EPA/63 O/P-02/002 A
May 2002
External Review Draft
A REVIEW OF THE REFERENCE DOSE AND
REFERENCE CONCENTRATION PROCESSES
Prepared for the
Risk Assessment Forum
U.S. Environmental Protection Agency
Washington, DC
Reference Dose/Reference Concentration (RfD/RfC) Technical Panel
Bob Benson (OPRA/Region 8) Edward Ohanian (OST/OW)
Gary Foureman (NCEA/ORD) Jennifer Orme-Zavaleta (NHEERL/ORD)
Lee Hofmann (OERR/OSWER) Deborah Rice (NCEA/ORD)
Carole Kimmel (NCEA/ORD)* Jennifer Seed (OPPT/OPPTS)
GaryKimmel (NCEA/ORD) Hugh Til son (NHEERL/ORD)
Susan Makris (OPP/OPPTS) Vanessa Vu (OSCP/OPPTS)
Deirdre Murphy (OAQPS/OAR)
* Technical Panel Chair
Technical Advisors
Amy Mills, IRIS Director, NCEA/ORD
Bill Wood, RAF Director, NCEA/ORD
Risk Assessment Forum
U.S. Environmental Protection Agency
Washington, DC 20460
-------
DISCLAIMER
This document is a draft for review purposes only. It has not been subjected to peer and
administrative review and does not constitute U.S. Environmental Protection Agency policy.
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
-------
LIST OF ACRONYMS AND ABBREVIATIONS
ACE II Angiotensin converting enzyme II
ADI Acceptable Daily Intake
AEGL Acute exposure guideline level
ARE Acute reference exposure
ATSDR Agency for Toxic Substances and Disease Registry
AUC Area under the curve
BMC Benchmark concentration
BMCL Benchmark concentration lower confidence limit
BMD Benchmark dose
BMDL Benchmark dose lower confidence level
BMR Benchmark response
CatReg Categorical Regression (software)
CFSAN Center for Food Safety and Nutrition
CNS Central nervous system
CSAF Chemical-specific adjustment factor
DAF Dosimetric adjustment factor
DNT Developmental neurotoxicity
ECE-1 Endothelin-converting enzyme-1
ELISA Enzyme-linked immunosorbent assay
FDA Food and Drug Administration
FQPA Food Quality Protection Act
FIFRA Federal Insecticide, Fungicide, and Rodenticide Act
GD Gestational day
GLP Good Laboratory Practices
HA Health Advisory
HEC Human equivalent concentration
HED Human equivalent dose
IPCS International Programme on Chemical Safety
IRIS Integrated Risk Information System
LOAEL Lowest-observed-adverse-effect-level
MF Modifying factor
MOE Margin of exposure
MRL Minimal risk level
NAAQS National Ambient Air Quality Standards
NCEA National Center for Environmental Assessment
NK Natural Killer
NOAEL No-observed-adverse-effect-level
OAR Office of Air and Radiation
OECD Organisation for Economic Cooperation and Development
OPP Office of Pesticide Programs
OPPTS Office of Prevention, Pesticides, and Toxic Substances
in
-------
OSWER Office of Solid Waste and Emergency Response
OW Office of Water
P Parental
PAD Population adjusted dose
PBPK Physiologically-based pharmacokinetic model
PFC Plaque-forming cell
PND Postnatal day
POD Point of departure
PRA Plasma renin activity
RDDR Regional deposited dose ratio
RGDR Regional gas dose ratio
RfC Reference concentration
RfD Reference dose
SAB Science Advisory Board
SPF Specific pathogen free
SRBC Sheep Red Blood Cells
TSCA Toxic Substance Control Act
TWA Time-weighted average
UF Uncertainty factor
IV
-------
LIST OF TABLES
Table 2-1. Duration definitions used for various reference values 2-11
Table 2-2. Uncertainty/safety factors for various reference values 2-12
Table 3-1. Approximate age at equivalent life stages in several species 3-4
Table 3-2. Systems/endpoints evaluated by routine toxicity
guideline testing protocols 3-6
Table 4-1. Factors for evaluation of the weight of evidence regarding
the likelihood of effects in humans 4-12
Table 4-2. Factors for evaluation of evidence regarding identification
and characterization of susceptible subpopulations 4-15
Table 4-3. DAFs based on BW3/4 for various species 4-34
Table 4-4. Derivation of reference values for chemical X - inhalation exposure 4-53
Table 4-5. Derivation of reference values for chemical X - oral exposure 4-55
Table B-l. Summary results of major studies in chemical X B-14
Table B-2. Exposure-response data of chemical X - inhalation B-17
Table B-3. Dose-response data of chemical X - oral exposure B-19
Table B-4. Derivation of reference values for chemical X - inhalation exposure B-22
Table B-5. Derivation of reference values for chemical X - oral exposure B-23
-------
LIST OF FIGURES
Figure 3-1. Exposures and endpoints related to general toxicity evaluations 3-18
Figure 3-2. Exposures and endpoints related to reproductive evaluations 3-19
Figure 3-3. Exposures and endpoints for neurotoxicity evaluations 3-20
Figure 3-4. Exposures and endpoints for immunotoxicity evaluations 3-26
Figure 3-5. Exposures and endpoints related to cardiovascular evaluations 3-27
Figure 3-6. Alternative acute toxicity protocol 3-33
Figure 3-7. Expanded chronic/carcinogenicity study 3-35
Figure 3-8. Unified screening study 3-37
Figure 4-1. Concentration-by-duration plot showing the effect
of the exponent in the C" x T = K on extrapolation across time 4-24
Figure 4-2. Current and proposed generalized procedures
for the derivation of HECs or HEDs from animal exposures 4-29
Figure 4-3. Exposure-response arrays for inhalation exposure to chemical X 4-34
Figure B-l. Exposure-response arrays for inhalation exposure to chemical X B-9
VI
-------
External Review Draft Do Not Cite or Quote
PREFACE
The U.S. Environmental Protection Agency (EPA) Risk Assessment Forum was
established to promote scientific consensus on risk assessment issues and to ensure that this
consensus is incorporated into appropriate risk assessment guidance. To accomplish this, the Risk
Assessment Forum assembles experts throughout EPA in a formal process to study and report on
these issues from an Agencywide perspective. For major risk assessment activities, the Risk
Assessment Forum has established Technical Panels to conduct scientific reviews and analyses.
Members are chosen to assure that necessary technical expertise is available.
The RfD/RfC Technical Panel (hereafter the Technical Panel) was established by EPA's
Risk Assessment Forum in early 1999 in response to a request from the Agency's 10X Task
Force1 to the Science Policy Council and the Risk Assessment Forum. In the process of
developing a strategy for implementation of the Food Quality Protection Act (FQPA) relative to
protecting children's health and application of the 10X safety factor, the 10X Task Force
produced two draft reports (one on toxicology and one on exposure data requirements [EPA,
1999a, b] that were used by the Office of Pesticide Programs (OPP) to develop a draft policy
document for implementation of the FQPA safety factor (EPA, 1999c). The draft 10X toxicology
report (EPA, 1999a) raised a number of issues that relate to the derivation of the oral reference
dose (RfD) and inhalation reference concentration (RfC). Examples of these issues include the
following. (1) Appropriate application of a database uncertainty factor (UF) or modifying factor
for studies that are considered necessary but are absent or judged inadequate that may show
children to be significantly more sensitive or susceptible than adults. Addressing this issue also
implicates aspects of other UFs that relate to children's health, including the factor for inter-
individual variability in humans (e.g., response of the aged versus response of the younger adult or
child), and the inter-species UF (e.g., young animals versus young humans). (2) How to account
for degree of concern for potential toxicity to children in the RfD/RfC process. Degree of
concern, as used in the 10X toxicology report, refers to the characterization of the database as to
the likelihood that the agent under review would have effects in humans, within the context of
dose, route, duration, and timing of exposure. (3) The use of developmental toxicity data as the
JThe 10X Task Force was created by the Administrator, EPA, to explore the adequacy of
current testing approaches for pesticides for protecting children's health, and to recommend
approaches for implementation of the additional 10X safety factor mandated by the 1996 FQPA.
vn
-------
External Review Draft Do Not Cite or Quote
basis for reference values2 of chronic duration (RfDs or RfCs) and the appropriate setting of
acute, short-term, and longer-term reference values, including the application of developmental
toxicity data for these shorter duration reference values. (4) The appropriateness and/or rationale
for adjustment of the no-observed-adverse-effect level (NOAEL) or the benchmark dose (BMD)
from developmental toxicity data with inhalation exposures using a concentration times time (C x
t) adjustment as is done for other study types.
The Technical Panel also was asked to consider the need for additional toxicity test
protocols related to children's health as recommended by the 10X Task Force, when they should
be required, and interpretation of the data for risk assessment purposes. These include (1)
collection of pharmacokinetic data, both in adults and at different developmental stages; (2)
direct dosing of neonates, especially when early exposure is of concern; (3) perinatal
carcinogenesis studies and appropriate triggers for when they should be required; (4)
developmental immunotoxicity testing and appropriate triggers; (5) advanced developmental
neurotoxicity (DNT) testing, in particular, cognitive testing that is more similar to that used in
humans; and (6) exposure assessments that are more compatible with the dose-response
assessment. See Appendix A for more a detailed discussion of the issues raised by the 10X Task
Force.
The Science Policy Council and the Risk Assessment Forum agreed that these issues
should be examined on a broader scale than just for pesticides, with input from various program
offices within the Agency and from the outside scientific/policy community. This charge was
expanded by the Forum to include a more in-depth review of a number of issues related to the
RfD/RfC process, in part because of several other Forum activities that were underway. These
activities included development of the Framework for the Harmonization of Cancer and
Noncancer Risk Assessment, revision of the Benchmark Dose Guidance Document, and revision
of the Cancer Risk Assessment Guidelines. In addition, the RfD/RfC derivation process had not
been evaluated in detail for a number of years, and several scientific issues concerning children's
health, e.g., neurotoxicity and immunotoxicity, have become increasingly important in risk
assessment. These various but related activities have prompted the need to re-examine the
RfD/RfC process and to coordinate these efforts with other related activities. In particular, it was
2The term reference value is used genetically here to refer to values such as the RfD, RfC,
acute reference exposure (ARE), Health Advisory (HA), acute exposure guideline level (AEGL),
minimal risk level (MRL), or other similar values.
Vlll
-------
External Review Draft Do Not Cite or Quote
important that efforts continue to focus on moving toward the goal of harmonization of risk
assessment approaches for all health endpoints.
IX
-------
External Review Draft
Do Not Cite or Quote
TABLE OF CONTENTS
LIST OF ACRONYMS AND ABBREVIATIONS iii
LIST OF TABLES v
LIST OF FIGURES vi
PREFACE vii
EXECUTIVE SUMMARY xiii
CHAPTER 1. INTRODUCTION, PURPOSE, AND SCOPE 1-1
CHAPTER 2. REVIEW OF THE CURRENT USE OF ACUTE, SHORT-TERM,
AND LONGER-TERM REFERENCE VALUES 2-1
A. Review of Current Less-Than-Lifetime Reference Values 2-1
A.I. ARE methodology 2-2
A.2. AEGL Program 2-3
A.3. Office of Pesticide Programs (OPP) procedures for
setting acute and intermediate RfDs 2-5
A.4. Office of Water (OW) Health Advisories (HAs) 2-7
A.5. Agency for Toxic Substances and Disease Registry (ATSDR)
Minimal Risk Levels (MRL) 2-8
B. Summary of Current Methods for Setting Acute, Short-term, and
Longer-term Reference Values 2-9
C. Recommendation 2-12
CHAPTER 3. REVIEW OF TESTING GUIDELINES WITH RESPECT
TO LIFE STAGE ASSESSMENT 3-1
A. Evaluation of Current Guideline Testing Protocols 3-2
A.I. Exposures and endpoints related to general toxicity testing 3-5
A.l.a. Acute and short-term toxicity studies 3-7
A.l.b. Subchronic and chronic toxicity studies 3-9
A.2. Exposures and endpoints related to evaluation of
reproductive toxicity 3-10
A.3. Exposures and endpoints related to evaluation of
neurotoxicity 3-13
-------
External Review Draft Do Not Cite or Quote
A.4. Exposures and endpoints related to evaluation of
immunotoxicity 3-16
A.5. Exposures and endpoints related to evaluation of
cardiovascular toxicity 3-24
B. Conclusions and Recommendations 3-27
B.I. Conclusions 3-28
B.2. Recommendations 3-29
C. Options for Alternative Testing Approaches 3-30
C.I. Alternative acute toxicity testing protocol 3-31
C.2. Alternative chronic toxicity testing protocols 3-31
C.2.a. The expanded chronic/carcinogenicity
study 3-34
C.2.b. The unified screening study 3-36
CHAPTER 4. FRAMEWORK FOR SETTING ACUTE, SHORT-TERM,
LONGER-TERM, AND CHRONIC REFERENCE VALUES 4-1
A. Definitions of Exposure Durations for Use in Setting
Reference Values 4-2
B. Proposed Changes in the Reference Value Definitions 4-3
C. Characterization of the Extent of the Health-Related
Database for Setting Reference Values 4-6
C.I. Review of studies 4-6
C.I.a. Adequacy of studies 4-6
C.2. Issues to be considered in the characterization of the
database for risk assessment 4-9
C.2.a. The weight-of-evidence approach 4-9
C.2.b. Use of human and animal data in risk
assessment 4-10
C.2.c. Characterization of effects in potentially susceptible
subpopulations 4-12
C.3. Characterization of the extent of the database 4-15
D. Derivation of Reference Values 4-19
D.I. Selection of endpoints to use as the POD for
Reference Values 4-20
D.2. Dose adjustment for duration of exposure 4-22
XI
-------
External Review Draft Do Not Cite or Quote
D.2.a. Duration adjustment procedures for inhalation
exposures to continuous-exposure scenarios 4-23
D.2.b. Duration adjustment for inhalation developmental
toxicity studies - a current exception 4-24
D.2.c. Duration adjustment for acute reference values -
discontinuous scenarios of 24 hours or less 4-25
D.3. Derivation of an HEC or an HED 4-27
D.S.a. PBPK models and derivation of HEDs and HECs:
estimating internal dose 4-28
D.S.b. Default procedures and derivation of HECs from
the RfC Methodology: derivation and application
of DAFs 4-30
D.S.c. HECs and children - a special case? 4-32
D.3.d. Derivation of an HED for oral and dermal exposure -
use of BW3/4 as a cross-species DAF 4-35
D.4. Other issues 4-35
D.5. Application of uncertainty/variability factors 4-36
D.S.a. Recommendations for application of UFs 4-37
D.S.b. Interspecies UF 4-39
D.S.c. Intraspecies UF 4-40
D.S.d. LOAEL-to-NOAEL UF 4-42
D.S.e. Database UF 4-42
D.S.f. Subchronic-to-chronic duration UF 4-43
D.S.g. MF 4-44
D.6. Future directions 4-44
D.6.a. CSAFs 4-44
D.6.b. Probabilistic approaches 4-46
D.7. Summary of key points from a case study on chemical X . 4-46
D.7.a. Narrative description of the extent of the database for
chemical X 4-47
D.7.b. Exposure-response array for chemical X 4-48
D.7.c. Derivation of reference values for chemical X .... 4-50
D.7.c.i. Acute exposure 4-50
D.7.c.ii. Short-term exposure 4-50
D.7.c.iii. Longer-term exposure 4-51
D.7.c.iv. Chronic exposure 4-52
CHAPTER 5. RECOMMENDATIONS 5-1
REFERENCES R-l
APPENDIX A: ISSUES RAISED BY THE 10X TASK FORCE A-l
xii
-------
External Review Draft Do Not Cite or Quote
APPENDIX B: CASE STUDY FOR CHEMICAL X B-l
GLOSSARY G-l
Xlll
-------
External Review Draft Do Not Cite or Quote
EXECUTIVE SUMMARY
This document summarizes the review and deliberations of the Technical Panel and its
recommendations for improvements in the process as well as additional efforts that are needed. It
discusses revisions to the framework for the derivation of reference values. The document is a
review, and not guidance, but does makes recommendations that should be considered in the
implementation of changes in the current process and/or development of needed guidance.
The Technical Panel reviewed most of the issues relating to hazard characterization for
developing reference values, and the need for developing reference values for different durations
of exposure, as well as the process of deriving reference values, but it did not go into detail on the
quantitative aspects of the dose-response process, which is being covered in other Forum
activities. The Technical Panel views the RfD/RfC process as one that should be continually
evolving as new information becomes available and new scientific and risk assessment approaches
are developed. This does not mean that current RfDs or RfCs are invalid, but these new scientific
issues should be included in the process of re-evaluation of current reference values.
This report reviews and discusses a number of issues and provides conclusions and
recommendations that are intended to improve the RfD/RfC process. The Technical Panel has
provided specific recommendations for the development of guidance in some cases and more
general conclusions and recommendations in others. In the latter cases, the Technical Panel felt
that development of specific recommendations was beyond the scope of its efforts or that policies
needed to be further developed before specific guidance could be written to implement the
recommendations.
The report is divided into five chapters:
Chapter 1 provides an introduction, background, purpose and scope for the project.
Chapter 2 reviews current approaches to developing acute, short-term, and longer-term
reference values as well as the chronic reference values, the RfD and the RfC. This chapter
incorporates the presentations and discussions on developing less-than-lifetime values from
briefings to the Technical Panel and a colloquium held August 2, 2000. These include discussions
of the proposed Acute Reference Exposure (ARE) methodology for acute inhalation exposures,
the Acute Exposure Guideline Level (AEGL) Program, the Office of Pesticide Programs' (OPP's)
procedures for setting acute and longer-term duration RfDs, the Office of Water's (OW's) Health
Advisories, and the Agency for Toxic Substances and Disease Registry's Minimal Risk Levels
xiv
-------
External Review Draft Do Not Cite or Quote
(MRLs). On the basis of its review of the various approaches to setting acute, short-term, and
longer-term reference values, the Technical Panel concurred with the recommendation of the 10X
Task Force that acute, short-term, and longer-term reference values should be set, where possible,
and that they should be incorporated into the Integrated Risk Information System (IRIS)
database. In addition, the Technical Panel recommended that this process be done in a consistent
manner, using standardized definitions for acute, short-term, longer-term, and chronic durations
that are consistent with current practice. These values can then be used by various program
offices, where applicable. A framework for deriving these additional values is presented in
Chapter 4.
Chapter 3 reviews the current Office of Prevention, Pesticides and Toxic Substances'
(OPPTS') harmonized health effects testing guidelines for the purpose of determining the data
available for setting various duration reference values. The point of this exercise was to
understand which target organs/systems are evaluated in current testing protocols and how
thorough the testing is with respect to life stage assessment, endpoint assessment, route, timing
and duration of exposure, and latency to response. These issues were all considered of
importance in evaluating potentially susceptible subpopulations, including life stages. The testing
guideline protocols were reviewed overall for these issues; in addition, four biological systems
were evaluated in depth, two that are fairly thoroughly evaluated (the reproductive and nervous
systems) and two that are evaluated to a more limited extent (the immune and cardiovascular
systems). In each case, an overview of the tests for the particular system is given, as well as a
more specific discussion of gaps in life stage of assessment, gaps in assessment endpoints, and
gaps in duration and latency assessment.
A primary goal of this review was to provide a basis for recommendations for innovative
alternative testing approaches and the use of such data in risk assessment. The Technical Panel is
not recommending additional testing for every chemical but is suggesting that alternative
strategies and guidance for testing approaches be developed that incorporate information on
pharmacokinetics and mode of action early in the process, thus allowing a more targeted testing
approach. In addition, alternative protocols are discussed that are aimed at more efficient use of
animals and resources in combined studies that would provide more extensive data on life stages,
endpoints and other factors not well characterized in current testing approaches.
Recommendations were also made about research areas that should be encouraged to aid in better
study design and interpretation of data for risk assessment.
xv
-------
External Review Draft Do Not Cite or Quote
The Technical Panel has made a number of recommendations concerning testing
guidelines: for example, to develop guideline study protocols for acute and short-term studies that
provide more comprehensive data for setting reference values and guidance on how and when to
use them; to modify existing guideline study protocols to provide more comprehensive coverage
of life stages for both exposure and outcomes and guidance on how and when to use them; to
encourage research to evaluate latency to effect and reversibility of effect from less-than-lifetime
exposures; to develop guideline study protocols that will provide more systematic information on
pharmacokinetics and guidance on how and when to use them; to encourage research on
mechanisms/modes of action and pharmacodynamics; to develop guideline study protocols to
more fully assess all types of toxicity, particularly immunotoxicity, carcinogen!city, neurotoxicity,
and cardiovascular toxicity at different life stages and guidance for how and when to use them;
and to explore the feasibility of setting dermal reference values for direct toxicity at the portal of
entry, including sensitization.
Finally, an example of an alternative testing protocol for acute exposure and evaluation
that incorporates the types of endpoints and evaluations optimal for setting acute reference values
is discussed. Two sample alternative protocols are presented for chronic exposures and options
are discussed for combining studies and evaluations to include a wider array of life stage and
endpoint assessments.
Chapter 4 discusses a number of modifications to the existing framework for use in the
derivation of reference values, both for the current chronic reference values (RfD and RfC) as
well as for acute, short-term, and longer-term reference values. In addition, a case study that
illustrates many of these concepts is summarized in this chapter and presented in more detail in
Appendix B. The Technical Panel recommended including the acute, short-term, longer-term, and
chronic reference values derived on the basis of the recommendations in this report in IRIS after
appropriate internal, external, and consensus review. Standard exposure durations are proposed,
as are definitions for the various reference values, including revision of the definitions for the
current RfD and RfC. In addition, standardization of the terminology for referring to reference
values that includes a designation for route and duration of exposure is proposed. The Technical
Panel recommends that endpoint-specific reference values should not be developed, including the
reference dose for developmental toxicity, RfDDT (EPA, 1991), but that all endpoints should be
considered in the derivation of various duration reference values that are applicable, and the
xvi
-------
External Review Draft Do Not Cite or Quote
reference values should be derived to be protective of all types of effects for that duration of
exposure.
An expanded approach to the evaluation of studies and characterization of the extent of
the database as a whole is recommended; in particular, several factors are discussed that should be
considered in a weight-of-evidence approach for characterizing hazard for the population as a
whole as well as for potentially susceptible subpopulations. Those considerations for assessing
level of concern raised by the Toxicology Working Group of the 10X Task Force (EPA, 1999a)
have been incorporated into this approach. In the context of this framework, the Technical Panel
recommends a somewhat different approach to characterizing the extent of the database for
reference values. Instead of specifying particular studies, this approach emphasizes the types of
data needed (both in terms of human and animal data) for deriving reference values, and it
recommends the use of a narrative description of the extent of the database rather than a single
confidence ranking of high, medium, or low. To characterize the database, the Technical Panel
has developed a description of a "minimal" database and a "robust" database as a way of
describing the range of data that can be used for deriving a reference value, and the Panel urges
the use of a great deal of scientific judgement in the process of summarizing the extent of the
database, including its strengths and limitations. The narrative approach is intended to emphasize
the types of data available (both human and animal data) as well as the data gaps that could
improve the derivation of reference values. This approach should encourage a wider range of
information to be used in deriving reference values, taking into consideration the issues of
duration, timing and route of exposure, the types and extent of endpoint assessment (i.e.,
structure and function), the life stages evaluated, and the potential for latent effects and/or
reversibility of effects.
Dosimetric adjustment of values for deriving a human equivalent concentration (HEC) for
inhalation exposure is discussed, as well as discussion of the derivation of a human equivalent
dose (HED) for oral or dermal exposure. The Technical Panel recommends that duration
adjustment of continuous exposures be used for inhalation developmental toxicity studies as for
other health endpoints. In addition, further evaluation of current dosimetric adjustments for
deriving HECs should be pursued to confirm or assess the relevance for population subgroups
(particularly for children).
Because of the recommendation for deriving several duration reference values, the
Technical Panel recommends that the data for the point of departure (POD) be evaluated based a
xvn
-------
External Review Draft Do Not Cite or Quote
comparison of all relevant endpoints carried through the derivation of sample reference values,
with selection of the limiting value(s) as the final step rather than based on selection of a single
"critical study" and "critical effect." To aid in this evaluation, the use of an exposure-response
array is recommended as a visual display of all relevant endpoints and durations of exposure in
order to determine the range of numerical values available for relevant endpoints.
The Technical Panel makes a number of recommendations concerning the application of
uncertainty factors (UFs) for reference value derivation. In particular, use of sound scientific
judgment is urged in the application of UFs, which are applied to the value chosen for the POD
derived from the available database (lower confidence limit on the benchmark dose [BMDL], no-
observed-adverse-effect level [NOAEL], or lowest-observed-adverse-effect level [LOAEL],).
Although default factors of 10 are recommended, with 3 used in place of half-power values (i.e.,
10°5) when occurring singly, the exact value of the UF chosen should depend on the quality of the
studies available, the extent of the database, and scientific judgment. The Technical Panel
recommends limiting the total UF applied to a chronic reference value for any particular chemical
to 3,000. If there is uncertainty in more than four areas of extrapolation, it is unlikely that the
database is sufficient to derive a reference value, and would need to be carefully evaluated in the
case of uncertainty in four areas. The Technical Panel supports and expands the recommendation
of the Toxicology Working Group of the 10X Task Force (EPA, 1999a) that reduction of the
intraspecies UF be considered only if data are sufficient to support the conclusion that the data set
on which the POD is based is representative of the exposure/dose-response data for the
susceptible subpopulation, including life stages. Given this, whether and how much the
intraspecies UF may be reduced must be linked to how completely the susceptible subpopulation
has been identified and its susceptibility described (e.g., versus assumed). At the other extreme, a
10-fold factor may sometimes be too small because of factors that can influence large differences
in susceptibility, such as genetic polymorphisms. The Technical Panel urges the development of
data to support the selection of the appropriate size of this factor, but recognizes that often there
are insufficient data to support a factor other than the default.
The Technical Panel urges continued research and evaluation of the similarities and
differences between the general population and susceptible subpopulations, particularly children
and the elderly, in their responses to particular agents. From such evaluations, the protectiveness
of the tenfold default factor should continue to be assessed. Given that there are several UFs that
can be used to deal with data deficiencies as part of the current reference value process, and given
xvm
-------
External Review Draft Do Not Cite or Quote
that these are assumed to overlap to some extent, the Technical Panel agrees with the 10X Task
Force Toxicology Working Group that the current interspecies, intraspecies, and database
deficiency UFs, if appropriately applied using the approaches recommended in this review, will be
adequate in most cases to cover concerns and uncertainties about children's health risks. If there
are residual concerns about toxicity and/or exposure, these can be dealt with in risk
characterization/risk management (e.g., by retention of all or part of the FQPA safety factor for
pesticides). The Panel considers the purpose of the modifying factor (MF) to be sufficiently
subsumed in the general database UF, and recommends discontinuance in the use of the MF. The
approach to using chemical-specific data for pharmacokinetic and pharmacodynamic components
of UFs has been discussed in the RfC methodology (EPA, 1994). The Technical Panel encourages
the Agency to develop its own guidance for chemical-specific adjustment factors (CSAFs) on the
basis of some of the available methodologies (e.g., International Programme on Chemical Safety
[IPCS]). Caution should be used, however, in that there are relatively few data available for many
substances that could serve as an adequate basis to replace defaults with CSAFs.
Several other issues discussed by the Technical Panel were considered more appropriate
for deliberation by other panels/committees, e.g., further consideration of the use of BMD
modeling approaches for deriving reference values; harmonization of the approaches for HEC and
HED derivation for all types of health effects; further evaluation of approaches such as
probabilistic analysis for characterizing variability and uncertainty in toxicity reference values;
further evaluation of appropriate adjustment of doses for duration of exposure for acute toxicity
data; and further evaluation of duration adjustment for short-term and longer-term reference
values analogous to the subchronic to chronic duration UF for chronic reference values.
Chapter 5 summarizes the recommendations of the Technical Panel.
xix
-------
External Review Draft Do Not Cite or Quote
CHAPTER 1
INTRODUCTION, PURPOSE, AND SCOPE
The RfD/RfC Technical Panel (hereafter the Technical Panel) was established by EPA's
Risk Assessment Forum in early 1999 to review the current reference dose and reference
concentration (RfD/RfC) processes, in particular with respect to how well children and other
potentially susceptible subpopulations are protected, to consider new scientific issues that have
become more important and of greater concern in risk assessment, and to raise issues that should
be explored or developed further for application in the RfD/RfC process. This document
summarizes the review and deliberations of the Technical Panel and its recommendations for
improvements in the process as well as additional efforts that are needed. It discusses revisions to
the framework for the derivation of RfDs and RfCs. The document is a review, not guidance, but
it does make recommendations that should be considered in the implementation of changes in the
current process and/or development of needed guidance.
Many of the recommendations made in this report are consistent with the agency's
commitment to harmonization of health risk assessment procedures, including the harmonization
of approaches for noncancer and cancer endpoints and making efficient use of animal testing to
achieve this goal. As noted several places in the document, all such topics have not been
discussed and resolved by the agency. For instance, the differences in scaling factors used for
cancer and noncancer derivations from oral exposure data is raised as an issue that has not been
resolved. Thus, there will likely be the need for revised or further guidance in the future on this
and other items. Although mixtures or multiple chemical exposures are not specifically discussed
in this review, most of the recommendations are applicable to the approach to risk assessment of
mixtures. The agency's mixtures risk assessment guidelines should be consulted for issues
specific to the evaluation of mixtures (EPA, 1986, 2000c). In addition, the agency has recently
issued a draft Framework for Cumulative Risk Assessment (2002c) that deals with the issue of
multiple stressors and their overall impacts on exposure-effect relationships. The risk assessment
approaches discussed within this framework are likely to be the subject of further guidance as
well.
The Technical Panel attempted to review most of the issues relating to hazard
characterization for developing reference values, the need for developing reference values for
1-1
-------
External Review Draft Do Not Cite or Quote
different durations of exposure, and the process of deriving reference values. The Technical Panel
did not go into detail on the quantitative aspects of the dose-response process, as this is being
covered in other Forum activities (e.g, the benchmark dose [BMD] guidance document and the
quantitative dose-response aspects of the cancer guidelines revision process). The Technical
Panel approached its review from the point of view that the RfD/RfC process has been and should
be a continually evolving process. Thus, as new information becomes available and new scientific
and risk assessment approaches are developed, they are incorporated into new RfDs and RfCs as
these values are developed or as current RfDs and RfCs are re-evaluated. This process of
incorporating new science does not invalidate current RfDs or RfCs, because consideration of
these new scientific issues is included in the re-evaluation of current values; higher or lower
values, or in some cases, no change in the current value may result.
This report provides conclusions and recommendations that are intended to improve the
RfD/RfC process. The audience for this review is primarily the Integrated Risk Information
System (IRIS) program, IRIS chemical managers, and other scientists within the Agency who are
involved in developing the RfDs and RfCs, as well as IRIS users and the program offices within
EPA who develop RfDs and RfCs or similar values (see Chapter 2), particularly resource
managers who may be impacted by the potential for additional workload due to several of the
recommendations. The Technical Panel has provided specific recommendations for guidance in
some cases and more general conclusions and recommendations in others. In the latter cases, the
Technical Panel felt that development of specific recommendations was beyond the scope of its
efforts or that policies needed to be further developed before specific guidance could be written to
implement the recommendations.
The methodology developed in the RfD document is considered generally applicable to
both cancer and noncancer endpoints where dose response relationships are thought to be either
nonlinear or consistent with a threshold. Although the emphasis in this document is on the
calculation of RfDs and RfCs, the same processes and considerations are applicable to the Margin
of Exposure, as discussed in the draft cancer risk assessment guidelines (EPA, 1999d).
A number of issues are discussed and recommendations made in this report by the
Technical Panel concerning a revised framework for the RfD/RfC process, with particular
emphasis on the extent to which children and other potentially susceptible subpopulations are
considered. The next three chapters summarize issues that the Technical Panel has discussed, and
several recommendations are made concerning those issues. Chapter 2 reviews current
1-2
-------
External Review Draft Do Not Cite or Quote
approaches to developing acute, short-term, and longer-term reference values as well as the
chronic reference values, the RfD and RfC. Chapter 3 reviews the current testing guidelines with
respect to life stage assessment and discusses the gaps in life stage assessment, endpoint
assessment, and assessment of duration and latency. Alternative testing protocols and strategies
as options for combining studies and evaluations are discussed. Chapter 4 provides constructive
commentary on the current framework used in the derivation of reference values and on the need
and possibilities for calculating reference values for different durations and routes of exposure. In
addition, an expanded approach to the evaluation of studies and characterization of the extent of
the database as a whole is presented and discussed, including dosimetric adjustment, selection of
the data to be used for the point of departure (POD) in deriving reference values, and the
application of uncertainty factors (UFs). The final chapter (chapter 5) summarizes all of the
recommendations of the Technical Panel. A case study illustrating several of the recommended
changes is also included as Appendix B.
1-3
-------
External Review Draft Do Not Cite or Quote
CHAPTER 2
REVIEW OF THE CURRENT USE OF ACUTE, SHORT-TERM,
AND LONGER-TERM REFERENCE VALUES
The Technical Panel considered the recommendation of the 10X Task Force that acute,
short-term, and longer-term reference values, as well as chronic reference values, should be set
for environmental agents (see Appendix A). It is likely that the endpoints critical for setting
acute, short-term, and longer-term reference values may differ from those for setting chronic
RfDs and RfCs, although studies that use acute and short-term exposure conditions from which
the appropriate data for many types of effects could be derived are not often available. Often data
on acute and short-term health effects must be derived from observations after the first exposure
in a repeated-exposure testing protocol. Several acute and short-term values currently are set for
various chemical types and media. For example, acute and chronic oral RfDs are set for
pesticides, with some intermediate values set for occupational and residential pesticide exposures.
Health Advisories of several durations have been developed for drinking water. In addition, the
Office of Solid Waste and Emergency Response (OSWER), the Office of Prevention, Pesticides,
and Toxic Substances (OPPTS), and other program offices and regional offices use values derived
through the interagency Acute Exposure Guidelines (AEGL) process for emergency response
planning. The National Center for Environmental Assessment (NCEA) is currently developing the
Acute Reference Exposure (ARE) methodology for acute inhalation exposures. These
developments are reviewed in more detail below.
A. Review of Current Less-Than-Lifetime Reference Values
The Technical Panel was briefed by representatives of several Agency offices on the
methods currently used to set various less-than-lifetime reference values. Subsequently, on
August 2, 2000, a Risk Assessment Forum colloquium was held on this topic (The CDM Group,
Inc., 2000). Each of the methods was presented and discussed. In addition, a recommendation
by the Technical Panel to begin deriving acute, short-term, and longer-term reference values, as
well as chronic values, and to standardize the definitions for each duration was presented and
discussed. Each method presented is summarized below.
2-1
-------
External Review Draft Do Not Cite or Quote
A.I. ARE methodology
The ARE methodology is being developed at the request of the Office of Air and
Radiation (OAR). It is intended for development of reference values for acute inhalation
exposures of various durations of 24 hours or less. The criteria air pollutants1 are not included,
because they are assessed within the National Ambient Air Quality Standards (NAAQS) setting
process. The ARE is defined as an inhalation exposure for 24 hours or less that is not likely to
cause noncancer adverse effects. The ARE can be applied to intermittent exposures or to
continuous exposures for 24 hours or less. AREs are being developed in order to address the
acute risk aspects of risk-related provisions of the hazardous air pollutant sections of the 1990
Clean Air Act Amendments. The ARE methodology is described in a 1998 EPA external review
draft document (EPA, 1998b). The method builds on the procedures of the RfC methodology.
The ARE method includes three approaches in order to accommodate the varying types of
data available for acute exposure. The first two approaches, no-observed-adverse-effect-level
(NOAEL) and benchmark concentration (BMC) are familiar, but the categorical regression
(CatReg) approach is newer. The NOAEL approach is useful for chemicals that have limited
available data and for which no or limited dose-response relationships have been established. The
BMC approach is suitable for analysis of studies that establish dose-response relationships. The
CatReg approach requires multiple studies that report not only dose and response, but also
duration; it is most applicable for data-rich chemicals. A feature of the CatReg approach is that
effects data are grouped into severity categories (e.g., mild or severe to lethal) to which
sophisticated regression procedures are then applied.
For derivation of ARE values of different durations (e.g., 15 min or 8 hr), adjustments are
made differently for the NOAEL and the BMC approach than for the CatReg approach. For any
approach used, the preferred adjustment procedure is to use a pharmacokinetic model, if available.
When the NOAEL and BMD approaches are used, the default procedure is to use the multiple of
C x t (Cn x t = k; ten Berge et al., 1986) to extrapolate from short to long duration, and to use the
same concentration as obtained for long duration to extrapolate from long to short duration.
When more than one duration is available, interpolation is performed. When the CatReg
approach is used, the procedure involves reading the values directly from the concentration
Criteria air pollutants are those air pollutants for which NAAQS have been established
under the Clean Air Act; at present, the six criteria air pollutants are paniculate matter, ozone,
carbon monoxide, nitrogen oxides, sulfur dioxide, and lead.
2-2
-------
External Review Draft Do Not Cite or Quote
duration curve that is generated by the CatReg software. These approaches are explained more
fully and illustrated in Chapter 4.
A minimal dataset has not been defined for the ARE. Also, extrapolation from the oral to
the inhalation route of exposure is not addressed in the ARE approach. UFs in the ARE approach
include a lowest-observed-adverse-effect-level- (LOAEL-) to-NOAEL UF of 10 and a default
value of 10 each for interspecies and intraspecies extrapolation. No factor is assigned for
database inadequacies and study quality.
In 1998, the Science Advisory Board (SAB) reviewed the ARE methodology document,
and made a number of comments, that addressed, among other things, issues about the NOAEL
and BMC approaches, the need for addressing protection of children, the dosimetry adjustment
and duration extrapolation, and the CatReg approach. With regard to the last point, the SAB
discussed the fact that the CatReg model, as currently set up, forces parallelism of the
concentration-duration curves for the various severity categories. In addition, there were
concerns about judging severity categories across various target organs and species, and there
was discussion about the reliability of the confidence limits around the maximum likelihood
estimate, and about the appropriateness of the approach used to accommodate group versus
individual data. In addition to revising the ARE methodology and CatReg software documents,
NCEA-Research Triangle Park will develop a framework for adding AREs to the IRIS database.
This methodology has since undergone an Agency review by the Risk Assessment Forum in
March of 2001. The principal comments from this review were to reevaluate whether CatReg
should remain as an approach in the ARE methodology and to further evaluate the procedures for
cross-species dosimetry adjustment. Revision of the ARE methodology is currently underway.
A.2. AEGL Program
The primary purpose of the AEGL program is to develop guideline levels for once in a
lifetime short-term exposures to airborne concentrations of acutely toxic chemicals (NRC, 2000).
AEGLs are needed for a wide variety of emergency planning, response, and prevention
applications. AEGLs represent threshold exposure limits for the general public and are applicable
to emergency exposure periods ranging from 10 minutes to 8 hours. Specific values are set for 10
minutes, 30 minutes, 1 hour, 4 hours, and 8 hours. It is believed that the recommended exposure
levels are applicable to the general population, including infants and children and other
individuals, e.g., asthmatics, who may be sensitive or susceptible. It is recognized that certain
2-3
-------
External Review Draft Do Not Cite or Quote
individuals who may be subject to unique or idiosyncratic responses could experience the effects
described at concentrations below the corresponding AEGL level.
AEGL-1, AEGL-2, and AEGL-3 are distinguished by varying degrees of severity of toxic
effects. With increasing airborne concentrations above each AEGL level, there is a progressive
increase in the likelihood of occurrence and the severity of effects described for each
corresponding AEGL level.
AEGL-1 is the airborne concentration of a substance above which it is predicted that the
general population, including susceptible individuals, could experience notable discomfort,
irritation, or certain asymptomatic, non-sensory effects. However, the effects would not be
disabling and would be transient and reversible upon cessation of exposure.
AEGL-2 is the airborne concentration of a substance above which it is predicted that the
general population, including susceptible individuals, could experience irreversible or other
serious, long-lasting adverse health effects, or an impaired ability to escape.
AEGL-3 is the airborne concentration of a substance above which it is predicted that the
general population, including susceptible individuals, could experience life-threatening health
effects or death.
Airborne concentrations below AEGL-1 represent exposure levels that could produce mild
and progressively increasing odor, taste, and sensory irritation, or certain asymptomatic, non-
sensory effects.
UFs are used for extrapolations. If there are no appropriate human data, an interspecies
UF of 1, 3, or 10 is used. The factors considered when deciding on a specific value include (1)
the species tested (type, appropriateness, and range); (2) the toxicological endpoint observed and
the likely mechanism of action; (3) the range of response in the species tested; (4) the variability
of response among the species tested; and (5) pharmacokinetic differences among the species
tested. An intraspecies UF of 1, 3, or 10 is also used. The factors considered when assigning a
specific value include (1) the toxicological endpoint observed and the likely mechanism of action;
(2) the range of response among humans and sub-populations; and (3) pharmacokinetic
differences among people. Individual factors of 3 are often used to ensure that the final values are
not overly conservative.
Adjustment for duration is conducted using the equation Cn x t = k. If data are available
for the endpoint of concern, the value of n is derived from regression analysis. If data are not
available for the endpoint of concern, then the value of n is usually derived from lethality data by
regression analysis and used for the other endpoints. If the study duration is greater than 1 hour,
2-4
-------
External Review Draft Do Not Cite or Quote
then the 10-minute value is usually assigned equal to the 30 minute value. If no data are available
to derive a value of n, then a value of 3 is used to extrapolate to shorter durations, and a value of
1 is used to extrapolate to longer durations. As mentioned above, this procedure is further
explained and illustrated in Chapter 4.
