United States       Science Advisory Board      EPA-SAB-EHC-99002
Environmental      Washington, DC        November 1998
Protection Agency              www.epa.gov/sab

AN SAB REPORT: REVIEW

OF THE RfC METHODS

CASE STUDIES
REVIEW OF CASE STUDIES
ASSOCIATED WITH THE
DOCUMENT METHODS FOR
DERIVATION OF INHALATION
REFERENCE CONCENTRATIONS
AND APPLICATION OF
INHALATION DOSIMETRY
(EPA/600/8-90/066F) BY THE
ENVIRONMENTAL HEALTH
COMMITTEE OF THE SCIENCE
ADVISORY BOARD (SAB)

-------
                                   November 17, 1998
EPA-SAB-EHC-99-003

Honorable Carol M. Browner
Administrator
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460

              Subject:      Review of the Sixteen Case Studies associated with the document
                           Methods for Derivation of Inhalation Reference Concentrations
                           and Application of Inhalation Dosimetry, October 1994,
                           (EPA/600/8-90/066F)

Dear Ms. Browner:

       At the request of the Office of Research and Development (ORD), National Center for
Environmental Assessment, the Environmental Health Committee (EHC) of the Environmental
Protection Agency's Science Advisory Board (SAB) reviewed the Agency's Inhalation Reference
Concentrations (RfC) Methods Case Studies. The Committee met on June 9-10, 1998 at the
EPA's Environmental Research Center in Research Triangle Park, North Carolina.

       The ORD's National Center for Environmental Assessment (NCEA) developed the
Inhalation Reference Concentrations in response to the Clean Air Act Amendments (CAAA) of
1990.  The CAA requires sources to demonstrate negligible risk and lack of residual risk (after
implementation of control technology) based on health risk estimates. It is anticipated that the
RfCs will be used for CAAA, as part of the determination of negligible and residual risk for
noncancer health effects of air toxics.  In addition, it is anticipated that air pollution control offices
at the regional, state and local level will continue to utilize RfC values in risk management
programs.

       The RfC is an estimate of a daily inhalation exposure to the human population, including
sensitive subgroups, that are likely to be without an appreciable risk of deleterious noncancer
effects during a lifetime. The methodology derives RfCs from dose-response estimates for
noncancer effects.  NCEA has estimated that the uncertainty associated with the RfC  may span an
order of magnitude.

       The SAB Environmental Health Committee (in 1989 and 1990) reviewed previous
versions of the inhalation RfC methodology.  The current version of the methodology, Methods
for Derivation of Inhalation Reference Concentrations and Application of Inhalation Dosimetry,
represents the Agency's response to the EHC's comments arising from these two reviews, which

-------
were incorporated in two reports: Science Advisory Board's Comments on the Use of
Uncertainty an Modifying Factors in Establishing Reference Dose Level., January 17, 1990
(EPA-SAB-EHC-90-005); and Science Advisory Board's Review of the Office of Research and
Development Document Interim Methods for Development of Inhalation Reference
Concentrations, April 29, 1991 (EPA-SAB-EHC-91-008).

       The EHC requested the opportunity to review case studies using the methods to
demonstrate the application of the dosimetric adjustments and to illustrate the methodology
applied to chemicals that are representative of the typical range of data available including those
with human occupational  or clinical information and those with databases considered to be
insufficient for quantitative dose-response estimation (i.e., "not-verifiable). This review was not
intended to be a review of the RfC methods themselves or a review of the science regarding
the specific chemicals addressed in the case studies, but rather of the conceptual framework
of the approach as applied to representative data.

       The case studies reviewed by the Committee fell into  one  of four categories: a) particle
case studies; b) category 1 gas case studies; c) category 3 gas case studies; and d) "not-verifiable"
case studies (which had less satisfactory databases than most  of the other case studies and for
which the EPA did not set RfCs).

       The Office of Research and Development requested that the EHC provide comment(s) on
each of the following aspects of the RfC Case Studies:

       a)      Overall, are the concepts and applications of the RfC methodology clearly
              articulated in the documentation provided for the case studies? Do the decisions
              and choices in these files attain the Agency's goal of being "transparent, clear, and
              reasonable? If not, what are specific examples within these files that could be
              instituted to better attain this goal?

       b)      In derivation of the RfC in the specific case studies,

              (1)     Are the study  summaries presented in sufficient detail for reader
                     evaluation?

              (2)     Are the designations of the critical effect and effects levels: no-observable-
                     adverse-effect-level, lowest-observable-adverse-effect level, benchmark
                     dose/concentration (NOAEL/LOAEL/BMC), based on rationales that are
                     clear and reasonable?

              (3)     Of the studies presented in either the IRIS Summary or Toxicological
                     Reviews of each chemical, has the Principal studies been selected in a
                     consistent and rational manner? Does  this  choice reflect consideration on
                     the current knowledge of potential human response?

-------
              (4)    Have the underlying assumptions of the duration and dosimetry
                    adjustments been presented clearly?

              (5)    Are the rationales presented for use of uncertainty factors clear, reasonable
                    and consistent?

              (6)    Do the confidence statements reflect the strengths and limitations (e.g.,
                    relevancy to humans, comprehensiveness of the data base) of the RfC
                    assessment in a manner consistent with the Agency's goals?

       c)      In the IRIS Summaries for the specific cases, numerous studies are included under
              the heading "Supporting/Additional Studies" that are meant to provide further
              support for designation of the critical effect (e.g., mechanistic data, human data) or
              for the effect level chosen in the Principal study, or to establish the completeness
              of the data base.  Is the depth of presentation in this section sufficiently
              comprehensive to provide information supportive of the decisions made in the
              assessment (such as uncertainly factors and confidence levels)?

       The Environmental Health Committee commends the Agency for demonstrating the
application of both the dosimetric adjustments and the RfC methodology to chemicals
representative of the typical range of data available (including those with human occupational or
clinical information as well as those with databases considered to be insufficient for quantitative
dose-response estimation).

       The EHC found the concepts and application of the RfC methodology to be clearly
articulated in some of the case studies and unclear in others.  The same finding also held true with
the derivation of the RfC. In general, the Committee found the concepts and application of the
RfC methodology in the recently updated or created case studies, such as those for vinyl chloride
and methyl methacrylate, to be clearly articulated. To the contrary, the case studies that did not
include the more recent research findings were criticized by the EHC. Examples of those case
studies include carbon disulfide and antimony trioxide. The IRIS Summaries for some of the case
studies were sufficiently comprehensive while others were not.  It is clear that the EPA knows
how to prepare comprehensive IRIS Summaries. The  main problem  is how to update the old
documents to meet current standards.  For some RfC case studies, there  was a difference of
opinion amongst the Committee regarding the clarity of the document, the rational for the RfC
derivation and the comprehensiveness of the summary. The Committee identified specific areas in
the RfC case studies that should  be improved and made the following overall recommendations:

       a)      In general, it was difficult to read the narrative for an RfC case study document
              and compare all the results from all the studies presented  for a given chemical.  To
              improve the clarity  of the documents, the Agency should  summarize some of the
              data using figures and tables.

-------
b)     When appropriate, the Agency should update the RfCs when more recent studies
       become available.

c)     In some cases, the Agency did not include available human data when developing
       the RfC.  The EPA should incorporate the human data, when available.

d)     In developing the RfCs, the Agency should expand their review of the literature to
       include new models, both laboratory animal-derived and mathematical.

e)     Each document should contain a section that addresses children and their unique
       exposures and vulnerabilities, and that delineates data gaps. Each document
       should also include a statement on whether the RfC is protective of children.

f)     The definition of susceptible human populations should be consistent throughout
       all of the RfC documents. The EPA should also include a  discussion on how it
       deals with susceptible subpopulations such as children with asthma or children with
       cystic fibrosis.

g)     All available data reviewed for a given document should be listed and the reasons
       for inclusion or exclusion from the derivation of the RfC should be provided in an
       appendix.

h)     Several of the RfC case studies used acronyms and scientific terminology that were
       not defined in the document.  Each RfC document should  stand on its own and
       should not require frequent referral to the RfC background documents.  Therefore,
       all of the documents should have a glossary, explaining the scientific terminology.
       In addition, the acronyms should be accompanied by their full names when they are
       first used in the RfC documents.

i)     Some of the calculations were unclear. An explanation of the calculations should
       be included, in some of the case studies,  and improved, in  other case studies.

j)     The units of measure should be consistent throughout the RfC documents.  If two
       units are used, one should be followed in parenthesis by the other. For example,
       408 ppm (68 mg/m3).

k)     Each RfC case study document should include an illustration of the respective
       chemical structure.

1)     The Agency should reassess the application of uncertainly  factors in the
       development of the RfC.  It appears that the Agency uses the same uncertainty
       factor whether or not there are data on other, possibly significant (but less frank)

-------
              endpoints, and when there is marginal dosimetry information or a fully developed
              PBPK model. This practice will tend to discourage additional research.

