Office of Pesticide Programs' Policy on
The Determination of The Appropriate
FQPA Safety Factor(s) For Use in The
Tolerance-setting Process:
RESPONSE TO PUBLIC COMMENTS
OPP Docket OPP-00610
Office of Pesticide Programs
U.S. Environmental Protection Agency
Washington, DC
February 28, 2002
-------
TABLE OF CONTENTS
LIST OF ABBREVIATIONS ii
INTRODUCTION 1
General FQPA Safety Factor Issues 2
Toxicology Issues 3
Exposure Issues 5
COMMENTS AND RESPONSES 7
I. General FQPA Safety Factor Issues 7
ISSUE 1. Applicability of FQPA Safety Factor to Conventional Chemistry
Pesticides 7
ISSUE 2. Weight-of-Evidence Approach 12
ISSUE 3. Overall Conservatism of OPP Risk Assessment Methodology 23
ISSUE 4. Relationship of FQPA Safety Factor to Traditional Uncertainty/Safety
Factors 37
II. Toxicology Issues 54
ISSUE 5. Definition of Core Toxicity Database 56
ISSUE 6. Approach to Calling in Neurotoxicity Data 61
ISSUE 7. Application of an FQPA Safety Factor Pending Receipt of Newly-
Required Studies 75
ISSUE 8. Criteria Indicative of Increased Concern for Pre- and Postnatal Hazard£6
ISSUE 9. Expectations about the Results of DNT Studies 92
ISSUE 10. Policy for Database Uncertainty Factors 96
ISSUE 11/12. Applying the FQPA Safety Factor Clause Relating to Potential Pre-
and Postnatal Toxicity 98
III. Exposure Issues 112
ISSUE 13. Conservatism of Exposure Assessments for Residues in Food ... 112
ISSUE 14. Conservatism of Drinking Water Exposure Assessments 120
ISSUE 15. Conservatism of Residential Exposure Assessments 125
ISSUE 16. Overall Conservatism of Exposure Assessments 130
ISSUE 17. Reliable Data 135
ISSUE 18. Policy or Rule 155
ISSUE 19. Additional Comments 165
REFERENCES 167
-------
LIST OF ABBREVIATIONS
ADME
Absorption, Distribution, Metabolism, and Excretion
APA
Administrative Procedures Act
BMD
Benchmark Dose
ChEl
Cholinesterase Inhibition
CWS
Community Water Systems
CSFII
Continuing Surveys of Food Intakes by Individuals
DCI
Data Call-In
DNT
Developmental Neurotoxicity Test
Db UF
Database Uncertainty Factor
EDos or 10
Effective Dose: central estimate on a dose associated with a 5% or a 10%
response adjusted for background
FQPA
Food Quality Protection Act
FFDCA
Federal Food and Drug Cosmetic Act
FIFRA
Federal Insecticide, Fungicide, and Rodenticide Act
FEL
Frank Effect Level
LOAEL
Lowest-Observed-Adverse-Effect Level
MOE
Margin of Exposure
MOS
Margin of Safety
MRM
Multiresidue Method
NAFTA
North American Free Trade Agreement
NAPIAP
National Agricultural Pesticide Impact Assessment Program
NASS
National Agricultural Statistics Service
NASQAN
National Stream Quality Accounting Network
NAWQPA
National Water Quality Assessment Program
NFCS
Nationwide Food Consumption Survey
NHEXAS
National Human Exposure Assessment Survey
NOAEL
No-Observed-Adverse-Effect Level
OP
Organophosphorus Pesticides
PDP
Pesticide Data Program
PBPK
Physiologically Based Pharmacokinetics
PHED
Pesticide Handlers' Exposure Database
POD
Point of Departure
PND
Post Natal Day
PAD
Population Adjusted Dose
aPAD
Acute Population Adjusted Dose
cPAD
Chronic Population Adjusted Dose
RfC
Reference Concentration
RfD
Reference Dose
SHEDS
Stochastic Human Exposure Dose Stimulation Model
ii
-------
SRM Single Residue Method
STORET Storage and Retrieval Database
SF Safety Factor
SOP Standard Operating Procedure
UF Uncertainty Factor
WOE Weight-of-evidence
Organizations:
AIHC American Industrial Health Council
AWWA American Water Works Association
CHPAC Children's Health Protection Advisory Committee
CDC Centers for Disease Control
CDPR California Department of Pesticide Regulation
CMA Chemical Manufacturers Association
FAO Food and Agriculture Organization of the United Nations
