot Plotting File:   C:\l\58_Sewall_1995_T4_HillCV_U_l.plt
                                            Tue Feb  16  19:54:31 2010
 Figure 1,  Saline noninitiated


   The form of  the response function is:

   Y[dose]  = intercept + v*doseAn/(kAn + doseAn)
   Dependent  variable = Mean
   Independent  variable = Dose
                                      G-611

-------
   rho is set to 0
   Power parameter is not restricted
   A constant variance model is fit

   Total number of dose groups = 5
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =      33.0913
                                                 Specified
      rho =
intercept =
        v =
        n =
        k =
       0
 30.6979
-12.2937
0.695384
 24.6674
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
alpha
alpha
intercept
V
n
k

-0.
0.
0.
-0.
1
0004
0059
0048
0059
intercept
-0.0004
1
-0.026
-0.44
0.07
V
0.0059
-0.026
1
0.77
-1
n
0.0048
-0.44
0.77
1
-0.82
k
-0.0059
0.07
-1
-0.82
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha          29.4396
41.6042
      intercept          30.6757
34.155
              v         -141.324
2215.33
              n         0.426599
0.940515
           Parameter Estimates



                  Std.  Err.

                    6.20653

                    1.77521

                     1202.4

                   0.262207
                   95.0% Wald

                Lower Conf. Limit

                        17.2751

                        27.1963

                       -2497.98

                     -0.0873175
                                     G-612

-------
1.5415e+006
                           31487
                         770429
                         -1.47853e+006
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
    0
  3.5
 10.7
   35
  125
30.7
27.9
25.9
23.6
18.4
30.7
27.8
26.1
23.3
18.5
4.66
7.17
6.81
5.38
4.12
5.43
5.43
5.43
5.43
5.43
 0.0123
 0.0279
 -0.137
  0.132
-0.0354
 Model Descriptions for likelihoods calculated


 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2


                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
     Log(likelihood)
       -98.583448
       -96.590204
       -98.583448
       -98.602701
      -109.013252
# Param' s
6
10
6
5
2
AIC
209.166896
213.180407
209.166896
207.205403
222.026503
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.
                                     G-613

-------
   Test
                     Tests of Interest
-2*log(Likelihood Ratio)   Test df
p-value
Test 1
Test 2
Test 3
Test 4
24.8461
3.98649
3.98649
0.0385071
8
4
4
1
0.001651
0.4078
0.4078
0.8444
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is greater than  .1.  A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1.
 to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data
                                 The modeled variance appears
                                 The model chosen seems
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =        16.5689

            BMDL =       1.90347
                                     G-614

-------
G.3.47.5. Figure for Additional Model Presented: Hill, Unrestricted



                                  Hill Model with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
        35
        30
        25
        20
        15
                   Hill
           E!MDL
BMD
                          20
            40
60

dose
80
100
120
   19:5402/162010
                                           G-615

-------
G.3.48. Shi et al. (2007): Estradiol 17B, PE9
G.3.48.1. Summary Table of BMDS Modeling Results
Model
Exponential (M2)
Exponential (M3)
Exponential (M4)a
Exponential (M5)
Hill
Linear
Polynomial, 4 -degree
Power
Hill, unrestricted
Power, unrestricted
Degrees of
freedom
3
3
2
2
2
3
3
3
1
2
X2 p-value
0.001
0.001
0.494
0.494
0.773
0.001
0.001
0.001
0.874
0.506
AIC
395.701
395.701
383.635
383.635
382.743
397.484
397.484
397.484
384.251
383.589
BMD
(ng/kg-day)
1.729E+01
1.729E+01
5.559E-01
5.559E-01
4.434E-01
2.243E+01
2.243E+01
2.243E+01
3.998E-01
3.409E-01
BMDL
(ng/kg-day)
8.956E+00
8.956E+00
2.236E-01
2.236E-01
error
1.523E+01
1.523E+01
1.523E+01
error
5.002E-03
Notes

power hit bound (d=\)

power hit bound (d=\)
n lower bound hit (n = 1)


power bound hit
(power =1)
unrestricted (n = 0.616)
unrestricted
(power = 0.155)
a Best-fitting model, BMDS output presented in this appendix.


G.3.48.2. Output for Selected Model: Exponential (M4)
Shi et al. (2007): Estradiol 17B, PE9
         Exponential Model.  (Version:  1.61;   Date:  7/24/2009)
         Input  Data File: C:\l\59_Shi_2007_Estradiol_Exp_l.(d)
         Gnuplot Plotting File:
                                            Tue  Feb  16 19:55:06 2010
 Figure  4  PE9  only
   The  form of the response function by Model:
      Model 2:      Y[dose]  = a * exp{sign  *  b  *  dose}
      Model 3:      Y[dose]  = a * exp{sign  *  (b  * dose)Ad}
      Model 4:      Y[dose]  = a * [c-(c-l)  *  exp{-b * dose}]
      Model 5:      Y[dose]  = a * [c-(c-l)  *  exp{-(b *  dose)Ad}]

    Note:  Y[dose]  is the median response for exposure  = dose;
           sign = +1 for increasing trend in  data;
           sign = -1 for decreasing trend.

      Model 2  is nested within Models 3 and  4.
      Model 3  is nested within Model 5.
      Model 4  is nested within Model 5.
   Dependent  variable = Mean
   Independent  variable = Dose
   Data  are assumed to be distributed: normally
   Variance Model:  exp(lnalpha +rho *ln(Y[dose])

                                      G-616

-------
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  * rho)

Total number of dose groups = 5
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 4
                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                          2.65881
                         0.913414
                              108
                         0.136287
                         0.340136
                                1
                  Parameter Estimates

                Variable          Model 4
                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                         1.81331
                         1.12126
                         100.526
                         1.53823
                        0.431796
                               1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
  0.143
  0.714
   7.14
   28.6
10
10
10
10
10
102.9
86.19
63.33
48.1
38.57
41.41
19.58
29.36
18.82
22.59
   Dose

      0
  0.143
  0.714
   7.14
   28.6
               Estimated Values of Interest

             Est Mean      Est Std     Scaled Residual
    100.5
    89.25
    62.45
    43.41
    43.41
32.83
30.71
25.14
 20.5
 20.5
 0.2245
-0.3147
 0.1108
  0.723
-0.7458
                                  G-617

-------
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e (ij)  } = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)

     Model  R:        Yij = Mu + e(i)
               Var{e(ij)} = SigmaA2
                     Model
             Likelihoods of Interest

             Log(likelihood)       DF
                                                                AIC
Al
A2
A3
R
4
-188.3615
-183.667
-186.1132
-203.3606
-186.8176
6
10
7
2
5
388.7231
387.3339
386.2263
410.7211
383.6352
   Additive constant for all log-likelihoods =     -45.95.  This  constant
added to the
   above values gives the log-likelihood including the term that  does  not
   depend on the model parameters.
                                 Explanation of Tests

   Test 1:  Does response and/or variances differ among Dose  levels?  (A2  vs.
R)
   Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
   Test 3:  Are variances adequately modeled?  (A2 vs. A3)

   Test 6a: Does Model 4 fit the data?  (A3 vs  4)
     Test

     Test 1
     Test 2
     Test 3
    Test 6a
         Tests  of Interest

-2*log(Likelihood Ratio)
                                                  D. F.
p-value
39.39
9.389
4.892
1.409
8
4
3
2
< 0.0001
0.05208
0.1798
0.4944
     The p-value for Test 1 is less than  .05.  There appears  to be  a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

                                     G-618

-------
  The p-value for Test 2 is less than .1.  A non-homogeneous
  variance model appears to be appropriate.

  The p-value for Test 3 is greater than .1.  The modeled
  variance appears to be appropriate here.

  The p-value for Test 6a is greater than  .1.  Model 4 seems
  to adequately describe the data.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =     0.555948

              BMDL =     0.223612
                                 G-619

-------
G.3.48.3. Figure for Selected Model: Exponential (M4)



                           Exponential_beta Model 4 with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
         140
         120
         100
          80
          60
          40
          20
Exponential
            BVIDL BMD
                                       10
                        15

                     dose
20
25
30
   19:5502/162010
                                          G-620

-------
G.3.49. Smialowicz et al. (2008): PFC per 106 Cells
G.3.49.1. Summary Table of BMDS Modeling Results
Model
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear
Polynomial, 4 -degree
Power3
Hill, unrestricted
Power, unrestricted1"
Degrees
of
freedom
3
3
2
2
2
3
2
3
1
2
x2
^-value
0.048
0.048
0.019
0.019
0.026
0.016
0.0001
0.016
0.183
0.446
AIC
903.586
903.586
905.578
905.578
904.975
905.992
1,198.471
905.992
901.442
899.282
BMD
(ng/kg-day)
8.234E+01
8.234E+01
8.032E+01
8.032E+01
1.617E+01
1.450E+02
1.375E+03
1.450E+02
8.297E+00
7.676E+00
BMDL
(ng/kg-day)
4.833E+01
4.833E+01
6.220E+00
6.220E+00
2.214E+00
1.102E+02
3.331E+01
1.102E+02
4.172E-01
4.087E-01
Notes

power hit bound (d = 1)

power hit bound (d = 1)
n lower bound hit (n = 1)


power bound hit (power =1)
unrestricted (n = 0.266)
unrestricted (power = 0.249)
a Alternate model, BMDS output also presented in this appendix.
b Best-fitting model, BMDS output presented in this appendix.
G.3.49.2. Output for Selected Model: Power, Unrestricted
Smialowicz et al. (2008): PFC per 106 Cells
         Power Model.  (Version:  2.15;   Date: 04/07/2008)
         Input Data File: C:\l\60_Smial_2008_PFCcells_PwrCV_U_l.(d)
         Gnuplot Plotting File:   C:\l\60_Smial_2008_PFCcells_PwrCV_U_l.plt
                                            Tue Feb  16  19:55:53 2010
 Anti  Response to SRBCs, PFC per  10to6 cells, Table  4


   The form of the response function is:

   Y[dose]  = control + slope *  doseApower
   Dependent variable = Mean
   Independent variable = Dose
   rho  is  set to 0
   The  power is not restricted
   A  constant variance model  is  fit

   Total number of dose groups = 5
   Total number of records with  missing values = 0
   Maximum number of iterations  = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has  been set to: le-008
                                      G-621

-------
                  Default Initial Parameter Values
                          alpha =       232385
                            rho =
                        control =
                          slope =
                          power =
            0
         1491
     -384.362
     0.215085
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

alpha
control
slope
power
alpha
1
-1.5e-009
-8.2e-009
-l.le-008
control
-1.5e-009
1
-0.79
-0.65
slope
-8.2e-009
-0.79
1
0.96
power
-l.le-008
-0.65
0.96
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha           220294
294893
        control          1470.38
1713.55
          slope         -282.777
1.64025
          power         0.248621
0.416462
Parameter Estimates



       Std.  Err.

         38061.1

          124.07

         145.113

       0.0856348
        95.0% Wald

     Lower Conf.  Limit

              145696

             1227.21

            -567.193

           0.0807799
     Table of Data and Estimated Values of Interest
Dose
Res .
0
1.07
10.7
107
321
N Obs Mean Est Mean
15
14
15
15
8
1.49e+003
1.13e+003
945
677
161
1.47e+003
1.18e+003
961
567
283
Obs Std Dev
716
171
516
465
117
Est Std Dev
469
469
469
469
469
Scaled
0.17
-0.429
-0.129
0.91
-0.735
                                     G-622

-------
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e (i) } = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -444.832859
             -425.402825
             -444.832859
             -445.641102
             -463.753685
# Param' s
6
10
6
4
2
AIC
901.665718
870.805651
901.665718
899.282205
931.507371
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
-2*log(Likelihood Ratio)  Test df
            76.7017
            38.8601
            38.8601
            1.61649
 p-value

<.0001
<.0001
<.0001
0.4456
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1.
non-homogeneous variance model

The p-value for Test 3 is less than .1.
different variance model
                              Consider running a
                              You may want to consider a
                                     G-623

-------
The p-value for Test 4 is greater than .1.  The model chosen seems
to adequately describe the data
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 7.67564


            BMDL = 0.408661
                                     G-624

-------
G.3.49.3. Figure for Selected Model: Power, Unrestricted

                              Power Model with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 OJ
         1500
         1000
          500
            BMDL BMD
                         50       100       150      200
                                            dose
250
300
   19:5502/162010
G.3.49.4. Output for Additional Model Presented: Power
Smialowicz et al. (2008): PFC per 106 Cells
        Power Model.  (Version: 2.15;   Date:  04/07/2008)
        Input Data  File:  C:\l\60_Smial_2008_PFCcells_PwrCV_l.(d)
        Gnuplot  Plotting File:  C:\l\60_Smial_2008_PFCcells_PwrCV_l.plt
                                            Tue Feb 16 19:55:53  2010
 Anti Response  to  SRBCs,  PFC per 10to6  cells,  Table 4


   The form of  the response function is:

   Y[dose] =  control  + slope * doseApower
   Dependent variable = Mean
   Independent  variable = Dose
                                      G-625

-------
   rho is set to 0
   The power is restricted to be greater than or equal to 1
   A constant variance model is fit

   Total number of dose groups = 5
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                          alpha =       232385
                            rho =
                        control =
                          slope =
                          power =
                 0
              1491
          -2925.99
         -0.136613
Specified
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -rho    -power
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
                  alpha

     alpha            1

   control     3.6e-009

     slope    -1.2e-008
 control        slope

3.6e-009    -1.2e-008

       1        -0.53

   -0.53            1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
          alpha           250878
335833
        control          1176.24
1317.86
          slope         -3.45384
-2.29332
          power                1
     Parameter Estimates



            Std.  Err.

              43345.1

              72.2586

             0.592114

                   NA
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
        95.0% Wald

     Lower Conf.  Limit

              165923

             1034.61

            -4.61436
                                    G-626

-------
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
    0    15  1.49e+003    1.18e+003
 1.07    14  1.13e+003    1.17e+003
 10.7    15        945    1.14e+003
  107    15        677          807
  321     8        161         67.6
                                          716
                                          171
                                          516
                                          465
                                          117
                 501
                 501
                 501
                 501
                 501
                   2.43
                 -0.325
                   -1.5
                     -1
                  0.528
 Model Descriptions for likelihoods calculated


 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e (i) } = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
                    Log(likelihood)
                     -444.832859
                     -425.402825
                     -444.832859
                     -449.996183
                     -463.753685
# Param's
      6
     10
      6
      3
      2
   AIC
901.665718
870.805651
901.665718
905.992366
931.507371
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2 )
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.
                  Tests of Interest

Test    -2*log (Likelihood Ratio)  Test df

Test 1              76.7017          8

                                  G-627
                                                    p-value

                                                   <.0001

-------
   Test 2              38.8601          4          <.0001
   Test 3              38.8601          4          <.0001
   Test 4              10.3266          3         0.01598

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  Consider running a
non-homogeneous variance model

The p-value for Test 3 is less than .1.  You may want to consider a
different variance model

The p-value for Test 4 is less than .1.  You may want to try a different
model
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 145.02


            BMDL = 110.161
                                     G-628

-------
G.3.49.5. Figure for Additional Model Presented: Power



                                 Power Model with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
          1500
          1000
           500
                                                                            300
   19:5502/162010
                                          G-629

-------
G.3.50. Smialowicz et al. (2008): PFC per Spleen
G.3.50.1. Summary Table of BMDS Modeling Results
Model
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear
Polynomial, 4 -degree
Power3
Hill, unrestricted
Power, unrestricted1"
Degrees
of
freedom
3
3
3
2
2
3
3
3
2
2
x2
^-value
0.133
0.133
0.133
0.061
0.069
0.044
0.044
0.044
0.0001
0.230
AIC
377.395
377.395
377.395
379.395
379.150
379.895
379.895
379.895
441.885
376.738
BMD
(ng/kg-day)
.320E+02
.320E+02
.320E+02
.320E+02
.401E+02
2.151E+02
2.151E+02
2.151E+02
7.545E-23
9.374E+01
BMDL
(ng/kg-day)
8.431E+01
8.431E+01
8.184E+01
8.184E+01
error
1.704E+02
1.704E+02
1.704E+02
error
2.088E+01
Notes

power hit bound (d = 1)

power hit bound (d = 1)
n lower bound hit (n = 1)


power bound hit (power =1)
unrestricted (n = 0.038)
unrestricted (power = 0.418)
a Alternate model, BMDS output also presented in this appendix.
b Best-fitting model, BMDS output presented in this appendix.
G.3.50.2. Output for Selected Model: Power, Unrestricted
Smialowicz et al. (2008): PFC per Spleen
         Power Model.  (Version:  2.15;   Date: 04/07/2008)
         Input Data File: C:\l\61_Smial_2008_PFCspleen_Pwr_U_l.(d)
         Gnuplot Plotting File:   C:\l\61_Smial_2008_PFCspleen_Pwr_U_l.plt
                                            Tue Feb 16  19:56:26 2010
 Anti  Response to SRBCs - PFC  x  10  to the 4 per spleen,  Table 4


   The form of the response  function is:

   Y[dose]  = control + slope * doseApower
   Dependent variable = Mean
   Independent variable = Dose
   The  power is not restricted
   The  variance is to be modeled  as  Var(i)
= exp(lalpha  +  log(mean(i))  * rho)
   Total  number of dose groups  =  5
   Total  number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence  has  been set to: le-008
   Parameter Convergence has been set to: le-008
                   Default Initial  Parameter Values
                                      G-630

-------
                         lalpha =
                            rho =
                        control =
                          slope =
                          power =
      4.76607
            0
         27.8
     -7.21601
     0.213905
           Asymptotic Correlation Matrix of Parameter Estimates
lalpha
lalpha
rho
control
slope
power

-0
0
-0
-0
1
.98
.25
.27
.23
-0

-0
0
0
rho
.98
1
.31
.28
.23
control
0
-0

-0
-0
.25
.31
1
.81
.74
slope
-0
0
-0

0
.27
.28
.81
1
.99
power
-0
0
-0
0

.23
.23
.74
.99
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha         0.747155
2.75494
            rho          1.36972
2.06962
        control          25.1733
30.9193
          slope         -1.98465
1.5847
          power         0.417867
0.696048
Parameter Estimates



       Std.  Err.