       The Committee appreciates the opportunity to review the RfC Methods Case Studies and
looks forward to receiving a written response from the Assistant Administrator for Research and
Development.

                                  Sincerely,
                                        /signed/
                                  Dr. Joan M. Daisey, Chair
                                  Science Advisory Board
                                  Dr. Emil A. Pfitzer, Chair
                                  Environmental Health Committee
                                  Science Advisory Board
                                  Dr. Mark J. Utell, Co-Chair,
                                  Environmental Health Committee
                                  Science Advisory Board

-------
                                       NOTICE
       This report has been written as part of the activities of the Science Advisory Board, a
public advisory group providing extramural scientific information and advice to the Administrator
and other officials of the Environmental Protection Agency.  The Board is structured to provide
balanced, expert assessment of scientific matters related to problems facing the Agency.  This
report has not been reviewed for approval by the Agency and, hence, the contents of this report
do not necessarily represent the views and policies of the Environmental Protection Agency, nor
of other agencies in the Executive Branch of the Federal government, nor does mention of trade
names or commercial products constitute a recommendation for use.

-------
                                     ABSTRACT

       The Environmental Health Committee (EHC) reviewed the EPA's Inhalation Reference
Concentration (RfC) Methods Case Studies for selected chemicals.  The Committee commends
the Agency's efforts to demonstrate the application of the dosimetric adjustments and to illustrate
the methodology.

       The EHC found the concepts  and application of the RfC methodology to be articulated
clearly in some of the case studies and unclear in others. Similarly, the Committee  concurred with
the derivation of the RfC in some case studies and had concerns about the derivation in others.
The same findings also held for the IRIS Summaries. For some of the case studies, there was a
difference in opinion amongst the EHC regarding the clarity of the documents, the  derivation of
the RfC and/or the comprehensiveness of the summary.

       The Committee made several  recommendations for improvement: a) improve the clarity of
the documents by summarizing some of the data using figures and tables; b) include more recent
studies in the RfC case studies; c) incorporate human data into the derivation of the RfC, when
available; d) expand the case studies to include a review of the newer models; e) include a
statement on children, and whether the RfC is protective of children; f) explain the  term
susceptible population; g) give reasons for including or excluding available data; h) define
scientific terminology used in the documents; i) clarify the calculations; j) make the units
consistent; k) provide chemical structures; and 1) reassess the application of uncertainty factors in
the development of the RfC.
       Keywords:   Inhalation Reference Concentration (RfC), RfC methods case studies,
                    Integrated Risk Information System (IRIS)

-------
                     U.S. Environmental Protection Agency
                             Science Advisory Board
                        Environmental Health Committee
                         RfC Case Studies Review Panel

Chair
Dr. Emil Pfitzer, Retired, Ramsey, NJ

Co-Chair
Dr. Mark J. Utell, Acting Chairman, Department of Medicine, Director, Pulmonary Unit, and
      Professor of Medicine and Environmental Medicine, University of Rochester Medical
      Center, Rochester, NY

Members
Dr. Cynthia Bearer, Assistant Professor, Division of Neonatology, Department of Pediatrics and
      Department of Neurosciences, Case Western Reserve University, Cleveland, OH

Dr. Adolfo Correa, Associate Professor, Department of Epidemiology, The Johns Hopkins
      University School of Hygiene and Public Health, Baltimore, MD (Did not attend meeting)

Dr. John Doull, Professor Emeritus, Department of Pharmacology, Toxicology and
      Therapeutics, University of Kansas Medical Center, Kansas City, KS

Dr. David G. Hoel, Distinguished University Professor, Department of Biometry and
      Epidemiology, Medical University of South Carolina, Charleston, SC

Dr. Abby A. Li, Toxicology Manager/Neurotoxicology Technical Leader, Monsanto Company,
      St. Louis, MO

Dr. Michele Medinsky, Toxicology Consultant, Durham, NC

Dr. Frederica Perera, Professor of Public Health, Division of Environmental Health Sciences,
      Columbia University, New York, NY (Did not attend meeting)

Dr. Lauren Zeise, Chief, Reproductive and Cancer Hazard Assessment Section, Office of
      Environmental Health Hazard Assessment, California Environmental Protection Agency,
      Berkeley, CA (Did not attend meeting)

Consultants
Dr. Kenny Crump, Vice President, KS Crump Group, Inc., Ruston, LA

Dr. Rogene Henderson, Senior Scientist, Inhalation Toxicology Research Institute,
      Albuquerque, NM

                                          iii

-------
Dr. Richard Schlesinger, Professor, Department of Environmental Medicine, New York
       University School of Medicine, Tuxedo, NY

Dr. Ron Wyzga, Technical Executive, Electric Power Research Institute, Palo Alto, CA1

Science Advisory Board Staff
Ms. Roslyn A. Edson, Designated Federal Officer, U. S. Environmental Protection Agency,
       Science Advisory Board (1400), 401 M Street, SW, Washington, DC 20460,

Mr. Samuel Rondberg, Designated Federal Officer, U. S. Environmental Protection Agency,
       Science Advisory Board (1400), 401 M Street, SW, Washington, DC 20460,2

Ms. Mary L. Winston, Management Assistant, Environmental Protection Agency, Science
       Advisory Board (1400), 401 M Street, SW, Washington, DC  20460
       JDid not attend the public meeting, but participated in the report preparation.
       2Did not attend the public meeting but provided editorial support for this report.
                                           IV

-------
                            TABLE OF CONTENTS


1.  EXECUTIVE SUMMARY  	1

2.  INTRODUCTION	4
      2.1 Background  	4
      2.2 Charge  	5

3.  RESPONSE TO THE CHARGE 	7
      3.1 General comments on the RfC Case Studies  	7
      3.2 Group 1: Particle Case Studies  	7
             3.2.1  Diphenylmethane Diisocyanate (MDI)	7
             3.2.2 Antimony trioxide	8
             3.2.3  Phosphoric Acid	11
      3.3 Group 2: Category 1 Gas Case Studies	13
             3.3.1  Methymethacrylate	13
             3.3.2 1,3-Dichloropropene	14
             3.3.3  Acetaldehyde  	14
             3.3.4 Chlorine dioxide	15
      3.4 Group 3: Category 3 Gas Case Studies	16
             3.4.1  Carbon disulfide	16
             3.4.2 n-Hexane  	17
             3.4.3  Hydrogen cyanide	18
             3.4.4 Vinyl chloride	19
             3.4.5  2-Ethoxyethanol	20
      3.5 Group 4: Not-Verifiable Case Studies	20
             3.5.1  Acrylamide	21
             3.5.2 Bis(chloromethyl)ether (BCME)	21
             3.5.3  Caprolactam	21
             3.5.4 Methoxychlor	22
      3.6 General Findings from RfC Case Studies	22

4.  SUMMARY OF RECOMMENDATIONS	23

APPENDIX A - ACRONYMS AND ABBREVIATIONS	A-l

REFERENCES	R-l

-------
                            1.  EXECUTIVE SUMMARY
       The Reference Concentration (RfC) is an estimate of a daily inhalation exposure to the
human population, including sensitive subgroups, that are likely to be without an appreciable risk
of deleterious noncancer effects during a lifetime. The uncertainty for the RfC estimates spans an
order of magnitude.  The EPA Office of Research and Development developed the RfC to meet
the Agency's mandate under the Clean Air Act to demonstrate negligible risk and lack of residual
risk (after implementation of control technology), based on human risk estimates.

       On June 9-10,  1998, the Environmental Health Committee met at the EPA's
Environmental Research Center in Research Triangle Park, North Carolina to review the
Agency's RfC Methods Case Studies.  There were four groups of case studies: a) particle case
studies; b) category  1 gas case studies; c) category 3 gas case studies; and d) not-verifiable case
studies. It is important to note that this was not a review of the science regarding the specific
chemicals addressed in the case studies. This review also did not focus on the RfC methodology.

       The Charge was to  provide comments on each of the following aspects of the RfC Case
Studies:

       a)     Overall, are the concepts and applications of the RfC methodology  clearly
              articulated in the documentation provided for the case studies? Do the decisions
              and choices in these files attain the Agency's goal of being "transparent, clear, and
              reasonable? If not, what are specific examples within these files that could be
              instituted to better attain this goal?

       b)     In derivation of the RfC in the specific case studies:

              (1)     are the study summaries presented in sufficient detail for the reader's
                     evaluation?

              (2)     are the designations of the critical effect and effects levels
                     (NOAEL/LOAEL/BMC) based on rationales that are clear and reasonable?

              (3)     of the studies presented in either the IRIS Summary or Toxicological
                     Reviews of each chemical, have the Principal studies been selected in a
                     consistent and rational manner? Does this choice reflect consideration on
                     the current knowledge of potential human response?

              (4)     Have the underlying assumptions of the duration and dosimetry
                     adjustments been presented clearly?

-------
              (5)    are the rationales presented for use of uncertainty factors clear, reasonable
                    and consistent?