HED Health Effects Division, Office of Pesticide Programs, USEPA
HIARC Hazard Identification Assessment Review Committee
HESI ILSI's Health and Environmental Sciences Institute
I LSI International Life Science Institute
ICCVAM Interagency Coordinating Committee for the Validation of Alternative
Methods
IWG Industry Working Group
NAS National Academy of Sciences
NCEA USEPA's National Center Environmental Assessment
NHEERL EPA's National Health and Environmental Effects Research Laboratory
NIEHS National Institute of Environmental Health Sciences
NRC National Research Council
NRDC National Resource Defense Council
OCHP EPA's Office of Children's Health Protection
OECD Organization for Economic Cooperation and Development
OPP EPA's Office of Pesticide Programs
OPPT EPA's Office of Office of Pollution Prevention and Toxics
ORD EPA's Office of Research and Development
OPPTS EPA's Office of Prevention, Pesticides, and Toxic Substances
PMRA Health Canada's Pesticide Management Regulatory Agency
SAP Scientific Advisory Panel
USGS U.S. Geological Survey
iii
-------
WHO World Health Organization
iv
-------
INTRODUCTION
The Food Quality Protection Act of 1996 (FQPA) significantly amended the Federal
Food Drug and Cosmetic Act (FFDCA) under which EPA establishes "tolerances" (or
maximum legal limits) for pesticide residues in food. In deciding to establish, modify, or
leave a tolerance in effect, EPA must conclude that the tolerance is "safe." FFDCA sec.
408(b)(2)(A)(i). The term, "safe," means that "there is a reasonable certainty that no harm
will result from aggregate exposure to the pesticide chemical residue, including all
anticipated dietary exposures and all other exposures for which there is reliable
information." FFDCA sec. 408(b)(2)(A)(ii). The FQPA amendments specifically directed
EPA, in making the "reasonable certainty of no harm" determination to ensure that infants
and children are adequately protected. FFDCA sec. 408(b)(2)(C)(ii)(l). The FQPA
amendments also added the following:
In the case of threshold effects, for purposes of clause [FFDCA sec. 408(b)(2)(C)]
(ii)(l) an additional 10-fold margin of safety for the pesticide chemical residue and
other sources of exposure shall be applied for infants and children to take into
account potential pre- and postnatal toxicity and completeness of the data with
respect to exposure and toxicity to infants and children. Notwithstanding such
requirement for an additional margin of safety, the Administrator may use a different
margin of safety for the pesticide chemical residue only if, on the basis of reliable
data, such margin will be safe for infants and children.
FFDCA sec. 408(b)(2)(C). The provision quoted above is referred to as the "FQPA Safety
Factor provision."
In March 1998, the U.S. Environmental Protection Agency (EPA) established an
Agency-wide "10X Task Force" to address the implementation of the FQPA Safety Factor
provision. Task Force members included high-level scientists primarily from the Office of
Prevention, Pesticides and Toxic Substances, the Office of Research and Development,
the Office of Solid Waste and Emergency Response, and the Office of Children's Health
Protection. The 10X Task Force formed two working groups-the Toxicology Working
Group and the Exposure Working Group. These groups each drafted a report containing
recommendations for the implementation of the FQPA Safety Factor in their respective
areas of review: Toxicology Data Requirements for Assessing Risks of Pesticide
Exposure to Children's Health (US EPA 1999a) and Exposure Data Requirements for
Assessing Risks from Pesticide Exposure of Children (USEPA 1999b).