A.3. Office of Pesticide Programs (OPP) procedures for setting acute and
intermediate RfDs
The OPP developed methodologies for acute dietary as well as occupational and
residential risk assessments during the process of re-registration, following the 1988 revision to
the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA). In 1998, a guidance document,
Toxicology Endpoint Selection Process, was presented to the FIFRA Scientific Advisory Panel
for review and comment (OPP, 1998). This document, which provided the basis for procedures
that are still in place, describes toxicology endpoint selection for less-than-lifetime dietary and
occupational/residential risk assessments for pesticides. It includes guidance on the evaluation of
toxicity studies that are relevant for use, the selection of appropriate endpoints for hazard
identification, the process of hazard identification, the influence of dermal absorption in hazard
identification, the criteria for the use of the NOAEL and LOAEL, and the use of Margins of
Exposure (MOEs) in risk assessments. Since this guidance was first issued, some changes have
evolved, such as the replacement of the acute MOE with the acute RfD and the addition of
standard consideration of short- and intermediate-term incidental nondietary ingestion exposures
for toddlers.
The OPP Toxicology Endpoint Selection Process document describes the types of studies
that are most likely to provide appropriate endpoints for the various exposure durations and risk
assessments that will be conducted for each pesticide. OPP can rely on the availability of a wide
variety of standard guideline toxicity studies from which to select endpoints, because they are
required by regulation for any pesticide registration (40 CFR Part 158). Additionally, OPP
considers other sources of toxicology data, such as studies published in the open literature, as
appropriate.
For the establishment of the acute RfD, OPP uses a weight of evidence approach in
evaluating all the available data. Three guideline studies have been found to be particularly useful
by OPP: the acute neurotoxicity study, the prenatal developmental toxicity study, and the
developmental neurotoxicity (DNT) study.
2-5
-------
External Review Draft Do Not Cite or Quote
Acute effects from subchronic and chronic dietary studies are also used in the
establishment of the acute dietary RfD. Careful scrutiny of toxicological data from early in the
first week of treatment can sometimes identify effects that can be described as acute. However,
for a number of reasons, this option has not often been used. These reasons include the absence
of detailed toxicological observations other than morbidity and mortality checks in subchronic and
chronic studies before the end of the first week of treatment (i.e., after 7 days of treatment), the
nature of the dietary exposure (i.e., each daily exposure results from an extended period of nightly
feeding rather than from a discrete acute dose), and the possibility that apparent adverse effects
during the first week of treatment may be related to palatability issues as the animals adjust to
treated feed.
OPP does not calculate short- or intermediate-term references doses. However, risk
assessments are conducted for incidental nondietary ingestion exposures to toddlers, a very
specific population subgroup, that result from the use of a pesticide in and around the home or
other non-occupational sources such as schools, parks, and golf courses. The post-application
risk assessment considers or accounts primarily for incidental ingestion of (1) the dry pesticide
materials (granules or pellets) used to treat outdoor residential areas, (2) pesticide residues in soil
that are ingested by toddlers who play in treated areas (e.g., yards, gardens, playgrounds) as a
result of normal mouthing activities, and (3) pesticide residues that are transferred to the skin of
toddlers playing in treated areas and are subsequently ingested as a result of hand-to-mouth
transfer. These risk assessments consider short-term (1 day to 1 month) and intermediate-term
(1-6 months) exposure durations. Risks are expressed as an MOE. The MOE approach is used
because these exposures are considered to be non-dietary in source and are based on high-end
values (or on assumptions when adequate site- or chemical-specific field data are unavailable).
OPP also conducts short-term, intermediate, and long-term (longer than 6 months) dermal
and inhalation risk assessments for occupational and residential exposures. The MOE approach is
also used to calculate the risk for these non-dietary exposure scenarios. A difficulty that OPP
often faces when conducting these risk assessments is that dermal absorption and inhalation
toxicity data are often not available for food-use pesticides; in that case, appropriate assumptions
are applied and the available oral toxicity data are converted for use in dermal and inhalation risk
assessment.
The Toxicity Endpoint Selection Process document does not address the use of UFs in
acute dietary risk assessment. In practice, however, the same 10-fold inter- and intraspecies UFs
are used in calculating the acute dietary RfD as are used for the chronic RfD. Other standard UFs
2-6
-------
External Review Draft Do Not Cite or Quote
may be used when appropriate (e.g., the LOAEL-to-NOAEL three-fold factor). Others are not
appropriate, e.g., the three-fold subchronic-to-chronic factor for an acute risk assessment.
However, no standard set of "core" studies has been defined for acute dietary risk assessment;
therefore, a database UF is not used. If appropriate endpoints and doses cannot be selected for
acute dietary risk assessment from the studies in the database, then an acute RfD is not calculated.
A.4. Office of Water (OW) Health Advisories (HAs)
The OW HA program was initiated in 1978 to provide guidance on unregulated
contaminants found in drinking water. Since then, HAs have also been developed for regulated
contaminants. HAs are developed for contaminants that are known or are likely to occur in
drinking water and that may cause adverse, noncarcinogenic health effects (Orme and Ohanian,
1991). The approach for developing HAs is based on recommendations from the National
Academy of Sciences (NAS, 1977). HAs are developed for specific exposure durations (1 day,
10 days, longer-term, and lifetime) that reflect different emergency contamination situations. HAs
are not legally enforceable, but they do serve as technical guidance to assist in emergency spills or
contamination situations or for determining unreasonable risks to health under sections 1415 and
1416 of the Safe Drinking Water Act. They also are issued at the request of State or local
governments or to fill a need for criteria, guidelines, or standards. HAs undergo scientific peer
review and can function as a preliminary risk assessment, if necessary.
The following assumptions are used in setting the various HAs. The 1-day HA represents
a concentration of the contaminant in drinking water that is considered protective of adverse
noncancer health effects in a 10 kg child. The 10 kg child serves as the protected individual for
the less than lifetime HAs because a child of this size is likely to receive a greater dose on a mg/kg
basis. This 1-day HA can serve as a guideline for each day, up to 5 consecutive days of exposure.
The 1-day HA is usually derived from experimental studies of 7 days duration or less.
The 10-day HA is considered protective of these effects in a 10 kg child for each day, up
to 14 days of continuous exposure and may be based on experimental studies of 30-day duration
or less.
The Longer-term HA, based on subchronic exposure studies covering 10% of an animal's
lifetime, is considered protective of an exposure period in humans of up to 7 years (i.e., 10% of an
individual's lifetime). The Longer-term HA is developed to protect both a 10 kg child and a 70 kg
adult.
The Lifetime HA is considered protective of lifetime exposures and is usually based on
chronic or subchronic or other more relevant experimental data. The Lifetime HA is based on the
2-7
-------
External Review Draft Do Not Cite or Quote
chronic oral RfD, adjusted for a 70 kg adult drinking 2 L water per day; the value is apportioned
by a relative source contribution, e.g., 20%.
The HA levels are generally based on available, well conducted studies that involve
humans or animals. Data from drinking water studies are preferred; however, data from dietary
or gavage studies can also be used. In the absence of oral data, studies by other routes of
exposure, such as inhalation or injection, are considered. Following identification of an
appropriate study to develop a HA, the NOAEL or LOAEL is adjusted for water consumption by
the protected individual. For a child, the assumed water consumption level is 1 L/day; for an
adult, 2 L/day is used.
When data are absent for setting a 1-day or a 10-day HA, OW uses scientific judgment on
how to handle any given situation based on the overall weight of evidence. In the absence of
short term toxicity studies, a subchronic or chronic study may be used to develop a less-than-
lifetime HA. Given the pressure under which HAs need to be calculated, many assessments are
based on whatever toxicological data are available and on scientific common sense. Although this
may be an overly conservative approach, OW considers the error to be protective of public health.
OW applies the same factors for minimum data as outlined in the Agency's RfD
methodology. For example, in emergency situations, missing data are accounted for by applying
another factor of 3 or 10. Or, for instance, where inhalation data might be applied to estimate a
HA based on water consumption, a factor may be applied to account for differences in absorption.
Such judgments based on toxicokinetic and toxicodynamic considerations are reached through
intensive consultation.
Calculation of HAs is straightforward and familiar, and in most cases, the NOAEL/UF
approach is used. For each of the less-than-lifetime HA values, it is assumed that all of an
individual's exposure to a contaminant comes from a drinking water source. The calculation of
the Lifetime HA differs from that of the less-than-lifetime values in that a relative source
contribution factor is included. This factor adjusts the exposure to reflect the portion that is likely
to be contributed from drinking water. Unless actual exposure data are available, a default factor
of 20% is used to reflect the assumed contribution to exposure from drinking water. Also, in
cases where there is limited evidence suggesting a carcinogenic potential of a contaminant, an
additional "policy" factor of 10 is applied in calculating the Lifetime HA.
A.5. Agency for Toxic Substances and Disease Registry (ATSDR) Minimal Risk
Levels (MRLs)
2-8
-------
External Review Draft Do Not Cite or Quote
The ATSDR is tasked with establishing MRLs which are defined as:
" ... an estimate of daily human exposure to a hazardous substance that is likely
to be without appreciable risk of adverse noncancer health effects over a
specified route and duration of exposure".
MRLs are considered by ATSDR to be substance-specific estimates intended to be screening
levels in the identification of contaminants and potential health effects that may be of concern;
they do not define clean-up or action levels. The derivation procedures for MRLs have many
similarities and parallels to the derivation of RfDs and RfCs; MRLs are based on careful scientific
consideration of noncancer health effects only, not on consideration of cancer effects. A list of
various procedural specifics employed in derivation of MRLs, including specific effects and the
level of severity, is codified in a Federal Register notice (ATSDR, 1996). The definition of an
MRL differs expressly from that of EPA's RfD or RfC in that both route and duration are
included. The current routes of concern for MRL derivation are oral and inhalation, not dermal.
The EPA procedures and methodologies discussed above address the issue of duration through a
variety of extrapolation procedures. For MRLs, however, duration is addressed by providing for
the designation of MRLs in three different duration categories: acute = <14 days, intermediate =
15-364 days, and chronic = >365 days. These duration categories are absolute and apply to all
species regardless of relative life span. Thus, it is possible for a contaminant to have a total of 6
different MRL values, two routes by three different durations.
The use of UFs is a parallel practice in RfD/C and MRL derivation. The UFs used by
ATSDR are as follows: intraspecies 1, 3, 10; interspecies 1, 3, 10; and LOAEL/NOAEL 3, 10.
The modifying factor (MF) can include database considerations, i.e., deficiencies in the data or
overestimates from bioaccumulative chemicals.
B. Summary of Current Methods for Setting Acute, Short-term, and Longer-term
Reference Values
In summary, several methods exist for setting acute, short-term, and longer-term reference
values that are used by various EPA programs. The definitions for each of the durations used for
the methods reviewed are included in Table 2-1. Because there are some differences in these
definitions, standardized definitions were discussed at the Risk Assessment Forum Colloquium
(The CDM Group, 2000), and these are shown in Table 2-1. Definitions for durations are further
discussed in Chapter 4.
2-9
-------
External Review Draft Do Not Cite or Quote
A comparison of the UFs applied for various reference values is shown in Table 2-2.
Although there is some variation in the UFs applied, those for animal-to-human extrapolation
(UA), for within-human variability (UH), and for LOAEL-to-NOAEL (UL) are fairly consistent.
Less consistent is the way in which database deficiencies (UD) are taken into consideration,
particularly for pesticides where the Food Quality Protection Act (FQPA) safety factor is used to
account for deficiencies in the database related to children's health risks.
2-10
-------
External Review Draft
Do Not Cite or Quote
Table 2-1. Duration Definitions Used for Various Reference Values
Reference Value
Definition
Acute:
ARE
AEGL
OPP Acute RfD
OW 1-day HA
ATSDR Acute MRL
Standardized definition3
Inhalation single continuous exposure values for durations <24
hrs (to be protective of intermittent exposures)
10 & 30 min, 1, 4 & 8 hours
Maximum 1-day dietary exposure
1 day (5 day successive daily doses)
< 14 days
24 hours or less
Short- Term:
ARE
AEGL
OPP Short-term RfD
OW 10-day HA
ATSDR MRL
Standardized definition3
NA
NA
1 day - 1 month
10 days (7-14 successive daily doses)
NA
More than 24 hours up to 30 days
Longer-term:
ARE
AEGL
OPP Intermediate RfD
OW Longer-term HA
ATSDR Intermediate MRL
Standardized definition3
NA
NA
1 month - 6 months
Longer-term - approximately 10% of lifespan in humans (90 days
to 1 year in test species)
Intermediate - 15-364 days
More than 30 days up to approximately 10% of the life
humans (>30 days-90 days in typically-used laboratory
span in
species)
3See Chapter 4 for further discussion of these definitions.
2-11
-------
External Review Draft
Do Not Cite or Quote
Table 2-2. Uncertainty/Safety Factors for Various Reference Values
Reference Value/lIF3
ARE
AEGL
OPP acute and
intermediate RfDs
OWHAs
ATSDR MRLs
UA
1,3, 10
1,3,10
10
1,3,10
1,3, 10
UH
1,3, 10
1,3,10
10
1,3,10
1,3, 10
UL
1,3, 10
3b
3,10
1,3,10
1,3, 10
UD
ND
NDC
NDd
case-
specific
NDC
FQPA
NA
NA
10+
NA
NA
3UA = animal-to-human UF; UH = within-human variability UF; UL = LOAEL-to-NOAEL UF; UD =
database deficiency UF; FQPA = additional safety factor required under FQPA.; ND = not done; NA = not
applicable.
bEndpoint = lethality, not really a LOAEL-to-NOAEL adjustment in this case
°Database deficiencies considered, and a factor may be included for intermediate RfDs if, for example, there
is no reproduction and fertility study.
Overlaps with the FQPA safety factor (see EPA, 1999c, 2002)
Duration extrapolation for each of these values was also reviewed. Some type of duration
adjustment of the NOAEL or BMD is done for the ARE and the AEGL methods, and there
appears to be consistency in the use of Cn x t for extrapolating from shorter to longer exposures,
but in using the same value (i.e., no duration adjustment) when extrapolating from longer to
shorter exposures. Duration extrapolation is not done for the OPP RfDs, OW HAs or the ATSDR
MRLs.
C. Recommendation
Based on its review of the various approaches to setting acute, short-term, and longer-
term reference values, the Technical Panel concurred with the recommendation of the 10X Task
Force that acute, short-term, and longer-term reference values should be set, where possible, and
that they be incorporated into the IRIS database. In addition, the Technical Panel recommended
that these values be set in a consistent manner, using standardized definitions for acute, short-
term, longer-term, and chronic durations that are consistent with current practice. They can then
2-12
-------
External Review Draft Do Not Cite or Quote
be used by various program offices, where applicable. A scheme for deriving these additional
values is presented in Chapter 4.
2-13
-------
External Review Draft Do Not Cite or Quote
CHAPTER 3
REVIEW OF TESTING GUIDELINES WITH RESPECT TO
LIFE STAGE ASSESSMENT
As a first step in determining the data necessary for setting various duration reference
values for protecting potentially susceptible subpopulations, the Technical Panel reviewed the
current OPPTS testing guidelines1 to determine what information is gathered in these studies.
The point of this exercise was to understand which target organ systems are evaluated in current
testing protocols and how thorough the testing protocols are with respect to life stage assessment,
endpoint assessment, route, timing and duration of exposure, reversibility, and latency to
response. These issues were all considered of importance in evaluating potentially susceptible
subpopulations, particularly children. The intent of this review is not to suggest that such data
should be collected for each and every chemical, but to indicate where testing protocols are and
are not available in the armamentarium of testing protocols from which the appropriate studies
for a given chemical might be selected. Another primary goal of this review was to provide a
basis for the development of innovative alternative testing approaches and the use of such data in
risk assessment. Alternative strategies and guidance for testing approaches are needed that
incorporate information on pharmacokinetics and mode of action early in the process, thus
allowing a more targeted testing approach. The recommendations for alternative testing
approaches are aimed at more efficient use of animals and resources in combined studies that
would provide more extensive data on life stages, endpoints and other factors not well
characterized in current testing approaches.
The review was conducted with respect to the life stages assessed both in terms of when
exposures occurred and when outcomes were evaluated. In addition, the route, timing and
duration of exposure were considered, as well as whether reversibility and latency to response
were covered at different life stages in current testing guideline protocols. The following
sections give an overview of the current testing protocols evaluated in this way and, for certain
organ/functional systems, provide a more in depth analysis as to whether and how current
JThe guidelines are available on the OPPTS web page (http://www.epa.gov/docs/
OPPTS_Harmonized/870_Health_Effects_Test_Guidelines/Series/).
3-1
-------
External Review Draft Do Not Cite or Quote
protocols address exposures and outcomes with respect to life stage, timing and duration of
exposure, reversibility and latency to response. The organs/functional systems that are examined
in greater detail included the reproductive and the nervous systems, selected to demonstrate
systems that are thought to be rather well-evaluated. The immune and the cardiovascular systems
were selected for review because the evaluation of these systems is limited. It should be noted
that testing guidelines were not originally designed with a focus on evaluations of different life
stages or different durations of exposure. Therefore, a number of gaps in life stage assessment,
endpoint assessment, timing and duration of exposure, reversibility, and latency to response were
noted for each organ system that is reviewed in depth. Recommendations for new testing
protocols, and for changes and alternatives to current testing approaches are discussed in sections
B and C of this chapter.
A. Evaluation of Current Guideline Testing Protocols
The following tables and figures summarize the exposures and endpoints covered in
current testing guidelines and what is covered for each organ system/endpoint measured, as well
as the relative depth of evaluation for each system/endpoint. In addition, the life stages covered
by exposures and outcomes are illustrated. The discussions that correspond to the figures give an
overview of the tests that are currently available and the gaps in assessment of life stages,
endpoints, timing and duration of exposure, and latency to response. Together, these analyses
provide a clear picture of the testing guidelines currently available, the systems/endpoints
measured, the life stages during which exposures and outcomes are measured, the timing and
duration of exposures included, and the degree of detail covered for both structural and
functional outcomes.
In order to make comparisons among laboratory animal species and humans in terms of
life stages covered, the approximate ages that correspond to specific events or life stages (e.g.,
birth, weaning, puberty, etc.) in different species are shown in Table 3-1, and these events/life
stages are indicated in the figures. In a few cases, no data could be found on appropriate ages
corresponding to particular life stages. In particular, the ages for mature adults and older adults
often were not available, and there is some controversy about what constitutes old age in today's
population. A background paper on aging discusses this issue to some extent (Versar Inc.,
200la). In animal studies, the use of dietary restriction has been shown to affect aging and
3-2
-------
External Review Draft Do Not Cite or Quote
lifespan to a significant extent, so the issue of what constitutes an older animal is also somewhat
controversial.
3-3
-------
External Review Draft
Do Not Cite or Quote
Table 3-1. Approximate Age at Equivalent Life Stages in Several Species
Rat
Embryonic
Fetal"
Neonateb
Weaning0
Young
Puberty
Sexual
Maturity
Mature Adult
Old Adult
GD 0-16
GD 16-22
(22-23 days)
PND 0-14
PND21
PND 22-35
PND 35-60
2.5-3 mos
5-18 mos
18 mos - 2
yrs+
Mouse
Embryonic
Fetal
Neonate
Weaning
Young
Puberty
Breeding age
Mature Adult
Old Adult
GD 0-15
GD 15-20
(18-22 days)
PND 0-14
PND 21
(19-28)
PND 21-35
PND 35-?
1.5-2 mos
Rabbit
Embryonic
Fetal
Neonate
Weaning
Young
Puberty
Breeding age
Mature Adult
Old Adult
GD 0-19
GD 19-32
(30-32 days)
PND 0-21?
PND 42
(42-56)
PND 42-
3-8 mos
6-9 mos
Beagle Dog
Embryonic
Fetal
Neonate
Weaning
Young
Puberty
Breeding age
Mature Adult
Old Adult
GD 0-30?
GD 30-63
(53-71
days)
PND 0-21
PND 42
1.5-5 mos
5-7 mos
12 mos
-15 yrs
Human
Embryonic
Fetal
Neonate
Infancy
Toddler
Preschool
Elementary
School Age
Adolescence
Young Adult
Mature Adult
Old Adult
GD 0-58
GD 58-267
PND 0-30
PND 30-
lyr
2-3 yrs
3 -6 yrs
6-12 yrs
12-21 yrs
2 1-40 yrs
40-65 yrs?
> 65 yrs?
ARange of gestation length in parentheses.
bSome neonatal events in rodents occur in utero in humans.
"Range of weaning ages in parentheses.
-------
External Review Draft Do Not Cite or Quote
A.I. Exposures and endpoints related to general toxicity testing
Table 3-2 provides an overview of the biological systems and other endpoints that are
evaluated by routine toxicity test designs. The table includes all of the routine test designs that
are available in Agency testing guidelines for evaluating toxicity, and includes most of the test
designs that focus on specific biological functions. The acute and subchronic studies are
intended to give general information on the potential toxicity of an agent by screening the major
organ systems, in particular, the liver, the kidney and the gastrointestinal tract. This information
can then be used to determine where to look in more detail at specific organ system structure and
function. The chronic studies, which are usually done in combination with a carcinogenicity
study, evaluate general toxicity in all major organ systems. Several testing guidelines have been
developed with the idea that certain systems should be evaluated frequently in more detail (e.g.,
neurotoxicity studies) or that the general toxicity studies do not provide any indication of a
potential for effects (e.g., reproductive and developmental toxicity studies). More detailed
information about specific aspects of guideline test designs for certain systems (e.g., life stages
covered, exposure periods, outcomes measured, etc.) is included in the figures.
Table 3-2 is shaded and marked to indicate the extent of the evaluation of a particular
system/endpoint within a particular test design. A ^^^^| indicates that the system/endpoint is
a primary focus of the particular test design and that detailed assessment of the dose-response
relationship of an exposure is carried out within some defined life stage and exposure period, for
major elements of the system/endpoint. AI ;4£Y Vindicates those systems/endpoints for which
some histopathology or clinical measure of system function is carried out. A indicates
those systems/endpoints that are assessed in some observational or gross manner. A "0"
indicates that the system/endpoint cannot be included, generally because of the design of the test.
A "blank" indicates that the system/endpoint is not presently included, but could be if the test
design were altered appropriately.
It is obvious from the table that few systems/endpoints are examined in any significant
detail. The systems/endpoints under the acute test designs are for the most part observational in
nature. The Acute Inhalation Toxicity with Histopathology Guideline (40CFR799.9135) was
developed under the Toxic Substances Control Act (TSCA) for characterizing the exposure-
response relationship for sensitive endpoints following acute inhalation exposure and the
toxicologic response following acute high exposures (see further discussion below in section
A. 1 .a.). Acute toxicity information is useful in establishing reference values for short-duration
5-5
-------
External Review Draft Do Not Cite or Quote
exposures and for establishing dose-ranges for subchronic and chronic studies. The subchronic
and chronic test designs evaluate most endpoints with somewhat greater detail than do the acute
-------
External Review Draft
Do Not Cite or Quote
Table 3-2. Systems/Endpoints Evaluated by Routine Toxicity Guideline Testing Protocols
Level X plus histopathology or some clinical measure of system function. The prenatal developmental toxicity study includes a more in-depth structural evaluation.
Includes some observational or gross endpoints
0
Cannot include major aspect
blank
Does not routinely include an aspect, but could
A Series 870 guideline(s) exists for conducting each of the above tests.
3-7
-------
External Review Draft Do Not Cite or Quote
endpoints include cage side observations, body weight at the end of the observation period, gross
test designs. Although the histopathology and/or clinical measures of system function are
screening in nature, there is greater confidence that with this level of examination the
dose-response relationship will be more clearly defined. Nevertheless, it should be recognized
that most systems/endpoints are evaluated at a screening level. Detailed analyses of pathology
and function are generally not carried out. Even in those test designs that do incorporate detailed
analyses, these analyses are limited in regard to the life stages, exposure periods, and measures
that are assessed
Figure 3-1 shows the study designs that are used for general toxicity testing superimposed
on a time line that indicates the life stages during which exposure occurs (hatched bars) and
endpoints are measured (indicated in the boxes). The guideline studies shown represent the
minimum requirement for derivation of a chronic oral RfD. Similar studies are required for the
chronic inhalation RfC, with appropriate endpoints for inhalation exposure and toxicity included.
In some cases, only a 90-day subchronic study is available instead of the chronic studies shown.
Because the relative length of time between life stages varies among species, the placement of
exposures and endpoints on the figures is not necessarily to scale. The following sections discuss
the studies that address acute and short-term toxicity as well as chronic toxicity. Similar figures
are shown in subsequent sections related to specific organ system toxicity testing.
A.I.a. Acute and short-term toxicity studies
Overview of tests. The primary purpose of the guideline acute toxicity tests (870.1100 acute oral;
870.1200 acute dermal; and 870.1300 acute inhalation) and other short-term studies (e.g., 14-to-
28 day studies, no OPPTS guidelines available) is to identify hazards (focusing on route-specific
lethality) from short-term exposure studies, to provide a basis for classification and labeling, and
3-8
-------
External Review Draft Do Not Cite or Quote
to select exposure ranges for longer-term studies2. Acute guideline studies are conducted in
young adult animals with a 14-day post-exposure observation period. Other than mortality, the
Alternative test protocols have been adopted by the Organization for Economic
Cooperation and Development (OECD) for acute toxicity testing for oral, dermal, and inhalation
exposure, including the fixed dose procedure, the acute toxic class method, and the up-and-down
procedure; all are designed to minimize animal usage and provide minimal hazard and dose-
response information for classification, labeling, and dose selection. EPA plans to put primary
reliance on the up-and-down procedure in the future for testing of technical grade pesticides,
although the other tests may be acceptable in some circumstances, e.g., testing of pesticidal
products. These studies are not designed to provide information for use in less-than-lifetime risk
assessment.
3-9
-------
Preconception
Emb/Fetal
weaning
Juvenile
Adolescence
Adulthood
Old Age
Neurotoxicity evaluation
Satellite group
followed 28 days
Satellite group
followed 28 days
Neurotoxicity evaluation
External Review Draft Do Not Cite or Quote
Figure 3-1. Exposures and Endpoints Related to General Toxicity Evaluations*
Life Stages
Guideline Study Designs:
Rodent Combined Chronic/
Carcinogenicity Study
Dog Chronic Study
Prenatal Developmental
Toxicity Study
Reproduction and
Fertility Study
Opthalmological exam, gross
structure & histopathology
Fetal survival, weight,
gross structure
Clinical observations, viability,
body weight, food & water
consumption, clinical pathology
(every 6 months)
P & F1 estrous cycles, sperm
measures, fertility, pregnancy
maintenance, parturition, organ
weight, gross structure,
histology
F1 & F2 litter size, sex ratio, F1 & F2
survival, weight, gross structure,
brain, spleen, thymus weights
F1 sexual
maturation
*Endpoints shown are for oral exposures; endpoints specific to inhalation and dermal exposure are included for studies by those routes of exposure.
-------
External Review Draft Do Not Cite or Quote
pathology changes at necropsy, and histopathological examination of organs showing evidence of
gross pathology in animals surviving 24 hours or more. Two other available guideline studies
include acute exposures followed by extensive assessment of a specific organ system. The first is
the acute inhalation toxicity study with histopathology (40CFR799.9135), which was developed
for hazardous air pollutants. This study includes assessments of liver, kidney, and broncho
alveolar lavage samples for several indicators of cellular damage (total protein, cell count,
percent leukocytes) and a phagocytosis assay to determine macrophage activity. For the
respiratory tract histopathology, detailed specifications are provided.
The second, expanded study includes observations following an acute exposure is the
acute neurotoxicity study (870.6200), which was developed for the evaluation of neurotoxic
chemicals and includes assessments of functional behavior and motor activity at the time of peak
effect and again at 14-days post-treatment, plus histopathology of the central and peripheral
nervous systems at 14-days post-treatment. The prenatal developmental toxicity study
(870.3700) in two species (typically rats and rabbits) and the DNT study (870.6300) can also
provide relevant data for acute risk assessment because maternal observations are often recorded
daily and because of the presumption that effects during development may result from a single
exposure.
Gaps in life stage of assessment. Acute/short-term testing is done only in prenatally
exposed animals and in young adults. No direct information is available from any of these
studies on acute or short-term exposure in postweaning young animals or aged animals.
Gaps in assessment endpoints. Data on only a limited number of toxicological endpoints
are available from guideline acute toxicity (lethality) studies, except in the case of the acute
inhalation toxicity guideline study with histopathology and the acute neurotoxicity study.
Consequently, these studies often are not suitable for use in deriving reference values unless
additional data, such as those from subchronic studies (e.g., hematological, clinical, histology of
more organs), are collected. Some data from animals examined at early times might be available
in guideline subchronic or chronic studies. These data could augment the results from guideline
acute studies.
Gaps in duration of exposure/latency to response assessment. There is no guideline study
for short-term toxicity testing, although the prenatal developmental toxicity studies in rats and
rabbits and the DNT study include repeated dosing of maternal animals for periods of less than
25 days. Because of the post-exposure observation period in acute guideline studies and in the
DNT study, some information on latency to effect and reversibility of effect may be available.
3-11
-------
External Review Draft Do Not Cite or Quote
A.l.b. Subchronic and chronic toxicity studies
The subchronic exposure studies (870.3100, 870.3150, 870.3200, 870.3250, 870.3465)
are used for setting chronic RfDs and RfCs when a chronic study is not available. The guideline
studies for chronic (870.4100, 870.4200, 870.4300) exposures (1 year in rodents, although the
typical study is a 2-year exposure combined chronic and carcinogenicity study) provide an in-
depth look at a number of organ systems, and in some cases they evaluate both structure and
function (see Figure 3-1). The chronic study in nonrodents, usually dogs, involves a 12-month
exposure with similar endpoints assessed as in rodents. The prenatal developmental toxicity
study (870.3700) in two species (typically rats and rabbits), the DNT study (870.6300), and the
reproduction and fertility study (870.3800), typically in rats, are also considered in setting
chronic RfDs or RfCs.
Gaps in life stage of assessment. The subchronic and chronic studies are conducted in
young adult animals, with exposure in the chronic/carcinogenicity study continuing into old age.
No information is available from chronic studies in pre- or postnatal animals. Exposures in
subchronic study protocols do not include pre- or postnatal development, although the
reproduction and fertility study does provide data on subchronic exposures in animals that are
exposed before birth, through prenatal and postnatal development up to mating of the Fl males
and females and through pregnancy (Fl young adult females). No subchronic toxicity
evaluations are conducted in aged animals. No chronic studies are conducted in pre- or postnatal
animals, although aged animals are exposed and evaluated as part of the chronic study protocol.
Gaps in assessment endpoints. The greatest gaps appear to be the lack of routine testing
for subchronic neurotoxicity in adults, immunotoxicity testing in adults, and more thorough
pharmacokinetics in animals at various life stages. Gaps in assessment endpoints during prenatal
and postnatal development are discussed in the next section. Assessment endpoints for routine
toxicity testing in old age are completely lacking, as is background information on endpoints
related to the aging process itself.
Gaps in duration/latency assessment. Chronic studies that include prenatal and postnatal
exposure into old age are lacking. The so-called chronic study in dogs is actually a short-term
study, as it does not cover at least 10% of the lifespan. Chronic studies that include a satellite
group in which exposure is stopped after 12 months in rodents do assess latency to response for a
brief period of time (28 days or more).
3-12
-------
External Review Draft Do Not Cite or Quote
A.2. Exposures and endpoints related to evaluation of reproductive toxicity
Overview of tests. The reproductive organs are examined structurally in a number of
general guideline screening studies, including the 90-day subchronic study (OPPTS 870.3100,
870.3150, 870.3250, 870.3465), chronic/carcinogenicity studies (OPPTS 870.4100, 870.4200,
870.4300), the prenatal developmental toxicity study (OPPTS 870.3700), and the two-generation
reproduction study (OPPTS 870.3800). In addition, extensive assessment of numerous
functional aspects of the reproductive system is conducted in the two-generation reproduction
study. Specific functional effects on the reproductive system of male animals can also be
assessed in the rodent dominant lethal assay (OPPTS 870.5450). As illustrated in Figure 3-2,
these studies include a variety of both structural and functional assessments of the reproductive
system over a wide sampling of life stages.
In guideline subchronic and chronic/carcinogenicity studies, gross structural evaluation
and general qualitative histopathology are conducted on reproductive organs and tissues. The
animals in these studies are adults, but they may be young (e.g., rats 45 days to 5 months of age
from a subchronic study), mature (e.g., rats 5-18 months of age from a reproduction study), or
old animals (e.g., rats 18 months to 2 years of age from a chronic study) at the time of organ
assessment, depending on the protocol.
Standard guideline prenatal developmental toxicity studies are designed to evaluate the
potential effects of the test substance on the developing fetus. Observations on the reproductive
capacity of the maternal animals in this study generally consist only of clinical observations
(including any abnormalities of pregnancy maintenance) and gross necropsy data (including
uterine). Selected fetuses are examined for gross structural changes to the internal reproductive
organs. In studies that employ methods of serial sectioning in the process of soft tissue
examination, a limited macroscopic evaluation of the internal structure and integrity of the
reproductive organs is performed; however, the fetal tissues are not examined microscopically.
Additionally, there are no assessments of organ function in this study design.
In the guideline reproduction study, rats are exposed to the test substance over the
duration of two generations, beginning when the first generation animals are young adults of
approximately 6-9 weeks of age. Daily exposure continues during all phases of development
and reproductive function; adult animals of both generations are killed as mature adults,
generally prior to reaching reproductive senescence (that is, the cessation of normal reproductive
function) or an age that would be considered geriatric in that species. Assessments of
3-13
-------
External Review Draft Do Not Cite or Quote
reproductive capability and function are conducted at least once in each generation. These
assessments include direct evaluation of the age of sexual maturation, estrous cyclicity
(immediately prior to mating), sperm measures (at termination), mating success, fertility and
fecundity, implantation, pregnancy maintenance, gestation duration, parturition, and success of
lactation (e.g., maternal nurturing and nesting behavior).
Indirect assessments of some reproductive functions are also evaluated. These
observations are based on evidence of normality in a structure, function, or process that is
dependent on normal functioning of the component parts, including, for example, hormonal
homeostasis, ejaculation, accessory gland function, placental function, milk production, pup
nursing behavior or ability, and, to some extent, reproductive senescence (although the adult
animals are terminated at the end of each generation, when they are only around 6 months of age;
therefore, there are no assessments conducted in older rats). Gross structural assessments of the
whole animal are conducted on adult and immature animals throughout the course of the study;
gross internal (organ) structural assessments are conducted on offspring that are killed at litter
standardization (postnatal day [PND] 4), weaning (PND 21), and termination of each generation
(mature adults). Histopathological evaluation of the reproductive organs (gonads and accessory
structures) is conducted only in the mature parental adult animals that are killed at the
termination of each generation. The guideline specifies a very focused pathological examination
of the reproductive organs in this study.
The dominant lethal assay is not conducted for every chemical, but it may be conducted
in response to a concern raised by other developmental or reproductive toxicity findings in the
database. In this study, sexually mature adult males are treated with the test substance to
determine whether there is an effect in the germinal tissue that does not cause dysfunction in the
gamete but is lethal to the fertilized egg or developing embryo. Exposed males are mated with
untreated females and uterine contents are evaluated. Evidence of pre- and/or postimplantation
loss is generally thought to be indicative of treatment-related chromosomal damage in germinal
tissue.
Gaps in life stage of assessment. Determination of gaps in the assessment of potential
effects of any chemical across all life stages requires consideration of both the exposure period
and the time of assessment. In the prenatal developmental toxicity study, animals are exposed
from implantation through gestation. The reproductive organs are examined for gross structural
changes, but no microscopic examination is conducted. There is no follow-up of the animals to
determine the functional consequences of prenatal exposure. In the two-generation reproductive
3-14
-------
External Review Draft Do Not Cite or Quote
toxicity study, the Fl animals are exposed from preconception throughout prenatal and postnatal
development until after mating. The reproductive organs are examined macroscopically at
3-15
-------
External Review Draft
Do Not Cite or Quote
Figure 3-2. Exposures and Endpoints Related to Reproductive Evaluations
Life Stages
NB/Pre-
Preconception |Emb/Fetal| weaning| Juvenile
Adolescence
Adulthood
Guideline Study Designs:
Prenatal Developmental
Toxicity Study
Developmental Neurotoxicity
Study
Reproduction and
Fertility Study
Litter size, sex ratio, fetal survival,
weight, gross structure
F1 &F2
survival, we ght
brain, spleen
litter size, sex ratio,
, gross structure,
, thymus weights
Rodent Dominant
Lethal Assay
Subchronic or Chronic
Toxicity Study
F1 sexual
maturation
Embryonic death
Old Age
Litter size, sex ratio, survival, weight,
gross structure, developmental
landmarks including sexual maturation
P & F1 Estrous cycles, sperm
measures, fertility, pregnancy
maintenance, parturition,
organ weight, gross structure,
histology
Organ weight, gross structure, histology
-------
External Review Draft Do Not Cite or Quote
weaning and adulthood. The maturation of the reproductive system is assessed, as is the function
of the reproductive system. Thus, the study provides a fairly thorough assessment of structure
and function following exposure during many critical periods of development. In the parental
generation, the animals are exposed as young adults, and the structure and function of the
reproductive organs are assessed.
The dominant lethal study, when conducted, assesses a single aspect of the function of the
reproductive system for one sex, although a detailed structural assessment is not conducted. In
the subchronic and chronic studies, the animals are exposed beginning as young adults, and the
structure - but not the function - of the reproductive organs is assessed. Therefore, the major
gaps include (1) the lack of functional assessment (particularly the age of onset of reproductive
senescence) in older adult animals following adult only exposures, and (2) the lack of structural
and functional assessments in older adult animals following developmental exposures.
The onset of reproductive senescence can be marked by findings such as altered hormonal
homeostasis, disruption of estrous cyclicity, diminished sperm measures (number, motility, or
morphology), or gonadal atrophy. Studies in rodents have demonstrated the adverse effects of a
number of agents (e.g., ionizing radiation, chemotherapeutic agents, polycyclic aromatic
hydrocarbons, and agents that form epoxides, such as 1-3 butadiene and 4-vinylcyclohexene) on
reproductive senescence (reviewed by Hoyer and Sipes, 1996). In humans, premature
reproductive senescence has been associated with cigarette smoking (Tick et al., 1977). In
addition to potentially diminishing fertility in individuals who are only slightly past prime
reproductive age, early reproductive senescence can adversely affect the general health of the
aged human. For example, hormonal alterations that are associated with early senescence have
been linked to abnormalities of cardiovascular function, osteoporosis, and even a predisposition
to early mortality.
Gaps in assessment endpoints. As described above, there are identifiable gaps in the
endpoints that are used to assess reproductive toxicity in guideline studies. Currently, there is no
assessment of functional endpoints in older animals following adult exposures, and there are no
structural or functional endpoints assessed in older animals following developmental exposures,
including reproductive senescence. In addition, concerns have recently been raised about the
ability to detect rare malformations of the reproductive organs and abnormalities in the
maturation of the reproductive system in the two-generation reproductive toxicity study. This
concern relates particularly to endocrine-active chemicals. In the current guideline, three
pups/sex/litter are examined macroscopically at weaning. Questions have been raised about
3-17
-------
External Review Draft Do Not Cite or Quote
whether these weanlings should be retained until day 45 (females) or day 60 (males) to ensure
that any later appearing gross or functional changes are detected. This issue is currently being
examined within the endocrine validation/standardization program.
Gaps in duration/latency assessment. There are no studies that include acute or chronic
exposures that can be used to assess the development of the reproductive system. As indicated
above, it has been suggested that animals be retained until older ages in the two-generation study
in order to assess later appearing structural or functional changes in reproductive organs. In
addition, there is no consideration of latent responses for reproductive toxicity, e.g., early onset
of reproductive senescence, as the result of an exposure earlier in life in any of the studies that
can be used to evaluate reproductive toxicity, except for a few endpoints in the DNT study.
A.3. Exposures and endpoints related to evaluation of neurotoxicity
Overview of tests. Observation of the animals for signs of overt toxicity and routine gross
pathological assessment of the nervous system is required under OPPTS acute, subchronic, and
chronic study protocols (870.100-870.400 series). In rat studies, age at initiation of testing is to
be 8 - 12 weeks under acute and subchronic testing protocols. In acute studies, cage-side
observation and gross neuropathology are the only endpoints required under 870.100 (oral,
dermal, or inhalation exposure). Motor activity, grip strength, and sensory reactivity and
neuropathology, are measured in the rodent oral study, the dermal 21 to 28- and 90-day
subchronic studies, and the 90-day inhalation study. In rodent subchronic studies, specific
assessment for neurotoxicity is performed at or near the end of the study, although observations
of the animals, including those for detection of overt neurotoxicity, are made routinely
throughout the study. No specific functional tests for neurotoxicity are required for nonrodent
subchronic studies, although observation and neuropathology are required.
Chronic toxicity studies (oral, dermal, inhalation) are to be performed in two species (one
rodent) over a 12-month period, regardless of the lifespan of the species. Exposure in rodents is
to begin no later than 8 weeks of age. Motor activity, grip strength, and sensory reactivity are to
be assessed at or near the end of the study, but no earlier than the 11th month. Clinical
observation is performed weekly throughout the study, and would presumably detect gross
neurological abnormality. In current practice, the chronic study is often combined with the
carcinogen!city test, in which dosing extends for 24 months in rats and 18 months for mice
(OPPTS 870.4300). Motor activity would be performed at 11 - 12 months only, as in the chronic
study, and not again until near the end of exposure.