          1.0244

        0.357098

         2.93169

         1.82113

        0.141932
   95.0% Wald

Lower Conf.  Limit

       -1.26063

        0.66982

        19.4273

         -5.554

       0.139686
     Table of Data and Estimated Values of Interest
Dose
Res .
0
1.07
10.7
107
321
N
15
14
15
15
8
Obs
27

17
12

Mean
.8
21
.6
.6
3
Est
25
23
19
11
3.
Mean
.2
.1
.8
.2
04
Obs St
13.
13.
9.
8.
3.
d Dev
4
6
4
7
1
Est S
13
12
11
7.
3.
td Dev
.2
.5
.2
59
11
Seal
0
-0
-0
0
-0.
ed
.769
.639
.768
.721
0353
                                     G-631

-------
 Model Descriptions for likelihoods calculated
 Model Al:        Yij
           Var{e(ij) }

 Model A2:        Yij
           Var{e(ij) }
             Mu(i)  + e(i j '
             SigmaA2

             Mu(i)  + e(i j '
             Sigma(i)A2
 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -190.565019
             -181.476284
             -181.900030
             -183.369059
             -204.636496
# Param' s
6
10
7
5
2
AIC
393.130038
382.952569
377.800059
376.738118
413.272993
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.
   Test

   Test 1
   Test 2
   Test 3
   Test 4
                     Tests of Interest
-2*log(Likelihood Ratio)  Test df
            46.3204
            18.1775
            0.84749
            2.93806
   p-value

  <.0001
0.001139
  0.8381
  0.2301
The p-value for Test 1 is less than  .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than  .1.
model appears to be appropriate
                              A non-homogeneous variance
The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is greater than  .1.  The model chosen seems

                                     G-632

-------
to adequately describe the data







               Benchmark Dose Computation




Specified effect =             1




Risk Type        =     Estimated standard deviations from the control mean




Confidence level =          0.95




             BMD = 93.7416







            BMDL = 20.8758
                                    G-633

-------
G.3.50.3. Figure for Selected Model: Power, Unrestricted

                             Power Model with 0.95 Confidence Level
       35  -
       30
       25
 o
 Q.
 o:
 c
 (0
 OJ
20  -
15
       10
   19:5602/162010
                                                                     300
G.3.50.4. Output for Additional Model Presented: Power
Smialowicz et al. (2008): PFC per Spleen
        Power Model.  (Version: 2.15;  Date:  04/07/2008)
        Input Data  File:  C:\l\61_Smial_2008_PFCspleen_Pwr_l.(d)
        Gnuplot  Plotting File:  C:\l\61_Smial_2008_PFCspleen_Pwr_l.plt
                                            Tue Feb 16 19:56:25  2010
 Anti Response  to  SRBCs - PFC x 10 to the  4  per spleen, Table  4


   The form of  the response function is:

   Y[dose] =  control  + slope * doseApower
   Dependent variable = Mean
   Independent  variable = Dose
                                      G-634

-------
   The power is restricted to be greater than or equal to 1
   The variance is to be modeled as Var(i)  = exp(lalpha + log(mean(i))

   Total number of dose groups = 5
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                                                   rho)
                  Default Initial Parameter Values
                         lalpha =
                            rho =
                        control =
                          slope =
                          power =
                 4.76607
                       0
                    27.8
                -54.5244
               -0.136501
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -power
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
lalpha
lalpha 1
rho -0.98
control 0.16
slope -0.48
rho
-0.98
1
-0.25
0.54
control slope
0.16 -0.48
-0.25 0.54
1 -0.88
-0.88 1
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
2.62213
            rho
2.24173
        control
24.674
          slope
-0.0450303
          power
  Estimate

  0.474614

   1.48709

   21.3571

-0.0574184

         1
Parameter Estimates



       Std.  Err.

         1.09569

        0.385029

         1.69233

      0.00632057

              NA
   95.0% Wald

Lower Conf.  Limit

        -1.6729

       0.732449

        18.0402

     -0.0698064
NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
                                     G-635

-------
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
    0
 1.07
 10.7
  107
  321
15
14
15
15
27.8
  21
17.6
12.6
   3
21.4
21.3
20.7
15.2
2.93
13.4
13.6
 9.4
 8.7
 3.1
12.3
12.3
12.1
 9.6
2.82
   2.02
-0.0898
  -1.01
  -1.05
 0.0745
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) +
           Var{e(ij)} = SigmaA2
 Model A2:
           Var{e(ij)} = Sigma(i)A2
 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e (i) } = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
              Log(likelihood)
               -190.565019
               -181.476284
               -181.900030
               -185.947278
               -204.636496
                       # Param's
                             6
                            10
                             7
                             4
                             2
                         AIC
                      393.130038
                      382.952569
                      377.800059
                      379.894555
                      413.272993
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
                                     G-636

-------
   Test    -2*log(Likelihood Ratio)  Test df        p-value
Test 1
Test 2
Test 3
Test 4
46.3204
18.1775
0.84749
8.0945
8
4
3
3
<.0001
0.001139
0.8381
0.0441
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than .1.  You may want to try a different
model
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 215.073


            BMDL = 170.412
                                     G-637

-------
G.3.50.5. Figure for Additional Model Presented: Power



                               Power Model with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
        35
        30
        25
20
15
        10
                  Power
   19:5602/162010
                                            BMDL
                                                   BMD
                        50
                          100
150

 dose
200
250
300
                                         G-638

-------
G.3.51. Smith et al. (1976): Cleft Palate in Pups
G.3.51.1. Summary Table of BMDS Modeling Results
Model3
Gamma
Logistic
Log-logistic a
Log-probit
Multistage, 5th degree
Probit
Weibull
Gamma, unrestricted
Log-logistic, unrestricted
Log-probit, unrestricted
Weibull, unrestricted
Degrees of
freedom
3
4
3
3
3
4
3
3
3
3
3
x2
p-value
0.4203
0.5057
0.4194
0.4132
0.4528
0.5721
0.43
0.4203
0.4194
0.4133
0.43
AIC
69.78
68.90
69.82
69.89
69.43
68.33
69.68
69.78
69.82
69.89
69.68
BMD
(ng/kg-day)
6.184E+02
9.754E+02
6.816E-KJ2
7.341E+02
4.829E+02
8.688E+02
5.908E+02
6.184E+02
6.816E+02
7.341E+02
5.908E+02
BMDL
(ng/kg-day)
2.205E+02
7.256E+02
1.842E-K)2
3.927E+02
2.277E+02
6.580E+02
2.223E+02
1.227E+02
1.705E+02
1.767E+02
1.432E+02
Notes











a Best-fitting model, BMDS output presented in this appendix.
G.3.51.2. Output for Selected Model: Log-Logistic
         Logistic Model.  (Version: 2.12; Date:  05/16/2008)
         Input  Data File:
C:\USEPA\BMDS21\la\76_Smith_1976_cleft_palate_LogLogistic_l.(d)
         Gnuplot Plotting File:
C:\USEPA\BMDS21\la\76_Smith_1976_cleft_palate_LogLogistic_l.plt
                                           Thu  Sep  01  12:46:35 2011
 Table  3  cleft palate
   The  form of the probability function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose))]
   Dependent  variable = DichEff
   Independent variable = Dose
   Slope  parameter is restricted as slope >=  1

   Total  number of observations = 6
   Total  number of records with missing values  =  0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to:  le-008
   Parameter  Convergence has been set to: le-008
   User  has  chosen the log transformed model
                                      G-639

-------
                  Default Initial Parameter Values
                     background =            0
                      intercept =     -7.91888
                          slope =            1
           Asymptotic Correlation Matrix of Parameter Estimates

             background    intercept        slope

background            1        -0.18         0.17

 intercept        -0.18            1           -1

     slope         0.17           -1            1
Confidence Interval
       Variable
Upper Conf. Limit
     background
*
      intercept
*
          slope
               Parameter Estimates

                                       95.0% Wald

      Estimate        Std.  Err.      Lower Conf. Limit

     0.0262471            *                *

      -15.6136            *                *

       2.05633            *                *
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.2701
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -29.9486         6
     -31.9094         3       3.92153      3
     -52.2767

      69.8188
1
44.6562
                            P-value
<.0001
                                  Goodness  of  Fit

0
1
10
100
1000
3000
Dose
.0000
.0000
.0000
.0000
.0000
.0000
Est
0.
0.
0.
0.
0.
0.
. Prob.
0262
0262
0263
0283
2175
7085
Exp
0
1
0
0
4
9
ected
.892
.076
.499
.482
.132
.918
Ob
0.
2.
0.
1.
4.
10.
served
000
000
000
000
000
000
Size
34
41
19
17
19
14
Scaled
Residual
-0.957
0.903
-0.716
0.758
-0.074
0.048
                                     G-640

-------
 ChiA2  =2.83
            d.f.  = 3
P-value  = 0.4194
   Benchmark Dose Computation



Specified effect =              0.1



Risk Type        =       Extra  risk



Confidence level =            0.95



              BMD =         681.581



             BMDL =         184.164





G.3.51.3. Figure for Selected Model: Log-Logistic



                            Log-Logistic Model with 0.95 Confidence Level
 T3

 £
 O
 c
 O
 •*=
 O
 (0
         0.8
         0.6
0.4
         0.2  -
          0  -  *-L
                                                                           3000
   12:4609/01 2011
                                       G-641

-------
G.3.52. Sparschu et al. (1971): Fetal Body Weight, Male
G.3.52.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)b
Hill
Linear
Polynomial, 3 -degree
Power
Hill, unrestricted
Power, unrestricted
Degrees of
freedom
3
3
2
1
1
3
3
3
1
2
X2
p-value
0.0001
0.0001
0.0002
<0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
AIC
-246.49
-246.49
-247.97
-246.36
-246.90
-245.45
-245.45
-245.45
-246.90
-245.65
BMD
(ng/kg-day)
6.665E+02
6.665E+02
5.744E+02
5.459E+02
5.105E+02
7.248E+02
7.248E+02
7.248E+02
5.105E+02
6.812E+02
BMDL
(ng/kg-day)
4.188E+02
4.188E+02
3.197E+02
1.296E-K)2
error
4.607E+02
4.607E+02
4.607E+02
error
3.949E+02
Notes










a Modeled variance model presented (p < 0.0001); variance not appropriately captured (p-test 3 = 0.008).
b Best-fitting model, BMDS output presented in this appendix.
G.3.52.2. Output for Selected Model: Exponential (MS)


         Exponential Model.  (Version: 1.61;  Date:  7/24/2009)
         Input Data File:
C:\USEPA\BMDS21\la\74_Sparschu_1971_pup_bw_male_Exp_l.(d)
         Gnuplot Plotting  File:
                                            Thu Sep  01 12:56:10 2011
 Table 4 males
    The form of the response  function by Model:
       Model 2:
       Model 3:
       Model 4:
       Model 5:
Y[dose]
Y[dose]
Y[dose]
Y[dose]
a
a
a
a
expfsign * b  *  dose}
exp{sign *  (b * dose)Ad}
[c-(c-l) * exp{-b * dose}]
[c-(c-l) * exp{-(b *  dose)Ad}]
     Note:  Y[dose] is the median response for  exposure = dose;
           sign = +1 for increasing trend in data;
           sign = -1 for decreasing trend.

       Model 2 is nested within Models 3 and 4.
       Model 3 is nested within Model 5.
       Model 4 is nested within Model 5.
    Dependent variable = Mean
    Independent variable =  Dose
    Data are assumed to be  distributed: normally
    Variance Model: exp(lnalpha +rho *ln(Y[dose]))
    The variance is to be modeled as Var(i) =  exp(lalpha + log(mean(i))  *  rho)

    Total number of dose groups = 5
                                       G-642

-------
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values
               Variable

                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                      Model 5

                          -4.28192
                           1.66816
                             4.347
                       0.000395512
                          0.312859
                                 1
                  Parameter Estimates

                Variable          Model 5
                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                        16.7441
                       -13.5393
                        4.04428
                     0.00167144
                       0.859252
                        1.18216
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
     30
    125
    500
   2000
117
 55
 66
 39
  3
4.03
4.14
3.85
3.86
2.72
0.37
0.26
0.35
0.61
0.25
   Dose

      0
     30
    125
    500
   2000
               Estimated Values of Interest

             Est Mean      Est Std     Scaled Residual
     4.044
     4.028
     3.962
     3.729
     3.484
0.3372
0.3465
0.3878
0.5845
0.9255
-0.458
 2.398
-2.336
 1.404
 -1.43
Other models for which likelihoods are calculated:

                                  G-643

-------
     Model Al:        Yij
               Var{e(ij) }

     Model A2:        Yij
               Var{e(ij) }

     Model A3:        Yij
               Var{e(ij) }

     Model  R:        Yij
               Var{e(ij) }
                         Mu(i) + e (ij)
                         SigmaA2

                         Mu(i) + e (ij)
                         Sigma(i)A2

                         Mu(i) + e (i j)
                         exp(lalpha + log(mean(i)) * rho)

                         Mu + e(i)
                         SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
5
126.4055
145.7666
137.4206
101.5293
129.1813
6
10
7
2
6
-240.8109
-271.5331
-260.8413
-199.0587
-246.3626
   Additive constant for all log-likelihoods =     -257.3.  This  constant
added to the
   above values gives the log-likelihood including the term that  does not
   depend on the model parameters.
R)
                              Explanation of Tests

Test 1:  Does response and/or variances differ among Dose levels?  (A2  vs.

Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)

Test 7a: Does Model 5 fit the data?  (A3 vs  5)
     Test

     Test 1
     Test 2
     Test 3
    Test 7a
                         Tests of Interest

                -2*log(Likelihood Ratio)
                                                  D. F.
p-value
88.47
38.72
16.69
16.48
8
4
3
1
< 0.0001
< 0.0001
0.0008177
< 0.0001
     The p-value for Test 1 is less than  .05.  There appears to be a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

     The p-value for Test 2 is less than  .1.  A non-homogeneous
     variance model appears to be appropriate.

                                     G-644

-------
     The p-value  for Test 3 is less  than .1.   You may want  to

     consider  a different variance model.