              (6)    do the confidence statements reflect the strengths and limitations (e.g.,
                    relevancy to humans, comprehensiveness of the data base) of the RfC
                    assessment in a manner consistent with the Agency's goals?

       c)      In the IRIS Summaries for the specific cases, numerous studies are included under
              the heading "Supporting/Additional Studies" that are meant to provide further
              support for designation of the critical effect (e.g.,  mechanistic data, human data) or
              for the effect level chosen in the Principal study, or to establish the completeness
              of the data base.  Is the depth of presentation in this section sufficiently
              comprehensive to provide information supportive of the decisions made in the
              assessment (such as uncertainty factors and confidence levels)?

       The EHC found the concepts and application of the RfC methodology to be clearly
articulated in some of the case studies and unclear in other case studies.  The EHC concurred with
the derivation of the RfC in some case studies and had concerns about the derivation of the RfC in
other case studies.  In general, the Committee found the concepts and application of the RfC
methodology in the recently updated or created case studies, such as those for vinyl chloride and
methyl methacrylate, to be clearly articulated. To the contrary, the case studies which did not
include the more recent research findings were criticized by the EHC. Examples of those case
studies include carbon disulfide and antimony trioxide. The IRIS Summaries for some of the case
studies were sufficiently comprehensive while others were not. It is clear that the EPA knows
how to prepare comprehensive IRIS Summaries.  The main problem is how to update the old
documents to meet current standards.  For some RfC  case studies, there was a difference of
opinion amongst the Discussants regarding the clarity of the document, the rational for the RfC
derivation and the comprehensiveness of the summary. The Committee identified specific areas in
the RfC case studies that should be improved and makes the following overall recommendations:

       a)      In general, it was difficult to read the narrative on an RfC case study document  and
              compare all the results from all the studies presented for a given chemical.  To
              improve the clarity of the documents, the Agency should summarize some of the
              data using figures and tables.

       b)      When appropriate, the Agency should update the RfCs when more recent studies
              become available.

       c)      In some cases, the Agency did not include available human data when developing
              the RfC.  The EPA should incorporate the human data, when available.

       d)      In developing the RfCs, the Agency should expand their review of the literature to
              include new models, both laboratory animal and mathematical.

-------
e)     Each document should contain a section on children and their unique exposures
       and vulnerabilities, and delineate data gaps. Each document should also include a
       statement on whether the RfC is protective of children.

f)     The definition of susceptible human populations should be consistent throughout
       all of the RfC documents. The EPA should also include a discussion on how it

       deals with susceptible subpopulations such as children with asthma or children with
       cystic fibrosis.

g)     All available data reviewed for a given document should be listed and the reasons
       for inclusion or exclusion from the derivation of the RfC should be provided in an
       appendix.

h)     Several of the RfC case studies used acronyms and scientific terminology that were
       not defined in the document. Each RfC document should stand on its own and
       should not require frequent referral to the RfC background documents. Therefore,
       all of the documents should have a glossary, explaining the scientific terminology.
       In addition, the acronyms should be accompanied by their full names when they are
       first used in the RfC documents.

i)     Some of the calculations were unclear. An explanation  of the calculations should
       be included, in some of the case studies, and improved,  in other case studies.

j)     The units of measure should be consistent throughout the RfC documents. If two
       units are used, one should be followed in parenthesis by the other.  For example,
       408 ppm (68 mg/m3).

k)     Each RfC case study document should include an illustration of the respective
       chemical structure.

1)     The Agency should reassess the application of uncertainty factors in the
       development of the RfC. It appears that the Agency uses the same uncertainty
       factor whether or not there are data on more discriminating endpoints and when
       there is marginal dosimetry information or a fully developed PBPK model.  This
       practice will tend to discourage additional  research.

m)    The underlying assumptions of the duration and dosimetry adjustments need
       further clarification in the calculation of a regional deposition dose ratio (RDDR)
       along with an explanation on the significance of the adjustments.

-------
                                2.  INTRODUCTION
2.1 Background

       The RfC is an estimate, with uncertainty spanning an order of magnitude, of a daily
inhalation exposure to the human population, including sensitive subgroups, that is likely to be
without an appreciable risk of deleterious noncancer effects during a lifetime.  The methodology
derives RfCs from dose-response estimates for noncancer effects. The RfCs were developed by
the Agency to meet its mandate under the Clean Air Act Amendments (CAAA) of 1990 to require
sources to demonstrate negligible risk and lack of residual risk, after implementation of control
technology, based on health risk estimates.

       The inhalation RfC methodology was developed according to the oral reference dose
(RfD) paradigm and emphasizes portal-of-entry considerations of comparative toxicity and
inhalation dosimetry for particles and gases. Extrapolation modeling was added and factors are
derived for adjustment of exposure concentrations that account for dosimetric differences
between experimental animal species and humans.  Previous versions of the RfC methodology
have undergone external peer review, including reviews by the Environmental Health Committee
in 1989 and  1990. These reviews generated two reports: Science Advisory Board's Comments on
the Use of Uncertainty an Modifying Factors in Establishing Reference Dose Level, January 17,
1990 (SAB,  1990); and Science Advisory Board's Review of the Office of Research and
Development Document Interim Methods for Development of Inhalation Reference
Concentrations, April 29, 1991 (SAB, 1991).

       The current version of the RfC methodology is described in Methods for Derivation of
Inhalation Reference Concentrations and Application of Inhalation Dosimetry (EPA, 1994).
This version incorporates the EPA's response to the EHC's previous comments.

       At its 1990 review of the RfC methodology, the EHC requested the opportunity to review
case studies using the methods in order to demonstrate the application of the dosimetric
adjustments and illustrate the methodology applied to chemicals representative of the typical
range of data available. The EHC recommended that the Agency include both case studies
incorporating human occupational or clinical information, and those with databases considered to
be insufficient for quantitative dose-response estimation ("not-verifiable").  The review was not
intended to be a review of the RfC methods themselves but rather one of the conceptual
framework of the approach as applied to representative data.

       The Committee reviewed the following case studies which were separated into four
categories:

       a)     group 1: particle case studies (diphenylmethane diisocyanate, antimony trioxide,
             and phosphoric acid)

-------
       b)     group 2: category 1 gas case studies (methylmethacrylate, 1,3-
              dichloropropene, acetaldehyde, and chlorine dioxide)

       c)     group 3: category 3 gas case studies (carbon disulfide, n-hexane, hydrogen
              cyanide, vinyl chloride, and 2-ethoxyethanol)

       d)     group 4: not-verifiable case studies (acrylamide, bis(chloromethyl)ether,
              caprolactam, and methoxychlor).

       In its briefing materials provided to the Committee, the Agency explained that there are
differences in the level of documentation amongst the case studies and that the differences
reflected changes made during a pilot program of the Integrated Risk Information System (IRIS)
process.  These changes are described in a paper by Mills and Foureman (in press).  For some of
the case studies, the background documentation describing the derivation of the RfC is embodied
by the IRIS Summary alone. In the newer case studies, the documentation is described in the
Toxicological Review. The complete IRIS file for the compounds that were reviewed is available
at http://www.epa.gov/iris.

2.2 Charge

       The review of the Inhalation Reference (RfC) Methods Case Studies was  focused on the
approach applied to the case studies.  The review was not intended to focus on the RfC
methodology or on the specific chemicals as independent critical reviews. However, in a few
cases, the opinions of the reviewers on methodology  or specific chemicals are included.  The case
studies that were reviewed by the Committee fell into one of four categories: a) particle  case
studies, b) category 1 gas case studies, c) category 3  gas case studies, and d) not-verifiable case
studies. The Committee was charged to provide comments on each of the following aspects of
the RfC Case Studies:

       a)     Overall, are the concepts and applications of the RfC methodology clearly
              articulated in the documentation provided for the case studies? Do the decisions
              and choices in these files attain the Agency's goal of being "transparent, clear, and
              reasonable?  If not, what are specific examples within these files that could be
              instituted to better attain this goal?

       b)     In derivation of the RfC in the specific case studies,

              (1)     are the  study summaries presented in sufficient detail for the reader
                     evaluation?

              (2)     are the  designations of the critical effect and effects levels
                     (NOAEL/LOAEL/BMC) based on rationales that are clear and reasonable?

-------
       (3)     of the studies presented in either the IRIS Summary or Toxicological
              Reviews of each chemical, have the Principal studies been selected in a
              consistent and rational manner? Does this choice reflect consideration on
              the current knowledge of potential human response?

       (4)     have the underlying assumptions of the duration and dosimetry adjustments
              been presented clearly?

       (5)     are the rationales presented for use of uncertainty factors clear, reasonable
              and consistent?

       (6)     do the confidence statements reflect the strengths and limitations (e.g.,
              relevancy to humans, comprehensiveness of the data base) of the RfC
              assessment in a  manner consistent with the Agency's goals?

c)     In the IRIS Summaries  for the specific cases, numerous studies are included under
       the heading "Supporting/Additional Studies" that are meant to provide further
       support for designation of the critical effect (e.g., mechanistic data, human data) or
       for the effect level chosen in the Principal study, or to establish the completeness
       of the data base. Is the  depth of presentation in this section sufficiently
       comprehensive to provide information supportive of the decisions made in the
       assessment (such as uncertainty factors and confidence levels)?