-------
The Office of Pesticide Programs (OPP) is responsible for implementing the
requirements of FQPA in making its pesticide regulatory decisions. Accordingly, OPP
developed interim guidance, which has been updated periodically since 1996, as to how it
will comply with FQPA concerning the FQPA Safety Factor for protecting infants and
children. In drafting the current guidance, OPP took into account the recommendations of
the 10X Task Force as embodied in the above-mentioned documents. OPP's guidance
consists of two revised documents: The Office of Pesticide Programs' Policy on
Determination of the Appropriate FQPA Safety Factor(s) for Use in the Tolerance-
Setting Process; draft version (USEPA 1999c) and Standard Operating Procedures for
the Health Effects Division FQPA Safety Factor Committee; draft version (USEPA
1999d). The former paper explains the general policies that OPP proposed to follow in
making determinations concerning the use of the FQPA Safety Factor, while the latter
paper specified the detailed procedures that OPP would use in following these policies.
On July 8, 1999, the Agency published a Notice in the Federal Register announcing the
availability of, and opportunity to comment on, the four documents relating to EPA's
implementation of the FQPA Safety Factor provision (64 £R 37001). Copies of the
documents made available for public comment also appear at:
http://www.epa.gOv/fedrastr/EPA-PEST/1999/Julv/Dav-08/p17315.htm. EPA subsequently
extended the original 60-day comment period by 30 days to October 7, 1999 (64 FR
48617). EPA specifically invited the public to comment on sixteen questions, grouped into
four broad categories:
General FQPA Safety Factor Issues
Question 1. The OPP guidance indicates that OPP will generally apply the FQPA
Safety Factor only to food-use pesticides of "conventional" chemistry. Please
comment on this approach. The guidance also indicates that different decisions
about the need for, and size of, an additional FQPA Safety Factor may be
appropriate for different durations of exposure and different exposed populations.
Please comment on this approach. Finally, the guidance indicates that it would be
appropriate to make only one FQPA Safety Factor decision for a single
population/exposure period, even though such exposure might occur by different
routes and pathways. Please comment on this approach.
2
-------
Question 2. Is a weight-of-the-evidence approach to making FQPA Safety Factor
decisions appropriate, taking into consideration the toxicology and exposure
databases for a pesticide and the potential risks for the developing fetus, infant and
child as well as other populations? If not, why not? Given the scope of the evidence
which OPP intends to consider, are there any other types of information that OPP
should consider in making its FQPA Safety Factor determinations?
Question 3. Do you agree with the view that the models and assumptions used by
OPP in the risk assessment process, together with reliable data available on
specific pesticides and other reliable, empirical data, typically do not understate
risk? If not, under what circumstances do you believe OPP's current approaches to
assessing risks from aggregate exposure to a single pesticide produce risk
assessments that understate the risks to infants and children?
Question 4. Do you agree with OPP's view that the FQPA Safety Factor should be
applied in addition to the interspecies and intraspecies uncertainty factors, but that
the FQPA Safety Factor should not be applied in a manner that results in
"double-counting" of uncertainties that are otherwise addressed in the toxicity and
exposure assessments through, for example, the database uncertainty factor or
conservative exposure models? If you disagree, why?
Toxicology Issues
Question 5. Please comment on OPP's proposed criteria for defining the core
toxicology database.
Question 6. After having considered the recommendations from the FIFRA
Scientific Advisory Panel and the Toxicology Working Group, OPP is beginning the
process of calling in data for three studies (the acute and subchronic neurotoxicity
studies in adult mammals and the developmental neurotoxicity study) for a subset of
conventional chemistry food-use pesticides known neurotoxicants. In addition, OPP
will be proposing to require the same set of studies for all conventional chemistry
food-use pesticides in the revision of the 40 CFR 158 regulations. Please
comment on this two-stage approach.
3
-------
Question 7. The OPP policy guidance indicates that one of the critical issues is
whether or not to apply an FQPA Safety Factor pending receipt of newly-required
studies. There are a variety of possible approaches. One possible approach
would be to apply the FQPA Safety Factor's database uncertainty component to
gaps related to new core data requirements only where there are specific concerns
regarding the pesticide pertaining to the data requirement. Alternatively, OPP
could apply the default 10X factor (or some other additional factor) whenever a new
data requirement is added and/or whenever a testing guideline is changed. Please
explain how you think the FQPA Safety Factor provision should be implemented
when OPP makes such changes. In commenting, please address whether OPP
should apply the default FQPA 10X factor, some different yet additional factor, or no
factor at all in the following circumstances:
¦ A minor change to testing guidelines.
¦ A major change to testing guidelines.
