United States ; v r: }
Environmental, Protectiofj
Agency ; i
j Off ice of Research and
pevelopment •
Washington DC 20460
EPA/630/R-00/003
December 2000
Summary Report for the
Workshop on Issues
Associated with Dermal
Exposure and Uptake
RISK ASSESSMENT FORUM
-------
-------
EPA/630/R-00/003
December 2000
Summary Report for the
Workshop on Issues Associated with
Dermal Exposure and Uptake
U.S. Environmental Protection Agency
Bethesda, MD
December 10-11,1998
Risk Assessment Forum
U.S. Environmental Protection Agency
Washington, DC 20460
Piloted on Recycled Paper
-------
NOTICE
This document has been reviewed in accordance with U.S. Environmental Protection Agency
(EPA) policy and approved for publication. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
This report was prepared by Eastern Research Group, Inc. (ERG), an EPA contractor (Contract
No. 68-C9-8148, Work Assignment No. 99-01) as a general record of discussion held during the
Workshop on Issues Associated with Dermal Exposure and Uptake (December 10-11,1998). As
requested by EPA, this report captures the main points and highlights of the meeting. It is not a
complete record of all details discussed, nor does it embellish, interpret, or enlarge upon matters
that were incomplete or unclear. Statements represent the individual views of each workshop
participant; none of the statements represent analyses by or positions of the Risk Assessment
Forum or the EPA.
n
-------
CONTENTS
Page
FOREWORD ; v
CHAIRPERSON'S SUMMARY ... .J. ... vi
1. INTRODUCTION .... 1
1.1 Workshop Purpose \
1.2 Workshop Participants .,.; 1
1.3 Workshop Agenda 2
1.4 Workshop Summary ...-.,." >...,. ..2
2. SUMMARY OF OPENING REMARKS .-..-: ...,,,...;..., .-•;., -..:.;.... 4
2.1 Welcome «...,.... :•..; ......;. '.;;.....:. 4
2.2 EPA Risk Assessment Forum's Role 4
2.3 Background on the Current Dermal Guidance 5
2.4 Charge to the Peer Consultants 8
3. DERMAL EXPOSURE TO CONTAMINANTS IN WATER 10
3.1 Presentation 10
3.2 Discussion 13
4. DERMAL EXPOSURE TO CONTAMINANTS IN SOIL 26
4.1 Presentation 26
4.2 Discussion 28
5. ADJUSTMENT OF TOXICITY FACTORS TO REFLECT ABSORBED DOSE 36
5.1 Presentation 35
5.2 Discussion 38
6. RISK CHARACTERIZATION AND UNCERTAINTY 46
6.1 Presentation 45
6.2 Discussion 49
7. NEXT STEPS: PLENARY DISCUSSION ON DERMAL EXPOSURE ISSUES . 58
111
-------
CONTENTS (cont.)
APPENDIX A: CHARGE TO PEER CONSULTANTS A-l
APPENDIX B: LIST OF PEER CONSULTANTS B-l
APPENDIX C: LIST OF OBSERVERS C-l
APPENDIXD: AGENDA D-l
APPENDIXES EPA PRESENTER OVERHEADS E-l
APPENDIX F: BREAKOUT GROUP OVERHEADS F-l
APPENDIX G: POST-MEETING COMMENTS G-l
IV
-------
FOREWORD
In January 1992, the EPA Office of Health and Environmental Assessment (now the National
Center for Environmental Assessment) completed an interim report entitled Dermal Exposure
Assessment: Principles and Application. This report provided guidance for conducting dermal
exposure and risk assessments. The conclusions of this report were summarized at the January
1992 National Superfund Risk Assessors Conference. During this meeting, Regional risk
assessors requested that a workgroup be formed to prepare an interim dermal risk assessment
guidance for the Superfund program. The purpose of this guidance would be to promote
consistency in the procedures used by the EPA Regions to assess risks from dermal exposure at
Superfund sites. In August 1992, a draft Superfund Dermal Guidance was circulated for review
and comment.
In 1995, a workgroup convened to address issues related to the August 1992 Superfund Dermal
Guidance and to redraft the document. The revised guidance was peer-reviewed in February
1998. Several issues related to dermal exposure and risk assessment were raised during the peer
review. The workgroup addressed some of these issues in a revised draft of the guidance. Other
issues raised during the peer review were broader in scope.
To address these broader issues, the EPA Risk Assessment Forum sponsored a workshop held on
December 10-11,1998, in Bethesda, Maryland. At this workshop, 20 peer consultants discussed
issues in four categories:
• Dermal exposure to contaminants in water.
• Dermal exposure to contaminants in soil.
• Adjustment of toxicity factors to reflect absorbed dose.
• Risk characterization and uncertainty analysis for dermal assessments.
In addressing these issues, the consultants were asked to consider:
• What is known today that can be applied to addressing the issue or providing
additional guidance on the topic?
• What short-term studies could be conducted to address the issue or provide
additional guidance?
• What longer-term research may be needed to address the issue or provide
additional guidance?
This report summarizes the discussions at the workshop.
BILL WOOD
-------
CHAIRPERSON'S SUMMARY
A workshop on Issues Associated with Dermal Exposure and Uptake was held to provide a
forum for discussion of generic technical issues raised during the February 1998 peer review of
the Superfund Dermal Guidance (SDG). The peer consultants who participated in the workshop
focused on four key areas of concern: estimating dermal absorption from water, estimating
dermal absorption from soil, the use of oral-dermal toxicity adjustment factors, and dermal risk
characterization/uncertainty. In each of these areas, the consultants attempted to identify and
categorize three kinds of issues: (a) those which should be addressed in the finalization of the
current SDG, (b) those which should be given immediate attention by the agency before any
future dermal guidance is prepared, and (c) those which should serve as the basis for an ongoing
research program to improve the accuracy and breadth of applicability of dermal risk assessment
methodologies. Overall, the peer consultants felt that the SDG document was generally well
written and provided a reasonable and justifiable basis for conducting dermal risk assessments,
given the current state of the art for the prediction of dermal absorption.
Specifically with respect to dermal exposure from water, the consultants endorsed the use in the
SDG of the predictive equation for skin permeability (Kp) of a chemical based on its octanol-
water partition coefficient (K^) and molecular weight (MW). However, the consultants felt that
the SDG should be revised to better document the derivation and implications of the 95%
confidence intervals and "effective predictive domain" (EPD) for the estimation of Kp. The
consultants strongly recommended that the agency give immediate attention to considering
alternative bases for regressions to estimate Kp, including the use of molar volume rather than
MW, as well as the use of molecular sub-structures, although it was recognized that these
alternatives could not be considered in the timeframe for publication of the current SDG
document. In addition, there was concern that the fact that the SDG gives preference to
predicted, rather than experimental, values for Kp could serve to discourage the collection of
experimental data on Kp for additional chemicals. Therefore, the agency should attempt to
encourage the collection of such data hi the future by the development of standard protocols
defining acceptable experimental determinations of Kp to replace the predicted values.
With regard to dermal exposure from soil, the use in the SDG of default soil absorption fractions
(Abs), which can be replaced with site-specific experimental data, was endorsed by the
consultants. There was general agreement that much more experimental data is needed to
improve this area of dermal risk assessment in the future. For this purpose, the agency should
develop standard protocols defining acceptable in vitro and in vivo methods for measuring
dermal absorption from soil. Factors that need to be considered include properties of the soil,
chemical composition/aging, and duration/nature of dermal contact. Additional data is also
needed on soil adherence and dermal transfer from surfaces (e.g., concrete).
It was the opinion of the consultants that the "toxicity factor" methodology for oral-to-dermal
extrapolation described hi the SDG is acceptable, but that the discussion hi the SDG should be
greatly expanded prior to publication to clarify the assumptions and limitations involved. In
VI
-------
particular, while the method described in the SDG is acceptable as a default, the guideline should
give preference to more desirable methods for route-to-route extrapolation, as discussed in a
previous EPA workshop (Gerrity and Henry, 1990). The oral bioavailability associated with the
critical toxicity study, which serves as the basis for the toxicity factor hi the SDG, should be
clearly distinguished from the human oral bioavailability used in oral exposure assessment.
The consultants felt that the discussion of risk characterization and uncertainty in the SDG was
generally adequate. It is recommended that some attempt be made to categorize the importance
of the various uncertainties listed in the document in at least a qualitative way (e.g., low,
medium, high). For the future, more quantitative sensitivity and uncertainty analysis should be
attempted.
The most consistent recommendation of the consultants was the need for the establishment of a
(funded) standing agency dermal working group. The functions of this working group would
include: (1) review of new experimental data on KpS or Abs to determine acceptability as a
replacement for the predicted/default values, (2) maintenance of a depository (preferably a
worldwide web site) for reviewed experimental values of Kp, Abs, etc., (3) development of
standard protocols, (4) fostering of exchange of information and standardization of dermal risk
assessment across agency programs as well as with other agencies, and (5) continuing evaluation
of progress hi the science of dermal absorption and the potential for its incorporation in agency
guidelines.
VII
-------
-------
1. INTRODUCTION
1.1 Workshop Purpose
The Workshop on Issues Associated with Dermal Exposure and Uptake was held on December
10 and 11, 1998, in Bethesda, Maryland, to discuss issues associated with estimating dermal
exposure and uptake of environmental contaminants. The workshop discussions focused on
generic technical issues raised during the February 1998 peer review of the Risk Assessment
Guidance for Superfund, Supplemental Guidance, Dermal Risk Assessment (hereafter known as
the Superfund Dermal Guidance). These issues are detailed in the charge to the peer consultants,
which is included as Appendix A of this report. Although the discussion topics detailed in the
charge were derived from the review of a proposed Superfund model, they are generically
applicable to the estimation of chemical uptake within many U.S. Environmental Protection
Agency (EPA) programs. Therefore, discussion of these issues at the workshop was intentionally
broader than the originating context of the Superfund Dermal Guidance.
1.2 Workshop Participants
The peer consultants for the workshop consisted of 20 experts in dermal exposure and uptake
from industry, academia, consulting, and state and Federal government agencies. Their expertise
covered a broad range of exposure and risk assessment topics including chemical principals,
dermal bioavailability, toxicity adjustments, and quantitative modeling. Over forty observers
also attended the workshop, including six members of EPA's Dermal Workgroup who had
authored the Superfund Dermal Guidance. The peer consultants and observers are listed in
Appendixes B and C, respectively.
-------
1.3 Workshop Agenda
The workshop agenda is provided in Appendix D. The workshop began with welcoming
remarks, presentations on EPA's Risk Assessment Forum and current dermal guidance, and a
review of the charge to the peer consultants. This was followed with a series of four discussion.
sessions on issues associated with:
• Dermal exposure to contaminants in water.
• Dermal exposure to contaminants in soil.
« Adjustment of toxicity factors to reflect absorbed dose.
• Risk characterization and uncertainty analysis for dermal assessments.
Each of these sessions began with a brief presentation, given by a member of the EPA Dermal
Workgroup, on background information relevant to that particular topic. The peer consultants
then divided into three breakout groups. (Breakout group chairs and members are listed in
Section 2.4.) For about an hour, the breakout groups separately discussed the specific issue
questions listed in the charge (Appendix A). Then all consultants reconvened in a plenary
session, during which each breakout group chair presented the key points from his or her
breakout group discussion and the consultants then discussed the issue further as a group.
Finally, observers were given an opportunity to comment. On the last afternoon of the workshop,
after the four issue areas had been discussed, the peer consultants held a plenary discussion of
dermal exposure issues and ideas for future workgroup activities.
1.4 Workshop Summary
This report summarizes the workshop presentations and discussions:
Section 2 of this report summarizes the opening presentations, which provided
background information and context for the workshop discussions.
-------
Sections 3 through 6 summarize the presentations and discussions in each of the
four issue areas. (The overheads used by the chairperson and EPA presenters are
provided in Appendix E. Overheads used by the breakout group chairs in
presenting the breakout group discussions can be found in Appendix F.)
Section 7 of this report summarizes the final plenary discussion on dermal
exposure issues. (The overheads developed during this discussion are inpluded in
Appendix F.) Following the workshop, two of the consultants submitted post-
meeting comments.These are included in Appendix G. ,
-------
2. SUMMARY OF OPENING REMARKS
2.1 Welcome
Jan Connery of Eastern Research Group, Inc. (ERG) opened the workshop by welcoming
participants and observers. She emphasized that the workshop was a peer consultation meeting
rather than a peer review meeting. While the Superfund Dermal Guidance would serve as a
resource for deliberations, it was not being peer-reviewed at the workshop. Ms. Connery stressed
to the peer consultants that the scope of their discussions in the four topic areas should extend
beyond Superfund to address issues generally applicable to estimating chemical uptake.
2.2 EPA Risk Assessment Forum's Role
Steve Knott, Exposure Science Coordinator for EPA's Risk Assessment Forum (RAF), provided
background on the RAF and its role in sponsoring this workshop. The RAF was established in
1984 in response to recommendations made by the National Research Council for improving risk
assessment practices in the Federal government. The mission of the Forum is to promote
agreement within the Agency on difficult risk assessment issues and to make sure that this
agreement is incorporated into Agency guidance. To do that, EPA assembles senior scientists
from the EPA program offices to participate in a formal process to study and report on issues
from an Agency-wide scientific perspective. Currently, 34 EPA senior scientists representing the
following offices are involved in the Risk Assessment Forum:
n Office of Prevention, Pesticides and Toxic Substances
» Office of Solid Waste and Emergency Response
» Office of Air Quality Planning and Standards
• Office of Water
• Office of Research and Development
-------
• Regions 1,2, 5, 6, 7, and 10
Mr. Knott explained that the projects these scientists take on are selected based on two criteria:
• They should involve controversial or cutting-edge issues.
• They should impact the risk assessment practices of the Agency as a whole,
affecting a multitude of Agency programs.
Mr. Knott then discussed the RAF's involvement with dermal uptake issues. In February 1998,
the Superfund Dermal Guidance was externally peer-reviewed. Two members of the Dermal
Workgroup, which considered the recommendations and comments of those peer reviewers, were
also RAF members—Kim Hoang and David Bennett. They recognized immediately that some of
the issues raised could be of concern in other dermal exposure initiatives within the Agency, so
they referred these issues to the RAF; this led to today's workshop. Specific issues and topics of
interest to the Agency that can be addressed in the workshop include:
• Aggregate exposure to pesticides
• Risks to children
» Dermal uptake of contaminants in drinking water
• Research planning
2.3 Background on the Current Dermal Guidance
Mark Johnson from EPA Region 5 and Mark Maddaloni from EPA Region 2 provided
background information and context concerning EPA's current dermal guidance. Mr. Johnson
outlined some crucial dermal issues that EPA thinks will help define and refine Agency guidance
documents. By refining its guidance documents, EPA hopes to aid risk assessors in performing
dermal pathway exposure assessments. Discussions during the workshop, and the resulting
-------
guidance, will hopefully be extremely valuable to EPA—not only to the Superfund risk
assessment process, but also to a diverse group of other Agency programs.
Mr. Johnson focused his overview on two topics: (1) the evolution of the Superfund Dermal
Guidance; and (2) how EPA plans to use this workshop's deliberations, and the information
gained from them, to refine its dermal exposure assessment methodology.
Beginning with the evolution of the Superfund Dermal Guidance, Mr. Johnson identified
significant dates and events associated with EPA dermal risk assessment:
• 1983: National Academy of Sciences (NAS) recommendations for risk
assessment methodology.
• 1989: Risk Assessment Guidance for• Superfund (RAGS).
• 1992: Dermal Exposure Assessment: Principles and Applications
(DBA)—EPA's Office of Research and Development (ORD).
• 1992: Superfund Interim Dermal Risk Assessment Guidance.
• 1995: Dermal Workgroup formed to update and finalize Superfund Guidance.
• June, 1997: Internal Peer Review of draft Superfund Guidance.
" January, 1998: External Peer Review of draft Superfund Guidance.
• August, 1998: Draft revised based on peer review comments.
• October, 1998: Discussion issues for this workshop were identified.
In 1983, NAS made recommendations for a risk assessment methodology that consisted of four
stages: hazard identification, toxicity assessment, exposure assessment, and risk
characterization. After adopting the NAS risk assessment paradigm, EPA's Superfund program
developed the Risk Assessment Guidance for Superfund (RAGS) in 1989. Superfund illustrated
the need for EPA to define a consistent risk assessment methodology (including dermal exposure
-------
guidance) for the dermal pathway. Superfund is a very decentralized program within EPA and its
risk assessment management decisions, methodologies, and practices vary greatly. Regional
Offices perform most Superfund risk assessments and dictate how to address each site. There are
thousands of sites and thousands of risk assessors, all of whom would benefit from EPA
guidance and consistent methodology for performing risk assessments.
RAGS was the Agency's first attempt to combine all of the elements of risk assessment into, a
guidance document. RAGS provided, guidance and a, consistent approach to risk assessments for
the EPA, as well as for the states, independent consultants, and others who performed
assessments. For example, the document provided quantitative recommendations for how to
estimate an absorbed dermal dose from water and soil. For some values (such as the chemical-
specific dermal permeability coefficient (Kp) for the estimated absorbed dose from water, and the
chemical-specific absorbed fraction from soil [ABS] and soil adherence factor [AF] for the
estimated absorbed dose from soil), RAGS suggests consulting the open literature. How.ever,
RAGS provides no guidance for doing this. Information is needed on the hundreds, of chemicals
that EPA characterizes at a Superfund site. Also, review and evaluation of the literature is
needed to develop recommendations for the dermal pathway.
For ABS and AF, RAGS suggests using conservative estimates when information is not
available. Because the Superfund program is highly decentralized, guidance to use conservative
estimates leads to various levels of aggressive approaches, such as assigning a single absorption
fraction to a whole class of chemicals. This has led to some degree of inconsistency across the
Regions.
In 1992, ORD developed the DEA, which provides the scientific basis for quantitative evaluation
of the dermal pathway. This document refined many of the parameters that were described
qualitatively in RAGS.
-------
Following development of the ORD document, Superfund developed the draft Superfund Interim
Dermal Risk Assessment Guidance, which distilled the key elements of the more technical ORD
document into practical guidance for Regional and state staff and their consultants. That draft
guidance was widely used.
The Dermal Workgroup was formed in 1995 to update and finalize the draft guidance. The
revised document was internally peer-reviewed at EPA in 1997 and externally peer-reviewed in
early 1998. The Workgroup revised the document based on the peer reviewer comments. In
addition, the peer reviewers raised a number of issues that EPA felt would best be addressed by a
separate peer consultation meeting. Many of these issues are captured in the charge for this
workshop (Appendix A).
Mr. Johnson pointed out that dermal risk assessments support many decisions. At Superfund
sites, for example, the dermal risk assessment may play a role in triggering cleanup and in
defining cleanup goals. EPA staff and many states rely on this type of guidance. EPA therefore
would like to increase the consistency and reduce the uncertainties of dermal risk assessment.
While uncertainty was to be specifically addressed as part of the fourth discussion topic at this
workshop, it also wove through the first three issue areas; Mr. Johnson expected that it might
arise throughout the discussions. Mr. Johnson emphasized that the Agency would like to reduce
the uncertainties in dermal risk assessment and he hoped that the workshop would help
contribute to that goal.
2.4 Charge to the Peer Consultants
Harvey Clewell of ICF Kaiser International, who served as the workshop chair, reviewed the
charge for the workshop. He suggested that, in approaching the meeting agenda, the peer
consultants adopt a mental framework of "Today, Tomorrow, and Future":
-------
"Today" includes issues and problems that could be immediately addressed or
fixed in the EPA document.
"Tomorrow" includes issues that research has not yet fully addressed. Some of,
these issues were touched upon in the EPA document, but more work needs to be
done before this information should be formally presented in a document. It seems
possible for these information gaps to be filled in the near future.
"Future" includes long-term needs and data gaps that should be addressed to
improve dermal risk assessment. It is unlikely that these issues can be addressed
in the near future, but they will eventually need to be examined in order to fully
understand dermal exposure assessment issues.
Mr. Clewell said that the topics to be featured in the "Next Steps" discussion at the end of the
workshop would be determined by the needs, interests, and desires of the peer consultants.
Therefore, he asked the consultants think about what they would like to see happen in the future.
Mr. Clewell divided the consultants into three breakout groups as follows:
' «. " * ' v,,^ ,?,*•
(Sroup 1 u> , "
' * S *• > -¥,^x
Annette Bunge*
Clay Frederick
Clint Skinner
Gerhard Raabe
Jim Knack
Val Schaeffer
Ron Brown
-^r^Grpup2 "• /-.'
Gary Diamond*
Kurt Enslein
Paul Chrostowski
Philip Leber
Stephen Di Zio
Robert Duff
„;„ \t : . JSroijgS^ f ^ & ;
John Kissel*
Jim Bruckner
Rosalind Schoof
Deborah Edwards
Bob Bronaugh
Lawrence Sirinek
* Chairs
-------
3. DERMAL EXPOSURE TO CONTAMINANTS IN WATER
3.1 Presentation
Kim Hoang from ORD presented background information for the first discussion topic: dermal
exposure to contaminants in water. (Dr. Hoang' s overheads are included in Appendix E.)
In the current approach for organic chemicals, the skin permeability (Kp) is estimated as a
function of a chemical's octanol/water partition coefficient (Kow) and its molecular weight (MW).
The relationship between Kp and these factors is based on a regression analysis of measured skin
permeabilities. For metals and inorganic chemicals, the Superfund Dermal Guidance
recommends using default Kp values. In the absence of measured values, a default of 0.001
cm/hr is recommended.
Dr. Hoang listed the methodologies in the ORD DBA that will not be changed, and then listed
those that will be changed. Methodologies that will not be changed include:
• Use of a two-compartment membrane model to represent the skin.
• Approximations of exact solutions.
» Use of the Dermal Absorbed Dose per event (DAevent) (estimated from Kp) for
event time (tevent) < time to reach steady state (t*), DAevent proportional to N tevent.
• Use of DAevent (estimated from Kp or Kp,max) for tevent > t*, DAevent proportional
toNtevent.