-------
                        3. RESPONSE TO THE CHARGE
3.1 General comments on the RfC Case Studies

       The Environmental Health Committee commends the Agency's efforts to demonstrate the
application of the dosimetric adjustments and to illustrate the methodology applied to chemicals
representative of the typical range of data available.

       The EHC found the concepts and application of the RfC methodology to be clearly
articulated in some of the case studies, but unclear in others.  Similarly, the Committee concurred
with the derivation of the RfC in some case studies and had concerns about the derivation in
others. The  same findings also held true for the IRIS Summaries.

       The Committee identified specific areas in the RfC case studies which should be improved
and makes the recommendations discussed below for each group of chemicals.

3.2 Group  1: Particle Case Studies

  3.2.1 Diphenylmethane Diisocyanate (MDI)

       a)     Is the RfC methodology clearly articulated? - The concepts and application of the
             RfC methodology are clearly put forth in the documentation.

       b)     Derivation of the RfC - The Reuzel et al. (1990; 1994) studies have the best
             characterization of exposure atmospheres in terms of size and concentration that
             can be related to observed effects in the animals.  In particular, Reuzel et al.
             (1994) is the appropriate  critical study for deriving the RfC. While the critical
             endpoint should be the production of asthma, there is no exposure information to
             allow the determination of critical exposure levels or duration for production of
             this  condition.  Thus, the endpoint selected for derivation of RfC (basal  cell
             hyperplasia in the olfactory epithelium) is reasonable and does have good exposure
             data associated with it. However, the lack of this effect in another study at a
             higher concentration is of some concern in attempting to understand the rationale
             for the critical level selected.

             The rationale for the critical effect, namely that protection against nasal  cell
             hyperplasia will likely protect against other effects, may be generally reasonable,
             unless of course the concentration needed to elicit this effect is much greater than
             that for responses in other areas of the lungs. The available database is  not clear in
             this  regard. Furthermore, it is possible that effects in the rat upper respiratory tract
             may occur to a greater extent than that in  humans, due to differences in structure
             between the two species.  Furthermore, for the same exposure concentration, there

-------
            could be greater effects in the lower respiratory tract in humans than in rats,
            depending upon inhaled particle size, due to lower efficiency of the human upper
            respiratory tract to collect particles. In this regard, there is concern with the
            statement that effects in the upper respiratory tract (ET) are reflective of the major
            health concern related to MDI exposure, which is asthma. This is not at all well
            justified in the toxicological review and this position could be a problem due to
            dosimetric differences between rat and human at the actual human exposure
            relevant size of 10 /
-------
of the dosimetry adjustments could be improved.  For example, the meaning of the
acronym, RDDR(TH), was unknown and the acronym was never defined in the
document.  The term, HEC, "human equivalent concentration" was understood
but the calculation of the HEC was unclear. The documentation for the calculation
of the dosimetry adjustment could also be improved. These calculations appear in
the section of the document with the subheadings "Scenario" and "Conversion
factors."  For example, it might be more logical to present the BMC(ADJ)
conversion first since it is not apparent in the "Scenario" how this value is
obtained. Additionally, it appears that the adjustment is to convert 6 hrs/day,  5
days/week into a continuous exposure (24 hr/day, 7 days/week).  However, the
conversion indicated that the hours and days are added/subtracted to the BMC
(e.g., +/- 6 hrs/24 hrs) instead of the BMC being multiplied by the 6/24 ratio.  A
brief narrative, beyond that presented, might also be useful for individuals who are
not familiar with these conversions to explain how the value of 0.46 (the
RDDR(TH)) was obtained.

Although the use of the Benchmark Dose approach is highly encouraged, it would
be useful for the narrative to contain a brief statement regarding the merits of both
approaches (NOAEL and Benchmark Dose) and the rationale for the choice of the
latter in this particular case. The calculation of the BMC also was unclear. The
first sentence under the section on the BMC states that the incidence of chronic
inflammation, etc. in the animals observed during the 1-year period were analyzed
for the BMC. Is this the first or the second year of the study? The subtleties  of
the difference between relative versus additional risk models may be known to
those in the field, but not to those outside the field.  For the calculations to be
transparent to a wider community than those who routinely do risk assessments,
EPA will have to explain the risk models more fully.  The Agency should define all
of the acronyms when they are first used in the document and the calculations
should be explained so that the antimony trioxide document stands on its own
without frequent referral to the RfC document.

The first complete sentence at the top of page 5 of the review document suggests
that standard operating procedures for calculating BMCs were not followed.  The
Agency should provide an explanation for not following the BMC standard
operating procedures. How does the EPA know that "the best curve fits and  the
lowest corresponding lower 95% confidence levels" were those obtained for
chronic inflammation in female rats if the Agency did not examine all of the
endpoints?  The EPA should revise this statement to clarify that this was the best
curve fit and lowest 95% confidence value for those endpoints for which the BMC
was calculated.

It was not clear as to how the 10% incidence level for calculating the BMC was
chosen. There are three sequential sentences in the top paragraph on page 5 of the

-------
       review document (These sentences appear on lines 11-16) describing the choice of
       the 10% incidence level for the BMC.  The subject text refers to studies by
       Faustman et al. (1994), Allen et al. (1993) and Allen and Chapman (1993), but
       does not state clearly which studies support the choice of the 10% incidence level.
       There is also a comment about the probability of response among a "subset," but
       this subset is not identified. The Committee suggests that this section of the
       report's text be revised to identify clearly the basis for the choice of the 10%
       incidence level.

b)     Derivation of the RfC - Individual reviewers within the Committee did not agree
       fully as to the level of confidence on the RfC for antimony trioxide.  Issues which
       concerned the whole Committee included the short duration of the principal study,
       and the lack of reproductive and developmental studies. As noted below, both of
       these issues generated uncertainty factors that did not appear to be well justified.
       Consequently, the EHC would like to see more details regarding the rationale for
       the choice of the uncertainty factors.

       The designation of the critical effect and the effects level were clear for the
       NOAEL and the LOAEL but the effect level for the BMC were merely stated (line
       16, page 5 of the review draft) making it difficult to evaluate the accuracy or
       reasonableness of the value.  The choice of the critical study seemed reasonable.
       The assumption of the duration and dosimetry adjustments were understood
       except for the calculation of HEC, which was not described. The rationales for the
       uncertainty factors were clearly  stated.

       It was noted that the choice of a ten-fold intraspecies extrapolation factor to be
       supportable because it is a "standard" default factor; the use, however, of a three-
       fold factor for an incomplete data base was not well justified.  It was unclear as to
       whether or not this uncertainly factor was an arbitrary default when
       reproductive/developmental studies have not been conducted regardless of the
       existence of data to indicate that this system may not be the most sensitive. For
       example, antimony trioxide appears to be a pulmonary toxicant as determined from
       the principal study.  The additional studies support this observation.  Toxicokinetic
       studies indicate long term retention in the lungs.  The various toxicity studies
       conducted have not determined  the reproductive organs to be target organs for
       absorbed antimony.  The one animal study conducted determined no teratogenic
       effects. The only evidence for potential reproductive effects appears to be a
       questionable human study.

       Likewise, the uncertainty factor  for study duration is  also not well justified.  The
       statement that steady state  concentrations of antimony were not reached in the
       lung conflicts with information presented in the narrative which states that the
       clearance half time for the  low dose in the Newton et al. (1994) study was 2.3
                                     10

-------
            months. With this half time, antimony levels in the lungs at the low dose should
            have been close to steady state.

            The Committee agreed with the case study narrative finding that the ocular effects
            in the Newton study were inappropriate for consideration, whereas the pulmonary
            inflammation was the appropriate response for the application of benchmark dose
            methodology (Newton et a/., 1994). The EHC also found that sufficient detail was
            presented with regard to the experimental protocol and major findings of the study
            selected for determination of the RfC (e.g., Newton et a/., 1994). The narrative
            related to the decision not to choose an endpoint (ocular toxicity, for example)
            along with the narrative related to the endpoint of choice for human extrapolation
            (pulmonary toxicity, chronic interstitial inflammation) was found to be useful. In
            conclusion, the presentation on both of these extremes was found to be useful and
            helped to convince the Discussant of the Agency's derivation of the RfC.

     c)     Presentation of Additional Studies - The Committee noted no problems with the
            depth of presentation of the additional studies. The supporting/additional studies
            provide a useful and valuable perspective for the determination  of the RfC.  Even
            though these studies cannot be used to determine the RfC, they do provide
            consistency regarding the toxic effects and mechanistic insights, particularly in the
            dosimetry area, that serve to support the choice of study and endpoint for effect.
            The Agency is encouraged to continue to include supporting studies, (particularly
            the inclusion of human studies) in the case studies.