¦ An addition of a new required test.
¦ An addition of a new required test to core requirements.
Question 8. In the absence of the results from any of the studies to be required
through Data Call-In (DCI) Notices (i.e., the acute and subchronic neurotoxicity
studies in adult mammals and the developmental neurotoxicity study), what
information from existing studies on a specific chemical would increase or
decrease the concerns about the potential for prenatal and postnatal hazard, in
general, and for neurotoxicity and developmental neurotoxicity, in particular?
Which, if any, of the seven criteria discussed in Section V.A.1 .a., footnote 4 and
associated text of the OPP policy document is appropriate forjudging whether
there is increased concern about the potential for a pesticide to cause
developmental neurotoxicity? Are there any other criteria which would be useful for
informing this judgment?
Question 9. Please comment on whether you would expect that developmental
neurotoxicity studies would, for a substantial number of chemicals, identify effects
that are not detected in other studies and more fully characterize the potential risks
of exposures during development. In addition, please comment on the sensitivity of
these tests vis-a-vis other studies required and used for age-related comparisons
for acute, intermediate, or chronic RfD derivation (e.g., prenatal developmental
toxicity or multi-generation reproduction study, subchronic and chronic studies, etc.).
Please explain the basis of your opinion.
4
-------
Question 10. OPP's Guidance states that currently five studies (a multi-generation
reproduction study, prenatal developmental toxicity studies in two species, and
chronic toxicity studies in a rodent and non-rodent species) comprise the toxicity
database necessary to produce a "high confidence RfD," and that some additional
database uncertainty factor will be imposed if the database on a pesticide lacks
one or more of these studies. OPP proposes to expand this core database to
include the subchronic neurotoxicity study. Eventually, OPP also includes the acute
neurotoxicity study in adult mammals and the developmental neurotoxicity study,
once these studies have met the criteria for inclusion in the core toxicity database.
Please comment on OPP's proposed approach to imposing a database uncertainty
factor of 3X if one key study is missing from the database and a factor of 1OX if
more than one is missing.
Question 11. OPP is proposing to adopt the framework and its criteria/factors for
assessing the degree of concern about the potential for prenatal and postnatal
effects as recommended by the Toxicology Working Group. Please comment on
the appropriateness of the proposed criteria/factors for use in this assessment
process, and OPP's proposed approach for accommodating its concerns in the
Reference Dose derivation and FQPA Safety Factor decision processes, in the
near term, and in the longer term.
Question 12. When the hazard to infants and children is well-characterized, and
the data show that infants and/or children are more susceptible than adults, under
what circumstances, if any, should this information lead OPP to employ an
additional Safety Factor?
Exposure Issues
Question 13. Subject to the qualifications expressed in the OPP policy document
and the report from the Exposure Working Group, OPP believes that each of the
Tiers for estimating exposure to a pesticide through food, in almost all instances,
will not underestimate exposure to infants and children. Please comment on this
conclusion, as it applies to each of the Tiers.
Question 14. OPP is developing a tiered approach to assessing the likelihood
and magnitude of contamination of drinking water and its sources by a pesticide.
As an interim approach, when direct assessment is not possible, is it reasonable
and protective to regard the estimates generated by OPP's current screening
methodology as upper bound pesticide concentrations for surface and ground
water and to assume that this concentration generally will not be exceeded in
drinking water?
5
-------
Question 15. OPP is developing approaches to assess the likelihood and
magnitude of exposure to pesticides in residential and other non-occupational use
scenarios. When direct assessment is not possible, is it reasonable and protective
to regard the estimates of exposure for the major residential and other
non-occupational exposure use scenarios developed by OPP as upper bound
estimates of the exposure received by infants and children from such use?
Question 16. In OPP's view, its aggregate exposure assessments generally do
not underestimate the exposure to infants and children because the aggregate
exposure is calculated by adding the high-end estimates of exposure to pesticides
in food, to the high-end estimates of exposure to pesticides both in water and as a
consequence of pesticide use in residential and similar settings. Please comment
on this view.