• Use of Kp correlation as a function of Kow and MW.
EPA believes that these methodologies are scientifically sound and do not need further review.
However, EPA would appreciate input on other methodologies in the guidance document that do
need further development and improved scientific focus, including for example:
10
-------
• Improving KJJ correlation for organics: •-..:.• , ,
— Using Flynn's database as the ORD DBA does.
— Taking out three in vivo data points (xylene, toluene, styrene).
— Using two predictors: Log Kow and MW.
— Calculating 95% confidence intervals (CIs) for both Flynn's data and the
200 chemical predictions. ...
• Establishing a 95% CI for predicted Kp of existing chemicals in the ORD DBA.
'•'•-.• ' - f •
• Establishing an effective predictive domain (EPD) for predicted Kp.
—- Statistical analysis of collinear data.
—; From the original experimental data set, allow the determination of an
/ EPD for extrapolation of unknown Kp. . • , .
• Determining Kpmax for chemicals outside of EPD/
• Determining Kp for inorganics and default values.
• Reassessing other default exposure assumptions.
Dr. Hoang noted that the dermal modeling approach used to derive a correlation equation was
based on the experimental Flynn database. The Flynn database includes estimated values for
over 200 chemicals. These estimated values are derived with uncertainty bounds (95%
confidence level) to extrapolate data from known experimental values. She explained that the
95% confidence level provides some idea about the range of Kp with which risk assessors are
faced. This allows site risk assessors to know the magnitude of uncertainty in their dermal risk
assessment exposure dose estimates. Outside the EPD, however, the estimates are not valid.
Knowing the magnitude of uncertainty should help risk assessors to (at least qualitatively)
improve their dermal assessments. Currently, dermal risk assessors have no reliable data set to
assist them in deciding which Kow to use.
11
-------
Dr. Hoang showed overheads that graphically depicted the EPD of ORD DBA predictions
compared to Flynn's database, bounded by the 95% CIs (see Appendix E). Virtually all DEA-
identified chemicals fell within the predicted EPD box of the Flynn database. Dr. Hoang showed
a list of chemicals with high and low Kow values that fell outside the EPD. She remarked that
xylene, toluene, and styrene should be removed from the data set. This is because these three
data points were collected using in vivo methodologies, which are not readily comparable to all
other chemical data points derived from in vitro studies.
Dr. Hoang said she would welcome any suggested improvements to the current approach. The
current methodologies use two predictor units (Log Kow and MW) that are not interchangeable.
There are no real guidelines for risk assessors concerning this dermal exposure issue. Dr. Hoang
also discussed the default KpS for inorganics, as well as other default exposure values such as the
recommended dermal exposure values for central tendency and RME residential scenarios for
water contact. Many of these default values may not accurately represent real-life scenarios.
Depending on the nature of the contaminant in site-specific conditions, some chemicals may
have a range of K^s. For these chemicals, using default values may over- or underestimate
dermal exposure dose estimates.
Comments would be particularly welcome on any of the following topics:
For organic chemicals:
H The database used to derive the correlation equation.
« The correlation equation (predictors Kow and MW) used to estimate the Kp and the
95% CI.
• The statistical analysis used to establish the EPD for the Kp correlation equation.
a TheuseofKpimax.
a The use of estimated K versus experimental data.
12
-------
For inorganic chemicals:
• The approach recommended for metals and inorganic chemicals.
• The other exposure default values.
• Using the model instead of a chemical-specific study—which one is better?
Why? (i.e., benefits, disadvantages of each).
The goal of the EPA Dermal Workgroup is to develop, for use at risk assessment sites, one
correlation for all Kp estimates of various chemicals. The process of developing this correlation
basically focuses on estimating absorption of a chemical into the skin. It does not include
inhalation exposure, nor has it yet looked at the chemical volatilization process. Dr. Hoang said
that this methodology essentially extends the previously used membrane model to create another
physiological model. She looked forward to hearing how the breakout groups discussed the
specific Kp issues listed in the charge (Appendix A). She hoped that the discussions would touch
upon current thinking about the experimental versus estimated Kp values.
3.2 Discussion
Following Dr. Hoang's presentation, the peer consultants divided into three breakout groups to
discuss dermal exposure to contaminants in water. Based on the charge (Appendix A), the
breakout groups focused their discussion on the following six topic areas:
1. Comment on the correlation equation used to estimate the skin permeability coefficient
(Kp) for organic chemicals. Is the approach used to estimate the Kp values and their 95%
confidence intervals plausible? Include in the discussion consideration of the database
analyzed to generate the correlation equation.
2. Comment on the statistical analysis used to establish the Effective Predictive Domain for
the Kp correlation equation (i.e. the range of Kow and MW where the predictive power of
the regression equation would be valid). Evaluate the new methodology for calculating
K max for chemicals outside of the Effective Predictive Domain.
13
-------
3. Comment on the use of Kp and Kp,max in the dermal absorption model (specifically the use
of Kp for all tevent (exposure time)t*).
4. Comment on the use of predicted Kp or Kp max vs. Chemical specific experimental values.
Consider the criteria used to select studies to develop the regression model (see
Appendix A of the Superfund Dermal Guidance). Should these and other criteria be used
to judge chemical specific experimental values? What are the minimum criteria that
should be satisfied before chemical specific experimental values can be used in lieu of
model predictions?
5. Comment on the approach recommended for metals and inorganic chemicals. Is the
default Kp (0.001 cm/hr), that was previously recommended in the 1992 Interim Guidance
for Dermal Exposure Assessment, still scientifically sound and defensible?
6. Comment on the other default exposure assumptions (see Table 3.2) recommended to
estimate the DAeV(!nt (e.g., tevent = 10 minutes for exposure in a shower). Are these events
scientifically sound and defensible?
Following the breakout discussions, the consultants reconvened in a plenary session. The
breakout group chairs summarized the discussions as follows. (Copies of the overheads used by
the chairs in making their presentations are included in Appendix F.)
Group 1
Chair: Annette Bunge, Colorado School of Mines
Discussion Area 1
Group 1 determined that the correlation equation used to estimate the Kp for organic chemicals is
probably sufficiently accurate, but that molar volume may be a better approach. Some members
of Group 1 felt uncomfortable using a 95% CI because it assumes an unknown error structure.
The database used to generate equations should include more relevant chemicals, particularly
high and low Log K^ chemicals. Procedures for collecting K^, data should be standardized.
Group 1 noted that Superfund is interested in many high production persistent chemicals.
Discussions emphasized the need to increase data generation, specifically focusing on persistent
14
-------
chemicals in water. Group 1 discussed the importance of these persistent water-borne
contaminants via the dermal pathway for Superfund and outside Superfund.:
Discussion Areas 2, .3, and 4
Commenting on the statistical analysis used to establish the EPD for the Kp correlation equation,
Group 1 decided that it was sufficiently accurate, but that risk assessors should not extrapolate
outside the EPD. Experimental Kp values are not a problem if they contain only small errors and
standard deviations. Group 1 felt that concern was warranted when there was only a single Kp ,
study, or when there were several experimental KpS with high variation. Group 1 also felt that
there should be a domain based on the properties of data in the database. The methodology for
calculating Kpmax for chemicals outside of the EPD was sufficiently accurate in the dermal
absorption model, but Group 1 felt that some questions still need to be answered hi this area.
Group 1 was uncomfortable with the use of Kp, and Kpmax in the dermal absqrption model because
they felt this method ignored data. Group 1 suggested that the "consensus". experimental values
(e.g., the average) could be reported as an alternative. Group 1, however, came to no definite
conclusions regarding the.use of Kp and Kp ^ in the dermal absorption model. They did not ,
decide what criteria Should be used to' judge chemical-specific experimental; values, or what >
minimum requirements should be satisfied before chemical-specific experimental values can be
usedin.lieuof model predictions. >•',.- • ., •'.'.;.-
Discussion Areas ,5 and 6 , , , , ,
As for the approach recommended for metals and inorganic chemicals, Group 1 felt that, the
default Kp (0.001 cm/hr) was sufficiently accurate, but noted that this default valpe did, not
account for chemical speciation. Group 1 discussed whether methyl mercury should be treated as
an organic chemical rather than a metal. They also discussed the differences in dermal exposure
from vapor mercury and water mercury. Group 1 asked why arsenic was not included in the ;
metals list. When discussing shower default exposure assumptions, Group 1 expressed no great
concerns or significant comments.
15
-------
Discussion Areas 4, 5, and 6
Group 1 suggested that the document contain a description of the issues it does not address; some
of these issues, though not relevant to Superfund, may be important on the broader scale. For
example, pesticide absorption from pesticide formulation is not treated in the document.
(Pesticide absorption is not a pathway of concern at most Superfund sites, but it is of concern in
other EPA programs.)
Group 2
Chair: Gary Diamond, Syracuse Research Corporation
Discussion Area 1
Group 2 expressed many ideas similar to those of Group 1. Group 2 expressed a concern that the
predictive model used to estimate Kp for organic chemicals may be wrong. Group 2
recommended several ways to improve the Kp estimation. First, they suggested that researchers
analyze and include information from the Vecchia database into their model. Group 2 also
suggested that the model include substructural parameters in the correlation analysis, explore
other K™ predictive models, explore nonlinear models for relating Kow and Kp, and use molar
volume in place of molecular weight in the prediction algorithm. Group 2 discussed the need to
consider variable dependency in the prediction algorithm (MW-KOW, molar volume-Kow) and to
consider modeling transformed data (e.g., Log). They emphasized the need for scientific review
of KOW values to create a high quality database of Kp values. Group 2 suggested including
information from the "Star List" into this data review. An attempt should be made to identify
any other available data which could be incorporated into the model.
Group 2 stressed the need for the dermal guidance document to explore experimental Kow values,
including how the predictive model may be wrong. Group 2 also said that the document needs to
clarify the derivation of the 95% upper confidence level. The document needs to explicitly state
this information.
16
-------
Discussion Area 2
Group 2 said that the document needs to clarify how it is defining "outliers of EPD" before they
can comment on the statistical analysis used to establish the EPD. Moreover, the EPD approach
needs to be clarified in the document to minimize confusion and disagreement about the EPD
approach. Group 2 felt that EPA needs an approach for how to replace the predicted Kpmax with
experimental values, noting that the predictive approach yields highly conservative values o
Discussion Area 3
Regarding exposure duration, Group 2 said that the document currently has no basis for
distinguishing between exposure time and the time it takes to reach steady-state absorption.
Exposure duration needs to be raised as an important uncertainty that is explicitly stated in the
document.
Discussion Area 4
Group 2 felt that the use of a predicted model value (instead of chemical-specific experimental
values) might not reward new data collection because researchers might see the model's value
and assume that it is well established and needs no further verification. If EPA "forced"
everyone to use its data, they might be discouraging people from going out and collecting new
data because potential sponsors of research (i.e., industry) will conclude that EPA will not accept
the use of new experimental data in site risk assessments. Therefore, Group 2 felt that EPA
should make an effort to assure researchers that future data can and will be incorporated into this
model and that verification of this model is greatly needed. Group 2 felt that EPA needs to
reward data collection, or at least the idea of data collection.
Group 2 believed that use of a predictive model will contribute to consistency among different
users at a diversity of sites. Group 2 stressed a need for criteria for evaluating experimental
values. Currently there is no basis for deciding when to use experimental data and when to use
estimated values. The group briefly discussed what criteria might be used to determine when to
use experimental values, but no criteria were forwarded to the peer review panel for discussion.
17
-------
Discussion Area 5
Group 2 recommended better documentation and explanation of how the workgroup derived
chemical-specific Kp values. Group 2 felt that the workgroup's methodology should consider
both ionized and non-ionized states of inorganics and whether or not they are using a value
arrived at via in vivo or in vitro studies. The workgroup should consider the speciation of metals
in its methodology. Group 2 asked how EPA deals with the ionization issue and felt that the
document should explain the workgroup's approach, including the use of models and an
explicitly stated assumption about relevant solution chemistry parameters (e.g., pH, ligand
concentrations, etc.)
Discussion Area 6
Regarding other default exposure assumptions, Group 2 felt that the workgroup needs to generate
a loss term for dermal factors (i.e., loss to air, exfoliation, etc.) and default exposure
assumptions. These factors should be generated on a chemical-specific basis. Group 2 felt that it
would be useful to explore probabilistic approaches for representing the other factors in the dose
algorithms. Group 2 members felt that it was possible, albeit unlikely, for shower default
absorption values to exceed drinking water exposure default values for volatile organic
compounds (VOCs). They felt that this was not a likely scenario for non-VOCs.
Other Issues
After discussing all issue topics, Group 2 discussed in vitro correlation, specifically the Potts and
Guy equation. They also discussed the EPD (K^ J, short-term versus long-term exposure
events (2.4 T)ag), and criteria for predicted Kp values versus measured values.
Group 3
Chair: John Kissel, University of Washington
18
-------
Discussion Areas 1-4 ,
Group 3 grouped the first four issues together in their discussion. Several consultants said that
they would like to see.the document clearly explain the workgroup's approach to dermal risk
assessment as it relates to maximum contaminant levels, and uncertainties involved in this
approach. Adding a discussion about maximum contaminant levels might put the dermal risk
assessment document ,in perspective, and help prevent risk assessors from using this information
in ways unintended by the workgroup. Group 3 determined that EPA guidance for dermal risk
assessment should encourage (or at least not actively discourage) data collection. The EPA ,
dermal risk assessment methodology needs to contain a standard protocol if siterspecific
experimentation is to be conducted. Group 3 emphasized that talking about models in the ,
absence of site-specific validation is short-sighted. EPA needs to seek validation via
experimentation. EPA also needs to seek consistency, and their guidance document should point
out where consistent and inconsistent topics exist in the area of dermal risk assessment.
Discussion Area 5
Group 3 criticized the default assumption value of 10'3 for inorganics. They raised questions
about mercury and concluded that the EPA workgroup needs more investigation and information
on mercury before they release the document. The default value for other metals seemed very
conservative. .
Discussion Area 6
Group 3 questioned whether a 10-minute default shower duration was adequate.
Additional Comments
Group 3 suggested that EPA needs to use residential versus recreational variables in Figure 1.1 of
the document. Group 3 also expressed concern that maximum contaminant concentrations must
be used. Finally, Group 3 suggested that the wprkgroup consider improving the clarity of the
document's writing. < ,
19
-------
General Discussion
Dr. Clewell summarized the three groups' discussions, saying that the groups apparently agreed
that the correlation equation used to estimate Kp for organic chemicals and the use of the
predicted Kp values were appropriate, but that they could be improved. The success of such
methodologies varies by contaminant. The three groups grudgingly accepted the EPD domain for
the Kp correlation model and the use of Kp and Kp>max in the dermal absorption model. Some peer
consultants felt the model should address lag times and distinguish between short-term versus
long-term exposure events. Other criteria to be considered when generating dermal data include
factors such as time to steady state, donor vehicle, receptor vehicle, temperature, pH, in vivo
versus in vitro, pore effects, and metabolism.
Peer consultants said that they would like to see EPA demonstrate an in vivo/in vitro correlation.
Several peer consultants noted that the higher molecular density compounds, such as halogenated
hydrocarbons, may be the most problematic compounds to model.
Peer consultants recognized that, for in vivo systems, species differences come up, as well as
individual factors like age, health, and skin region. Other differences arise between in vivo and
in vitro systems as well. For example, dead human skin has low glucose levels compared to live
in
skin cells. In addition, metabolism in the skin must be considered.
The workgroup model currently does not include metabolism or chemical metabolic factors into
its calculations. Subsets of chemicals influenced by metabolic factors, however, have been
identified (e.g., metabolism for organophosphate pesticides). For such chemicals, evidence
shows that pore effects and metabolism in the skin are significant. One consultant suggested that
EPA look at chemicals in the body, where they go after they enter the body, and how they are
metabolized by various parts of the body. Currently, there is a complete data break between how
much of a chemical is absorbed and what happens to the chemical once it enters the body. Peer
consultants agreed that there may be different criteria, depending on the nature of the chemical,
20
-------
as to what might be worth considering in the experimental design. It is probably going to be
necessary to set criteria that are chemical-class-specific, but this does not prevent the workgroup
from coming up with protocols.
Peer consultants noted that the EPA workgroup probably needs to create different criteria
depending on the nature of the chemical. They briefly discussed tissue issues (e.g., split versus
full, in vivo versus in. vitro), and noted that some comparative studies have already been
conducted and data are already available. Currently, some data exist on the effect of different
kinds of skin tissue preparations that can be helpful in ascertaining whether the skin preparation
is going to significantly impact the experiment. For example, the temperature effects on uptake
and the pH effects on uptake are already well documented.
Annette Bunge suggested that EPA try not to use anything in the system that would alter the skin
permeability more than contact with water would. For example, when conducting dermal
experiments, try not to include surfactants; use only water so that additional variables and
potentially confounding factors can be eliminated. Peer consultants agreed that a Volpo "no
barrier" receptor vehicle would be a good standard receptor vehicle for these experiments
because Volpo does not affect skin permeability.
Mr. Clewell asked if peer consultants were generally comfortable with the conclusion that a 10-
minute shower may lead to more chemical exposure than a full day's drinking water. No peer
consultants strongly disagreed with this, although several said that it was highly improbable.
One peer consultant noted that inhalation in a shower is an extremely important pathway. More
empirical demonstrations have been done to demonstrate the significance of inhalation exposure
than have been done for dermal exposure. Current information on completed dermal pathway
exposure in a shower consists of only one data set for one chemical (work done by Jo and
coworkers). Researchers have no idea how important the dermal pathway is for non-VOC
chemicals. Peer consultants agreed that EPA needs to gather more shower data, especially for
nonvolatile organic contaminant compounds.
21
-------
Val Schaeffer suggested consulting the Organization for Economic Cooperation and
Development (OECD) for information on in vitro guidelines and related dermal exposure issues.
He said that OECD is currently developing standardized methods to assess dermal exposure and
that this workgroup should be aware of OECD progress. OECD's standardized methods involve
many criteria this workgroup has addressed. Dr. Schaeffer suggested that there may already be
an in vitro method being considered and harmonized by OECD. The OECD project sounds like
it has initiated the process to select the criteria needed for data acceptance of in vitro studies.
Clay Frederick clarified that there are two types of data collection. The first is obtaining water
partition coefficients, K,,w from available literature and experimental data. These coefficients are
values based on physical and chemical properties. They should be reproducible values, assuming
researchers use the same protocol. The second type of data collection, for Kp, is more
problematic and justifies modeling extrapolations of existing data.
This brought the group to discuss criteria/credibility issues. Dr. Schaeffer said that standard
criteria methods for dermal studies are being developed right now. Dr. Frederick emphasized,
however, that it would be worthwhile to support the creation of the criteria. Other peer
consultants agreed that this should be emphasized.
Dr. Bunge commented on the correction of model guidelines. She said that EPA could do the
same thing for molar volume measurements that it has done for molecular weight in the model.
Dr. Bunge said that chemicals with larger MW for their size (e.g., halogenated hydrocarbons)
tend to be outliers of the current EPA correlation for estimating permeability coefficients. Based
on the small number of chemicals that have been analyzed and modeled using both molecular
weight and molar volume, participants felt that this issue belonged in the "Tomorrow" category.
Dr. Frederick said that he would like to see this workgroup establish a dermal database/website
on EPA's website similar to EPA's Integrated Risk Information System (IRIS). This website
could be updated regularly with the most recent available scientific information for reference by
22
-------
risk assessors and; other interested parties. Several other,peer consultants endorsed this idea of a
central reference point and information clearinghouse to. help ensure thatrisk assessor ^11 use the
same data sets, and to provide consistency throughout EPA's dermal risk assessment
methodologies. This website could be set up similarly to EPA's Right-to-Know initiatives. A
dermal website could potentially have a very strong influence on enhancing the quality of
experimental data, improving data compatibility between studies, and strengthening risk
assessments involving dermal pathway exposure. Dr. Frederick felt that this database/website
might be fairly simple to start up and could fall into the "Today" category. Other peer
consultants felt that this project was likely to be fairly labor-intensive and would better fall in the
"Tomorrow" category.
Observer Comment
Leonard Kieffer, EPA
Leonard Kieffer commented on the OECD in vitro guideline, saying that OECD is probably not
as far along as Dr. Schaeffer suggested. Canada and the U.S. are currently in major disagreement
with the rest of the OECD. Some work is being done or will be started that tries to correlate in
vivo and in vitro studies to achieve better guidelines. Mr. Kieffer recommended that this
workgroup put the in vitro guideline initiative in the "Future" category.
Brail Berattum, BASF Corporation
Brail Berattum had one generic comment applicable to the entire meeting. The Chemical
Manufacturers' Association (CMA) is currently funding a research program with $25 million per
year for dermal funding. This funding is spread among ten broad areas. One of the areas is
exposure assessment, led by Mike Jajak from Rohm and Haas Company. In the Exposure
Assessment Group^ne of the key issues is exposure parameters specific to the dermal pathway.
The Exposure Assessment Group met the week of December 14, 1998, and reviewed the CMA
requests for proposals for these broad-scale dermal issues, including identifying a methodology
t
23
-------
and nonchemical-specific criteria for dermal exposure. Mr. Berattum volunteered to serve as a
liaison between CMA and peer consultants, EPA, and other parties interested in this funding.
24
-------
4. DERMAL EXPOSURE TO CONTAMINANTS IN SOIL
4.1 Presentation
Mr. Johnson from EPA Region 5 presented background information concerning the second
discussion topic: dermal exposure to contaminants in soil. (Mr. Johnson's overheads are
included in Appendix E.)
Mr. Johnson began by saying that the Dermal Workgroup's recommendations for estimating
dermal exposure and uptake of chemical contaminants in soil are the same as those presented in
the Superfund Dermal Guidance. This approach estimates an absorbed systemic dose for dermal
contact with soil based on limited data. In the absence of measurements, a default absorption
fraction of 10% is recommended for semivolatile organic compounds (SVOCs). For inorganic
chemicals, a default of 1 % is recommended.