3-18
-------
External Review Draft Do Not Cite or Quote
The neurotoxicity screening battery (870.6200) is designed to be included in acute,
subchronic, or chronic toxicity studies (Figure 3-3). The endpoints examined extend those
required in the 870.100 series, although there is no guidance as to when these extended batteries
would be required. The Functional Observation Battery includes a ranking system for general
reactivity, activity, and gait abnormalities, as well as forelimb and hindlimb grip strength,
landing foot splay, sensorimotor reactivity to sensory stimuli, and pain reception. Motor activity
and a more detailed neuropathological observation are also required in this battery. For acute
studies, assessments are made before initiation of dosing, at the estimated peak of activity within
8 hours of dosing and at 7 and 14 days post-dosing. For subchronic studies, assessments are
performed pre-exposure and at 7, 8, and 13 weeks of exposure. For chronic studies, assessment
is at pre-exposure and every 3 months post-exposure. There is no specific guidance regarding the
assessment schedule for the combined chronic/carcinogenicity study, but presumably the
schedule required for the chronic study would be maintained.
The DNT study protocol (870.6300) currently requires dosing of the dams from
gestational day (GD) 6 through PND 10, although the requirement may soon be extended to PND
21 (i.e., until weaning). Motor activity is measured at PND 13, 17, 21, and 60. Auditory startle
is measured around weaning and at PND 60, as is a test of learning and memory, which may be
the same test or different tests at the two time-points. Cage-side observation of both dams and
pups is required, and neuropathology in the pups is required at PND 11 and at the termination of
the study (usually PND 60). The prenatal developmental toxicity study (870.3700) requires
dosing of the dams and from GD 6 through 20 in rats, GD 6 through 29 in rabbits. Gross
structural evaluation of the nervous system is evaluated as part of the fetal examinations
conducted in this study.
Gaps in life stage of assessment. One of the most significant gaps revealed by Figure 3-3
is the lack of exposure or assessment under any protocol during old age. For example, following
acute exposure, assessment is for 14 days in juvenile or young adult animals. The chronic
exposure protocol extends exposure into adulthood; the combined chronic/carcinogenicity
protocol extends exposure up to approximately the aged period in the rat, but neurotoxicology
assessments are not performed in aged animals. Thus, none of the protocols assess potential
effects of chemicals on aging as a function of exposure during development. This may be
important because studies in animals have shown that developmental exposure to agents that
cause neurotoxicity, such as trimethyl tin, can accelerate the onset of cognitive deficits measured
later in life. Other studies with methyl mercury have documented early onset sensory
3-19
-------
External Review Draft Do Not Cite or Quote
dysfunction in monkeys exposed during development. Furthermore, current testing protocols do
not provide information collected at different life stages, i.e., comparison of effects of exposure
during infancy, adulthood, or old age. This is important, because life-stage dependent differences
3-20
-------
External Review Draft Do Not Cite or Quote
Figure 3-3. Exposures and Endpoints for Neurotoxicity Evaluations
to
Life Stages
Preconception
NB/Pre-
Emb/Fetal weaning] Juvenile
Guideline Study Designs:
Neurotoxicity
Screening
Battery
acute
subchronic
(90 days)
chronic
(12 months)
Chronic/carcinogenicity
(24 months)
Developmental Neurotoxicity
Study
Prenatal Developmental
Toxicity Study
Delayed Neurotoxicity of
Organophosphorous
Pesticides (adult hen)
acute
28-day
°vn
\
Adolescence
A
Auditory startle,
learning & memory
Neuropathology
Gross structure of the CMS
Adulthood
Old Age
Neuropathology
Motor activity, FOB (includes grip
strength, landing foot splay, gross
sensorimotor reactivity, pain
perception
Schedule-controlled
operant behavior,
peripheral nerve function,
sensory evoked
potentials
May be requested/required
b
Neuropathy target esterase
(NTE), neuropathology, gross
observation, motor function,
acetylcholinesterase (AChE)
-------
External Review Draft Do Not Cite or Quote
in pharmacokinetic and, possibly, pharmacodynamic parameters could result in quantitatively or
qualitatively different effects at different life stages.
Under the DNT protocol, there currently is no requirement to perform kinetic studies to
ascertain either in utero or postnatal exposure. There is no mechanism to guarantee exposure
postnatally (i.e., direct dosing of pups), because the compound may not be excreted into breast
milk, or it may be excreted only at very low concentrations. This is of particular importance
because the early postnatal period in the rodent is equivalent to a prenatal life stage in humans.
There is no long-term follow-up assessment to detect delayed neurotoxic effects, a situation that
is arguably more worrisome for developmental exposure than for exposure later in life.
Gaps in assessment endpoints. The nervous system is one of the most fully assessed
organ systems in the EPA/OPPTS 870 guidelines. Nonetheless, most of the endpoint
assessments are designed to be screening procedures rather than sensitive assessments of nervous
system function. In addition, the assessments required are different in the neurotoxicity
screening battery than in the DNT study. The adult neurotoxicity screening battery does not
require assessment of learning and memory or auditory startle. The lack of assessment of
cognitive function in the neurotoxicity screening battery constitutes an omission for which there
is no scientific justification. It may also be pointed out that even in the developmental protocol,
the tests that are used to assess learning and memory may be very simple, potentially revealing
only relatively gross deficits. In addition, although potentially more sensitive cognitive, sensory,
and motor tests are available (Figure 3-3), there is no guidance as to what would trigger a
requirement for these assessments. Except for the protocol for delayed neurotoxicity for
organophosphorous pesticides in the hen, there is no assessment of neurochemical endpoints.
Additionally, the required neuropathological assessments may also be considered screening.
Minimal morphometric analysis is required in the DNT study, consisting of the thickness
of "representative" layers in the neocortex, hippocampus, and cerebellum. No morphometric
analyses are required in the adult neurotoxicity testing protocols. Although more sophisticated
tests would presumably not be performed on all agents, more sophisticated measures could be
triggered by results from screening tests. It may also be advisable to require more sensitive tests
in instances of particular concern, e.g., adding more extensive morphometric analysis to the DNT
protocol.
In summary, although the nervous system is one of the most thoroughly assessed systems
in the 870 test guideline studies, it must be kept well in mind when interpreting the results that
3-22
-------
External Review Draft Do Not Cite or Quote
these are screening tests. Positive findings must be viewed as indicative of relatively overt
toxicity, not so-called subtle effects.
Gaps in duration/latency assessment. One of the principles in the neurotoxicity risk
assessment guidelines (EPA, 1998c) is that neurotoxicity could occur after one or a few
exposures, such as in the case of an organophosphate insecticide that produces a delayed
neuropathy, or only after a series of repeated exposures, as in the case of acrylamide. In the case
of DNT, it is assumed that a single exposure to a chemical during a critical period of
development could result in an adverse effect on the developing nervous system. There are,
however, few data that compare the effects of a single exposure to a chemical with the effects of
the same chemical given multiple times during development.
A.4. Exposures and endpoints related to evaluation of immunotoxicity
Overview of tests. Examination of the macro- and/or microscopic structural anatomy of immune
system organs and tissues is performed in a number of general guideline screening studies,
including the acute inhalation toxicity with histopathology guideline (40CFR799.9135), the 90-
day subchronic study (OPPTS 870.3100, 870.3150, 870.3250, 870.3465), the
chronic/carcinogenicity studies (OPPTS 870.4100, 870.4200, 870.4300), the prenatal
developmental toxicity study (OPPTS 870.3700), and the two-generation reproduction study
(OPPTS 870.3800). In addition, functional assessments of the immune system are evaluated in
the skin sensitization study (OPPTS 870.2600) and the immunotoxicity testing guideline (OPPTS
870.7800) (see Figure 3-4).
In the guideline immunotoxicity study, young adult rats (6-8 weeks of age) are exposed
to the test substance for 28 days, at which time they are terminated. The spleen and thymus are
examined macroscopically, and organ weights are recorded; histopathological evaluation is not
performed. Assessments of immune system function include an evaluation of the response to T-
cell-dependent antigen, sheep red blood cells (SRBC). The SRBC antigen response assays can
be conducted either by an antibody plaque-forming cell (PFC) assay or by immunoglobulin
quantification by enzyme-linked immunosorbent assay (ELISA). In addition, an assessment of
natural killer (NK) cell activity and/or enumeration of splenic or peripheral blood total B cells,
total T cells, and T cell subpopulations may be required on a case-by-case basis.
The skin sensitization study is generally conducted in guinea pigs. It involves an initial
intradermal and/or epidermal exposure of the test animal to a substance, followed by an
challenge exposure at least 1 week later. Sensitization is determined by examining the reaction
3-23
-------
External Review Draft Do Not Cite or Quote
to the challenge exposure and comparing this reaction with that of the initial induction exposure.
Histopathological evaluation of the skin is not required but may be conducted. No other immune
system endpoints or organs are evaluated in this study.
-------
External Review Draft
Do Not Cite or Quote
OJ
to
Figure 3-4. Exposures and Endpoints for Immunotoxicity Evaluations
Life Stages
Guideline Study Designs:
Immunotoxicity Study
Skin Sensitization Study
(guinea pig preferred species)
Subchronic Study
Chronic/carcinogenicity Study
Prenatal Developmental
Toxicity Study
Reproduction and
Fertility Study
Preconception
Emb/Fetal
NB/Pre-
weaning
Juvenile
Adolescence
Adulthood
Old Age
*/..
V
Gross structure & weights of thymus/spleen,
response to T cell-dependent antigen, SRBC
(RFC or ELISA). Case-by-case - NK cell activity,
total B cells, T cells and T cell subpopulations
Reaction to initial induction or challenge dose
Gross structure, thymus/spleen weights,
spleen/thymus/lymph node (2 sites) histology
Clinical pathology (differential WBC,
serum immunoglobulin every 6 months
Gross thymus/spleen structure
Organ weights, gross
structure, histology
Spleen/thymus
weights
5-25
-------
External Review Draft Do Not Cite or Quote
In guideline subchronic and chronic/carcinogenicity studies, an evaluation of macroscopic
structure and general qualitative histopathology are conducted on only a few immune system
tissues. In studies that include young adult animals (e.g., rats 45 days to 5 months of age from a
subchronic study), the spleen, thymus, and lymph nodes from two locations (one near to and the
other distant from the site of administration) are examined; the spleen and thymus are weighed.
In chronic and carcinogenicity study guidelines, there is no requirement that the thymus be
examined and/or weighed. For rodents (e.g., rats or mice 18 months to 2 years of age), it is
reasonable to assume that the thymus would have undergone normal age-related atrophy by study
termination. However, the thymus might be present at early interim sacrifices of rodents (e.g., at
6 months or 12 months of study) during a long-term study, and it would certainly be present at
study termination in a canine chronic study (at which point the dogs are young adults of only
approximately 1.5 years of age).
Differential white cell counts in the circulating blood are examined at study termination
in the subchronic study and at approximately 6-month intervals in long-term studies. Serum
immunoglobulin levels may be measured at the same intervals. Perturbations may indicate
increased immune system response to some unspecified initiator, but this information does not
address the adequacy of immune system function. In the same manner, histopathological
evaluation of other organ systems in the subchronic and chronic/carcinogenicity studies may
identify cellular alterations that are nonspecific indicators of an effect on immune response, for
example, the presence of increased numbers of macrophages in lung tissue or an increased
incidence of inflammatory dermal lesions.
In the two-generation reproduction study in rats, a macroscopic evaluation of all organ
systems is conducted in a sample of offspring at weaning and in the mature adult parental
animals at the termination of each generation. Additionally, the spleen and thymus are weighed
for those pups that are necropsied at weaning; these measurements are intended to provide
information on the need for further evaluation of immunotoxic potential of a chemical to the
immature animal.
In the prenatal developmental toxicity study, an evaluation of the macroscopic structure
of the thymus and spleen is conducted in at least half of the fetuses from each litter.
Gaps in life stage of assessment. In the available guideline studies, assessments of organs
with immune system function are conducted in fetuses following prenatal exposure and in
weanling animals following pre- and postnatal exposure and at a variety of time points in young
and/or mature adult animals. With prenatal exposures and evaluation at early life stages, these
3-26
-------
External Review Draft Do Not Cite or Quote
assessments consist entirely of the evaluation of macroscopic changes, with no microscopic
examination. Pharmacokinetic data that characterize the exposure in the young (i.e., exposure of
the fetus to the chemical or its metabolites via the placenta or of the neonate via breast milk) are
not routinely required and are seldom available.
Some detailed structural assessment (histopathology) is conducted in mature or older
adult animals. Indirect assessment of immune system function is conducted in adult animals of
various ages via the evaluation of peripheral blood cells and chemistry. Direct functional
assessments of the immune system are conducted only in young adult animals; generally this age
group is selected for assessment because of the anticipated robustness of the immune response.
There is no guideline that examines potential perturbation of immune system function
following early pre- and/or postnatal exposure (often referred to as a developmental
immunotoxicity study). Comparisons of immune effects following exposure at various life
stages (i.e., during in utero or postnatal development, adulthood, or old age), including data that
analyze whether these effects are more severe in one age group or if the effects are persistent, are
not required. To achieve even a minimal assessment of immune system structure and function, a
broad variety of studies would need to be conducted and assessed; yet there could still be
relatively low confidence in the ability of the results of these combined studies to predict the
outcome of age-specific insults to the immune system.
Gaps in assessment endpoints. There are identifiable gaps in the endpoints that are used
to assess immunotoxicity in guideline studies. For example, for fetuses, immature animals, and
old animals (rodents), assessments are composed entirely of the evaluation of macroscopic
structural changes, with no histopathological or functional evaluations. In mature adult animals,
thorough macroscopic and microscopic structural assessments, as well as routine hematological
testing (e.g., blood cell counts), are performed; however, those assessments are generally very
limited in young animals, and guideline requirements do not consider species differences. The
only assessments of functional integrity of the immune system are provided by the guideline
sensitization study and the 28-day immunotoxicity study. These studies are conducted only in
young adult animals, and they include only a few examples of potential immune system response
(e.g., hypersensitivity, humoral immunity, or nonspecific cell-mediated immunity). In very
young and very old animals, there is no direct assessment of immunological function. No
assessment of autoimmune effects is conducted in any of the current guideline protocols.
Gaps in duration/latency assessment. Latent effects on immune function that result from early
lifetime exposure are not assessed; these can include effects in aged animals that result from in
3-27
-------
External Review Draft Do Not Cite or Quote
utero, neonatal, or young adult exposure. Exacerbation of effects in relation to aging and
response to subsequent immunological challenge are not routinely or systematically assessed to
any extent. The two-generation reproduction study offers an opportunity to evaluate
immunotoxic response in adulthood that resulted from prenatal or early postnatal exposure. In
the chronic toxicity studies in rodents, aged animals are available for evaluation. However, in
both cases, there is little focus on the evaluation of the immune system. Only indirect evidence
of perturbation of the immune system may be observed through macroscopic and microscopic
evaluation of various organs; corollary functional assessment is not performed. Response to an
immunological challenge is examined only in the guinea pig hypersensitization study, and even
when the results from this study are positive, no further specific assessment of the immune
system is pursued.
A.5. Exposures and endpoints related to evaluation of cardiovascular toxicity
Overview of tests. Gross observation of the heart and major vessels augmented by
conditional standard pathology is mentioned in most applicable OPPTS Series 870 Health Effect
guidelines (Fig. 3-5).
Gaps in life stage of assessment. The period from birth to maturity is essentially without
toxicological monitoring of cardiovascular endpoints for both repeated chronic and single acute
exposure regimes.
Gaps in assessment endpoints. Gross observation only of the heart is provided for in
most OPPTS guidelines. Functional clinical, or histopathological cardiac examination is not
currently part of any testing guideline. Even gross pathology could be improved and brought into
line with current cardiovascular evaluation by separating, weighing, and constructing right and
left ventricle/BW ratios to give an evaluation of cardiac hypertrophy. Also, guidelines regarding
sectioning procedures for the heart, either number or plane, could be provided.
No simple cardiac functional evaluation is currently available, including even systolic or
diastolic blood pressures. It should be noted that telemetric in-dwelling echocardiograms
(ultrasound examinations of the heart) can be used to detect occlusions and atherosclerosis and to
detect alterations in cardiac output. Combination echocardiograms and electrocardiogram
analysis can detect cardiac wave forms as well as heart rate variability in high- and low-
frequency power ranges (i.e., beat-to-beat changes in heart rate ascribed to varying control by the
autonomic nervous system). Heart rate variability may be critical in explaining toxicity, as was
shown in recent work associating exposures to fine particulate matter with decreases in heart rate
3-28
-------
External Review Draft Do Not Cite or Quote
variaiblity in elderly humans (Creason et al., 2001). Both echocardiograms and
electrocardiograms can be done on rats down to 100 g, well within the size range of juvenile and
adolescent rats.
Chemicals can produce degenerative and/or inflammatory changes in the peripheral blood
vessels as a consequence of an excessive pharmacologic effect or by an interaction with a
vascular structural or functional macromolecule. As a result of sustained arterial
vasoconstriction, peripheral arterial lesions consisting of intimal proliferation and medial
degenerative changes could result in gangrene. Also, chemicals can induce or enhance atheroma
formation characterized by endothelial damage with increased permeability, monocyte adhesion,
and endothelial proliferation. Selected representative techniques to study the peripheral vascular
system consist of flow measurement techniques (Smith et al. 1994), such as electromagnetic
flowmetry, pulsed Doppler flowmetry, transit time flowmetry, laser Doppler fluxmetry, and laser
scanner methods. These techniques allow investigation of blood flow in vessels as large as the
aorta and as small as the capillary, determination of the level of perfusion in tissues, and
calculation of the derived hemodynamic variable of resistance. The two major noninvasive
techniques for determining microvascular velocity are the flying spot technique and the dual-slit
technique. External ultrasound may be used to examine internal vascular dimensions. A
noninvasive assessment of arterial flow in rodents and monkeys can be performed using Doppler
spectrum analysis (duplex ultrasound technology) (Leopold et al., 1997). This test detects
arterial compromise in extremities, functional severity, and the hemodynamic significance of
vascular lesions. In most cases, the locations in the arteries involved can be designated.
Information regarding the extent and effectiveness of collateral circulation can also be gained.
This testing is a valuable tool for monitoring early flow compromise secondary to chronic
reoccurrence of anastomotic or distal disease.
Several blood/plasma tests for clinical assessment are in active use in cardiovascular
research. In general, these are tests that may be used to document a cardiovascular accident
(within 48 - 96 hours). Their utility for risk assessment have yet to be evaluated. Specific
enzymes currently being used by the research community for these purposes include LDH-I,
creatinine kinase-II, and troponin. Other enzymes useful as prognostic indicators of risk of a
cardiovascular accident include the angiotensin converting enzyme (ACE II), plasma renin
activity (PRA), endothelin-converting enzyme-1 (ECE-1), and catecholamines (epinephrine and
norepinephrine).
3-29
-------
External Review Draft
Do Not Cite or Quote
Figure 3-5. Exposures and Endpoints Related to CardiovascularEvaluations
NB/Pre-
OJ
o
Life Stages
Guideline Study Designs:
Preconception
Emb/Fetal
weaning
Juvenile
Adolescence
Adulthood
Old Age
Acute Toxicity Test
Subchronic Toxicity Test
Combined Chronic Toxicity/
Carcinogenicity Test
Prenatal Developmental
Toxicity Study
!^
Gross
pathology,
no
sectioning
instructions
stipulated
Gross structure
-------
External Review Draft Do Not Cite or Quote
B. Conclusions and Recommendations
A review of current testing guidelines was conducted to determine the types of data
available for setting reference values. The approach used was to evaluate testing guidelines from
the point
of view of life stages covered, of endpoints assessed generally and for specific organ systems, of
timing and duration of exposure, and of evaluation of reversibility and latency to response.
The relevance of these issues to the health evaluation of children and other potentially
susceptible subpopulations should be apparent from the gaps identified in each of the above
sections regarding life stage assessment, endpoints assessed, timing and duration of exposures
included in guideline studies, reversibility, and latency to response. Although a number of areas
of toxicity testing have been discussed, this review should not be considered exhaustive, and
other health effects may be of as much or more importance for particular chemicals than those
reviewed in detail here. Of particular concern for children's health that have not been discussed
in great detail here are effects related to asthma and other respiratory tract toxicity. For both
children and the elderly, renal and liver function can be a major factor in the disposition and
excretion of chemicals, and therefore their toxicity. Thus, the evaluation of toxicity and the
interpretation of data in terms of its completeness will always require scientific judgment about
whether or not adequate data have been collected on effects of importance at the appropriate life
stages, timing and durations of exposure, etc., for a given agent.
Effects seen at the termination of a chronic study may be due to cumulative damage from
a continued repeated chemical insult, but they could also be a latent response from a single or
short-term multiple earlier exposure. Thus, latent effects might be revealed in chronic studies,
but it would not be clear whether they were the result of acute/short-term exposure or the chronic
exposure. Specific information on the latency of a response would follow only from a clearer
understanding of the mechanism of the effect and from actual "stop exposure" protocols (e.g., the
satellite studies depicted in Fig, 3-1) or from shorter-term exposures with follow-up over a much
longer period of time. It thus follows that any chemical database that does not have exposure-
response studies of lifetime duration or any specific exposure-latency protocols would not cover
the possibility of latent effects.
Effects that persist throughout a designated post-exposure period may be considered
irreversible; those that do not are reversible. For chronic lifetime exposures, designation of an
effect as irreversible or reversible is academic, as exposure is presumed to be lifetime, i.e., there
is no post-exposure period. For shorter-term values (e.g., acute, short-term) where an appreciable
3-31
-------
External Review Draft Do Not Cite or Quote
period of time post-exposure is anticipated, designation of an effect as reversible or irreversible
becomes more relevant. Derivation of a reference value based on shorter-term exposure
guideline protocols would have to fully consider the aspect of reversibility in interpretation of the
data. It is important to understand the difference between an endpoint that is truly reversible and
one that is related to or is a precursor of other adverse effects. For example, low birth weight
may be "reversible" through catch-up growth postnatally, but it also may be related to
developmental delays or other health outcomes that result from prenatal growth
reducti on/retardati on.
B.I. Conclusions
From this review, the Technical Panel reached the following major conclusions:
1. There are a number of gaps in life stages covered in current guideline testing protocols,
particularly in terms of the exposure periods included. In particular, there is minimal evaluation
of aged animals, especially after exposures that include early development.
2. There are a number of gaps in the evaluation of endpoints included for certain systems;
for example, the evaluations of the cardiovascular and immune systems in various guideline
studies were reviewed as examples of systems that are minimally covered. Other systems, e.g.,
the reproductive and nervous systems, are evaluated in more detail, but even in these systems
there are gaps that need to be considered. Notably, functional evaluations are not always
included or integrated with structural evaluations of particular systems.
3. Acute and short-term exposure duration studies are either not available or include only
gross effects, so that the data needed to derive acute and short-term reference values are often not
available.
4. Latency to response and reversibility are only rarely evaluated directly. These types of
effects could have a major impact on hazard characterization, especially in designing acute and
short-term test guideline protocols and ultimately on the risk management options that can be
used for intervention or prevention.
5. Although not more specifically discussed, it is clear that there is a lack of information on
pharmacokinetics. The data that are available are generally limited to studies that are conducted
in young adult animals, buth there are no guideline protocols for pharmacokinetic evaluations
during development or in older age related to exposures and outcomes.
6. The underlying assumption that the internal dose of the active form of an agent to the
target site is the relevant measure of dose clearly underscores pharmacokinetics as an essential
3-32
-------
External Review Draft Do Not Cite or Quote
tool that must be used in both hazard identification and dose-response evaluations. This should
not only continue to be a central and critical area of exploration, it should be an area of direct
application to assessment activities to address various issues including, but not limited to (a)
design of studies, (b) delivery to the fetus/neonate, (c) dose scaling, (d) pharmacokinetic and
pharmacodynamic considerations, and (e) route extrapolation.
A white paper on pharmacokinetics commissioned by this Panel (Versar Inc., 200Ib) is
meant to serve as a technical resource for the application of pharmacokinetics to these and other
issues addressed throughout this document. Another white paper on aging also addresses issues
of changing pharmacokinetics during this life stage (Versar Inc., 200la).
7. Portal-of-entry effects (i.e., respiratory, gastrointestinal, dermal) are acknowledged as
being important in the effects of chemicals, and they may preclude systemic toxicity as being
sentinel. Chronic oral RfDs and inhalation RfCs have been developed for a number of agents,
but rarely have dermal RfDs been derived. In some cases, oral RfDs and oral cancer potency
factors have been used to assess systemic toxicity from dermal exposures. However, the dermal
route of exposure can result in different patterns of distribution, metabolism, and excretion than
those that occur from the oral route. Dermal contact with a chemical may also result in direct
dermal toxicity, such as allergic contact dermatitis, urticaria reactions, chemical irritation, and
skin cancer. The dose-response relationship for the portal-of-entry effects in skin is likely to be
independent of any associated systemic toxicity exhibited by a particular chemical. Therefore,
there is a long-term need for the development of dermal RfDs that consider both the systemic
toxicity effects and the portal-of-entry effects of individual chemicals. In addition, there is a
need for data on the dermal uptake of chemicals from soil, water, and air, including information
about specific chemical forms and bioavailability from different soil types that contribute to
variations in uptake. Different exposure duration RfDs, such as acute chemical injury to the skin,
need to be developed.
B.2. Recommendations
Based on the review of the guideline toxicity studies, the Technical Panel makes the
following recommendations.
1. Develop protocols for acute and short-term studies that provide more comprehensive
data for setting reference values (see section C). Develop guidance for how and when to use the
guidelines. The existing protocols for acute studies (except for the acute inhalation protocol with
3-33
-------
External Review Draft Do Not Cite or Quote
histopathologic evaluation) generally collect data only on what could be called frank effects
which may not be protective of more subtle effects.
2. Modify existing guideline study protocols to provide more comprehensive coverage of life
stages (see section C). Develop guidance for how and when to use the guidelines. Existing
guideline studies do not include, for example, the evaluation of toxic effects that may occur in
old age from prenatal or early postnatal exposure (including carcinogenesis) or premature aging
from exposure earlier in life.
3. Collect more information from less-than-lifetime exposure to evaluate latency to effect
and to evaluate reversibility of effect. Develop guidance for how and when to collect such
information. Existing guideline studies, with the exception of the acute tests and some
developmental toxicity studies, expose animals up to the time of testing. Some form of "stop
exposure" studies would provide useful information that could increase or decrease the level of
concern for an observed toxic event.
4. Develop guidelines or guideline study protocols that will provide more systematic
information on pharmacokinetics andpharmacodynamics. Develop guidance for how and when
to use the guidelines. Such studies could provide information that would be relevant to
susceptible subpopulations, including life stages (that is, inform the selection of the intraspecies
UF). Such studies also could provide information on species differences (that is, inform the
selection of the interspecies UF). Finally, such studies can provide information to conduct route-
to-route extrapolations and reduce the number of route- specific tests required to derive a
reference value.
5. Develop guidelines or guideline study protocols to assess immunotoxicity, developmental
immunotoxicity, cardiovascular toxicity. Develop guidance for how and when to use the
guidelines. These endpoints are presently looked at only in a cursory manner. There is a need to
integrate functional measurements into our evaluation of these and other systems.
6. Explore the feasibility of setting dermal reference values for direct toxicity at the portal
of entry, including sensitization. Reference values have been derived for lesions in the
gastrointestinal and respiratory tracts from direct exposure. The lack of procedures for dealing
with similar effects on the skin is a glaring omission.
C. Options for Alternative Testing Approaches
The Technical Panel explored alternative testing protocols for acute toxicity testing, as
well as alternative protocols for sub chronic/chronic toxicity testing. These are offered here as
3-34
-------
External Review Draft Do Not Cite or Quote
alternatives that may be used, depending on the agent being tested or the type of reference values
needed.
C.I. Alternative acute toxicity testing protocol
The current EPA test guidelines for acute toxicity focus on the determination of an LD50
in adult test species. A gross necropsy is conducted on the animals, and histologic evaluation of
target organs may or may not be conducted. Therefore, very limited information is obtained from
the current protocol that would be useful for determining an acute reference value. However, a
number of alternative study designs are available that would provide information for
consideration in establishing the acute reference value (Gad and Chengelis, 1998). One basic
study design is shown in Figure 3-6. In this protocol, a control group and a minimum of three
dose groups are used with 10 animals per sex per group. The animals are dosed once on day 1
and followed for 2 weeks. Clinical signs of toxicity are recorded daily, food consumption and
body weights are recorded on days 1 - 4, 8 and 14. There is an interim sacrifice of five
animals/sex/group at 3 days after dosing and a final sacrifice of the remaining animals at 2 weeks
after dosing. At both sacrifices, hematological and clinical chemistry analyses are conducted, as
is a urinalysis. The animals are necropsied, organ weights are recorded, and the organs are
examined histologically. Because the purpose of this study design is to provide hazard and dose-
response information rather than determination of an LD50, the dose levels should be chosen
accordingly. This study would initially be conducted on adult animals. As information is
obtained from other toxicology and/or toxicokinetic studies, it may be necessary to conduct the
study with animals at different life stages, and it may be necessary to include other endpoints.
C.2. Alternative chronic toxicity testing protocols
As stated, a review of currently available EPA guideline toxicology studies (OPPTS 870
series) demonstrates there is no single protocol that addresses continuous exposure through all
life stages of any test species. To address this issue, two possible alternative study designs were
considered: the "expanded chronic/carcinogenicity study" and the "unified screening study."
These are described in some detail below and illustrated in accompanying figures. The intent of
this discussion is to demonstrate the advantages (and disadvantages) of exploring nontraditional
testing paradigms; however, such discussion does not comprise a recommendation for
3-35
-------
External Review Draft Do Not Cite or Quote
implementation. For many chemicals, the existence of adequate (by Agency standards) stand-
alone studies would preclude the need for further testing, with or without expanded or combined
protocols such as those described below. In any case, any proposal to use alternative study
designs in a regulatory setting should be thoroughly discussed by Agency and Registrant
scientists prior to study initiation.
3-36
-------
External Review Draft
Do Not Cite or Quote
Figure 3-6. Alternative Acute Toxicity Protocol
Acclimatization
c
A
Body weight,
Food consumption
Body weight,
Food consumption
Body weight,
Food consumption
Study day
1234567 8 9 10 11 12 13 14 15
t
Dosing
Gross necropsy,
organ weight,
histopathology,
hematology,
clinical chemistry
Gross necropsy,
organ weight,
histopathology,
hematology,
clinical chemistry
-------
External Review Draft Do Not Cite or Quote
C.2.a. The expanded chronic/carcinogenicity study
An example of a study design that would incorporate lifetime (in utero through old age)
exposure is the expanded chronic/carcinogenicity study (shown in Figure 3-7). This study could
serve as a replacement for a standard guideline chronic/carcinogenicity study in rats.
In this study, female rats are assigned to test groups, mated, and treated with test
substance throughout gestation and lactation. When pups are weaned on PND 21, they are
assigned individual animal numbers within their established test group. The early exposure to
test substance in this study is similar to that required for the in utero carcinogenicity study that is
used to evaluate food additive chemicals for regulation by the Food and Drug Administration
(FDA) Center for Food Safety and Nutrition (CFSAN). In the study design discussed here,
however, the study duration is also extended to a period of 3 years (versus a typical chronic
duration of 2 years for rats), with interim sacrifices scheduled at yearly intervals. The total
number of animals used in this expanded study is greater than the number used for a standard
guideline chronic/carcinogenicity study because of the additional interim sacrifice; for each
annual segment, the sacrifice of 25 rats/sex/group is required. To reduce this number, the study
could be conducted with fewer animals per segment (e.g., 20/sex/group), or only two sacrifices
could be scheduled (i.e., at 1.5 and 3 years). Of course, such actions will either reduce the power
of the evaluation for tumor data or will eliminate examination of an important life phase.
Parameters typical to a guideline chronic/carcinogenicity study are examined in this
expanded study (e.g., mortality, clinical observations, body weight, food consumption, clinical
chemistry and hematology, ophthalmology, gross pathology, and histopathology). In addition,
neurological and immunological evaluations are performed in the adult animals at multiple
intervals into old age, which, along with the fact that the animals are exposed to the chemical
during all life stages, contributes to the superiority of this study design.
Although the temporal linear nature of this study protocol makes it less complicated to
conduct in the laboratory, this attribute also results in the inability to easily assess some other
important endpoints within specific targeted organ systems, e.g., prenatal developmental
assessment, reproduction and endocrine function, and DNT. Additionally, by 3 years of age,
when this study would be terminated, survival in laboratory rats may be compromised; therefore,
it may be necessary to consider using feed restriction to maximize the number of animals
available for in vivo and post mortem assessment of aged animals. In addition, housing from
birth in specific-pathogen-free (SPF) facilities may be necessary to maintain sufficient viable
3-38
-------
External Review Draft Do Not Cite or Quote
animals for such an extended period of time (see the background white paper on aging, Versar
Inc., 2001a).
3-39
-------
External Review Draft
Do Not Cite or Quote
Figure 3-7. Expanded Chronic/Carcinogencity Study
tation, Lactatioi
Neurobehavior
lyr
Neurobehavior 2 yr
Neurobehavior
3yr
1 1 1
Clin. Path.
Wean
Assign Fl
SacP
Immunotox
screen
Clin. Path.
Interim sac 1
Immuno. screen
Clinical path.
Hstopathology
Neuropathology
Clin path.
Interim sac 2
Immuno. screen
Clinical path.
Hstopathology
Neuropathology
Terminal sac
Immuno. screen
Clinical path.
Hstopathology
Neuropathology
-------
External Review Draft Do Not Cite or Quote
C.2.b. The unified screening study
An alternative study design, the unified screening study, is illustrated in Figure 3-8. This
study is composed of at least four segments, each of which could be conducted as a separate
study: two-generation reproduction, expanded chronic/carcinogenicity, developmental toxicity,
and DNT studies. Additionally, an optional continuous-breeding reproduction study could be
added to the design. When conducted in the rat, the unified screening study assesses all life
stages of the animals and provides a means to evaluate prenatal developmental toxicity, DNT,
reproduction, and endocrine function, all within animal subjects that are derived from the same
gene pool and are evaluated within two generations of the progenitor rodents that are initially
placed on study.
The unified screening study begins as a typical two-generation reproduction study, with
10 weeks of treatment, mating, gestation, and lactation phases being conducted according to
OPPTS 870.3800. The Fl weanlings are selected for either the second generation of the
reproduction study or the expanded chronic/carcinogenicity study. (As a point of clarification, at
any point that animals are selected and/or assigned to a different study phase, it is assumed that
the treatment group remains constant for each animal.) The parental (P) animals from the first
generation are not immediately terminated; rather, they are transferred to a prenatal
developmental study phase. After a short rest, they are mated. The P males can be terminated at
any time point, and the P females are continued through to caesarian section on approximately
GD 20. The resulting Fib fetuses are processed and examined for external, soft tissue, and
skeletal abnormalities, as is typical to an OPPTS 870.3700 study. At necropsy, however, the P
generation animals receive an extended postmortem examination, according to the procedures for
the two-generation reproduction study, that includes sperm measures for the males and extensive
histopathology of the reproductive and other organ systems for both sexes.
The expanded chronic/carcinogenicity study, using Fl animals, would continue as
described above concurrently with all other phases of the unified screening study but continuing
well past the time that the others have been terminated. The other Fl pups, which are selected as
second generation P animals for the two-generation reproduction phase, are treated for 10 weeks
and then undergo the standard reproductive functional assessments as specified in the OPPTS
870.3800 guideline. Because a number of F2 pups from this generation will continue on into the
DNT study phase, some additional observations are required during the lactation segment of the
second generation. Specifically, F2 pups are selected and assigned for neurobehavioral
assessments on PND 4 (at the time of litter standardization). Preweaning observations include
3-41
-------
External Review Draft
Do Not Cite or Quote
Figure 3-8. Unified Screening Studya
OJ
to
lyr
Chronic/Carcinogenicity Study Segment
2yr
Gestatior
Term sac F2
.
.
Mate C-section
Histopathology
Fib fetal evaluations
Prenatal Developmental
Toxicity Study Segment
...
'
Developmental Neurotoxicity
Study Segment
.,••••"
Gestation
Gestationl iGestationl iGestationl [Gestation! Lactation C^ A "^
cfP
Mate Fl M & F - continuous cohabitation
Syr
Initiate
treatment
10 weeks
Qua
(P Generation)
h
late
Gestation
;
A
Lactation
..P
^ (F 1 Generation)
Interim sac 1
^ 8t Reproduction ;
i\<§^^ Study Se
10 weeks Gestation
F 1 (F 1 Generation)
Mate
Interim sac 2
Termir
md Fertility
*ment
\\
Lactation ^^% p>Tp ^
..••'' 1 Neuro.behavioral assessments
...-••"' g^ (F2 Generation)
3Studv lines are not drawn to scale
Optional Continuous Breeding Study Segment
-------
External Review Draft Do Not Cite or Quote
weekly age-appropriate clinical/functional behavioral observations conducted outside of the
home cage and motor activity assessments on PND 13 and 17. Additional assessments of
physical, reflex, and sensory development may also be conducted during this period.
At the time of weaning of the second generation at PND 21, those pups preselected for
neurobehavioral assessment continue into the DNT phase, while other weanlings are sacrificed
for postmortem evaluations that address the considerations of both the reproduction protocol
(including organ weight data) and the DNT protocol (requiring in situ perfusion fixation of
tissues and neuropathology, including morphometric analysis). The DNT phase F2 animals are
evaluated as per OPPTS 870.6300, which includes multiple assessments of clinical and
functional observations, motor activity, auditory startle habituation, and learning and memory.
They are maintained until termination (with postmortem evaluations, including neuropathology
following perfusion fixation) at approximately PND 60.
Also at the time of weaning of the second generation (F2) pups, a decision could be made
to either sacrifice the Fl P animals immediately (with the usual sperm measures and postmortem
evaluations) or to maintain them through a continuous breeding reproduction study phase,
sequentially mating the Fl adults for the production of five litters (the pups from which are
terminated in early lactation). This continuous-breeding study phase, which would extend the
reproduction study for about 100 additional days, uses a standardized assessment protocol that
has been well characterized in the peer-reviewed literature (Lamb, 1985; Lamb et al., 1985;
Morrissey et al., 1989) but does not have a corresponding OPPTS guideline.
As previously stated, in this unified study protocol, the animals are both exposed and
assessed during all life stages, and the evaluation of both structural and functional endpoints for
multiple organ systems are maximized in the overall design, e.g., by the inclusion of
immunotoxicity and neurotoxicity endpoints. There is one notable exception to this statement in
that reproductive senescence is not standardly examined. Nevertheless, if the two-generation
reproduction study phase identifies problems with fertility or cyclicity, this could be pursued
more rigorously by the addition of testing during the second or third year of the expanded
chronic/carcinogenicity study, e.g., evaluating cyclicity in aged female rats and/or evaluating
ovarian follicular counts and atrophy at sacrifice.
Another benefit of using the unified screening study design is that it results in the
purchase and use of many fewer naive animals for study initiation, and it increases the efficient
utilization of animals, particularly of the second generation (F2) offspring from the reproduction
study.
3-43
-------
External Review Draft Do Not Cite or Quote
Although there are obvious benefits in using a unified screening study, there are also a
number of concerns or potential problems involved with its conduct. Although it is assumed that
treatment levels and route of administration will remain constant across all study phases in the
unified screening study, this approach to dose-setting and route selection may not always be
optimal for every phase. Generally, a temporal nonlinear design of this nature makes such a
study more difficult to manage in the laboratory. The strain of rats generally used in toxicity
studies is the Sprague-Dawley, while the Fischer 344 rat is often used in the standard
chronic/carcinogenicity study. Fischer 344 rats have not typically been used in reproductive and
developmental toxicity studies. The use of either strain for the unified study could compromise
the use of historical data for comparison, e.g., for the chronic/carcinogenicity study if the
Sprague-Dawley is used and for the reproductive and developmental toxicity studies if the F344
is used.
As study complexity increases, so does the opportunity for error. In some cases, a serious
technical error in one study phase could compromise subsequent study phases and result in an
extensive waste of animals and resources. As with the stand-alone expanded
chronic/carcinogenicity study, survival during this study phase may need to be enhanced via feed
restriction. Also, if the test substance interferes with reproduction or results in increased
mortality, the number of offspring that are available for assignment to subsequent study phases
(e.g., the selection of F2 animals for the DNT phase) may be critically reduced. An additional
but similar problem could arise when selecting Fl animals for the expanded
chronic/carcinogenicity study at the same time as for the second generation of the reproduction
study, because there is the need for a large number of offspring to be available all at one time.
Additionally, the offspring that are assigned to the chronic/carcinogenicity segment should be
genetically diverse within each group and should originate from as many litters as possible (not
be siblings).
A number of possible solutions, which could be used alone or in combination to increase
the number of Fl pups available for selection in other study phases include the following:
1. Reducing the number of animals needed for the expanded chronic/carcinogenicity
segment by examining fewer animals at each serial sacrifice or by abandoning the final year of
evaluation, as described above.
2. Reducing the number of animals assigned to the second generation of the reproduction
study; however, this could compromise the number of Fl offspring that would be available for
the DNT study segment.
3-44
-------
External Review Draft Do Not Cite or Quote
3. Standardizing litters to 10 rather than 8 pups per sex and assuming that no litter has less
than 10 pups and that no pups die during lactation. Because some small litters and neonatal pup
deaths almost always occur, even in controls, it is wiser to design the study more conservatively
in order to avoid discovering that there are not enough Fl pups to assign to the next segment(s).