     The p-value  for Test 7a is less  than  .1.   Model 5 may  not  adequately

     describe  the data;  you may want  to  consider another model.
   Benchmark  Dose  Computations:



     Specified  Effect = 1.000000



            Risk Type = Estimated standard deviations from  control



     Confidence Level = 0.950000



                   BMD =      545.876



                 BMDL =      129.551
G.3.52.3. Figure for Selected Model: Exponential (MS)



                         Exponential_beta Model 5 with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 HI
         3.5
        2.5
                        Exponential
               BMDL
   12:5609/01 2011
BMD
                             500
            1000

            dose
1500
2000
                                      G-645

-------
G.3.53. Sparschu et al. (1971): Fetal Body Weight, Female
G.3.53.1. Summary Table of BMDS Modeling Results
Model3
Exponential (M2) b
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear
Polynomial, 3 -degree
Power
Hill, unrestricted
Power, unrestricted
Degrees of
Freedom
3
3
2
2
2
3
3
2
1
2
fp-
Value
0.0278
0.0278
0.0147
0.0147
0.0151
0.0245
0.0245
0.0025
0.0038
0.0146
AIC
-229.517
-229.517
-228.188
-228.188
-228.244
-229.239
-229.239
-224.657
-226.278
-228.180
BMD
(ng/kg-day)
.033E-KJ3
.033E+03
.057E+03
.057E+03
.073E+03
.050E+03
.050E+03
.860E+03
.073E+03
.077E+03
BMDL
(ng/kg-day)
6.479E+02
6.479E+02
5.759E+02
5.759E+02
5.800E+02
6.749E+02
6.749E+02
5.877E+02
5.828E+02
6.192E+02
Notes










a Modeled variance model presented (p = 0.001); variance not appropriately captured (p-test 3 = 0.005).
b Best-fitting model, BMDS output presented in this appendix.
G.3.53.2. Output for Selected Model: Exponential (M2)


         Exponential Model.  (Version:  1.61;  Date:  7/24/2009)
         Input Data File:
C:\USEPA\BMDS21\la\75_Sparschu_1971_pup_bw_male_Exp_l.(d)
         Gnuplot Plotting File:
                                            Thu Sep  01  13:43:52 2011
 Table  4  females
   The  form of the response  function by Model:
      Model 2:
      Model 3:
      Model 4:
      Model 5:
Y[dose]
Y[dose]
Y[dose]
Y[dose]
a
a
a
a
exp{sign * b  *  dose}
exp{sign *  (b * dose)Ad}
[c-(c-l) * exp{-b * dose}]
[c-(c-l) * exp{-(b *  dose)Ad}]
    Note:  Y[dose]  is the median  response for exposure  = dose;
           sign = +1 for increasing trend in data;
           sign = -1 for decreasing trend.

      Model 2 is nested within Models 3 and 4.
      Model 3 is nested within Model 5.
      Model 4 is nested within Model 5.
   Dependent variable = Mean
   Independent variable = Dose
   Data  are assumed to be distributed:  normally
   Variance Model: exp(lnalpha  +rho *ln(Y[dose]))
   The variance is to be modeled as Var(i)  = exp(lalpha + log(mean(i))

   Total number of dose groups  = 5
                                                         rho)
                                      G-646

-------
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values
               Variable

                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                      Model 2

                          -7.22746
                           4.02075
                           3.75712
                       0.000140769
                                 0
                                 1
                  Parameter Estimates

                Variable          Model 2
                 Inalpha
                     rho
                       a
                       b
                       c
                       d
                        10.6901
                       -9.26779
                        3.89584
                    0.000100525
                              0
                              1
         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
     30
    125
    500
   2000
129
 60
 58
 54
  4
3.89
3.98
3.71
3.78
2.69
0.39
0.35
0.37
0.54
0.19
   Dose

      0
     30
    125
    500
   2000
               Estimated Values of Interest

             Est Mean      Est Std     Scaled Residual
     3.896
     3.884
     3.847
     3.705
     3.186
0.3842
0.3896
0.4072
0.4849
0.9753
-0.1727
  1.907
 -2.566
  1.139
 -1.018
Other models for which likelihoods are calculated:

                                  G-647

-------
     Model Al:        Yij
               Var{e(ij) }

     Model A2:        Yij
               Var{e(ij) }

     Model A3:        Yij
               Var{e(ij) }

     Model  R:        Yij
               Var{e(ij) }
                         Mu(i) + e (ij)
                         SigmaA2

                         Mu(i) + e (ij)
                         Sigma(i)A2

                         Mu(i) + e (i j)
                         exp(lalpha + log(mean(i)) * rho)

                         Mu + e(i)
                         SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
2
123.0729
132.131
123.3163
100.5646
118.7583
6
10
7
2
4
-234.1458
-244.262
-232.6326
-197.1292
-229.5166
   Additive constant for all log-likelihoods =     -280.3.  This  constant
added to the
   above values gives the log-likelihood including the term that  does not
   depend on the model parameters.
R)
Test 1:

Test 2:
Test 3:
Test 4:
                     Explanation of Tests

Does response and/or variances differ among Dose levels?  (A2  vs.

Are Variances Homogeneous?  (A2 vs. Al)
Are variances adequately modeled?  (A2 vs. A3)
Does Model 2 fit the data?  (A3 vs. 2)
     Test

     Test 1
     Test 2
     Test 3
     Test 4
                         Tests of Interest

                -2*log(Likelihood Ratio)
                                                  D. F.
                                                    p-value
63.13
18.12
17.63
9.116
8
4
3
3
< 0.0001
0.001171
0.0005244
0.02779
     The p-value for Test 1 is less than  .05.  There appears to be a
     difference between response and/or variances among the dose
     levels, it seems appropriate to model the data.

     The p-value for Test 2 is less than  .1.  A non-homogeneous
     variance model appears to be appropriate.
                                     G-648

-------
     The p-value  for Test 3 is less  than .1.   You may want  to

     consider  a different variance model.



     The p-value  for Test 4 is less  than .1.   Model 2 may not  adequately

     describe  the data;  you may want  to  consider another model.
   Benchmark  Dose  Computations:



     Specified  Effect = 1.000000



            Risk Type = Estimated standard deviations from  control



     Confidence Level = 0.950000



                   BMD =      1032.78



                 BMDL =      647.855
G.3.53.3. Figure for Selected Model: Exponential (M2)


                         Exponential_beta Model 2 with 0.95 Confidence Level
 c
 o
 Q.
 (/)
 0)
 o:

 c
 (0
 OJ
         3.5
         2.5
   13:4309/01 2011
                        Exponential
                              BMDL
   BMD
                             500
1000

dose
1500
2000
                                      G-649

-------
G.3.54. Toth et al. (1979): Amyloidosis
G.3.54.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage, 3 -degree
Probit
Weibull
Gamma, unrestricted
Log-logistic,
unrestricted13
Log-probit, unrestricted
Weibull, unrestricted
Degrees of
freedom
2
2
2
2
2
2
2
2
2
2
2
X2
p-value
0.022
0.013
0.028
0.007
0.022
0.014
0.022
0.917
0.847
0.811
0.882
AIC
150.666
152.187
149.984
153.479
150.666
152.040
150.666
140.208
140.370
140.458
140.287
BMD
(ng/kg-day)
2.296E+02
4.088E+02
1.759E+02
4.402E+02
2.296E+02
3.846E+02
2.296E+02
7.687E-01
8.465E-01
8.545E-01
8.179E-01
BMDL
(ng/kg-day)
1.460E+02
3.125E+02
9.729E+01
2.965E+02
1.460E+02
2.911E+02
1.460E+02
7.637E-04
1.565E-03
2.334E-03
1.140E-03
Notes
power bound hit
(power =1)

slope bound hit
(slope = 1)
slope bound hit
(slope = 1)
final B = 0

power bound hit
(power =1)
unrestricted
(power =0.187)
unrestricted
(slope = 0.238)
unrestricted
(slope = 0.135)
unrestricted
(power =0.2 12)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.


G.3.54.2. Output for Selected Model: Log-Logistic
Toth et al. (1979): Amyloidosis
         Logistic Model.  (Version: 2.12;  Date:  05/16/2008)
         Input Data File:  C:\l\62_Toth_1979_Amylyr_LogLogistic_l.(d)
         Gnuplot Plotting  File:   C:\l\62_Toth_1979_Amylyr_LogLogistic_l.plt
                                             Tue  Feb 16 19:56:59  2010
 Table  2
   The  form of the probability function  is:

   P[response] = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose) ) ]
   Dependent variable  =  DichEff
   Independent variable  = Dose
   Slope parameter is  restricted as slope  >= 1

   Total number of observations = 4
                                       G-650

-------
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =            0
                      intercept =     -6.90711
                          slope =            1
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -slope
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )

             background    intercept

background            1        -0.47

 intercept        -0.47            1
                                 Parameter Estimates
Confidence Interval
       Variable
Upper Conf. Limit
     background
*
      intercept
*
          slope
 Estimate

0.0848984

 -7.36716
Std.  Err.
                                       95. 0% Wald

                                    Lower Conf. Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.00691
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
      -68.017         4
     -72.9918         2        9.9496      2
     -82.0119

      149.984
1
                                     G-651
                           27.99
3
                            P-value
<.0001

-------
                                  Goodness  of  Fit

Dose
0.0000
1.0000
100.0000
1000.0000

Est. Prob.
0.0849
0.0855
0.1393
0.4392

Expected
3.226
3.761
6.128
18.884

Observed
0.000
5.000
10.000
17.000

Size
38
44
44
43
Scaled
Residual
-1.878
0.668
1.686
-0.579
 ChiA2 = 7.15      d.f. = 2        P-value = 0.0280







   Benchmark Dose Computation




Specified effect =            0.1




Risk Type        =      Extra risk




Confidence level =           0.95




             BMD =        175. 903




            BMDL =        97.2899
                                     G-652

-------
G.3.54.3. Figure for Selected Model: Log-Logistic

                          Log-Logistic Model with 0.95 Confidence Level
        0.6
        0.5
        0.4
 o
 '•8
 ro
        0.3
        0.2
        0.1
                        200
400         600
     dose
800
1000
  19:5602/162010
G.3.54.4. Output for Additional Model Presented: Log-Logistic, Unrestricted
Toth et al. (1979): Amyloidosis
        Logistic Model.  (Version:  2.12; Date:  05/16/2008)
        Input Data File: C:\l\62_Toth_1979_Amylyr_LogLogistic_U_l.(d)
        Gnuplot Plotting File:   C:\l\62_Toth_1979_Amylyr_LogLogistic_U_l.plt
                                            Tue  Feb 16 19:57:00  2010
 Table 2
   The  form of the probability function is:

   P[response]  = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose) ) ]
   Dependent  variable = DichEff
   Independent variable = Dose
   Slope parameter is not restricted
                                       G-653

-------
   Total number of observations = 4
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =            0
                      intercept =     -2.10894
                          slope =     0.227921
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -background
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
              intercept

 intercept            1

     slope        -0.89
             slope

             -0.89

                 1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background                0
*
      intercept         -2.15753
*
          slope         0.238304
                                 Parameter Estimates
                      Std.  Err.
                     95.0% Wald

                  Lower  Conf.  Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.8455
  Reduced model
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
      -68.017         4
     -68.1848         2       0.33571      2
     -82.0119
    1

G-654
27.99
3
                                P-value
<.0001

-------
           AIC:          140.37
                                  Goodness  of  Fit

Dose
0.0000
1.0000
100.0000
1000.0000

Est. Prob.
0.0000
0.1036
0.2573
0.3749

Expected
0.000
4.560
11.321
16.119

Observed
0.000
5.000
10.000
17.000

Size
38
44
44
43
Scaled
Residual
0.000
0.218
-0.456
0.277
 ChiA2 = 0.33      d.f. = 2        P-value = 0.8471







   Benchmark Dose Computation




Specified effect =            0.1




Risk Type        =      Extra risk




Confidence level =           0.95




             BMD =       0.846547




            BMDL =     0.00156534
                                     G-655

-------
G.3.54.5. Figure for Additional Model Presented: Log-Logistic, Unrestricted



                               Log-Logistic Model with 0.95 Confidence Level
 O


 I


 O

 13
 (0
          0.6
          0.5
          0.4
0.3
          0.2
          0.1
                             Log-Logistic
            BMDL
                              200
                                 400           600

                                       dose
800
1000
   19:5702/162010
                                           G-656

-------
G.3.55. Toth et al. (1979): Skin Lesions
G.3.55.1. Summary Table of BMDS Modeling Results
Model
Gamma
Logistic"
Log-logistic
Log-probit
Multistage, 3 -degree
Probit
Weibull
Gamma, unrestricted
Log-logistic,
unrestricted13
Log-probit, unrestricted
Weibull, unrestricted
Degrees of
freedom
2
2
2
2
2
2
2
2
2
2
2
X2
p-value
0.009
0.002
0.029
0.001
0.009
0.003
0.009
0.882
0.630
0.558
0.762
AIC
159.223
162.974
156.567
164.598
159.223
162.684
159.223
147.287
147.969
148.218
147.581
BMD
(ng/kg-day)
1.181E+02
2.709E+02
6.750E+01
2.446E+02
1.181E+02
2.522E+02
1.181E+02
error
1.137E+00
1.096E+00
1.077E+00
BMDL
(ng/kg-day)
8.308E+01
2.147E-KJ2
4.057E+01
1.626E+02
8.308E+01
2.015E+02
8.308E+01
error
5.477E-02
6.847E-02
4.080E-02
Notes
power bound hit
(power =1)

slope bound hit
(slope = 1)
slope bound hit
(slope = 1)
final B = 0

power bound hit
(power =1)
unrestricted
(power =0.251)
unrestricted
(slope = 0.351)
unrestricted
(slope = 0.202)
unrestricted
(power =0.3)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.


G.3.55.2. Output for Selected Model: Logistic
Toth et al. (1979): Skin Lesions
         Logistic Model.  (Version: 2.12;  Date:  05/16/2008)
         Input Data  File:  C:\l\63_Toth_1979_SkinLes_Logistic_l.(d)
         Gnuplot Plotting  File:  C:\l\63_Toth_1979_SkinLes_Logistic_l.plt
                                             Tue Feb 16  19:57:29 2010
 Table  2


   The  form of the probability function  is:

   P[response] = I/[1+EXP(-intercept-slope*dose)]
    Dependent variable  =  DichEff
    Independent variable  = Dose
    Slope parameter is  not restricted

    Total number of observations = 4
                                       G-657

-------
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                     background =            0   Specified
                      intercept =     -2.53484
                          slope =   0.00299511
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)  -background
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
              intercept

 intercept            1

     slope        -0.67
             slope

             -0.67

                 1
Confidence Interval
       Variable
Upper Conf. Limit
      intercept
-1.39061
          slope
0.00312686
      Estimate

      -1.91768

    0.00230499
Parameter Estimates



       Std.  Err.

         0.26892

     0.000419329
         95. 0% Wald

      Lower Conf.  Limit

             -2.44475

           0.00148312
       Model
     Full model
   Fitted model
0.0003459
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -71.5177         4
      -79.487         2       15.9387      2
     -95.8498

      162.974
       1
48.6642
3
                                   P-value
<.0001
     Dose     Est._Prob.