Estimates of soil-to-skin adherence must be used with estimates of absorption fraction from soil
to calculate the Dermal Absorbed Dose per event (DAevent). Recommendations on the use of soil-
to-skin adherence factors (AFs) also are presented in the Superfund Dermal Guidance. Soil-to-
skin AFs vary by activity and soil moisture content. Mr. Johnson briefly discussed the two
methods used to evaluate dermal absorption from soil:
• Fraction absorbed (percent of applied dose absorbed into blood)
• Flux model (rate of migration of chemical in skin)
Mr. Johnson raised the question about whether or not the fraction-absorbed approach was the
most appropriate at this time. He reviewed the calculation the workgroup used to estimate an
absorbed dose per exposure event (see overhead in Appendix E) and suggested that the
consultants think about how the monolayer theory impacts the dermal absorption fraction value.
25
-------
Mr. Johnson also showed the consultants an overhead of activity-specific surface-area-weighted
AFs, noting that an adjustment for Kow was not yet included in the estimates.
Mr. Johnson also showed an overhead of recommended dermal absorption factors from soil for a
variety of inorganic compounds. Recommended dermal absorption factors, however, were only
available for a handful of chemicals and chemical families. For these chemicals, the dermal
absorption factors were average numbers based on small data sets with very limited (if any)
statistical rigor. Chemicals for which experimental data are unavailable include most metals and
SVOCs. The generic defaults for screening the metals and SVOCs were 0.001 and 0.1,
respectively. Mr. Johnson hoped that the peer consultants would have a chance to discuss the
appropriateness of these screening values.
Mr. Johnson listed the soil AFs that can influence absorbed systemic doses for dermal contact
with soil:
• Soil properties influence skin adherence (e.g., moisture content, particle size, soil
type).
a Soil adherence varies across different body parts.
• Soil adherence varies with exposure activity.
The workgroup evaluated several exposure scenarios for the dermal pathway, including
residential child, residential adult, commercial/industrial worker, and recreational. Mr. Johnson
listed numerous activities that fell into these four categories of exposure scenarios. A full list of
these activities is included in Appendix E, and includes such possibly exposed groups as children
playing, gardeners, farmers, and soccer players.
Mr. Johnson described the workgroup's method for weighting soil AFs for a variety of body parts
into one estimate of exposure for the whole body. An overhead depicted body part-weighted
averages using several shaded boxes enclosed in one large rectangle. The rectangle represented
26
-------
the potentially exposed body, with each shaded box representing a specific body part (e.g., leg,
arm, face). The shade intensity of each shaded box represented a chemical's loading density; an
unshaded box indicated an unexposed or unevaluated body part. Using these methods,, the
workgroup obtains a body part-weighted average AFs for the entire body that is easily
incorporated into dermal risk assessment calculations. Such a body part-weighted average is
convenient'because the exposed average can be used as a surrogate estimate of other exposed
areas lacking quantitative data.
To conclude his presentation, Mr. Johnson showed an overhead of estimated activity-specific
surface-area-weighted soil AFs. These weighted soil AFs were based on exposure to face,
forearms, hands, lower legs; and/or feet. The soil AFs also accounted for activities arid exposure
scenarios. Weighted soil AFs were reported for the 50th and 95th percentiles. Mr. Johnson felt
that the 95th percentile represented a reasonable maximum exposure estimate. Dr. Clewell
raised the issue of how to adjust the soil AF to account for site-specific conditions. For example,
the soil AF could factor soil particle size into the model. '
' " •' ' ' ' ' • •'..'.• ' ' ' '.
4.2 Discussion
Following Mr. Johnson's presentation, the peer consultants divided into the three breakout
groups to discuss dermal exposure to contaminants in soil. Based on the charge (Appendix A),
the breakout groups focused their discussion on the following five topic areas:
Discuss the current absorption fraction approach as applied to the dermal absorption of
chemical contaminants from soil. Consider such factors as the duration of soil contact, the
soil particle size, and the level of soil moisture and whether these factors should be used
to adjust the absorption estimate. Overall, is the proposed methodology scientifically
sound and defensible? , ,
Comment on the soil absorption values presented in Table 3.4 of the Superfund Dermal
Guidance. Are these estimates supported by the available data? Are they scientifically
sound and defensible?
27
-------
3. Comment on the proposed default absorption fraction of 10% for organic compounds in
soil. Is the rationale for selecting this default clear and transparent? Is the estimate
scientifically sound and defensible? Is there enough supporting evidence to allow this
estimate to be characterized as representative of the average?
4. Comment on the proposed default absorption fraction of 1% for inorganic chemicals in
soil. Is the rationale for selecting this default clear and transparent? Is the estimate
scientifically sound and defensible? Is there enough supporting evidence to allow this
estimate to be characterized as representative of the average?
5. Comment on the proposed approach to calculate a total body soil-to-skin AF
based on the surface area weighted AFs for each body part. Is this a scientifically sound
approach? Could other methodologies be recommended?
Following the breakout discussions, the consultants reconvened in a plenary session. The
breakout group chairs summarized the discussions as follows. (Copies of the overheads used by
the chairs in making their presentations are included in Appendix F.)
Group 1
Chair: Annette Bunge, Colorado School of Mines
Discussion Area 1
Group 1 felt that the workgroup's dermal absorption fraction approach may be the best available
approach given current theoretical limitations and data limitations. They noted that the data
assumed 24 hours of exposure and that not all data in the literature are included. Group 1 raised
the issue of how new data should be included into the table. The group suggested that all data
listed in the table should include and adjust for monolayer cover factors (and be noted as such).
Monolayer cover or less represents what realistically occurs. Group 1 expressed concern that the
experiments currently do not account for numerous factors that affect dermal absorption,
including the effects of exposure time, mechanisms of transfer from soil to skin, and soil transfer
to the skin. Specifically, these variables include sweating, direct contact, vapor pressure (even
for nonvolatiles), and turnover versus no turnover.
28
-------
Discussion Area 2
Group 1 said that the workgroup's accepted default values for soil absorption were not
completely scientifically defensible because numerous factors were not incorporated into the
table's data. For example, the table did not account for:
• Properties of applied soil (e.g., organic carbon, moisture content)
• Particle size distribution
• Amount of applied soil relative to amount required for monolayer coverage
• Where monolayer coverage occurred (i.e., adjustment to monolayer)
• Contamination procedure
• Default chemical groups
Group 1 recommended that the workgroup create chemical default values based on groups or
categories of chemicals. These chemical groups could be defined by chemical vapor pressures or
by chemical ability to transfer from the soil. Group 1 recognized that understanding mechanisms
of transfer in order to group chemicals may require more experimental data.
Discussion Area 3
Group 1 felt that the rationale for the default absorption fraction of 10% for organic compounds
was not clear and transparent. Moreover, they did not believe that the 10% default value was
scientifically supportable because the data are weak.
Discussion Area 4
Group 1 said that no clear rationale existed for the default absorption fraction of 1 % for inorganic
compounds. They felt that there was not enough supportable evidence for the 1% default
assumption.
29
-------
Discussion Area 5
Group 1 had no comment on the workgroup's proposed approach to calculate a total body soil-to-
skin AF based on the surface-area-weighted AFs for each body part.
General Discussion
Group 1 summarized their conclusion by saying that researchers need to gather and generate
more data on dermal exposure to contaminants in soil. Group 1 suggested that part of the dermal
workgroup charge should be to create standardized procedures so that people can really start
gathering more data that can be integrated with wide-ranging applications.
Group 2
Chair: Gary Diamond, Syracuse Research Corporation
Discussion Area 1
Group 2 said that more experimental data are needed on dermal exposure to contaminants in soil
to fully assess the adequacy of the workgroup's approach. Specifically, Group 2 would like to
see more experimental data generated on kinetics, of adsorption and absorption of chemicals
from soil. There may need to be a time adjustment to the absorption factor to account for rate
effects. An exploration of the modeling of adsorption and a screening approach to estimate
adsorption (e.g., solvent extraction assays) would be useful. Group 2 recommended creating
guidelines for experimental data colleption. They would like to see a standard soil
characterization for experimental data that would provide specific guidelines about soil particle
size, loading, and aging factors. Current experimental data appear biased toward large particle
sizes and unaged soil. Group 2 also felt it was important to create guidelines for in vivo
experiment preparation and protocol.
Discussion Area 2
Group 2 identified numerous chemical-specific problems with the soil absorption values cited in
Table 3-4 . They did not elaborate on these chemical-specific issues, but said that available data
30
-------
are limited and more data are necessary for a variety of contaminants in order to make these soil
absorption values scientifically sound and defensible. ,
Discussion Area 3 . , . . , • • ,
Group 2 felt that the proposed default absorption fraction of 1Q% for organic compounds in soil
is adequate ("better than nothing") given the limited data set. They recognized that the data set is
limited and that this 10% may not be the best value, but that it appears to be the best given
available data limitations. Group 2 emphasized, however, that the document does not adequately
explain the basis and empirical support for the 10% default value. Current empirical support for
this default value is not overwhelming: existing data are biased and limited. Group 2
emphasized the need for more experimental data on absorption fraction. They suggested that the
workgroup may want to present these default values to risk assessors as a distribution or range of
values reflecting uncertainity, rather than as a point estimate. ...... ,. .,. .
Group 2 also suggested that experimental data for soil extraction using a solvent might be a basis
for departing from theuse of these default values. Therefore, the document might want to ,
include a discussion about when and how these default values should be used fpr screening or
baseline risk assessments. Group 2 said that the document, should explicitly state the limitations
of using these default values for screening purposes. , . ,
Discussion Area 4
Group 2 expressed little confidence in the proposed default absorptipn fraction of 1% for
inorganic compounds in soil, primarily because the 1% value is based on only two studies
(arsenic, cadmium). Group 2 said that they need more data in order to reach agreement on
inorganic default values. They asked the workgroup to clearly explain the basis and rationale for
the cadmium value, specifically why cadmium was set apart from other inorganic compounds.
Group 2 suggested that the workgroup may want to present these default values to risk assessors
as a distribution or range of values rather than as a point estimate. .
31
-------
Discussion Area 5
Group 2 felt that the proposed approach to calculate a total body soil-to-skin AF based on the
surface-area-weighted AFs for each body part was "a great leap forward." Group 2 endorsed the
general concept, but for the most part felt uncomfortable with using percentages in these
calculations. Instead of percentages, they would prefer to do more analysis of underlying
probabilities to determine central tendencies and the underlying confidence level of such
estimates. Group 2 suggested that the workgroup explore various statistical approaches for
estimating the central tendencies and reasonable maximum exposure (RME) values.
Group 3
Chair: John Kissel, University of Washington
Discussion Area 1
Group 3 said that, for the short term, the workgroup's dermal absorption fraction approach
appears acceptable, but that the current approach is limited because it relies entirely on
experimental data for absorption at 24 hours. This constraint may be problematic when the
workgroup tries to increase the complexity of their approach to consider the actual duration of
exposure events. For the long term, Group 3 would like to expand the data set to incorporate
experimental data of all kinds, including data for less than 24 hours. This will require further
experimental research and data collection.
Group 3 said that current default numbers for soil contact rates appear to be consistent with the
empirical observations of one of the group's members. These current default values appear
approximately right even though they appear to have been derived at least in part on inaccurate
assumptions. Therefore, Group 3 recommended that the workgroup not change the individual
parameters in their current absorption fraction approach for soil unless they plan on addressing
and overhauling their entire approach. Due to the magnitude of such an initiative, which would
have to incorporate factors such as time-dependent absorption kinetics and soil characteristics
32
-------
(e.g., particle size, aging, moisture content), Group 3 recommended that this project be placed in
the "Future" category.
Discussion Area 2
According to Group 3, the table's soil absorption values were biased and heavily dominated by in
vivo (e.g., Wester et al.) data. Such dependence on in vivo data may not be scientifically sound
and defensible, because these studies:
• Used particle sizes that were too large.
* Involved a pattern of contact that may or may not better correspond to real
exposures than in vitro methods.
• Did not age the contaminant on the soil
Discussion Area 3
Group 3 said that the proposed default absorption fraction of 10% for organic compounds in soil
is "good enough." They recognized that data to verify this default assumption are limited and
that more data are needed.
Discussion Area 4
Group 3 said that the proposed default absorption fraction of 1% for inorganic compounds in soil
seems conservative. Group 3 recognized that extremely limited data are available for inorganic
compounds, but that this default assumption probably overestimates exposure. They raised the
issue that not all metals are well absorbed into the body orally either, and asked how the
workgroup might account for gastrointestinal absorption adjustments when performing route-to-
route extrapolation.
Discussion Area 5
The general concept and proposed approach for the total body soil-to-skin AF received Group 3's
support. Group 3 said that the approach was generally good, but that they would like to see the
33
-------
model account for site-specific conditions. Briefly, Group 3 discussed hand-loading and other
variables that may need to be addressed in the model so that exposures are not unnecessarily
over- or underestimated.
General Discussion
Jim Bruckner raised the issue of factors influencing the dermal availability of chemicals in soil.
It was noted that absorbed doses are altered by the solubility of a chemical (e.g., for different
compounds of the same metal). The expert consultants also discussed how the dermal model
should account for sweating, the rate of transfer, soil dust adherence to skin, and contaminant
transfer in skin. Additionally, higher contamination levels generally exist in fine-particle soils
than in coarser material. Therefore, site-specific conditions are likely to influence dermal
absorption. Mr. Clewell raised the example of arsenic inhalation, saying that the composition of
a chemical in the environment significantly affects the chemical's absorption into the body.
Therefore, the state of a chemical and other site-specific issues are. relevant factors that should be
incorporated into the ABS values for dermal contact with soil. Dr. Leber suggested using total
body adherence factor (TBAF) extraction methodology to determine chemical availability. Most
peer consultants endorsed using TBAF methodologies, and agreed that if a chemical is not
available to the solvent then it will not be available to the skin.
Observer Comment
No observers commented on this topic.
34
-------
5. ADJUSTMENT OF TOXICITY FACTORS
TO REFLECT ABSORBED DOSE
5.1 Presentation
Mark Maddaloni from EPA Region 2 began the toxicity adjustment discussion by presenting an
equation: risk = dose x toxicity. He emphasized that in this equation, the units need to be
harmonious. Typically, dose and toxicity are represented as administered doses and experimental
data on dermal doses are not available. Therefore, the question arises of how to adjust toxicity
values derived from oral dosing studies to correspond to dermal exposures. The methodologies
described in the Superfund Dermal Guidance recommended that "adjustment of oral toxicity
values should be considered when characterizing the risk associated with the dermal exposure
route." The Superfund Dermal Guidance also provided a summary of some gastrointestinal
absorption efficiencies and a table of recommendations pertaining to whether adjustment of the
oral toxicity factor might be necessary. The apparent solution to this dilemma is to adjust the
toxicity factor to reflect the actual absorbed dose in agiven scenario. ; ' ; • :
Adjusting toxicity factors to reflect actual absorbed doses can be done using the following
equation: Doseabs = doseadm x Fractionabs. The workgroup endeavored to incorporate this " •
adjustment into their dermal guidance document. Using this equation, complete chemical
absorption (~ 100%) (i.e., the absorbed dose equals the administered dose) requires no toxicity
adjustment. Poor chemical absorption, however, means that the absorbed dose is relatively small
compared to the administered dose, thereby requiring a change in toxicity factor to accurately •
represent internal dose. /
To accurately incorporate toxicity adjustments into risk assessment guidance, the workgroup
must address two issues that may be problematic due to limited available data. First, the
workgroup must obtain absorption estimations from critical studies; second, the workgroup must
practically apply toxicity adjustments for risk assessments.
35
-------
Mr. Maddaloni defined a "critical study" as one that defines reference dose concentrations, and
therefore is the basis for deriving a toxicity factor. A critical study is essentially a toxicity
assessment based on a dose/response relationship that rarely accounts for bioavailability
determinations. Critical study results can vary depending on host characteristics and study
regimen (i.e., how the dose was administered). Most critical studies contain minimal, if any,
information about absorbed dose. Rather than redoing critical studies to obtain needed
experimental data, the workgroup performed an extensive literature review of chemical-specific
bioavailability studies. They focused their literature search on bioavailability studies that most
closely mimicked the methodologies and variables of critical studies, specifically similar host
characteristics (e.g., species, age, sex) and dosing regimens (e.g., route, vehicle, dosage). From
this information, the workgroup compiled a table of various compounds, their oral absorption
(bioavailability), and whether or not a toxicity factor adjustment was necessary. This table can
be found in the Superfund Dermal Guidance document.
As for the practical application of toxicity adjustments for risk assessments, the workgroup
recognized that, theoretically, a toxicity adjustment would be necessary whenever absorption in a
critical study was less than 100%. Because, however, critical studies have limited precision and
vary in their results and approaches, Mr. Maddaloni said, adjusting for toxicity factors when
analyzing chemicals with high absorption dose rates (e.g., 95%) may result in "lexicological hair
splitting." A high degree of uncertainty is often involved in assessing chemical bioavailability.
Therefore, using toxicity adjustments for high-absorption rate chemicals could imply a false level
of accuracy and precision regarding absorbed doses. For these reasons and other practical
purposes, the workgroup proposed a "50% rule": toxicity adjustments should only be applied
when absorption in a critical study was less than 50%. Mr. Maddaloni asked the expert
consultants to comment on the proposed 50% rule and its implications on policy decisions for
managing uncertainty.
After Mr. Maddaloni concluded his presentation, Mr, Clewell asked the consultants also to think
about toxicity factors in the context of:
36
-------
• Metabolic effects.
• Organic versus inorganic issues.
• Overestimating versus underestimating risks.
5.2 Discussion
Following Mr. Maddaloni's presentation, the peer consultants divided into three breakout groups
to discuss the adjustment of toxicity factors to reflect absorbed doses. Based on the charge
(Appendix A), the breakout groups focused their discussion on the following two topic areas:
In cases where the critical study, which forms the basis of the toxicity factor, hasn't
provided adequate information on oral absorption, the Superfund Dermal Guidance
attempts to identify appropriate bioavailability studies in the peer reviewed literature.
Such studies are reflected based on their resemblance to the critical study in terms of
dosing regimen (e.g., route, vehicle, and dosage) and host characteristics (e.g., species,
age, and sex), in order to provide sufficient information on the oral absorption of the
chemical in question. Comment on the factors considered in the selection of appropriate
bioavailability studies. Should other factors be considered? Also, comment on the studies
summarized in Table 4.1. Are the estimates of gastrointestinal absorption the best
available?
The information in Table 4.1 in Chapter 4 is then used to determine if an adjustment in
the toxicity factor is necessary to account for the difference in the estimated dose between
the oral and dermal routes. The Superfund Dermal Guidance recommends making
quantitative adjustments to toxicity factors only when there is evidence to indicate that
the oral absorption in the critical study was significantly less than complete. An oral
absorption fraction of 50% is recommended as the cut-off for this purpose. This avoids
making minor adjustments for chemicals that exhibit relatively efficient absorption (80-
90%). Further, the 50% cut-off is intended to reflect the inherent variability associated
with measuring bioavailability. Comment on the approach for deciding when to make
adjustments to the oral toxicity factor. Are these recommendations scientifically sound
and defensible? Is the rationale for selecting a 50% cut-off clear and transparent? Is the
cut-off estimate scientifically sound and defensible?
37
-------
Following the breakout discussions, the consultants reconvened in a plenary session. The
breakout group chairs summarized the discussions as follows. (Copies of the overheads used by
the chairs in making their presentations are included in Appendix F.)
Group 1
Chair: Annette Bunge, Colorado School of Mines
Discussion Area 1
Group 1 did not enthusiastically support the workgroup's approach of adjusting toxicity values
using selected bioavailability studies. Group 1 suggested that it might be better to use
gastrointestinal absorption factors for estimating oral bioavailability and creating a dermal
reference dose. Given current data constraints, gastrointestinal absorption factors might prove
more scientifically feasible for inorganic compounds. For organic compounds, Group 1
suggested, a reference dose based on experimental dermal data would be most preferable. In the
absence of such data, the workgroup should consider adjusting its toxicity factors based on
inhalation- or intravenous-pathway studies, provided that these data are available.
Discussion Area 2
Group 1 was not enthusiastic about the workgroup's 50% cut-off value, but they admitted that
50% was acceptable. The group said that the document's discussion of absorption factors should
consider and note first-pass metabolic transformations. In some instances, when metabolic
transformation factors are accounted for, dermal contact can be more toxic than oral ingestion for
the same dose (i.e., inactivation). In other instances, oral ingestion is more toxic than dermal
contact for the same dose (i.e., activation). The only way to assess such route-to-route
extrapolations is to consider metabolic transformations. Group 1 recommended to the
workgroup that they address this issue by including language in the document that says it is
appropriate to use a gastrointestinal absorption factor for chemicals with efficient absorption
rates (e.g., 80-100%), except for those chemicals with known or expected or possible first-pass
metabolic effects.
38
-------
When discussing route-to-route extrapolations, Group 1 said that the effective absorbed dose rate
(i.e., timing of exposure) must be taken into account by the workgroup?1 Specifically, the
workgroup should address whether or not the toxicity factor was derived from short- or long-term
ingestion studies, and whether or not these toxicity factors are being applied to short- qr long-
term dermal exposure perjods. The workgroup could address such issues by identifying ingestion
study methodologies (e.g.; short- or long-term gavage, feeding, or drinking studies) and dermal
exposure scenarios (e.g., short- or long-term shower, pica behavior, or occupational exposure).
Group 1 acknowledged that it is difficult to relate toxicity for one dosing procedure to another.
Single administered doses (e.g., injection) cannot often successfully be compared to divided
administered doses (e.g., feeding studies) because they fail to consider chemical .concentrations
per unit period of time. Therefore, dermal risk assessment is not only a matter of determining
exposure time, but also relating, this exposure to the time-frame and dose administration periods
involved in the critical study.
Lastly, Group 1 felt that the document needs to describe Table 4.1's toxicity factor data better. ;
Specifically, the workgroup should discuss the derivations and limitations qf such toxicity factor
data. Without such qualification of the table data, Group 1 anticipates that,ri;sk assessors may be
tempted to misuse the data. Group 1 suggested that the document include.the derivation qf the .
reference dose from the critical study, since the reference dose value is the basis.for
recommending adjustment for using dermal absorption fractions. , . .