4. Assigning additional females to the two-generation reproduction study in order to
produce extra Fl pups for selection. Although even a modest increase in the number in each
group would increase the probability of producing a sufficient number of Fl pups, a larger
number of litters would generally be required in order to ensure genetic diversity among the
weanlings that are assigned to the chronic/carcinogenicity phase. This could be accomplished by
placing additional P generation females or breeding pairs on study, perhaps combined with 2:1
mating procedures, or by mating the males with the reproduction study segment females first and
then with an extra set of females. One adverse consequence of placing additional females on
study so that their litters can be used for selection of genetically diverse offspring for the
chronic/carcinogenicity study is that this method results in a larger number of excess Fl weanling
pups that would not be used for evaluations in this protocol. However, these pups could be used
for other evaluations, e.g., immunotoxicity, specialized neurotoxicity tests, or adult onset disease
or diseases of aging.
Some of the above options appear to be more advantageous and preferable than others;
however, no recommendation is proffered, because the list is presented only to illustrate some of
the many possibilities that could be used in a customized study design. It should be noted that
simply combining the two-generation study and the DNT study when a two-generation study is
not already available greatly reduces the total number of animals required as compared to
conducting the two studies individually. No additional animals are required over the two-
generation study alone, and there is greater efficiency in the use of the F2 offspring when the
DNT study is conducted in that group.
3-45
-------
External Review Draft Do Not Cite or Quote
CHAPTER 4
FRAMEWORK FOR SETTING ACUTE, SHORT-TERM, LONGER-TERM, AND
CHRONIC REFERENCE VALUES
As noted in Chapter 2, the Technical Panel is recommending that EPA begin deriving
acute, short-term, and longer-term reference values in addition to chronic reference values. The
approach to reference values discussed here is intended for use in risk assessments for health
effects known or assumed to be produced through a nonlinear and/or threshold mode of action.
Although there has been a dichotomy between cancer and noncancer risk assessment in terms of
the underlying assumption about the linearity or nonlinearity of the dose-response curve, there is a
move toward harmonization among approaches for all health effects (Bogdanffy et al., 2001).
This includes the possibility that some carcinogenic agents may work through nonlinear
mechanisms (EPA, 1999d), whereas some agents that produce other types of effects may work
through linear mechanisms (see discussion in EPA, 1998a). Thus, the decision to use a linear
extrapolation approach or a reference value approach should take into consideration the
underlying mode of action and presumed dose-response relationship.
The approach described here is the default approach to be used when the assumption is a
nonlinear and/or threshold mode of action except for cases where other methods have been
developed (e.g., in support of the NAAQS). This approach can and should be improved upon or
replaced when more specific data on pharmacokinetics and mode of action are available to allow
the development of a chemical-specific or a biologically-based dose-response model for prediction
of risks to humans, and to susceptible individuals within the population. The acute, short-term,
longer-term, and chronic reference values derived on the basis of the recommendations in this
report should be included in IRIS after appropriate internal, external, and consensus review.
These values would then be available for use by program offices, where appropriate.
In this chapter, we discuss the definitions of the exposure durations and the proposed
changes in the definitions of the corresponding reference values. In addition, several issues are
discussed regarding the adequacy of studies and characterization of the extent of the database
with regard to the sufficiency of the data for deriving reference values. The derivation of
reference values also is discussed with regard to dosimetric adjustment and application of UFs. A
number of recommendations are made with regard to this process. In particular, the Technical
4-1
-------
External Review Draft Do Not Cite or Quote
Panel recommends incorporation of the concept of life stage and expansion of the endpoints
evaluated as well as consideration of duration of exposure and latency to response in
characterizing the extent of the database used for setting reference values. The Technical Panel
strongly encourages the use of a narrative description of the database including strengths and
limitations, rather than a single confidence statement for support of a reference value.
The adjustments required for derivation of the HED for oral and dermal exposure and the
HEC for inhalation exposure are described and discussed. This is followed by recommendations
about the evaluation and comparison of data for the POD, based on an analysis of each potentially
limiting endpoint carried through the reference value derivation process, followed by selection of
the appropriate health-protective reference value.
Finally, the Technical Panel emphasizes that considerable use of scientific judgment is
advisable and necessary in practically all phases of the process, especially in the application of
UFs. This review and its recommendations build on the principles in the Agency's Handbook on
Risk Characterization (EPA, 2000a), which calls for transparency, clarity, consistency, and
reasonableness in the risk assessment process.
A. Definitions of Exposure Durations for Use in Setting Reference Values
The Technical Panel proposes the following definitions of exposure duration as a first step
in the development of consistent approaches for the Agency. These definitions are based on
exposure durations for humans; analogous exposure durations for rodents are indicated for the
longer-term and chronic durations. The definitions are not intended to be rigid specifications, but
simply general descriptions of the relevant exposure time period.
The definitions were developed on the basis of the review of values currently set by
various program offices (see Chapter 2), and they have been standardized to be compatible with
those definitions currently used by various program offices within the Agency. The definitions for
various durations were discussed at an EPA Risk Assessment Forum Colloquium (The CDM
Group Inc, 2000).
Acute: Exposure by the oral, dermal, or inhalation route for 24 hours or less
Short term: Repeated exposure1 by the oral, dermal, or inhalation route for more than 24
JA repeated exposure may be either continuous, periodic, or intermittent. A continuous
exposure is a daily exposure for the total duration of interest. A periodic exposure is one
occurring at regular intervals, e.g., inhalation exposure 6 hours/day, 5 days/week or oral exposure
4-2
-------
External Review Draft
Do Not Cite or Quote
hours, up to 30 days.
Longer term: Repeated exposure by the oral, the dermal, or the inhalation route for
more than 30 days, up to approximately 10% of the life span in humans2 (more than 30 days up to
approximately 90 days in typically used laboratory animal species3).
Chronic: Repeated exposure by the oral, dermal, or inhalation route for more than
approximately 10% of the life span in humans (more than approximately 90 days to 2 years in
typically used laboratory animal species).
B. Proposed Changes in the Reference
Value Definitions
In the process of considering
definitions for different duration reference
values, the Technical Panel discussed
several issues that have been raised about
the current definitions of the chronic RfD
and RfC (see Box 4-1). The following
items describe the issues and the
recommended changes.
1. The parenthetical statement in the
current RfD and RfC definitions - "with
uncertainty spanning perhaps an order of
magnitude" - has been variously
interpreted by risk assessors and risk
BOX 4-1. Current Definitions for the Chronic Oral
RfD and Inhalation RfC
RfD: An estimate (with uncertainty spanning perhaps an
order of magnitude) of a daily oral exposure to the human
population (including sensitive subgroups) that is likely to
be without appreciable risk of deleterious effects during a
lifetime. It can be derived from a NOAEL, LOAEL, or
BMD, with UFs generally applied to reflect limitations of
the data used. Generally used in EPA's noncancer health
assessments.
RfC: An estimate (with uncertainty spanning perhaps an
order of magnitude) of a continuous inhalation exposure to
the human population (including sensitive subgroups) that
is likely to be without appreciable risk of deleterious effects
during a lifetime. It can be derived from a NOAEL,
LOAEL, or BMD, with UFs generally applied to reflect
limitations of the data used. Generally used in EPA's
noncancer health assessments.
5 days/week. An intermittent exposure is one in which there is no effect of one exposure on the
effect of the next; this definition implies sufficient time for the chemical and its metabolites to
clear the biological system before the subsequent exposure, that is, non-cumulative
pharmacokinetics. A periodic exposure may or may not be intermittent.
2The lifespan value used depends on the situation under consideration. For example, an
average of 70 years has been the typical default used for chronic exposures, but the average life
span based on US census data is 75.5 years (EPA, 1997a).
3Typically-used laboratory animal species refers to rats, mice, and rabbits, for example.
4-3
-------
External Review Draft Do Not Cite or Quote
BOX 4-2. Proposed Revisions in the Reference Value Definitions
Acute [Oral, Dermal, Inhalation] Reference Value: An estimate of an exposure for 24 hours or less to the
human population that is likely to be without an appreciable risk of adverse effects for a lifetime (including
susceptible subgroups3). It can be derived from a BMD, a NOAEL or a LOAEL, with uncertainty/variability15
factors generally applied to reflect limitations of the data used. The application of these factors is intended to
provide an estimate centered within an order of magnitude.
Short-Term [Oral, Dermal, or Inhalation] Reference Value: An estimate of an exposure for up to 30 days to
the human population that is likely to be without an appreciable risk of adverse effects for a lifetime (including
susceptible subgroups). It can be derived from a BMD, a NOAEL or a LOAEL, with uncertainty/variability factors
generally applied to reflect limitations of the data used. The application of these factors is intended to provide an
estimate centered within an order of magnitude.
Longer-term [Oral, Dermal, or Inhalation] Reference Value: An estimate of an exposure for up to
approximately 7 years (10% of the average life span) to the human population that is likely to be without an
appreciable risk of adverse effects for a lifetime (including susceptible subgroups). It can be derived from a BMD,
a NOAEL or a LOAEL, with uncertainty/variability factors generally applied to reflect limitations of the data used.
The application of these factors is intended to provide an estimate centered within an order of magnitude.
Chronic [Oral, Dermal, or Inhalation] Reference Value: An estimate of an exposure for up to the average life
span of the human population that is likely to be without an appreciable risk of adverse effects for a lifetime
(including susceptible subgroups). It can be derived from a BMD, a NOAEL or a LOAEL, with
uncertainty/variability factors generally applied to reflect limitations of the data used. The application of these
factors is intended to provide an estimate centered within an order of magnitude.
Susceptible subgroups may refer to life stages, e.g., children or the elderly, or to other segments of the population,
e.g., asthmatics or the immune-compromised, but they are likely to be somewhat chemical-specific, and may not
be consistently defined in all cases. See below (Section C.2.c) for further discussion.
bSee discussion later in this chapter (Section D.5) on application of uncertainty/variability factors.
managers to mean that the estimate is at the upper end, the lower end, or in the middle of the
range of an order of magnitude. In an attempt to be clearer, the revised definitions (see Box 4-2)
have been reworded to indicate that the reference value is intended to provide an estimate
centered within an order of magnitude, further emphasizing that the estimate is not a bright line
but has some range of variability that may be considered by risk managers in decision making.
2. The term "deleterious" is considered ambiguous by some, so it has been replaced with the
term "adverse," because the latter is more commonly understood in the context of data evaluation
and selection of endpoints for setting reference values.
3. In the spirit of harmonization of risk assessment approaches for human health effects, it
has been recommended that health effects no longer be categorized as "cancer" or "noncancer"
for the purposes of hazard characterization and dose-response analysis (EPA, 1997c, 1998a;
Bogdanffy et al., 2001). As indicated earlier, the approach to reference values discussed here is
4-4
-------
External Review Draft Do Not Cite or Quote
intended for risk assessments for any type of health effect known or assumed to be produced
through a nonlinear and/or threshold mode of action (which may include U-shaped or other
nonmonotonic dose-response curves as well as thresholds). In light of this recommendation, the
term "noncancer" has been removed from the definitions, denoting the move toward defining
approaches for low-dose estimation or extrapolation based on mode of action. It is recommended
that this issue be considered further in the deliberations by the Risk Assessment Forum's
Technical Panel on a Framework for Harmonization of Approaches for Human Health Risk
Assessment.
To fulfill the need for consistency in the designation of various duration reference values,
the Panel recommends that the terminology for reference values be standardized. Rather than
continuing to use RfD and RfC only to denote chronic oral and inhalation reference values,
respectively, standardized terminology should be developed that denotes both duration and route
of exposure. Although Technical Panel members did not come to agreement on the best way to
do this (and we welcome alternative suggestions), the terminology shown below is offered as an
example of the way in which consistent labels could be developed and used. Either new standard
terminology, e.g., reference value could be used, or RfD and RfC could continue to be used, but
these would always need to be accompanied by the qualifying duration of exposure and, in the
case of the RfD, by the route of exposure. Thus, the following alternatives for terminology are
offered.
Acute [Oral, Dermal] Reference Value or Dose, Acute [Inhalation] Reference
Value or Concentration:
RfVAO, RfVAu, RfVAI; RfDAO, RfD^, RfCAI or RfCA
Short-term [Oral, Dermal] Reference Value or Dose; Short-term [Inhalation]
Reference Value or Concentration:
RfVso, RfVSD, RfVSI; RfDso, RfDSD, RfCSI or RfCs
Longer-term [Oral, Dermal] Reference Value or Dose; Longer-term [Inhalation]
Reference Value or Concentration:
RfVLO, RfVLD, RfVLI; RfDLO, RfDLD, RfCLI or RfCL
Chronic [Oral, Dermal] Reference Value or Dose; Chronic [Inhalation] Reference
Value or Concentration:
RfVco, RfVCD, RfVCI; RfDco, RfDCD> RfQi or RfCc
4-5
-------
External Review Draft Do Not Cite or Quote
The Panel recommends that endpoint or life-stage specific reference values such as the
RfDDT (Reference Dose for Developmental Toxicity), which were originally proposed in the
Guidelines for Developmental Toxicity Risk Assessment (EPA, 1991), not be derived. Reference
values are intended to protect the population as a whole, including potentially susceptible
subgroups, and the RfDDT concept of a critical window of exposure for some health effects is
addressed in the adoption of the less-than-chronic reference values. Thus, it is recommended that
the RfDDT not be used because developmental toxicity endpoints would be considered along with
other relevant endpoints in the derivation of most, if not all, of the reference values suggested
here. This does not preclude, however, using specific common endpoints in the assessment of
cumulative risk for chemicals with a common mode of action or for risk management purposes.
C. Characterization of the Extent of the Health-Related Database for Setting Reference
Values
A necessary first step in hazard characterization is the critical evaluation of all pertinent
and relevant human and animal data that are available in the open literature as well as data
submitted to the Agency in response to various regulatory standards, data call-ins, or other
requirements and agreements.
C.I. Review of studies
Data will be available from a wide variety of sources, including studies conducted
according to EPA guidelines, studies conducted by industry using OECD or other protocols,
experimental studies conducted by academic researchers, epidemiology studies, case reports or
series, or controlled clinical studies in volunteers.4 These studies will be of widely differing
quality. EPA must evaluate each study to determine whether it is of acceptable quality.
C.l.a. Adequacy of studies
The following list of questions could be helpful in the process of evaluating data from
animal and human studies.
4Currently, OPP is reviewing its policy concerning use of human data from studies in
which there is intentional pesticide exposure, and it has asked the National Academy of Sciences
for input on the acceptability of such studies and ethical criteria for their use under the Protection
of Human Subjects Rule (the "Common Rule") (EPA, 2001b).
4-6
-------
External Review Draft Do Not Cite or Quote
All types of studies:
What was the purpose of the study and is there a clearly delineated hypothesis?
• Is there sufficient description of the protocol, statistical analyses, and results to make an
evaluation?
Were the appropriate endpoints assessed in the study?5 Were the techniques used for the
assessment scientifically sound?
• Were appropriate statistical techniques applied for each endpoint? Was the power of the
study adequate to detect effects?\
Did the study establish dose-response relationships? Was a BMD lower confidence level
(BMDL), LOAEL orNOAEL established?
• Is the shape of the dose-response curve consistent with the known pharmacokinetics of
the test compound?
• Do effects fit with what is known about mode of action?
Is the dose-response curve for precursor events consistent with the dose-response curve
for clinical effects?
• Are the results of the study biologically plausible?
What uncertainties exist? Do the results of the study indicate the need for follow-up
studies to reduce uncertainties?
• Are the study conclusions supported by the data?
Human studies:
What were the data sources for exposure, health status, and risk factors (e.g.,
questionnaires, biological measurements, exposure/work history record reviews, or
exposure/disease registries) and what were their strengths and limitations?
What methods were used to control, measure, or reduce various forms of error (e.g.,
5 A chemical may cause a variety of toxic effects depending on the amount, duration,
timing, and pattern of exposure (i.e., continuous, periodic, or intermittent). These effects may
range from severe effects, such as death, to more subtle biochemical, physiological, or
pathological changes in one or more organ systems. In addition, the effects will vary depending
on their latency following exposure and when the observations are made. Primary attention is
given in risk assessment to those effects in the lower exposure range and/or the effects most
biologically appropriate for a human health risk assessment.
4-7
-------
External Review Draft Do Not Cite or Quote
misclassification or interviewer bias, confounding factors and potential effect modifiers)
and their potential impact on the findings? What is the validity (accuracy) and reliability
(reproducibility) of the methods used to determine exposure and outcome? What were the
response rates?
What major demographic and other personal factors were examined, e.g., age, sex, ethnic
group, socioeconomic status, smoking status, and occupational exposure? What other
climate, or life stage factors were important for the endpoints and exposures assessed?
• Were the findings examined for biologic plausibility, internal and external consistency of
the findings, and the influence of limitations of the design, data sources, and analytic
methods?
Animal studies:
• Was the study sufficiently documented (e.g., conducted in accordance with good
laboratory practices [GLPs])?
Were appropriate analytical techniques used to measure the stability, homogeneity, and
actual level of the test substance in the study (in the water, feed, air, etc.)?
• Was an appropriate animal species used?6 Was an appropriate number of animals used?
Both sexes? Age?
Were the dose levels appropriate? What was the basis for choosing the dose levels?
• Was an appropriate method used to assign the animals to dose groups?
• Was an appropriate route and matrix of exposure employed?7
6The laboratory animals used most often are the rat, mouse, rabbit, guinea pig, hamster,
dog, or monkey. When reviewing these studies, the risk assessor makes judgments about the
ability of the study to predict the potential for toxicity in humans and tries to select data from the
species that is most relevant to humans using the most defensible biological rationale. When
available, comparative pharmacokinetics can be used to support this decision. Absent a clearly
most-relevant species, the most sensitive mammalian species is used, i.e., the species that shows
toxicity at the lowest exposure level.
7The most appropriate route of exposure is the route for which an evaluation is to be
made. The toxicity of the chemical may differ with route of exposure because of differences in
mechanism of action or pharmacokinetics (absorption, distribution, metabolism, and excretion).
Development of data to establish dosimetry for the purpose of route-to-route extrapolation is
encouraged; however, route-to-route extrapolation is inappropriate when based exclusively upon
default assumptions regarding exposure and pharmacokinetics. Even within the same route of
4-8
-------
External Review Draft Do Not Cite or Quote
Was the duration of exposure adequate for the particular study design?
• Were possible alterations in metabolism considered at the higher exposure levels?
Professional judgment is required to decide, on the basis of a thorough review of all
available data and studies, whether any observed effect is adverse and how the results fit with
what is known about the underlying mode of action. These judgments require the input of
experts trained in toxicology, statistics, epidemiology and, often, of specialists in the structure and
function of the target organ systems. Both the biological and the statistical significance of the
effects are considered when making these judgments. Biological significance is the determination
that the observed effect (a biochemical change, a functional impairment, or a pathological lesion)
is likely to impair the performance or reduce the ability of an individual to function or to respond
to additional challenge from the agent. Biological significance is also attributed to effects that are
consistent with steps in a known mode of action. Statistical significance quantifies the likelihood
that the observed effect is not due to chance alone. Precedence is given to biological significance,
and a statistically significant change that lacks biological significance is not considered an adverse
response.
For many discrete or quantal endpoints, (e.g., birth defects, tumors, or some discrete
pathological changes), this judgment is more straightforward because criteria have been
established for deciding what type and incidence of effects are to be considered to be adverse, and
an increase above the background rate can be judged using statistical tools. In the case of
continuous measures (e.g., body weight, enzyme changes, physiological measures), this tends to
be more difficult, because the amount of change to be considered adverse has not been defined by
toxicologists or health scientists. Consequently, the endpoint is often decided in the context of
the endpoint itself, the study, and the relationship of changes in that endpoint to other effects of
the agent. Decisions about the amount of change to consider adverse must always be made using
professional judgment and must be viewed in light of all the data available on the endpoint of
concern. All toxicological data on a chemical must be reviewed before deciding whether an effect
is biologically significant and adverse. Using a default cutoff value to define adversity for
continuous measures may result in an inappropriate interpretation of data and less than optimum
exposure, responses may differ due to alterations in pharmacokinetics, e.g., dietary or water
exposure versus oral gavage.
4-9
-------
External Review Draft Do Not Cite or Quote
evaluation of a chemical's effects.
C.2. Issues to be considered in the characterization of the database for risk
assessment
C.2.a. The weight-of-evidence approach
A weight-of-evidence approach should be used in assessing the database for an agent
(e.g., EPA's RfC Methodology [1994]; EPA's Proposed Guidelines for Carcinogen Risk
Assessment [1999d]). This approach requires a critical evaluation of the entire body of available
data for consistency and biological plausibility. Potentially relevant studies should be judged for
quality and studies of high quality given much more weight than those of lower quality. When
both epidemiological and experimental data are available, similarity of effects between humans
and animals is given more weight. If the mechanism or mode of action is well characterized, this
information is used in the interpretation of observed effects in either human or animal studies.
"Weight-of-evidence" is not to be interpreted as simply tallying the number of positive and
negative studies, nor does it imply an averaging of the doses or exposures identified in individual
studies that may be suitable as PODs for risk assessment. The study or studies used for the POD
are identified by an informed and expert evaluation of all the available evidence.
C.2.b. Use of human and animal data in risk assessment
Adequate human data are the most relevant for assessing risks to humans. When sufficient
human data are available to describe the exposure-response relationship for an adverse
outcome(s) that is judged to be the most sensitive effect(s), reference values should be based on
human data. Much more data on a wide range of endpoints typically are required to establish
confidence that there are no effects of exposure. If sufficient human data are not available to
provide the basis for reference values, data from animal studies must be employed. It is
advantageous if some human data are available to compare with effects observed in animals, even
if the human data are not adequate for quantitative analysis. Availability of data on effects in
humans at least allows qualitative comparison with effects observed in animals for determining
whether toxicity occurs in the same organ systems and whether the nature of the effects is similar
or different. If no human data are available, reliance must be exclusively on animal data. In that
case, attention should be paid to whether data are available in more than one species, and if so,
whether the same or similar effects occur in different species and possible sources of any observed
4-10
-------
External Review Draft Do Not Cite or Quote
differences. EPA's risk assessment guidelines include as one of the major default assumptions
that animal data are relevant for humans (e.g., EPA, 1991, 1996, 1998c). Such defaults are
intended to be used in the absence of experimental data that can provide direct information on the
relevance of animal data.
Several types of information should be considered when determining the relevance or non-
relevance of effects observed in animal models for humans. This information is used in a variety
of ways, from determining the role of metabolism in toxicity (is the parent chemical or a
metabolite responsible for toxicity?), to assessing whether homologous activity would be expected
across species (do humans share the sensitivity of the animal model, or is the response due to
some species-specific idiosyncratic reaction?), to determining whether or not a threshold is likely
to exist for the response (are repair mechanisms capable of maintaining a homeostatic process?).
All of this information must be weighed in light of the known heterogeneity of the human
population versus the relatively inbred status of laboratory animals used in toxicity testing studies
and housed under carefully controlled environmental conditions.
Table 4-1 presents several factors to consider when evaluating the weight of evidence
about the likelihood of the occurrence of effects in humans that is based on animal data (in
conjunction with human data, if available). The table is not necessarily intended to delineate all
factors that may need to be considered, but rather to provide a framework for evaluation and
interpretation. It is important to evaluate the database in a holistic manner, determining strengths
and weaknesses that are relevant to the overall assessment. Each chemical and database presents
a unique set of issues that must be evaluated critically and thoughtfully.
The dose-response nature of the data is an important characteristic of the data base or
individual study. When data are dose-related, that is, when the incidence and/or intensity of
response changes in an orderly manner as a function of dose, the effect should be considered to be
of greater importance than when there is no apparent association between exposure and toxicity.
Note, however, that the dose-response relationship need not be monotonic. U-shaped (or
inverted-U-shaped) dose-response functions are not uncommon in toxicology. For example, a
chemical may induce an enzyme at low doses and inhibit it at high doses. Similarly, many solvent-
like chemicals (including alcohol) produce increased motor activity at lower doses and depressed
activity at high doses.
Similarly, comparative pharmacokinetic/metabolism data that suggest qualitative and
quantitative comparability to that in humans would support the relevancy of animal data.
4-11
-------
External Review Draft Do Not Cite or Quote
Evidence suggesting a difference in pharmacokinetics/metabolism would require additional
exploration regarding whether the difference(s) results in a major qualitative or quantitative
difference in internal dose in humans.
The similarity of effects between species is also an important aspect that needs to be
considered in the characterization of the database. Similar effects in more than one species
indicate that the effect provides increased weight of evidence for the risk assessment process,
even if such data are not available in humans. In contrast, response data that show inconsistency
of effects among studies and/or species that cannot be explained by differences in
pharmacokinetics/metabolism, or timing and/or magnitude of exposure, may suggest that less
emphasis be placed on the effect. "Similarity" does not necessarily require identical effects
between species. For example, changes in motor activity in animals evaluated in the neurotoxicity
screening test and cognitive effects in humans would generally be considered similar, since both
are indicative of changes in nervous system function.
Mode of action information is also important in understanding whether a particular effect
may be important for humans. For example, a transient reduction in anogenital distance in the
postnatal animal following perinatal exposure to an anti-androgen has increased weight if the
chemical is also known to act as an anti-androgen in humans. Likewise, the interpretation of
increased skeletal variants observed following exposure to many chemicals would be enhanced by
data indicating that the mechanistic pathways for these agents and the overall biological
significance defined were also a possibility in humans. Mode-of-action data are also important in
determining whether various chemicals work by common modes or mechanisms of action which
would then be considered in a cumulative risk assessment.
4-12
-------
External Review Draft
Do Not Cite or Quote
Table 4-1. Factors for evaluation of the weight of evidence regarding the likelihood of
effects in humans.
Factor
Dose-response
relationship
Pharmacokinetics /
metabolism
Similarity of effects
Mode of action
Temporal relationship
Increased weight
Orderly change in effect as a function
of exposure (need not be monotonic)
Qualitative and quantitative
comparability between humans and
animals
Similar effects in more than one
animal species, or in animals and
humans
Demonstration of homologous mode
of action in animal model and
humans
Consistent temporal relationship
between exposure and effect
Decreased weight
No identified relationship between
exposure and magnitude of effect
Qualitative and quantitative differences
between humans and animals
Inconsistency of effects among studies
and/or species that cannot be explained by
differences in timing and/or magnitude of
exposure, or pharmacokinetics/metabolism
Evidence suggesting the mode of action is
species-specific and irrelevant to humans
Lack of temporality between exposure and
effect
Another criterion that is important in evaluating data is the temporal relationship
between exposure and effect. The exposure should precede the effect at an interval that is
consistent with what is known about the toxicokinetics and mode of action of the agent. It may
be the case, however, that higher doses produce a shorter latency to effect than do lower doses.
C.2.c. Characterization of effects in potentially susceptible subpopulations
A dose-response analysis for potentially susceptible subpopulations should be done as part
of the overall dose-response analysis for health effects in general. "Susceptible" in this context
means a differential (greater) response at the same internal dose in a particular segment of the
population due to intrinsic (possibly unknown) factors. "Susceptible subpopulations" is used here
to refer both to life stages and to other factors that may predispose individuals to greater response
to an exposure. Life stages may include the developing individual before and after birth up to
maturity (e.g., embryo, fetus, young child, adolescent), adults, or aging individuals. Other
4-13
-------
External Review Draft Do Not Cite or Quote
susceptible subpopulations may include people with specific genetic polymorphisms that render
them more vulnerable to a specific agent, or people with specific diseases or pre-existing
conditions (e.g., asthmatics). The term may also refer to gender differences, lifestyle choices, or
nutritional state. It is important to recognize that little basis currently exists for a priori
identification of susceptible subpopulations for many chemicals. Without other data to raise
suspicions, only the evaluation of effects in various segments of the population, such as those
mentioned above, can identify susceptible subpopulations for a particular chemical and a particular
set of exposure conditions.
In some situations, differential exposure rather than differential susceptibility per se may
be the critical issue (e.g., hand-to-mouth activity in toddlers). Economic differences may also
result in differential exposure and susceptibility.
A great deal of focus has been given in recent years to the issue of children as a
susceptible subpopulation. Several approaches have been proposed for characterizing the
database concerning the potential pre- and postnatal toxicity of a particular chemical and
providing some guidance as to the weight of evidence or degree of concern for children's health.
However, each approach has been developed for a slightly different purpose and, as such, is
generally complementary to the other approaches but not the same. The EPA developmental
toxicity (1991) and reproductive toxicity (1996) risk assessment guidelines describe an approach
that characterizes the database as sufficient or insufficient to judge whether a chemical does or
does not pose a hazard within the context of dose, route, duration, and timing of exposure. The
International Programme on Chemical Safety (IPCS) (1995) proposed an approach that was
based on the quality of information gathered in developmental and reproductive toxicity studies,
and what types of data were not available from these studies. The EPA draft 10X toxicology
report (EPA, 1999a) further extended the recommendations for characterizing risks to children's
health within the context of the FQPA by discussing issues that would increase or decrease the
level of concern.
The present report endorses and extends the recommendations of the 10X Toxicology
Working Group's report by incorporating the issues dealing with level of concern into a
framework for evaluating the evidence regarding the identification and characterization of
susceptible subpopulations (see below). A workshop was held recently to discuss aspects of a
Framework for Children's Health Risk Assessment and to emphasize a broader perspective on the
issues that should be considered in hazard characterization, dose-response assessment, exposure
4-14
-------
External Review Draft Do Not Cite or Quote
assessment, and risk characterization for children as a susceptible subpopulation (ILSI RSI, 2001)
In contrast to the attention paid to children and to asthmatics as potentially susceptible
subpopulations in recent years, little attention has been focused on risk assessment for other
potentially susceptible subgroups. As outlined in Chapter 3, there currently are no requirements
in EPA animal study protocols for exposure during old age or for outcome evaluations near the
end of the life span following earlier lifestage exposures. Similarly, healthy animals that are more
genetically homogeneous than humans are used in standard toxicity testing protocols, and
information on pre-existing conditions or genetic polymorphisms is largely unavailable from
animal studies.
Human studies also usually employ healthy nonelderly individuals, although some studies
in more susceptible populations have been conducted, e.g., study of the effects of air pollutants in
asthmatics. Individuals with identified risk factors that are not the focus of a study are usually
excluded from the study sample. It is important to consider such characteristics of the database if
human data are used as the basis for the risk assessment.
As can be seen in Table 4-2, several issues must be considered in assessing the potential
for some subpopulations, including different life stages, to have greater susceptibility to a
chemical than others. These include the timing (life-stage)-response relationship indicating
greater susceptibility to exposure at some life stages than others, whether effects are identified
that are of a different type in identifiable subgroups of the population as well as the dose-
response relationship, i.e., whether effects are observed at different levels of exposure in
different subpopulations. Another important consideration is whether effects are observed at the
same dose but with a shorter latency in different subpopulations. Additionally, differences among
groups in terms of the severity of the effect and reversibility must be considered. For example,
an agent may produce relatively mild and reversible neurological effects in adults, but produce
permanent behavioral impairment following in utero exposure. It is also important to keep in
mind that effects that may initially appear to be reversible may re-appear later or be predictive of
later adverse outcomes. This is probably best exemplified by certain outcomes following a
developmental exposure; e.g., an initial depression in birth weight or weight gain or subtle
developmental retardation may be indicators of more serious abnormalities later in life.
4-15
-------
External Review Draft
Do Not Cite or Quote
Table 4-2. Factors for evaluation of evidence regarding identification and
characterization of susceptible subpopulations"
Factor
Timing (life-stage)
response relationship
Type of effect
Dose-response
relationship
Latency of effect
Seriousness/
reversibility of effects
Increased weight
Effects occur at greater magnitude at
one or more life stage(s)
Different types of effects in specific
subpopulations
Effect occurs at lower exposures in one
(or more) subpopulation(s)
Latency to observed effect different in
specific subpopulations
Effects different in seriousness or
degree of reversibility in specific
subpopulations, and/or differences in
later consequence of an initially
reversible effect
Decreased weight
No difference in effects at different life
stage(s)
Same effect(s) across all potential
subpopulations
No evidence for differential dose-
response across different subpopulations
No difference between subpopulations in
latency to effect
No differences between subpopulations
in seriousness and/or reversibility of
effects, or in later consequences of an
initially reversible effect
a Subpopulations may be defined by gender, individuals at different life stages (fetus, child, adult, elderly),
differences in genetic polymorphisms, and/or pre-existing diseases or conditions that may result in differential
sensitivity to adverse effects from exposure to a specific toxic agent.
C.3. Characterization of the extent of the database
The derivation of an RfD or an RfC is a multifaceted process that involves the
coordination of data gathering and evaluation, analysis and judgment in varying proportions, and
integration of all the information available. A vital part of the chronic RfD and RfC derivation
process that relies heavily on judgment, for example, is the current approach to characterizing the
database. For example, the minimum dataset for low-confidence and high-confidence RfDs and
RfCs has been specifically defined (EPA, 1994, 2001b) as follows: minimum dataset for a low
confidence chronic RfD or RfC is a single subchronic study. The minimum dataset for a high
confidence chronic RfD or RfC is a chronic study in two species, a single two-generation
reproductive toxicity study, and a developmental toxicity study in two species by the appropriate
route of exposure.
The Technical Panel is recommending a somewhat different approach. Instead of
specifying particular studies, this approach emphasizes the types of data needed (in terms of both
4-16
-------
External Review Draft
Do Not Cite or Quote
human and animal data) for deriving reference values and recommends the use of a narrative
description of the extent of the database rather than a single confidence statement. The Technical
Panel believes that this approach encourages the use of a wider range of information to be used in
deriving reference values, that take into consideration the issues of duration and route of
exposure, the timing of exposures, the types and extent of endpoint assessment (i.e., structural
and function), the susceptible subpopulations evaluated, and the potential for latent effects and/or
reversibility of effects. In addition, this approach encourages the identification of the data that
would be needed or useful for
improving the risk assessment for a
particular chemical or group of
chemicals.
To characterize the database,
the Technical Panel has developed a
description of a "minimal" database
and a "robust" database as a way of
describing the range of data that can
be used for deriving a reference value
(see Box 4-3). A great deal of
scientific judgment is necessary when
evaluating the extent of the database
for a particular chemical. Defining
the extent of the database requires an
overall evaluation and judgement as
to where in the minimal - robust
continuum the available database
should be characterized. The
Technical Panel purposely did not
define additional categories between
minimal and robust, e.g., moderate, and the Panel has serious concerns about developing such
categories because of the tendency to try to characterize a database with single word descriptors.
Instead, we strongly support a narrative description of the extent of the database, with emphasis
on the strengths and limitations of the data. It should also be noted that a database that is less
Box 4-3. Description of Minimal and Robust Databases
Minimal Database: No human data available, route-specific
toxicity data are limited to dose-response data applicable to the
duration in question with assessment of endpoints other than
mortality. A study showing only effect levels for mortality or
other extremely severe toxicity would not be sufficient to set a
reference value.
Robust Database: Includes extensive human and/or animal
toxicology data that cover route-specific information on many
health endpoints, durations of exposure, timing of exposure, life
stages and susceptible subpopulations. In the absence of complete
human data, mechanistic and other data show the relevance of the
animal data for predicting human response. Specifically, the
dose-response data for the reference value in question includes
endpoint-specific data (e.g., developmental toxicity, neurotoxicity)
coupled with pharmacokinetic information as needed for route to
route extrapolation. The toxicity studies include the evaluation of
a variety of endpoints (e.g., hematological, clinical, histology of
target organs) and endpoints specific to any known hazard
characterization. The database for a reference value of less-than-
chronic duration has also addressed the issue of reversibility of
effects and latency to response, taking into consideration the
possibility that less-than-chronic exposure may lead to effects
some period of time after exposure. Biological and chemical
characteristics of the exposure and outcomes, as well as known
limits on reserve capacities and repair of damage, form the basis
for determining the appropriate length of follow-up.
4-17
-------
External Review Draft Do Not Cite or Quote
than minimal should not be used to derive a reference value.
Rather than presenting separate "minimal" and "robust" database descriptions for each
type of reference value that might be derived, the descriptions in Box 4-3 are intended to apply
generally across the various reference value types (e.g., acute, short-term, longer-term, or chronic
durations for oral, dermal, or inhalation routes of exposure). Additionally, it is expected that the
different types of reference values for a particular chemical will be developed within the same
assessment. In this manner, the entire database for a chemical may be relied upon in the
development of each of the different values (e.g., important and relevant insights may be gleaned
from toxicity studies for exposure durations other than those directly corresponding to the type of
reference value being developed).
A minimal database as defined above can be used to set reference values, but the
limitations of such a database should be clearly recognized and discussed in the narrative
description. For example, a minimal database may provide data on only one duration or route of
exposure, or it may be specific to only one endpoint or organ system. Thus, the uncertainties
related to such a database will be great and should be reflected in the size of the UFs applied for
reference value derivation (see further discussion below). On the other hand, a robust database
would address issues of potential toxicity in humans and animals, and include data on several
durations and routes of exposure as well as a thorough assessment of a variety of health
endpoints. It would also include sufficient data on pharmacokinetics and mode action to provide
extensive information for extrapolation of effects to humans, including potentially susceptible
subpopulations. A complete database on a single health endpoint that does not contain
information on other endpoints of possible relevancy, would not necessarily constitute a robust
database, nor would a database that provides complete information on one route and/or duration
of exposure be considered robust. It is clear that a robust database represents a "gold standard"
that will rarely, if ever, be available. However, a lack of robustness does not mean that the
database is deficient to the extent that a reference value could not be derived or that large UFs
would need to be applied. Sound scientific judgement will be required to determine which UFs
are appropriate in each case.
A critical assessment of the extent and quality of the database will inform the selection of
the endpoints to be used to derive the reference values and the appropriate UFs. A reference
value based on a single study would likely have a high degree of uncertainty. As more
information from additional toxicology studies becomes available, pharmacokinetic studies,
4-18
-------
External Review Draft Do Not Cite or Quote
structure-activity relationships, and human data, EPA can have greater assurance that the
appropriate species, route of exposure, and target organ system(s) are known for each duration
reference value needed for a human health risk assessment. As this additional information
becomes available, the use of UFs will likely decrease. The ultimate objective is to account for all
human health endpoints resulting from exposures over all life stages from before conception to the
elderly adult.
The optimum assessment considers subtle effects that impact an individual's quality of life,
as well as so-called "frank" effects (death and major disease). The evaluation should encompass
immediate health outcomes as well as those that are a delayed response to an exposure (i.e., latent
responses), although most current testing guidelines do not explicitly evaluate latency to response.
The following series of questions can help guide the assessment process.
Extent of the Database:
Have adequate studies been conducted to establish the target organs/endpoints?
• Have the effects been characterized for both sexes and all life stages?
• Are data pertaining to potentially susceptible subpopulations available?
Are the responses consistent across species? Are the results of the studies biologically
plausible?
• Is the route and matrix of exposure relevant to the specific reference value being derived?
• Is the duration of exposure appropriate for the specific reference value being derived?
Is the animal species and strain appropriate for extrapolation to humans?
To what degree may the biological endpoints be extrapolated (qualitatively and
quantitatively) to humans?
• Are pharmacokinetic data available? For both sexes, for relevant life stages, for other
susceptible subpopulations?
• Is the shape of the dose-response curve consistent with the known pharmacokinetics of
the test compound?
• Are the metabolism and pharmacokinetics in the animal species similar to those of
humans?
Has the dose-response curve been replicated by or is it consistent with data from other
laboratories and other test species?
• Have the data for all relevant endpoints been adequately modeled by the BMD or other
4-19
-------
External Review Draft Do Not Cite or Quote
appropriate quantitative analysis to determine the most sensitive endpoint(s)?
How well is the toxicity characterized? Do the results of all the studies indicate the
possibility of effects on particular systems that have not yet been explored sufficiently, and
they indicate that additional studies may reveal effects not yet characterized?
D. Derivation of Reference Values
After the database has been thoroughly evaluated for quality and extent as outlined above,
several decisions must be made and procedures applied before the final derivation of a reference
value. This section summarizes the current procedures and points out assumptions made and
areas for improvement and clarification. A variety of factors related to derivation of reference
values is discussed, including the selection of relevant endpoints for the POD for various duration
reference values (Section D.I). Adjustment of the study dose/exposure for duration is described
in Section D.2., and derivation of an HED or HEC is discussed in Section D.3. Other issues are
discussed briefly in Section D.4., e.g., varying levels of response at the BMDL, BMC lower
confidence limit (BMCL), or NOAEL due to varying study designs and test sensitivity and
considerations of adversity and severity (i.e., nature of the response) for choosing the benchmark
response (BMR) level. The nature and application of uncertainty/variability factors and MFs are
discussed and critiqued in Section D.5, and future directions are briefly discussed in Section D.6.
Section D.7. summarizes the key points from a case study that are discussed in more detail in
Appendix B.