    0.0000     0.1281
                Goodness  of  Fit

          Expected    Observed     Size

            4.869     0.000          38

                  G-658
                                Scaled
                               Residual
                               -2.363

-------
    1.0000
  100.0000
 1000.0000

 ChiA2 =  12.19
       0.1284
       0.1561
       0.5956
           d.f.  = 2
 5.649      5.000           44
 6.870     13.000           44
25.612     25.000           43

      P-value = 0.0023
-0.292
 2.546
-0.190
   Benchmark Dose Computation

Specified  effect =             0.1

Risk Type         =       Extra risk

Confidence level =            0.95

              BMD =         270.917

             BMDL =          214.66


G.3.55.3. Figure for Selected Model: Logistic

                              Logistic Model with 0.95 Confidence Level
 T3
 0)
 •5

 I
 C
 o
 •*=
 o
 (0
0.7


0.6


0.5


0.4


0.3


0.2


0.1
                      Logistic
                        BMDL
                       BMD
                          200
                              400          600
                                   dose
                                800
      1000
   19:5702/162010
                                       G-659

-------
G.3.55.4.  Output for Additional Model Presented: Log-Logistic, Unrestricted
Toth et al. (1979): Skin Lesions
        Logistic Model.  (Version: 2.12; Date:  05/16/2008)
        Input Data File: C:\l\63_Toth_1979_SkinLes_LogLogistic_U_l.(d)
        Gnuplot Plotting File:  C:\l\63_Toth_1979_SkinLes_LogLogistic_U_l.plt
                                          Tue  Feb  16  20:01:56  2010
 Table 2
   The form of the probability function is:

   P[response] = background+(1-background)/[1+EXP(-intercept-
slope*Log(dose) ) ]
   Dependent variable = DichEff
   Independent variable = Dose
   Slope parameter is not restricted

   Total number of observations = 4
   Total number of records with missing values  =  0
   Maximum number of iterations = 250
   Relative Function Convergence has been  set to:  le-008
   Parameter Convergence has been set to:  le-008
   User has chosen the log transformed model
                  Default Initial Parameter Values
                     background =             0
                      intercept =     -2.14055
                          slope =     0.332409
           Asymptotic Correlation Matrix of  Parameter  Estimates

            ( *** The model parameter(s)  -background
                 have been estimated at a boundary  point,  or  have been
specified by the user,
                 and do not appear in  the correlation  matrix  )

              intercept        slope

 intercept            1         -0.9

     slope         -0.9            1
                                     G-660

-------
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
     background                0
*
      intercept         -2.24241
*
          slope         0.350932
                                 Parameter Estimates
                      Std. Err.
                      95.0% Wald

                   Lower Conf. Limit
  - Indicates that this value is not calculated.
       Model
     Full model
   Fitted model
0.6271
  Reduced model

           AIC:
      Analysis of Deviance Table

Log(likelihood)   # Param's  Deviance  Test d.f.
     -71.5177         4
     -71.9844         2       0.93345      2
     -95.8498

      147.969
     1
48.6642
3
                                 P-value
<.0001
                                  Goodness  of  Fit

Dose
0.0000
1.0000
100.0000
1000.0000

Est. Prob.
0.0000
0.0960
0.3483
0.5453

Expected
0.000
4.224
15.327
23.448

Observed
0.000
5.000
13.000
25.000

Size
38
44
44
43
Scaled
Residual
0.000
0.397
-0.736
0.475
 ChiA2 = 0.93
 d.f.  = 2
P-value = 0.6295
   Benchmark Dose Computation

Specified effect =            0.1

Risk Type        =      Extra risk

Confidence level =           0.95

             BMD =         1.1374

            BMDL =      0.0547689
                                     G-661

-------
G.3.55.5. Figure for Additional Model Presented: Log-Logistic, Unrestricted



                                Log-Logistic Model with 0.95 Confidence Level
 T3

 £
 O
 c
 O
 •*=
 O
 (0
0.7




0.6




0.5




0.4




0.3




0.2




0.1
                             Log-Logistic
            BMDLBMD
                              200
                                 400           600

                                       dose
800
1000
   20:01 02/162010
                                           G-662

-------
G.3.56. van Birgelen et al. (1995): Hepatic Retinol
G.3.56.1. Summary Table of BMDS Modeling Results
Model
Exponential (M2)
Exponential (M3)
Exponential (M4)a
Exponential (M5)
Hill
Linear
Polynomial, 5 -degree
Power
Hill, unrestricted
Power, unrestrictedb
Degrees
of
freedom
4
4
3
3
3
4
0
4
2
3
X2/7-value
0.0001
0.0001
<0.0001
0.0001
0.044
0.0001
N/A
0.0001
0.269
0.025
AIC
164.340
164.340
148.052
148.052
128.757
178.734
283.606
178.734
125.273
129.990
BMD
(ng/kg-day)
2.912E+02
2.912E+02
1.151E+02
1.151E+02
1.314E+01
7.815E+02
2.481E+03
7.815E+02
5.561E+00
4.205E-01
BMDL
(ng/kg-day)
error
error
7.098E-H)!
7.098E+01
error
5.997E+02
error
5.997E+02
error
8.504E-03
Notes

power hit bound (d = 1)

power hit bound (d = 1)
n lower bound hit (n = 1)


power bound hit (power =1)
unrestricted (n = 0.571)
unrestricted (power =0.118)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.
G.3.56.2. Output for Selected Model: Exponential (M4)
van Birgelen et al. (1995): Hepatic Retinol
         Exponential Model.  (Version:  1.61;  Date:  7/24/2009)
         Input Data File: C:\l\65_VanB_1995a_HepRet_Exp_l.(d)
         Gnuplot Plotting File:
                                            Tue  Feb  16 20:03:05 2010
 Tbl3,  hepatic retinol
   The  form of the response  function by Model:
      Model 2:     Y[dose] = a  *  exp{sign * b * dose}
      Model 3:     Y[dose] = a  *  exp{sign *  (b *  dose)Ad}
      Model 4:     Y[dose] = a  *  [c-(c-l) * exp{-b  *  dose}]
      Model 5:     Y[dose] = a  *  [c-(c-l) * exp{-(b *  dose)Ad}]

    Note:  Y[dose]  is the median response for exposure  = dose;
           sign = +1 for increasing trend in data;
           sign = -1 for decreasing trend.

      Model 2 is nested within  Models 3 and 4.
      Model 3 is nested within  Model 5.
      Model 4 is nested within  Model 5.
   Dependent variable = Mean
   Independent variable = Dose
   Data  are assumed to be distributed: normally
   Variance Model: exp(lnalpha  +rho *ln(Y[dose])

                                      G-663

-------
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  * rho)

Total number of dose groups = 6
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008

MLE solution provided: Exact
               Initial Parameter Values

               Variable          Model 4

                 Inalpha             -1.16065
                     rho              1.53688
                       a               15.645
                       b           0.00625117
                       c            0.0365247
                       d                    1
                  Parameter Estimates

                Variable          Model 4

                 Inalpha           -0.882225
                     rho             1.82707
                       a             10.5294
                       b          0.00720346
                       c           0.0688661
                       d                   1


         Table of Stats From Input Data

  Dose      N         Obs Mean     Obs Std Dev
      0
     14
     26
     47
    320
   1024
               Estimated Values of Interest

   Dose      Est Mean      Est Std     Scaled Residual
14
8
8
5
2
0
.9
.4
.2
.1
.2
.6
8
3
2
0.
0.
0.
.768
.394
.263
8485
8485
5657
      0         10.53        5.526            2.237
     14         9.589        5.073          -0.6628
     26         8.855        4.717          -0.3926
     47         7.714        4.159           -1.778
    320         1.703        1.046            1.343

                                  G-664

-------
      1024
               0.7313
0.4833
-0.7681
   Other models for which likelihoods are calculated:

     Model Al:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = SigmaA2

     Model A2:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = Sigma(i)A2

     Model A3:        Yij = Mu(i) + e(ij)
               Var{e(ij)} = exp(lalpha + log(mean(i)) *  rho)

     Model  R:        Yij = Mu + e(i)
               Var{e(ij)} = SigmaA2
                     Model
                             Likelihoods of Interest

                             Log(likelihood)      DF
                                                                AIC
Al
A2
A3
R
4
-87.1567
-47.28742
-55.32422
-109.967
-69.02619
7
12
8
2
5
188.3134
118.5748
126.6484
223.934
148.0524
   Additive constant for all log-likelihoods =     -44.11.   This  constant
added to the
   above values gives the log-likelihood including the  term  that  does  not
   depend on the model parameters.
R)
                              Explanation of Tests

Test 1:  Does response and/or variances differ among Dose  levels?  (A2  vs.

Test 2:  Are Variances Homogeneous?  (A2 vs. Al)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)

Test 6a: Does Model 4 fit the data?  (A3 vs  4)
     Test

     Test 1
     Test 2
     Test 3
    Test 6a
                         Tests of Interest

                 -2*log(Likelihood Ratio)
                                                  D. F.
                                 p-value
125.4
79.74
16.07
27.4
10
5
4
3
< 0.0001
< 0.0001
0.002922
< 0.0001
     The p-value for Test 1 is less than  .05.  There appears  to be  a

                                     G-665

-------
  difference between response and/or variances among the dose
  levels, it seems appropriate to model the data.

  The p-value for Test 2 is less than .1.  A non-homogeneous
  variance model appears to be appropriate.

  The p-value for Test 3 is less than .1.  You may want to
  consider a different variance model.

  The p-value for Test 6a is less than .1.  Model 4 may not adequately
  describe the data; you may want to consider another model.
Benchmark Dose Computations:

  Specified Effect = 1.000000

         Risk Type = Estimated standard deviations from control

  Confidence Level = 0.950000

               BMD =      115.128

              BMDL =       70.981
                                 G-666

-------
G.3.56.3. Figure for Selected Model: Exponential (M4)

                        Exponential_beta Model 4 with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 OJ
       20
        15
10
                       Exponential
              BMDL   BMP
                         200
   20:0302/162010
                             400         600
                                   dose
800
1000
G.3.56.4. Output for Additional Model Presented: Power, Unrestricted
van Birgelen et al. (1995): Hepatic Retinol
         Power Model.  (Version:  2.15;   Date:  04/07/2008)
         Input Data File: C:\l\65_VanB_1995a_HepRet_Pwr_U_l.(d)
         Gnuplot Plotting File:   C:\l\65_VanB_1995a_HepRet_Pwr_U_l.plt
                                            Tue  Feb  16 20:03:11 2010
 Tbl3, hepatic retinol


   The form of the response  function is:

   Y[dose]  = control + slope  *  doseApower
   Dependent variable = Mean
   Independent variable = Dose
                                       G-667

-------
   The power is not restricted
   The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i))  * rho)

   Total number of dose groups = 6
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                  Default Initial Parameter Values
                         lalpha =      2.76506
                            rho =            0
                        control =         14.9
                          slope =     -3.78637
                          power =     0. 191713
           Asymptotic Correlation Matrix of Parameter Estimates

lalpha
rho
control
slope
power
lalpha
1
-0.8
-0.047
0.042
0.065
rho
-0.8
1
-0.085
-0.0029
-0.11
control
-0.047
-0.085
1
-0.95
-0.81
slope
0.042
-0.0029
-0.95
1
0.96
power
0.065
-0.11
-0.81
0.96
1
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha         -1.02622
-0.263475
            rho          1.68421
2.07466
        control          16.9577
21.2918
          slope         -7.19097
-3.27676
          power         0.117935
0.162111
Parameter Estimates



       Std.  Err.

        0.389164

        0.199212

         2.21133

         1.99708

       0.0225396
   95.0% Wald

Lower Conf.  Limit

       -1.78897

        1.29376

        12.6235

       -11.1052

      0.0737578
     Table of Data and Estimated Values of Interest
                                     G-668

-------
 Dose
Res .
                 Obs Mean
                              Est Mean
                              Obs Std Dev  Est Std Dev
                                      Scaled
    0
   14
   26
   47
  320
 1024
       14.9
        8.4
        8.2
        5.1
        2.2
        0.6
   17
 7.14
  6.4
 5.63
 2.76
0.672
 8.77
 3.39
 2.26
0.849
0.849
0.566
 6.49
 3.13
 2.86
 2.57
 1.41
0.428
-0.896
  1.14
  1.78
-0.588
 -1.12
-0.475
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e (i) } = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
              -87.156698
              -47.287416
              -55.324218
              -59.994980
             -109.967018
           # Param's
                 7
                12
                 8
                 5
                 2
             AIC
          188.313395
          118.574833
          126.648436
          129.989960
          223.934036
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
-2*log(Likelihood Ratio)  Test df
            125.359
            79.7386
         10
          5
         p-value

        <.0001
        <.0001
                                     G-669

-------
   Test 3              16.0736          4        0.002922
   Test 4              9.34152          3         0.02508

The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is less than .1.  You may want to consider a
different variance model

The p-value for Test 4 is less than .1.  You may want to try a different
model
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 0.420475


            BMDL = 0.00850422
                                     G-670

-------
G.3.56.5.  Figure for Additional Model Presented: Power, Unrestricted



                                Power Model with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
                           200
400          600

       dose
800
1000
   20:0302/162010
                                          G-671

-------
G.3.57. van Birgelen et al. (1995): Hepatic Retinol Palmitate
G.3.57.1. Summary Table of BMDS Modeling Results
Model
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill
Linear"
Polynomial, 5 -degree
Power
Hill, unrestricted
Power, unrestrictedb
Degrees of
freedom
4
4
3
3
3
4
0
4
3
3
X2 p-value
0.0001
0.0001
0.0001
0.0001
0.0001
<0.0001
N/A
0.0001
0.0001
0.348
AIC
467.446
467.446
454.087
454.087
563.579
488.446
573.977
488.446
522.322
408.062
BMD
(ng/kg-day)
error
error
error
error
error
1.420E-H)3
error
1.420E+03
2.418E-12
3.765E-02
BMDL
(ng/kg-day)
error
error
error
error
error
9.889E-KJ2
error
9.889E+02
2.418E-12
1.208E-05
Notes

power hit bound (d = 1)

power hit bound (d = 1)



power bound hit (power =1)
unrestricted (n = 0.452)
unrestricted (power = 0.054)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.
G.3.57.2. Output for Selected Model: Linear
van Birgelen et al. (1995): Hepatic Retinol Palmitate
         Polynomial Model.  (Version:  2.13;  Date:  04/08/2008)
         Input Data File: C:\l\66_VanB_1995a_HepRetPalm_Linear_l.(d)
         Gnuplot Plotting File:   C:\l\66_VanB_1995a_HepRetPalm_Linear_l.plt
                                            Tue Feb  16  20:03:46 2010
 Tbl3,  hepatic retinol palmitate


   The  form of the response  function is:

   Y[dose]  = beta 0 + beta l*dose  + beta 2*doseA2  +
   Dependent variable = Mean
   Independent variable = Dose
   Signs  of the polynomial  coefficients are not restricted
   The  variance is to be modeled as Var(i) = exp(lalpha  + log(mean(i)) *  rho)

   Total  number of dose groups  = 6
   Total  number of records  with missing values =  0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                   Default Initial  Parameter Values
                          lalpha  =       9.57332

                                      G-672

-------
                            rho =
                         beta_0 =
                         beta 1 =
                      0
                177.506
              -0.204775
           Asymptotic Correlation Matrix of Parameter Estimates

                 lalpha          rho       beta 0       beta 1

    lalpha            1        -0.95       -0.017        0.022

       rho        -0.95            1      0.00019      -0.0048

    beta_0       -0.017      0.00019            1           -1

    beta 1        0.022      -0.0048           -1            1
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
0.527811
            rho
2.54093
         beta_0
212.363
         beta_l
-0.0835018
 Estimate

-0.723216

  2.26615

  150.535

-0.143931
Parameter Estimates



       Std.  Err.

        0.638291

        0.140196

         31.5457

       0.0308317
   95.0% Wald

Lower Conf.  Limit

       -1.97424

        1.99137

        88.7064

       -0.20436
     Table of Data and Estimated Values of Interest
Dose
Res .
0
14
26
47
320
1024
N
8
8
8
8
8
8
Obs Mean
472
94
107
74
22
3
Est Mean
151
149
147
144
104
3.15
Obs Std Dev
272
67.9
76.4
39.6
22.6
2.83
Est Std Dev
204
201
199
194
135
2.56
Scaled
4.45
-0.766
-0.567
-1.02
-1.73
-0.166
 Model Descriptions for likelihoods calculated
 Model Al:        Yij
           Var{e(ij) }
 Mu(i)  + e(i j '
 S i gma A 2
                                     G-673

-------
 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
            Log(likelihood)
             -250.554817
             -196.755746
             -197.383174
             -240.223107
             -276.789644
# Param' s
7
12
8
4
2
AIC
515.109634
417.511491
410.766347
488.446215
557.579287
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.