            There are some editorial comments.  On page 10, line 16, "smelters" should be
            changes to "smelter workers." Also, care should be taken to describe the particle
            size in each exposure.  This was not done consistently.  On page 8, the term
            "Ferret's diameter" was unfamiliar.  Is there a more common term?

3.2.3  Phosphoric Acid

     a)     Is the RfC methodology clearly articulated? - The study summaries are given in
            sufficient detail.  In addition, the rationale for the designations of the critical effect
            and the effects levels (NOAEL etc.) are presented clearly.  However, the case
            study, as written, was somewhat difficult to read and, therefore, was not
            transparent.  Who is the audience for these reports?  Even assuming some degree
            of toxicology background, some clarification is still in order.  The Committee
            recommends that the abbreviations be defined when they are first used,
            calculations be explained, and where applicable, the importance of calculations
            should be noted.  For example,

                   (1)    Abbreviations need to be defined  (e.g. RDDR, MMAD, and
                         BMC 10 (HEC)).
                                          11

-------
       (2)    It is not clear where some numbers used in the calculations come from,
              example, in the scenario, the BMC10 (HEC) = 5.4 mg/m3 X 0.64.

       (3)    Some explanation of the importance of certain calculations would be
              clarifying. For example, there is discussion on the nonaerodynamic
              diameter of the aerosol particles at the lesion site. How is this related or
              important to the calculation of the BMC 10?

       (4)    The underlying assumptions of the duration and dosimetry adjustments
              need further clarification in the calculation of an RDDR along with an
              explanation on the significance of the adjustments.

       Also, the rationale was unclear on how the Agency applied the information on the
       hygroscopic growth of the acid aerosols to estimate the diameter of the particles
       (based upon initial mass median aerodynamic diameter (MMAD) of the aerosol) at
       the proposed site of the critical lesion, which is the tracheobronchial tree.

       Phosphoric acid is being compared in a number of instances with another
       particulate acid for which there is more data, namely sulfuric acid. In the case of
       the latter, high concentration exposures are associated with alveolitis, bronchial
       and/or bronchiolar edema and epithelial desquamation. There is little evidence for
       bronchiolar fibrosis. Also, the sensitivity of the morphological endpoints is
       dependent upon the animal species, with the rat the least sensitive to sulfuric  acid.
       Thus, the critical dose for rat may be much higher than that for other species,
       including humans; the latter may respond with similar lesions at lower
       concentrations. The EPA should note for clarification purposes that the maximum
       uncertainly factor for both dosimetry  response is 3.  Together these factors
       comprise the interspecies uncertainty  factor of 10.

b)     Derivation of the RfC - The study summaries are given in sufficient detail.  The
       rationale for the designations of the critical effect and effects levels (NOAEL, etc.)
       are presented clearly.  Given those studies described, the principal study appears
       appropriately chosen.  However, the choice of the appropriate  study, and indeed
       the studies which are presented, lack any consideration of the current knowledge
       of the potential human response in that no discussion of developmental effects is
       given.  There is potential that the most sensitive endpoint (i.e.,  a developmental
       endpoint) has not been studied. Given that the newborn lung is primarily made up
       of terminal bronchioles with alveolarization only beginning, and given that children
       breathe more air per unit body weight than adults, the potential for lung
       developmental effects is great. This issue should be addressed under uncertainty
       factors, where little explanation is given. There was a concern about the adequacy
       of the uncertainty factor of 10. The Agency should clarify who are considered
                                     12

-------
              sensitive human subpopulations, if children are considered a sensitive human
              subpopulation, and if children with chronic lung disease such as
              bronchopulmonary dysplasia, or children with asthma, are considered sensitive
              human subpopulations.

3.3 Group 2:  Category 1 Gas Case Studies

  3.3.1 Methymethacrylate

       a)      Is the RfC methodology clearly articulated? -  This document was described as
              "by far, the clearest in its presentation of the methodology" during the Committee's
              discussions. It could serve as a model for the other compounds.  However, there
              were still some areas that could be made clearer.  On page 7, the abbreviations
              used in the footnote to the table  should be defined. At the bottom of page 7, the
              exposure concentrations should be rounded to 100 and 400 ppm as they are on the
              next page as opposed to the numbers cited in the review document, 99.79 and
              396.07.  The same holds for the  calculated values for the concentrations in mg/m3.
              On page 8, line 11, "effect" should be "affect." The chemical structure for
              methylmethacrylate should be added to the document.

              It is confusing that the critical study is described as having 70 rats of each sex per
              group but the observations described at the bottom of page 7 refer to incidences
              per 10 animals or even in one case in only 4 animals. This inconsistency should be
              corrected or explained. On page 9,  the data for degeneration  of the olfactory
              epithelium in males is described for groups ranging in number from 38 to 48.
              What happened to the rest of the animals? On page 9, in the middle paragraph,  on
              the 4th from the bottom line: the EPA should explain what is meant by an
              "environmentally protective RfC?"  The RfC is meant to protect human health, not
              the environment.  On the top of page 10, the Agency should define the
              abbreviations here. Otherwise, the paragraph in which this appears is quite well
              written and close to being the transparent report that the EPA desires.

              Concepts and application of RfC methodology are explained in detail in the
              documentation in a sort of boiler-plate manner. However, when it comes to
              establishing the RfC, there are numerous data gaps and places where there are
              more than enough data but no clear indication of how the  data were used to
              establish the value. Documentation needs to be more reader-friendly. Some
              figures and tables that summarize the data and make comparisons  are
              recommended.  To meet its goal of improving the utility and credibility of the IRIS
              process, these documents need to become more understandable to all levels of
              users of these documents. The methylmethacrylate document was clearly one of
              the best of the IRIS documents included in the Committee's background material
              but even it needs editing. As recommended for other RfC documents, the
                                           13

-------
            methylmethacrylate Document should use full description with first encounter
            (e.g., adjusted for ADJ, and human exposure concentration for HEC).

     b)     Question 2; Derivation of the RfC - The study summaries were in sufficient detail.
            The choice of critical effect and effect level seemed reasonable. The choice took
            into account current knowledge of human responses.  The underlying assumption
            of the duration and dosimetry adjustments were presented quite clearly in this
            document - much more so than for acetylaldehyde and hydrogen cyanide. The
            considerations for the choice of uncertainty factors were clearly presented and
            appeared to be carefully thought through.  The statement on confidence was clear
            and logical.

            The data needed for the recommendation is included and can be "teased out"  by
            the diligent reader. Some summary tables and figures would be helpful.  The
            crucial study is the Hazelton inhalation rat study with effects at 400 ppm,
            histopathological changes at 100 ppm and NOAEL at 25 ppm.  The uncertainty
            factor of 10 (intraspecies) and modifying factor (MF) of 1 seem appropriate.

     c)     Presentation of Additional Studies - The depth of the presentation of the additional
            studies was adequate except for the lack of references for the second paragraph
            under "Supporting Studies." Supporting studies do not add a great deal.  It would
            be helpful to have specific sensitization data in humans and some methodology to
            quantitate an exposure-response for methylmethacrylate.

3.3.2  1,3-Dichloropropene

     a)     Is the RfC methodology clearly articulated? - The case study for 1,3-
            dichloropropene is transparent  and clear.

     b)     Derivation of the RfC - The Agency should explain, in the document, why it did
            not use the benchmark dose.

     c)     Presentation of Additional Studies - The depth of the presentation in this document
            is sufficiently  comprehensive, providing information supportive of the decisions
            made in  the assessment.  However, a statement on children, and whether the RfC is
            protective of children, should be added.

3.3.3  Acetaldehyde

     a)     Is the RfC methodology clearly articulated? - The acetaldehyde document is more
            typical of IRIS documents and  when compared with the Agency for Toxic
            Substances and Disease Registry (ATSDR) documents or the occupational
            exposure limit documentation, is much less readable.  In general, the Discussants
                                         14

-------
            found the methodology to be well articulated. However, the methodology could
            be made clearer in several ways. The acronyms used at the top of page 2 should
            be defined. The method for calculating the HEC should be described. On page 1,
            the term, critical dose, would be more appropriately called, critical exposure
            concentration.  A value that has the units "mg/m3" should not be called a dose.
            Why is the critical effect listed twice (as 1  and 2)?  Perhaps it refers to the two
            critical studies.  However, it is confusing to list the same effect twice.  The critical
            effect was degeneration of olfactory epithelium, period.

     b)     Derivation of the RfC - The study summaries were presented in sufficient detail.
            The designation of critical effects were clear and reasonable.  The Discussants
            agree with the choice of a critical effect. The duration  and dosimetry adjustments
            were clear for the most part, but the method  of conversion to an HEC was not
            described. The description of both the uncertainty factors and the confidence
            statements were clear and reasonable.