D.I. Selection of the endpoints to use as the POD for Reference Values
Currently, the "critical effect" is used as the basis for the POD, and various UFs are
applied to the dose at the critical effect for derivation of the RfD or the RfC. The critical effect is
defined as "the first adverse effect, or its known precursor, that occurs to the most sensitive
species as the dose rate of an agent increases" (EPA, 2002a). The underlying assumption is that if
the RfD or the RfC is derived to prevent the critical effect from occurring, then no other effects of
concern will occur; in addition, this approach assumes that the relationship of various health
effects for a particular chemical is maintained across species. The Technical Panel is concerned
that presenting only a single critical effect and the critical study from which it was derived in the
IRIS summary table that appears at the beginning of each RfD or RfC file may not provide
4-20
-------
External Review Draft Do Not Cite or Quote
enough information to the reader who is unfamiliar with risk assessment, and thus could be
misleading. Presentation of a single endpoint as a POD for a systemic effect, for example, cannot
capture the nature of the dose-response curve for that particular endpoint. Nor does the
presentation of a single endpoint convey the possibility that other more serious endpoints may
have a dose-response character markedly different from the less serious endpoint. For example,
an agent may have a clear progression of responses with increasing dose that is seen as one type
of effect at the lowest exposure level (e.g., proteinuria in the case of cadmium) but at a higher
level it produces additional effects (proteinuria PLUS GFR decrements), and at the highest level
even more types of effects (proteinuria PLUS GFR decrements PLUS osteomalacia). Each of
these effects could have a markedly different dose-response character. Focusing on a single
critical effect also does not reflect the situation in which other types of effects may be found at
similar levels of exposure or the variety of health outcomes that may result when an exposure
significantly exceeds the RfD or the RfC. Most importantly, in light of the Technical Panel
recommendations for deriving an expanded number of reference values for different durations and
routes of exposure, the limitations of focusing only on the critical effect become apparent because
the most sensitive endpoint may be different for different durations or routes of exposure.
Layered upon this complex consideration of dose-response is the further complication that
all of the exposure levels producing these effects are or should be adjusted to a human equivalent
exposure at the time of their comparison. These adjustments may profoundly affect what is
considered the most sensitive organ or system. Effects that occur at the same external inhaled
concentration but in different organs in the same exposed animals (e.g., effects in the liver and
effects in the nasal cavity) may have quite different HECs based on the current RfC methodology
(EPA, 1994), as the underlying basis for the adjustment used for systemic effects is markedly
different from that used for portal-of-entry effects between animals and humans. This adjustment
procedure is discussed further below but is noted here because of its interrelationship with
identification of what is to be considered a critical effect.
These aspects all support the case that a more comprehensive approach to setting
reference values requires a more extensive and systematic analysis of endpoints than has typically
been conducted in the past. In the approach proposed here, the selection of the POD would be
similar to the current critical effect approach (e.g., EPA, 1994), and include the use of sound
scientific judgment in evaluating the strength and validity of studies and the extent of the database,
as described in Section C above. In this approach, however, the selection of the POD would be
4-21
-------
External Review Draft Do Not Cite or Quote
based on consideration of all relevant and appropriate endpoints carried through the derivation of
sample reference values, with selection of the limiting value(s) protective of all endpoints as the
final step (the same approach would be used for deriving a POD for low-dose modeling as
discussed in the proposed cancer risk assessment guidelines, EPA, 1999d). For example, the
dose-response curves would be modeled for several adverse endpoints and the corresponding
BMDs and BMCs and their lower 95% confidence limits (BMDLs/BMCLs) calculated (EPA,
2000b) or NOAELs determined if dose-response modeling is not possible. Next, duration
adjustment to the continuous exposure scenario would be performed for each endpoint with
further adjustment to the corresponding HECs using the RfC methodology (EPA, 1994) or
adjusted BMDLs or NOAELs for oral or dermal exposures (see Section D.3 for further
discussion). These adjusted values would represent the POD for each relevant endpoint. Then
uncertainty/variability factors that take into account a variety of issues, including chemical-specific
data, such as known pharmacokinetic differences between the laboratory animal species tested and
humans, and mode of action information would be applied to the adjusted values for each relevant
endpoint. The sample reference values would then be compared across endpoints and organ
systems to determine which are the most relevant for use in deriving the final reference value for
each exposure duration that will be protective of the human population (including susceptible
subgroups).
The Technical Panel recommends the use of a more visual and graphic exposure-response
array to depict the PODs for all relevant endpoints for various routes and durations of exposure,
somewhat like those shown in the ATSDR Toxicology Profiles, but with appropriate changes for
the purpose of deriving reference values. The exposure-response array of the PODs would
facilitate the evaluation and comparison of relevant endpoints and values. (See an example of the
proposed approach in a case study on Chemical X, Section D.7. below and Appendix B.)
D.2. Dose adjustment for duration of exposure
Available studies from which reference values are derived seldom if ever match the intent
of the reference value regarding species or duration. For example, chronic RfD and RfC values
are intended by definition to be for "a continuous exposure to the ... human population." Doses
or exposures from studies in which animals are exposed for less than a lifetime or in which worker
populations are exposed only during working hours require adjustment in order to be concordant
with the intended duration of the reference value. This section describes various procedures that
4-22
-------
External Review Draft Do Not Cite or Quote
are currently used by the Agency to adjust a LOAEL, a NOAEL or a BMDL with regard to
duration. The basis for these adjustments is discussed as is the applicability of these procedures to
various routes of exposure. The Agency has invested considerable time and effort into exploring
these aspects for the inhalation route. A major point that will become apparent in this discussion
is that methodologies for duration adjustment via the inhalation route are currently in place as part
of the existing methodology for the chronic RfC and as proposed for acute inhalation exposure
(ARE) derivations, whereas no comparable documents exist yet for the oral or dermal routes of
exposure.
D.2.a. Duration adjustment procedures for inhalation exposures to continuous-
exposure scenarios
Duration adjustment to continuous exposure is regularly applied to studies of repeated-
exposure but not single-exposure inhalation toxicity studies in animals and humans to adjust
discontinuous exposure scenarios to those applicable to the Agency's intent for human
assessments, i.e., a daily continuous exposure (EPA, 1994). Operationally, this is accomplished
by applying a C x t product both for the number of hours in a daily exposure period and for the
number of days per week that the exposures are performed. In an inhalation study in which
animals are exposed to 100 ppm for 6 hours, 5 days per week, the adjustment to a continuous
exposure concentration would consider both hours/day and days/week:
100 ppm x 6 hrs = 25 ppm x 24 hrs,
25 ppm x 5/7 days/week =17.9 ppm
with 17.9 ppm being the concentration adjusted for continuous exposure.
Human occupational airborne exposures are often reported as 8-hour time-weighted
averages (TWA). These, too, are converted to a continuous concentration for use in reference
values. In this case, half of the daily ventilatory capacity of a human is assigned (10 m3 of 20 m3
total) to the 8-hour occupational exposure, based on the assumption that activity levels are higher
in this setting than in others such as at rest or asleep (i.e., instead of 1/3 or 8/24 hrs) (ICRP,
1994). The average airborne concentrations are then multiplied by this factor, 10/20 m3, and the
product is considered to be an average continuous airborne concentration. In parallel with the
animal studies, an adjustment for days/week (usually 5/7 days/week) is also made, if applicable.
4-23
-------
External Review Draft Do Not Cite or Quote
These adjustment procedures imply that the C x t product and not C alone is associated
with the endpoints observed; this may be restated as implying that it is the area under the curve
(AUC), C x t, and not the peak concentration, C alone, that is the dosimeter associated with
toxicity. Although neither of these dosimeters may be demonstrable experimentally to be the
appropriate measure of dose, the Agency uses adjustment to a continuous inhalation exposure
based on the C x t relationship as a matter of policy. When applied to a discontinuous inhalation
exposure regimen from an experimental study, adjustment to a continuous exposure will always 1)
result in a lower value of C, and 2) maintain a measure of total exposure, i.e., C x t. Thus,
application of this procedure provides an automatic margin of protectiveness for chemicals for
which C alone may be appropriate, and it reflects the maximum dose for agents for which total or
cumulative dose is the appropriate measure. When considered in this way, this policy can be
regarded as being protective of public health. However, assessors should be encouraged, to look
for data on specific chemicals that support the use of C x t, or that offer alternative models for
adjustment of exposure duration.
D.2.b. Duration adjustment for inhalation developmental toxicity studies - a
current exception
A notable exception to duration adjustment occurs with inhalation developmental toxicity
studies in which this practice historically has not been done. Although the Guidelines for
Developmental Toxicity Risk Assessment (EPA, 1991) recommended against dosimetric
adjustment on the basis that developmental effects were more likely to depend on peak exposure
concentration, further evaluation indicates that developmental effects for a number of agents are a
function of AUC (Weller et al., 1999), although the effects of some agents have been shown to be
more a function of peak concentration (Nau, 1991), and in some cases the same agent may be
more related to AUC or peak concentration depending on the timing of exposure (Terry et al.,
1994).
On the basis of this information and the rationale used for dosimetric adjustment for other
health effects (i.e., that exposure adjustment based on C x t tends to be more health protective),
the Technical Panel recommends that duration adjustment procedures to continuous exposures be
used for inhalation developmental toxicity studies as for other health effects from inhalation
exposure. The Panel also urges continued development of data, modeling, and improved
procedures for dose-duration adjustments related to developmental toxicity.
4-24
-------
External Review Draft Do Not Cite or Quote
D.2.c. Duration adjustment for acute reference values - discontinuous scenarios of
24 hours or less
As discussed above, the magnitude of response to a toxic chemical exposure usually
depends on both the concentration and the duration of the exposure such that the combination of
these components, C * t, is determinative of the response and, by logical extension, of the internal
dose of a chemical at the target tissue. In the derivation of acute, short-term, or longer-term
reference values, there may be a need to specifically adjust or to present these values under
alternative C x t combinations. For example, an acute reference value may be required for both a
1-hour duration and an 8-hour duration, whereas the data available on which to base the acute
value is from a 4-hour exposure. Currently, the available guidance on this issue is contained in the
draft methodology for development of AREs (EPA, 1998b). This section presents the adjustment
procedures recommended in the draft ARE methodology.
Because of the recognized limitations of the C x t model, a modification has been
developed such that Cn x T = K, with n being empirically derived. The consequences of varying
the values of the "n" exponent are shown in Figure 4-1. This figure, which was derived from the
current version of the Agency's ARE methodology, is based on the data often Berge et al.
(1986). These investigators were able to empirically derive values of "n" that ranged from 0.8 to
3.5 for a number of chemicals on the basis of acute lethality. A value of 1 for the exponent "n"
would indicate that the relationship described by Haber's law holds and that the response is
related to total dose.
4-25
-------
External Review Draft
Do Not Cite or Quote
Figure 4-1. Concentration-by-duration plot showing the effect of the exponent in the C" x
T = K on extrapolation across time (adapted from the data often Berge et al.,1986).
1,000
Time (Hours)
Note that for any degree of downward slope with increasing duration (lines marked with n
= 1 or n = 0.8), an extrapolation from a longer to a shorter duration (i.e., from right to left) would
result in a higher value for C. Extrapolating from a shorter to a longer duration (i.e., from left to
right), however, would have a different consequence in that with any degree of downward slope,
C would always be lower for the longer duration. Several possible approaches for extrapolation
in this situation could be envisioned. One approach would be to assume a value of 1 for "n,"
such that Cn x T = K and lower values of C would always result; this approach is likely to be the
4-26
-------
External Review Draft Do Not Cite or Quote
actual case because the value of "n" for most chemicals so far examined have been found to show
an appreciable downward slope (e.g., 0.8 < n < 3.5; ten Berge et al., 1986).
The optimal approach for extrapolating from one dose-duration response situation to
another is with the use of a physiologically-based pharmacokinetic model (PBPK) model. The
principle of using PBPK models as the basis for describing the correlations between level and
duration of exposure, internal dose, and biological effect has been stated clearly by Andersen et al.
(1987). Integration of information using PBPK models requires a chemical database that is rich
in toxicity data; therefore, this approach is not applicable to most chemicals for which
pharmacokinetic data are scarce or nonexistent.
In the absence of such a database to support the development of a PBPK model, the
approach recommended by the draft ARE methodology is to use chemical-specific data on
duration dependence of the effect concentration, for example, to interpolate from other adequate
but longer duration data, if they exist (e.g., in extrapolating to 28 days from 7-day data, include
90-day repeated-dose data). This is considered a conservative approach because the duration
adjustment approach (i.e., averaging to continuous exposure) when applied to multiple exposure
studies always results in decreased values for C, i.e., extrapolation would be from left to right,
from shorter to longer durations on the curves in Figure 4-1.
In the absence of chemical-specific data to inform duration adjustment, the response has
most often been related to the simple C x t product. This is also the default in the draft ARE
methodology for adjustment to longer durations. For adjustment to shorter durations, the ARE
methodology conservatively recommends that there be no change in concentration.
Further investigation would increase confidence in the basic assumptions made for the
latter two methods of duration adjustment, including the applicability of the C x t relationship
over spans of exposure from months to years and assessing the "conservativeness" of these
approaches in relation to public health. Further investigation of C x t relationships relative to life-
stage is also recognized as a research need.
D.3. Derivation of an HEC or an HED
Animal data often form the basis for dose-response assessment. By definition, the IRIS
risk values are for humans, thereby making animal to human extrapolation requisite. The specific
point of this extrapolation is to estimate from animal exposure information the human exposure
scenario that would result in the same response. The simplest manner in which this may be done
4-27
-------
External Review Draft Do Not Cite or Quote
is application of an animal-to-human UF (discussed further below), typically with a value of 10; in
application this means that humans are assumed to be more sensitive to effects by a factor of 10
than are animals. Much of the RfC methodology (EPA, 1994) focused on improving the science
underlying the animal to human UF, segregating it into pharmacokinetic and pharmacodynamic
components, and providing generalized procedures to derive dosimetric adjustment factors
(DAF). Application of DAFs to the animal airborne exposure values yields estimates of the
concentration that would result in the same concentration to humans, i.e., the FffiC. Application
of a DAF in the calculation of a FtEC is considered to address the pharmacokinetic aspects of the
animal to human UF (i.e., to estimate from animal exposure information, the human exposure
scenario that would result in the same dose to a given target tissue). The current Agency practice
is to accommodate uncertainty about the remaining pharmacodynamic component through
application of a partial animal-to-human UF (10°5, which is typically rounded to 3). The
theoretical basis for derivation of DAFs used in calculating HECs, along with recommendations
for improvement of this process, is discussed in this section.
Currently, no procedures parallel to the inhalation RfC methodology exist for derivation of
either oral or dermal human equivalents from animal data. Default factors (usually of 10) are
routinely applied to address the issue of animal-to-human extrapolation. Thus, no parallel to the
HEC, i.e., an FED, is derived or other adjustments applied to the animal oral /dermal dose.
This section recommends that dose adjustments similar to those by which FtECs are
estimated be explored in deriving FtEDs for oral and dermal exposures. This would be
accomplished in a manner parallel to the HEC derivation, by instituting and applying a DAF to
animal oral or dermal exposures. Specific recommendations are also presented and discussed
concerning the basis for deriving DAFs for FLED calculation. These recommendations, along
with current procedures for estimating human equivalent values, are illustrated in Figure 4-2.
This figure also demonstrates how calculation of the HEC through application of a DAF is
considered to address the pharmacokinetic, but not the pharamacodynamic component of the
animal-to-human extrapolation. Procedures outlined in this figure for derivation of an FIEC may
be applied to any animal inhalation exposure, regardless of whether it is a BMDL, a NOAEL, a
LOAEL, or another effect level.
D.S.a. PBPK models and derivation of HEDs and HECs: estimating internal dose.
The preferred option for calculating an FLED or FIEC is to use a PBPK model
4-28
-------
External Review Draft
Do Not Cite or Quote
parameterized for the species and regions (e.g., respiratory tract) involved in the toxicity, as
shown on the left-hand side in Figure 4-2. When sufficiently parameterized, a PBPK model is
capable of calculating internal doses to a target organ from any exposure scenario in an animal
and then estimating what human exposure would result in this same internal dose, i.e. the HED or
HEC. A formal DAF is not calculated in this process; rather, the model itself serves as a DAF in
estimating HECs or HEDs. However, constructing a PBPK model is an information-intensive
process that requires much chemical-specific data, including route-specific data. Such
sophisticated data and models are available usually for only a subset of chemicals that have
extensive databases.
Figure 4-2. Current and Proposed Generalized Procedures for the Derivation of HECs or
HEDs from Animal Exposures
Oral/Dermal/Inhalation
Exposure
Chemical-
Specific PBPK
Model
Inhalation Exposure Oral/Dermal Exposure
Exposure conditions,
e.g., mg/m3, hr/d, d/wk
Adjust to
Continuous Exposure
24 hr/d, 7 d/wk
Application of DAF
(for pharmacokinetics)
Exposure conditions,
e.g., mg/kg-day, d/wk
Adjust to
Continuous Exposure
(7 d/wk)
Application of DAF
(for pharmacokinetics)
I HED
DAF = Dosimetric Adjustment Factor
4-29
Proposed
-------
External Review Draft Do Not Cite or Quote
It should be noted that even these sophisticated models are often parameterized on the
basis of adult members of the species. Many of the parameters critical to PBPK model solutions
are sensitive to life stages, such as lung function/development in humans (Pinkerton and load,
2000), for which there are no or few data available. Thus, these models are available but often
cannot specifically address species differences at life stages other than mature adults (and then
usually males). The Technical Panel encourages research and data gathering to support the
construction of PBPK models, endorses attempts to produce PBPK models that are sensitive to
life stages, and supports fully attempts to produce template models for suites of related chemicals,
as has recently been done by Barton et al. (2000).
D.S.b. Default procedures and derivation of HECs from the RfC Methodology:
derivation and application of DAFs
The next lower level of complexity in deriving HECs is less data-intensive than the PBPK
approach. Shown also in Figure 4-2, this procedure involves the use of species-specific
physiologic and anatomic factors relevant to the form of pollutant (e.g., particle or gases) and
categorized with regard to elicitation of response either locally (i.e., within the respiratory tract)
or remotely. These factors are all employed in determining the appropriate DAF. For HECs,
DAFs are applied to the "duration-adjusted" concentration to which the animals were exposed
(e.g., to a weekly average). The generalized DAF procedures may also employ chemical-specific
parameters, such as mass transport coefficients, when available. In lieu of such data, however,
default procedures that yield generalized adjustments are recommended. Although these
generalized procedures were developed from the existing scientific understanding of the relevant
processes, they have not been comprehensively evaluated (e.g., using data from humans and
animals). They are explained fully in the RfC Methodology (EPA, 1994).
For example, the manner in which a HEC is calculated for a reactive gas that elicits an
effect in the extrathoracic region of the respiratory tract (i.e., the nasal tract) of a rat is by creating
a surface area/ventilation ratio for both humans and rats and applying it to the external exposure
concentration for rats. The current default values used for both the human and rat extrathoracic
surface area are single estimates from the literature and are apparently estimated from adult
specimens. The ventilation measure for humans is set at a default value of 20 m3, and the
ventilation measure for rats is based on an algorithm of body weight (from EPA, 1988). A major
assumption made in this particular adjustment is that the distribution of a gas in the region of
4-30
-------
External Review Draft Do Not Cite or Quote
interest is uniform, although it is known to be highly nonuniform (Kimbell et al., 1993, 1997).
Data are not available to address this simplified assumption directly. Use of the method, for
example on effects in the extrathoracic region, results in a DAF of about 0.2, such that the
resultant HECs are about 20% of the animal duration adjusted concentration. Although
information is not yet available to address this assumption, indications are that resolution with
actual data may produce DAFs that are much closer to unity, i.e., that are near the animal adjusted
concentration.
In comparison to the procedure for gases that elicit respiratory effects, calculation of an
HEC for a category 3 gas (i.e., a gas that is relatively water-insoluble and unreactive in the
respiratory tract, and for which the site of toxicity is general remote to the site of absorption in
the pulmonary region) is usually accomplished by creating a ratio of the blood:gas partition
coefficient for the laboratory animal species to the human value. The ratio is used as the DAF and
applied to the experimental exposure concentration. In lieu of data on the values for blood:gas
partition coefficients for the chemical, or when the data indicate the ratio to be >1, the default
assumption is that the ratio of animal coefficient to human coefficient is 1, and therefore the DAF
would be 1. However, available data on partition coefficients for a number of compounds
indicate that the animal/human ratio is usually >1 (Gargas et al., 1989; Jepson et al., 1994) such
that the DAF would also be >1. In the context of substituting data-derived values for UFs, the
Technical Panel recommends further investigation into using data-derived values in constructing
the animal/human ratios, even when much greater than 1, in place of the default.
The default dosimetric adjustment procedure for particulate substances is an empirical
model that estimates regional deposition only, although it is recognized that with the development
of the relevant data, clearance and the retained dose may be used as a DAF (EPA, 1994). The
DAF for particles is more specifically termed the regional deposited dose ratio (RDDR) and is
derived from a normalizing factor (surface area being the recommended factor for all three
regions of the respiratory tract), the ratio of animal to human minute volumes (where the human
default value is the traditional adult value of 13.8 L versus the adult value for the relevant animal),
and the ratio of animal to human regional fractional deposition. Physiological parameters used in
estimating the regional deposition include body weight, minute volume, and surface area for the
three areas of the respiratory tract. Defaults for the human values are based on adult data (e.g.,
70 kg body weight, 13.8 L minute ventilatory volume, etc); the animal values are also traditionally
based on adult data. To evaluate protectiveness of these default calculations for different life
4-31
-------
External Review Draft Do Not Cite or Quote
stages, it may be appropriate to perform ratio calculations using data for other life stages.
As a general recommendation, the Technical Panel encourages further consideration of the
existing animal-to-human extrapolation procedures described in current methodologies (e.g., the
chronic RfC methodology, EPA, 1994) and the development of procedures for inhalation
adjustment to incorporate the most current scientific thought and data to address, as needed,
issues of variability due to life stage and other intrinsic factors. This consideration would include
examination of the extent to which calculation of an HEC (or any recommended HED) addresses
cross-species pharmacokinetics and identification and parallel investigation into issues of
pharmacodynami cs.
D.S.c. HECs and children - a special case?
Children are often characterized as constituting a potentially susceptible subgroup, as they
could be at greater risk than adults for inhaled toxic agents, including both gases and particulates,
for reasons relating to either pharmacokinetics or pharmacodynamics. It is clear for any of a
variety of reasons related to pharmacokinetics, that an adult and a child breathing the same
concentration of an agent such as a reactive gas may receive different doses to the body or to the
lungs. A generalized theoretical approach to judging whether greater doses would be received by
children than by adults with both breathing the same concentration of a reactive gas, for example,
would be to compare the amount of gas breathed in (which would be directly proportional to the
ventilatory volume) with the overall surface area in the respiratory tract on which the gas may
impinge. The current Agency default assumption used in the derivation of HECs for particles and
reactive gases eliciting respiratory effects is that the surface area of the total respiratory tract of an
adult male, estimated at 54.3 m2, is exposed to a total daily air intake of 20 m3, a volume for an
adult male derived from a combination 24-hour activity pattern in ICRP (ICRP, 1994) of sitting
awake for 8 hours, light exercise for 8 hours and sleeping for 8 hours.
It has been well established that the human respiratory system passes through several
distinct stages of maturation and growth that involve branching morphogenesis and cellular
differentiation during the first several years of life and into adolescence (Pinkerton and load,
2000). The proportion of surface area to ventilation volume may be markedly different during
these developmental stages. The significance of these disproportions with regard to toxicant
exposure overall or to the sites of active cellular differentiation have yet to be elucidated.
The Technical Panel recommends that issues involving dose to the young from inhalation
4-32
-------
External Review Draft Do Not Cite or Quote
exposures be pursued both theoretically and experimentally in order to establish the basis on
which children should be considered as a susceptible subpopulation for inhalation exposures. It
should also be reiterated that this is an estimate of the pharmacokinetic aspect of dose only, and
pharmacodynamic differences between the lungs of young children and adults are not addressed.
D.S.d. Derivation of an HED for oral and dermal exposure - use of BW3/4 as a cross-
species DAF
As indicated above, current Agency procedures do not provide for the calculation of an
HED for oral or dermal exposure scenarios that would be parallel to the inhalation HEC. Instead,
assumptions are made regarding the comparability of ingested or applied dose, based on a mg/kg
body weight basis, and there is no adjustment for portal of entry alterations to internal dose or on
portal of entry versus systemic effects. The Technical Panel recognizes the work of an
interagency workgroup to develop and propose dosimetric adjustment procedures for both dermal
and oral routes of exposure in order to address those aspects of cross-species dosimetric
adjustment that are missing in Figure 4-2. Some of these proposals have already appeared in
abstract form (Jarabek, 2000; Hanna and Jarabek, 2000; Hubal et al., 2000; Rigas et al., 2000).
Figure 4-2 demonstrates that dosimetric adjustment procedures for estimating human
equivalents from animal values are not consistent for different exposure routes. Other
procedures, both from within and external to the Agency, could be explored for the purposes of
deriving a DAF and employing it to estimate an HED. For example, in the absence of more
sophisticated physiologically based models, the Agency has endorsed scaling of doses for
carcinogens between species according to body mass raised to the 3/4 power (BW3/4) (EPA,
1992). This procedure presumes that equal doses in these units (i.e., in mg/kg3/4/day) when
administered daily over a lifetime will result in equal lifetime cancer risks across mammalian
species. This same relationship, i.e., BW3/4, has been affirmed to apply across entire phyla,
including plants (Gillooly et al., 2001), for general metabolic rates. The basis for less than the full
power relationship for general metabolic processes (i.e., < BW1) is thought to be related to
species differences in exchange surfaces and distribution networks that constrain concentration
and flux of metabolic reactants (West et al., 1997; Enquist et al., 1998). Thus, when this
procedure is applied to animal data, the resulting scaled human dose may be viewed as a valid
cross-species relationship not only of cancer potency but also for general metabolic processes and,
by extension, for other phenomena involving the fundamental determinants of concentration and
4-33
-------
External Review Draft
Do Not Cite or Quote
flux, the same ones that drive basic pharmacokinetics.
This brief analysis of the BW3/4 cross-species relationship and pharmacokinetic processes
and the Agency's endorsement of this procedure for carcinogenic agents makes this process a
possible candidate for estimating cross-species pharmacokinetic relationships in the absence of
adequate pharmacokinetic information. That is, BW3/4 factors could be applied as DAFs for
deriving an HED. This procedure would parallel that used for deriving the HEC. As with the
HEC, however, this process applies only to pharmacokinetic aspects of cross-species
extrapolation and does not address pharmacodynamic differences that may exist between species.
As with the HEC, consideration of PD is proposed to be through application of a portion of the
animal to human extrapolation (10°5, which is typically rounded to 3). Table 4-3 shows the
general magnitude of the DAFs that would be applied to various species to obtain the HED,
along with the default UF of 3 to cover pharmacodynamic differences.
Table 4-3. DAFs based on BW3/4 for Various Species
Species
Mouse
Rat
Guinea pig
Rabbit
Human
Wt(Kg)
0.03
0.25
0.5
2.5
70
DAFa
7
4
O
2
1
"Derived based on BW3/4 relationship. All variables in BW3/4
relationship containing time will scale BW"1/4, such that animal BW"1/4
/ human BW1/4 = DAF.
The Technical Panel encourages consideration of cross-species extrapolation procedures
for oral and dermal reference values, including evaluation of the most current scientific thought
and data to address, as needed, issues of variability due to life stage and other intrinsic factors.
This consideration would include examination of the extent to which calculation of an HED
addresses cross-species pharmacokinetics and identification and parallel investigation into issues
of pharmacodynamics.
4-34
-------
External Review Draft Do Not Cite or Quote
D.4. Other issues
The Technical Panel considered several other issues that are related to the application of a
factor (data-derived or default) to the BMDL, the BMCL, the NOAEL, or the LOAEL selected
as the POD from data considered adequate for risk assessment. In particular, there was
controversy about the application of such a factor on the basis of the level of response at the
BMD, the BMC, the NOAEL, or the LOAEL. For example, the use of a quantitative dose-
response modeling approach results in the calculation of a BMD or a BMC, which is based on a
particular level of response, i.e., the BMR. The BMR is usually selected to be at the low end of
the observable range of the data, which is dependent on the power of the study to detect changes
from control values. The limit of sensitivity for most long-term bioassays is in the range of 10%
as determined from both the typical number of animals used in bioassays (~50/group) and a low
spontaneous background rate (e.g., 0.1%) for a given effect (Haseman, 1984; Haseman et al.,
1989). For other types of studies, however, the limit of sensitivity may be lower or higher than
10%. For example, in an analysis of a large number of standard prenatal developmental toxicity
studies with an average sample size of 15 - 20 litters, the limit of sensitivity averaged 5% for the
proportion of pups affected/litter, whereas when the quantal endpoint, i.e., the number of litters
affected, was analyzed in dams from the same studies, the limit of sensitivity averaged 30% (Allen
et al., 1994). For data from some human studies, e.g., high-quality, large epidemiology studies,
the limit of sensitivity may be in the range of 1 to 5%.
In the BMD Guidance Document (EPA, 2000b), the BMDL or BMCL is recommended
for the POD in order to ensure that a majority of the population is below the selected BMR.
However, a concern has been raised that a BMD or BMC based on a response rate of >10% may
not be appropriate to use in deriving an exposure to the human population (including sensitive or
susceptible subgroups) likely to be without appreciable risk of deleterious or adverse effects
(from current and proposed reference value definitions [Boxes 4-1 and 4-2]) without application
of a factor to extrapolate to a lower dose/exposure level considered to reflect a more appropriate
level of risk (e.g., <10%).
Similarly, the NOAEL is not necessarily a no-effect level and it depends on the study
design, including sample size, background rate, and response variability, which can be used to
determine the limit of detection for a particular study. Thus, a NOAEL may be equivalent to no
response, or it may actually represent a substantial response rate. Previously, there has been no
4-35
-------
External Review Draft
Do Not Cite or Quote
attempt to apply a factor to the NOAEL on the basis of power calculations, sample size, or
response variability for the derivation of a POD, although professional judgment is recommended
in deciding whether the study is acceptable for use in deriving a POD.
Adjustment for the steepness of the dose-response curve has been noted as another critical
aspect of the dose-response character that is not currently considered in the choice of a response
level using either a BMD/BMC or a
NOAEL approach.
The Technical Panel was unable to
fully evaluate these issues or to reach
agreement about any recommendation for
change to the current methodology, and it
recommends that they be considered further
by the Agency. The Technical Panel also
recommends that factors such as the
response rates at the BMD or the NOAEL,
the power of the study, and the slope of the
dose-response curve be included in the
description of the database, where possible,
as part of risk characterization.
D.5. Application of
uncertainty/variability factors
Reference values are derived in a
way that attempts to account for both the
uncertainty and the variability in the data
available (see Box 4-5). The existing
definition of UF in the IRIS glossary mixes
the above concepts. The present definition for UF is as follows.
Uncertainty Factor: One of several, generally 10-fold factors, used in operationally
deriving the RfD and RfC from experimental data. UFs are intended to account for(l) the
variation in sensitivity among the members of the human population, i.e., interhuman or
intraspecies variability; (2) the uncertainty in extrapolating animal data to humans, i.e.,
Box 4-5. Variability and Uncertainty3
Variability refers to true heterogeneity or diversity. For
example, among a population that drinks water from the
same source and with the same contaminant
concentration, the risks from consuming the water may
vary. This may be due to differences in exposure (i.e.,
different people drinking different amounts of water, and
having different body weights, different exposure
frequencies, and different exposure durations) as well as
differences in response (e.g., genetic differences in
resistance to a chemical dose). Those inherent differences
are referred to as variability. Differences among
individuals in a population are referred to as inter-
individual variability, while differences for one individual
over time is referred to as intra-individual variability.
Uncertainty occurs because of a lack of knowledge. It is
not the same as variability. For example, a risk assessor
may be very certain that different people drink different
amounts of water but may be uncertain about how much
variability there is in water intakes within the population.
Uncertainty can often be reduced by collecting more and
better data, while variability is an inherent property of the
population being evaluated. Variability can be better
characterized with more data, but it cannot be reduced or
eliminated. Efforts to clearly distinguish between
variability and uncertainty are important for both risk
assessment and risk characterization.
4-36
-------
External Review Draft Do Not Cite or Quote
interspecies variability; (3) the uncertainty in extrapolating from data obtained in a study with less-
than-lifetime exposure to lifetime exposure, i.e., extrapolating from subchronic to chronic
exposure; (4) the uncertainty in extrapolating from a LOAEL rather than from a NOAEL; and (5)
the uncertainty associated with extrapolation from animal data when the data base is incomplete.
Following the logic above, the LOAEL-to-NOAEL extrapolation, the subchronic-to-
chronic extrapolation, and the database deficiency factors are UFs. The variation in susceptibility
among members of the human population is a variability factor. When a default factor is used for
intra-human variability, however, this factor also contains some degree of uncertainty, because the
range of uncertainty is not really known, although it is presumed to be no more than 10-fold.
Rather than adding a new definition of variability factor, we propose to modify the wording of the
UF definition as follows.
Uncertainty /Variability Factor: One of several, generally 10-fold default factors, used
in operationally deriving the RfD and RfC from experimental data. The factors are intended to
account for (1) the variation in sensitivity among the members of the human population, i.e., inter-
individual variability; (2) the uncertainty in extrapolating animal data to humans, i.e., interspecies
uncertainty; (3) the uncertainty in extrapolating from data obtained in a study with less-than-
lifetime exposure to lifetime exposure, i.e., extrapolating from subchronic to chronic
exposure; (4) the uncertainty in extrapolating from a LOAEL rather than from a NOAEL; and (5)
the uncertainty associated with extrapolation when the database is incomplete.
D.S.a. Recommendations for application of UFs
The exact value of a UF chosen will depend on the factors indicated above and will require
scientific judgment. The default factors typically used cover a single order of magnitude (i.e.,101).
The Technical Panel recommends that departure from the default factors of 10 be based on data
(qualitative and/or quantitative), and that the basis for the departure be made clear in the
assessment. By convention, in the Agency, a value of 3 is used in place of Va power, i.e., 10as,
when appropriate. The Technical Panel recommends that these half-power values be factored as
whole numbers when they occur singly but as powers or logs when they occur in tandem. A
composite UF of 3 and 10 would be expressed as 30 (3 xlO1), whereas a composite UF of 3 and 3
would be expressed as 10 (10°'s x 10°'5 = 101). The exact value of the UF chosen should depend
on the quality of the studies available, the extent of the database, and scientific judgment. It is
imperative that the IRIS documentation contain a justification for the individual UFs selected for a
4-37
-------
External Review Draft Do Not Cite or Quote
particular agent. It should be noted, in addition, that rigid application of log or !/2 log units for
UFs could lead to an illogical set of reference values; therefore, the Technical Panel emphasizes
that application of scientific judgment is critical to the overall process.
The Technical Panel recognizes that there is overlap in the individual UFs and believes that
the application of five UFs of 10 for the chronic reference value (yielding a total UF of 100,000)
is inappropriate. In fact, in cases where maximum uncertainty exists in all five areas, it is unlikely
that the database is sufficient to derive a reference value. Uncertainty in four areas may also
indicate that the database is insufficient to derive a reference value. In the case of the RfC, the
maximum UF would be 3000, whereas the maximum would be 10,000 for the RfD. This is
because the derivation of RfC s and RfDs have evolved somewhat differently. The RfC
methodology (EPA, 1994) recommends dividing the interspecies UF in half, one-half each (10°5)
for pharmacokinetic and pharmacodynamic considerations, and it includes a DAF to account for
pharmacokinetic differences in the calculation of the HEC, thus reducing the interspecies UF to 3
for pharmacodynamic issues. RfDs, however, do not incorporate a DAF for derivation of an
HED, and the interspecies UF of 10 is typically applied.
The Technical Panel recommends limiting the total UF applied for any particular chemical
to no more than 3,000 and avoiding the derivation of a reference value that involves application of
the full 10-fold UF in four or more areas of extrapolation. This maximum of 3,000 applies only to
the UFs discussed in the following sections, and it does not include the various adjustment factors
that have been discussed previously (Chapter 4, Sections D.2. and D.3.). Similar concerns would
need to be considered for the less-than-lifetime reference values, taking into account those UFs
that are appropriate for each duration reference value.
An additional "safety" factor was mandated by the FQPA for pesticide tolerances. This
FQPA safety factor has been discussed in a draft document by the Toxicology Working Group of
the 10X Task Force (EPA, 1999a) as well as in policy by OPP (EPA, 1999c, 2002b). The FQPA
safety factor is to be considered in assessing the risks to infants and children to take into account
the potential for pre- and postnatal toxicity and the completeness of the toxicity and exposure
databases. The statute authorized EPA to replace this additional 10X factor with a different
factor only if, on the basis of reliable data, the resulting MOE would be safe for infants and
children. In its policy guidance document (EPA, 2002b), OPP considered the FQPA factor to
overlap with several of the traditional UFs, but to be in addition to the interspecies and
intraspecies UFs. The overlap is with several traditional UFs that account for data gaps
4-38
-------
External Review Draft Do Not Cite or Quote
(extrapolation from the LOAEL when a NOAEL is not available, extrapolation from a subchronic
study to a chronic-exposure scenario when no chronic study data are available, and application of
a database UF when there are gaps in the data considered essential for setting a reference value,
including lack of data on children). Given this overlap the Technical Panel agrees with the 10X
Task Force draft Toxicology Working Group report (EPA, 1999a) that the current interspecies,
intraspecies, LOAEL-to-NOAEL, subchronic-to-chronic, and database-deficiency UFs, if
appropriately applied using the approaches recommended in this review, will be adequate in most
cases to cover concerns and uncertainties about children's health risks. In other words, the
currently available factors should be sufficient to account for uncertainties in the database from
which the reference values are derived (and does not exclude the possibility that these UFs may be
decreased or increased from the default value of 10). As part of the risk characterization process,
the adequacy and acceptability of the MOE are considered. If there are residual concerns for
control of risks (either toxicity or exposure risks) to children, the MOE can be increased by
applying part or all of the FQPA factor to the RfD, taking into account the traditional UFs that
have already been applied. This FQPA-corrected RfD is called the population adjusted dose
(PAD). Although this approach is consistent with procedures used in the past for managing
potential risks, the FQPA has brought a significant new focus on improving the process of risk
assessment relative to children's health risks from environmental exposures.
Guidance is needed on the use of developmental toxicity data in all reference values,
including the appropriate application of UFs, because of the assumption that a single exposure
during development may produce an effect (EPA, 1991), and the concomitant recognition that
multiple exposures may result in effects at lower doses in many cases or cause tolerance in other
cases. These issues are chemical-specific, and scientific judgement about when and how to apply
UFs must include consideration of pharmacokinetics/metabolism, as well as the mode of action for
each agent.
D.S.b. Interspecies UF
The interspecies UF is applied to account for the extrapolation of laboratory animal data
to humans, and it generally is presumed to include both pharmacokinetic and pharmacodynamic
aspects. The pharmacokinetic aspects of this factor were addressed earlier in the section on
derivation of the HEDs and the HECs (Chapter 4, Section D.3). This UF is intended also to
account for differences in species sensitivity, i.e., pharmacodynamics, between the laboratory
4-39
-------
External Review Draft Do Not Cite or Quote
animal species used for testing and humans. Seldom are there data available to inform
pharmacodynamic differences. One-half the default 10-fold interspecies UF (i.e., 10°5) is assumed
to account for such differences, but more specific data should be used when available (see
discussion of chemical-specific adjustment factors [CSAFs] below), and the flexibility for
applying a factor greater than 10 should be recognized. Unless data support the conclusion that
the test species is more or equally as susceptible to the pollutant as are humans, and in the absence
of any other specific pharmacokinetic or pharmacodynamic data, the default factor of 3 (in
conjunction with HEC derivation) or 10 is applied.
D.S.c. Intraspecies UF
The intraspecies UF is applied to account for variations in susceptibility within the human
population and the possibility (given a lack of relevant data) that the database available is not
representative of the dose/exposure-response relationship in the most susceptible subpopulations
among the human population. In general, the Technical Panel recommends that reduction of the
intraspecies UF from a default of 10 be considered only if data are sufficiently representative of
the exposure/dose-response data for the most susceptible subpopulation(s).
Various authors have evaluated the intraspecies UF using data from animal or human
studies, as summarized by Dourson et al. (1996), who concluded that the 10-fold default factor
appeared to be protective when starting from a median response, by inference a NOAEL assumed
to be from an average group of humans. Renwick and Lazarus (1998) considered data on
toxicokinetics and toxicodynamics to support the idea that the 10-fold intraspecies factor can be
divided into two factors to account for kinetics and dynamics. When they evaluated the
composite 10-fold factor to account for variability in both kinetics and dynamics, they concluded
that a 10-fold factor would cover the vast majority of the population (>99%). These evaluations,
however, did not specifically consider children as part of the range of human variability when
evaluating the adequacy of the intraspecies UF.
In papers that have evaluated this factor for the general population as well as for specific
subpopulations, including children (Renwick and Lazarus, 1998; Renwick, 1998) and the elderly
(Abdel-Megeed et al., 2001), the 10-fold intraspecies factor appears to be sufficient in most cases,
and chemical-specific factors often indicate a requirement for less than a 10-fold factor. Renwick
(1998) indicated that the 10-fold factor is more likely to be sufficient if developmental toxicity
data are available on the specific agent. Calabreese (2001) reviewed the data available on a
4-40
-------
External Review Draft Do Not Cite or Quote
number of chemical classes and concluded that the young are often more susceptible than adults
but that there is a not-infrequent occurrence of greater susceptibility in adults. The sometimes
greater sensitivity among the elderly than among mature adults appears to be related primarily to
reduced renal clearance (Abdel-Megeed et al., 2001; Skowronski and Abdel-Rahman, 2001).
The Technical Panel urges continued research and evaluation of the similarities and
differences between the general population and susceptible subpopulations, particularly children
and the elderly, in their responses to specific agents. From such evaluations, the protectiveness of
the 10-fold default factor can continue to be assessed.