                     Tests of Interest
   Test

   Test 1
   Test 2
   Test 3
   Test 4
-2*log(Likelihood Ratio)  Test df
            160.068
            107.598
            1.25486
            85.6799
10
 5
 4
 4
 p-value

<.0001
<.0001
 0.869
<.0001
The p-value for Test 1 is less than  .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than  .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is less than  .1.  You may want to try a different
model
             Benchmark Dose Computation
                                     G-674

-------
Specified effect =



Risk Type



Confidence level =



              BMD =





             BMDL =
                          1



                 Estimated standard  deviations  from the control mean



                       0.95



                    1419.81





                    988.945
G.3.57.3. Figure for Selected Model: Linear



                               Linear Model with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
 700




 600




 500




 400




 300




 200




 100




   0




-100
                    Linear
                                                      BMDL
                                    BIVHD
                       200
                        400
600      800

    dose
1000     1200
1400
   20:0302/162010
                                       G-675

-------
G.3.57.4. Output for Additional Model Presented: Power, Unrestricted
van Birgelen et al. (1995): Hepatic Retinol Palmitate
        Power Model.  (Version:  2.15;   Date:  04/07/2008)
        Input Data  File:  C:\l\66_VanB_1995a_HepRetPalm_Pwr_U_l.(d)
        Gnuplot Plotting  File:   C:\l\66_VanB_1995a_HepRetPalm_Pwr_U_l.plt
                                           Tue Feb 16 20:03:50 2010
 Tbl3, hepatic retinol palmitate


   The form of the response  function  is:

   Y[dose] = control + slope  *  doseApower
   Dependent variable = Mean
   Independent variable =  Dose
   The power is not restricted
   The variance is to be modeled  as  Var(i)  = exp(lalpha + log(mean(i))  * rho)

   Total number of dose groups  =  6
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence  has been set to:  le-008
   Parameter Convergence has been set  to:  le-008
                  Default  Initial  Parameter Values
                         lalpha  =       9.57332
                             rho  =             0
                         control  =           472
                           slope  =      -315.054
                           power  =     0.0586881
           Asymptotic Correlation  Matrix of Parameter Estimates

lalpha
rho
control
slope
power
lalpha
1
-0.95
0.29
-0.31
-0.3
rho
-0.95
1
-0.4
0.39
0.29
control
0.29
-0.4
1
-0.98
-0.82
slope
-0.
0.
-0.

0.
31
39
98
1
91
power
-0.3
0.29
-0.82
0.91
1
                                     G-676

-------
                                 Parameter Estimates
Confidence Interval
       Variable         Estimate
Upper Conf. Limit
         lalpha        0.0734958
1.7386
            rho          1.80632
2.18774
        control          465.497
635.845
          slope          -318.06
-156.534
          power        0.0540573
0.0771278
                 Std. Err.

                  0.849559

                  0.194602

                    86.914

                   82.4127

                 0.0117709
    95.0% Wald

 Lower Conf. Limit

        -1.59161

         1.42491

         295.149

        -479.586

       0.0309869
     Table of Data and Estimated Values of Interest
Dose
Res .
0
14
26
47
320
1024
N
8
8
8
8
8
8
Obs Mean
472
94
107
74
22
3
Est Mean
465
98.7
86.2
73.8
31.1
2.86
Obs Std Dev
272
67.9
76.4
39.6
22.6
2.83
Est Std Dev
266
65.6
58.1
50.5
23.1
2.68
Scaled
0.069
-0.201
1.01
0.0086
-1.11
0.145
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e (i) } = SigmaA2
            Model
             Al
Likelihoods of Interest

Log(likelihood)    # Param's
 -250.554817            7
   AIC
515.109634
                                     G-677

-------
             A2
             A3
         fitted
              R
             -196.755746
             -197.383174
             -199.031154
             -276.789644
       12
        8
        5
        2
  417.511491
  410.766347
  408.062307
  557.579287
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?
          (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:   When rho=0 the results of Test 3 and Test 2 will be the same.
   Test

   Test 1
   Test 2
   Test 3
   Test 4
                     Tests of Interest
-2*log(Likelihood Ratio)   Test df
            160.068
            107.598
            1.25486
            3.29596
10
 5
 4
 3
 p-value

<.0001
<.0001
 0.869
0.3482
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is greater than  .1.  The modeled variance appears
 to be appropriate here

The p-value for Test 4 is greater than  .1.  The model chosen seems
to adequately describe the data
               Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =          0.95

             BMD = 0.0376489


            BMDL = 1.20769e-005
                                     G-678

-------
G.3.57.5. Figure for Additional Model Presented: Power, Unrestricted



                                Power Model with 0.95 Confidence Level
 o
 Q.
 o:

 c
 (0
 OJ
         700
         600
         500
         400
300
         200
         100
                   Power
            BMDLBMD
                            200
400         600

      dose
                                                        800
1000
   20:0302/162010
                                         G-679

-------
G.3.58. White et al. (1986): CH50
G.3.58.1. Summary Table of BMDS Modeling Results
Model
Exponential (M2)
Exponential (M3)
Exponential (M4)
Exponential (M5)
Hill3
Linear
Polynomial, 6 -degree
Power
Hill, unrestricted15
Power, unrestricted
Degrees of
freedom
5
5
4
4
4
5
3
5
3
4
x2
p-value
0.001
0.001
0.001
0.001
0.001
0.0001
0.0001
0.0001
0.058
0.131
AIC
391.472
391.472
392.128
392.128
391.223
396.430
643.059
396.430
381.943
379.574
BMD
(ng/kg-day)
4.480E+02
4.480E+02
3.126E+02
3.126E+02
2.042E-K)2
8.065E+02
9.600E+02
8.065E+02
9.677E-01
7.186E-01
BMDL
(ng/kg-day)
2.844E+02
2.844E+02
1.140E+02
1.140E+02
3.585E-K)!
5.899E+02
error
5.899E+02
1.900E-01
1.157E-02
Notes

power hit bound (d = 1)

power hit bound (d = 1)
n lower bound hit (n = 1)


power bound hit (power =1)
unrestricted (n = 0.2 11)
unrestricted (power = 0.188)
a Best-fitting model, BMDS output presented in this appendix.
b Alternate model, BMDS output also presented in this appendix.
G.3.58.2. Output for Selected Model: Hill
White et al. (1986): CH50
         Hill Model.  (Version:  2.14;   Date: 06/26/2008)
         Input Data File: C:\l\71_White_1986_CH50_Hill_l.(d)
         Gnuplot Plotting File:   C:\l\71_White_1986_CH50_Hill_l.plt
                                            Tue Feb  16  20:06:45 2010
  [insert  study notes]


   The  form of the response  function is:

   Y[dose]  = intercept + v*doseAn/(kAn + doseAn)
Dependent variable = Mean
Independent variable = Dose
Power parameter  restricted to be greater  than 1
The variance  is  to be modeled as Var(i) = exp(lalpha
                                                              rho * In (mean (i)
   Total  number of dose groups  =  7
   Total  number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence  has  been set to: le-008
   Parameter Convergence has been set to: le-008
                   Default Initial  Parameter Values
                          lalpha  =       5.60999

                                      G-680

-------
                            rho =
                      intercept =
                              v =
                              n =
                              k =
        0
       91
      -74
0.0969998
       10
           Asymptotic Correlation Matrix of Parameter Estimates

           (  *** The model parameter(s)   -n
                 have been estimated at a boundary point, or have been
specified by the user,
                 and do not appear in the correlation matrix )
lalpha
lalpha
rho
intercept
V
k

-0
0
0
-0
1
.99
.19
.13
.22
rho intercept
-0.99
1
-0.2
-0.14
0.23
0.19
-0.2
1
0.33
-0.7
0.
-0.
0.

-0.
V
13
14
33
1
86
k
-0.22
0.23
-0.7
-0.86
1
                                 Parameter Estimates
Confidence Interval
Variable
Upper Conf. Limit
lalpha
7.47574
rho
1.19246
intercept
82.212
V
-33.4163
n
k
1342.9

Estimate

4.34761

0.381496

71.6585

-62.7464

1
441.016


Std. Err.

1.59601

0.413764

5.38454

14.9646

NA
460.151

NA - Indicates that this parameter has hit a bound
     implied by some inequality constraint and thus
     has no standard error.
                                                         95.0% Wald

                                                      Lower Conf. Limit

                                                              1.21948

                                                            -0.429467

                                                               61.105

                                                             -92.0765


                                                             -460.864
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled
                                    G-681

-------
   0
  10
  50
 100
 500
1000
2000
91
54
63
56
41
32
17
71
70
65
60
38
28
20
.7
.3
.3
.1
.3
.1
.2
14
8.
11
25



.1
49
.3
.5
17
17
17
19.
19.
19.
19.
17.
16.
15.
9
8
5
2
6
6
6
2
-2
-0.
-0.
0
0.
-0.
.75
.33
329
598
.43
661
589
Model Descriptions for likelihoods calculated
Model Al:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = SigmaA2

Model A2:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = Sigma(i)A2

Model A3:        Yij = Mu(i) + e(ij)
          Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
    Model A3 uses any fixed variance parameters  that
    were specified by the user

Model  R:         Yi = Mu + e(i)
           Var{e(i)} = SigmaA2
                      Likelihoods of Interest
           Model
            Al
            A2
            A3
        fitted
             R
Log (likelihood)
-181.340979
-175.820265
-181.238690
-190.611743
-212.367055
# Param' s
8
14
9
5
2
AIC
378.681959
379.640529
380.477380
391.223485
428.734109
                  Explanation of Tests

Test 1:  Do responses and/or variances differ  among  Dose  levels?
         (A2 vs. R)
Test 2:  Are Variances Homogeneous?  (Al vs A2)
Test 3:  Are variances adequately modeled?  (A2 vs. A3)
Test 4:  Does the Model for the Mean Fit?  (A3  vs.  fitted)
(Note:  When rho=0 the results of Test 3 and Test  2  will  be  the  same.

                    Tests of Interest
  Test

  Test 1
  Test 2
  Test 3
-2*log(Likelihood Ratio)  Test df
            73.0936
            11.0414
            10.8369
12
 6
 5
  p-value

 <.0001
 0.0871
0.05471
                                    G-682

-------
   Test 4
18.7461
0.0008815
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data

The p-value for Test 2 is less than .1.  A non-homogeneous variance
model appears to be appropriate

The p-value for Test 3 is less than .1.  You may want to consider a
different variance model

The p-value for Test 4 is less than .1.  You may want to try a different
model
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =        204.214

            BMDL =       35.8504
                                     G-683

-------
G.3.58.3. Figure for Selected Model: Hill

                               Hill Model with 0.95 Confidence Level
 o
 Q.
 o:
 c
 (0
 OJ
        100
         80
         60
         40
         20
                   Hill
            EiMDL
BMD
                             500
                      1000
                      dose
1500
2000
   20:0602/162010
G.3.58.4. Output for Additional Model Presented: Hill, Unrestricted
White etal. (1986): CH50
        Hill  Model.  (Version:  2.14;   Date: 06/26/2008)
        Input Data File: C:\l\71_White_1986_CH50_Hill_U_l.(d)
        Gnuplot Plotting File:   C:\l\71_White_1986_CH50_Hill_U_l.plt
                                            Tue  Feb  16 20:06:46 2010
  [insert  study notes]


   The  form of the response  function is:

   Y[dose]  = intercept + v*doseAn/(kAn + doseAn)
   Dependent  variable = Mean
   Independent variable = Dose
                                      G-684

-------
   Power parameter is not restricted
   The variance is to be modeled as Var(i) = exp(lalpha

   Total number of dose groups = 7
   Total number of records with missing values = 0
   Maximum number of iterations = 250
   Relative Function Convergence has been set to: le-008
   Parameter Convergence has been set to: le-008
                                        rho * In(mean(i))
                  Default Initial Parameter Values
                         lalpha =
                            rho =
                      intercept =
                              v =
                    5.60999
                          0
                         91
                        -74
                              n =    0.0969998
                              k =
                         10
           Asymptotic Correlation Matrix of Parameter Estimates
                 lalpha
              rho    intercept
    lalpha
-0.022
               -1
       rho
0.019
 intercept
0.0069
-0.91
-0.35
  -1
0.17
                   0.22
                  -0.42
                 -0.022
-0.17
            -0.22
             0.42
            0.019
              0.17
             -0.17
              0.16
             -0.58
            0.0069
  0.22
 -0.22
  0.16
-0.048
 -0.91
-0.42
 0.42
-0.58
             -0.048
-0.35
Confidence Interval
       Variable
Upper Conf. Limit
         lalpha
10.8266
            rho
0.821941
      intercept
100.576
              Parameter Estimates

                                      95.0% Wald

     Estimate        Std.  Err.      Lower Conf.  Limit

      6.62767          2.14235              2.42875

    -0.266376         0.555274             -1.35469

       89.579          5.61106              78.5815
                                     G-685

-------
330.93

0.309273

9.94061e+007
    -458.615

    0.210614

9.00638e+006
     402.837

   0.0503369

4.61231e+007
     -1248.16

     0.111956

-8.13933e+007
 Dose
Res .
     Table of Data and Estimated Values of Interest

            N    Obs Mean     Est Mean   Obs Std Dev  Est Std Dev
    0
   10
   50
  100
  500
 1000
 2000
                                                Scaled
91
54
63
56
41
32
17
89.6
65.4
56.3
51.5
37.9
30.8
22.9
14.1
8.49
11.3
25.5
17
17
17
15.1
15.8
16.1
16.3
16.9
17.4
18.1
0.266
-2.04
1.18
0.777
0.516
0.191
-0.927
 Model Descriptions for likelihoods calculated
 Model Al:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = SigmaA2

 Model A2:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = Sigma(i)A2

 Model A3:        Yij = Mu(i) + e(ij)
           Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
     Model A3 uses any fixed variance parameters  that
     were specified by the user

 Model  R:         Yi = Mu + e(i)
            Var{e(i)} = SigmaA2
                       Likelihoods of Interest
            Model
             Al
             A2
             A3
         fitted
              R
Log (likelihood)
-181.340979
-175.820265
-181.238690
-184.971691
-212.367055
# Param's
8
14
9
6
2
AIC
378.681959
379.640529
380.477380
381.943382
428.734109
                   Explanation of Tests

 Test 1:  Do responses and/or variances differ among Dose levels?

                                     G-686

-------
           (A2 vs. R)
 Test 2:  Are Variances Homogeneous?  (Al vs A2)
 Test 3:  Are variances adequately modeled?  (A2 vs. A3)
 Test 4:  Does the Model for the Mean Fit?  (A3 vs. fitted)
 (Note:  When rho=0 the results of Test 3 and Test 2 will be the same.
   Test

   Test 1
   Test 2
   Test 3
   Test 4
                     Tests of Interest
-2*log(Likelihood Ratio)   Test df
            73.0936
            11.0414
            10.8369
              7.466
12
 6
 5
 3
  p-value

 <.0001
 0.0871
0.05471
0.05844
The p-value for Test 1 is less than .05.  There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1.
model appears to be appropriate

The p-value for Test 3 is less than .1.
different variance model
                              A non-homogeneous variance
                              You may want to consider a
The p-value for Test 4 is less than .1.  You may want to try a different
model
        Benchmark Dose Computation

Specified effect =             1

Risk Type        =     Estimated standard deviations from the control mean

Confidence level =           0.95

             BMD =       0.967689

            BMDL =      0.189992
                                     G-687

-------
G.3.58.5. Figure for Additional Model Presented: Hill, Unrestricted

                                 Hill Model with 0.95 Confidence Level
         100
          80  -
 o
 Q.
 o:
 c
 (0
 OJ
          60
          40
          20
            BMDLBMD
                                                                              2000
   20:0602/162010
G.4. REFERENCES
Amin, S: Moore, RW: Peterson, RE; Schantz, SL. (2000). Gestational and lactational exposure to
       TCDD or coplanar PCBs alters adult expression of saccharin preference behavior in
       female rats. Neurotoxicol Teratol 22: 675-682. http://dx.doi.org/10.1016/S0892-
       0362(00)00094-5.
Bell PR: Clode. S: Fan. MQ: Fernandes. A: Foster. PM: Jiang. T: Loizou. G: MacNicoll A:
       Miller. BG: Rose. M: Tran. L: White. S. (2007). Toxicity of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin in the developing male Wistar(Han) rat. II: Chronic dosing causes developmental
       delay. Toxicol Sci 99: 224-233. http://dx.doi.org/10.1093/toxsci/kfml41.
Cantoni, L; Salmona, M; Rizzardini, M. (1981). Porphyrogenic effect of chronic treatment with
       2,3,7,8-tetrachlorodibenzo-p-dioxin in female rats. Dose-effect relationship following
       urinary excretion of porphyrins. Toxicol Appl Pharmacol 57: 156-163.
Crofton, KM; Craft ES: Hedge, JM; Gennings, C: Simmons, JE; Carchman, RA; Carter, WH, Jr;
       DeVito, MJ. (2005). Thyroid-hormone-disrupting chemicals: Evidence for dose-
       dependent additivity or synergism. Environ Health Perspect 113: 1549-1554.
                                         G-688