            The use of the Appleman studies to define critical effects is adequate but there is a
            lot of human data that could be used (Appleman et a/.,  1982; Appleman and
            Woutersen, 1986).  The use of human data would avoid the interspecies
            extrapolation factor of 10.  If the EPA uses the human  data to derive the RfC for
            acetaldehyde, the factor of 10 between subchronic and  chronic may also vanish.

     c)     Presentation of Additional Studies - The presentation of the additional studies was
            in sufficient depth for the evaluation of the choice of the critical study.  The
            Committee agree with the low confidence value.  For a compound  of this
            importance (since it is an IARC carcinogen) the IRIS document should be
            improved.

            No discussion of the common genetic polymorphism of acetaldehyde
            dehydrogenase is presented. The inability  or relative inability to metabolize
            acetadeldehyde is common amongst Asian populations. Are there  any data that
            suggest that such individuals are at increased risk of an adverse effect to
            acetaldehyde gas exposure? The Agency should address this issue.

3.3.4  Chlorine dioxide

     a)     Is the RfC methodology clearly articulated? - The RfC  methodology was found to
            be clearly articulated. However, the definition of thoracic respiratory effects
            should be included in the document.

     b)     Derivation of the RfC - The Committee was unclear about the significance of
            deriving an RfC when there was clearly zero confidence. During the public
                                          15

-------
              meeting, the Agency responded that it was required by the Clean Air Act to set an
              exposure limit for chlorine dioxide.

       c)      Presentation of Additional Studies - The chlorine dioxide presentation provided in
              the Agency's document is sufficiently comprehensive to provide information
              supportive of the decision made in the assessment.

3.4 Group 3: Category 3 Gas Case Studies

  3.4.1 Carbon disulfide

       a)      Is the RfC methodology clearly articulated? - This case study was very clearly
              written. However, the units were not consistent.  This should be corrected.

       b)      Derivation of the RfC - The methods used to calculate the BMC is not consistent
              with methods described for continuous data sets in other EPA documents.  EPA
              has recommended either a "hybrid approach" or conversion of continuous data to
              quantal data. EPA may want to revisit this and compare its methods for
              calculating the BMC with other guidance the EPA has put forward for continuous
              data.

              It will be helpful to report the magnitude of change in the maximum motor
              conduction velocity (MCV) and to clearly state that the original authors (and not
              just a peer reviewer) questioned the biological significance of the findings in the
              study. The authors stated that "all reductions are within range of clinical normal
              values."

              An additional 10-fold uncertainty  factor was added because of effects seen in a
              developmental neurotoxicity study.  In the n-hexane RfC, (described in  section
              3.4.2), a 10-fold uncertainty factor was added because there were no data at all.
              Yet in this carbon disulfide case study, an experiment was conducted with more
              sensitive endpoints at different stages of development.  The EPA added a 10-fold
              factor because minimal effects were noted. If the developmental  effects are truly a
              concern, then the RfC should be based on these effects, rather than adding a 10-
              fold uncertainly factor to a different endpoint. It appears that the Agency adds the
              same uncertainty factor whether or not additional data are available. This practise
              will discourage registrants from better understanding mechanisms and evaluating
              more discriminating endpoints.

              On the other hand, another Committee Member concluded that the direct acting
              mechanism of carbon disulfide toxicity should not lessen an uncertainty factor for
              susceptible populations. The rationale was that a susceptible populations does not
              just mean metabolic polymorphisms and that an axon which is growing  and then
                                           16

-------
            cross-linked has more potential for an adverse outcome(s) than one which is static.
            The EPA should define the term "sensitive subpopulations" and use it consistently
            throughout the documents. It was also concluded that, in the Tabacova studies, an
            increase in narrow-path crossing slips and falls is reported at 10 mg/m3, although
            not at higher concentrations, and that this may be enough evidence to require an
            uncertainty factor for developmental effects of at least three if not ten.

     c)     Presentation of Additional Studies - As explained above, the depth of the
            presentation for carbon disulfide is not sufficiently comprehensive to provide
            information that is supportive of the uncertainty factors or the confidence  limits.

3.4.2  n-Hexane

     a)     Is the RfC methodology clearly articulated? - The concepts and applications of the
            RfC methodology are clearly articulated in the documentation. However, it was
            not transparent to the reviewers that the RfC is based on  a study in which the
            effects are of questionable biological significance.  The change in MCV was less
            than a 5% change.  The authors of the study state that there were "no apparent
            abnormalities in any one person" (Sanagi and Seki, 1980). Thus, on an individual
            basis in the present study, no objective signs indicating damage to the nervous
            system that could have been due to the n-hexane exposure were detected."  The
            authors also state that "it cannot be determined that even on a group basis, toxic
            effects on the peripheral nervous system were demonstrated at the given exposure
            level." Finally, they state that the "number of subjects examined and the
            differences noted between the groups were really too small to come to any decisive
            conclusion." These conclusions from the original authors need to be included in
            the documentation so it can be clearly understood that the RfC is based on very
            questionable changes.

            The EPA's decision to use these "effects" as a LOAEL from which to set the RfC
            is not completely consistent with the Agency's approach to setting the RfC for
            carbon disulfide. In the carbon disulfide example (addressed in section 3.4.1), the
            benchmark dose approach was used to extrapolate above the dose range to a 10%
            change because the EPA concluded that the minimal effects observed in the study
            were not considered biologically meaningful. If the Agency chooses to use less
            than a 5% change in MCV for risk assessment purposes because these findings are
            consistent with other changes seen at higher doses, then it is important for the
            Agency  to state this and make transparent the uncertainties surrounding the effect
            observed in Sanagi and Seki (1980).

     b)     Derivation of the RfC - The 3-fold uncertainty factor was added for lack of data on
            reproductive and chronic respiratory effects. There have been several studies
            submitted to the Agency on commercial hexane (55% n-hexane) including
                                          17

-------
            developmental studies in two species, a 2-generation reproductive study, and a
            chronic study (Ladefoged and Perbellin, 1986; Ladefoged etal., 1989; Ladefoged
            et a/., 1994; and Patten et a/., 1986). The Agency should review these studies and
            consider removing the 3-fold uncertainty factor since the database is quite rich.

            A 10-fold uncertainty factor was added for the use of a LOAEL rather than a
            NOAEL. The full 10-fold factor should be reduced to take into account the fact
            that the "LOAEL" was actually based on effects that are of questionable biological
            significance which were considered to be a "NOAEL" in the carbon disulfide case.
            In addition, these effects, if real, could be a result of exposure to both acetone and
            hexane.  This leaves an uncertainty factor of 10 for sensitive populations, which is
            appropriate and should not be reduced.

            The definition of teratogen would be helpful since sometimes it is defined to mean
            functional defects  (which were not examined in this study).  It is not clear from the
            discussion whether the brain or spinal cord (for pyramidal tracts or ascending
            tracts) were examined. It is not clear whether any developmental neurotoxicity
            testing was done.  The Agency should clarify all of these issues in the n-hexane
            RfC document.

     c)     Presentation of Additional Studies - The n-hexane RfC document did not address
            the significant acetone exposure. There is evidence to suggest that acetone might
            potentiate the effects of n-hexane  (Ladefoged and Perbellin, 1986; Ladefoget et
            a/., 1989, 1994; and Patten etal., 1986).  Thus, if one makes the conservative
            assumption that the statistically significant effects in the Sanagi and Seki (1980)
            study are biologically meaningful, these effects may be a result of exposure to both
            acetone and hexane.  Based on this evidence, the minimal changes in MCV would
            likely be observed at levels above 58 ppm for n-hexane alone.

3.4.3  Hydrogen cyanide

     a)     Is the RfC methodology clearly articulated? - The Committee was disappointed
            with the hydrogen cyanide RfC document in terms of the uncertainty factors and
            the focus on rat data rather than human studies.  The ATSDR document on
            cyanide has a more extensive human database and uses the trans-species
            information in a more rational manner.

     b)     Derivation of the RfC - The dose response relationship for cyanide is steeper than
            for almost any other chemical. It is known that hydrogen cyanide acts almost
            instantaneously by producing cyanohemoglobin, and that all species react similarly.
            Thus, an uncertainty factor of 1000 is not considered to be justified.
                                          18

-------
     c)     Presentation of Additional Studies - The Agency should revise the hydrogen
            cyanide document to focus on the human studies.

3.4.4  Vinyl chloride

     a)     Is the RfC methodology clearly articulated? - In contrast to the document on
            carbon disulfide, this document is clearly presented and defended. The IRIS
            summary is a most valuable support to the toxicology summary.  The IRIS
            Summary and Toxicological Profiles are more up to date than most IRIS
            documents and generally do a better job than most such documents.