Although EPA made four documents concerning the FQPA Safety Factor provision
available for review and public comment, the Agency encouraged the public to focus
particularly on the OPP Policy document. OPP also noted that, while it used the two
papers produced by the Toxicology and Exposure Working Groups of the Agency's 10X
Task Force in developing its guidance, the 1OX Task Force was not planning to revise and
reissue these documents following public comment. In addition, since the OPP Standard
Operating Procedure was largely derived from the OPP guidance document, any changes
in it following public comment should reflect changes in the revised guidance document.
OPP received about 825 public comments in response to the Notice of Availability.
The comments came from a wide range of organizations and individuals interested in
pesticide regulation including representatives of pesticide companies; organizations
representing growers and other pesticide users; academicians and consultants; public
health, environmental, animal welfare, and children's advocacy groups, as well as from
foreign and state governments. The Agency also presented earlier drafts (which did not
differ substantively from those that were made available for public comment) of the four
documents for review by its FIFRA Scientific Advisory Panel (SAP), which also submitted
comments (USEPA 19991).
EPA has reviewed all of the comments and has grouped similar comments together.
The remainder of this document contains EPA's summary of the comments and its
responses to the comments. The comments are generally organized to follow the sixteen
questions contained in the original Notice of Availability; the sequence of the questions has
been changed to a slight degree to make the Agency's reasoning easier to understand.
6
-------
COMMENTS AND RESPONSES
OPP received over 800 comments on this draft policy. These comments, along with a
listing of each commentor and their affiliation, can be found at the FIFRA Docket which is
located at: Crystal Mall #2, Room 119; 1921 Jefferson Davis Highway; Arlington, Virginia
(http://www. epa. aov/pesticides/docket/).
I. General FQPA Safety Factor Issues
ISSUE 1. Applicability of FQPA Safety Factor to Conventional
Chemistry Pesticides
The OPP guidance indicates that OPP will generally apply the FQPA Safety Factor
only to food-use pesticides of "conventional" chemistry. Please comment on this
approach.
Comment 1. A number of commentors (369, 761, 771, 773 and 776) questioned
the proposal that the FQPA Safety Factor would generally apply only to food-use
pesticides of "conventional" chemistry, apparently, in some cases, interpreting the
proposal to say that only this category of substance would be covered and offering
the opinion that this was too narrow a designation.
Response. EPA, in proposing that the need for making an FQPA Safety
Factor decision would generally apply to food-use pesticides of conventional
chemistry, meant those pesticides having a defined chemical structure. For the
purpose of FQPA and the scope of the 10X policy, the term "pesticide" covers
both active and other (i.e., inert) ingredients. The Agency had proposed that it
would be possible to make an FQPA Safety Factor decision only in those cases
where the required and necessary toxicology database would be sufficiently
robust to allow or support the derivation of a hazard value, such as an acute or
chronic reference dose (RfD), arguing that without such a hazard value, it would
be inappropriate to conduct a quantitative safety factor analysis (i.e.,
determination to retain, remove, reduce, or increase the 10X FQPA Factor).
It was not the Agency's intention to constrain the FQPA Safety Factor
decision process only to pesticides of "conventional chemistry," meaning only
those with a defined chemical structure. At the time, OPP did believe, however,
that the toxicity database for such chemicals would best lend themselves to
quantitative FQPA Safety Factor decisions. Upon further reflection, and with
additional experience, the Agency has realized that it, in fact, can and should
(and, has) made FQPA Safety Factor decisions for a much broader range of
7
-------
pesticides (both active ingredients and others). A sizeable proportion of these
decisions will be only "qualitative" in nature (i.e., an evaluation of available
information to determine the hazard concern for the young); others would be
"quantitative," in that RfD or RfC values would be derived, then modified
upward, downward or left unchanged, depending upon the nature and fullness of
the available information.
In the U.S., at the present time, there are nearly 1000 pesticides registered
as active ingredients and about 2500 pesticides registered as "other"
ingredients. Food use pesticides, both actives (about 450 in number) and
others, belong to many chemical classes. Some are mixtures. Some
pesticides are registered both as an active and as an inert ingredient, albeit
probably in different formulations (e.g., phenol and formaldehyde). Others are
either whole foods or naturally-occurring constituents of food (e.g., soybean oil,
ethylene and l-glutamic acid) or are food additives (e.g., oil of orange).