The cases on IRIS in which the intraspecies UF has been reduced from the default of 10-
fold were documented by Dourson et al. (1996). These included 2/46 RfCs and 13/346 RfDs
(overall frequency 3.6%). In those cases where developmental effects were the most sensitive
endpoint (0 RfCs, 6 RfDs), reduction of the intraspecies UF from 10 to 3 was based on data
derived either from human data showing which age groups or time periods were most susceptible
(e.g., methyl mercury exposure to the developing fetus) or from an animal study with support
from strong human or other data (e.g., Aroclor 1016 in utero exposure in monkeys, strontium-
induced rachitic bones in young rats). In three cases, the intraspecies UF was reduced to 1, based
on very specific data about the particular vulnerability of infants and children within certain age
ranges to an agent (e.g., nitrate, nitrite, fluorine/soluble fluoride). However, even within these
populations, it is possible that some variability exists, based on genetics, lifestyle, or other factors.
In cases where the susceptible subpopulation is quite specifically defined (e.g., through
knowledge of the chemical's mode of action) so that the resultant RfC is truly applicable to the
susceptible subpopulation (although not necessarily to hypersensitive individuals), reduction of the
intraspecies UF is warranted. Thus, the Technical Panel supports and expands the
recommendation of the Toxicology Working Group of the 10X Task Force (EPA, 1999a) that
reduction of the intraspecies UF from a default of 10 be considered only if data are sufficient to
support the conclusion that the data set on which the POD is based is representative of the
exposure/dose-response data for the susceptible subpopulation(s). Given this, whether and how
much the intraspecies UF may be reduced must be linked to how completely the susceptible
subpopulation has been identified and their sensitivity described (e.g., versus assumed). At the
other extreme, a 10-fold factor may sometimes be too small because of factors that can influence
large differences in susceptibility, such as genetic polymorphisms. The Technical Panel urges the
development of data to support the selection of the appropriate size of this factor, but recognizes
4-41
-------
External Review Draft Do Not Cite or Quote
that often there are insufficient data to support a factor other than the default.
D.S.d. LOAEL-to-NOAEL UF
A UF (default 10) is typically applied to the LOAEL when a NOAEL is not available. The
size of the LOAEL-to-NOAEL UF may be altered, depending on the magnitude and nature of the
response at the LOAEL. It is important to take into consideration the slope of the dose-response
curve in the range of the LOAEL in making the determination to reduce the size of the LOAEL-
to-NOAEL UF. Several papers have described the magnitude of the difference between the dose
at the LOAEL and at the NOAEL. For example, Lewis et al. (1990) and Faustman et al. (1994)
showed that the ratio of the LOAEL-to-NOAEL in many cases was approximately 3-fold, but in a
few cases, the difference was as much as 10-fold. In general, the ratio of the doses at the LOAEL
and the NOAEL is likely to vary considerably among studies and may not be informative. This is
because the lowest dose in a study is often selected to ensure that no statistically significant
response above control is observed and the next higher dose is selected to ensure that some
significant response is observed, rather than selecting doses that will give a maximum NOAEL
and a minimum LOAEL. Data should be carefully evaluated, taking into consideration the level
of response at the LOAEL and the NOAEL and the slope of the dose-response curve before
reducing the size of the UF applied to the LOAEL.
D.S.e. Database UF
The database UF is intended to account for the potential for deriving an underprotective
RfD/RfC as a result of an incomplete characterization of the chemical's toxicity. In addition to
the identification of toxicity information that is lacking, review of existing data may also suggest
that a lower RfD/RfC might result if additional data were available. Consequently, in deciding to
apply this factor to account for deficiencies in the available data set, and in identifying its
magnitude, the assessor should consider both the data lacking and the data available for particular
endpoints and life stages.
In many respects, the additional 10-fold factor for infants recommended by the National
Research Council (1993) by Schilter et al. (1996) and called for in the 1996 FQPA is similar to
the database UF. Often a factor of 3 is applied if either a prenatal toxicity study or a two-
generation reproduction study is missing, or a factor of 10 may be applied if both are missing
(Dourson et al., 1996). Dourson et al. (1992) examined the use of the database UF by analyzing
4-42
-------
External Review Draft Do Not Cite or Quote
ratios of NOAELs for chronic dog, rat, and mouse studies and reproductive and developmental
toxicity studies in rats. They concluded that reproductive and developmental toxicity studies
provide useful information for establishing the lowest NOAEL, and if one or more bioassays are
missing, a factor should be used to address this scientific uncertainly in deriving a chronic RfD.
If there are data from the available toxicology studies that raise suspicions of
developmental toxicity and signal the need for other types of testing, e.g., specialized DNT
studies, developmental immunotoxicity studies, developmental carcinogenesis studies, or
developmental endocrine toxicity studies, then the database factor should take into account
whether or not these data have been collected and used in the assessment and their potential to
affect the POD for the particular duration RfD or RfC under development. The size of the factor
to be applied will depend on other information in the database and on how much impact the
missing data may have on determining the toxicity of a chemical and, consequently, the POD.
D.S.f. Subchronic-to-chronic duration UF
As indicated earlier, a duration adjustment currently in use is the application of a UF when
only a subchronic duration study is available to develop a chronic reference value such as the RfC
or the RfD (EPA, 1994). A default value of 10 for this UF is applied to the NOAEL/LOAEL or
BMD/BMC from the subchronic study under the assumption that effects from a given compound
in a subchronic study occur at a 10-fold higher concentration than in a corresponding (but absent)
chronic study. This factor would be applied subsequent to the adjustment of the exposures from
intermittent to continuous, as above.
The specific use of a UF applied to a subchronic study in the derivation of a chronic
reference value is reasonable. Some work has been published on this aspect of extrapolation
(Lewis et al., 1990; Pieters et al., 1998). Guidance for replacement of the default factor of 10 by
CSAFs may be forthcoming. It would be appropriate to incorporate such data into applicable
assessments. In the current practice, this factor is applied when a chronic reference value is
derived from a database in which the critical study is of subchronic duration. No chronic
reference value is derived if neither a subchronic or chronic study is available. The application of
a UF to less than sub-chronic studies is not part of the current practice, but further exploration of
this issue may be appropriate. For short-term and longer-term reference values, the application of
a UF analogous to the subchronic to chronic duration UF also needs to be explored, as there may
be situations in which data are available and applicable but they are from studies in which the
4-43
-------
External Review Draft Do Not Cite or Quote
dosing period is considerably shorter than that for the reference value being derived.
D.S.g. MF
A clear definition of intended usage for an MF is lacking. The only comments located
about the MF are in the RfC methodology (EPA, 1994), and they indicate that the MF is intended
to account for scientific uncertainties in the study or database that are not explicitly treated by
other UFs. It is further stated that use of the factor depends principally on professional judgment
and assessment. Some example applications are also given, such as accounting for small sample
size or for poor exposure characterization in the principal study. The definition in the IRIS
glossary gives similar examples.
The description of the database UF shows substantial similarity to that of the MF. Text on
the database UF indicates that this factor attempts to recognize that without a comprehensive
array of endpoints there is uncertainty as to whether all possible toxicologic endpoints at the
various life stages are adequately addressed. Without this information, uncertainty remains as to
whether the critical effect chosen for RfD or RfC derivation is either the most sensitive or the
most appropriate. There are only seven cases in IRIS for which an MF has been applied: RfDs for
chromium III, chromium VI, nitrite, 1,1-biphenyl, and manganese, and RfCs for methyl ethyl
ketone and acetonitrile. The rationale for these varies considerably, but all cases appeared to be
for reasons that could be considered under other UFs.
Recent developments in the IRIS process include the obligation for risk characterization
within the assessments. A central aspect of risk characterization includes discussing confidence
and uncertainties in the quality of data used and the "clarity, transparency, consistency and
reasonableness" of the assessment (EPA, 2000a). Within the risk characterization, the assessor
has a pathway provided to discuss and analyze all aspects of uncertainty about the database,
including the adequacy or limitations of the data base, directly in the assessment.
The Panel considers the purpose of the MF to be sufficiently subsumed in the general
database UF. The Panel also notes that the risk characterization section of assessments may be
used to provide a full and complete characterization of all uncertainly, including any residual
uncertainty that may not be addressed by the other UFs. In view of these factors, the Panel
recommends the discontinuance of the use of the MF.
4-44
-------
External Review Draft Do Not Cite or Quote
D.6. Future directions
D.6.a. CSAFs
There is growing support for the use of CSAFs that provide an incentive to fill existing
data gaps (Murray and Anderson, 2001; Meek, 2001, Meek et al., 2001; Bogdanffy et al., 2001).
Additional chemical-specific data permit the replacement of components of interspecies or
interindividual variation with data-derived values in the context of the traditional default
framework as developed by Renwick (1993) and revised by the IPCS (1994). The following is a
brief discussion of available methodologies that promote the use of CSAFs in risk assessment.
Renwick (1993) described the use of toxicokinetic and toxicodynamic data as a means of
replacing the traditional 10-fold safety factors for human sensitivity and experimental animal-to-
human extrapolation in developing acceptable daily intakes (ADIs). His data-derived approach
assigns default values for both toxicokinetic and toxicodynamic differences within each traditional
10-fold safety factor. Specifically, Renwick proposed dividing both the interspecies and the
interindividual UFs into a factor of 2.5 for toxicodynamics and a factor of 4.0 for toxicokinetics.
IPCS (1994) has adopted the data-derived approach initially proposed by Renwick (1993), with a
slight modification in the UF for interindividual variation (3.16 for toxicodynamics and 3.16 for
toxicokinetics). IPCS has used this approach in several of its recent risk assessments (e.g., EHC
for Boron, IPCS, 1998), and EPA is proposing a similar approach for boron (EPA, 200la).
IPCS has developed a draft guidance document (IPCS, 2001) to assist risk assessors in the
use of experimental data in deriving CSAFs for interspecies differences and human variability in
dose/concentration response assessment. CSAFs have been adopted because they describe better
the nature of the refinement to the usual default approach.
For several years, EPA used a more qualitative approach to modify the usual 10-fold default
values (Dourson et al., 1996). Recently, it has used a data-derived approach as one of the methods
to derive a UF for boron (EPA, 200la).
EPA has not yet established guidance for the use of data for derivation of UFs, but the
division of UFs into toxicodynamic and toxicokinetic components has been used in the RfC
methodology (EPA, 1994). EPA's assessments of data assume a division of both interspecies and
intraspeciesUFs into toxicokinetic and toxicodynamic components assigned default values of 3.16
(10as) each. The Agency will develop its own guidance for the use of CSAFs in risk assessment,
based on some of the available methodologies (e.g., IPCS).
The Technical Panel would like to caution the user that for many substances there are
4-45
-------
External Review Draft Do Not Cite or Quote
relatively few data available to serve as an adequate basis to replace defaults for interspecies
differences and human variability with more informative CSAFs. Currently, relevant data for
consideration are often restricted to the component of uncertainly related to interspecies
differences in toxicokinetics. Although there are fewer relevant data with which to address the
other four components, namely interspecies (animal-to-human) differences in toxicodynamics,
intraspecies (human) variability in toxicokinetics, intraspecies (human) variability in
toxicodynamics, and adequacy of the database, it is anticipated that availability of such
information will be needed to apply CSAFs. Specifically, the data-derived CSAF approach for
any single substance is necessarily determined principally by the availability of relevant data. The
extent of data available is, in turn, often a function of the economic importance of the substance,
and this is frequently related to the extent of potential human exposure.
D.6.b. Probabilistic approaches
Another approach to quantifying uncertainty in RfD or RfC derivation when data are not
sufficient to develop a chemical-specific or biologically based dose-response model is probabilistic
analysis. When the available data are sufficient to meaningfully characterize the distributions of
interest, a probabilistic approach would provide results as a distribution rather than as a single
measure for the dose/concentration-response. For example, distributions could be used for inputs
into a pharmacokinetic model to derive a distribution of internal dose metrics. Also, the
approaches described in the draft IPCS guidance document are amenable to probabilistic analysis
(IPCS, 2001).
Probabilistic analysis for human health assessments generally has been confined to the
exposure variables. In the derivation of human health toxicity reference values, interindividual
variability in pharmacokinetics and pharmacodynamics is usually represented with a UF because
data are insufficient to support a more quantitative representation of these sources of
interindividual variability. The Technical Panel recommends that the Agency further evaluate
approaches such as probabilistic analysis for characterizing variability and uncertainty in toxicity
reference values.
D.7. Summary of key points from a case study on Chemical X
A case study was developed for Chemical X, a hypothetical synthetic halogenated aliphatic
alkene that is a nonflammable, volatile liquid at room temperature. A detailed summary of the
4-46
-------
External Review Draft Do Not Cite or Quote
case study is provided in Appendix B. Below are key sections from the case study that
demonstrate the use of the proposed framework. First, the data are reviewed and characterized
on the basis of the hazard and dose-response data, including consideration of the weight-of-
evidence factors discussed in Section C.2. above. A narrative statement is used to describe the
extent of the database for both inhalation and oral exposure, as well as the gaps in information
that would make the database more robust (Section D.7.a). Dosimetric adjustments are made to
derive HECs, whereas the adjustments for oral exposure are made on a BW1 basis and do not
incorporate the BW3/4 scaling factor or other DAF, as further work is needed on the
harmonization of approaches for derivation of oral and dermal HEDs. The data are presented
both in tabular form (see Tables B-2 and B-3 in Appendix B), and in graphical form (Section
D.7.b) as an exposure response array (Figure 4-3) to provide a visualization of the data applicable
to each duration of exposure. Then, the reference values are derived (Section D.7.c) by
considering all of the relevant data for each duration reference value, developing sample values on
the basis of various endpoints considered for each duration and selecting a final reference value
for each route of exposure and duration on the basis of an evaluation of each of the relevant
endpoints rather than on a single critical study and critical effect (Tables 4-4 and 4-5).
D.7.a. Narrative description of the extent of the database for Chemical X
The database for inhalation exposure is minimal but adequate to derive reference values.
No information is available on possible modes of action or pharmacokinetics. Some human data
on acute, short-term, and longer-term exposures are available, although the range of endpoints
evaluated and the dose-response information for different durations of exposure are limited. The
animal data include acute, short-term, longer-term, and chronic studies with exposures beginning
in young adult animals. The acute and short-term data are limited to clinical signs of morbidity
and mortality; the short-term, longer-term, and chronic studies include some histopathology as
well. There is a study of DNT with prenatal exposure in rats limited to GD 7 - 13 (as opposed to
more extensive exposure throughout a major part of central nervous system (CNS) development,
e.g., GD 6 to PND 11 in the standard DNT study testing protocol). No other studies of prenatal
or postnatal developmental toxicity study were done except for evaluations of survival and
growth in a two-generation reproduction study in rats. The protocol used, however, was one in
which reproductive development (e.g., timing of puberty or anogenital distance) and adult
reproductive function (semen quality, estrous cyclicity) were not evaluated and organ weights
4-47
-------
External Review Draft Do Not Cite or Quote
were not measured. No studies were conducted that considered issues related to the toxicity of
the agent in old age, either from earlier exposures or from exposures in aged animals.
The database for oral exposure is much more limited than the database for inhalation
exposure, with acute data in humans on neurotoxicity at a single, high-dose level and chronic data
on birth defects but no dose information. The animal data are likewise very limited, with a single-
dose acute toxicity study in rats in which clinical signs of morbidity and mortality were evaluated
and subchronic (90-day) and chronic toxicity data in rats and mice that included histopathology.
Prenatal developmental toxicity data were available in rats following exposure on GD 6-19, and
an evaluation of adult neurotoxicity was conducted in mice following postnatal developmental
exposure on days 10-17 of age. No other developmental toxicity data were available, and no
information on reproductive toxicity or adult neurotoxicity was available. No studies were
conducted that considered issues related to the toxicity of the agent in old age, either from earlier
exposures or from exposures in aged animals.
D.7.b. Exposure-response array for Chemical X
In addition to displaying the data in tabular form, an exposure-response array can be a
useful way of visually displaying the data (see Figure 4-3) to show what data are available for
each duration of exposure. The points in the graph are for HECs, based on the dosimetric
adjustments discussed earlier in this chapter. Dosimetric adjustment of the developmental toxicity
inhalation data are included here as was done for other types of toxicity data (see Section D.2.b).
4-48
-------
External Review Draft
Do Not Cite or Quote
Figure 4-3. Exposure-Response Arrays for Inhalation Exposure to Chemical X
Acute Exposure
Short-Term Exposure
40 -
O
LU 30
I
Eon
Q.
Q.
10 -
-^ 0
•
A
A
D A
A
O
LJJ
E
Q.
Q.
L Neuro- Devel Repro
\O tox Tox Tox
Longer-Term Exposure
9E>
9n
LU 15 -
I
E 10 -
Q.
o.
5 -
0 -
A
A
1 •
A A A
A 2 2
A
Neuro- Devel Repro Liver Renal
O
LU
E
Q.
Q.
tox Tox Tox Tox Tox
20
15
10 -
K
n
, A
A
A
" A
Neuro- Devel Repro
tox Tox Tox
Chronic Exposure
25
o r\
15 -
m -
5 -
o J
A
A
1 A t I
a A a a
Neuro- Devel Repro Liver Renal
tox Tox Tox Tox Tox
Q Human NOAEL g Human LOAEL £ Rat NOAEL ± Rat LOAEL Q Mouse NOAEL « Mouse LOAEL
-------
External Review Draft Do Not Cite or Quote
D.7.c. Derivation of reference values for Chemical X
D.T.c.i. Acute exposure
Inhalation exposure. Results of available studies indicate that acute inhalation exposure
to Chemical X can result in neurotoxic effects in human adults with a LOAELjjgc of 50 ppm
(NOAELjjEc of 4 ppm). Animal studies also show that Chemical X has the potential to cause
DNT and reproductive effects at comparable HECs with LOAELj^cS of 900 and 1000 ppm
(NOAELjjEc of 2.5 ppm and 5 ppm), respectively.
Because animal studies indicate that the developing nervous system is vulnerable to
Chemical X exposure and the NOAELjjEc for that endpoint is most protective, a NOAELjjEc of
2.5 ppm in the developmental toxicity study is used as the basis for deriving an acute reference
value for inhalation exposure. Default UFs of 101/2 (animal-to-human extrapolation), 10
(interindividual differences), and 101/2 (database deficiencies: no adequate prenatal developmental
toxicity studies in two species, no adequate DNT study) are applied. The resultant reference
value for acute inhalation exposure is 0.03 ppm (Table 4-4).
Oral exposure. Acute oral exposure to Chemical X can result in neurotoxic effects in
human adults (LOAEL of 100 mg/kg/d). However, dose-response data are not available in
humans. A single study in mice indicates that Chemical X (dosing on PND 10 - 17; equivalent to
approximately 1 month to l-!/2 years of age in humans) also has the potential to cause DNT, with
a LOAEL of 300 mg/kg/d (NOAEL of 50 mg/kg/d). Applying default UFs of 10 and 10 to
account for animal-to-human extrapolation and interindividual differences, as well as a database
UF of 10 due to the limitations of the available data, results in a reference value for acute oral
exposure of 0.05 mg/kg (Table 4-5).
D.7.c.ii. Short-term exposure
Inhalation exposure. The reference value for short-term inhalation exposure is based on
the human data as well as the DNT and reproductive toxicity data with NOAELjjgcS of 2 ppm, 2.5
ppm and 5 ppm, respectively (LOAELjjEcS of 10 ppm, 100 ppm, and 300 ppm). Using the human
NOAELjjEc of 2 ppm and applying a 10-fold default UF for intraspecies uncertainty and variability
and a 101/2-fold UF for database deficiencies would result in a reference value for short-term
inhalation exposure of 0.07 ppm. However, a default factor of 101/2 (interspecies), 10
(intraspecies) and 101/2 (database deficiencies) would be applied to the HECs for the animal data
on DNT and reproductive toxicity (2.5 and 5 ppm, respectively), resulting in reference values of
4-50
-------
External Review Draft Do Not Cite or Quote
0.03 and 0.05 ppm. Given the close range of values, the reference value of 0.03 ppm, would be
used because it is more protective of the developing individual as well as the adult (Table 4-4).
Oral exposure. The reference value for short-term oral exposure would be the same as for
acute exposure, which is based on DNT as discussed above (Table 4-5).
D.T.c.iii. Longer-term exposure
Inhalation exposure. Subchronic and chronic inhalation exposure to Chemical X can
result in multiple health effects. Available studies demonstrate neurotoxicity in adult humans.
However, dose-response information is not available, and the presumed LOAEL (20 ppm) for
neurotoxicity in humans is somewhat higher than the HECs for other health endpoints
(developmental, reproductive, liver, and renal effects) observed in animal studies, where the
LOAELjjEcS range from 7 ppm to 22.5 ppm (NOAELjjgcS range from 2.5 ppm to 5 ppm). Dose-
response data for these health endpoints in animal studies can be used as the basis for deriving a
longer-term inhalation reference value for Chemical X. UFs of 101/2 (interspecies), 10
(intraspecies), and 101/2 (database deficiencies) were applied to NOAELjjgcS for the various
endpoints in deriving sample reference values. If an additional factor of 3 were applied to the rat
developmental toxicity data to account for the marked difference in exposure duration in the study
itself (7 days of exposure: GD 7 - 13), a longer-term sample reference value of 0.01 ppm would
result. Without this additional factor, the HEC from the developmental toxicity study was still the
lowest value (0.03 ppm) and was used in the reference value derivation (Table 4-4). Whether an
additional factor should be applied to the developmental toxicity data or to other data of much
shorter duration should be explored further.
Oral exposure. Available animal data indicate that longer-term oral exposure to Chemical
X can cause liver, renal, and developmental effects, with LOAELs ranging from 300-2114
mg/kg/d (NOAELs ranging from 50 mg/kg/d to 71 mg/kg/d). Application of default UFs of 10
(interspecies), 10 (intraspecies) and 10 (database deficiencies) to the data from the subchronic
studies would result in longer-term oral reference values of 0.05 mg/kg/d (Table 4-5). If an
additional factor of 3 was applied to the mouse developmental toxicity data to account for short-
term to longer-term exposure and a total UF of 3000 applied, a sample reference value of 0.02
mg/k/d would be calculated, which is less than the other values derived from subchronic exposure
data. As indicated above, whether an additional factor should be applied to the developmental
toxicity data or to other data of much shorter duration should be explored further.
4-51
-------
External Review Draft Do Not Cite or Quote
D.T.c.iv. Chronic exposure
Inhalation exposure. For the chronic inhalation reference value, the NOAELjjgcS range
from 2 ppm to 5 ppm (LOAELj^s range from 10 ppm to 300 ppm), and UFs of 101/2
(interspecies), 10 (intraspecies), and 101/2 (database deficiencies) applied to the chronic exposure
NOAELjjEcS for neurotoxicity, and liver and kidney toxicity, and reproductive toxicity data results
in sample reference values of 0.02 - 0.05 ppm (Table 4-4). Applying these UFs to the NOAELj^c
of 2.5 ppm for developmental toxicity yields a sample reference value of 0.03 ppm, falling within
the range of chronic study-based values. In this example, the chronic study neurotoxicity is the
limiting endpoint, providing a chronic inhalation reference value of 0.02 ppm. If, contrary to
current practice, an additional 10 for subchronic to chronic duration were applied to the
developmental NOAELjjgc, the resultant sample reference value would be 0.003 ppm. As
mentioned in Section D.S.f, this issue may need further exploration.
Oral exposure. As chronic dosing studies are available with NOAELs of 36 mg/kg/d
(LOAELs of 50 mg/kg/d), application of default UFs of 10 (interspecies), 10 (intraspecies) and 10
(database deficiencies) would result in chronic oral reference values of 0.04 mg/k/d (Table 4-5).
Applying these same factors to the developmental NOAEL of 50 mg/kg/d yields a slightly higher
value of 0.05 mg/kg/d. Applying an additional UF of 10 to the mouse developmental toxicity data
to account for the difference between short-term and chronic exposure would result in a total UF
of 10,000.
4-52
-------
External Review Draft Do Not Cite or Quote
Table 4-4. Derivation of Reference Values for Chemical X - Inhalation Exposure
Exposure
Duration
Acute
Short-term
Longer-term
Chronic
HEC
(ppm)
4
2.5
5
2
2.5
5
20Ld
4
5
2.5
5
4
5
4
5
20L
2
2
2.5
5
2
2
2
2
Species
Human
Rat
Rat
Human
Rat
Rat
Human
Rat
Mouse
Rat
Rat
Rat
Mouse
Rat
Mouse
Human
Rat
Mouse
Rat
Rat
Rat
Mouse
Rat
Mouse
Type of
Effect3
NT
DT
RT
NT
DT
RT
NT
NT
NT
DT
RT
LT
LT
KT
KT
NT
NT
NT
DT
RT
LT
LT
KT
KT
Uncertainty Factors'"
Total
30
100
100
30
100
100
300
100
100
100
100
100
100
100
100
300
100
100
100
100
100
100
100
100
A
1
o
J
3
1
3
o
5
i
o
J
3
o
J
3
o
J
3
o
J
3
1
3
o
J
3
o
J
3
o
J
3
o
3
H
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
L
1
1
1
1
1
1
10
1
1
1
1
1
1
1
1
10
1
1
1
1
1
1
1
1
s
1
1
1
1
1
1
1
1
1
le
1
1
1
1
1
1
1
1
le
1
1
1
1
1
D
3
o
J
3
o
J
Reference Value
(ppm)c
Sample
0.1
0.03
0.05
0.07
0.03
0.05
0.07
0.04
0.05
0.03
0.05
0.04
0.05
0.04
0.05
0.07
0.02
0.02
0.03
0.05
0.02
0.02
0.02
0.02
Final
0.03
0.03
0.03
0.02
a NT = neurotoxicity; DT = developmental toxicity; RT = reproductive toxicity; LT = liver toxicity; KT = kidney
toxicity
b A = animal-to-human (interspecies); H = interindividual (intraspecies); L = LOAEL-to-NOAEL; S = subchronic-to-
chronic duration; D = database deficiency
0 Sample = Reference value based on that particular endpoint, species, duration; Final = reference value for the entire
database for a particular duration of exposure.
dL indicates that this value is the HEC based on the LOAEL.
eA duration UF was not applied to the data from the developmental toxicity study, but should be considered further.
4-53
-------
External Review Draft Do Not Cite or Quote
Table 4-5. Derivation of Reference Values for Chemical X - Oral Exposure
Exposure
Duration
Acute
Short-term
Longer-term
Chronic
HED
(mg/k/d)
100Ld
1000L
900L
50
900L
50
900L
50
71
71
900L
50
36
36
36
36
Species
Human
Rat
Rat
Mouse
Rat
Mouse
Rat
Mouse
Rat
Mouse
Rat
Mouse
Rat
Mouse
Rat
Mouse
Type of
Effect3
NT
NT
DT
DT
DT
DT
DT
DT
LT
LT
DT
DT
LT
LT
KT
KT
Uncertainty Factors'5
Total
~
-
-
1000
~
1000
~
1000
1000
1000
~
1000
1000
1000
1000
1000
A
-
~
~
10
-
10
-
10
10
10
-
10
10
10
10
10
H
~
-
-
10
~
10
~
10
10
10
~
10
10
10
10
10
L
-
~
~
1
-
1
-
1
1
1
-
1
1
1
1
1
s
~
-
-
1
~
1
~
lf
1
1
~
lf
1
1
1
1
D
10
10
10
10
Reference Value
(mg/k/d)c
Sample
No D-Re
NoD-R
NoD-R
0.05
NoD-R
0.05
NoD-R
0.05
0.07
0.07
NoD-R
0.05
0.04
0.04
0.04
0.04
Final
0.05
0.05
0.05
0.04
3 NT = neurotoxicity; DT = developmental toxicity; RT = reproductive toxicity; LT = liver toxicity; KT =
kidney toxicity
b A = animal-to-human (interspecies); H = interindividual (intraspecies); L = LOAEL-to-NOAEL; S =
subchronic-to-chronic duration; D = database deficiency
0 Sample = Reference value based on that particular endpoint, species, duration; Final = reference value for
the entire database for a particular duration of exposure.
dL indicates that this value is the HED based on the LOAEL.
eNo D-R = no dose-response data; usually only one dose in the study.
fA duration UF was not applied to the data from the developmental toxicity study, but should be considered
further.
4-54
-------
External Review Draft Do Not Cite or Quote
CHAPTER 5
RECOMMENDATIONS
A number of recommendations have been made in other parts of this report. This chapter
summarizes those recommendations, based on the Technical Panel's review of the RfD and RfC
process. The Technical Panel assumes that some of the recommendations will be possible to
implement in the near future, given adequate resources and personnel, whereas others will require
additional effort. In particular, the development of additional or alternative testing guidelines is
the responsibility of OPPTS, which together with scientists in other parts of the Agency will
develop such guidelines as part of the Harmonized Health Effects Test Guidelines (870 Series).
In addition, testing strategies are needed to determine when to implement new testing guidelines
in the process of developing a data package on a particular chemical.
As part of its deliberations, the Technical Panel has considered the recommendations of
the Toxicology Working Group of the 10X Task Force (EPA, 1999a, and Appendix A). The
Technical Panel endorses those recommendations and extends and expands them to deal with a
broader view of life stages, timing and duration of exposure, and evaluation of endpoints, both
structural and functional. The recommendations are presented here in the order of the chapters in
which they appear. In many cases, further discussion of the specific recommendations can be
found in the earlier chapters.
Chapter 2
The Technical Panel concurred with the recommendation of the 10X Task Force that
reference values should be derived, where possible, for acute, short-term, and longer-term as well
as chronic exposures for oral, dermal, and inhalation routes and that they be included in the IRIS
database for use by EPA programs, where applicable. The definitions for duration should be
standardized but left flexible so they can be adjusted depending on the exposure situation of
concern.
Chapter 3
The Technical Panel reviewed and evaluated current testing guidelines and approaches
testing approaches as a follow-up to its recommendation in Chapter 2 concerning the derivation
5-1
-------
External Review Draft Do Not Cite or Quote
of less-than-lifetime reference values. This review was undertaken to determine what information
is currently gathered with regard to life stage assessment, endpoint assessment, route and duration
of exposure, and latency to response. A primary goal of this review was to provide a basis for
recommendations for the development of innovative alternative testing approaches and the use of
such data for risk assessment. The Technical Panel is not recommending additional testing for
every chemical but is suggesting that alternative strategies and guidance for testing approaches be
developed that incorporate information on pharmacokinetics and mode of action early in the
process, thus allowing a more targeted testing approach. In addition, alternative protocols are
discussed that are aimed at more efficient use of animals and resources in combined studies that
would provide more extensive data on life stages, endpoints and other factors not well
characterized in current testing approaches. Recommendations were also made about research
areas that should be encouraged to aid in better study design and interpretation of data for risk
assessment.
1. Develop OPPTS guideline study protocols for acute and short-term studies that provide
more comprehensive data for setting reference values.1 Develop guidance for how and when to
use the guidelines.
2. Modify existing OPPTS guideline study protocols to provide more comprehensive
coverage of life stages for both exposure and outcomes. Develop guidance for how and when to
use the guidelines. Several alternative protocol options are discussed.
3. Encourage research to evaluate latency to effect and reversibility of effect from less-than-
lifetime exposures.
4. Develop OPPTS guideline study protocols for more systematic information on
pharmacokinetics, including at different life stages. Develop guidance for how and when to use
the guidelines.
5. Encourage research on mechanisms/modes of action and pharmacodynamics for
application in reference value derivation.
6. Develop OPPTS guideline study protocols to more fully assess all types of toxicity,
particularly immunotoxicity, carcinogenicity, neurotoxicity, and cardiovascular toxicity at
different life stages. Develop guidance for how and when to use the guidelines.
programs that can require testing, the OPPTS guideline study protocols should be
used.
5-2
-------
External Review Draft Do Not Cite or Quote
1. Explore the feasibility of setting dermal reference values for direct toxicity at the portal of
entry, including sensitization.
Chapter 4
The Technical Panel discussed a number of modifications to the existing framework for
reference value derivation in Chapter 4, both for the current chronic reference values (RfD and
RfC) as well as for the acute, short-term, and longer-term reference values. In addition, a case
study illustrating many of these concepts is summarized in Chapter 4 and provided in detail in
Appendix B. The recommendations for improvement and expansion of the existing approaches
are aimed at taking a broader approach to the characterization of the entire database and what
impact that will have on the dose-response assessment and risk characterization of a chemical.
These include recommendations for setting several less-than-lifetime reference values, more
broadly characterizing the database instead of using a checklist of a minimum set of studies for
setting a reference value, using an exposure-response array and carrying through the derivation of
sample reference values for all relevant endpoints before deciding which endpoint(s) to use for the
POD, deriving reference values in a way that is protective of all relevant endpoints rather than
setting reference values on particular endpoints (e.g., the RfDDT), but using a process that
facilitates the evaluation of risk to particular subgroups for specific program office needs,
including cumulative risk assessment.
The specific recommendations follow.
1. Include the acute, short-term, longer-term, and chronic reference values derived on the
basis of the recommendations in this report in IRIS after appropriate internal, external, and
consensus review.
2. Use consistent definitions for the duration of exposure in deriving acute, short-term,
longer-term, and chronic reference values.
3. Use the revised definitions for the reference values shown in Chapter 4. These definitions
are aimed at clarifying (a) that the reference value is intended to provide an estimate that is
centered within an order of magnitude, further emphasizing that the estimate is not a bright line,
but has some range of variability that may be considered by risk managers in decision making; (b)
the term "deleterious" has been replaced with the term "adverse," because the latter is more
commonly used and understood in data evaluation and selection of endpoints for setting reference
5-3
-------
External Review Draft Do Not Cite or Quote
values; and (c) the term "noncancer" has been removed from the definitions in the spirit of
harmonization of risk assessment approaches for human health effects because it has been
recommended that health effects no longer be categorized as "cancer" or "noncancer" for the
purposes of hazard characterization and dose-response analysis. This change denotes the move
toward defining approaches for low dose estimation or extrapolation based on mode of action.
4. For consistency in the designation of various duration reference values, the Panel
recommends that the terminology for reference values be standardized; this standardized
terminology should be reflect both duration and route of exposure. Consistent terminology
recommendations for reference values are proposed in this report, but additional suggestions are
sought.
5. The Technical Panel recommends that endpoint-specific reference values not be
developed, including the RfDDT, as originally proposed in the Guidelines for Developmental
Toxicity Risk Assessment (EPA, 1991). Rather, all relevant endpoints should be considered in
the derivation of various duration reference values that are applicable, and the reference values
should be derived to be protective of all types of effects for that duration of exposure.
6. An expanded approach to the evaluation of studies and characterization of the extent of
the database as a whole is recommended; in particular, several factors are discussed that should be
considered in a weight-of-evidence approach for characterizing hazard for the population as a
whole as well as for potentially susceptible subpopulations. Those considerations for assessing
level of concern raised by the Toxicology Working Group of the 10X Task Force (EPA, 1999a)
have been incorporated into this approach.
7. A narrative approach should be used in describing the extent of the database instead of
using a confidence ranking of high, medium, or low. The extremes for the extent of the database,
i.e., minimal or robust, are defined in Chapter 4. The narrative approach is intended to emphasize
the types of data available (both human and animal data) as well as the data gaps that could
improve the derivation of reference values, and it should encourage a wider range of information
to be used in deriving reference values, taking into consideration the life stages evaluated, the
issues of timing, duration and route of exposure, the types and extent of endpoint assessment (i.e.,
structural and function), and the potential for latent effects and/or reversibility of responses.
8. Duration adjustment procedures to continuous exposures for inhalation developmental
toxicity studies should be done in the same way as for other health endpoints.
5-4
-------
External Review Draft Do Not Cite or Quote
9. Additional consideration of the HEC and the HED derivation methodology is needed to
confirm or assess the relevance for all population subgroups (particularly including children).
10. An exposure-response array should be used as a visual display of all relevant endpoints
and durations of exposure, as shown in the case study. This array can be used to evaluate the
range of exposure-response data for different durations of exposure in order to determine the
range of numerical values available for each route and duration of exposure.
11. The POD should be selected on the basis of an evaluation of all relevant endpoints carried
through to reference value derivation with selection of the limiting value(s) as the final step rather
than on a single "critical study" and "critical effect."
12. Sound scientific judgment should be used in the application of UFs to derive reference
values which are applied to the value chosen for the POD derived from the available database
(BMDL, NOAEL, or LOAEL). Although default factors of 10 are recommended, with 3 used in
place of half-power values (i.e., 10°-5) when occurring singly, the exact value of the UF chosen
should depend on the quality of the studies available, the extent of the database, and scientific
judgment. It is imperative that the IRIS documentation contain a justification for the individual
factors selected for a particular chemical because rigid application of UFs could lead to an
illogical set of reference values.
13. The Technical Panel recommends limiting the total UF applied to a chronic reference value
for any particular chemical to 3,000. If there is uncertainty in more than four areas of
extrapolation, it is unlikely that the database is sufficient to derive a reference value, and would
need to be carefully evaluated in the case of uncertainty in four areas. This maximum of 3,000
applies only to the UFs and does not include the various DAFs discussed in Chapter 4.
14. The Technical Panel supports and expands the recommendation of the Toxicology
Working Group of the 10X Task Force (EPA, 1999a) that reduction of the intraspecies UF be
considered only if data are sufficient to support the conclusion that the data set on which the POD
is based is representative of the exposure/dose-response data for the susceptible subpopulation(s).
Given this, whether and how much the intraspecies UF may be reduced must be linked to how
completely the susceptible subpopulation has been identified and its susceptibility described (e.g.,
versus assumed). At the other extreme, a 10-fold factor may sometimes be too small because of
factors that can influence large differences in susceptibility, such as genetic polymorphisms. The
Technical Panel urges the development of data to support the selection of the appropriate size of
5-5
-------
External Review Draft Do Not Cite or Quote
this factor, but recognizes that often there are insufficient data to support a factor other than the
default.
15. The Technical Panel urges continued research and evaluation of the similarities and
differences between the general population and susceptible subpopulations in their responses to
particular agents, particularly children and the elderly. From such evaluations, the protectiveness
of the tenfold default factor can continue to be assessed.
16. Given that there are several UFs that can be used to deal with data deficiencies as part of
the current reference value process, and given that these are assumed to overlap to some extent,
the Technical Panel agrees with the 10X Task Force Toxicology Working Group (EPA, 1999a)
that the current interspecies, intraspecies, and database deficiency UFs, if appropriately applied
using the approaches recommended in this review, will be adequate in most cases to cover
concerns and uncertainties about children's health risks. Rather, any residual concerns about
toxicity and/or exposure can be dealt with in risk characterization/risk management (e.g., by
retention of all or part of the FQPA safety factor for pesticides).
17. The Panel considers the purpose of the MF to be sufficiently subsumed in the general
database UF. Therefore, the Panel recommends the discontinuance in use of the MF.
18. The EPA has not yet established guidance for the use of specific data to replace UFs (i.e.,
CSAFs), but the division of UFs into pharmacodynamic and pharmacokinetic components has
been used in the RfC methodology (EPA, 1994). The Agency is encouraged to develop its own
guidance, based on some of the available methodologies (e.g., IPCS), but caution should be used
in that there are relatively few data available for many substances to serve as an adequate basis to
replace defaults with CSAFs.
The following issues were discussed by the Technical Panel but were considered more
appropriate for discussion and recommendation by other panels/committees.
1. There have been inconsistencies in the use of BMD modeling approaches to deriving RfDs
and RfCs currently in IRIS. The Technical Panel was unable to fully evaluate these issues or to
reach agreement about any recommendation for change to current methodology and recommends
that they be considered further by the Agency. The Technical Panel also recommends that factors
such as the response rates at the BMD or NOAEL, the power of the study, and slope of the dose-
response curve be included in the description of the database, where possible, as part of risk
characterization.
5-6
-------
External Review Draft Do Not Cite or Quote
2. The Technical Panel recommends harmonization of the approaches for HEC and HED
derivation for all types of health effects. Development of the appropriate adjustment procedure is
referred to the Harmonization Framework Technical Panel.
3. The Technical Panel recommends that the Agency further evaluate approaches such as
probabilistic analysis for characterizing variability and uncertainty in toxicity reference values.
4. The Technical Panel recommends further evaluation of appropriate adjustment of doses
for duration of exposure. The method derived from ten Berge et al. (1986) is raised as a
possibility for acute exposures on the basis of its recommendation in the ARE methodology.
Duration adjustment for short-term and longer-term reference values analogous to the subchronic
to chronic duration UF for chronic reference values is raised in the case study and should be
explored further.
5-7
-------
External Review Draft Do Not Cite or Quote
REFERENCES
Abdel-Mageed, MA; Suh, DH; Abdel-Rahman, MS. 2001. Intra-species extrapolation in risk
assessment of different classes of antimicrobials. Human Ecolog. Risk Assess., 7:15-36.
Allen, BC; Kavlock, RJ; Kimmel, CA; Faustman, EM. 1994a. Dose-response assessment for
developmental toxicity: II. Comparison of generic benchmark dose estimates with NOAELs.
Fund. Appl. Toxicol., 23:487-495.
Andersen, ME; Gargas, ML; Clewell, HJ, III; Severyn, KM. 1987. Quantitative evaluation of the
metabolic interactions between trichloroethylene and 1,1-dichloroethylene in vivo using gas
uptake methods. Toxicol. Appl. Pharmacol., 89:149-57.
ATSDR (Agency for Toxic Substances and Disease Registry). 1996. Minimal Risk Levels for
Priority Substances and Guidance for Derivation. Fed. Regist. 61:25873-25881.
Barton, HA; Deisinger, PJ; English, JC; Gearhart, JN; Faber, WD; Tyler, TR; Banton, MI;
Teeguarden, J; Andersen, ME. 2000 Family approach for estimating reference
concentrations/doses for series of related organic chemicals. Toxicol Sci., 54(1):251-61.
Bogdanffy, MS; Daston, G; Faustman, EM; Kimmel, CA; Kimmel, GL; Seed, J; Vu, V. 2001.