-------
DeCaprio. AP: McMartin. DN: O'Keefe. PW: Rej. R: Silkworm. JB: Kaminsky. LS. (1986).
       Subchronic oral toxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin in the guinea pig:
       Comparisons with a PCB-containing transformer fluid pyrolysate. Fundam Appl Toxicol
       6: 454-463. http://dx.doi.Org/10.1093/toxsci/6.3.454.
Franc. MA: Pohjanvirta. R: Tuomisto. J: Okev. AB. (2001). Persistent, low-dose 2,3,7,8-
       tetrachlorodibenzo-p-dioxin exposure: effect on aryl hydrocarbon receptor expression in
       a dioxin-resistance model. Toxicol Appl Pharmacol 175: 43-53.
       http://dx.doi.org/10.1006/taap.2001.9222.
Hoio. R: Stern. S: Zareba. G: Markowskl VP: Cox. C: Kost JT: Weiss. B. (2002). Sexually
       dimorphic behavioral responses to prenatal dioxin exposure. Environ Health Perspect
       110:247-254.
Kattainen, H; Tuukkanen, J: Simanainen, U: Tuomisto, JT: Kovero, O: Lukinmaa, P, -L:
       Alaluusua. S: Tuomisto. J: Viluksela. M. (2001). In Utero/Lactational 2,3,7,8-
       Tetrachlorodibenzo-p-dioxin Exposure Impairs Molar Tooth Development in Rats.
       Toxicol Appl Pharmacol  174: 216-224.  http://dx.doi.org/10.1006/taap.2001.9216.
Keller, JM: Huet-Hudson, YM: Leamy, LJ. (2007). Qualitative effects of dioxin on molars vary
       among inbred mouse strains. Arch Oral  Biol 52: 450-454.
       http://dx.doi.0rg/10.1016/i.archoralbio.2006.10.017.
Keller, JM: Huet-Hudson, Y; Leamy, LJ. (2008a). Effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       on molar development among non-resistant inbred strains of mice: A geometric
       morphometric analysis. Growth Development and Aging 71:  3-16.
Keller, JM: Zelditch, ML; Huet YM: Leamy, LJ. (2008b). Genetic differences in sensitivity to
       alterations of mandible structure caused by the teratogen 2,3,7,8-tetrachlorodibenzo-p-
       dioxin. Toxicol Pathol 36: 1006-1013. http://dx.doi.org/10.1177/0192623308327409.
Kociba. RJ: Keves. DG: Beyer. JE: Carreon. RM: Wade. CE: Dittenber. DA: Kalnins. RP:
       Frauson. LE: Park. CN: Barnard. SD: Hummel RA: Humiston. CG. (1978). Results of a
       two-year chronic toxicity and oncogenicity study of 2,3,7,8-tetrachlorodibenzo-p-dioxin
       in rats. Toxicol Appl Pharmacol 46: 279-303. http://dx.doi.org/10.1016/0041-
       008X(78)90075-3.
Kuchiiwa, S: Cheng, SB; Nagatomo, I; Akasaki, Y; Uchida, M; Tominaga, M; Hashiguchi, W;
       Kuchiiwa, T. (2002). In utero and lactational exposure to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin decreases serotonin-immunoreactive neurons in raphe nuclei of male mouse
       offspring. Neurosci Lett 317: 73-76. http://dx.doi.org/10.1016/S0304-3940(01)02434-X.
Latchoumycandane, C: Mathur, PP.  (2002). Effects of vitamin E on reactive oxygen species-
       mediated 2,3,7,8-tetrachlorodibenzo-p-dioxin toxicity in rat testis. J Appl Toxicol 22:
       345-351. http://dx.doi.org/10.1002/iat.866.
Li, B: Liu, HY;  Dai, LJ: Lu, JC: Yang, ZM; Huang, L. (2006).  The early embryo loss caused by
       2,3,7,8-tetrachlorodibenzo-p-dioxin may be related to the accumulation of this compound
       in the uterus. Reprod Toxicol 21: 301-306.
       http://dx.doi.0rg/10.1016/i.reprotox.2005.09.008.
Li. X: Johnson. DC: Rozman. KK. (1997). 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
       increases release of luteinizing hormone and follicle-stimulating hormone from the
       pituitary of immature female rats in vivo and in vitro. Toxicol Appl Pharmacol  142: 264-
       269. http://dx.doi.org/10.1006/taap.1996.8044.
Markowski, VP: Zareba,  G:  Stern, S: Cox, C: Weiss, B. (2001). Altered operant responding for
       motor reinforcement and the determination of benchmark doses following perinatal

                                         G-689

-------
       exposure to low-level 2,3,7,8-tetrachlorodibenzo-p-dioxin. Environ Health Perspect 109:
       621-627.
Miettinen, HM; Sorvari, R; Alaluusua, S: Murtomaa, M; Tuukkanen, J: Viluksela, M. (2006).
       The Effect of Perinatal  TCDD exposure on caries susceptibility in rats. Toxicol Sci 91:
       568-575. http://dx.doi.org/10.1093/toxsci/kfil58.
Murray. FJ: Smith. FA: Nitschke. KD: Humiston. CG: Kociba. RJ: Schwetz. BA. (1979). Three-
       generation reproduction study of rats given 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
       in the diet. Toxicol Appl Pharmacol 50: 241-252. http://dx.doi.org/10.1016/0041-
       008X(79)90149-2.
NTP (National Toxicology Program). (1982). NTP Technical Report on carcinogenesis bioassay
       of 2,3,7,8-tetrachlorodibenzo-p-dioxin in Osborne-Mendel rats and B6C3F1 mice
       (gavage study). Research Triangle Park, NC.
       http://ntp.niehs.nih.gov/?obiectid=07060172-DEB2-6542-D7CD537BAB5B2ACD.
NTP (National Toxicology Program). (2006). NTP technical report on the toxicology and
       carcinogenesis studies of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (CAS no. 1746-
       01-6) in female harlan Sprague-Dawley rats (gavage studies). (NTP TR 521; NIH
       Publication No. 06-4468). Research Triangle Park, NC.
Ohsako, S: Miyabara, Y; Nishimura, N; Kurosawa, S: Sakaue, M: Ishimura, R: Sato, M: Takeda,
       K; Aoki, Y; Sone, H; Tohyama, C: Yonemoto, J. (2001). Maternal exposure to a low
       dose of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) suppressed the development of
       reproductive organs of male rats: Dose-dependent increase of mRNA levels of 5a-
       reductase type 2 in contrast to descrease of androgen receptor in the pubertal ventral
       prostate. Toxicol  Sci 60: 132-143.
Sewall CH: Flagler. N: Vanden Heuvel JP: Clark. GC:  Tritscher. AM: Maronpot RM: Lucier.
       GW. (1995). Alterations in thyroid function in female Sprague-Dawley rats following
       chronic treatment with 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Appl Pharmacol
       132: 237-244.
Shi. Z: Valdez. KE: Ting. AY: Franczak. A: Gum. SL: Petroff. BK. (2007). Ovarian endocrine
       disruption underlies premature reproductive senescence following environmentally
       relevant chronic exposure to the aryl hydrocarbon receptor agonist 2,3,7,8-
       tetrachlorodibenzo-p-dioxin. Biol Reprod 76: 198-202.
       http://dx.doi.org/10.1095/biolreprod.106.053991.
Smialowicz, RJ: DeVito, MJ: Williams, WC: Birnbaum, LS. (2008). Relative potency based on
       hepatic enzyme induction predicts immunosuppressive effects of a mixture of
       PCDDS/PCDFS andPCBS. Toxicol Appl Pharmacol 227: 477-484.
       http://dx.doi.0rg/10.1016/i.taap.2007.ll.018.
Smith. FA: Schwetz. BA: Nitschke. KD. (1976). Teratogenicity of 2,3,7,8-tetrachlorodibenzo-p-
       dioxin in CF-1 mice. Toxicol Appl Pharmacol 38: 517-523.
       http://dx.doi.org/10.1016/0041-008X(76)90183-6.
Sparschu.  G. . L.: Dunn. F. . L.: Rowe. V. . K. (1971). Study of the teratogenicity of 2,3,7,8-
       tetrachiorodibenzo-p-dioxin in the rat. Food Cosmet Toxicol 9: 405-412.
       http://dx.doi.org/10.1016/0015-6264(71)90045-9.
Toth, K: Somfai-Relle, S: Sugar, J: Bence, J. (1979).  Carcinogenicity testing of herbicide 2,4,5-
       trichlorophenoxyethanol containing dioxin and of pure dioxin in Swiss mice. Nature 278:
       548-549.
                                        G-690

-------
Van Birgelen, AP; Van der Kolk, J: Fase, KM; Bol, I; Poiger, H; Brouwer, A; Van den Berg, M.
       (1995). Subchronic dose-response study of 2,3,7,8-tetrachlorodibenzo-p-dioxin in female
       Sprague-Dawley rats. Toxicol Appl Pharmacol 132: 1-13.
       http://dx.doi.org/10.1006/taap.1995.1080.
White, KL, Jr; Lysy, HH; McCay, JA; Anderson, AC. (1986). Modulation of serum complement
       levels following exposure to polychlorinated dibenzo-p-dioxins. Toxicol Appl Pharmacol
       84: 209-219. http://dx.doi.org/10.1016/0041-008X(86)90128-6.
                                        G-691

-------
O EPA                                 EPA/600/R-10/038F
\^d /	\                                   www.epa.gov/iris
                 APPENDIX H

 Endpoints Excluded From Reference Dose
Derivation Based on Toxicological Relevance
                     January 2012
             National Center for Environmental Assessment
                Office of Research and Development
               U.S. Environmental Protection Agency
                      Cincinnati, OH

-------
 CONTENTS—Appendix H: Endpoints Excluded from Reference Dose Derivation Based
                          on Toxicological Relevance
APPENDIX H. ENDPOINTS EXCLUDED FROM REFERENCE DOSE
      DERIVATION BASED ON TOXICOLOGICAL RELEVANCE	H-l
      H.l.BURLESONET AL. (1996)	H-l
      H.2. DEVITOETAL. (1994)	H-2
      H.3.HASSOUNETAL. (2003; 2002; 2000; 1998)	H-2
      H.4. HONG ETAL. (1989)	H-2
      H.5.KITCHIN AND WOODS (1979)	H-3
      H.6. LATCHOUMYCANDANE ETAL. (2003)	H-3
      H.7. LUCIER ETAL. (1986)	H-4
      H.8. MALLY AND CHIPMAN (2002)	H-4
      H.9. SEW ALL ETAL. (1993)	H-5
      H.10. SLEZAK ET AL. (2000)	H-5
      H.ll. SUGITA-KONISHI ET AL. (2003)	H-6
      H.12. TRITSCHER ETAL. (1992)	H-7
      H.13. VANDENHEUVEL ETAL. (1994)	H-7
      H. 14. REFERENCES	H-8
                                   H-ii

-------
     APPENDIX H. ENDPOINTS EXCLUDED FROM REFERENCE DOSE
               DERIVATION BASED ON TOXICOLOGICAL RELEVANCE
       The National Academy of Sciences committee commented on the low dose model
predictions and the need to discuss the biological significance of the noncancer health effects
modeled in the 2003 Reassessment. In selecting point of departure (POD) candidates from the
animal bioassays for derivation of the reference dose (RfD), U.S. Environmental Protection
Agency (EPA) had to consider the toxicological relevance of the identified endpoint(s) from any
given study. Often endpoints/effects may be sensitive, but lack general toxicological
significance due to not being clearly adverse (defined in the Integrated Risk Information System
glossary as a biochemical change, functional impairment, or pathologic lesion that affects the
performance of the whole organism, or reduces an organism's ability to respond to an additional
environmental challenge), being an adaptive response, or not being clearly linked to downstream
functional or pathological alterations.  It is standard EPA RfD derivation policy not to base a
reference value on endpoints that are not adverse or not obvious precursors to an adverse effect.
For select studies, a rationale for lack of toxicological relevance of particular endpoints reported
is listed here.  These endpoints were not considered for derivation of the RfD.

H.l. BURLESON ET AL. (1996)
       Burleson et al. (1996) analyzed the effect of a 2,3,7,8-tetrachlorodibenzo-^-dioxin
(TCDD) on viral host resistance following a single gavage dose of TCDD by measuring
mortality mediated by influenza virus challenge in B6C3Fi female mice.  The study authors
found that TCDD at >10 ng/kg-day increased influenza-induced mortality. The experimental
design calls for a 30% mortality in untreated animals (15% was achieved); mortality, itself, is not
a direct result of TCDD exposure. None of the other immunologically-relevant measures were
affected by TCDD treatment in this study, and no other effects were reported.  The interpretation
of these results with respect to humans is problematic. Furthermore, the findings were not
reproduced by Nohara et al. (2002) using the same experimental design (see  Section 2.4.2 and
Table 2-4). Therefore, this endpoint is not considered relevant as a POD candidate.
                                          H-l

-------
H.2. DEVITO ET AL. (1994)
       Devito et al. (1994) assessed the activity of CYP1A1 and CYP1A2, the amount of
phosphorylation of phosphotyrosyl proteins (pp32, pp34, and pp38), and the levels of estrogen
receptor in the liver, uterus, lung and skin tissue of female B6C3Fi mice administered TCDD for
5 days a week for 13 weeks. The authors hypothesized that these measurements may be
sensitive biomarkers for exposure to TCDD.  Body weights were also recorded weekly.
Induction of CY1A1 and CYP1A2, as well as increased phosphorylated forms of pp32, pp34,
and pp38 were sensitive indicators of TCDD  exposure, with statistically significant changes seen
at 1.07 ng/kg-day. 7-ethoxyresorufin-O-deethylase (EROD) activity in the ling, skin, and liver
was also observed with significant increases at this dose. However, the authors did not find a
change in rat body or terminal organ weights, nor did they note any pathology in the animals at
this dose level. The role of cytochrome P450s (CYPs) and phosphorylated pp32, pp34, and pp38
in TCDD-mediated toxicity is unknown, and  changes in the activity or function of these proteins
are not considered adverse. Therefore, these  endpoints are not considered suitable as PODs.

H.3. HASSOUN ET AL.  (2003; 2002; 2000; 1998)
       In multiple studies by Hassoun et al. (2003: 2002: 2000: 1998). various indicators of
oxidative stress were measured in hepatic and brain tissue of female B6C3Fi mice and
Sprague-Dawley rats following 13 or 30 weeks of TCDD gavage dosing (5 days a week).
Biomarkers for oxidative stress included production superoxide anion, lipid peroxidation, and
DNA single-strand breaks.  The authors report a statistically significant effect on several
oxidative stress markers as a result of TCDD  exposure, the lowest dose producing an effect being
0.32 ng/kg-day (1998).  In this study, all oxidative stress markers were significantly affected, but
no other indicators of brain pathology were assessed.  Thus, it is impracticable to link the
markers of oxidative stress to a toxicological  outcome in the brain, and this study and its
endpoints are not considered relevant POD candidates.

H.4. HONG ET AL. (1989)
       Hong  et al. (1989) studied the immunotoxicity of TCDD in female adult rhesus monkeys
administered 0.12 or 0.67 ng/kg-day TCDD in feed for 4 years. Additionally, offspring from
exposed mothers were examined. In adult monkeys, an increased number of T lymphocytes
                                         H-2

-------
were observed in the 0.67 ng/kg-day dose group, but there was not a proportional increase in
each of the T cells subsets.  Macrophage depletion in the 0.12, and 0.67 ng/kg-day groups
resulted in the absence of amplification in a mixed lymphocyte response assay, compared to a
fivefold amplification in control monkeys. In the offspring, there was an immune
hyperresponsiveness to tetanus toxoid immunization which correlated with TCDD tissue levels.
Although a thorough immunological investigation, in the absence of any relevant
immunotoxicity endpoints or functional decrements of immune function following TCDD
exposure, there are no suitable endpoints for consideration as candidate PODs in this study.