Group 2
Chair: Gary Diamond, Syracuse Research Corporation
Discussion Area 1
Group 2 felt that the document needs to state explicitly, that the values for absorption factors
shown in Table 4.1 apply to the species and study design on which the RfD was based, and that
these absorption factors may not be applicable to humans (see Appendix G). Some Group 2
members felt that if better human data were available, they should be included in the table. All
39
-------
Group 2 members agreed that the table should include toxicity factors; the table should specify
whether their values are applicable to reference doses or cancer slope factors. Group 2 also noted
that values in the table indicate a level of precision that may not be realistic (e.g., ±0.1%). They
also said that the default assumption value of 50% ("other organics") may not apply to larger-
molecular-weight chemicals (e.g., petroleum hydrocarbons). One member of Group 2 suggested
that if EPA had better biomarkers of exposure, then the Agency would not need to compile such
a data table. Another consultant pointed out that the cadmium and nickel values in Table 4.1 do
not reflect the best available data on cadmium and nickel absorption fractions. This may have
resulted from the need to select estimates that best reflect the experimental design of the study on
which the RfD or cancer slope factor was based. The group also discussed how the workgroup
might best extrapolate oral-to-dermal exposures using Haber's assumption. Group 2 felt that the
document needs to consider dose-rate differences in the various studies and exposure scenarios.
Discussion Area 2
Group 2 discussed the "50% rule" recommended cut-off value for.toxicity factor adjustments.
Most Group 2 members endorsed the 50% rule, although there was some concern about how to
assess data to determine if 50% is met. Group 2 said that EPA needs to create an oral absorption
guidance document similar to the Agency's dermal guidance document. This document should
guide researchers and risk assessors on how to evaluate the quality and usefulness of dermal data.
This, in turn, will facilitate the incorporation of this data into risk assessment decisions.
Group 3
Chair: John Kissel, University of Washington
Discussion Area 1
Many of Group 3's responses reaffirmed points made by Group 1. Group 3 felt that all the
incorporated information in the table was appropriate, but noted that the document text did not
explicitly state the factors considered in the selection of the appropriate bioavailability studies.
Group 3 suggested that the workgroup include information about the selected studies and their
40
-------
resemblance to a critical study in terms of dosing regimen (e.g., route, vehicle, and dosage) and
host characteristics (e.g., species, age, and sex). Group 3 recommended that the workgroup
rework the document so that it explicitly states the importance of considering dosing regimens,
host characteristics, measures of bio-availability, and the potential for first-pass
activatipn/inactivation; all these factors are highly relevant to making route-to-route
extrapolations.
Discussion Area 2
Group 3 said that the workgroup's current approach for deciding when to make adjustments to
the oral toxicity factor, using an oral absorption fraction of 50%, may lead to a false sense of
confidence. Risk assessors may see this information, which is often based on very little
experimental data and contains a high degree of uncertainty, and believe that the toxicity factor
and oral absorption fractions are more precise measurements than they really are. This could
unwittingly lead risk assessors to misuse such information. Therefore, Group 3 felt that the
document should include additional language describing relevant uncertainties regarding the
adjustment of toxicity factors.
The 50% cut-off for adjusting a chemical's oral toxicity factor was not very popular with Group
3. They suggested that 80% may be a more reasonable cut-off point.
Group 3 was concerned that the workgroup's entire approach to bioavailability might not be
applicable to most chemicals. They emphasized that researchers cannot assess route-to-route
extrapolations if they do not have chemical-specific information, and currently much of these
data do not exist. Therefore, the document text needs to qualitatively address these uncertainties
about availability. Addressing these unknowns and uncertainty issues regarding toxicity factors
will help to put the dermal guidance in perspective, helping risk assessors to recognize any
potential imprecision in the data.
41
-------
Group 3 felt that acknowledging these uncertainties is especially important when it comes to
adjusting the toxicity factors of environmental levels of contaminants. The workgroup also felt
that acknowledging skin metabolic factors is important. Currently, numerous gaps exist in data
on the importance of skin metabolism of chemicals. For example, no one knows what happens
with the release of the metabolites (i.e., what happens to metabolites in the skin). The impact of
metabolites may be small, or it may be significant, but researchers currently have no
understanding of these processes due to limited data.
General Discussion
Mr. Clewell briefly reviewed the three groups' discussions. He began by mentioning the expert
consultants' desire that the document more thoroughly discuss and put into perspective the
studies and information about bioavailability. Mr. Clewell recapped the general acceptance of
using a 50% cut-off level for toxicity factor adjustment. He then asked the consultants to try to
distinguish "Today" versus "Tomorrow" issues regarding the adjustment of toxicity factors to
reflect an absorbed dose.
Dr. Bunge and other experts agreed that the "Today" category should include descriptive text for
Table 4.1 that clearly and completely shows what assumptions the workgroup used to derive the
table's toxicity factor values. Specifically, the document should clearly state which critical
studies were used, and how they were administered, so that people will not use the table's values
for purposes other than those originally intended by the workgroup.
One expert said that EPA's approach effectively removes toxicology from risk assessment in an
attempt to simplify the process. In the "Tomorrow" category, the experts hope that the
workgroup will be able to better incorporate toxicology (e.g., metabolism, human versus animal
factors) and epidemiology into their toxicity adjustment and dermal risk assessment processes.
42
-------
Dr. Frederick felt that assessing toxicity factors using data-from human studies is a practical way
to address oral bioavailability-and metabolic effects. Dr. Frederick alsp emphasized the
practicality of the 50% cut-off value, saying .that it is appealing because it will enable; researchers
to effectively ignore toxicity adjustments for organic chemicals and only, assign toxicity factor
adjustments to the inorganic, compounds. .... . ..-,.-.,
Mr. Clewell emphasized that to a large extent, many of the consultants' concerns, can.be
addressed just as EPA addressed local toxicity issues—by explicitly acknowledging the need to,,
consider the possibility of route-specific metabolism. Many of the constraints for route-to-route
extrapolation are also evident in traditional dosimetry approaches, that inherently incorporate
uncertainty into their species and route extrapolation assumptions. There was general agreement
that more data are needed for the oral-to-dermal extrapolation. Several expert consultants., ,..,,,
strongly recommended that the workgroup add text to the document that explicitly describes
when to use the toxicity values in this dermal guidance document table and when to.use other
(e.g., human) default values. This will help put the default values into their proper perspective,
so that people do not think that the dermal guidance document has 100%-accurate answers.
The expert consultants discussed other shortcomings of the toxicity adjustment approach,
specifically that it does not account for exposure times, first-pass metabolic effects, and site-
specific conditions of environmental availability. The consultants recognized that incorporating
all of these theoretical ideas, into a "nice" table or text discussion, would be complex. They felt,
however, that even if EPA is not sure how to resolve these issues, that the workgroup must at
least identify these data gaps: these unknowns are crucial to the overall dermal guidance
approach. Gary Diamond said that the EPA might be able to achieve this goal simply by creating
a fuller presentation of its species-specific and human absorption fraction estimates. Dr.
Diamond also suggested that the document should more explicitly state that its goal is to estimate
bioavailability in the critical study.
43
-------
All expert consultants agreed that the document needs to encourage data collection. They
discussed Haber's law (concentration x time = constant) and its implications of exposure time
versus bioavailability time. In general, for prolonged periods of exposure time, toxicity
approaches zero. However, the consultants agreed that a chemical's bioavailability may further
prolong exposure, thereby confounding results. Further difficulties arise when guidance is used
in multi-pathway assessment (e.g., when risk assessors are adding systemic dose received via the
dermal route to other routes of exposure to achieve a total dose of one or more chemicals in the
body). In these cases, risk assessors are not necessarily relying on a prolonged low-level dermal
exposure as that which is producing the critical toxic endpoint; such exposure usually adds to the
effects of exposures by other routes.
Observer Comment
No observers commented on this topic.
44
-------
6. RISK CHARACTERIZATION AND UNCERTAINTY
6.1 Presentation
EPA's 1997 Policy for Use of Probabilistic Analysis in Risk Assessment and 1995 Policy for
Risk Characterization call for greater clarity, transparency, reasonableness, and consistency in
Agency risk assessments. To address these objectives, all major uncertainties and an evaluation
of their influence on the outcome of the risk assessment should be discussed in a risk
characterization. The expert consultants discussed many of the major uncertainties identified in
the Superfund Dermal Guidance.
To begin their discussions about risk characterization and uncertainty in dermal risk assessment,
Anne-Marie Burke outlined the four steps of the, risk assessment process: hazard identification,
exposure assessment, dose-response assessment, and risk characterization. Ms. Burke discussed
EPA's policy for risk characterization, as defined in 1995, as striving to achieve "greater clarity,
transparency, reasonableness and consistency in Agency risk assessments." Essentially, EPA's
goal was to identify clearly where conclusions were based on science, where conclusions were
based on science policy, where default assumptions were used, and how these default
assumptions impacted the risk assessment process. To achieve these measures, EPA decided to:
• Discuss their assumptions and uncertainties and how they potentially influence the
outcome of risk assessment.
• Present several types of risk information.
— range of exposures (high-end, central tendency risk)
— sensitive subgroups
• Assist and improve communication between risk assessors and risk managers.
45
-------
Ms. Burke reminded the expert consultants of the charge questions. She also asked the
consultants to consider whether current data are robust enough to develop standard assumptions
for dermal risk characterization, and if so, what these standard assumptions might be. Then Ms.
Burke gave a brief overview of the elements of the four-step risk assessment process.
Hazard identification involves identifying a subset of chemicals detected which are most likely
to result in adverse health effects. During hazard identification, risk assessors consider
information about occurrence and distribution of chemicals in the environment; fate, mobility,
and persistence in the environment; chemical concentrations; and toxicity based on animal and/or
human studies. For the dermal-water pathway, EPA retains a chemical in risk assessment if the
dermal pathway contributes at least 10% of the dose from the oral route. For the dermal-soil
pathway, EPA retains a chemical in risk assessment if it has a high dermal absorption value (i.e.,
is readily bioavailable).
As defined in EPA's risk assessment process, exposure assessment determines the conditions
under which individuals could be exposed to contaminants and the doses exposed people could
receive. Ms. Burke reviewed the exposure assessment equations for dermal-water and dermal-
soil pathways.
Ms. Burke then addressed uncertainties associated with the exposure assessment step of dermal
risk assessment. Major uncertainties identified by EPA for the dermal-water pathway include:
Model for DAe
vent
Concentration term for water
Exposure time
46
-------
Ms. Burke .talked about EPA's kinetics of absorption uncertainty assumptions. She stressed that
the Agency needs a more formal process to address uncertainties involved in the dermal risk
assessment process.
Major uncertainties identified by EPA for the dermal-soil pathway include:
• AF
• Concentration term for soil
• Dermal-soil absorption values ;
» Model for DAevent
• Frequency
• Surface area '
• Default absorption values for classes of chemicals
Ms. Burke restated the conclusion, drawn from prior consultant discussions, that most experts at
the meeting accepted the 10% default value for organic compounds in soil. The expert
consultants were not, however, very confident in the 1% default value for inorganic compounds
in soil.
Dose-response assessment (the third step of the risk assessment process) involves evaluating
toxicity information and characterizing the relationship between the dose of the contaminant
received with the incidence of adverse health effects in the exposed population. Dose-response
assessments develop chemical-specific reference doses and slope factors. Ms. Burke noted that
the same approach for these assessments is used for dermal-water and dermal-soil pathways.
Uncertainties associated with the dose-response step of dermal risk assessment include:
The lack of reference doses and cancer slope factors specific for the dermal
pathway.
47
-------
• The lack of dermal slope factor for cPAHs (most slope factor information is
derived from direct dermal application).
For risk characterization, the fourth stage of the dermal risk assessment process, Ms. Burke
reviewed the equations for cancer endpoints (excess cancer risk) and noncancer endpoints
(hazard quotient). She pointed out several assumptions and uncertainties regarding risk
characterization, including the lack of information on gastrointestinal absorption and the lack of
information about toxicity at the skin surface.
Ms. Burke concluded her presentation by asking the expert consultants to consider whether or not
it seems reasonable that the dermal pathway can drive the risk assessment process in certain
circumstances.
6.2 Discussion
Following Ms. Burke's presentation, the peer consultants divided into three breakout groups to
discuss risk characterization and uncertainty issues. Based on the charge (Appendix A), the
breakout groups focused their discussion on the following four topic areas:
1. Have the factors which make the most significant contribution to uncertainty been
identified in this guidance? Is the discussion of the uncertainties complete?
2. How should these uncertainties be characterized in a dermal risk assessment in order to
effectively communicate the results to risk managers and the public?
3. Using the default assumptions in this guidance, the estimated risks associated with dermal
exposure are often greater than the risks for the ingestion or inhalation routes, particularly
for contaminants in soil. How does the magnitude of the uncertainty for estimating
dermal risks compare to the uncertainty for these other routes of exposure? How should
this information be used to characterize the uncertainty for the dermal route?
4. How can the magnitude of the uncertainties be reduced in order to improve the overall
quality of risk assessment?
48
-------
Following the breakout discussions, the consultants reconvened in a plenary session. The
breakout group chairs summarized the discussions as follows. (Copies of the overheads used by
the chairs in making their presentations are included in Appendix F.)
Group 1
Chair: Annette Bunge, Colorado School of Mines
Discussion Area 1
Group 1 emphasized that more data, especially soil data, are needed to reduce uncertainties in the
dermal guidance document. Group 1 said that a 10% default value for organic compounds in soil
was a threshold starting place, but that much more research is needed. The group mentioned
studies on 4-cyanophenol, as evidence that absorption from soils can exceed the 10% default.
Group 1 felt that the guidance document identified all important uncertainties, but recommended
that the document discuss some uncertainties in greater detail. Group 1 also suggested that the
ionization of chemicals in water not necessarily be regarded as an uncertainty in the context of
this dermal guidance document. They also raised questions about the impact of pH factors on the
dermal risk assessment process.
Discussion Area 2
Addressing the characterization of uncertainty, Group 1 suggested that the workgroup consider
labeling uncertainty with a numerical order of magnitude value (e.g., 2-, 10-, 100-fold).
Representing uncertainty with order of magnitude values may help to put the relative importance
of various factors into perspective. In other words, it will be easier for risk assessors to
determine which uncertainties are more likely to be more significant and which are likely to be
less significant in the dermal risk assessment process. Group 1 recommended that the workgroup
consider categorizing uncertainties by:
Exposed dose
Absorbed dose
49
-------
a Kp, tcxp, Absorption Factor
« Dermal to oral (species) extrapolation
Discussion Area 3
Group 1 said that it may sometimes be possible for dermal risks to exceed oral or inhalation
exposure risks, but more research needs to be done to identify chemicals and/or chemical classes
for which this makes sense. Some consultants in Group 1 suggested that dermal risks may
outweigh other exposure pathway risks for high-molecular-weight chemicals such as
hydrocarbons.
Discussion Area 4
Group 1 felt that the magnitude of uncertainties can be reduced by more data collection.
Specifically, Group 1 would like to see more dermal data generated for chemicals in soil and for
high-molecular-weight chemicals in water.
Group 2
Chair: Gary Diamond, Syracuse Research Corporation
Discussion Area 1
Group 2 felt that variables that could be represented as reflecting variability include: shower
time, surface area, and absorption fractions. Available data are not adequate to support
distributions of KpS for individual chemicals.
According to Group 2, the dermal guidance document identified several tools for characterizing
uncertainty:
• Conducting sensitivity analysis (ranking of importance)
» Identifying research needs
50
-------
• Identifying what distributions are needed
Group 2 identified other uncertainties, including temperature effects, aging of skin, concentration
terms, summing across pathways, different absorption dose rates, pathway-specific toxicologic
targets, and K,, prediction uncertainty. Group 2 felt that the workgroup identified these
uncertainties in the guidance and that the next step was to quantify these uncertainties to
determine which factors are most important.
Discussion Area 2
To effectively communicate uncertainty regarding dermal risk assessment, Group 2
t '.f ' •' , ' •
recommended that the workgroup quantitatively identify the uncertainties by level of importance.
Group 2 also raised the question of how the collection of site-specific data might reduce
uncertainty, but did not draw specific conclusions or make recommendations regarding site-
specific data collection. Group 2 recommended that the workgroup consider addressing
uncertainty with an order of magnitude or range value, saying that sometimes a higher confidence
can be achieved with a distribution than a point estimate.
• - ' • t • • . • ' =- • • ; v : • '
Discussion Area 3
Group 2 said that if the dermal risk drives soil pathway risk assessments, then the model and
default assumptions for estimating risks are probably flawed. Currently, data do not exist to
support the conclusion that health effects at Superfund sites are associated with the dermal
pathway. Group 2 suggested that it would be good to have a dermal reference dose for a select
few problem chemicals (e.g., PAHs, organic chlorine pesticides). Such chemical-specific data
could serve as benchmarks that may enable researchers and risk assessors to determine whether
or not dermal pathway exposure is a problem.
Discussion Area 4
To improve the overall quality of dermal risk assessment and reduce uncertainty, Group 2
recommended that the workgroup focus on evaluating and validating models. The workgroup
51
-------
should more thoroughly evaluate the types of studies used to create the models. The workgroup
might also consider comparing existing data using biomarkers of exposure (e.g., biological
exposure index [BEI]) with estimated dermal doses. Group 2 suggested that the workgroup
apply models for estimating KpS and dermal doses to Pharmaceuticals. For example, the
workgroup could evaluate the KpS predictive model and dermal dose model by conducting
relatively simple nicotine or nitroglycerin human studies. Human studies such as these will
enable the workgroup to compare their predicted dermal doses with in vivo estimated dermal
doses, which will in turn help them to evaluate and validate their models. Group 2 felt that the
current validation studies cited in the dermal guidance document are not adequate to support a
confidence statement.
The weakest part of the model's variable distributions was the lack of soil data. The group
agreed that information gathering and generating should be a priority to improve the overall
quality of dermal risk assessment. They said that the workgroup might want to approach data
gathering with a probabilistic-distribution mind set. Even if these.data are of poor quality, it is
highly probable that researchers will obtain more information from scientifically unsound
probability distributions than they would from scientifically unsound single-point estimates.
Group 3
Chair: John Kissel, University of Washington
i
Discussion Area 1
Overall, Group 3 felt that the dermal guidance document addressed most of the factors that
significantly contribute to uncertainty. Group 3 distinguished between identifying the
uncertainties and emphasizing the uncertainties in order to put dermal risk assessment
uncertainties into perspective. The expert panelists felt that the document should emphasize
some of the identified uncertainties over others. They suggested that the workgroup place
particular emphasis on uncertainties regarding route-to-route extrapolations and dermal
52
-------
absorption of inorganics from soil. For example, the document should discuss route-to-route
extrapolations and toxicity factor uncertainties in its introduction of Section 5.2.
Discussion Area 2
Group 3 suggested qualitatively ranking uncertainties based on Agency confidence levels, using
"low," "medium," and "high" ratings of parameter confidence. The experts emphasized
communicating "comparative" risk estimates to risk managers and the public.
Discussion Area 3
Group 3 said that the magnitude of uncertainty regarding dermal exposure is greater and more
problematic than other exposure-route uncertainties. The two uncertainties highlighted by Group
3 were route-to-route extrapolations and exposure estimates for inorganics from soil. Group 3
suggested that the workgroup consider adding monolayer adjustment (with an appropriate
explanation of how and when to use it) in the document text. Group 3 also felt that the dermal
guidance needs more discussion and recommendations regarding sediments and sediment
exposure. Group 3 discussed replenishment of contaminated soil on the skin, concluding that it
was another uncertainty area that the workgroup might want to address.
Discussion Area 4
Group 3 felt that more data acquisition will lower the magnitude of dermal exposure
uncertainties and improve the overall quality of risk assessments. Group 3 recommended •
creating standard protocols for dermal studies. Consultants in Group 3 also suggested that the
workgroup search for opportunities to validate their assumptions and approaches used in the
dermal models and guidance.
General Discussion
Mr. Clewell summarized the group's discussions about dermal risk characterization and
uncertainty. He mentioned factors such as the oral toxicity factors, exclusion of skin toxicity,
53
-------
predicted KpS versus experimental data, the lack of soil data, and other dermal data gaps and
uncertainties. The expert consultants briefly discussed the benefits and costs of conducting in
vitro versus in vivo experiments for dermal data generation. Mr. Clewell said that the most
important variable in chemical availability appears to be chemical form (composition); the
consultants agreed that chemical form and compound variations are more important than site-
specific soil conditions.
The panelists discussed the implications of the fact that there are so many data gaps; they
recognized that for many risk assessors, the level of uncertainty regarding dermal exposure to
inorganics in soil is so high that pathway is often not included in the risk assessment process.
For example, the State of New Jersey does not examine the dermal pathway as part of its risk
assessment process, primarily because the State does not want to let dermal uncertainty drive risk
management. Several expert consultants expressed concern about functionally ignoring dermal
exposure. This concern generally focused on dermal exposure to organic chemicals.
Dr. Frederick said that he is not overly concerned with dermal exposure to inorganic chemicals
because the metal ions do not readily pass through soil into skin. Paul Chrostowski emphasized
that all chemicals have an octanol/water partition coefficient, even those that do not readily pass
through the skin. Several panelists said that they do not feel comfortable dismissing the dermal
pathway for inorganics, even though they have doubts that such exposure is significant. Dr.
Chrostowski and Rosalind Schoof summarized the conversation by saying that the object of this
dermal guidance is to get good data, even if the values for exposure are very low. Therefore, the
goal of the workgroup should be to encourage data generation. The consultants recognized that
the impetus to research dermal exposure can be enhanced with Agency encouragement and
support—for example, if EPA were to integrate new research into their dermal guidance
document. Several consultants also recommended developing a protocol for site-specific data
generation in order to facilitate and enhance the integration of dermal exposure pathway data and
to reduce scientific uncertainty.