Harmonization of cancer and non-cancer risk assessment: proceedings of a consensus-building
workshop. Toxicol. Sci., 61:18-31.
Calabreese, EJ. 2001. Assessing the default assumption that children are always at risk. Human
Ecolog. Risk Assess., 7:37-59.
The CDM Group, Inc. 2000. Colloquium on Setting Acute and Intermediate Reference Values,
Summary Report, September 22. Contract Number 68-D6-0060. Chevy Chase, MD.
Creason, J; Neas, L; Walsh, D; Williams, R; Sheldon, L; Liao, D; Shy, C. 2001. Paniculate
mater ans heart rate variability among elderly retirees: the Baltimore 1998 PM study. J.
Expo. Anal. Environ. Epidemiol., 11(2): 116-122.
Dourson, ML; Knauf, LA; Swartout, JC. 1992. On Reference Dose (RfD) and its underlying
toxicity database. Toxicol. Ind. Health, 8:171-189.
Dourson, ML; Felter, SP; Robinson, D. 1996. Evolution of science-based uncertainty factors in
noncancer risk assessment. Regul. Toxicol. Pharmacol. 24:108-120.
R-l
-------
External Review Draft Do Not Cite or Quote
Dourson, ML; Andersen, ME; Erdreich, LS; MacGregor, JA. 2001. Using human data to protect
the public's health. Regulat. Toxicol. Pharmacol., 33:234-256.
Enquist, BJ; Brown, JH; West, GB. 1998. Allometric scaling of plant energetics and population
density. Nature, 395 (No. 6698): 163-165.
EPA (U.S. Environmental Protection Agency). 1986. Guidelines for Health Risk Assessment of
Chemical Mixtures. Fed. Regist, 51: 34014-34025. Available at http://www.epa.gov/ncea/raf/.
EPA (U.S. Environmental Protection Agency). 1988. Recommendations for and documentation
of biological values for use in risk assessment. EPA/600/6-87/008. Environmental Criteria and
Assessment Office, Office of Health and Environmental Assessment, ORD/EPA, Cincinnati, OH.
Available from the National Technical Information Service, Springfield, Virginia 22161, (703)
605-6000, http://www.ntis.gov.
EPA (U.S. Environmental Protection Agency). 1991. Guidelines for Developmental Toxicity
Risk Assessment; Notice. Fed. Regist., 56:63798-63826. Available at
http://www.epa.gov/ncea/raf/.
EPA (U.S. Environmental Protection Agency). 1992. Draft Report: a cross-species scaling
factor for carcinogen risk assessment based on equivalence of mg/kg3/4 /day. Federal Register 57,
Issue 109, p. 24152.
EPA (U.S. Environmental Protection Agency). 1994. Methods for Derivation of Inhalation
Reference Concentrations and Application of Inhalation Dosimetry, EPA/600/8-90/066F.
Available from the U.S. EPA Risk Information Hotline at telephone 1-513-569-7254, fax 1-513-
569-7159, email RIH.IRIS@epamail.epa.gov and/or the National Technical Information Service,
Springfield, VA 22161, (703) 605-6000, http://www.ntis.gov.
EPA (U.S. Environmental Protection Agency). 1996. Guidelines for Reproductive Toxicity Risk
Assessment. Fed. Regist, 61:56274-56322. Available at http://www.epa.gov/ncea/raf/.
EPA (U.S. Environmental Protection Agency). 1997a. Exposure Factors Handbook, Vols. I, II,
and III. EPA/600/P-95/002Fa. Available at http://www.epa.gov/ncea/.
EPA (U.S. Environmental Protection Agency). 1997b. Guiding principles for Monte Carlo
analysis. Risk Assessment Forum. EPA/630/R-97/001. Available from National Technical
Information Service, Springfield, VA; PB97-188106 and http://www.epa.gov/ncea/raf/.
EPA (U.S. Environmental Protection Agency). 1997c. Summary of the U.S. EPA Colloquium
on a Framework for Human Health Risk Assessment, Vol. 1, EPA/600/R-99/001. Available at
http://www.epa.gov/nceawwwl/colloquium.htm.
R-2
-------
External Review Draft Do Not Cite or Quote
EPA (U.S. Environmental Protection Agency). 1998a. Summary of the U.S. EPA Colloquium
on a Framework for Human Health Risk Assessment, Vol. 2, EPA/600/R-98/155. Available at
http://www.epa.gov/nceawwwl/colloquium.htm.
EPA (U.S. Environmental Protection Agency). 1998b. Methods for Exposure-Response
Analysis for Acute Inhalation Exposure to Chemicals, Development of the Acute Reference
Exposure. EPA/600/R-98/051.
EPA (U.S. Environmental Protection Agency). 1998c. Guidelines for Neurotoxicity Risk
Assessment. Fed. Regist. 63:26926-26954. Available at http://www.epa.gov/ncea/raf/.
EPA (U.S. Environmental Protection Agency). 1999a. Toxicology Data Requirements for
Assessing Risks of Pesticide Exposure to Children's Health - Report of the Toxicology Working
Group of the 10X Task Force (April 28, 1999 draft). Available at
http://www.epa.gov/scipoly/sap/1999/index.htmtfmay.
EPA (U.S. Environmental Protection Agency). 1999b. Exposure Data Requirements for
Assessing Risks of Pesticide Exposure to Children's Health - Report of the Exposure Working
Group of the 10X Task Force (March 8, 1999 draft). Available at
http://www.epa.gov/scipoly/sap/1999/index.htmtfmay.
EPA (U.S. Environmental Protection Agency). 1999c. The Office of Pesticide Programs' Policy
on Determination of the Appropriate FQPA Safety Factor(s) for Use in the Tolerance-Setting
Process (May 10, 1999 draft). Available at http://www.epa.gov/scipoly/sap/1999/index.htmtfmay.
EPA (U.S. Environmental Protection Agency). 1999d. Guidelines for Carcinogen Risk
Assessment. SAB Review Draft, July, NCEA-F-0644. Available at
http://www.epa.gov/ncea/raf/.
EPA. (U.S. Environmental Protection Agency). 2000a. Science Policy Handbook on Risk
Characterization, EPA 100-B-00-002. Available at http://www.epa.gov/ORD/spc/2riskchr.htm.
EPA. (U.S. Environmental Protection Agency). 2000b. Benchmark Dose Technical Guidance
Document, EPA/63O/R-00/001, External Review Draft, October. Available at
http ://www. epa. gov/ncea/bnchmrk/bmds_peer. htm.
EPA. (U.S. Environmental Protection Agency). 2000c. Supplementary Guidance for Conducting
Health Risk Assessment of Chemical Mixtures, EPA/630/R-00/002. 01 August 2000. Available at
http ://cfpub. epa. gov/ncea/cfm/.
R-3
-------
External Review Draft Do Not Cite or Quote
EPA. (U.S. Environmental Protection Agency). 2001a. Toxicological Review of Boron and
Compounds (CAS No. 7440-42-8), External Review Draft. Available at
http://www.epa.gov/ncea/boron.htm.
EPA. (U.S. Environmental Protection Agency). 2001b. Press release: Agency requests National
Academy of Sciences input on consideration of certain human toxicity studies; announces interim
policy. Friday, December 14 (contact David Deegan, 202-564-7839; deegan.david@epa.gov).
EPA (U.S. Environmental Protection Agency). 2002a. Integrated Risk Information System
(IRIS) online: http://www.epa.gov/iris/.
EPA (U.S. Environmental Protection Agency). 2002b. Determination of the Appropriate FQPA
Safety Factor(s) for Use in the Tolerance-Setting Process. Office of Pesticide Programs, Office of
Prevention, Pesticides, and Toxic Substances. Available at
http: //www. epa. gov/oppfead 1 /trac/sci ence/# 10_fol d.
EPA 2002c. Framework for Cumulative Risk Assessment (EXTERNAL REVIEW DRAFT). 23
APRIL 2002. U.S. Environmental Protection Agency, Risk Assessment Forum, Washington, DC,
120 pp. Available at http://cfpub.epa.gov/ncea/cfm/nceahome.cfm.
Faustman, EM; Allen, BC; Kavlock, RJ; Kimmel, CA. 1994. Dose-response assessment for
developmental toxicity: I. Characterization of data base and determination of NOAELs. Fund.
Appl. Toxicol., 23:478-486.
Gad, SC; Chengelis, CP. 1998. Acute Toxicology Testing, 2nd edition, Academic Press, San
Diego, 534pp.
Gargas, ML; Burgess, RJ; Voisard, DE; Cason, GH; Andersen, ME. 1989. Partition coefficients
of low-molecular-weight volatile chemicals in various liquids and tissues. Toxicol. Appl.
Pharmacol., 98:87-99.
Gillooly, JF; Brown, JH; West, GB; Savage, VM; Charnov, EL. 2001. Effects of size and
temperature on metabolic rate. Science, 293:2248-2251.
Haseman, JK. 1984. Statistical issues in the design, analysis and interpretation of animal
carcinogenicity studies. Environ. Health Perspect, 58:385-92.
Haseman, JK; Huff, JE; Rao, GN; Eustis, SL. 1989. Sources of variability in rodent
carcinogenicity studies. Fundam. Appl. Toxicol., 12:793-804.
R-4
-------
External Review Draft Do Not Cite or Quote
Hanna, LM; Jarabek AM. 2000. Proposed Inhalation Route Model Suites:
Mode-of-Action-Motivated Dosimetry Project. Society of Risk Analysis, 2000 Annual Meeting
(abstract).
Hoyer, PB; Sipes, IG. 1996. Assessment of follicle destruction in chemical-induced ovarian
toxicity. Annu. Rev. Pharmacol. Toxicol., 36:307-331.
Hubal, EA; Bunge, A; Hanna, LM; McDougal, JN; Jarabek, AM. 2000. Proposed Suite of
Models for Estimating Dose Resulting from Exposures by the Dermal Route. Society of Risk
Analysis, 2000 Annual Meeting (abstract).
ICRP (International Commission on Radiological Protection). 1994. Human Respiratory Tract
Model for Radiological Protection: a report of a task group of the International Commission on
Radiological Protection. Oxford, United Kingdom: Elsevier Science Ltd. ICRP Publication no
66., Annals of the ICRP; 24 (1-3), Pergamon Press, Elsevier Science, Tarrytown, NY.
ILSI RSI (International Life Sciences Research Institute Risk Sciences Institute). 2001.
Workshop to Develop a Framework for Assessing Risks to Children from Exposure to
Environmental Agents, Stowe, VT, July 30-August 2.
IPCS (International Programme on Chemical Safety). 1994. Assessing human health risks of
chemicals: Derivation of guidance values for health-based exposure limits. Geneva, World Health
Organization, IPCS (Environmental Health Criteria No. 170).
IPCS (International Programme on Chemical Safety). 1995. IPCS/OECD Workshop on the
Harmonization of Risk Assessment of Reproductive and Development Toxicity. BIBRA
International, Carshalton, Surrey, UK, October 17-21, 1994, IPCS/95.25.
IPCS (International Programme on Chemical Safety). 1998. Boron. Geneva, World Health
Organization, IPCS (Environmental Health Criteria No. 204).
IPCS (International Programme on Chemical Safety). 2001. Guidance Document for the Use of
Data in Development of Chemical-Specific Adjustment Factors (CSAF) for Inter-species
Differences and Human Variability in Dose/Concentration Response Assessment. WHO/
PCS/01.4, July, draft final. Available at http://www.ipcsharmonize.org/CSAFsummary.htm.
Jarabek, AM. 2000. Mode of Action: Framework for Dosimetry Model Development. Society
of Risk Analysis, 2000 Annual Meeting (abstract).
Jepson, GW; Hoover, DK; Black, RK; McCafferty, JD; Mahle, DA; Gearhart, JM. 1994. A
partition coefficient determination method for nonvolatile chemicals in biological tissues.
Fundam. Appl. Toxicol., 22:519-24.
R-5
-------
External Review Draft Do Not Cite or Quote
Tick, H; Porter, J; Morrison, AS. 1977. Relation between smoking and age of natural
menopause. Lancet, 1:1354-55.
Kimbell, JS; Gross, EA; Joyner, DR; Godo, MN; Morgan, KT. 1993. Application of
computational fluid dynamics to regional dosimetry of inhaled chemicals in the upper respiratory
tract of the rat. Toxicol. Appl. Pharmacol., 121:253-263.
Kimbell, JS; Godo, MN; Gross, EA; Joyner, DR; Richardson, RB; Morgan, KT. 1997.
Computer simulation of inspiratory airflow in all regions of the F344 rat nasal passages. Toxicol.
Appl. Pharmacol., 145:388-398.
Kimmel, CA; Francis, EZ. 1990. Proceedings of the workshop on the acceptability and
interpretation of dermal developmental toxicity studies. Fund. Appl. Toxicol., 14:386-398.
Lamb, JC. Reproductive toxicity testing: Evaluating and developing new testing systems. J. Am.
Coll. Toxicol., 4:163-171, 1985.
Lamb, JC, IV; Maronpot, RR; Gulati, DK; Russell, VS; Hommel, L; Sabharwal, PW. 1985.
Reproductive and developmental toxicity of ethylene glycol in the mouse. Toxicol. Appl.
Pharmacol., 81:169-173.
Leopold, PW.; Rivens, ML; Leach, MO. 1997. Noninvasive diagnostic techniques in vascular
disease. J. Vase. Technol., 11:183-186.
Lewis, SC; Lynch, JR; Nikiforov, AI. 1990. A new approach to deriving community exposure
guidelines from "no-observed-adverse-effect-levels." Reg. Toxicol. Pharmacol., 11: 314-330.
Meek, ME. 2001. Categorical default uncertainty factors - interspecies variation and adequacy of
database. Hum. Ecol. Risk. Assess., 7:157-163.
Meek, ME; Renwick, A; Ohanian, EO; Dourson, M; Lake, B; Naumann, B; Vu, V. 2001.
Guidelines for application of Compound Specific Adjustment Factors (CSAF) in
dose/concentration response Assessment. Comments in Toxicology (in press).
Morrissey, RE; Lamb, JC, IV; Morris, RW; Chapin, RE; Gulati, DK; Heindel, JJ. 1989. Results
and evaluations of 48 continuous breeding reproduction studies conducted in mice. Fund, and
Appl. Toxicol., 13:747-777.
Murray, JF; Andersen, ME. 2001. Data derived uncertainty factors; boric acid (BA) as a case
study. Hum. Ecol. Risk. Assess., 7:125-138.
R-6
-------
External Review Draft Do Not Cite or Quote
NAS (National Academy of Sciences). 1977. Drinking Water and Health. Safe Drinking Water
Committee. National Academy Press. Washington, DC.
NRC (National Research Council). 1993. Pesticides in the Diets of Infants and Children.
National Academy Press, Washington, DC.
NRC (National Research Council). 2000. Acute Exposure Guideline Levels for Selected
Airborne Chemicals, Volume 1. Washington, DC: National Academy Press.
Nau, H. 1991. Pharmacokinetic considerations in the design and interpretation of developmental
toxicity studies. Fundam. Appl. Toxicol., 16(2):219-21.
OPP (Office of Pesticide Programs, EPA). 1998. Hazard identification - Toxicology Endpoint
Selection Process. Dated August 11. Reviewed by the SAP.
Orme, J; Ohanian, EV. 1991. Health Advisories for Pesticides. In: Chemistry, Agriculture and
the Environment. Ed. ML Richardson. Royal Society of Chemistry. Cambridge, UK, pp. 429-
443.
Pieters, MN; Kramer, HJ; Slob, W. 1998. Evaluation of the uncertainty factor for subchronic-to-
chronic extrapolation: statistical analysis of toxicity data. Reg. Toxicol. Pharmacol., 27:108-111.
Pinkerton, KE; load, J. 2000. The mammalian respiratory system and critical windows of
exposure for children's health. Environ. Health Perspect, 108 (Suppl 3):457-462
Renwick, AG. 1993. Data-derived safely factors for the evaluation of food additives and
environmental contaminants. Food Addit. Contain., 10:275-305.
Renwick, AG. 1998. Toxicokinetics in infants and children in relation to the ADI and TDI.
Food Additiv. Contamin., 15 (Suppl.): 17-35.
Renwick, AG; Lazarus, NR. 1998. Human variability and noncancer risk assessment - an
analysis of the default uncertainty factor. Regul. Toxicol. Pharmacol., 27:3-20.
Rigas, ML; Roth, WL; Jarabek, AM. 2000. Proposed Models for Estimating Relevant Dose
Resulting from Exposures by the Gastrointestinal Route. Society of Risk Analysis, 2000 Annual
Meeting (abstract).
Schilter, B; Renwick, AG; Huggett, AC. 1996. Limits for pesticide residues in infant food: A
safety-based approach. Regul. Toxicol. Pharmacol., 24:126-140.
R-7
-------
External Review Draft Do Not Cite or Quote
Skowronski, G; Abdel-Rahman, MS. 2001. Relevance of the 10X uncertainty factor to the risk
assessment of drugs used by children and geriatrics. Human Ecolog. Risk Assess., 7:139-152.
Smith, TL; Koman, LA; Mosberg, AT. 1994. Cardiovascular physiology and methods for
toxicology. In: Principles and Methods of Toxicology. Ed. AW Hayes. Raven Press, New York,
pp. 917-935.
Ten Berge, WF; Zwart, A; Appelman, LM. 1986. Concentration-time mortality response
relationship of irritant and systemically acting vapours and gases. J. Hazard. Mater., 13:301-309.
Terry, KK; Elswick, BA; Stedman, DB; Welsch, F. 1994. Developmental phase alters
dosimetry-teratogenicity relationship for 2-methoxyethanol in CD-I mice. Teratology,
49:218-27.
Versarlnc. 200la. Report on Exploration of Aging and Toxic Response Issues. Prepared for
the Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC, under
EPA Contract 68-C-99-238.
Versar Inc. 2001b. Report on Exploration of Perinatal Pharmacokinetic Issues. Prepared for the
Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC, under EPA
Contract 68-C-99-238.
Weller, E; Long, N; Smith, A; Williams, P; Ravi, S; Gill, J; Henessey, R; Skornik, W; Brain, J;
Kimmel, C; Kimmel, G; Holmes, L; Ryan, L. 1999. Dose-rate effects of ethylene oxide exposure
on developmental toxicity. Toxicol. Sci., 50:259-270.
West, GB; Brown, JH; Enquist, BJ. 1997. A general model for the origin of allometric scaling
laws in biology. Science, 276 (No.5309): 122-126.
R-8
-------
External Review Draft Do Not Cite or Quote
APPENDIX A
ISSUES RAISED BY THE 10X TASK FORCE
A number of issues were raised by the 10X Task Force1 in its discussions of the
requirements for protecting children's health and application of an additional 10X safety factor, as
mandated by the 1996 FQPA. The Task Force felt that these issues, which include the following,
should be discussed on a broader Agency-wide basis, as well as with the outside community, for
both pesticides and other agents.
1. Appropriate application of the database modifying factor for additional required
developmental and adult toxicity studies. It appears from the data available that the default
intraspecies 10-fold uncertainty factor may be adequate in the majority of cases for protecting
children's health. However, when data specific to children's health are missing or inadequate for
a particular agent, application of the database modifying factor in addition to the intraspecies
variability factor may be sufficient to account for the possibility that children may be significantly
more sensitive than adults. This issue needs further examination.
2. How to account for the level of concern in the RfD/RfC process. Criteria for assessing the
level of concern for children's health were developed by the Toxicology Working Group of the
10X Task Force and include factors such as (a) human data on pre- and postnatal toxicity; (b)
pre- and postnatal toxicity in animal studies, including effects of a different or similar type as
those in adults; (c) dose-response nature of the experimental animal data, including the dose-
related incidence of response, relative potency of response, slope of the dose-response curve
when the MOE is small, and how well the NOAEL or BMD is defined; and (d) relevance of the
experimental animal data to humans, including toxicokinetics, similarity of the biological response,
knowledge of the mechanism of action. For each of these areas, criteria are given for estimating a
level of concern for children's health as high, moderate, or low. The level of concern may be
taken into account in the uncertainty and modifying factors applied to the RfD, although there is
currently no formal process for doing so.
10X Task Force documents: Toxicology Data Requirements for Assessing Risks of
Pesticide Exposure to Children's Health (EPA, 1999a) and Exposure Requirements for Assessing
Risks from Pesticide Exposure to Children's Health (EPA, 1999b).
A-l
-------
External Review Draft Do Not Cite or Quote
3. As indicated in the Toxicology Document appended to the Task Force report, the current
default recommended for using developmental toxicity data for different duration reference
values is to apply most endpointsfor all durations. This is because it is assumed that most
endpoints of developmental toxicity can be caused by a single exposure. If, however,
developmental effects are more sensitive than those seen after longer-term exposures, then even
the chronic RfD/RfC should be based on such effects to reduce the risk of potential greater
sensitivity in children. Because the standard studies currently conducted for developmental
toxicity involve repeated exposures, data are not often available on which endpoints may be
induced by acute, subacute, subchronic, or chronic dosing regimens and, therefore, which should
be used in setting various duration reference values. Further consideration of the appropriate
application of developmental toxicity endpoints to various duration reference values is
recommended. As part of this recommendation, an in-depth review of the FLED document on
Hazard Identification - Toxicology Endpoint Selection System, should be undertaken;
4. Appropriate setting of intermediate RfDs/RfCs for pesticides and other agents. The focus
of the RfD and the RfC has been on chronic exposure reference values. Acute RfDs are also set
for pesticides, and intermediate reference values are set for residential exposures as well as for
drinking water. Data on developmental toxicity will often be a greater factor in calculating the
acute and intermediate reference values, and exposures to children are more often of this type as
well. Consideration should be given to setting intermediate reference values for environmental
agents. In addition, the question of whether or when to set RfDs/RfCs specific for children
should be considered.
5. Appropriate adjustment of the NOAEL or the BMDfrom inhalation exposure studies for
extrapolation of developmental toxicity data using less than continuous exposure to a continuous
exposure scenario. Currently, NOAELs/BMDs from inhalation exposure studies other than those
for developmental toxicity using, e.g., a 6-hr/day exposure regimen, are adjusted to a continuous
(24 hr/day) exposure for calculating RfDs/RfCs. The developmental toxicity risk assessment
guidelines (EPA, 1991) recommended against making this adjustment, because it was assumed
that there was a threshold above which exposure would have to occur before an effect would
result. This recommendation needs to be reconsidered, along with the adjustment of
NOAELS/BMDs in general.
Several improvements in testing approaches were also proposed for consideration in the
10X Task Force report as a way to improve the assessment of potential risks to children. The
A-2
-------
External Review Draft Do Not Cite or Quote
Technical Panel was asked to consider the need for such tests, when they should be required, and
interpretation of the data for risk assessment purposes. These include
• pharmacokinetics that include data from different developmental stages, perhaps done in a
tiered approach as suggested in Kimmel and Francis (1990);
direct dosing of neonates, especially when early exposure is of concern, since this is the
time when differences in metabolic capability are greatest;
• perinatal carcinogenesis studies and appropriate triggers for when they should be required;
• developmental immunotoxicity testing and appropriate triggers;
advanced DNT testing, in particular, cognitive testing that is more similar to that used in
humans.
An additional issue was how to make exposure assessments compatible with the dose-
response assessment. For example, how should the appropriate durations of exposure be
determined for toxic endpoints of concern? Should standard exposure durations be used?
A-3
-------
External Review Draft Do Not Cite or Quote
APPENDIX B
CASE STUDY FOR CHEMICAL X
EVALUATION AND SELECTION OF HEALTH ENDPOINTS FOR
DERIVATION OF REFERENCE VALUES
Chemical X, a synthetic halogenated aliphatic alkene, is a nonflammable volatile liquid at
room temperature. The chemical enters the air through its industrial and commercial use,
primarily as a solvent. It is also found in surface and ground water and soil upon disposal. The
most important routes of human exposure are inhalation of the chemical in the ambient and indoor
air and ingestion of contaminated drinking water. Because of its high volatility, dermal exposure
to the chemical is expected to be minimal.
The health effects information for Chemical X is considered adequate for deriving
reference values for inhalation and oral exposures. As summarized below, the combined results
from available studies in humans and animals indicate that Chemical X has the potential to induce
neurotoxicity, liver toxicity, kidney toxicity, developmental toxicity, reproductive effects, and
cancer. The toxicity profile of the chemical is dependent on the dose, duration, and route of
exposure. The mode(s) of action of the observed toxicities in treated animals is not fully known.
This case study illustrates the use of single or multiple endpoints for deriving reference
values for different durations of exposure following oral and inhalation exposure. For the purpose
of illustration, results of key studies are summarized in Table B-l. Tables B-2 and B-3 present
dose-response data for different health endpoints relevant to different durations of exposure via
inhalation and ingestion, respectively. Table B-2 shows the HECs adjusted for duration and
cross-species differences. Table B-3 shows the HEDs adjusted for duration (see discussion of
HEC and HED derivation in Chapter 4).
SUMMARY OF HEALTH EFFECTS INFORMATION
Absorption, Distribution, Metabolism, and Elimination
There is very little information on the absorption and distribution of Chemical X in humans
and laboratory animals following oral, inhalation, and dermal exposure. However, similar effects
B-l
-------
External Review Draft Do Not Cite or Quote
are seen by oral and inhalation exposures, suggesting that Chemical X or its metabolites reach
their target sites after absorption from either exposure route. Available in vitro metabolic studies
indicate that Chemical X is extensively metabolized in target tissues including the liver and kidney
of rats and mice. Limited in vitro studies with human tissues show a similar pattern of
metabolism. As discussed below, much of Chemical X-induced toxicity appears to be due to its
metabolites. These metabolites have been detected in the urine of rats and mice following
inhalation and oral exposure to the parent chemical.
Neurological Effects
The ability of Chemical X to cause neurotoxic effects in humans and animals from acute
and longer-term inhalation exposure is well documented. In contrast, only limited information is
available regarding the potential neurotoxic effects of Chemical X by ingestion. Available data
indicate that humans are more sensitive to the CNS effects of Chemical X than are rats and mice.
The mechanism of action for the CNS effects has not been clearly established but is believed to be
related to effects of the parent compound on lipid and fatty acid composition of the membranes.
Inhalation Exposure
Several clinical studies available in the open literature reported dose-dependent clinical
signs of CNS symptoms in healthy adults exposed acutely and subacutely via inhalation to
Chemical X. Male and female human volunteers exposed acutely to high concentrations of
Chemical X (500 ppm to 2000 ppm for 2 hours) showed dose-dependent effects, including
headache, dizziness, incoordination, drowsiness, and anesthesia. No effect was reported
following acute exposure to 50 ppm. Similar effects were observed in adult humans at lower
concentrations (20, 100, 150 ppm) for 6 hours per day for up to 7 days, with a NOAEL of 10
ppm.
Long-term and chronic neurotoxic effects have been reported in several studies of
occupational exposure of workers to Chemical X in different industries. Exposure data were not
provided in these reports; however, it can be presumed that these workers were exposed to a daily
TWA exposure of 20 ppm. Subjective neurological symptoms, including dizziness and
forgetfulness, were consistently reported across studies. No other health effects information was
collected in these studies.
B-2
-------
External Review Draft Do Not Cite or Quote
Concentration-dependent clinical signs of neurological effects including hyperactivity,
ataxia, hypoactivity, and finally loss of consciousness, have also been reported in rats, and mice
following acute (800, 1500, 3000 ppm for 4 hours) and short-term inhalation exposure (500,
1000, 2000 ppm 6 hours per day for 2 weeks) to Chemical X at high concentration. Similar
effects were observed in rats exposed to Chemical X at 400 and 800 ppm and in mice at 200 and
400 ppm for 13 weeks. The subchronic NOAELs for rats and mice were 200 and 100 ppm,
respectively. Chronic exposure to Chemical X at lower concentrations resulted in less serious
clinical signs of CNS effects in rats (200 or 400 ppm) and mice (100 or 200 ppm). The chronic
NOAELs for rats and mice in these studies were 100 ppm and 50 ppm, respectively. It should be
noted that neurological endpoints examined in these animal studies are limited to clinical signs and
histopathology. In a special study, changes in fatty acid composition of the brain were observed
in rats exposed at 300 ppm (the only tested concentration) for 90 days.
Oral Exposure
Acute neurological effects in adult humans after ingestion of Chemical X are similar to
those seen after inhalation. Accidental exposure of approximately 6 - 8 ml (or about 100 mg/kg/d)
resulted in narcotic effects.
Single oral gavage treatment of Chemical X to adult rats (1,000 mg/kg) caused ataxia.
Ataxia was also observed in pregnant rats treated by gavage at 900 mg/kg on GD 6-19. No
CNS effects were reported in a chronic oral gavage study in rats and mice at 50, 100, or 300
mg/kg/day. Neurological endpoints examined in these studies were limited to clinical signs and
histopathology.
Liver Effects
The combined results of available human and animal studies indicate that chronic
inhalation exposure to Chemical X has the potential to cause liver toxicity and cancer. Chemical
X induces liver effects through its metabolites. Liver tumors in mice are apparently mediated by
induction of hepatocellular peroxisomes. Mice appear to be more sensitive to the liver effects of
Chemical X than are rats and humans. This is supported by the observation that mice produce
more of the active metabolites than rats and humans and that the peroxisome proliferation
response in mice is more pronounced than in rats and humans. It is likely that Chemical X-
induced liver toxicity and cancer in humans result from a mechanism(s) that differs from the
mechanism that produces liver effects in mice.
B-3
-------
External Review Draft Do Not Cite or Quote
Inhalation Exposure
Several studies reported changes in serum levels of liver enzymes in workers exposed to
the chemical at a daily TWA exposure concentration over an 8 hour work shift of about 20 ppm.
These workers, however, did not exhibit any clinical symptoms of liver dysfunction. Exposure-
related increased incidences of liver cancer (with increasing exposure, duration of exposure,
and/or increased time since first exposure) were observed in several epidemiological studies. No
other health endpoints were investigated in these occupational studies.
Dose-related liver effects (liver hypertrophy, vacuolization of hepatocytes, necrosis) have
been observed in exposed mice following subchronic exposure (13 weeks) to Chemical X at 200,
400 ppm with a NOAEL of 100 ppm. Dose-related liver toxicity and tumors were also found in
mice following chronic exposure at 100 and 200 ppm. The NOAEL for liver toxicity in mice in
this chronic study was 50 ppm.
Rats showed similar liver responses, but at higher exposure concentrations following
subchronic exposure (400 or 800 ppm) with a NOAEL of 200 ppm. Liver toxicity were also
observed in rats in a chronic study at 200 or 400 ppm. Liver tumors were not induced in rats. The
NOAEL for liver effects in rats was 100 ppm. It should be noted that liver effects examined in
these subchronic and chronic studies were limited to clinical chemistry, morphology, and
histopathology.
Oral Exposure
Similar liver effects were observed in mice and rats treated subchronically (100, 300, 500
mg/kg/day) or chronically (50,100, 300 mg/kg/day) with Chemical X via oral gavage. Mice
showed more severe effects than did rats. Dose-related liver tumors were also found in treated
mice in a chronic study.
Renal Effects
Available human and animal studies indicate that Chemical X also has the potential to
cause renal toxicity and cancer. The mechanism for the development of kidney effects in humans
is not known. However, the preponderance of the evidence suggests that the cysteine conjugates
and reactive metabolites generated from their beta-lyase metabolism are likely responsible for the
kidney toxicity and tumorigenicty in the rat.
B-4
-------
External Review Draft Do Not Cite or Quote
Inhalation Exposure
Symptoms of renal dysfunction (proteinuria, hematuria) have been associated with
accidental exposure to anesthetic concentrations of Chemical X. Subtle or no renal effects were
reported in workers exposed chronically to Chemical X. Increased urinary levels of lysozyme and
beta-glucuronidase suggestive of mild renal tubular damage have been observed in workers
exposed for an average of 15 years to a daily average TWA concentration of 20 ppm. Several
epidemiologic studies of workers exposed to Chemical X showed increased incidences of renal
cell carcinoma. No other health endpoints were examined in these studies.
Dose-related renal toxicity (cytomegaly, toxic nephrosis of tubular epithelial cells in the
inner renal cortex) were induced in rats (400, 800 ppm) and mice (200, 400 ppm) exposed to
Chemical X for 13 weeks. Subchronic NOAELs for renal effects in rats and mice were 200 ppm
and 100 ppm, respectively.
Similar renal effects were observed in a chronic study of Chemical X in rats (100, 200,
400 ppm) and mice (50, 100, 200 ppm). Low incidences of renal tubular cell adenomas and/or
adenocarcinoma were also induced in rats and mice following chronic exposure to Chemical X at
the two highest concentrations. Chronic NOAELs for renal effects in rats and mice were 100 and
50 ppm, respectively.
Oral Exposure
Dose-related toxic nephropathy characterized by degenerative changes in the proximal
convoluted tubules and necrosis of the tubular epithelium were found in rats and mice treated with
Chemical X via oral gavage for 90 days at 100, 300, or 500 mg/kg/day and for 2 years at 50, 100,
or 300 mg/kg/day. Subchronic and chronic NOAELs for renal effects in both rats and mice were
at 100 and 50 mg/kg/day, respectively.
Developmental Effects
Available studies in humans and animals indicate that Chemical X has the potential to
cause developmental effects by inhalation and oral ingestion. Limitations of human studies could
not resolve whether the observed developmental effects are causally related to the chemical or a
result of chance or bias. However, the epidemiologic findings are supported by animal studies
with exposure to Chemical X by inhalation and oral gavage showing that the developing nervous
system is the most sensitive target in rats and mice.
B-5
-------
External Review Draft Do Not Cite or Quote
Inhalation Exposure
Epidemiologic studies of women occupationally exposed to Chemical X and other related
solvents have reported elevated risk of cardiac anomalies in their offspring. Due to limitations of
these studies, exposure-response could not be established. No other health endpoints were
investigated in these studies.
Pregnant rats were exposed by inhalation to Chemical X at 0, 100, or 900 ppm for 6
hrs/day on days 7-13 of gestation. Decreased performance for neuromuscular ability was
observed in pups from dams exposed to 900 ppm. A NOAEL of 100 ppm for developmental
effects was identified in this study.
Oral Exposure
Neural tube defects and eye anomalies have been reported in studies of residents exposed
to drinking water contaminated with Chemical X and other solvents. Exposure-response could
not be determined from these studies.
An increased incidence of micro/anophthalmia were observed in the offspring of rats
treated with Chemical X by gavage at 900 mg/kg/day on GDs 6-19. In a study that investigated
the effect of Chemical X on the developing nervous system, male mouse pups were treated by
gavage with Chemical X at 50 or 300 mg/kg/day for seven days (age 10-17 days). Hyperactivity
was reported in animals during adulthood at the high dose. No effects were found at the low dose.
Reproductive Effects
Available studies in humans and animals suggest that Chemical X may have the potential
to cause reproductive effects. The underlying mechanism of action for potential reproductive
effects is not known.
Inhalation Exposure
There is suggestive evidence of spontaneous abortion and menstrual disorders among
women occupationally exposed to Chemical X. However, no definitive conclusions can be made
because of the limitations associated with these studies.
In a two- generation reproduction inhalation study, reduced litter size and reduced
survival of offspring were reported in rats exposed to Chemical X at 1,000 ppm, a concentration
that also resulted in sedation and renal effects. No reproductive effects were identified at 300
ppm. The protocol used, however, was one in which reproductive development (e.g., timing of
B-6
-------
External Review Draft Do Not Cite or Quote
puberty or anogenital distance) and adult reproductive function (semen quality, estrous cyclicity)
were not evaluated, nor were organ weights measured.
Oral Exposure
No information is available on the potential reproductive effects of Chemical X in animals
via oral exposure.
B-7
-------
External Review Draft Do Not Cite or Quote
SELECTION OF HEALTH ENDPOINTS AND DERIVATION OF REFERENCE
VALUES
Narrative Description of the Extent of the Database
The database for inhalation exposure is limited but adequate for deriving reference values.
No information is available on possible modes of action or pharmacokinetics. Some human data
on acute, short-term, and longer-term exposures are available, although the range of endpoints
evaluated and the dose-response information for different durations of exposure are limited. The
animal data include acute, short-term, longer-term, and chronic studies with exposures beginning
in young adult animals. The acute and short-term data are limited to clinical signs of morbidity
and mortality, while the short-term, longer-term, and chronic studies include some histopathology
as well. There is a study of DNT with prenatal exposure in rats limited to GDs 7-13 (as opposed
to more extensive exposure throughout a major part of CNS development, e.g., GD 6 to [prenatal
day] PND 11 in the standard DNT study testing protocol). No other studies of prenatal or
postnatal developmental toxicity study were done except for evaluations of survival and growth in
a two-generation reproduction study in rats. The protocol used, however, was one in which
reproductive development (e.g., timing of puberty or anogenital distance) and adult reproductive
function (semen quality, estrous cyclicity) were not evaluated, nor were organ weights measured.
No studies were conducted that considered issues related to the toxicity of the agent in old age,
either from earlier exposures or from exposures in aged animals.
The database for oral exposure is much more limited than the database for inhalation
exposure, with acute data in humans on neurotoxicity at a single, high-dose level and chronic data
on birth defects but no dose information. The animal data are likewise very limited, with a single-
dose acute toxicity study in rats in which clinical signs of morbidity and mortality were evaluated
and subchronic (90-day) and chronic toxicity data in rats and mice that included histopathology.
Prenatal developmental toxicity data were available in rats following exposure on GD 6-19, and
an evaluation of adult neurotoxicity was conducted in mice following postnatal developmental
exposure on days 10-17 of age. No other developmental toxicity data were available, and no
information on reproductive toxicity or adult neurotoxicity was available. No studies were
conducted that considered issues related to the toxicity of the agent in old age, either from earlier
exposures or from exposures in aged animals.
B-8
-------
External Review Draft Do Not Cite or Quote
Figure B-l. Exposure-Response Arrays for Inhalation Exposure to Chemical X
25
O
uj -I/
E
Q.
Q.
10
Acute Exposure
Longer-Term Exposure
A
A
A
A
Neuro- Devel
tox Tox
A
A 2
Repro Liver
Tox Tox
A
Renal
Tox
Short-Term Exposure
td
VO
o
I
E
Q.
Q.
30 -
90 -
1 n
o
A
A
D A
A
LJJ
X
E
Q.
Q.
Neuro- Devel Repro
tox Tox Tox
20
15
10 -
c
^
0
• A
A
A
g "
D A
Neuro- Devel Repro
tox Tox Tox
Chronic Exposure
~^ 9n
o
m 15
£1 n
I U
Q.
°- 5
n
A
A
a A
Neuro- Devel
tox Tox
A S
a
Repro Liver
Tox Tox
a
Renal
Tox
Human NOAEL m Human LOAEL A Ra* NOAEL A Ra* LOAEL Q Mouse NOAEL * Mouse LOAEL
-------
External Review Draft Do Not Cite or Quote
Exposure-Response Array
In addition to displaying the data in tabular form (Tables B-l, B-2, and B-3), an exposure-
response array can be a useful way of visually displaying the data (see Figure B-l) to show what
data are available for each duration of exposure. The points in the graph are for HECs and HEDs
based on the dosimetric adjustments discussed in Chapter 4, including dosimetric adjustment of
the developmental toxicity data as was done for other types of toxicity data.
Acute Exposure
Inhalation exposure
Results of available studies indicate that acute inhalation exposure to Chemical X can
result in neurotoxic effects in human adults with a LOAELjjgc of 50 ppm (NOAELjjgc of 4 ppm).
Animal studies also show that Chemical X has the potential to cause developmental neurotoxicity
and reproductive effects at comparable doses with LOAELj^cS of 900 and 1000 ppm (NOAE!^,.
of 2.5 ppm and 5 ppm), respectively.
Because animal studies indicate that the developing nervous system is vulnerable to
Chemical X exposure and the NOAELjjgc for that endpoint is most protective, a NOAELjjgc of
2.5 ppm in the developmental toxicity study is used as the basis for deriving an acute reference
value for inhalation exposure. Default UFs of 101/2 (animal-to-human extrapolation), 10 (inter-
individual differences), and 101/2 (database deficiencies: no adequate prenatal developmental
toxicity studies in two species, no adequate developmental neurotoxicity study) are applied. The
resultant reference value for acute inhalation exposure is 0.03 ppm (Table B-4).
Oral exposure
Acute oral exposure to Chemical X can result in neurotoxic effects in human adults
(LOAEL of 100 mg/kg/d). However, dose-response data are not available in humans. A single
study in mice indicates that Chemical X (dosing on days 10-17 postnatally; equivalent to
approximately 1 month to l-!/2 years of age in humans) also has the potential to cause
developmental neurotoxicity, with a LOAEL of 300 mg/kg/d (NOAEL of 50 mg/kg/d). Applying
default UFs of 10 and 10 to account for animal-to-human extrapolation and interindividual
differences, as well as a database UF of 10 due to the limitations of the available data, results in a
reference value for acute oral exposure of 0.05 mg/kg (Table B-5).
B-10
-------
External Review Draft Do Not Cite or Quote
Short-term Exposure
Inhalation exposure
The reference value for short-term inhalation exposure is based on the human data as well
as the developmental neurotoxicity and reproductive toxicity data with NOAELj^s of 2 ppm, 2.5
ppm, and 5 ppm, respectively (LOAELj^cS of 10 ppm, 100 ppm, and 300 ppm). Using the human
NOAELjjEc of 2 ppm, and applying a 10-fold default UF for intra-species uncertainty and
variability and a 101/2-fold UF for database deficiencies would result in a reference value for short-
term inhalation exposure of 0.07 ppm. However, a default factor of 101/2 (interspecies), 10
(intraspecies) and 101/2 (database deficiencies) would be applied to the HECs for the animal data
on developmental neurotoxicity and reproductive toxicity (2.5 and 5 ppm, respectively), resulting
in reference values of 0.03 and 0.05 ppm. Given the close range of values, the reference value of
0.03 ppm would be used because it is more protective of the developing individual as well as the
adult (Table B-4).