H.5. KITCHIN AND WOODS (1979)
       Kitchin and Woods (1979) administered female Sprague-Dawley rats a single gavage
dose of TCDD and measured CYP levels and benzo[a]pyrene hydroxylase (BPH) activity as a
marker of hepatic microsomal cytochrome P448-mediated enzyme activity.  They found a
statistically significant increase in BPH at doses >2 ng/kg and a significant increase in
cytochrome P450 levels at doses >600 ng/kg. Aryl hydrocarbon hydrolase and EROD were both
significantly increased 3 months after exposure; however the elevation did not maintain
statistical significance at 6 months. No other indicators of hepatic effects were analyzed.  CYP
induction alone is not considered a significant lexicologically adverse effect given that CYPs are
induced as a means of hepatic processing of xenobiotic agents. Additionally, the role of CYP
induction in hepatotoxicity and carcinogenicity of TCDD  is unknown, and CYP induction is not
considered a relevant POD  without obvious pathological significance.

H.6. LATCHOUMYCANDANE ET AL. (2003)
       Latchoumycandane et al. (2003) examined the induction of oxidative stress in epididymal
sperm of male Wistar rats.  The activities of antioxidant enzymes including superoxide dismutase
(SOD), catalase (CAT), glutathione reductase (GRX), and glutathione peroxidase (GPX), as well
as the oxidative stress indicators hydrogen peroxide (H2O2) and lipid peroxidation (LPX)  were
measured in epididymal sperm, caput epididymis, corpus epididymis, and cauda epididymis
following gavage dosing of 0, 100, 1,000, and  10,000 ng/kg-day TCDD for 4 consecutive days.
No significant changes in epididymal sperm counts were evident at any dose tested compared to
control. SOD, CAT, GRX, and GPX activities were significantly decreased at doses
                                          H-3

-------
> 1,000 ng/kg-day in epididymal sperm. H2O2 and LPX were significantly increased at all doses
tested. SOD, CAT, GRX, and GPX activities were significantly decreased only at the highest
dose in the caput epididymis and corpus epididymis, but were significantly decreased at all doses
tested in the cauda epididymis.  Conversely, H2O2 and LPX were significantly increased only at
the highest dose in the caput epididymis and corpus epididymis, but were significantly increased
at all doses tested in the cauda epididymis. Although several oxidative stress indicators were
significantly changed in this study, sperm count was not altered, and no other indices of sperm
function were assessed; it is unfeasible to link the markers of oxidative stress to a
TCDD-induced toxicological outcome. Therefore, these endpoints are not considered relevant as
POD candidates.

H.7. LUCIER ET AL. (1986)
       Because TCDD had been detected in the soil of contaminated locations,  determining the
bioavailability of TCDD from ingested soil may be important to the calculation  of safe exposure
levels. Lucier et al. (1986) fed adult female Sprague-Dawley rats TCDD contaminated soil or
gave them TCDD in corn oil at various doses and compared the effects of TCDD on biochemical
parameters from liver tissue.  They found that equivalent doses of TCDD in corn oil and soil
produced similar increases in hepatic aryl hydrocarbon hydroxylase activity (AHH) and uridine
diphosphate (UDP) glucuronyltransferase activity. They determined that AHH was statistically
induced 1.8-fold at 15 ng/kg in corn oil and 40 ng/kg in soil. Cytochrome P450 was significantly
increased at higher doses.  No clinical signs of acute toxicity or changes in body weight were
observed.  The association between AHH activity and TCDD-mediated hepatotoxicity is
unknown and no adverse endpoints were measured. Thus, this endpoint is not suitable as a POD
candidate.

H.8. MALLY AND CHIPMAN (2002)
       Mally and Chipman (2002) evaluated the effect of TCDD on gap junctions,
hypothesizing that as a nongenotoxic carcinogen,  TCDD may induce tumor formation by
disturbing tissue homeostasis. Female F344 rats were dosed with TCDD by oral gavage for
either 3 consecutive days or 2 days a week for 28  days. Gap junction connexin (Cx) plaque
expression and hepatocyte proliferation was measured. The study authors report a decrease in
                                          H-4

-------
Cx32 plaque number and area in the liver of rats exposed to 0.7 ng/kg-day and higher, however
they did not find an associated increase in hepatocyte proliferation. No clinical signs of toxicity
were observed, and histological examination of the liver revealed no abnormalities.  In the
absence of additional indicators of hepatotoxicity,  a decrease in Cx32 plaque formation is not
clearly linked to TCDD-mediated hepatotoxicity or hepatocarcinogenicity, nor is it considered an
adverse effect. This endpoint is not considered a lexicologically relevant POD.

H.9. SEW ALL ET AL. (1993)
       Sewall et al. (1993) investigated alterations in the epidermal growth factor receptor
(EGFR) pathway in a two-stage initiation promotion model of TCDD hepatic cancer. EGFR
signaling has been implicated in the altered cell growth induction by tumor promoters.  Female
Sprague-Dawley rats were administered TCDD biweekly by oral gavage for 30 weeks following
initiation by a single dose of diethylnitrosamine (DEN). A group also received TCDD without
prior DEN initiation. Livers were harvested and fixed from sacrificed animals and sections
tested for EGFR binding, autophosphorylation, immunolocalization, and hepatic cell
proliferation.  The authors report a significant dose-dependent decrease in plasma membrane
EGFR maximum binding capacity in TCDD-exposed rats beginning at 3.5 ng/kg-day. However,
at this same dose, the authors note a statistically significant decrease in cell proliferation (as
measured by DNA replication labeling), with increases in proliferation only occurring at higher
doses (125 ng/kg-day).  No other indicators of hepatic toxicity or tumorigenicity were assessed.
The role of EGFR in TCDD-mediated hepatotoxicity and hepatocarcinogenicity is unknown, and
as such, this endpoint cannot be unequivocally linked to TCDD-induced hepatic effects nor
labeled as adverse.  Thus,  it is not suitable as a POD  candidate.

H.10. SLEZAK ET AL. (2000)
       Slezak et al. (2000) studied the impact of subchronic TCDD exposure on oxidative stress
in various organs of B6C3Fi female mice.  The oxidative stress indicators superoxide anion
(SA), lipid peroxidation (measured through formation of thiobarbituric acid reactive substances
[TEARS]  in tissue homogenates), ascorbic acid (AA), and total glutathione (GSH) were
measured in liver, lung, kidney, and spleen following gavage dosing for 13 weeks (5 days a
week). Tissue TCDD concentrations also were measured. Significant TCDD-induced changes
                                          H-5

-------
in the liver included decreased SA and GSH at 0.15 ng/kg-day, increased GSH at
0.45 ng/kg-day, increased SA and AA at 15 and 150 ng/kg-day, and increased GSH and TEARS
at 150 ng/kg-day. Unlike the liver, there was no significant increase in SA in the lung, but SA
was significantly decreased at 0.45,  15, and 150 ng/kg-day. Lung GSH and AA were decreased
at 0.15 ng/kg-day, while AA was increased at 15 and 150 ng/kg-day.  In the kidney, SA was
increased at 15 and 150 ng/kg-day.  Renal  GSH, like the liver and the lung, was decreased at
0.15 ng/kg-day with this trend continuing at 0.45 and 1.5 ng/kg-day, and AA levels were lower at
all doses except 1.5 ng/kg-day.  In the spleen, SA was unchanged, GSH was increased at
150 ng/kg-day, and AA was decreased at 0.15, 1.5, and 150 ng/kg-day. Although several
oxidative stress indicators were significantly changed in this study, no other indices of liver,
lung, kidney, or spleen pathology were measured, and it is unfeasible to link the markers of
oxidative stress to a TCDD-induced toxicological outcome in the organs assessed.  Therefore,
these endpoints are not considered relevant as POD candidates.

H.11. SUGITA-KONISHI ET AL. (2003)
       Sugita-Konishi et al. (2003)  investigated the change in host resistance of mice offspring
lactationally exposed to TCDD.  Pregnant  C57BL/6NCji mice were administered TCDD via
drinking water from parturition to weaning of the offspring (17 days).  One group of offspring
was then infected with Listeria monocytogenes and blood and spleen samples were collected
various time points post infection. Uninfected, TCDD exposed offspring were weighed and their
spleens and thymuses removed for assay of cellular content and protein expression. TCDD
exposure caused a statistically-significant decrease in relative spleen weight and a
statistically-significant increase in thymic CD4+ cells in the high-dose group (11.3 ng/kg-day).
Offspring infected with Listeria following TCDD exposure exhibited a statistically significant
increase in serum tumor necrosis factor alpha 2 days after infection in both sexes in the low-
(1.14 ng/kg-day) and high-dose groups.  The authors conclude that exposure to TCDD disrupted
the  host resistance of the offspring at the lowest dose tested, despite the primary immune
parameters being unaffected. Without an obvious  association between TCDD and immune
function, however, this endpoint is not suitable for identification of a
lowest-observed-adverse-effect level (LOAEL). Thus, the LOAEL for this study is
11.3 ng/kg-day, and the no-observed-adverse-effect level is 1.14 ng/kg-day.
                                          H-6

-------
H.12. TRITSCHER ET AL. (1992)
       Tritscher et al. (1992) performed an initiation-promotion study in female
Sprague-Dawley rats. Rats were initiated with an i.p. injection of DEN or saline, followed
2 weeks later by promotion with biweekly administration of TCDD via gavage for 30 weeks.
Hepatic cytochrome P450 levels (CYP1 Al and CYP1A2) and EROD activity were quantified,
and immunohistochemical detection of CYP1A1 and CYP1A2 in liver was also conducted.
Liver TCDD concentrations were also analyzed.  A dose-response trend for increased liver
CYP1A1 and CYP1A2 protein was observed in initiated and noninitiated rats as assessed by
microsomal quanitification and immunihistochemical staining.  A strong relationship between
liver TCDD concentration and CYP1 Al and CYP1A2 protein levels and EROD activity was also
observed in DEN/TCDD-treated rats.  CYP induction alone is not considered a significant
lexicologically adverse effect given that CYPs are induced as a means of hepatic processing of
xenobiotic agents. Additionally, the role of CYP induction in the hepatotoxicity and
carcinogenicity of TCDD is unknown, and CYP induction is not considered a relevant POD
without obvious pathological significance.

H.13. VANDEN HEUVEL ET AL. (1994)
       Vanden Heuvel et al. (1994) analyzed changes in hepatic messenger ribonucleic acid
(mRNA) following a single administration of TCDD to female Sprague-Dawley rats by oral
gavage. Four days after treatment, animals were sacrificed and livers were excised. Using
reverse transcriptase-polymerase chain reaction on hepatic ribonucleic acid, they compared
levels of "dioxin responsive" mRNA's (CYP1A1, UDP-glucuronosyltransferase I, plasminogen
activator inhibitor 2, and transforming growth factor a) at various doses of TCDD and at control
(baseline) levels. They determined that CYP1 Al  elicited the most sensitive response to  TCDD,
with a statistically significant increase (threefold) in mRNA from rat livers exposed to
1 ng/kg-day TCDD. Induction of CYP1A1 expression is not considered an adverse effect, as the
role of CYP1A1 in TCDD-mediated carcinogenicity is unsettled. Therefore,  in the absence of
other indicators of hepatoxicity, increases in liver CYP1 Al cannot be considered lexicologically
relevant for a POD candidate.
                                         H-7

-------
H.14. REFERENCES
Burleson. GR: Lebrec. H: Yang. YG: Ibanes. JD: Pennington. KN: Birnbaum. LS. (1996). Effect
       of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on influenza virus host resistance in
       mice. Fundam Appl Toxicol 29: 40-47.
DeVito, MJ: Ma, X; Babish, JG: Menache, M; Birnbaum, LS. (1994). Dose-response
       relationships in mice following subchronic exposure to 2,3,7,8-tetrachlorodibenzo-p-
       dioxin: CYP1A1, CYP1A2, estrogen receptor, and protein tyrosine phosphorylation.
       Toxicol Appl Pharmacol 124: 82-90.
Hassoun. EA: Wilt SC: Devito. MJ: Van Birgelen. A: Alsharif NZ: Birnbaum. LS: Stohs. SJ.
       (1998). Induction of oxidative stress in brain tissues of mice after subchronic exposure to
       2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Sci 42: 23-27.
       http://dx.doi.0rg/10.1093/toxsci/42.l.23.
Hassoun, EA: Li, F; Abushaban, A; Stohs,  SJ. (2000). The relative abilities of TCDD and its
       congeners to induce oxidative stress in the hepatic and brain tissues of rats after
       subchronic exposure. Toxicology 145: 103-113.
Hassoun, EA: Wang, H; Abushaban, A; Stohs, SJ. (2002). Induction of oxidative stress
       following chronic exposure to TCDD, 2,3,4,7,8-pentachlorodibenzofuran, and
       2,3',4,4',5-pentachlorobiphenyl. J Toxicol Environ Health A 65: 825-842.
Hassoun, EA: Al-Ghafri, M; Abushaban, A. (2003). The role of antioxidant enzymes in TCDD-
       induced oxidative stress in various brain regions of rats after subchronic exposure. Free
       Radic Biol Med 35: 1028-1036. http://dx.doi.org/10.1016/50891-5849(03)00458-1.
Hong, R; Taylor, K; Abonour, R. (1989). Immune abnormalities associated with chronic TCDD
       exposure in rhesus. Chemosphere 18: 313-320. http://dx.doi.org/10.1016/0045-
       6535(89)90136-7.
Kitchin, KT: Woods. JS. (1979). 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) effects on hepatic
       microsomal cytochrome P-448-mediated enzyme activities. Toxicol Appl Pharmacol 47:
       537-546. http://dx.doi.org/10.1016/0041-008X(79)90524-6.
Latchoumycandane, C: Chitra, KC: Mathur, PP.  (2003). 2,3,7,8-Tetrachlorodibenzo-p-dioxin
       (TCDD) induces oxidative stress in the epididymis and epididymal sperm of adult rats.
       Arch Toxicol 77: 280-284. http://dx.doi.org/10.1007/s00204-003-0439-x.
Lucier, GW: Rumbaugh, RC: McCoy. Z: Hass, R: Harvan, D: Albro, P. (1986). Ingestion of soil
       contaminated with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) alters hepatic enzyme
       activities in rats. Fundam Appl Toxicol 6: 364-371.
Mally, A; Chipman, JK. (2002). Non-genotoxic carcinogens: Early effects on gap junctions, cell
       proliferation and apoptosis in the rat. Toxicology 180: 233-248.
Nohara. K: Izuml H: Tamura. S: Nagata. R: Tohyama. C. (2002). Effect of low-dose 2,3,7,8-
       tetrachlorodibenzo-p-dioxin (TCDD) on influenza A virus-induced mortality in mice.
       Toxicology 170: 131-138.
Sewall, C: Lucier, G: Tritscher, A;  Clark, G. (1993). TCDD-mediated changes in hepatic
       epidermal growth factor receptor may be a critical event in the hepatocarcinogenic action
       of TCDD. Carcinogenesis 14: 1885-1893.
Slezak. BP: Hatch. GE: DeVito. MJ: Diliberto. JJ: Slade. R: Crissman. K: Hassoun. E:
       Birnbaum, LS. (2000).  Oxidative stress in female B6C3F1 mice following acute and
       subchronic exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Toxicol Sci 54:
       390-398.