Examples of classes more traditionally thought of as pesticide active
ingredients are the organophosphorus and pyrethroid insecticides, the triazine
herbicides, and the conazole fungicides. There also are several classes of
biopesticide active ingredients (e.g., biochemicals such as formic acid;
pheromones; plant growth regulators; natural insect growth regulators; and
microbial pesticides such as bacteria and fungi). Most of the roughly 2500 other
ingredients are divided up into four lists, categorized primarily by known or
expected toxicity potential (List 1 having the potential for being the most
hazardous and List 4 being the least, with List 3 substances being largely
undefined as to their hazard potential). Tolerance exemptions have been
granted for many actives and most inerts.
The chemical characteristics and anticipated toxicity potential of the
pesticide (along with the proposed use pattern) dictate the kinds of toxicity data
that would be needed to characterize its hazard profile. For those categories of
food-use pesticides for which only a minimum toxicity database is deemed
necessary, a "reasonable certainty [of] no harm" and the FQPA Safety Factor
finding would be accomplished only in the qualitative sense, particularly for
those for which an exemption from a tolerance would be granted. The Agency
believes that there are many examples of substances that might be subjected
only to a qualitative finding, such as the active components in plant-incorporated
protectants, microbial and some other biopesticides, as well as many inert
ingredients.
8
-------
Comment 2. One commentor (369) did not feel it is appropriate to limit the risk
assessment policies outlined in the policy only to food-use pesticides, stating that
they felt it is inappropriate to set two different sets of safety standards for the
protection of children-one for food use pesticides and one for nonfood-use
pesticides. They also believed that the same considerations should be applied to
indirect exposures of fetuses and infants which may occur via pregnant or nursing
women who may be occupationally exposed to either food-use or nonfood-use
pesticides.
Response. The 1996 amendments to FFDCA (the tolerance-setting
provisions) state that the Agency shall assess risk to infants and children and
consider the FQPA 10X Safety Factor when "establishing, modifying, leaving in
effect, or revoking a tolerance or exemption for a pesticide chemical residue..."
Thus, under FFDCA section 408, the 10X Safety Factor provision is only
applicable to those pesticide uses that need tolerances. Under FIFRA, the
FFDCA section 408 safety standard, including the 10X Safety Factor provision,
is only applicable to pesticide uses posing a dietary risk due to residue in or on
food. Pesticides having only nonfood uses are not subject to these provisions.
In addition, the FFDCA provisions state that, in assessing the aggregate
exposure to the pesticide residue, aggregate exposure is considered to include
dietary sources and other nonoccupational sources. The assessment of the
risk that may accrue from occupational exposures should be conducted
separately, under the provisions of FIFRA. The applicable FIFRA standard is
that the use must not cause "any unreasonable risk to man or the environment,
taking into account the economic, social, and environmental costs and benefits
of the use..."
Comment 3. Commentor 781 disagreed with the exclusion of consideration of
exposures to fetuses of pregnant farmworkers in the risk assessment process for
tolerance-setting. Exposures to these unborn children are not "occupational" and,
hence, must be considered in determining the applicability of the 10X safety factor.
Congress expressed concern for all children's pre- and postnatal development,
without excluding any prenatal exposure which may have occurred in the course of a
parent's employment. Indeed, section 408(b)(2)(C) of the Act makes no mention of
excluding occupational exposures. Since Congress expressly excluded
occupational exposures in other provisions, the absence of any "occupational"
exclusions here should be regarded as dispositive of Congress' intent to fully
protect all children.
9
-------
Response. EPA believes that, for the purpose of implementation of the FQPA
statutory language as it pertains to tolerance-setting, any exposures to the fetus
or the child that result from its parent's employment are occupational exposures.
To adopt the interpretation suggested by the Farmworker Justice Fund would
essentially read the explicit limitation on considering occupational exposure out
of the statute. Further, the limitation in section 408(b)(2)(D)(vi) regarding
occupational exposures does apply to assessing children's risk to pesticides
under tolerances because section 408(b)(2)(D)(vi) explicitly applies to all
tolerance actions under the FFDCA.