Oral exposure
The reference value for short-term oral exposure would be the same as for acute exposure,
which is based on developmental neurotoxicity, as discussed above (Table B-5).
Longer-term Exposure
Inhalation exposure
Subchronic and chronic inhalation exposure to Chemical X can result in multiple health
effects. Available studies demonstrate neurotoxicity in adult humans. However, dose-response
information is not available, and the presumed LOAEL (20 ppm) for neurotoxicity in humans is
somewhat higher than the HECs for other health endpoints (developmental, reproductive, liver,
and renal effects) observed in animal studies, where the LOAELjjgcS range from 7 ppm to 22.5
ppm (NOAELjjEcS range from 2.5 ppm to 5 ppm). Dose-response data for these health endpoints
in animal studies can be used as the basis for deriving a longer-term inhalation reference value for
Chemical X. UFs of 101/2 (interspecies), 10 (intraspecies), and 101/2 (database deficiencies) were
applied to NOAELjjgcS for the various endpoints in deriving sample reference values. If an
additional factor of 3 were applied to the rat developmental toxicity data to account for the
marked difference in exposure duration in the study itself (7 days of exposure: GD 7 - 13), a
longer-term sample reference value of 0.01 ppm would result. Without this additional factor, the
HEC from the developmental toxicity study was still the lowest value (0.03 ppm) and was used in
B-ll
-------
External Review Draft Do Not Cite or Quote
the reference value derivation (Table B-4). Whether an additional factor should be applied to the
developmental toxicity data or to other data of much shorter duration should be explored further.
Oral exposure
Available animal data indicate that longer-term oral exposure to Chemical X can cause
liver, renal, and developmental effects, with LOAELs ranging from 300-2114 mg/kg/d (NOAELs
ranging from 50 mg/kg/d to 71 mg/kg/d). Application of default UFs of 10 (interspecies), 10
(intraspecies) and 10 (database deficiencies) to the data from the subchronic studies would result
in longer-term oral reference values of 0.05 mg/kg/d (Table B-5). If an additional factor of 3 was
applied to the mouse developmental toxicity data to account for short-term to longer-term
exposure and a total UF of 3000 applied, a sample reference value of 0.02 mg/k/d would be
calculated, which is less than the other values derived from subchronic exposure data. As
indicated above, whether an additional factor should be applied to the developmental toxicity data
or to other data of much shorter duration should be explored further.
Chronic Exposure
Inhalation exposure
For the chronic inhalation reference value, the NOAELjjEcS range from 2 ppm to 5 ppm
(LOAELjjEcS range from 10 ppm to 300 ppm), and UFs of 101/2 (interspecies), 10 (intraspecies),
and 101/2 (database deficiencies) applied to the chronic exposure NOAELjjgcS for neurotoxicity,
liver and kidney toxicity, and reproductive toxicity data results in sample reference values of 0.02
- 0.05 ppm (Table B-4). Applying these UFs to the NOAE!^,-, of 2.5 ppm for developmental
toxicity yields a sample reference value of 0.03 ppm, falling within the range of chronic study-
based values. In this example, the chronic study neurotoxicity data is the limiting endpoint,
providing a chronic inhalation reference value of 0.02 ppm. If, contrary to current practice, an
additional 10 for subchronic to chronic duration were applied to the developmental NOAELjjEc,
the resultant sample reference value would be 0.003 ppm. As mentioned in section D.S.f, this
issue may need further exploration.
Oral exposure
As chronic dosing studies are available with NOAELs of 36 mg/kg/d (LOAELs of 50
mg/kg/d), application of default UFs of 10 (interspecies), 10 (intraspecies) and 10 (database
deficiencies) would result in chronic oral reference values of 0.04 mg/k/d (Table B-5). Applying
these same factors to the developmental NOAEL of 50 mg/kg/d yields a slightly higher value of
0.05 mg/kg/d. Applying an additional UF of 10 to the mouse developmental toxicity data to
B-12
-------
External Review Draft Do Not Cite or Quote
account for the difference between short-term and chronic exposure would result in a total UF of
10,000.
B-13
-------
External Review Draft Do Not Cite or Quote
Table B-l. Summary Results of Major Studies on Chemical X
Species
Exposure Duration &
Frequency
Concentration/Dose
Results
td
Human
(Male&
Female)
(Female)
(Male&
Female)
2hrs
5 hrs/d for 7 d
occupational (> 15 yrs)
occupational
accidental exposure
accidental exposure
chronic exposure via
drinking water
50, 500, 2000 ppm
10,20, 150 ppm
TWA of 20 ppm
not available
not available (inhalation)
about 100 mg/kg (oral)
not available
headache, dizziness, incoordination, drowsiness
anesthesia at 2000 ppm; no effects at 50 ppm
headache, dizziness, incoordination, drowsiness
no effect at 10 ppm
dizziness, forgetfulness; changes in serum liver enzymes
increased urinary levels of lysozymes, beta-glucuronidase
menstrual disorders; spontaneous abortion
cardiac anomalies in children of workers
narcosis, proteinuria, hematuria
narcosis
neural tubular defects
eye anomalies
-------
External Review Draft Do Not Cite or Quote
Table B-l. Summary Results of Major Studies on Chemical X (continued)
Species
Exposure Duration
and Frequency
Concentration/
Dose
Results
td
Rat (M & F)
Mouse ((M & F)
Rat (M & F)
Mouse (M & F)
Rat (M & F)
Rat (M & F)
Mouse (M & F)
Rat (M & F)
Mouse (M & F)
4hrs
6 hrs/d, 5 d/wk for
2 wks
6 hrs/d, 5 d/wk for
13 wks
6 hrs/d, 5 d/wk for
13 wks
6 hrs/d, 5 d/wk for
13 wks
6 hrs/d, 5 d/wk for
104 wks
6 hrs/d, 5 d/wk for
78 wks
0, 800, 1500, 3000
ppm
0, 500, 1000, 2000
ppm
0, 300 ppm
0, 200, 400, 800 ppm
0, 100, 200, 400 ppm
0, 100, 200, 400 ppm
0, 50, 100, 200 ppm
dose-related hyperactivity, ataxia, hypoactivity, narcosis
dose-related hyperactivity, ataxia, hypoactivity, narcosis
changes in fatty acid composition of the brains
dose-related hyperactivity, ataxia; liver hypertrophy,
vacuolization of hepatocytes, necrosis; cytomegaly, toxic
nephrosis of tubular epithelial cells
dose-related hyperactivity, ataxia; liver hypertrophy,
vacuolization of hepatocytes, necrosis; cytomegaly, toxic
nephrosis of tubular epithelial cells
dose-related liver hypertrophy, vacuolization of hepatocytes,
necrosis; cytomegaly, toxic nephrosis of tubular epithelial
cells; renal cell adenoma/carcinoma
dose-related liver hypertrophy, vacuolization of hepatocytes,
necrosis; cytomegaly, toxic nephrosis of tubular epithelial
cells; renal cell adenoma/carcinoma
-------
External Review Draft Do Not Cite or Quote
Table B-l. Summary Results of Major Studies on Chemical X (continued)
Species
Rat (F)
Rat (M & F)
Rat (F)
Mouse (M)
Rat (M & F)
Rat (M & F)
Mouse (M & F)
Rat (M & F)
Mouse (M & F)
Exposure Duration
& Frequency
GD 7 - 13, 6 hrs/day
two-generation
reproductive study, 6
hrs/day, 5 days/wk
GD 6 - 19
days 10 - 17 of age
single dose
90 days
104 weeks
78 wks
Concentration/
Dose
0, 100, 900 ppm
0, 300, 1000 ppm
0, 900 mg/kg/d
0, 50, 300 mg/kg/d
0, 1000 mg/kg
0, 100, 300, 500
mg/kg/d
0, 50, 100, 300
mg/kg/d
Results
decreased performance for neuromuscular ability in pups
ataxia in dams at high dose
reduced litter size and reduced survival of offspring at high
dose
increased incidence of micro/anophthalmia
hyperactivity in exposed animals during adulthood at high
dose
ataxia
dose-related liver hypertrophy, vacuolization of hepatocytes,
necrosis; degenerative changes in the proximal convoluted
tubules, and necrosis of the tubular epithelium
dose-related liver hypertrophy, vacuolization of hepatocytes,
necrosis, liver tumors (mice only); degenerative changes in
the proximal convoluted tubules, and necrosis of the tubular
epithelium
td
-------
External Review Draft Do Not Cite or Quote
Table B-2. Exposure-Response Data of Chemical X- Inhalation
Exposure Duration
Acute
Human (NOAEL)3
(LOAEL)
Rat (NOAEL)
(LOAEL)
Mouse (NOAEL)
(LOAEL)
Short-Term
Human (NOAEL)
(LOAEL)
Rat (NOAEL)
(LOAEL)
Mouse (NOAEL)
(LOAEL)
Longer-term
Human (NOAEL)
(LOAEL)
Rat (NOAEL)
(LOAEL)
Mouse (NOAEL)
(LOAEL)
Neurotoxicity
50 ppm (4 ppm)b
500 ppm (40 ppm)
N/AC
800 ppm (13 ppm)
N/A
800 ppm (33 ppm)
10 ppm (2 ppm)
20 ppm (4 ppm)
N/A
500 ppm (9 ppm)
N/A
500 ppm (22 ppm)
N/A
20 ppm
200 ppm (4 ppm)
400 ppm (7 ppm)
100 ppm (5 ppm)
200 ppm (9 ppm)
Developmental
Effects
100 ppm (2.5 ppm)
900 ppm (22.5 ppm)
100 ppm (2.5 ppm)
900 ppm (22.5 ppm)
100 ppm (2.5 ppm)
900 ppm (22.5 ppm)
Reproductive
Effects
300 ppm (5 ppm)
1000 ppm (18 ppm)
300 ppm (5 ppm)
1000 ppm (18 ppm)
300 ppm (5 ppm)
1000 ppm (18 ppm
Liver Effects
N/A
20 ppm
200 ppm (4 ppm)
400 ppm (7 ppm)
100 ppm (5 ppm)
200 ppm (9 ppm)
Renal Effects
200 ppm (4 ppm)
400 ppm (7 ppm)
100 ppm (5 ppm)
200 ppm (9 ppm)
td
-------
External Review Draft
Do Not Cite or Quote
Table B-2. Exposure-Response Data of Chemical X- Inhalation (continued)
Exposure Duration
Chronic
Human (NOAEL)
(LOAEL)
Rat (NOAEL)
(LOAEL)
Mouse (NOAEL)
(LOAEL)
Neurotoxicity
N/A
20ppm
100 ppm (2 ppm)
200 ppm (4 ppm)
50 ppm (2 ppm)
100 ppm (5 ppm)
Developmental
Effects
100 ppm (2.5 ppm)
900 ppm (22.5 ppm)
Reproductive
Effects
300 ppm (5ppm)
1000 ppm (18 ppm)
Liver Effects
N/A
20 ppm
100 ppm (2 ppm)
200 ppm (4 ppm)
50 ppm (2 ppm)
100 ppm (5 ppm)
(CEL)d
Renal Effects
N/A
10 ppm
100 ppm (2 ppm)
200 ppm (4 ppm)
(CEL)
50 ppm (2 ppm)
100 ppm (5 ppm)
(CEL)
td
i
oo
3NOAELs and LOAELs were used for the derivation of HECs and reference values rather than BMCLs here because of ease of determination.
However, the Technical Panel strongly encourages the use of dose-response modeling and calculation of BMCLs.
b Underlined values are the lowest HEC. Values in parentheses are approximate HECs (based on the NOAEL or LOAEL adjusted for duration and
cross-species differences; human values were duration adjusted only) using the RfC methodology for Category 3 Gases. The RfC cross-species
adjustment is approximately 10-fold lower than the NOAEL or LOAEL. Adjustment for duration included multiplying the value by 6/24 and 5/7,
depending on the period of exposure in the study (see Table B-l).
°Not Available
d Cancer Effect Level
-------
External Review Draft Do Not Cite or Quote
Table B-3. Dose-Response Data of Chemical X- Oral Exposure
Exposure
Duration
Acute
Human (NOAEL)a
(LOAEL)
Rat (NOAEL)
(LOAEL)
Mouse (NOAEL)
(LOAEL)
Short-Term
Human (NOAEL)
(LOAEL)
Rat (NOAEL)
(LOAEL)
Mouse (NOAEL)
(LOAEL)
Neurotoxicity
N/Ab
lOOme/ke/d
N/A
lOOOmg/kg/d
Developmental
Effects
N/A
900 mg/kg/d
50 ma/ka/d
300 mg/kg/d
N/A
900 mg/kg/d
50 mg/kg/d
300 mg/kg/d
Reproductive
Effects
Liver Effects
Renal Effects
td
-------
External Review Draft
Do Not Cite or Quote
Table B-3. Dose-Response Data of Chemical X- Oral Exposure (continued)
Exposure
Duration
Longer-term
Human (NOAEL)
(LOAEL)
Rat (NOAEL)
(LOAEL)
Mouse (NOAEL)
(LOAEL)
Chronic
Human (NOAEL)
(LOAEL)
Rat (NOAEL)
(LOAEL)
Mouse (NOAEL)
(LOAEL)
Neurotoxicity
Developmental
Effects
N/A
900 mg/kg/d
50 mg/kg/d
300 mg/kg/d
N/A
900 mg/kg/d
50 mg/kg/d
300 mg/kg/d
Reproductive
Effects
Liver Effects
100 mg/kg/d
(7]_mg/kg/d)c
300 mg/kg/d
(2 14 mg/kg/d)
100 mg/kg/d
(Tlmg/kg/d)
300 mg/kg/d
(214 mg/kg/d)
50 mg/kg/d
(36 mg/kg/d)
100 m/kg/d
(71 mg/kg/d)
50 mg/kg/d
(36 mg/kg/d)
100 mg/kg/d (CELd)
(71 mg/kg/d)
Renal Effects
50 mg/kg/d
(36 mg/kg/d)
100 m/kg/d (CEL)
(71 mg/kg/d)
50 mg/kg/d
(36 mg/kg/d)
100 mg/kg/d
(CEL)(71 mg/kg/d)
td
to
o
-------
External Review Draft Do Not Cite or Quote
"NOAELs and LOAELs were used for the derivation of HEDs and reference values rather than BMDLs here because of ease of
determination. However, the Technical Panel strongly encourages the use of dose-response modeling and calculation of BMDLs.
b Underlined values are the lowest HED. Values in parentheses are approximate HEDs (based on the NOAEL or LOAEL adjusted for
duration). Adjustment for duration included multiplying the value by 6/24 and 5/7, depending on the period of exposure in the study (see
Table B-l).
cNot Available
d Cancer Effect Level
B-21
-------
External Review Draft Do Not Cite or Quote
Table B-4. Derivation of Reference Values for Chemical X - Inhalation Exposure
Exposure
Duration
Acute
Short-term
Longer-term
Chronic
HEC
(ppm)
4
2.5
5
2
2.5
5
20L"
4
5
2.5
5
4
5
4
5
20L
2
2
2.5
5
2
2
2
2
Species
Human
Rat
Rat
Human
Rat
Rat
Human
Rat
Mouse
Rat
Rat
Rat
Mouse
Rat
Mouse
Human
Rat
Mouse
Rat
Rat
Rat
Mouse
Rat
Mouse
Type of
Effect3
NT
DT
RT
NT
DT
RT
NT
NT
NT
DT
RT
LT
LT
KT
KT
NT
NT
NT
DT
RT
LT
LT
KT
KT
Uncertainty Factors'"
Total
30
100
100
30
100
100
300
100
100
100
100
100
100
100
100
300
100
100
100
100
100
100
100
100
A
1
3
3
1
3
3
1
3
3
3
3
3
3
3
3
1
3
3
3
3
3
3
3
3
H
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
L
1
1
1
1
1
1
10
1
1
1
1
1
1
1
1
10
1
1
1
1
1
1
1
1
s
1
1
1
1
1
1
1
1
1
r
1
1
1
1
1
1
1
1
r
i
i
i
i
i
D
3
3
3
3
Reference Value
(ppm)c
Sample
0.1
0.03
0.05
0.07
0.03
0.05
0.07
0.04
0.05
0.03
0.05
0.04
0.05
0.04
0.05
0.07
0.02
0.02
0.03
0.05
0.02
0.02
0.02
0.02
Final
0.03
0.03
0.03
0.02
a NT = neurotoxicity; DT = developmental toxicity; RT = reproductive toxicity; LT = liver toxicity; KT =
kidney toxicity
b A = animal-to-human (interspecies); H = interindiyidual (intraspecies); L = LOAEL-to-NOAEL; S =
subchronic-to-chronic duration; D = database deficiency
c Sample = reference value based on that particular endpoint, species, duration; Final = reference value for
the entire database for a particular duration of exposure.
"L indicates that this value is the HEC based on the LOAEL.
eA duration was not applied to the data from the developmental toxicity study, but should be considered
further.
B-22
-------
External Review Draft Do Not Cite or Quote
Table B-5. Derivation of Reference Values for Chemical X - Oral Exposure
Exposure
Duration
Acute
Short-
term
Longer-
term
Chronic
HED
(mg/k/d)
100Ld
1000L
900L
50
900L
50
900L
50
71
71
900L
50
36
36
36
36
Species
Human
Rat
Rat
Mouse
Rat
Mouse
Rat
Mouse
Rat
Mouse
Rat
Mouse
Rat
Mouse
Rat
Mouse
Type
of
Effect3
NT
NT
DT
DT
DT
DT
DT
DT
LT
LT
DT
DT
LT
LT
KT
KT
Uncertainty Factors'5
Total
~
~
~
1000
~
1000
~
1000
1000
1000
~
1000
1000
1000
1000
1000
A
~
~
~
10
~
10
~
10
10
10
~
10
10
10
10
10
H
~
~
~
10
~
10
~
10
10
10
~
10
10
10
10
10
L
~
~
~
1
~
1
~
1
1
1
~
1
1
1
1
1
s
~
~
~
1
~
1
~
lf
1
1
~
lf
1
1
1
1
D
10
10
10
10
Reference Value
(mg/k/d)c
Sample
No D-Re
NoD-R
NoD-R
0.05
NoD-R
0.05
NoD-R
0.05
0.07
0.07
NoD-R
0.05
0.04
0.04
0.04
0.04
Final
0.05
0.05
0.05
0.04
a NT = neurotoxicity; DT = developmental toxicity; RT = reproductive toxicity; LT = liver toxicity;
KT = kidney toxicity
b A = animal-to-hum an (interspecies); H = interindividual (intraspecies); L = LOAEL-to-NOAEL; S
= subchronic-to-chronic duration; D = database deficiency
c Sample = reference value based on that particular endpoint, species, duration; Final = reference
value for the entire database for a particular duration of exposure.
dL indicates that this value is the HED based on the LOAEL.
eNo D-R = no dose-response data; usually only one dose in the study.
*A duration was not applied to the data from the developmental toxicity study, but should be
considered further.
B-23
-------
External Review Draft Do Not Cite or Quote
GLOSSARY
NOTE: The following terms are used in this document. To the extent possible,
definitions were taken from other EPA sources, e.g., IRIS, the Children's Health Research
Strategy, the RfC Methodology. In some cases, the definitions have been revised from the
originals in IRIS for the sake of clarity or to be consistent with usage in this document. Those
terms and definitions that are changed and/or newly proposed in this document to be added to
IRIS are shown in italics and the definition(s) they are proposed to replace is indicated in
brackets. A number of other terms are included in the IRIS glossary that are not listed here,
simply because they were not used in this document.
Acute Exposure: One dose or multiple doses of short duration spanning less than or
equal to 24 hours, [current IRIS definition]
Acute Exposure: Exposure by the oral, dermal, or inhalation route for 24 hours or less.
[Proposed definition to replace the current Acute Exposure definition on IRIS.]
Adverse Effect: 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.
Benchmark Dose (BMD) or Concentration (BMC): A statistical lower confidence limit
on the dose that produces a predetermined change in response rate of an adverse effect (called the
benchmark response or BMR) compared to background, [current IRIS definition]
Benchmark Dose (BMD) or Concentration (BMC): A dose or concentration that
produces a predetermined change in response rate of an adverse effect (called the benchmark
response or BMR) compared to background. [Proposed definition to replace the current
definition on IRIS.]
BMDL or BMCL: A statistical lower confidence limit on the dose or concentration at the
BMD or BMC, respectively. [A new definition to be added to IRIS.]
Benchmark Response (BMR): An adverse effect, used to define a benchmark dose from
which an RfD (or RfC) can be developed. The change in response rate over background of the
BMR is usually in the range of 5-10%, which is the limit of responses typically observed in
well-conducted animal experiments.
Bioassay: An assay for determining the potency (or concentration) of a substance that
causes a biological change in experimental animals.
G-l
-------
External Review Draft Do Not Cite or Quote
Unavailability: The degree to which a substance becomes available to the target tissue
after administration or exposure.
Biologically Based Dose Response (BBDR) model: A predictive tool used to estimate
potential human health risks by describing and quantifying the key steps in the cellular, tissue and
organismal responses as a result of chemical exposure, [current IRIS definition]
Biologically Based Dose Response (BBDR) Model: A predictive model that describes
biological processes at the cellular and molecular level linking the target organ dose to the
adverse effect. [Proposed definition to replace the current definition on IRIS.]
Blood-to-air Partition Coefficient: A ratio of a chemical's concentration between blood
and air when at equilibrium.
Chronic Exposure: Multiple exposures occurring over an extended period of time, or a
significant fraction of the animal's or the individual's lifetime, [current IRIS definition]
Chronic Exposure: Repeated exposure by the oral, dermal, or inhalation route for more
than approximately 10% of the life span in humans (more than approximately 90 days to 2 years
in typically used laboratory animal species). [Proposed definition to replace the current
definition for Chronic Exposure on IRIS.]
Chronic Study: A toxicity study designed to measure the (toxic) effects of chronic
exposure to a chemical.
Critical Effect: The first adverse effect, or its known precursor, that occurs to the most
sensitive species as the dose rate of an agent increases.
Critical Study: The study that contributes most significantly to the qualitative and
quantitative assessment of risk. Also called Principal Study.
Developmental Toxicity: Adverse effects on the developing organism that may result
from exposure prior to conception (either parent), during prenatal development, or postnatally
until the time of sexual maturation. The major manifestations of developmental toxicity include
death of the developing organism, structural abnormality, altered growth, and functional
deficiency.
Dose: The amount of a substance available for interactions with metabolic processes or
biologically significant receptors after crossing the outer boundary of an organism. The potential
dose is the amount ingested, inhaled, or applied to the skin. The applied dose is the amount
presented to an absorption barrier and available for absorption (although not necessarily having
G-2
-------
External Review Draft Do Not Cite or Quote
yet crossed the outer boundary of the organism). The absorbed dose is the amount crossing a
specific absorption barrier (e.g., the exchange boundaries of the skin, lung, and digestive tract)
through uptake processes. Internal dose is a more general term denoting the amount absorbed
without respect to specific absorption barriers or exchange boundaries. The amount of the
chemical available for interaction by any particular organ or cell is termed the delivered or
biologically effective dose for that organ or cell. [New definition proposed to be added to IRIS]
Dose-Response Assessment: A determination of the relationship between the magnitude
of an administered, applied, or internal dose and a specific biological response. Response can be
expressed as measured or observed incidence, percent response in groups of subjects (or
populations), or as the probability of occurrence within a population, [current IRIS definition]
Dose-Response Assessment: A determination of the relationship between the magnitude
of an administered, applied, or internal dose and a specific biological response. Response can
be expressed as measured or observed incidence or change in level of response, percent response
in groups of subjects (or populations), or the probability of occurrence or change in level of
response within a population. [Proposed definition to replace the current definition on IRIS.]
Dose-Response Relationship: The relationship between a quantified exposure (dose),
and the proportion of subjects demonstrating specific, biological changes (response), [current
IRIS definition]
Dose-Response Relationship: The relationship between a quantified exposure (dose),
and the proportion of subjects demonstrating specific, biological changes in incidence or in
degree of change (response). [Proposed definition to replace the current definition on IRIS.]
Endpoint: An observable or measurable biological event or chemical concentration (e.g.,
metabolite concentration in a target tissue) used as an index of an effect of a chemical exposure.
Epidemiology: The study of disease patterns in human populations.
Epidemiology - The study of the distribution and determinants of health-related states or
events in specified populations, and the application of this study to the control of health
problems.
Exposure: Contact made between a chemical, physical, or biological agent and the outer
boundary of an organism. Exposure is quantified as the amount of an agent available at the
exchange boundaries of the organism (e.g., skin, lungs, gut).
G-3
-------
External Review Draft Do Not Cite or Quote
Exposure Assessment: An identification and evaluation of the human population exposed
to a toxic agent, describing its composition and size, as well as the type, magnitude, frequency,
route and duration of exposure.
Exposure Pathway: The physical course an environmental agent takes from the source to
the individual exposed.
Extrapolation, Low Dose: An estimate of the response at a point below the range of the
experimental data, generally through the use of a mathematical model.
Hazard: A potential source of harm.
Hazard Assessment: The process of determining whether exposure to an agent can
cause an increase in the incidence of a particular adverse health effect (e.g., cancer, birth defect)
and whether the adverse health effect is likely to occur in humans.
Hazard Characterization: A description of the potential adverse health effects
attributable to a specific environmental agent, the mechanisms by which agents exert their toxic
effects, and the associated dose, route, duration, and timing of exposure. [New definition
proposed to be added to IRIS]
Human Equivalent Concentration (HEC): The human concentration (for inhalation
exposure) of an agent that is believed to induce the same magnitude of toxic effect as the
experimental animal species concentration. This adjustment may incorporate toxicokinetic
information on the particular agent, if available or use a default procedure.
Human Equivalent Dose (HED): The human dose (for other than the inhalation routes
of exposure) of an agent that is believed to induce the same magnitude of toxic effect as the
experimental animal species dose. This adjustment may incorporate toxicokinetic information on
the particular agent, if available, or use a default procedure, such as assuming that daily oral doses
experienced for a lifetime are proportional to body weight raised to the 0.75 power.
Incidence: The number of new cases of a disease that develop within a specified
population over a specified period of time.
Incidence Rate: The ratio of new cases within a population to the total population at risk
given a specified period of time.
Latency Period: The time between exposure to an agent and manifestation or detection
of a health effect of interest.
G-4
-------
External Review Draft Do Not Cite or Quote
Linear dose response: A pattern of frequency or severity of biological response that
varies proportionately with the amount of dose of an agent, [current IRIS definition]
Linear Dose Response: A pattern of frequency or severity of biological response that
varies directly with the amount of dose of an agent. This linear relationship holds only at low
doses in the range of extrapolation. [Proposed definition to replace the current definition on
IRIS.]
Longer-Term Exposure: Repeated exposure by the oral, the dermal, or the inhalation
route for more than 30 days, up to approximately 10% of the life span in humans (more than 30
days up to approximately 90 days in typically used laboratory animal species). [Proposed new
definition to be used relative to the Longer-Term Reference Value. Similar to the current
definition for Subchronic Exposure. Because subchronic exposure studies will continue to be
used in risk assessment, the latter term should be retained as well, but replaced with the definition
for Longer-Term Exposure.]
Lowest-Observed-Adverse-Effect Level (LOAEL): The lowest exposure level at which
there are statistically or biologically significant increases in frequency or severity of adverse
effects between the exposed population and its appropriate control group. Also referred to as
lowest-effect level (LEL). [current IRIS and RfC Methodology definition]
Lowest-Observed-Adverse-Effect Level (LOAEL): The lowest exposure level at which
there are biologically significant increases in frequency or severity of adverse effects between
the exposed population and its appropriate control group. [Proposed to replace the current
definition in IRIS and the RfC Methodology, EPA, 1994]
Margin of Exposure (MOE): The LED10 or other point of departure divided by the
actual or projected environmental exposure of interest.
Mechanism of Action: The complete sequence of biological events that must occur to
produce the toxic effect.
Mode of Action (MOA): A less-detailed description of the mechanism of action in which
some but not all of the sequence of biological events leading to a toxic effect is known.
Modifying Factor (MF): A factor used in the derivation of a reference dose or reference
concentration. The magnitude of the MF reflects the scientific uncertainties of the study and
database not explicitly treated with standard uncertainty factors (e.g., the completeness of the
overall database). A MF is greater than zero and less than or equal to 10, and the default value for
theMFisl.
G-5
-------
External Review Draft Do Not Cite or Quote
No-Observed-Adverse-Effect Level (NOAEL): An highest exposure level at which
there are no statistically or biologically significant increases in the frequency or severity of adverse
effect between the exposed population and its appropriate control; some effects may be produced
at this level, but they are not considered adverse, nor precursors to adverse effects, [current IRIS
and RfC Methodology definition]
No-Observed-Adverse-Effect Level (NOAEL): The highest exposure level at which there
are no biologically significant increases in the frequency or severity of adverse effect between
the exposed population and its appropriate control; some effects may be produced at this level,
but they are not considered adverse, nor precursors to adverse effects. [Proposed to replace the
current definition in IRIS and the RfC Methodology, EPA, 1994]
Non-linear dose response: A pattern of frequency or severity of biological response that
does not vary proportionately with the amount of dose of an agent. When mode of action
information indicates that responses may not follow a linear pattern below the dose range of the
observed data, non-linear methods for determining risk at low dose may be justified, [current IRIS
definition]
Non-Linear Dose Response: A pattern of frequency or severity of biological response
that does not vary directly with the amount of dose of an agent. When mode of action
information indicates that responses may fall more rapidly than dose below the range of the
observed data, non-linear methods for determining risk at low dose may be justified. [Proposed
definition to replace the current definition on IRIS.]
Pharmacodynamics: The determination and quantification of the sequence of events at
the cellular and molecular levels leading to a toxic response to an environmental agent (also called
toxicodynamics). [New definition proposed to be added to IRIS]
Pharmacokinetics: The determination and quantification of the time course of absorption,
distribution, biotransformation, and excretion of chemicals (also called toxicokinetics). [New
definition proposed to be added to IRIS]
Physiologically Based Pharmacokinetic (PBPK) Model: Physiologically based
compartmental model used to characterize pharmacokinetic behavior of a chemical. Available data
on blood flow rates, and metabolic and other processes which the chemical undergoes within each
compartment are used to construct a mass-balance framework for the PBPK model, [current
IRIS definition]
Physiologically Based Pharmacokinetic (PBPK) Model: A model that estimates the dose
to a target tissue or organ by taking into account the rate of absorption into the body,
distribution among target organs and tissues, metabolism, and excretion. [Proposed definition
to replace the current definition on IRIS.]
G-6
-------
External Review Draft Do Not Cite or Quote
Point of Departure: The dose-response point that marks the beginning of a low-dose
extrapolation. This point is most often the upper bound on an observed incidence or on an
estimated incidence from a dose-response model, [current IRIS definition]
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 or LOAELfor an observed
incidence, or change in level of response. [Proposed definition to replace the current definition
on IRIS.]
Ppb: A unit of measure expressed as parts per billion. Equivalent to 1 x 10-9.
Ppm: A unit of measure expressed as parts per million. Equivalent to 1 x 10-6.
Prevalence: The proportion of disease cases that exist within a population at a specific
point in time, relative to the number of individuals within that population at the same point in
time.
Reference Concentration (RfC): An estimate (with uncertainty spanning perhaps an
order of magnitude) of a continuous inhalation exposure to the human population (including
sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a
lifetime. It can be derived from a NOAEL, LOAEL, or benchmark concentration, with uncertainty
factors generally applied to reflect limitations of the data used. Generally used in EPA's noncancer
health assessments, [current IRIS definition]
Reference Dose (RfD): An estimate (with uncertainty spanning perhaps an order of
magnitude) of a daily oral exposure to the human population (including sensitive subgroups) that
is likely to be without an appreciable risk of deleterious effects during a lifetime. It can be derived
from a NOAEL, LOAEL, or benchmark dose, with uncertainty factors generally applied to reflect
limitations of the data used. Generally used in EPA's noncancer health assessments, [current IRIS
definition]
Reference Value (RjV): An estimate of an exposure for [a given duration] to the human
population that is likely to be without an appreciable risk of adverse effects for a lifetime
(including susceptible subgroups). It can be derived from a benchmark dose, NOAEL or
LOAEL, with uncertainty/variability factors generally applied to reflect limitations of the data
used. The application of these factors is intended to provide an estimate centered within an
order of magnitude. [Durations include acute, short-term, longer-term, and chronic, and are
defined individually in this glossary. This definition is proposed to replace those for the Reference
Dose (RfD) and Reference Concentration (RfC). A subscript would be used with the RfV to
denote route and duation, e.g., RfVAO, for the Acute Oral Reference Value.]
G-7
-------
External Review Draft Do Not Cite or Quote
Regional Deposited Dose (RDD): The deposited dose of particles calculated for a
respiratory tract region of interest (r) as related to an observed toxicity. For respiratory effects of
particles, the deposited dose is adjusted for ventilatory volumes and the surface area of the
respiratory region affected (mg/min-sq. cm). For extra respiratory effects of particles, the
deposited dose in the total respiratory system is adjusted for ventilatory volumes and body weight
(mg/min-kg).
Regional Deposited Dose Ratio (RDDR): The ratio of the regional deposited dose
calculated for a given exposure in the animal species of interest to the regional deposited dose of
the same exposure in a human. This ratio is used to adjust the exposure effect level for
interspecies dosimetric differences to derive a human equivalent concentration for particles.
Regional Gas Dose: The gas dose calculated for the region of interest as related to the
observed effect for respiratory effects. The deposited dose is adjusted for ventilatory volumes and
the surface area of the respiratory region affected (mg/min-sq.cm).
Regional Gas Dose Ratio (RGDR): The ratio of the regional gas dose calculated for a
given exposure in the animal species of interest to the regional gas dose of the same exposure in
humans. This ratio is used to adjust the exposure effect level for interspecies dosimetric
differences to derive a human equivalent concentration for gases with respiratory effects.
Risk (in the context of human health): The probability of injury, disease, or death from
exposure to a chemical agent or a mixture of chemicals. In quantitative terms, risk is expressed in
values ranging from zero (representing the certainty that harm will not occur) to one (representing
the certainty that harm will occur). The following are examples of how risk is expressed within
IRIS: E-4 or 10-4 = a risk of 1/10,000; E-5 or 10-5 = 1/100,000; E-6 or 10-6 = 1/1,000,000.
Similarly, 1.3 E-3 or 1.3 x 10-3 = a risk of 1.3/1,000=1/770; 8 E-3 or 8 x 10-3 = a risk of 1/125
and 1.2 E-5 or 1.2 x 10-5 = a risk of 1/83,000. [current IRIS Definition]
Risk: The probability of adverse effects resulting from exposure to an environmental
agent or mixture of agents. [Proposed definition to replace the current definition on IRIS.]
Risk Characterization: The integration of information on hazard, exposure, and dose-
response to provide an estimate of the likelihood that any of the identified adverse effects will
occur in exposed people. [New definition proposed to be added to IRIS]
Risk Assessment (in the context of human health): The determination of potential
adverse health effects from exposure to chemicals, including both quantitative and qualitative
expressions of risk. The process of risk assessment involves four major steps: hazard
identification, dose-response assessment, exposure assessment, and risk characterization, [current
IRIS Definition]
G-8
-------
External Review Draft Do Not Cite or Quote
Risk Assessment: The evaluation of scientific information on the hazardous properties of
environmental agents (hazard characterization), the dose-response relationship (dose-response
assessment), and on the extent of human exposure to those agents (exposure assessment). The
product of the risk assessment is a statement regarding the probability that populations or
individuals so exposed will be harmed and to what degree (risk characterization). [Proposed
definition to replace the current definition on IRIS.]
Short-Term Exposure: Multiple or continuous exposure to an agent for a short period of
time, usually one week, [current IRIS Definition]
Short-Term Exposure: Repeated exposure by the oral, dermal, or inhalation route for
more than 24 hours, up to 30 days. [Proposed definition to replace the current definition for
Short-Term Exposure on IRIS.]
Statistical Significance: The probability that a result likely to be due to chance alone. By
convention, a difference between two groups is usually considered statistically significant if
chance could explain it only 5% of the time or less. Study design considerations may influence the
a priori choice of a different statistical significance level, [current IRIS definition]
Statistical Significance: The probability that a result is not likely to be due to chance
alone. By convention, a difference between two groups is usually considered statistically
significant if chance could explain it only 5% of the time or less. Study design considerations
may influence the a priori choice of a different level of statistical significance. [Proposed
definition to replace the current definition on IRIS]
Subchronic Exposure: Exposure to a substance spanning approximately 10% of the
lifetime of an organism, [see note for Longer-Term Exposure]
Subchronic Study: A toxicity study designed to measure effects from subchronic
exposure to a chemical.
Supporting Studies: Studies that contain information useful for providing insight and
support for conclusions.
Susceptible Subgroups: May refer to life stages, e.g., children or the elderly, or to other
segments of the population, e.g., asthmatics or the immune-compromised, but are likely to be
somewhat chemical-specific, and may not be consistently defined in all cases. [New definition
proposed to be added to IRIS]
Susceptibility: Increased likelihood of an adverse effect, often discussed in terms of
relationship to a factor that can be used to describe a human subpopulation (e.g., lifestage,
demographic feature, or genetic characteristic). [New definition proposed to be added to IRIS]
G-9
-------
External Review Draft Do Not Cite or Quote
Systemic Effects or Systemic Toxicity: Toxic effects as a result of absorption and
distribution of a toxicant to a site distant from its entry point, at which point effects are produced.
Not all chemicals that produce systemic effects cause the same degree of toxicity in all organs.
[current IRIS definition]
Systemic Effects or Systemic Toxicity: Toxic effects as a result of absorption and
distribution of a toxicant to a site distant from its entry point. [Proposed definition to replace the
current definition on IRIS.]
Target Organ: The biological organ(s) most adversely effected by exposure to a chemical
substance, [current IRIS definition]
Target Organ: The biological organ(s) most adversely affected by exposure to a
chemical or physical agent. [Proposed definition to replace the current definition on IRIS.]
Threshold: The dose or exposure below which no deleterious effect is expected to occur.
Toxicity: The degree to which a chemical substance elicits a deleterious or adverse effect
upon the biological system of an organism exposed to the substance over a designated time
period, [current IRIS definition]
Toxicity: Deleterious or adverse biological effects elicited by a chemical, physical, or
biological agent. [Proposed definition to replace the current definition on IRIS.]
Toxicology: The study of harmful interactions between chemicals and biological systems.
[current IRIS definition]
Toxicology: The study of harmful interactions between chemical, physical agents, or
biological agents and biological systems. [Proposed definition to replace the current definition
on IRIS.]
Toxic Substance: A chemical substance or agent which may cause an adverse effect or
effects to biological systems, [current IRIS definition]
Toxic Substance: A chemical, physical, or biological agent that may cause an adverse
effect or effects to biological systems. [Proposed definition to replace the current definition on
IRIS.]
Uncertainty: Occurs because of a lack of knowledge. It is not the same as variability.
For example, a risk assessor my be very certain that different people drink different amounts of
water, but may be uncertain about how much variability there is in water intakes within the
population. Uncertainty can often be reduced by collecting more and better data, while variability
G-10
-------
External Review Draft Do Not Cite or Quote
is an inherent property of the population being evaluated. Variability can be better characterized
with more data, but it cannot be reduced or eliminated. Efforts to clearly distinguish between
variability and uncertainty are important for both risk assessment and risk characterization. [New
definition proposed to be added to IRIS]
Uncertainty Factor (UF): One of several, generally 10-fold factors, used in operationally
deriving the RfD and RfC from experimental data. UFs are intended to account for (1) the
variation in sensitivity among the members of the human population, i.e., interhuman or
intraspecies variability; (2) the uncertainty in extrapolating animal data to humans, i.e.,
interspecies variability; (3) the uncertainty in extrapolating from data obtained in a study with
less-than-lifetime exposure to lifetime exposure, i.e., extrapolating from subchronic to chronic
exposure; (4) the uncertainty in extrapolating from a LOAEL rather than from a NOAEL; and (5)
the uncertainty associated with extrapolation from animal data when the data base is incomplete.
[current IRIS definition]
Uncertainty/Variability Factors (UFs): One of several, generally 10-fold default
factors, used in operationally deriving the RfD and RfC from experimental data. The factors are
intended to account for (1) the variation in sensitivity among the members of the human
population, i.e., inter-individual variability; (2) the uncertainty in extrapolating animal data to
humans, i.e., interspecies uncertainty; (3) the uncertainty in extrapolating from data obtained in
a study with less-than-lifetime exposure to lifetime exposure, i.e., extrapolating from subchronic
to chronic exposure; (4) the uncertainty in extrapolating from a LOAEL rather than from a
NOAEL; and (5) the uncertainty associated with extrapolation when the database is incomplete.
[Proposed definition to replace the current one for Uncertainty Factor on IRIS.]
Variability: Refers to true heterogeneity or diversity. For example, among a population
that drinks water from the same source and with the same contaminant concentration, the risks
from consuming the water may vary. This may be due to differences in exposure (i.e., different
people drinking different amounts of water, having different body weights, different exposure
frequencies, and different exposure durations) as well as differences in response (e.g., genetic
differences in resistance to a chemical dose). Those inherent differences are referred to as
variability. Differences among individuals in a population are referred to as inter-individual
variability, while differences for one individual over time is referred to as intra-individual
variability. [New definition proposed to be added to IRIS]
G-ll
------- |