54
-------
The expert consultants recommended that EPA keep the derrnal guidance workgroup as a
standing working group to provide technical information when needed.
Several panelists endorsed the concept of using the Internet to work toward consistency for the
national point of view for dermal risk assessment guidance. An Internet webpage could be
created to describe the workgroup. One panelists suggested routing the dermal workgroup
webpage through the Risk Assessment Forum. Consultants strongly endorsed creating such
Internet links for the dermal workgroup, including a dermal guidance database that would be
regularly updated with the most recently available data (similar to EPA's IRIS website).
Observer Comment
Garrett Keating, of the Lawrence Livermore Lab, noted that the words "more data" appeared on
nearly every overhead and asked what kind of data, specifically, is needed. Mr. Keating said that
risk assessors need guidance in the dermal-soil pathway. Currently, virtually any dermal-soil
protocol is considered equally acceptable. Regarding dermal-water exposure, there are .,
potentially two kinds of data: steady-state data (the model needs improving) and validation data
' ' ' • ' • • • '•• ' ••'•.•.'•'.'.'..' •• ii. , ••••.' it
(the model needs testing). Mr. Keating asked the consultants to,address these issues while the
document is being developed. He also recommended that, as more in vivo data become
available, EPA consider ways to incorporate the new information in the document.
55
-------
7. NEXT STEPS: PLENARY DISCUSSION ON
DERMAL EXPOSURE ISSUES
Before concluding the two-day meeting, Mr. Clewell asked the expert consultants to prioritize
the key points they want to convey to EPA, ranking them using three levels: "Today" (short-
term), "Tomorrow" (medium-term), and "Future" (long-term).
"Today"
In the short term, the expert consultants would like to see the following points incorporated into
and/or more fully addressed in the Superfund Dermal Guidance document:
• Clarify the derivation of the 95% CI for Kp calibration.
a Clarify the basis and derivation of EPD boundaries.
• Clarify the document's text, especially the discussions involving uncertainties.
• Address the fact that the document "underpredicts halogenated chemicals"—
consider adjustment based on density.
• Insert the monolayer correction into the document (where appropriate).
• Insert an acknowledgment and citation of the reference (Gerrity and Henry, 1990)
for the preferred route-to-route extrapolation.
• Perform sensitivity/uncertainty analysis and rank the uncertainties using
qualitative/quantitative measurements (e.g., low, medium, high).
• Define, illustrate, and bound how to calculate lag time (T,).
• Define recreational exposure assumptions (including sediment and surface water
recreational criteria).
• Establish the workgroup as a standing dermal workgroup (with funding).
56
-------
• Acknowledge in the document that there may be a loss of VOCs during shower
scenarios because the water chemical concentrations lessen when the water
contacts air.
• Add vehicle to toxicity factor table.
"Tomorrow"
1 *•'
In the early stages of the "tomorrow" medium-term period, meaning in the next couple of
months, the consultants hope to see the workgroup:
• Perform a regression analysis for Kp using the Vecchia database.
• Perform the regression for Kp using Kow and molar volume rather than K and
MW.
After this is done, the consultants recommend that the workgroup:
Perform the regression for Kp using molecular substructures and compare with
current method.
Continue ranking uncertainties using sensitivity/uncertainty analysis of
parameters.
Establish a standard resource depository and clearinghouse (e.g., a website) for
reviewed Kp values and other important dermal data to encourage consistency.
Develop standard criteria for dermal-soil exposure protocols and Kp
determinations that allow risk assessors to use chemical-specific experimental
(and soil AFs) instead of predicted KpS. Criteria should be developed for both
retrospective and prospective use in evaluating and planning.
— in vivo protocols
— in vitro protocols
Refine soil adherence estimates using Dr. Kissel's information.
57
-------
Address transfer factors, including dermal exposure estimates from contact with
concrete, utility poles, and other surfaces.
Incorporate pesticide deposition and absorption (neat vs. aqueous) issues.
Create an Agency Dermal Working Group (with funding) to address dermal issues
spanning all EPA divisions.
Transfer and integrate dermal data across EPA programs.
Future
Beginning in the next millennium and continuing far into the future, the consultants hope to see
the workgroup and other dermal researchers:
• Generate more dermal data, especially for inorganics from soil. This data should
include information about specific chemical forms and soil types.
• Collect data, for priority chemicals, on toxicity (portal-of-entry effects). When
justified, derive dermal reference doses considering both systemic toxicity effects
and portal-of-entry type effects.
• Focus on in vitro dermal-soil studies, then progress to in vivo studies: develop in
vitro (and later in vivo) study protocols, then use the generated in vitro data to
learn how dermal exposure is impacted by soil variability, chemical form, and
other variables. Expert consultants noted that this may be difficult because broad
generalizations may not be accurate and risk assessors may need site-specific data
to properly assess dermal exposure issues.
• Use in vitro dermal studies to create and/or refine protocols that risk assessors and
researchers should use in addressing various classes of chemicals (e.g., inorganics,
SVOCs, pesticides, persistent chemicals). In vitro studies may also be used to
develop screening tests for risk assessors. The panelists noted that the protocols
and screening tests may vary by chemical class.
• Conduct human in vivo studies with internal biomarkers. For example, look at
dermal exposure to benzopyrene or nicotine in order to correlate environmental
exposure with measurable systemic exposure. This will help to compare modeled
versus experimental data.
58
-------
• Develop more complete mathematical models (pharmacokinetic and kinetic) for
extrapolating to human from animal-study data. Use human studies with
butoxyethanol or isopropanol for validation.
• Address how risk assessors can assess the influence of chemical mixtures in the
dermal pathway.
• Examine chemical disposition in different tissues and the ensuing effects of
chemicals on various target tissues.
• More thoroughly research the effects of skin metabolism, skin reservoir effects,
and chemical metabolism in skin.
After completing the "Today," "Tomorrow," and "Future" lists, Mr. Clewell adjourned the
meeting by thanking the expert consultants, the EPA workgroup, and observers for their
participation in the Risk Assessment Forum's Workshop on Issues Associated with Dermal
Exposure and Uptake.
59
-------
-------
APPENDIX A
CHARGE TO PEER CONSULTANTS
A-l
-------
-------
Workshop on Issues Associated with Dermal Exposure and Uptake
U.S. Environmental Protection Agency
Washington, D.C.
December 1998
Charge to Experts/Discussion Issues
This workshop is being held to discuss issues associated with estimating dermal exposure
and uptake of environmental contaminants. The workshop discussions will focus on generic
technical issues raised during the February 1998 peer review of the Risk Assessment Guidance
for Superfund, Supplemental Guidance, Dermal Risk Assessment (hereafter lenpwn as the
.Superfund Dermal Guidance). Although the workshop issues were derived from the review pf a
proposed Superfund model, they are genetically applicable to the estimation of chemical uptake
within many U.S. Environmental Protection Agency (EPA) programs. Therefore, discussion of
these issues at the workshop should not be limited to the context of the Superfund Dermal
Guidance.
Background
In January 1992, the EPA Office of Health and Environmental Assessment (now ttyp
National Center for Environmental Assessment, NCEA) completed an interim report entitled
Dermal Exposure Assessment: Principles and Applications: This report provides guidance for
conducting dermal exposure and risk assessments. The conclusions of this report were
summarized at the January 1992 National Superfund Risk Assessors' Conference. During this
meeting, Regional risk assessors requested that a workgroup be formed to prepare an interim
dermal risk assessment guidance for the Superfund program. The purpose of this guidance would
be to promote consistency in the procedures used by the EPA Regions to assess risks from
dermal .exposure at Superfund sites. In August 1992, a draft Superfund Dermal Guidance was
circulated for review and comment.
In 1995, a workgroup convened to address issues related to the August 1992 Superfund
Dermal Guidance and to redraft the document. The revised guidance was peer reviewed in
February 1998. Several issues related to dermal exposure and risk assessment were raised during
the peer review. The workgroup addressed some of these issues in a revised draft of the
guidance. Other issues raised during the peer review were broader in scope. The EPA Risk
Assessment Forum is organizing the present workshop to discuss some of these broader, more
generic issues.
A-3
-------
Discussion Issues
The generic technical issues identified during the February 1998 peer review of the
Superfund Dermal Guidance can be organized into four categories: issues associated with dermal
exposure to contaminants in water, issues associated with dermal exposure to contaminants in
soil, issues associated with the adjustment of toxicity factors to reflect absorbed dose, and issues
related to risk characterization and uncertainty analysis for dermal assessments. These issues
will be the focal point for discussions during this workshop. For each category, workshop
participants are referred to specific sections of the November 1998 draft of the Superfund Dermal
Guidance for background and technical details.
The questions within each category are intended to help structure and guide the workshop
discussions. In addressing these questions, workshop participants are asked to consider: what do
we know today that can be applied to answering the question or providing additional guidance on
the topic; what short term studies could be conducted to answer the question or provide
additional guidance; and what longer term research may be needed to answer the question or
provide additional guidance.
Dermal Exposure to Contaminants in Water
Proposed approaches for estimating dermal exposure and uptake of chemical
contaminants in water are presented in Chapter 3 of the November 1998 draft Superfund Dermal
Guidance. In the approach for organic chemicals, a skin permeability coefficient (Kp) is
estimated as a function of a chemical's octanol/water partition coefficient (Kow) and its molecular
weight (MW). The relationship is based on a regression analysis of measured skin
permeabilities. For metals and inorganic chemicals, the Superfund Dermal Guidance
recommends using measured Kp values. In the absence of measured values, a default of 0.001
cm/hr is recommended.
1. Comment on the correlation equation used to estimate the skin permeability
coefficient (Kp) for organic chemicals. Is the approach used to estimate the Kp
values and their 95% confidence intervals plausible? Include in the discussion
consideration of the database analyzed to generate the correlation equation.
2. Comment on the statistical analysis used to establish the Effective Predictive
Domain for the Kp correlation equation (i.e., the range of Kow and MW where the
predictive power of the regression equation would be valid). Evaluate the new
methodology for calculating Kpimax for chemicals outside of the Effective
Predictive Domain.
3. Comment on the use of Kp and Kp>max in the dermal absorption model (specifically
the use of Kp for all tevent (exposure time) < t* (time to reach steady state
absorption), and the use of Kp or Kp_max when tevent > t*).
A-4
-------
6.
Comment on the use of predicted Kp or Kpimax vs. chemical specific experimental
values. Consider the criteria used to select studies to develop the regression
model (see Appendix A of the Superfund Dermal Guidance). Should these and
other criteria be used to judge chemical specific experimental values? What are
the minimum criteria that should be satisfied before chemical specific
experimental values can be used in lieu of model predictions?
Comment on the approach recommended for metals and inorganic chemicals. Is
the default Kp (0.001 cm/hr), that was previously recommended in the 1992
Interim Guidance for Dermal Exposure Assessment, still scientifically sound and
defensible?
Comment on the other default exposure assumptions (see Table 3.2)
recommended to estimate the Dermal Absorbed Dose per event (DAem!f) (e.g.,
tevau = 10 minutes for exposure in a shower). Are these defaults.scientifically
sound and defensible?
Dermal Exposure to Contaminants in Soil
Recommendations for estimating dermal exposure and uptake of chemical contaminants
in soil are presented in Chapter 3 of the Superfund Dermal Guidance. The approach is very
briefly summarized in this chapter and the reader is referred to Chapter 6 of the 1992 Dermal
Exposure Assessment: Principles and Applications for details on the methodology. Table 3.4 in
Chapter 3 lists recommended absorption fractions for two metals and eight organic
compounds/classes of compounds. These estimates are based on data. Table 3.4 also lists two
default absorption fractions to be used in the absence of measurements. For semivolatile organic
compounds, a default absorption fraction of 10% is recommended. For inorganic chemicals, a
default of 1% is recommended.
Estimates of soil-to-skin adherence must be used with estimates of absorption fraction
from soil to calculate the Dermal Absorbed Dose per event (DACTeJ. Recommendations on the
use of soil-to-skin adherence factors also are presented in Chapter 3 of the Superfund Dermal
Guidance. Table 3.3 in Chapter 3 list some soil-to-skin adherence factors by activity and soil
moisture content.
1. Discuss the current absorption fraction approach as applied to the dermal
absorption of chemical contaminants from soil. Consider such factors as the
duration of soil contact, the soil particle size, and the level of soil moisture and
whether these factors should be used to adjust the absorption estimate. Overall, is
the proposed methodology scientifically sound and defensible?
A-5
-------
2. Comment on the soil absorption values presented in Table 3.4 of the Superfund
Dermal Guidance. Are these estimates supported by the available data? Are they
scientifically sound and defensible?
3. Comment on the proposed default absorption fraction of 10% for organic
compounds hi soil. Is the rationale for selecting this default clear and transparent?
Is the estimate scientifically sound and defensible? Is their enough supporting
evidence to allow this estimate to be characterized as representative of the
average?
4. Comment on the proposed default absorption fraction of 1% for inorganic
chemicals hi soil. Is the rationale for selecting this default clear and transparent?
Is the estimate scientifically sound and defensible? Is their enough supporting
evidence to allow this estimate to be characterized as representative of the
average?
5. Comment on the proposed approach to calculate a total body soil-to-skin
adherence factor based on the surface area weighted adherence factors for each
body part. Is this a scientifically sound approach? Could other methodologies be
recommended?
Adjustment of Toxicity Factors to Reflect Absorbed Dose
Chapter 4 of the Superfund Dermal Guidance provides a discussion on adjusting toxicity
values derived from oral dosing studies. The methodologies described hi the Superfund Dermal
Guidance attempt to estimate absorbed or internal dose following dermal exposure. However,
many toxicity factors (such as the cancer slope factors and RfDs reported hi IRIS) are derived
from an administered oral dose. Therefore, it is recommended in Chapter 4 that "adjustment of
oral toxicity values should be considered when characterizing the risk associated with the dermal
exposure route." Table 4.1 hi Chapter 4 provides a summary of some gastrointestinal absorption
efficiencies that are available in the published literature. Also provided in the table are
recommendations pertaining to whether adjustment of the oral toxicity factor would be
necessary.
1. In cases where the critical study, which forms the basis of a toxicity factor, hasn't
provided adequate information on oral absorption, the Superfund Dermal
Guidance attempts to identify appropriate bioavailability studies in the peer
reviewed literature. Such studies are selected based on their resemblance to the
critical study in terms of dosing regimen (e.g., route, vehicle, and dosage) and host
characteristics (e.g., species, age, and sex), hi order to provide sufficient
information on the oral absorption of the chemical in question. Comment on the
factors considered hi the selection of appropriate bioavailability studies. Should
A-6
-------
2.
other factors be considered? Also, comment on the studies summarized in Table
4.1. Are the estimates of gastrointestinal absorption the best available?
The information in Table,4.1 in Chapter 4 is then used to determine if an
adjustment in the tpxicity factor is necessary to account for the difference in the
estimated dose between the oral and dermal,routes. The Superfund Dermal
Guidance.recommends making quantitative.adjustments to toxicity factors only
when there.is evidence to.indicate that the oral absorption in the critical study was
significantly less than complete. An oral absorption fraction of 50% is
recommended as the cut-off for this purpose. This avoids making minor
adjustments for chemicals that exhibit relatively efficient absorption (80-90%).
Further, the 50% cut-off is intended to reflect the inherent; variability associated
with measuring bioavailability. Comment on the approach for deciding when to
make adjustments to the oral toxicity factor. Are these recommendations
scientifically sound and defensible? Is the rationale for selecting a 50% cut-off
clear and transparent? Is the cut-off estimate scientifically sound and defensible?
Risk Characterization and Uncertainty
The EPA's 1997 Policy for Use of Probabilistic Analysis in Risk Assessment and 1995
Policy for Risk Characterization call for greater clarity, transparency, reasonableness, and
consistency in Agency risk assessments. To address these objectives, all major uncertainties and
an evaluation of their influence on the outcome of the risk assessment should be discussed in a
risk characterization. Some of the major uncertainties in dermal risk assessments are identified
in Chapter 5 of the Superfund Dermal Guidance. These include:
the reliance on adjusted oral toxicity factors to estimate toxicity from the dermal route of
exposure;
the exclusion of toxic effects at the skin surface in the risk assessment;
the recommended use of permeability coefficients for water that are based on model
predictions, rather than measured values;
the lack of quantitative information for the dermal absorption of chemicals in soil; and
variability and uncertainty in dermal exposure parameters for soil contact, such as skin
surface area exposed, soil-to-skin adherence, and frequency of exposure.
1) Have the factors which make the most significant contribution to uncertainty
been identified in this guidance? Is the discussion of the uncertainties complete?
A.-1
-------
2) How should these uncertainties be characterized in a dermal risk assessment in
order to effectively communicate the results to risk managers and the public?
3) Using the default assumptions in this guidance, the estimated risks associated with
dermal exposure are often greater than the risks for the ingestion or inhalation
routes, particularly for contaminants in soil. How does the magnitude of the
uncertainty for estimating dermal risks compare to the uncertainty for these other
routes of exposure? How should this information be used to characterize the
uncertainty for the dermal route?
4) How can the magnitude of these uncertainties be reduced in order to improve the
overall quality of the risk assessment?
A-8
-------
APPENDIX B
LIST OF PEER CONSULTANTS
B-l
-------
-------
&EPA
United States
Environmental Protection Agency
Risk Assessment Forum
Workshop on Issues Associated with
Dermal Exposure and Uptake
Bethesda Ramada
Bethesda, MD
December 10-11,1998
Peer Consultants
James Bruckner
Professor of Pharmacology & Toxicology
Department of Pharmaceutical
& Biomedical Sciences
College of Pharmacy
University of Georgia
Athens, GA 30602-2352
706-542-5405
Fax: 706-542-3398
E-mail: bruckner@merc.rx.uga.edu
Annette Bunge
Professor
Department of Chemical
Engineering & Petroleum Refining
Colorado School of Mines
Golden, CO 80401
303-273-3722
Fax:303-273-3730
E-mail: abunge@mines.edu
Paul Chrostowski
Director
The Weinberg Group, Inc.
1220 19th Street, NW - Suite 300
Washington, D.C. 20036
202-833-8077
Fax:202-833-4157
E-mail: pach@weinberggroup.com
Harvey Clewell
Senior Project Manager
K.S. Crump Division
ICF Kaiser International
602 East Georgia Avenue
Ruston, LA 71270
318-255-4800
Fax:318-255-4960
E-mail: hclewell@linknet.net
Gary Diamond
Senior Scientist
Environmental Science Center
Syracuse Research Corporation
6225 Running Ridge Road
North Syracuse, NY 13212
315-452-8409
Fax:315-452-8440
E-mail: diamond@syrres.com
Stephen DiZio
Senior Toxicologist
California Department of Toxic Substances Control
400 P Street - 4th Floor
Sacramento, CA 95812
916-327-2517
Fax:916-327-2509
E-mail: stevedizio@aol.com
B-3
-------
Robert Duff
Public Health Advisor
Off ice of Toxic Substances
New Market Industrial Campus
Washington State Department of Health
7171 Cleanwater Lane - Building 3
Tumwater, WA 98501
360-236-3371
Fax: 360-236-3383
E-mail: md0303@doh.wa.gov
Deborah Edwards
Senior Staff Toxicologist
Exxon Biomedical Sciences, Inc.
Mettlers Road (CN2350)
East Millstone, N J 08875-2350
732-873-6300
Fax: 732-873-6009
E-mail: daedwar@fpe.erenj.com
Kurt Enslein
Vice President of Technology
Health Designs, Inc.
183 East Main Street
Rochester, NY 14604
716-546-1464, Ext.: 23
Fax:716-546-3411
E-mail: kenslein@oxmol.com .
Clay Frederick
Senior Research Fellow in Biochemical Toxicology
Toxicology Department
Rohm and Haas Company
727 Norristown Road
Spring House, PA 19477-0904
215-641-7496
Fax:215-619-1621
E-mail: cfrederick@rohmhaas.com
John Kissel
Associate Professor
Department of Environmental Health
University of Washington
1705 Pacific Street, NE - Health Sciences (E179A)
Seattle, WA 98105
206-543-5111
Fax:206-543-8123
E-mail: jkissel@u.washington.edu
James Knaak
Assistant Adjunct Professor
Department of Pharmacology & Toxicology
State University of New York at Buffalo
3435 Main Street -102 Farber Hall
Buffalo, NY 14214-3000
716-688-5690
Fax:716-829-2801
E-mail: jbknaak@aol.com
Philip Leber
Project Manager
The Goodyear Tire & Rubber Company
1485 East Archwood Avenue
Akron, OH 44306
330-796-1046
Fax:330-796-1069
E-mail: apleber@goodyear.com
Gerhard Raabe
Director, Health Risk Assessment
Mobil Global Medical Services
2215 Aquetong Road
New Hope, PA 18938
215-862-5718
Fax:215-862-3551
E-mail: gerhard_k_raabe@ email.mobil.com
Rosalind Schoof
Principal
Exponent
15375 Southeast 30th Place - Suite 250
Bellevue.WA 98007
425-643-9803
Fax: 425-643-9827
E-mail: schoofr@exponent.com
Clint Skinner
President
Skinner Associates
3985 Shooting Star Road
Creston, CA 93432
805-238-1096
Fax:805-238-1228
E-mail: skinal ©tcsn.net
Val Schaeffer
Department of Labor
Occupational Safety and Health Administration
Health Standards Programs
200 Constitution Ave, NW
RoomN3718
Washington, DC 20210
Phone: 202-693-2092 X32279
Lawrence Sirinek
Environmental Scientist
Ohio Environmental Protection Agency
Lazarus Government Center
122 South Front Street
Columbus, Ohio 43215
614-644-2902
Fax:614-485-1636
E-mail: larry.sirinek@epa.state.oh.us
B-4
-------
APPENDIX C
LIST OF OBSERVERS
C-l
-------
-------
&EPA
United States
Environmental Protection Agency
Risk Assessment Forum
Workshop on Issues Associated with
Dermal Exposure and Uptake
Bethesda Ramada
Bethesda, MD
December 10-11,1908
Final List of Observers
Robert Bellies
Toxicologist
National Center for Environmental Assessment
U.S. Environmental Protection Agency
401 M Street, SW (8623)
Washington, DC 20460
202-564-3273 , . .