            The RfC of 0.1 mg/m3 is clearly defined as limited to non-cancer effects
            applications. Since the quantitative estimate of cancer risk is 0.000001 per
            microgram/cu.m., it appears that a cancer risk of 0.0001 is present at the RfC. If
            this is the case, the Agency should make this explicit in the RfC  document, since
            most readers of the document will assume that the RfC is a "no adverse effects"
            level

     b)     Derivation of the RfC - Vinyl chloride clearly presents health problems. It
            produces angiosarcoma in both rats and workers. However, since the
            Occupational Safety and Health Administration (OSHA) reduced the Permissible
            Exposure Limit (PEL) to 1 ppm several decades ago, there has  not been a single
            new case of angiosarcoma in the registry at Louisville. This clearly indicates that
            there is at least a pragmatic or practical threshold for vinyl chloride angiosarcoma
            in man and the RfC methodology is more appropriate for assessing noncancer risks
            than the slope factor approach using a linear multistage model.  Storm and
            Rozman (1997) have compared the dose response data for rats and man to vinyl
            chloride and demonstrated that the linear multistage approach is inappropriate, that
            rats are more sensitive than man and that there is clearly a threshold for
            angiosarcoma in both rats and man. This paper needs to be incorporated into the
            database for vinyl chloride.

            Uncertainty factors seem to be limited to 1, 3 and 10 (although the first two 3's
            (3x3) are equated to 10). Vinyl chloride is given a factor of 3 based on the failure
            of PBPK modeling to address the potential for different tissue sensitivities between
            species. While this sounds reasonable it also appears that other chemicals with the
            same factor of 3 have more serious deficiencies. The Agency should consider
            using a further gradation in uncertainty factors.  Given the significant effort
            required to develop and apply a PBPK model, there does not seem to be a
            practical benefit for the use of these models if no reduction in uncertainly (and
            corresponding uncertainty factors) is obtained.  The confidence level of medium
            for the database also seems more severe than the same confidence level for other
                                         19

-------
              chemicals. This is particularly true based on the stated unlikelihood of
              reproductive effects.

       c)      Presentation of Additional Studies - The EHC found that the use of a two-year
              dietary admix study in rats as the principal study for an inhalation RfC is justified
              based on the qualitative and quantitative dose-response relationship with a one-
              year inhalation study in rats, the accepted reliability of PBPK modeling to calculate
              levels of metabolites per liver volume in rats and humans, and the lower doses of
              the oral study.  One Member was concerned that the inhalation RfC is based upon
              an oral feeding study when inhalation exposure study data are available and  would
              like to see this issue discussed in the document and a justification presented  for the
              use of an ingestion study.  It may be the most appropriate, but this decision  should
              be defended. In the case of vinyl chloride, human data are not only available but
              are adequate to establish causality with exposure-response information.  In such
              cases the rat data become irrelevant or are of value only as corroborative informa-
              tion.  It is informative with vinyl chloride to note the comparison of the rat-human
              data and it helps us define the utility of animal tests as predictors of human effects.

  3.4.5  2-Ethoxyethanol

       a)      Is the RfC methodology clearly articulated? - The RfC methodology was clearly
              articulated with the exception of the inconsistency in the units.  The units of ppm
              or mg/m3 need to be made consistent throughout the document.  In the description
              of the Doe (1984) study, the parameters which assess fetal toxicity  included fetal
              body weight, and external, visceral and skeletal examinations. The  Agency  should
              clarify whether the term "visceral" includes  the brain and whether this study was
              the standard toxicological assessment defined by Wilson.

       b)      Derivation of the RfC - In one study, the NOAEL (HEC) was 47 mg/m3.  The
              EHC was a concerned that this value was lower than the adjusted value for the
              testicular effects. The rationale for choosing the testicular atrophy  as the critical
              endpoint should be explained in this piece.

       c)      Presentation of Additional Studies - The RfC for ethoxyethanol would be more
              convincing if it were based on the human data (Clapp and Smallwood, 1987) rather
              than on the rabbit data.  In setting the RfCs, the Agency's goal is not to protect
              rabbits or rats but to improve human health. The Committee recommends that the
              EPA pay more attention to data from the target species available.

3.5 Group 4: Not-Verifiable Case Studies

       These following four chemicals have less satisfactory databases than most  of the other
case studies; subsequently, the EPA did not set RfCs for these chemicals.
                                           20

-------
  3.5.1 Acrylamide

       Although there is some useful information in this document, there was agreement that
there are insufficient data for developing an RfC.  A chronic oral (drinking water) study in rats has
been used to estimate carcinogenic potency for both the oral and inhalation routes.  The Agency
should explore the possibility of using the noncancer findings in that study to estimate an RfC.
The Agency should also consider giving more prominence to the dermal route for this chemical
and other chemicals where there is insufficient inhalation data. In the acrylamide case, it was
shown that peripheral neuropathy has been produced with exposure by the dermal route.

  3.5.2 Bis(chloromethyl)ether (BCME)

       In  general, the Discussants also agreed that the information on bis(chloromethyl)ether was
insufficient to develop an RfC. However, a six-month inhalation study to BCME vapors showed
no noncancer effects at the highest dose tested.  It may be possible to use this to estimate an RfC.
While the  RfC is clearly designed for noncancer endpoints, the RfC should reference
carcinogenicity when cancer is the crucial endpoint as it is for BCME.  There is little justification
for an RfC based on available data.

  3.5.3 Caprolactam

       The Committee agreed that an RfC should not be derived at this time, given the paucity of
data on caprolactam. Although an oral RfD from animal studies and some worker information is
available, very limited human inhalation data are available. Several workplace studies with a small
number of employees are presented; however, no meaningful exposure information is  available. It
should be  also noted in the caprolactam document that an oral RfD has been derived based on
reduced body weight.

       In  order to make the report more understandable to the reader, several clarifications are
required in the caprolactam RfC document:

       a)      On page 6 of 15 (paragraphs #3 and #4), does the reference to intubation with
              Caprolactam or water in these two paragraphs refer to oral or inhalation  studies?
              Although at first glance these would appear to be gavage studies, it is not clear
              from the remainder of the paragraphs. The EPA should clarify this.

       b)      On page 7 of 15, line  4: how do you have menstrual disturbances in pregnant
              women? This statement also needs clarification.
                                           21

-------
  3.5.4  Methoxychlor

       The Committee concurs that the database is inadequate to derive an RfC for methoxychlor
at this time.  There is only one case report of an aplastic anemia after using a tomato dust
pesticide containing methoxychlor.

3.6 General Findings from RfC Case Studies

       It was difficult to read the narrative on an RfC case study document and compare all the
results from all the studies presented for a given chemical.  It is suggested that the EPA consider
using tabular or pictorial presentation of data similar to ATSDR in addition to using narrative.

       In discussing critical effects that are statistically significant, the magnitude and hence
severity of the effect should be reported.  It was the absence of this information in the n-hexane
review that led one of the reviewers to check the original paper and discover that the effect was
less than a 5% effect that would not typically be regarded as biologically significant. Inclusion of
this type of data will make the risk assessment process more transparent and improve the scientific
rigor of the discussion.

       The Agency should reconsider how uncertainty factors are added when the critical effects
have variable levels of concern. The carbon disulfide and n-hexane case studies illustrate this
problem in two different ways:

       a)     A 10-fold uncertainty factor was added for n-hexane because there were no
              developmental toxicity data.  Yet a 10-fold safety factor was added to carbon
              disulfide which had developmental toxicity data using functional endpoints because
              slight equivocal effects were noted.  This is not a consistent application of
              uncertainty factors.

       b)     A 10-fold uncertainty factor was added to n-hexane because the critical effect was
              a LOAEL.  Although this effect was statistically significant it was not regarded by
              the authors as biologically  meaningful. All values were within normal clinical
              range. If the critical effect is an equivocal effect then it is difficult to justify the full
              10-fold safety factor to extrapolate from LOAEL and NOAEL. In fact, one might
              challenge the need for an additional safety factor at all.

       In conclusion, there should be less uncertainty if experiments were conducted to  elucidate
mechanisms and as more sensitive endpoints are used that refine the NOAEL.  This should be
considered in the selection of uncertainty  factors.
                                            22

-------
                   4.  SUMMARY OF RECOMMENDATIONS
       The Committee had the following thoughts and recommendations for EPA to consider in
revising the document and their approaches to RfC calculations:

       a)     In general, it was difficult to read the narrative on an RfC case study document and
             compare all the results from all the studies presented for a given chemical. To
             improve the clarity of the documents, the Agency should summarize some of the
             data using figures and tables.

       b)     The Agency should include more recent studies to derive the RfCs which were
             based on studies that were done in the past and may be outdated..

       c)     In some cases, the Agency did not include  available human data when developing
             the RfC. The EPA should pay more attention to the human data, when available.

       d)     In developing the RfCs, the Agency should expand their review of the literature to
             include new models, both laboratory and mathematical.

       e)     Each document should contain a section on children and their unique exposures
             and vulnerabilities, and delineate data gaps. Each document should also include a
             statement on whether the RfC is protective of children

       f)     The definition of susceptible human populations should  be consistent throughout
             all of the RfC documents. The EPA should also include a discussion on how it
             deals with susceptible subpopulations such as children with asthma or children with
             cystic fibrosis.

       g)     All available data reviewed for a given document should be listed and the reasons
             for inclusion or exclusion from the derivation of the RfC should be provided in an
             appendix.

       h)     Several of the RfC case studies used acronyms and scientific terminology that were
             not defined in the document. Each RfC document should stand on its own and
             should not require frequent referral to the RfC background documents.  Therefore,
             all of the documents should have a glossary, explaining the scientific terminology.
             In addition, the acronyms should be accompanied by their full names when they are
             first used in the RfC documents.

       i)     Some of the calculations were unclear. An explanation of the calculations should
             be included in some of the case studies and improved, in other case studies.
                                           23

-------
j)      The units of measure should be consistent throughout the RfC documents. If two
       units are used, one should be followed in parenthesis by the other. For example,
       408 ppm (68 mg/m3).

k)     Each RfC case study document should include an illustration of the respective
       chemical structure.