                                         H-8

-------
Sugita-Konishi. Y: Kobayashi. K: Naito. H: Miura. K: Suzuki. Y. (2003). Effect of lactational
       exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin on the susceptibility to Listeria infection.
       Biosci Biotechnol Biochem 67: 89-93.
Tritscher. AM: Goldstein. JA: Portier. CJ: McCoy. Z: Clark. GC: Lucier. GW. (1992). Dose-
       response relationships for chronic exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin in a rat
       tumor promotion model: Quantification and immunolocalization of CYP1A1 and
       CYP1A2 in the liver. Cancer Res 52: 3436-3442.
Vanden Heuvel JP: Clark. GC: Tritscher. A: Lucier. GW. (1994). Accumulation of
       polychlorinated dibenzo-p-dioxins and dibenzofurans in liver of control laboratory rats.
       Fundam Appl Toxicol 23: 465-469. http://dx.doi.Org/10.1093/toxsci/23.3.465.
                                          H-9

-------
 C EZDA                                       EPA/600/R-10/038F
\Xd r	V                                        www.epa.gov/iris
                     APPENDIX I
             Literature Search Terms
                        January 2012
               National Center for Environmental Assessment
                   Office of Research and Development
                  U.S. Environmental Protection Agency
                         Cincinnati, OH

-------
               CONTENTS—Appendix I: Literature Search Terms
APPENDIX I.  LITERATURE SEARCH TERMS	1-1
      I.I. REFERENCES	1-10
                                   I-ii

-------
                APPENDIX I.    LITERATURE SEARCH TERMS

       The U.S. Environmental Protection Agency (EPA) has developed a literature database of
peer reviewed studies on 2,3,7,8-tetrachlorodibenzo-/>-dioxin (TCDD) toxicity, including in vivo
mammalian dose response studies and epidemiologic studies for use in quantitative TCDD
dose-response assessment and supporting qualitative discussions.  An initial literature search for
studies published since the 2003 Reassessment was conducted by the U.S. Department of
Energy's Argonne National Laboratory (ANL) through an Interagency Agreement with EPA.
ANL used the online National Library of Medicine database (PubMed) and identified studies
published between the year 2000 and October 31, 2008.
       EPA published the initial literature search results in the Federal Register on November
24, 2008 (U.S. EPA. 2008) and invited the public to review the list and submit additional peer
reviewed in vivo mammalian dose response studies for TCDD, including epidemiologic studies
that were absent from the list (U.S. EPA, 2008).  Submissions were accepted by the EPA through
an electronic docket, email  and hand delivery, and were evaluated for use in TCDD dose-
response assessment.
       This appendix contains the search terms utilized by ANL in conducting the literature
search.
                                          1-1

-------
                                                     LITERATURE SEARCH TERMS
1746-01-6
2,3,7,8-TCDD; TCDD
dioxin
absorbed, absorbed dose
absorbed, absorption
accident
acetylcholine
acetylcholinesterase
acute
acute myocardial infarction
adenocarcinoma
adenoma
adipose
administered
administered, administered dose
adrenal
adrenal (gland, cortex)
adverse
age
agent orange
agonist
Ah, aryl hydrocarbon, Ah receptor
AhR, arylhydrocarbon receptor
alveolar
alveolar duct
alveoli
AMI
anamnestic response
anemia
animal, animal stud
antibody
antigen
antigen presenting cell
antigenic
aorta
apoptosis
arcuate nucleus
area under curve
artery
atheromatous plaque
atria
atrioventricular
atrioventricular fistula
atrioventricular node
atrioventricular opening
atrioventricular valve
atrium
atrophy
AUC, area under the curve
autoimmune
Bcell
                                                   B-cell
beagle
behavior
behavioral
behavioral abnormalities
benchmark (see BMC, BMP, others)
benign
bicuspid
bicuspid valve
bile
bile, biliary
bile, biliary
biliary
binding
bioaccumulation
bioavailability, bioavailable
bioavailable, bioavailability
biochem, biochemical
biological half-life
biotransformation
blind
blood
blood cells
blood concentration
blood pressure
blood, blood concentration
BMC, benchmark concentration
BMP, benchmark dose
BMDL
BMR, benchmark response
body burden
body weight
bolus
                                                   bone
bowel
brain
brain aromatase
brain stem
brain tissue
brain tissues
brainstem
breast milk
breast milk, lactation, milk
                                                                                                      bronchi
bronchial
bronchial tree
bronchiole
CA, cancer, carcino, carcinogen

-------
                                             LITERATURE SEARCH TERMS (continued)
cancer
carcinogen
carcmogenesis
carcinogenic
carcinoma
cardiac
cardiac arrest
cardiac cycle
cardiac notch
cardio
cardio (myopathy), cardiovascular, CV
cardiogenic
cardiogenic plate
cardiomyopathy
cardiovascular
cardiovascular disease
case report
CD4
CDS
cell, cell line, cell proliferation
cell-mediated immune response
central nervous system
cerebellar
cerebral
cerebrum
chloracne
cholesterol
chordae tendineae
chronic
chronic lymphocytic leukemia
chronic obstructive pulmonary disease
cirrhosis
cirrhotic
cleft
                                                   clinical
cognition
cognitive
cognitive abnormalities
                                                   cohort
colitis
colon
compartment
concentration, peak
conjugate
contaminant, contamination, contaminated
control
COPD
COPD, chronic obstructive pulm disease
coplanar, coplanarPCB(s)
                                                   cornea
corneal
                                                   coronary
cortical
cortical asymmetry
cortical cells
cortical thickness
count
critical
culture, tissue culture
cuspid
cutaneous
CV
CVD
CVD (CV), cardiovascular disease
CYP, cytochrome P450
cytochrome, CYP (1A1, 1A2)
cytokine
dam
                                                   deficit
                                                                                                      defoliant
degeneration
delayed-type hypersensitivity reaction
                                                   dendrite
dendritic
dentition
depot
depot
dermal
dermal, dermis, transdermal
dermal, transdermal, skin
dermis
developing
developmental
                                                   developmental, developmental effect
diabetes
                                                   diabetic
dialysis
diaphragm
diastole
diet, dietary
dietary, ingestion
differentiation, cell differentiation
diffusion, permeability
disease
disposition
distribute, distributed, distribution
DLC, dioxin-like compound
dog
dorsal raphe nuclei
dose response, dose-response
dose, dose metric, dose-dependent
dose, dose-dependent

-------
                                              LITERATURE SEARCH TERMS (continued)
dose-dependent
duodenum
dysplasia
ED, effective dose
edema
effect, effect level
eliminate, eliminated, elimination
embryo
embryo, embryotox(ic), embryonic
embryonic
embryotoxic
endo, endocrine, endocrine disrupt(or/ion)
endocarditis
endocrine
endocrine disrupter
endocrine disrupting
endocrine disruption
endocrine disruptor
endocrinology
endometrial
endometriosis
endometriosis
enterohepatic
enzyme
epidemiol, epidemiologic
epidermal
epidermis
equilibrium
ER
EROD
EROD, ethoxyresorufin-o-deethylase
estrogen
estrogen receptor
estrogen, ER, estrogen receptor
ethoxyresorufm-O-deethylase
excrete(d), excretion
excrete, excreted, excretion
eye
fat
fat, fatty
fate
fatty
fecal
fecal, feces
feces
fecundity (2 spellings?)
FEL, frank effect, frank effect level
female
fertility
fetal
fetal, feto, fetotox, fetotoxic, fetus
fetotoxic
fetus
FEV
fish
foci
food consumption
forced expiratory volume
forebrain
fraction
fraction, ratio
function
furan, furans
gastritis
gastrointestinal
gastrointestinal, GI, gut
gastrointestine
gastrointestine, gastrointestinal, GI, gut
gavage
gavage, bolus
GD
gender
genotox, genotoxicity
genotoxic
genotoxicity
gerbil
gestation
gestation, gestational, gestational day, GD
gestational	
gestational day
GI
glial cells
glomerular
glomerulus
glucagon
gonadotropin
granule neuroblast
gravid
growth hormone
gut
haematology
haematopoiesis
haemopoeisis
haemopoeitic
half-life, half life, half-lives
half-life, half-lives
hamster
hamster (Syrian golden)
HDL
HDL, high-density lipoprotein
health

-------
                                             LITERATURE SEARCH TERMS (continued)
heart
heart attack
heart disease
heart murmur
hematology
hematopoiesis
hemoglobin
hemopoeisis
hemopoeisis, hematopoiesis / poeitic
hemopoeitic
hemorragic
hemorrhage
hemorrhage, hemorragic
hemotoxin
hepatic
hepatic enzyme
hepatic, hepato(cyte), hepatotox(ic)(ity)
hepatic, liver
hepatocyte
hepatoma
hepatotoxicity
hepatoxic
herbicide
high blood pressure
high density lipoprotein
high-density lipoprotein
hippocampus
histologic, histopathologic, histopath
Hodgkins (2 spellings)
hormone, hormone
hospital
human
human, human stud
humoral immune response
hydronephrosis
hydroxylate(ion)
hyperglycemia
hyperglycemia, hypoglycemia
hyperglycemic
hyperplasia
hyperplasia, hypertrophy
hypersensitivity reaction
hypersensitized
hypertension
hypertrophy
hypertrophys
hypoglycemia
hypoglycemic
hypotension
hypothalamus
hypothalamus-preoptic area
IL
IL 5, interleukin 5
ileitis
ileum
immune
immune regulation
immune response
immune suppression
immune system
immune, immuno, immunological
immunocompromised
immunoglobulin
immunologic
immunological
immunology
immunosuppression
immunosuppressive
immunotox, immunotoxicity
immunotoxic
immunotoxicity
implantation
impurity, impurities, impure
in vitro, in vivo
individual
induce(d), inducible, induction
induce(d), inducible, induction, indue
infant
infection
infertility
inflammation
inflammatory
inflammatory lesion
inflammatory, inflammation
influenza
ingestion
inhal, inhalation
inhibition
injection
                                                  instillation
instillation, trachea! instillation
                                                  insulin
                                                  interleukin
intermediate
intermediate, reactive intermediate
intestinal
intestine
intraperitoneal, ip
intravenous, iv
involuntary muscle
IP, intraperitoneal
islets of Langerhorn

-------
                                             LITERATURE SEARCH TERMS (continued)
IV, intravenous
jaw
jejunum
keratitis, keratitic, keratin(ized), kerat
kidney
kinetic
Kupffer
lactat(ion), lactate, lactational
lactation
lactational
large intestine
LC, lethal concentration
LD, lethal dose
LDL
LDL, low-density lipoprotein
lesion
lethality
leukemia
leukemia, leukemic
leukemic
lipid
lipophilic
lipophilic, lipophilicity
lipophilicity
liver
liver enzyme
LOAEL, LOEL
lobes
low blood pressure
low density lipoprotein
low-density lipoprotein
low-dose
lung
lymph node
lymph, lymphatic
lymphocyte
lymphoid
lymphoid organs
lymphoma
                                                  macaque
macrophage
major histocompatibility complex
male
malignancy
malignant
malignant, malignancy
mammal
                                                  mammary
mammary gland
mammary, mammary gland
man
mandible
marker
mating behavior
mechanism, mechanistic (see MO A)
median raphe nuclei
men
metabolic
metabolism, metabolite, metabolize
metabolite
metaplasia
methoxyresorufin-O-deethylase
MHC
MI
mice (several strains)
microsome, microsomal
mink
mitral
mitral regurgitation
                                                                                                    mitral valve
MO A, mode (mechanism) of action
model
molar
                                                  monkey (rhesus)
mortality
motor development
mouse (incl. Swiss)
MR
MROD
Mrp, multidrug resistance-assoc protein
                                                  mucosa
                                                  mucosa, mucosal, oral mucosa
mucosal
                                                                                                    muscosa
                                                  muscosal
muta, mutagen, mutation
mutagen
mutation
myeloid leukemia
myocardial
myocardial infarction
myocardium
myocyte
nasal
nasal (turbinates)
nasal turbinates
natural killer
neocortical
neonatal
neoplasia
neoplasm
neoplasm, neoplast, neoplastic, neoplasia

-------
                                             LITERATURE SEARCH TERMS (continued)
neoplastic
nephron
nerve
nerve conductance
nerve conduction
nerves
neural
neural activity
neuro, neurologic
neuroblast
neuroblastoma
neurochemical
neurodevelopment
neurological
neuropathy
neuropeptides
neuropsychological
neurotox, neurotoxic, neurotoxicity
neurotoxic
neurotoxicity
neurotransmitters
neurotrophic factor
neutrophil
NK
NOAEL, NOEL
nonca, noncancer, noncarcmogemc
non-Hodgkins lymphoma (4 spellings)
NTS
nuclear receptor
nucleus of solitary tract
occupational
ocular
olfactory bulb
oncogen
oncogene
oncogemc
optic
oral
oral mucosa
                                                  organ
osteo
osteoblast
osteosarcoma
                                                  ovary
palate
palate, palat
pancreas
pancreatic
pancreatitis
papillary muscle
papilloma
paraventricular nucleus
parent
parenteral
partition, partitionong
pathol, pathology
pathway
patient
PB, physiol, physiologically based
                                                  PBPK
PCB, polychlorinated biphenyl
PD, pharmacodynamic
peak
peak, peak dose
people
percent
pericardium
perinatal
peripheral nervous system
peripheral neuropathy
                                                                                                    person
pesticide
physiological
                                                  pig, guinea pig (Hartley)
pituitary
pituitary hormone
PK, pharmacokinetic
                                                  plasma
PND
PND, postnatal day
POD, point of departure
polymorphism, polymorph
polyneuropathy
POP, persistent organic pollutant
population
porphyrin, porphyria
postnatal
postnatal day
potency, potent
pregnancy
pregnant
pregnant, pregnancy
prenatal
                                                  preoptic area
primate
product, production
profile
progesterone
proliferation
promotion, promoter, promote, promoting
public	
pulmonary

-------
                                                   LITERATURE SEARCH TERMS (continued)
oo
     pulmonary artery
     pulmonary edema
     pulmonary embolism
     pulmonary epithelium
     pulmonary valve
     pulmonary vein
     pulmonary, transpulmonary
     pup
     pup survival
     rabbit
     rat (several strains)
     rate
     rate, time, time-dependent
     ratio, fraction
     reactive (intermediate)
     reactive oxygen species
     receptor, receptor mediated
     red blood cells
     regenerate, regeneration, regen
     regeneration
     renal
     repro, reproductive, reprotox
     reproduction
     reproductive
     reprotoxic
     respiration
     respiratory
     respiratory, respired air
     respired air
     response
     retina
     retinal
     rhabdomyosarcoma
     risk, risk analysis, risk assessment
rodent
ROS
sarcoma
SCC
SCC, squamous cell carcinoma
sensitive, sensitivity
sequestration
                                                        serum
sex
sex ratio
sheep red blood cells
short term
sight
signal, signaling
skeletal
skeleton
skin
skin
small intestine
smooth muscle
soft tissue sarcoma
somatic sensory cortex
species
sperm
sperm abnormality
sperm count
spleen
sprayed area
squamous cell carcinoma
                                                        SRBC
SRBC, sheep red blood cell
steady state
stomach
storage, stored
strain
subacute
                                                   subchronic
subcutaneous, sc
substantia negra
supenor vena cava
superoxide anion
                                                   superoxide dismutase
suprachiasmatic nucleus
susceptible, susceptibility
synapse
synaptic
system
systole
Tcell
T3
T4
T-cell
TD, toxicodynamics
teeth
TEF, toxic equivalency factor
TEQ, toxic equivalent
teratogen
teratogen, teratogenic(ity)
teratogenic
teratogenicity
testes
testes, testicular, testic
testicular
                                                   testosterone
TG
TG, triglyceride
TH
TH, thyroid hormone

-------
                                             LITERATURE SEARCH TERMS (continued)
threshold
thymi
thymic atrophy
thymocyte
thymus
thymus involution
thymus, thymic, thym
thyroid
thyroid function
thyroid hormone
thyroid stimulating hormone
thyroid, thyroid function
thyroxine
thyroxine, T4; T3, triiodothyronine
time
time, time-dependent
time, time-weighted
tissue
tissue, target tissue
TK, toxicokinetics
tooth
toxic, toxicity, toxico, toxicological
trachea
transcutaneous
transdermal
transduction
transformation
transpire(d) air
transpulmonary
tricuspid
tricuspid valve
triglyceride
triiodothyronine
TSH
TSH, thyroid stimulating hormone
tubular
tubule
tumor
tumor, tumorigenic
tumorigenic
turbinates
uncertainty
unnary, unne
urine, unnary
uterine
uterus
uterus, uterine
variability
vascular
vascular disease
vehicle
                                                  vein
ventricle
ventricular
ventromedial hypothalamus
vision
visual cognition
visual motion
visual, visual acuity
vital capacity
vitamin A
vitamin D
vulnerable
vulnerable plaque
wasting syndrome
WBC
weight
white blood cell
white blood cells
                                                  women
worker

-------
1.1.  REFERENCES
U.S. EPA (U.S. Environmental Protection Agency). (2008). 2,3,7,8-Tetrachlorodibenzo-p-dioxin
       (TCDD) dose-response studies: Preliminary literature search results and request for
       additional studies. (EPA/600/R-08/119). Washington, DC: U.S. Environmental Protection
       Agency, National Center for Environmental Assessment.
       http://cfpub.epa. gov/ncea/cfm/recordisplay. cfm?deid= 199923.
                                          1-10

-------