Fax:202-564-0049
E-mail: beliles.robert@epa.gov .
Robert Benson
Toxicologist ,
Municipal Systems ,
OPRA Division
U.S. Environmental Protection Agency
999 18th Street - Suite 500 (8P-W-MS)
Denver, CO 80202-2466 ; '
303-312-7070 , ,
Fax:303-312-6131
E-mail: benson.bob@epamail.epa.gov
Christopher Carroll
Industrial Hygenist :
U.S. Army Center for Health & Preventive Medicine
(Attn: MCHB-TS-OFS)
5158 Blackhawk Road
Aberdeen Proving Ground, MD 21010
410-436-5465
Fax:410-436-5471 '";
E-mail: christepher_carrol@chppm-
ccmail.apgea.army.mil ..',''
Hsieng-Ye Chang
Research Associate
Environmental Health Risk
Assessment & Risk Community Program
U.S. Army Center for Health Promotion &
Prevention
5158 Blackhawk Road (Attn: MCHB-TS-EHR)
Aberdeen Proving Ground, MD 210105422
410-436-2025
Fax:410-436-8170
E-mail: hsieng-ye_chang@cnp>pm-
ccmail.apgea.army.mil
Nancy Chin
Toxicologist
U.S. Environmental Protection Agency
401 M Street, SW
Washington, DC 20460
202-260-7587 '
Fax:202-260-1036
E-mail: chin.nancy@epamail.epa.gov
Selene Chow
Environmental Health Scientist
Toxicology Division
Agency for Toxic Substances and Disease Registry
1600 Clifton Road (MS-E29)
Atlanta, GA 30333
404-639-5287
Fax:404-639-6315
C-3
-------
Linda Cullen
Unit Supervisor
Environmental Toxicology and Risk Assessment
New Jersey Department of Environmental
Protection
P.O. Box 413
Trenton, NJ 08625
609-984-9778
Fax: 609-633-2360
E-mail: lcullen@dep.state.nj.us
Michael Dellarco
Environmental Health Scientist
Exposure Analysis and Rsk
National Center for Environmental Assessment
U.S. Environmental Protection Agency
401 M Street, SW (8623D)
Washington, DC 20460
202-564-3239
Fax: 202-565-0079
E-mail: dellarco.mike@epa.gov
JanineDinan
Environmental Health Scientist
Superfund Division
U.S. Environmental Protection Agency
401 M Street, SW (5202G)
Washington, DC 20460
703-603-8824
Fax:703-603-9133
E-mail: dinan.janine@epamail.epa.gov
Donald Dobbs
Inter-Disciplinary Scientist
Division of OTC Drug Products
U.S. Food & Drug Administration
5600 Fishers Lane (HFD-560)
CDER/DOTCDP
Rockville, MD 20857
301-827-2283
Fax:301-827-2316
Victor Fan
Ph. D. Candidate/Lab Scientist .
Environmental Health
Johns Hopkins University
P.O. Box2191
Baltimore, MD 21203
410-502-6528
Fax:410-955-9334
E-mail: vfan@jhsph.edu
Marcus Garcia
Senior Toxicologist
Risk Assessment Division
ENSR
35 Nagog Park
Acton, MA 01720
978-635-9500 .
Fax:978-635-9180
E-mail: mgarcia@ensr.com
William Gilvertson
Associate Director for Monographs
Center for Drug Evaluation & Research
U.S. Food & Drug Administration
9201 Corporate Boulevard
Rockville, MD 20850
301-827-2304
'Fax: 301-827-2317
Deborah Hastings
Epidemiologist
Environmental Health Risk
Assessment & Risk Community Program
U.S. Army Center for Health Promotion &
Prevention
(Attn: MCHB-TS-EHR)
5158 Blackhawk Road
Aberdeen Proving Ground, MD 21010-5422
410-436-2953
Fax:410-436-8170
E-mail: deborah_hastings@chppm-
ccmail.apgea.army.mil
Lee Hofmann
Environmental Health Scientist
Office of Solid Waste & Emergency Response
U.S. Environmental Protection Agency
401 M Street, SW (5202G)
Washington, DC 20460
703-603-8874
Fax:703-603-9133
E-mail: hofmann.lee@epamail.epa.gov
William Itterly
Scientist II
Novartis Crop Protection
P.O. Box 18300
410 Swing Road
Greensboro, NC 27419
336-632-7524
Fax: 336-632-7645
E-mail: william.itterly@cp.novartis.com
C-4
-------
PauVine Johnston
Chemist
Center for Pollutant & Source Guidance
Indoor Environments Division
U.S. Environmental Protection Agency
401 M. Street, SW (6604J)
Washington, DC 20460
202-564-9425
Fax: 202-565-2038
E-mail: johnston.pauline@epa.gov
Garrett Keating
Research Scientist
Health and Ecological Assessment
Lawrence Livermore National Laboratory
P.O. Box 808 (L-396)
Livermore, CA 94611
925-422-0921
Fax: 925-422-5673
E-mail: keating2@llnl.gov
Russell Keenan
Vice President
Ogden Environmental and Energy Services
15 Franklin Street
Portland, ME 04101
207-879-4222
Fax: 207-879-4223
E-mail: rekeenan@oees.com
Leonard Keifer
Chemist
New Chemical Screening and Assessment Branch
Risk Assessment Division
U.S. Environmental Protection Agency
401 M Street, SW (7403)
Washington, DC 20460
202-260-1548
Fax:202-260-1279
E-mail: keifer.leonard@epa.gov
Daniel Kelley
Booz-Allen & Hamilton Inc.
1725 Jefferson Davis Highway - Suite 1100
Arlington, VA 22202
703-412-7769
Fax:703-412-7689
E-mail: kelley_daniel@bah.com
Steven Knott
Project Manager
Off ice'of Research and Development
Environmental Protection Agency
401 M. Street, SW (8601-D)
Washington, DC 20460
202-564-3359
Fax: 202-565-0062
E-mail: knott.steven@epa.gov
Gary Krieger
Vice-President
Health Services Branch
Dames & Moore Group, Inc.
633 17th Street - Suite 2500
Denver, CO 80202 ,
303-299-7848
Fax: 303-299-7901
E-mail: dengrk@dames.com
Mick Major
Chemist
USACHPPM DTOX
U.S. Army Center for Health
Promotion & Preventive Medicine
5158 Blackhawk Road (Attn: MGHB TS THE)
Aberdeen Proving Ground, MD 21010-5422
410-436-7159
Fax:510-436-6710
E-mail: mmajor@aeha1 .apgea.army.mil
Robert McGaughy
Senior Scientist
Office of Research and Development
Environmental Protection Agency
401 M. Street, SW (H623-D)
Washington, DC 20460
202-564-3244
Fax:202-565-0079
E-mail: mcgaughy.robert@epamail.epa.gov
Tom Mueller
Study Director
Novartis Crop Protection Inc.
P.O. Box 18300
Greensboro, NC 27419
336-632-7642
Fax:336-632-7645
E-mail: tommueller® novartiscropprotection.com
Thomas Mueller
Scientist
Novartis Crop Protection
410 Swing Road
Greensboro, NC 27419
336-632-6078
Fax:336-632-7645
E-mail: thomas.mueller@cp.novartis.com
C-5
-------
Nicole Paquette
Toxicologist PhD
Health Effects Division
Office of Pesticide Program
U.S. Environmental Protection
1921 Jefferson Davis Highway - EPA CM 2 (7509C)
Arlington, VA 22202
703-035-6380
Fax:703-308-7157
E-mail: paquette.nicole @ epamail.epa.gov
Ralph Pared
Manager of Toxicology
BASF Corporation
1 Heritage Place - Suite 300
Southgate, Ml 48195
734-324-6212
Fax: 734-324-5226
E-mail: parodr@basf.com
Brian Polgar
Staff Scientist
Karch & Associates, Inc.
1701 K Street, NW - Suite 1000
Washington, DC 20006
202-463-0400
Fax: 202-463-0502
E-mail: bpolgar@karch-inc.com
Susan Reith
Senior Science Advisor
Environmental Corporation
4350 North Fairfax Drive
Arlington, VA 22203
703-516-2300
Fax:703-516-2393
E-mail: srieth@environcorp.com
Timothy Roy
Consultant
Petrotec, Inc.
527 East Ravine Avenue
Langhorne, PA 19047
609-224-4629
Fax: 609-224-4652
E-mail: tim_a_roy@email.mobil.com
John Schaum
Chief
Office of Research and Development
U.S. Environmental Protection Agency
808 17th Street, NW
Washington, DC 20460
202-564-3237
Marc Stifelman
Environmental Scientist
Risk Evaluation Unit
Office of Environmental Assessment
U.S. Environmental Protection Agency
1200 Sixth Avenue (OEA-095)
Seattle, WA 98101
206-553-6979
Fax:206-553-0119
E-maii: stifelman.marc@epa.gov
Robin Streeter
Program Director for Risk Assessment
ERMWest ,
90 Harvey Road
Chadds Ford, PA 19317
610-558-3327
Fax:610-558-3541
E-mail: robin_streeter@erm.com
P.V. Shah
Toxicologist
Office of Pesticide Program Division
Environmental Protection Agency
401 M. Street, SW (7509C)
Washington, DC 20640
703-308-1846
Fax: 703-305-5147
E-mail: shah.pv@epamail.epa.com
Marlin Visage
Executive Vice President
The Bishop Group
25 Burgher Road - P.O. Box 799
Highland Lakes, NJ 07422-0799
973-764-1500
Fax:973-764-1566
E-mail: bishopgp@host.warwick.net
Cassi Walls
Associate Chemist
Novigen Sciences Inc.
1730 Rhode Island Avenue, NW - Suite 1100
Washington, DC 20036
202-293-5374
Fax: 202-293-5377
E-mail: cwalls@novigensci.com
Susan Youngren
Manager, Aggregate Risk Assessment
Novigen Sciences, Inc.
1730 Rhode Island Avenue, NW - Suite 1100
Washington, DC 20036
202-293-5374
Fax:202-293-5377
E-mail: syoungren@novigensci.com
C-6
-------
Valerie Zartarian
Environmental Engineer
Human Exposure Analysis Branch
National Exposure Research Laboratory
U.S. Environmental Protection Agency
12201 Sunrise Valley Drive
55 National Center
Reston, VA 20192
703-648-5538
Fax: 703-648-4290
E-mail: zartarian.valerie@epamail.epa.gov
C-7
-------
-------
APPENDIX D
AGENDA
D-l
-------
-------
&EPA
United States
Environmental Protection Agency., ,
Risk Assessment Forum '* '" • ' '
Workshop on Issues Associated with
Dermal Exposure and Uptake
Bethesda Ramada
Bethesda, MD
December 10-11,1998
Agenda
- *'4 • t *', r "''••! f ^ ' '/" '' /V -
•. v.".' :j,;-r;;>:••*'!!.*! l^jn-^fi i* ,:-i't;; ..'^
Workshop Chair: Harvey Clewell, ICF Kaiser, International
Workshop Facilitator: Jan Connery, Eastern Research Group, Inc.
THURSDAY, DECEMBER 1 0, 1 998 r! " i • \ -.., •'••;
8:OOAM Registration/Check-ln
8:30AM Welcome Remarks, Meeting Structure, Objectives, and Peer Oatraadtaitton
Harvey Clewell and Jan Connery
9:OOAM U.S. EPA Risk Assessment Forum's Role , , ;
Steve Knott, U.S. Environmental Protection Agency (U.S. EPA), Risk Assessment Forum
9:10AM Background on the Current Dermal Guidance
Mark Johnson, U.S. EPA, Region V and Mark Maddaloni, U.S. EPA, Region II
9:40AM Charge to the Peer Consultants ^ , ( , ' , ,
Harvey Clewell •><••. ,-.. ^-., <• • '/'.''' '
9:55AM Presentation on Discussion Issue One: ,
Dermal Exposure to Contaminants in Water i
Kim Hoang, U.S. EPA, Office of Research and Development (ORD)/National Center for
Environmental Assessment West (NCEA-W)
10:10AM BREAK •>'''' ' =•--•* •
10:25AM Breakout Group Discussion of Issue One u ; MJ
11:25AM Summary of Discussion Issue One
12:10PM ObserverCommentsi w*>^-* '*fl • " •-•••" ••'"••!"" •:••' •.-'-; -' •: ' >.-••••:-• ;-...,
12:20PM - LUNCH ' ; •"
D-3
-------
THURSDAY, DECEMBER 10, 1998 (continued)
1:20PM Presentation on Discussion Issue Two:
Dermal Exposure to Contaminants in Soil
Mark Johnson
1:35PM Breakout Group Discussion of Issue Two
2:35PM Summary of Discussion Issue Two
3:20PM Observer Comments
3:30PM BREAK
3:45PM Presentation on Discussion Issue Three:
Adjustment of Toxicity Factors to Reflect Absorbed Dose
Mark Maddaloni
4:OOPM Breakout Group Discussion of Issue Three
5:OOPM Summary of Discussion Issue Three
5:45PM Observer Comments
6:OOPM ADJOURN
FRIDAY, DECEMBER 11, 1998
8:45AM Planning and Logistics
Harvey Clewell
9:OOAM Presentation on Discussion Issue Four:
Risk Characterization and Uncertainty
Ann-Marie Burke, U.S. EPA, Region I
9:25AM Breakout Group Discussion of Issue Four
10:15AM BREAK
10:30AM Summary of Discussion Issue Four
11:15AM Observer Comments
11:30AM LUNCH
12:30PM Next Steps: General Discussion on Dermal Exposure Issues
2:30PM Workshop Summary
Harvey Clewell
3:OOPM ADJOURN
D-4
-------
APPENDIX E
CHAIRPERSON AND EPA PRESENTER OVERHEADS
E-l
-------
Peer Consultant Introduction
Harvey Clewell
E-2
-------
cd
I
CO
e outco
•t
1
CO
HH
HH
g
-------
00
a
o
CO
CO
CO
i
CO
1
O
c/i
«i-H
Q
CO
O
'•• S
0) -'fe
II
E-4
-------
U.S. EPA Risk Assessment Forum's Role
Steve Knott
E-5
-------
1 a
&&
42 so
CO
"ii to
8.a.
•8 a
co O .
tS*
to o • •
a .&
IB
•S'S
*3 >^ co
i §^
1&12
=4 . S5 O
2 & d
rg ^ °
PH co
U8§
E-6
-------
E-7
-------
Background on the Current Dermal Guidance
Mark Johnson
E-8
-------
CO
0
03 g
E 03
o
LLJ
0 •=
Qo
Q_ c
LU O
CO
CO
0
CO
CO
(0
E
3*.
O
c
S w
CO Q
o
o
CO
11
E^j3
4f%
Q)
< E
«w ^
O
E-9
-------
tfl
-------
BM &
C'j*- co
O .£ ^*
** C£ CO
•
c
0)
E
(0
0)
o
0)
(0
<
J£
0)
E
_fg
1^1
E o ro
% EiS
flj *5 > C
CO CO O
^ o to
.^ £ <£
00^
•S c co
2 P °
O I- Q
^ ^i
0
J^ O
c
0
E
CO
CO
0)
CO
3
^_
2.
- £,' '
! J
";v, .
!. \
'C,'/ .!'•-. ,, ; '.,
•«.- -«c
>.,,.. ?,o
^
" -3
9
'••/ ,;, - »
E-ll
-------
CO
E
CD
CL
UJ
o
O
LU
0
co
0>
S
0
o
•S-s
c .2
0) ^
O CO
co
CO
<
o *-
co o
< E
?. 8
3§
o> <2
T- <
T3
C
€
O
£L
C0
O
O
CO
3
O
o
E
<0
0
O
(ft
0
(0
00
o>
E-12
-------
• mm.
«
CO
CO
o
E
O
JJ^j
o
SOL &•
rtzl! CD
5 15
*% 5*
o ^^
<0
c
o
o
• •••• -
(0
E
o
o
LL
0
X
Q
LL!
x
LL,
LL!
X
H
LJJ
X
0
OL
x
CO
X
0
^^
"c
«tf%
CO
c
o
•
~
H 05
< 0
x £
^ 0
mf^^^ mf^
m CL
CO
0
T3
.2
"o
II £
0 CO
CO -L
O 05
"O O
E ® ^
0 ^4
CO • • f"
/"^ f ^ c""
< Q_ ^
E-13
Rco
0
"-T
mmf
"c5
£
0
-5
\g
. c
0
CL
O
"5
CO
o
JO
• •
8
lta*»
C
(C
'5 .
O
CL
IU
-------
'o
CO
e
to
(0
o
£
0)
O
HM*
"i
Contact
Q
LJJ
X
u_
LJJ
X
LJJ
X
CO
GO
X
X
<
X
CO
O
ii
0
C/3
>-H
5
X
CD ^
45
adherence
CO
i surface ar
"o
CO
o
V|_
c
g
CO
0
^
o
^ ^^.
^
^
1
• c
0
a.
o
0
^
ice: consul
c
0
"I
aT
0
""3
mmf
15
LL
••^•^•i
^^^
^/
c
CO
CO
CO
o
"o
0
ative estimates
«^r
0
C
O
0
13
•*
^^
CO
>w
1
c
CO
£
o
0
(/)(/> CO
II II II
LL, < CO
< CO GO
o
O
o
0
LLJ "O
E-14
-------
co
i* a>
ro g
E as
Q_
LU
o
o
o
LLJ
CO
CO
0
CO
CO
(0
•0.2
C n
E o
E£
O O
£ E
CO o
< 1
Z CO
.„ CO
•o
i
O
Q.
3
(0
0)
O
C
as
«E
co
C
~ .SS! O
o
CO
CO
o
CO
CO
CO
2 co
00
0>
<—-•'*•
^^^^ ^^^^
CO
J}
£L
"S
C
O
E
o
CO
CO
So
3 a;
CO O
o ^
a •
x co
ui
NMMnV ^"^"
C9 '43
C Ctf
E .S>
-------
J*
CO
E
&•§
O
CO
c
•2 >»
«j O)
•S o
C VQ
0 -n
E o
E£
o o
0 E
£ ~
0 §
1983: NA
Assessm
•o
c
€
CD
Q.
3
CO
|M ,
£
0)
o
c
CO
^o
^•^
0
, o
CO
CO
o
CO
0
CO <
o> f£
CO
0
Q.
"o
c
"C
Q.
• •
ll
o
c
CO
CO
0)
CO
CO
So
|o
^M^
&«
w§
CO "45
ECO
v, 0
si
«5
o ^
o c
T- CO
V
CO
E
•MM
(Q
E
o
Q
E
m ' . Jn|%
|0
JE co
ii
i2
cu* c
£L O
1992: Su
Assessm
E-16
-------
^
,co
"co
E
0
Q
Q.
LU
"o
LU
0
O
CO
.2
O
"c
0
E
CO
CO
0
CO
CO
TJ
CD
•o
Q.
3
Q.
0
T3
Q_
13
O
D)
|
15
E
Q
iH
o>
o>
8
C
CD
O
c
«M
«i
0
Q.
CO
0
15
c
8
CD
,-g
•V
O
1
Q
*o
0
o:
0
0
0.
15
0
'*-«
N!
(Q^ -*-J
June, 19!
documen
CD
Q
i
0
a:
0
0
D_
15
c
1
^>
LU
do
a
o
January,
d
0
O
o
T5
^•fe.
Guidance
1
0)
o;
L.
0
0
CL
c
O
T>
0
CO
CD
jQ
1
Q
0
_co
0
a:
• •
CO
o>
a>
t5
0)
3
CO
comment
0
u.
CO
0
CO
.£2
|
.g
o
"c
0
2
B •
00
o
0>
October,
a.
0
JZ
CO
u.
^^
1
c
o
"s
CO
•MB
O
O
0
0
a.
0
D)
M^^
CD
Q
CO
CD
discussio
E-17
-------
(0
3
(0
to
c
0)
(0
0)
(0
E
75
E
O
ell
CM
E-18
-------
Presentation on Discussion Issue One:
Dermal Exposure to Contaminants in Water
KimHoang
E-19
-------
O
g
O
g
a
CO
CO
GO
CO
2
i
fl
o
O
E-20
-------
s
o
CO
U
CO
O
d
o
ctf
1
O
O
8
1
a
• fH
CO
*rt
o
• ^H
s
J*
o
CO
i"H
XI
VH
O,
VH
O
so
0s*
ON
(N
ON
PH
w
Q
(D
O
^^
Q
a
o
Q
Effective Pre
Q
CLj
w
(4-1
o
1/5
o
CO
o .2
0
CO
S
CD
1
CO
o
§
CO
G
0
CO
CO
C^J
§
CO
o
^ o
E-21
-------
«p\
Ofi
h
O
fi
o
1
"3
o
U
w
CO
c«
CD
CO
CO
s
s
§
S-
%
•&
GO
.ET
o
PH
Ctf
I
d
I
1
PH H
O
f-H
*a-
o
O
CO
hJ ON
CO
§
CU
CO
E-22
-------
a
*«
E
e
Q
«M
ation
S3
•« .9
ti
data
:s.
•J-H
f—t.