1)      The Agency should reassess the application of uncertainty factors in the
       development of the RfC. It appears that the Agency uses the same uncertainty
       factor whether or not there are data on other, possibly significant (but less frank)
       endpoints and when there is marginal dosimetry information or a fully developed
       PBPK model.  This practice will tend to discourage additional research.
                                    24

-------
            APPENDIX A - ACRONYMS AND ABBREVIATIONS
ACGffl
ADJ
ATSDR
BCME
BMC
BMD
CAAA
CFD-PBPK
EHC
ET
HEC
IARC
IRIS
LOAEL
MMAD
MAK
MCV
MDI
MEK
MF
mg/m3
MLE
Mva
Mvh
NCEA
NOAEL
ORD
OSHA
PBPK
PEL
ppm
RDDR
REL
RfC
RfD
RGDR
Sa
SAB
Sh
TLV
American Conference of Governmental Industrial Hygienists
Adjusted
Agency for Toxic Substances and Disease Registry
Bis(chloromethyl)ether
Benchmark Concentration
Benchmark Dose
Clean Air Act Amendments
Computational Fluid Dynamics-Physiologically-Based Pharmacokinetic
Environmental Health Committee
Upper Respiratory Tract
Health Equivalent Concentration
International Agency for Research on Cancer
Integrated Risk Information System
Lowest-Observed-Adverse Effect Level
mass median  aerodynamic diameter
German Maximum Allowable Concentration
maximum motor conduction velocity
Diphenylmethane Diisocyanate
Methyl Ethyl Ketone
Modifying Factor
milligrams per cubic meter
Central Estimate
minute volume for the laboratory animal
minute volume for humans
National Center for Environmental Assessment
No-Observed-Adverse Effect Level
Office of Research and Development
Occupational Safety  and Health Administration
Physiologically-Based-Pharmacokinetic
Permissible Exposure Limit
parts per million
Regional Deposition Dose Ratio
Recommended Exposure Limit
Inhalation Reference Concentration
Inhalation Reference Dose
regional gas deposition ratio
surface area for the laboratory animal
Science Advisory Board
surface area for humans
Threshold Limit Value
                                        A-l

-------
                                   REFERENCES
Allen, B. C., and A. Chapman.  1993.  Statistical analysis of lung and cataract responses in rats
       exposed to antimony trioxide (Unpublished material prepared for J.S. Gift, Environmental
       Criteria and Assessment Office, Office of Research and Development, U.S. EPA)

Allen, B. C., Kavlock, R. J., Kimmel, C. A. and E. M. Faustman.  1993.  Dose-response
       assessment for developmental toxicity. II. Comparison of generic benchmark dose
       estimates with no observed adverse effect levels.  Fund. Appl. Toxicol.., Vol. 23, No. 4,
       pp. 487-495.

Appleman, L. M., Woutersen, R. A., and V. M. Feron.  1982.  Inhalation toxicity of acetaldehyde
       in rats. I. Acute and subacute studies.  Toxicology., Vol. 23, pp. 293-307.

Appleman, L. M. and R. A. Woutersen. 1986. Effect of variable versus  fixed exposure levels on
       the toxicity of acetaldehyde in rats. J. Appl. Toxicol, Vol. 6, No. 5 pp. 331-336

Clapp, D. E. and A. W. Smallwood. 1987. Workplace assessment of exposure to 2-
       ethoxyethanol, App. Ind. Hyg., Vol. 2, No. 5, pp. 183-187.

Doe, J. E. 1984. Ethylene glycol monoethyl  ether and ethylene glycol monoethyl ether acetate
       teratology studies. Environ Health Perspect.  Vol. 57, pp. 33-41.

EPA. 1990. Interim Methods for Development of Inhalation Reference  Concentrations, August
       1990 (EPA/600/8-90/066).

EPA. 1994. Methods for Derivation of Inhalation Reference Concentrations and Application of
       Inhalation Dosimetry, October 1994 (EPA/600/8-90/066F)

Faustman, E. M., Allen, B. C., Kavlock, R. J., and C.A. Kimmel.  1994.  Dose-response
       assessment for developmental toxicity. I. Characterization of database and determination
       of no observed effect levels. Fund. Appl. Toxicol.  Vol. 23, No. 4, pp. 478-486.

Jarabek, A.  M.  1994. Inhalation RfC methodology: Dosimetric adjustments and dose-response
       estimation of noncancer toxicity in the upper respiratory tract. Inhal. Toxicol. Vol. 6:
       301-325.

Jarabek, A.  M.  1995a. The application of dosimetry  models to identify key processes an
       parameters for default dose-response assessment approaches. Toxicl. Lett.  Vol.79: 171-
       184.
                                          R-l

-------
Jarabek, A. M.  1995b.  Interspecies extrapolation based on mechanistic determinants of chemical
       disposition. Human andEcol. Risk Assessment.  Vol. 1, No. 5, pp. 641-662.

Ladefoged, O. and L. Perbellin. 1986. Acetone-induced changes in the toxicokinetics of 2,5-
       hexanedione in rabbits. ScandJ Work Environ Health.  Vol. 12: 627-629

Ladefoged, O., Roswall, K., and J-J. Larsen.  1994.  Acetone potentiation and influence on the
       reversibility of 2,5-hexanedione-induced neurotoxicity studied with behavioral and
       morphometrick methods in rats. Pharmacol Toxicol. 74: 294-299.

Ladefoged, O., Haas, U. and  L. Simonsen. 1989. Neurophysiological and behavioral effects of
       combined exposure to 2,5-hexanedione and acetone or ethanol in rats. Pharmacol
       Toxicol. Vol. 65: 372-375.

Mills, A. and G.L. Foureman. In press. U.S. EPA's IRIS pilot program: establishing IRIS as a
       centralized, peer-reviewed data base with agency consensus. Toxicology (in press)

Newton, P. E., Bolte, H. F., and I. W. Daly. 1994. Subchronic and chronic inhalation toxicity of
       antimony trioxide in the rat. Fund. AndAppl. Tox. Vol. 22, pp. 561-576.

Patten C. J., Ning, S. M., and A. Y. H. Lu. 1986. Acetone-inducible cytochrome P-450:
       Purification, catalytic activity, and interaction with cytochrome b5.  Arch Biochem
       Biophys. Vol. 251, pp. 629-638.

Reuzel, P. G. J., and J. H. E Arts.  1990.  Chronic toxicity/carcinogenicity inhalation study of
       polymeric methylene diphenyl diisocyanate aerosol in rats, Prepared by the Civo Institute
       for the International Isocyanate Institute. Report No. V, Volume 88, page 122, 1990.

Reuzel, P. G. J.  and J. H. E. Arts.  1994.  Chronic inhalation and carcinogenicity study of
       respirable polymeric methylene diphenyl diisocyanate (polymeric MDI) aerosol in rats.
       Fundam Appl Toxicol. Vol. 22, pp. 19 5 -210.

Sanagi, S., and Y. Seki. 1980. Peripheral nervous system functions of workers exposed to n-
       hexane at a low level. Int Arch Occup Environ Health.  Vol. 47, pp. 69-79.

SAB. 1990.  Science Advisory Board's comments on the use of uncertainty an modifying factors
       in establishing Reference Dose level, January 17, 1990 (EPA-SAB-EHC-90-005),
       USEPA, Science Advisory Board, Washington, DC.

SAB. 1991.  Science Advisory Board review of the Office of Research and Development
       document Interim Methods for Development of Inhalation Reference Concentrations,
       April 29, 1991 (EPA-SAB-EHC-91-008), USEPA, Science Advisory Board, Washington,
       DC.
                                          R-2

-------
Storm, J. E. and K. K. Rozman. 1997. Evaluation of Alternative methods for establishing safe
      levels of occupational exposure to vinyl halides. Regulatory Toxicology and
      Pharmacology. Vol. 25, pp. 240-255.
                                          R-3

-------
                              DISTRIBUTION LIST

Administrator
Deputy Administrator
Assistant Administrators
EPA Regional Administrators
Director, National Center for Environmental Assessment
Director, Office of Science Policy, ORD
EPA Laboratory Directors
EPA Headquarters Library
EPA Regional Libraries
EPA Laboratory Libraries
Library of Congress
National Technical Information Service
Congressional Research Service

-------