O
o
<4H
O
CO
•rH
CO
£ s
^ 13
*T3 O
CO
ctf
Q<
X
GO
*CJ
T T
S
PH
f
2
;ft
!o :'•
!:•
•fl>'
fa
O
O
CD
O
a
fi
(D
E-23
-------
IHOFIAV Hvinoanow
E-24
-------
^
"
o
Q
AH
to
a
o
=
OS
VI
2
VI
VI
VI
E-25
-------
Chemicals from Table 5-8 of DEA (EPA, 1992) identified to be
outside of the Effective Predictive Domain of the Flynn database
Log Kow < -1 and MW < 60
Chemicals
Water
Urea
Hydrazine H-suIfate
Log
Kow
-1.38
-2.11
-2.07
MW
18
60
32
Log Kow > 4 and 150 < MW < 350,
andMW>600
Chemicals
Benzo-a-anthracene
Benzo-a-pyrene
Benzo-b-fluoranthene
Chrysene
DDD
DDE
DDT
Decanol
Dibenzo(a,h)anthracene
Fluoranthene
Hexachlorobenzene
IndenoQpyrene
Nitrofen
PCB-chlorobiphenyl
PCB-hexachlorobiphenyl
Pentachlorophenol
TCDD
Tris(2J-dibromopropyl)
phosphate
Log
Kow
5.66
6.10
6.12
5.66
5.80
5.69
636
4.11
6.84
4.95
5.31
6.58
5.53
6.50
6.72
5.86
6.80
4.98
MW
228
250
252
228
320
318
355
158.3
278
202.3
284.8
2763
284.1
292
361
266
322
697.6
E-26
-------
.max
•
estimation for outliers
Kasting and Robinson (1993): K^jMsr bounded by
absorption through epidermis and blood flow rate:
deviation from DBA membrane model: -: , •-.•.
1
K
1
1
K
p,sc
• ;
1
P,ve
where:
ax: upper limit for Kp>w in aqueous layer \' % ; ' h ;
sc: steadyrstate permeability coefficient through, the st^atu^ cprneum
(sc) (obtained from correlation)
. ',..'' li/-;^^}::•„• .,^i ?;,-•' 'V^^-I •
Kb/v : blood-to-vehicle partition coefficient
qi,: cutaneous blood flow rate per unit area of skin
Kp>ve: steady-state permeability coefficient through the viable epidermis
(ve)
Kp.max used only when tevent > t*? for steady-state
absorpion through all layers of skin
'' ' '
E-27
-------
Kn for Inorganics
• r
Table 3.1 Permeabflity Coefficients for Inorganics
xompound
Permeability Coefficient Kp
(cm/hr)
Cadmium
Chromium (+6)
Chromium (+3)
Cobalt
Lead
Mercury (+2)
Methyl mercury
Mercury vapor
Nickel
Potassium
Silver
Zinc chloride
All other inorganics
lO'3
2xlO-3
1 x 10-3
4 x ID"4
1 x ID"4
1 x 10'3
1 x ID"3
0.24
2 x 10"4
2x1-6*
6 x 10-4
1 x lO'3
E-28
-------
o
cu
M
o>
O>
o
cy Sc
I
o
Expos
toi)
co
CQ
.f
PQ
o
I
CO
I
o
&
.3
CO
o
-------
Charges (1)
ForOrganics
o Comment on the database used to derive
the correlation equation
o Comment on the correlation equation
(predictors KQW and MW) used to
estimate the skin permeability
coefficient (Kp) and the 95% CI
o Comment on the statistical analysis used
to establish the Effective Predictive
Domain for the Kp correlation equation
o Comment on the use of Kpjmax
o Discuss the use of estimated Kp vs.
experimental data
E-30
-------
Charges (2)
For Inorganics: comment on the approach
recommended for metals and inorganic
chemicals.
-...»..- "> i ,.;;:'" '. . i ..; . -, ,.v • -•
Comment on the other exposure default
values
Discuss the Issue of Using Model Instead of
Chemical Specific Study?
•E-31
-------
Presentation on Discussion Issue Two:
Dermal Exposure to Contaminants in Soil
Mark Johnson
E-32
-------
o
CO
••••I
tJ
:
o
o
<5
0)
o
Ul
2
I
Q>
§
OQ
o
o
0)
JD
2
0)
II II II II II II II
I
Q I
< << LL O
Q Q CO HI 111 LLJ OQ
E-33
-------
o
j»
03
m __
(D
I
0)
o
Q.
5
CM
^^
ll
C3)
= E
|
ii
CO
— c co
I
/is ^^ **""
L. CD *—' O
& o »- ,
Q. C
2
«fc
i =
o o .2
*s O -5
cl^o
o o Q ^
'to c T- ro
i- CD
CD CO
^^
T3 C
CD "r™
JQ E
$<3<351S
II II II II
, LJL
Q O O <
II
0?
CD
E-34
-------
0
o
CO
o
CD
CD
O
C
CD
CD
•• •••B^^^M
o
00
0)
0
CO O
c
,* ®
CO N
0 "co
c. J
0 O
CO
Q.
0 C
"•E S
0
Q.
0
CO ^
CO
CO
g
O
CO
CO
0
CO
C
X
co
0 CX
O >,
c -o
0 O
0
•o
CO
= J>
o it
co =0
CO
o
Q.
X
0
CO
0
CD
0
O
O
CO CO
E-35
-------
o
0)
Q.
£ O
^ •"•
> fc
(0
(
o
o
(0
0)
o
x
LLJ
'*-• ^
2 ;§ •§
£ -o c
o n ^ -=
w -75
<0
o
'8
CD
2
Q. w
.g'.&^TJ |g I §21
E-36
-------
CD
CD
OS
0
CD
CO
CD
CO
Q.
O
m
E-37
-------
Table 33 Activity Specific-Surface Area Weighted Soil Adherence Factors
CfflLDREN1
[Mldren Playing (dry soil)
Daycare Kids
Children Playing (wet soil)
Cids-in-mud
RESIDENTIAL ADULTS2
Ace(yr)
8-12
1-6.5
8-12
9-14
jroundskeepers >18
Landscape/Rockery >18
•gflrdfinSTS ^lO
COMMER/INDUSTR ADULTS3
jroundskeepers >18
Landscape/Rockery >18
irrigation Installers >18
jardeners . >1'6
Construction Workers >18
Equip. Operators >18
[Jtility Workers >18
OTHER RECEPTORS4
SoccerNo. 1 (teens:moistconditicttis)13-15
Soccer Kos. 2&3 (adults) >18
^rcheqlogists . ->16
Farmers >18
Rugby >18
teed Gatherers >18
Weighted AF
(me/cm2)
50th % 95th %
0.04
0,06
0.2
22
0.01
0.04
0.07
0.02
0,04
0.08
0.1
0.1
0.2
0;2
0.04
0.01
0.09
0.1
0.1
0.3
0.2
0.2
2.7
123
0.5
0.1
0.3
0.7
0.1
0.2
0.4
0.3
0.6
0.8
0.2
0.07
0.3
0.4
0.6
6.3
1 Weighted AF based on exposure to face, forearms, hands, lowerlegs, & feet.
* Weighted AF based on exposure to fece, forearms, hands, & lowerlegs.
3 Weighted AF based on exposure to face, forearms, & hands.
Note: tiiis results in different weighted AFs for similar activities between
residential and commercial/industrial exposure scenarios.
'Weighted AF based oh all body parts for which data were available.
E-38
-------
Presentation on Discussion Issue Three:
Adjustment of Toxicity Factors To Reflect Absorbed Dose
MarkMaddaloni
E-39
-------
ON
oo
OJ
c
Process
t—i
•8
CO
OH
E-40
-------
3
"S
o
H
X
O
Q
II
8
J
O
g
P
S
o
E-41
-------
CO
«j
CO
CO
•9
t
O
o
T-H
II
o
8
3
d
t
CO
O
-------
s
to
3
nd
<
£
o
'S
o
H
(D
C/)
GO
CO
3
CO
O
O
d
fl
o
CO
-------
>>
a
GO
O
B
a
•t-H
a
o
o
C3
PH
a
o
•TH
&
o
CO
PI
O
o < i^j
e
JD
CO
rrt
S .o
O
-------
03
t3
Q
c
o
s
bO
a
PH
O
a
'
§
X O
0-42
o
(D
3
CO
« ^
•fi
§ g
CD
• i-4
I
d>
o
f
f
cS
o
c/)
O
81
PH
a
o
•1-H
C/3
•1—4
O
•Sta
>> o
O
, 1
o
PH
O
E-45
-------
Presentation on Discussion Issue Four:
Risk Characterization and Uncertainty
Ann-Marie Burke
E-46
-------
% ' '
E-47
-------
.^w^^55Jj^v^vv*A^x^V^A-^Vf^>^^S^\^ N^X^^* \ / % ^>V^^' %'VUs
-? rH- fX , s<"^ * ^A'PH^* '<'« o o- :fe-$K'-.\'-'tC
^^^vl^^^™^./%, ,'£•>,'$,'',"!." » s' "" ' *• ^ -»A!«\ ' ," *< « >, 't^-t ,
>\ \->s* ' ' <.c>iv ,, ^ -- -•. ,sv""—I've *v>/^ -:r\ */vs ?t & -f
T-Svf'f '••s** " X" ' "' ' " " " ' "';'<"'%">^'-< " *'*'f*, T*t v»"J" S5^*^ ^ "'Sf*
:^r-s -,;C*:TT? , 13 - *-• ,-
'. . ^ s % ^ s* x-/~ t1- ^9E- ^ f f
&*\.juf . •••" ffJ. \f.-,^ «.>.J9ffW -.'•.!. VS
E-48
-------
''"•••''' -
-------
•\v.
>•• '«"«*. *l
«^v^<
'"Vf •*. r>f& j*.,.
^!
5- ,J&S>
«&.»«5
•sZlt* •£
.<•-»«
:^:
Hut
J" '
— **S, '
*?~tw
jhHWMr ^^^
fci^-'S
got
^^2*
*?£
"x;
t»:,>^ ,55^,^:
|:jg5:,;''/^'JK'%r'':
[ gf :^v;fe'%^
t M.*." , -, %rt ...-1 -'^
:o>
«:
«J^
c»
%
«l
3
&:
» &
i
,S*Sr
;^fe B>"H '" -c s' * ' ^ .».
.••w, s ^^^ %-4b«rx
ii--^''§
_; ""•&- fi?
M^'Wps«?.*S^>;'"N'*f/- s ':?WTO -^'
7r^At^^U^"A^'
* s •#•*"**• *
s*lM'^Cf^A- x ^ ""', ^ /
' ^f:lf^4^-;-.^^f>" ••-'- %\%s<*. * ^
E-50
^tl
N!S
»««(
'•.«',
P*i
^^? "-1
-------
E-51
-------
sH";
\*??s
v.-. * " , ':* „ s f T;;"** ^ r ' -x.'sa,-"",. ^S*^
* ' „ ,- ,xr#tf*<** ,,, '•^f,"«-y^r,^;fo<
•^ >','" >r~ / t't*},* 't/fr'f* -. ^' J?',
^^^'^ /" %^vx^^.' ^^%'
iftitii,'*^,; >"' '"'/ ;>*
&HK-
TO&"LC?'
,v^&r
|::^S^J/\>; ^>
»»&'j^»->' -> ^ -**^H- & s,*, " ^ *«, ? \;;-. - --. ::, jt >>';? ^- «;^mt,;;,,.
I.s!Mlt--A^..VV^-/^v«^- '^r<-4\ -;v ,}- -, ""- ' 7:^**'; ' %*%"\ \//%- - j'^?^*%-*?,
?i?:S?S^"-" dv^vww - **" ' \'^r?-*"' -'-' • ™~ W ,* '\' -f H^y^* •
^ttiJTI'"" --"r "^'^^^ "'" -•."••-> ^ ' " ' ' H°" ' •»<*<' v '' V'- -/>4$5Kv;,
f \- rt^-S^^^ '-*^»% ^ -1st'' ,-; :»: ,r//*"'% ^i^-
,-
idJVJ
M;« \£ Si™- "•
au^TSSsr^J
v**o -
?'r *oi'" ^
jSsr<\X,x V^J'
"? ' S.*, V- '
'" '\?tX>'.
••••* 2j
s *«t W, *
^,\, 4k,
:^^>5?:
&
;'"•• ,\ >
\\ssx-,,- -* ' &
•« '^ i.
'V*
(vt.^ Vn
:*?.
.E-52
-------
E-53
-------
*/••*•*..*<*' ~.« <~- ' >, -, »»> Tftt"' '' '> % *»>*•*' ''' "' '''''',,, '', ; /*'' \*
V^ -- - , ^ , ,*, -*'''„ ,- ^,; --, „ ^<;~ -/,;?,', ,„ ,,,/f,;~:;, * ;-^'-;
^^^^^?^|^^~,,s " , ^ "^ - <^ - , ' . „„ ,"4 ,„ ',
*«X>*J 3f.!j?f ^ « *% "•
s^ ''^S^ ' */ '''''Jm&Ht ^^^&*•*
.^'J^%<&*^ff& "V|
* ' - ,^fe iKiiwifm , ,„ ; ^T^ -" #• - 'Cf'^^^ ~-/'^?vX^*T"f;"
- ,-^ : J]' J?t 4 jft|'' 5* * A* '" * ''"-"*'"'' ^ ''' fCI^ "*^-S: '^'^%'^'^
, ;- "r-yrsv^C *^g «w - , £"U - ' ';$» -\>' , v^W'V-.v
''
,>--' s ^
V*"*" ^^""' ''"'''i'J/. "' /-.T/JEf.t
€i& *»5^'-;''' * ^.^ '-^ - i , / ^*
«. - ' » ^>;•',',- ~,-;-i?.'f-z-:t*'r"S"
5*!*'
~fs%
- , ,' "<,">'''« '
f f ffff s vAtf > '
' *\«. \ i""^;"\^s
' x -/ " ,
fl». *W--
A.tJi \V' T^k'**1^^ . , s"; ••••';v ,' ,'*""'-
^. 'f-J- V\fc •>*> s S-s " s^ /•'«•.•> •• * ' , '
N^;^ *^:;^'^"' "v, „ -.
.?^^t>:-^ ^>-,, - v %% ' ^^- -- -
E-54
-------
E-55
-------
-. VVA> '•'•'•
•• . s«/s.;x- ',.
N<&^-•s &s v *•%<(; s,
' -./^^ s *s' *"• ^ ' •* ' 5" \ •• ** ^""' Sl"
^12^"',-''-" ^ - ,- ',-* >f^''l" ^issr*'* 'i-,£r*'''^
' »(•••»^'^x'/^/', '%i| ^\x»'^ -v^> ^ " ^7^'^'^*%^
•A-v',.*^^ '-"»'=«J'"^'' " -,"'''- <^ *&*'•>*>*$&.''#&&
•'-.?.*
•'/, .'.{ •*>
*?< \
« t/
E-56
-------
Next Steps: Plenary Discussion on Dermal Exposure Issues
(typedfrom Harvey Clewett's handwritten notes)
E-57
-------
TODAY
(SOD)
• 95%€IforKp&EPD
• "Underpredicts halogenated chemicals"—can use C to assist
• Monolayer correction
• Acknowledge preferred RTR (G & H)
• Sensitivity/uncertainty analysis (qualitative/quantitative) H, M, L
• Define/illustrate/bound
TU ;.-..' - : . •••.:,.-• .:'"••, . ,. '.
• Define recreational exposure assumptions
• Establish standing SDWG
— w/funding
• Acknowledgement of loss to air of VOCs in shower
• Add vehicle to RfD table
E-58
-------
TOMORROW
A.M.
P.M.
Regression analysis for Kp
— Vecchia database
— MV,T
P.M. SUBSTRUCTURE
Sensitivity/uncertainty analysis of parameters
/
Depository (website) for reviewed Kp vaKes etc. (not another IRIS)
Standard criteria for exposure Kp protocols (& soil)
— retrospective & prospective
(OECD—K^)
REFINED SOIL ADHERENCE (KISSEL)
CONCRETE & OTHER SURFACES
(TRANSFER FACTORS)
DEPOSITION (pesticides)
— pesticide absorption vs. water
STANDING AGENCY DERMAL W.G; ,
TRANSFER/INTEGRATION OF DATA ACROSS PROGRAMS
E-59
-------
THE NEXT MILLENNIUM
MORE DATA.
• INORG. FROM SOIL
— CHEM. FORM
— SOIL TYPE
» DERMAL TOX. (P.O.E.>-in situ
— DERMAL RfDS
— SYSTEMIC &P.O.E.
— human in vivo studies
— soil vs. existing data
— internal biomarkers
— PK/kinetic models
— human validation
-T- butoxyethanol
— isopropanol
— in vitro soil studies
— . variability of soil
— chemical form
— contamination/preparation
— classes of chemical
— inorg.
— SVOCs
— pesticides
— persistent
— develop screening tests
—• mixtures
— move beyond absorbed dose to target tissue
— skinmetab.
— reservoir .
E-60
-------
APPENDIX F
BREAKOUT GROUP OVERHEADS
F-l
-------
-------
BREAKOUT GROUP DISCUSSION OF ISSUE ONE:
DERMAL EXPOSURE TO CONTAMINANTS IN WATER
GROUP 1
3.
4.
Comment on correlation equation for Kp
Kp equation? — OK, molar volume may be better
95% CI? -- uncomfortable wifh this — assumes an error structure that is unknown
Database used to generate equation —
More data generation .-.,...•
— relevant chemicals
— Log Kow , (high & low)
— standardized procedures
=0 Superfund interests are persistent chemicals
also included in high production chemicals
Statistical analysis of EPD — OK
There should be a domain based on the properties of data in the database
Extrapolation outside the EPD — No ,
Kp,max?OK
K,, & Kp,max-OK? ' ..
' - ' . ... *..',, L •.,''•'.
Kp/Kpmax vs. exp. values.
=€> uncomfortable with ignoring data
=f> options:
report correlation value
report "consensus" exp. values. Average (?)
Metals — OK, but could be improved. Some issues are:
speciation is not considered
methyl mercury — organic rather than metal? . ; - .
mercury vapor
arsenic is missing in metals list ,
No significant comment
May want to state in the document dermal absorption issues of importance to EPA that are not
treated. For example, pesticide absorption from pesticide formulation is not treated in document.
F-3
-------
GROUP 2
Issue 1: In Vitro Correlation
1. Database—analyze Vecchia
2. Include substructural parameters in corr. analysis—better r2
Explore other Kow predictive models
3. Explore nonlinear models for relating Kow-Kp
4. Use molar volume in place of MW (software avail.)
5. Consider variable dependency (MW-KOW, molar volume-Kow)
5a. Consider data transformation in model (e.g., Log)
6. Need high quality database of Kows—reviewed
e.g. "Star List"; other data experience
7. UCL,,5—doc needs to show how UCL95 was derived
8. Need to explore experimental Kows, e.g., HPLC
• Predictive model may be wrong j
Need clarification of UCL^ I NOW
Issue 2: K^ EPD
1. Need clarification of "outliers of EPD"
2. Need approach for replacing predicted Kpmax w/ experimental
• Predictive approach yields highly conservative values of K,,
Issue 3: EXP Duration
1. No basis for distinguishing tc < or > tss
This needs to be explicitly stated in doc.
Issue 4: Predicted vs. Experimental
1. Use of model might not reward new data collection
2. Use of model contributes to consistency
3. Need criteria for evaluating experimental values
F-4
-------
Issue 5: Inorganic
1. Better documentation of derivation of chemical-specifiaKpS
9. Methodology needs to consider both ionized & non-ionized states of inorganics
How did EPA deal w/this issue
Use models—water parameters
Issue 6: DBF EFs
1. Need loss term for Cw—Doc needs to address Cw term on chemical-specific basis
2. PRA
F-5
-------
GROUPS
1.-4. Guidance should encourage (at least not discourage) data collection
— need protocols if site specific experimentation is to be conducted
=& Seek validation
1. Seek consistency across EPA groups
5.
6.
10"3 for inorganics
— Hg?
— Others don't drive?
10-minute shower
— some doubt as to validity as screen (NHAPS?)
— Fig. 1.1
residential vs. recreational
F-6
-------
GENERAL DISCUSSION
• 1. In vitro correlation (Potts & Guy equation)
2. Effective predictive domain
3. Short-term vs. long-term exposure events
- 2.4tlag - .,, .. . ,
• 4. Predicted Kp vs. measured
— criteria
5. Metals/inorganics
— default Kp
• 6. Exposure defaults
— shower > D.W.
— VOCs
Dermal data
Criteria:
S.S.
Donor vehicle "just H2O"
Receptor vehicle
— "no barrier"
— VOLPO
T,pH
In vivo vs. in vitro
— receptor (high Log Kow)
— pore effect (low Kow)
— metab. (OPs)
OECD
F-7
-------
BREAKOUT GROUP DISCUSSION OF ISSUE TWO:
DERMAL EXPOSURE TO CONTAMINANTS IN SOIL
GROUP1
MOREDATA
1. Current Absorption Factor Approach vs. Rate Approach
Given the theoretical limitations and data limitations maybe this is the best we can do
now—
— Note that data assume 24 hour exposure always
— Not all data in the literature are included—
— How should data be added—?
— Data listed in table should be for less than monolayer coverage (and noted as such)
Concerns
a) time effect
default values—b) mechanisms of transfer from soil to skin (expected vs. actual):
sweating, direct contact, vapor pressure
c) Soil transfer to skin; turn-over vs. no turn-over
2. Data used—
— soil properties of applied soil (org. carbon, moisture)
— particle size
— amount of soil (cp. to monolayer)
— adjustment to monolayer?
— contamination procedure
— default-chemical groups
3. 10% organic compound default?
No, not clear
4. 10% for inorganic
No, not clear
<=£> Solvent extraction (cp. to max. cone, in solvent)
5. No comment
F-8
-------
GROUP 2
3.
4.
3. &4.
Need more experimental data:
kinetics
=z> rate approach >
!=£> time adjustment of absorption fraction •. .
absorption fraction
screen based on extraction—modeling of adsorption
Need guidance for experimental:
std. soils
— particle size ]
— loading ] current data biased towards:
— aging ] — large particle size
— unaged
expt. prep.
— in vitro
— in vivo est. of absorption
Table 3-4
Numerous chemical-specific issues with values in table
10% organic default =£> /. "better than nothing"
— basis not clearly presented
— empirical support is not overwhelming
(need more data)
(existing data biased)
— Soil extraction (solvent) might be a basis for departing from default
— state limitations of use for screening
1% organic default
— low confidence in default
— based on 2 studies
— need more data
— explanation of basis of value for cadmium
Present default as distribution
(e.g., range)
5. SA weighting—"A great leap forward"
Estimates for multiple activities
— need to explore various statistical approaches for estimating CT & RME
F-9
-------
GROUP 3
1. ABS
— short term yes
— ' long term, move to AT * 24 hours (requires knowledge of ABS as f[t])
— current implicit adult SCR ~ good, can't change parameters in isolation (contact
duration)
— particle size, aging—yes (LT), moisture -no
2. Table 3.4
— dominated by Wester et al. data
— particle size too big
— in vivo contact?
— aging
3. 10% org. default
OK
4. 1% metal default
No GI adjustment?
5. Total body AF
— generally good
— site specific?
F-10
-------
BREAKOUT GROUP DISCUSSION OF ISSUE THREE:
ADJUSTMENT OF TOXICITY FACTORS TO REFLECT ABSORBED DOSE
GROUP 1
1. GI Absorption Factors for estimating oral bioavailability .
— dermal RfD best , :
— IV or inhalation better
Almost all absorbs— ,, ; , , , • .. ,,, . . ;
— but should consider oral 1st pass metabolic transformations
— oral inactivation (dermal would be more toxic than oral)
— oral activation (oral more toxic)
. r }
Recommend language saying OK to use GI Absorption Factor (~ 100%), except for chemicals with
known or expected or possible 1- pass effects.
Route-to-route extrapolation—absorbed dose rate
ingestion
— gavage
— feeding
— drinking
dermal ••.-,-
, ', "! f
— showering exposures=€>shot time
— soil or occupational exposures=£>long time
Ch. 4 needs to describe better what the data in Table 4.1 are & are not. Anticipate the temptation
for misuse of the data—may want to include RfD from the critical study since the absorption
fraction is tied to this.
F-ll
-------
GROUP 2
RAGS:
oral-'dermal
THQ =
water-soil
IHQ =
RfD-AFl
I-RAFI
RfD
.-. NEED lower & upper bounds & CTE (central tendency estimate)
HQ,
.oral
Options
!•
u,
dermal
RfD-AFH
Udermal
2. HQdennai
Otion 1
NOAELsp • AFsp • UF
U
dermal
1.
NOAELsp • AFsp • UF
Revision of RfD prompts revision (review) of AFsp
2. AFsp/AFH is (may) be a component of UF
3. AFsp are tox. factor-specific (RfD, CSF)
F-12
-------
4.
5.
AFsp * AFH
AF proliferation:
— AFsp, RfD (1998, 1999, 2005, etc.)
— AFsp, CSF (1998, 1999, 2005, etc.)
- AFH
Option 2
U,
HQde
der
1.
RfD-AFH
Consistent w/definition of RfD
2. Requires EPA to develop one set of AF values, i.e. AFH
3. Update of AFH is prompted by new AF data, not new RfD
4. AFH can be assessed using "standard" approach:
— weight of evidence
— UFs—inter- & intraspecies
— lower and upper bounds etc.
5. EPA has to develop AFH values anyway to support extrapolations across chemical species
and matrix (e.g., source =£> soil)
6. Avoids "AF proliferation"
7. Prompts data collection on AFH needed to support EXP-UPT-biokinetic modeling
TABLE 4.1 VALUES:
— Speciation of inorganics may not be comparable across routes
— Need to be clear that values are for oral=4>dermal extrapolation and apply to the NOAELsp
and not necessarily to humans
— If better data for AFH are available, should it be represented in table 4.1 ?
— Need to specify to which tox. factor the values in Table 4.1 apply (RfD or CSF)
— Values indicate level of precision that may not be realistic (e.g., l.# %)
— "Other organics"—may not apply to TPH
F-13
-------
— If we had better biomarkers of exposure, we wouldn't need Table 4.1
— Cd & NJ values do not reflect best data on AF N; & Cd
— Need to consider dose-rate differences
— oral=§>dermal (i.e., Haber's assumption)
MADD. RULE:
— OK
— How do you assess the data to determine if 50% criteria is met?
MISC.:
Need guidance on how to assess AFora,
There is no "oral" guidance analogous to dermal
F-14
-------
GROUP 3
Route, vehicle, dosage
species, age, sex
noted in text<=t>should be added to table
+ dosing regimen
+, measure of availability „
+ 1- pass activation/inactivation (and dose)
Availability adjustment w/o PK adjustment
=£> danger of false confidence
50% adjustment not popular (80%?)
F-15
-------
BREAKOUT GROUP DISCUSSION OF ISSUE FOUR:
RISK CHARACTERIZATION AND UNCERTAINTY
GROUP 1
ft/IORE DATA—Especially from soil and, for chemicals with a large Log Kow, from water
1. Have all of the factors been identifiecH>laundry list
Yes—
=0 ionization of chemicals in water are not necessarily uncertain if pH is known or can
be estimated
2. Characterization of uncertainty?
Order of magnitude indications?
2,10,100 fold?
Relative importance—
which uncertainties are more likely to be large/small
May want to group uncertainties in:
(1) exposed dose
(2) absorbed dose
Kp, texp, Abs. Factor
(3) dermal to oral (species) extrapolation
3. Is it possible that dermal risks really could be larger than oral or inhalation?
Yes, sometimes—
Identify chemicals (classes) that this makes sense for?
F-16
-------
GROUP2
Model variable distributions ; .
• Shower time—high conf. in variability est.
• SA " " " .-,..- .'.• , 0.. '...-': :J
• ADH factor—when "Kissel is ready" ; ,
• K,, "don't go there yet" ,
• sometimes you have higher confidence in a distribution than a point estimate
UFs
• sensitivity analysis
— ranking of importance
— ID research needs
— what distributions are needed
• other UFs:
— temperature effects
— aging of skin
— cone, term
— summing across pathways
— different abs. dose rates
— pathway specific tox. targets
— Kp prediction uncertainly
U Communication
• Quantify uncertainty—
=£> Which factors are most important?
=£> Can collection of data @ the site reduce uncertainty?
Dermal UNC Relative to Other Pathways
• If dermal risk drives soil pathway risk:
"Houston, we have a problem"
• No data supporting health effects at SF sites associated w/dermal pathway
" It would be good to have dermal RfDs for a few problem chemicals (e.g., PAHs)
— Benchmarks—is the dermal pathway a problem?
F-17
-------
Reducin g Uncertainty
• Compare biomarkers of exposure (e.g., BEI) w/ estimated dermal dose , ',
• Apply models for estimated K,, & dermal dose to Pharmaceuticals: nicotine, nitroglycerin
— Kp model evaluation
— dermal dose model evaluation
• Compare predicted dermal dose w/in vivo estimated dermal dose •.-...'i •
.-. Model(s) evaluation/validation is needed !
— validation studies cited are not adequate to support a confidence statement
F-18
-------
GROUPS
• How does this approach relate to MCLs
— concern that MCLs might be used nonconservatively
1. ~ Yes? ID vs. emphasize
— route-to-route (but not PK, metabolism)
=£ dermal absorption of inorganics from soil
• section 5.2 intro
=f> no mention of route-to-route
• L, M, H ratings of parameter confidence
• check list with refs. of PDFs
2. Emphasize "comparative"
3. Dermal uncertainty vs. other routes
clearly worse
=t> route-to-route
=£> inorganics from soil
• add monolayer adjustment
• sediments?
• replenishment? =§> ABS
4. • more data (standard protocols?)
• hunt for validation opportunities
F-19
-------
-------
APPENDIX G
POST-MEETING COMMENTS
G-l
-------
Note: Post-meeting comments were not required but were submitted by two of the peer
consultants, Dr. Gary Diamond and Dr. Kurt Enslein.
G-2
-------
Comments on November 1998 Draft of RAGS Supplemental; Guidance:
•;•••• , Dermal Risk Assessment ,;
by Dr. Gary Diamond
1. General
The document is very important and will be extremely useful to risk assessors. The Dermal
Workgroup should be commended for their efforts to develop and bring this guidance document
forward. " ": •..•"::'.'.•.••••/. y.v .••;,'! .••,,., ,.• ,
2. Table 3.1 ,
The bases for the recommended values of K,, are not obvious from the document. It would be very
useful to the reader if the derivation could be made available in a supplemental document or file.
This would facilitate updates to the values should new data become available (i.e., the reader needs
to know what data were considered in the derivations to identify "new" data). The 1992 ORD report
cites Moore et al. (1980) as the basis for a K,, value of 4E-06 cm/hr (page 5-83). The 1998 guidance
document cites 5E-07 cm/hr for lead acetate, presumably based on Moore et al. (1980), although this
is not stated (Table 3.1). I attempted to derive this value from the Moore et al. (1980) and came up
with values ranging between 3.5E-07 - 4.9E-07 cm/hr, defending on which time points were
averaged. ,. s
3. Table 3.4
The value for the dermal absorption fraction of cadmium in soil is cited as 0.01,"from Wjester'et al.
(1992). However, Wester et al (1992) seems to support a values that rangefrpni 0.0002, - 0.0007..
4- Adjustment of Toxicity Factors
The oral pathway Hazard Quotient (HQ) is calculated'as:'' '"' ; ••; ;i-
"ral
Dfn
KJDoral
where I is intake and RfD is the oral dose that is not expected to produce' an adverse effect in
humans, given the associated uncertainties in our understanding of the toxicology of the chemical.
' ••'• • ' '<•:• ' ..•-•.- , •:'..,. * , ,- - .f -. ;=,-.,:.,.. :, •.-;•„;-...'.•, - f:<
G-3
-------
In the calculation of the dermal pathway HQ, intake (I), is replaced with uptake (U), the "absorbed
dermal dose". The issue at hand is the corresponding term that replaces the RfD:
Eq.2
HQdermat ~'
u
dermal
Two very different approaches are:
Eq.3
u.
dermal
RfD-AFH
Eq.4
HQoral =
U
dermal
NOAELSp-AFSp-UF
In Eq 3, the RfD, the "non-adverse intake in humans", is factored by an estimate of the oral
absorption fraction in humans (AFH). In Eq 4, the NOAEL estimate in the species (Sp) used in the
"critical study" is factored by an estimate of the absorption fraction in that species (AFSp) (and for
that experimental design).
These two approaches are conceptually very different and require different types of supporting data.
The data needed to support Eq 3 are data that would support a best estimate of the oral absorption of
the chemical in humans. The data needed to support Eq 4 are data needed to estimate the oral
absorption of the chemical in the bioassay (or in exposure scenario, if the RfD was based on a human
epidemiology study).
In my opinion, Eq 4 is problematic in the following ways:
A change in the "critical study" will require EPA to derive an AFSp for the new
study design (species, route, dose, etc).
The "critical study" provides a quantitative launching point for the estimate of the
NOAEL in the most sensitive human population (i.e., the RfD). This is usually
achieved by applying uncertainty factors to a LOAEL or NOAEL observed in a
bioassay or epidemiological study. These uncertainty factors account for
uncertainties in our understanding of interspecies and intra(human)species
variability in absorption. If we factor the NOAELSp with AFSpS we need to consider
an adjustment to the UF applied to the NOAELsp.
G-4
-------
Values for AFSp are not only chemical-specific, but they must be toxicity factor
specific. That is, if the oral toxicity and oral cancer bioassays are in different species
(or are of different exposure designs), different AFSp values may be needed in the
estimate of the dermal pathway HQ and cancer risk..
The values for AFSp recommended by EPA must not be confused with values for AFH
used to adjust the intake (I) in the estimate of HQs for exposure pathways not
represented by the RfD (e.g., using an RfD for water soluble arsenic to estimate the
HQ exposure to less soluble forms of arsenic in soil). I fear that the list of AFSps in
Table 4-1 of the dermal guidance will be erroneously interpreted as reasonable
estimates of AFH, they are not (at least this is true for the values presented for nickel
and cadmium).
5. Table 4.1 and Supporting Documentation
Given the above comments on toxicity factor adjustments, some additional comments on Table 4.1
follow:
The table should identify the toxicity factor (e.g., RfD, CSF) to which the adjustment
applies.
The document should provide a more complete evaluation of AF values for each
chemical, and should note the subset of the data set that is being used to estimate
AFsp. For example. Elakhovskay (1972) is the basis for an estimate of AFsp for
nickel, presumably because it is the only long-term oral dosing study in which nickel
absorption estimates could be derived (the estimate is based on urinary nickel
measurements made during a 6-month daily gavage study). However, there are at
least 10 other studies from which estimates of AF can be derived, seven of which are
studies in human subjects or populations (Sunderman et al., 1989; Cronin et al.,
1980; Christensen and Lagassoni, 1981; Gawdrodger et al., 1986; Menne et al., 1978;
Spuit and Bongarrts, 1977; Horak and Sunderman, 1977; McNeely yet al., 1972; Ho
and Furst, 1973; Jasim and Tjalve, 1986a,b; Tjalve and Stahl, 1984).
The table cites AF values for cadmium in food and water. However, there is really
no basis for unique values in the two media (Diamond et al., 1994); and the same can
be said for lead and cadmium (James et al., 1985; Maddaloni et al., 1998;
Sunderman et al., 1989). Whether the metal is in food or water makes little
difference. What does make a difference is when the metal is ingested with respect to
the ingestion of food; thus, the absorption of the metal in water will be high if it is
ingested after a fast that continues for several hours after the metal is ingested;
absorption will be lower if the metal is ingested in water right before, during or after
a meal. Presumably, the reason there are unique food and water AF values for
cadmium in Table 4.1 is because there are unique food and water RfDs for cadmium
on IRIS. The basis for the two RfDs is the assumption of higher absorption of
cadmium when it is ingested in water than when it is ingested in food; the rationale
for this assumption is similarly flawed, in my opinion. As an aside, there are four
G-5
-------
high quality studies of cadmium absorption in humans, in addition to the McLellan
et al. (1978; the latter citation is listed as McLellan, 1978 in Table 4.1): Flanagan et
al., 1978; Newton et al., 1984; Shaikh and Smith, 1980; Rahola et al., 1972).
A document that provides guidance on how to evaluate data on oral
bioavailability/gastrointestinal absorption, of similar depth and quality as the
dermal document, would be of great benefit to risk assessment.
References
Christensen, O.B. and Lagessoni, V. (1981) Nickel Concentration of Blood and Urine After Oral
Administration, Ann. Clin. Lab. Sci. 11, 119-125.
Cronin, R, Di Michiel and S.S. Brown, S.S. (1980) Oral Challenge in Nickel-Sensitive Women with
Hand Eczema, In: S.S. Brown, F.W. Sunderman, Jr. (eds.), Nickel Toxicity (Academic Press, New
York, 1980), pp. 149-152.
Elakhovskaya, N.P. (1972) On the Metabolism of Nickel Entering the Organism With Drinking
Water, Gig. Sanit. 6, 20-22 [Russian].
Gawkrodger, D.J., Cook, S.W., Fell, G.S. and Hunter, J.A.A. (1986) Nickel Dermatitis: The Reaction
to Oral Nickel Challenge, Br. J. Dermatol. 115, 33-38. ......
Flanagan, P.R., McLellan, J.S., Haist, J., Cherian, G., Chamberlain, M.J., and Valbergi L.S. (1978).
Increased dietary cadmium absorption in mice and human subjects with iron deficiency.
Gastroenterology 7, 841-846.
Ho W. and Furst, A. (1973) Nickel Excretion by Rats Following a Single Treatment, Proc. West.
Pharmacol. Soc. 16,245-248.
Horak E. and Sunderman, F.W. Jr., (1973) Fecal Nickel Excretion by Healthy Adults, Clin. Chem.
19,429-430.
James, H.M., Milburn, M.E. and Blair, J.A., 1985. Effects of Meals and Meal Times on Uptake of
Lead from the Gastrointestinal Tract of Humans. Human Toxicology, 4: 401-407, 1985.
Jasim S. and Tjalve, H. (1986a) Effect of Sodium Pyridinethione on the Uptake and Distribution of
Nickel, Cadmium and Zinc in Pregnant and Non-pregnant mice, Toxicol. 38, 327-350.
Jasim S. and Tjalve, H. (1986b) Effect of Zinc Pyridinethione on the Tissue Disposition of Nickel
and Cadmium in Mice. Acta. Pharmacol. Toxicol. 59,204-208.
Maddaloni, M., Lolacono, N., Manton, W., Blum, C., Drexler, J. and Graziano, J., 1998,
Bioavailability of soil-borne lead in adults by stable isotope dilution. Environ. Health Perspect., 106:
in press.
G-6
-------
McLellan, J.S., Flanagan, P.R., Chamberlain, M.J., and Valberg, L.S. (1978). Measurement of
dietary cadmium absorption in humans. /. Toxicol. Environ. Health 4, 131-138.
Menne, T.^Mikkelsen, H.I. and Solgaard, P. (1978) "Nickel Excretion in Urine After Oral
Administration," Cont. Dermatol. 4, 106-108. ;,,.,':
Moore, M.R., Meredithy, P.A., Watson, W.S., Sumner, D.J., Taylqr, M.K. and Goldberg, A. 1980.
the percutaneous absorption of lead-203 in humans from cosmetic preparations containing lead
acetate as assessed by whole-body counting techniques. Food Cosmet. Toxicol. 18: 399-405.
Newton, D., Johnson, P., Lally, A.E., Pentreath, R.J. and Swift, DJ., I9&t 'Tte u$iate by man of
cadmium ingested in crab meat. Hum. Toxicol. 3, 23-28. . >r ;; M;;,»
• ..;• ' •' ,!';,' ;•• . ;•.'!, •'• '>.;
Rahola, T., Aaran, R.-K., and Miettmen, J.K. (1972). Half-time studies of mercury and cadmium by
wholebody counting. In Assessment of Radioactive Contamination in Man, IAEA-SM-150/13, pp.
553-562. International Atomic Energy Agency, Unipublisher, New, York, , , ;
Ruoff, W.L., Diamond, G.L., Velazquez, S.F., Stiteler, W.M. and Gefell, D (f994). Bioavailability
of cadmium in food and .water: A case study on the derivation of relative bioavailability factors for
inorganics arid their relevance to the reference dose. Reg. Toxieol. Pharmacol 20, 139-160.
Shaikh, Z.A., and Smith, J.C. (1980). Metabolism of orally ingested cadmium in humans; In
Mechanism of Toxicity and Hazard Evaluation. B.Holmstedt, R. Lauwerys, M. Mercier, M.
Roberfroid, eds. Eisevier/North-Holland biomedical Press. PP. 569-574:
' ' . ' " ' ' ' ' ' ' . - • ' !
Spruit, D.and Bongaarts, P J.M. (1977) Nickel Content of Plasma, Urine and Hair in Contact
Dermatitis, Derrriatolpgiea 154, 291-300. •','•. "•". ''.'"'•",'..
Sunderman, W-F- Jr., Hopfer, S.M., Sweeney, K.R., Marcus, A.H., Most, B.M. and Creason, J.
(1989) Nickel Absorption :and Kinetics in Human Volunteers, Soc. Exp. Bibl. Med. 191, 5-1 1.
Tjalve, H. and Stahl, K, (1984) Effect of 5-Chloro-7-hydroxy-quinoline (Clioquinol) on the Uptake
and Distribution of Nickel, Zinc and Mercury in Mice, Acta Pharmacol. Toxicol. 55, 65-72.
Wester, R.C., Maibach, H.L, Sedik, L., Melendres, J., DiZio, S. and Wade, M. (1992) In vitro
percutaneous absorption of cadmium from water and soil into human skin. Fund. Appl. Toxicol. 19:
G-7
-------
Health Designs, Inc. . 183 Eaet Main Street . Rochestar, NY 14604 USA
716/546-1464 Vpice * 716/546-3411 Fax
Prom: Kurt BMMn (716) 540-1464#20 Vote*
SubJactDirmtl Exposure Workshop
D«tt : Dicembir 14, 1996
No- of Ptfles: cover
Tb: Or, Robert Bdife (202) 364-3273 Qffiea
US EPA
NCEA
401 MSOTitSW (6823)
Washington, DO 20460
(716) 84*3411 FAX
(202)
Dear Bob:
lummarizcd them fairly
seems obviau* that tt« model snoukl he redeveloped wlrti tofff* .
wbat more e««rive.bur more ifflporajitly,thcjn«5luUcddi^«lwJmcci
ihc Vecchia datibose doe» include some of Flynf s compounds.
There will remain the problem tbit the d«*b» will still be Waged towards certain chemical
groups . The only wiuticn to that problem ii tewing.
eev* a
wd hcrrinj .ft would be mr.»*~»*
G-8
-------
software • Clay Frederick mentioned the Web.
I also brought up the point that (here may be better log? estimators than the one being used
currently. I specifically cited a papw by Corabar and Ensjcin (no bias of course !} which
describes such an estimator.
Improving the model would reduce ihc variability of estimates, and result in more intelligent
regulation.
3. EPD. I was the fellow who suggested the technique in 1992. We have developed some
improvements, and the remit has been patented. The way the technique has been implemented is
not quite correct - otherwise all the compound from which the equation was developed would be
within the EPD. EPA staff used some approximations; I believe this should not be done.
Incidentally, we now have » patent on the implementation of the procedure.
A. 95% Confidence intervals. The intervals arc much too optimistic. They are based on the error
terms in the equation. They should be based on the assumption that the compounds to be
estimated art not pan of the same distribution as the design database. This requires the use of
unity in the Hat matrix, and will result in much wider intervals. They may look pretty bod, but
that's life.
5. Predicted vs. measured Kp. The argument in die draft ij that, because of the great variability of
measured Kp's, an estimate from the model would be more secure. I would agree with that, and,
in fact reinforce that judgment However, if the reasons for the large variation in measured
values could be identified, one nay wish to chnage the opinion. 1 know that Vecchia has
attempted to identify all the causes of variability, but I am tint convinced. I am nut convinced
because I have worked with many thousands of data points, and usually find that the reason for
deviations can be found in most instances.
6. What'* next? Toward the enduftfce meeting I tried to determine as tahow the suggestions
from the peer review group would be implemented, inasmuch as the EPA itaff that has been
responsible tor the draft is now scattered among th« regional offices. I am concerned that
inadquate resources will be available to implement the necessary changes. Do you have any
information as to this topic?
Best personal regards.
Kurt Hnsletn
VJP.. Technology
http://www,ojffl)oLconVprods/topkat
G-9
-------
-------
-------
------- |