Comment 4. Commentor 778 argued that FQPA mandates that EPA consider
identifiable, highly exposed/sensitive subpopulations in making its statutory
determinations. The commentor believes that farm children should be designated
as such a subpopulation, and their exposures and risks should be separately
evaluated under FQPA. Citing a 1998 NRDC petition and report "Trouble on the
Farm," the commentor noted that the subgroup includes 320,000 children under the
age of six, and hundreds of thousands of children who play or attend schools on or
near agricultural land or have family members who work on farms. There are
approximately one million children of adult farmworkers living in the United States.
These children are exposed in ways other children generally are not and their
exposures are higher than other children. Recognition of the unique exposures and
risks of farm children is also consistent with environmental justice concerns for
minorities and low income groups.
Response. By definition, there are subpopulations which may be more highly
exposed or more sensitive to the effects of exposure than the average. Farm
worker/farm family children may fall into this category by meeting one or both
criteria. While EPA has not formally designated this group as a special
category, it has been working to better understand the nature and magnitude of
exposures that this group experiences, and, where data show differences, EPA
will incorporate such information into chemical-specific aggregate exposure and
risk assessments, as appropriate. This consideration would extend to
cumulative risk assessment, should the chemical(s) under evaluation be
deemed to share a common mechanism of toxicity. Overall, however, OPP
believes that its exposure and risk assessment methodologies are
conservative, i.e., do not generally underestimate exposure, for this subgroup.
See also responses to Issue 16 under Comments 5 and 7.
10
-------
The guidance also indicates that different decisions about the need for, and size of,
an additional FQPA Safety Factor may be appropriate for different durations of
exposure and different exposed populations. Please comment on this approach.
Comment 1. Two commentors (773 and 369) expressed support for the policy of
determining the need for, and size, of the FQPA Safety Factor for different
durations of exposure and different exposed subpopulations.
Response. These commentors agreed with the Agency approach and the
Agency continues to believe that this approach is appropriate and necessary.
Finally, the guidance indicates that it would be appropriate to make only one FQPA
Safety Factor decision for a single population/exposure period, even though such
exposure might occur by different routes and pathways. Please comment on this
approach.
Comment 1. Both commentors on this question (773 and 369) disagreed with the
Agency's proposal of making only one FQPA Safety Factor decision for each
population/exposure period combination. Both noted that there may be different
endpoints of toxicity and differing qualities of the toxicity and exposure databases
for the three possible routes of (aggregate) exposure (oral, dermal, and inhalation).
They argued that, in light of these possibilities, it may be necessary (and, more
appropriate) to make a different decision regarding the need for/size of the FQPA
safety factor, on a route-specific, not an aggregate, basis.
Response. The Agency has identified three major subpopulations (i.e., infants,
children, and women of child bearing age) that may require FQPA Safety Factor
decisions. These three subpopulations were characterized, in part, because it
was expected that different endpoints of toxicity might be of importance to each
of these groups. Thus, tailored decisions for these subpopulations with regard
to the nature and quality of the toxicity database and with regard to these
specific endpoints would be addressed in the FQPA Safety Factor decision
process.
The Agency disagrees with the commentors on how to handle the exposure
questions. The Agency has interpreted as its mandate to the conduct of
exposure assessment in the aggregate and, therefore, also to judge the nature
and quality of the exposure database in the aggregate. Furthermore, it is
indicated in OPP's guidance document on aggregate exposure risk
assessment (USEPA 2001 b) that only similar toxicological effects found for
11
-------
different routes of exposure will be combined or aggregated. If the process is
conducted properly, the decision outcome should be the same as if one were to
make separate findings on a route-specific basis and then integrate these
separate findings into an overall finding or on an aggregate basis from the
beginning. In either case, if a deficiency were to be identified for one or more
route/pathway, it should be highlighted and both a qualitative and quantitative
judgment made as to the impact of this deficiency on the exposure assessment
as a whole.
This particular discussion of decision-making in risk assessment brings to
light how very important it is to develop a clearly-articulated, transparent risk
characterization that includes a thorough description of the decision logic used
to reach a conclusion, not simply a set of numbers which lack a context for their
development.
ISSUE 2. Weight-of-Evidence Approach
Is a weight-of-the-evidence approach to making FQPA Safety Factor decisions
appropriate, taking into consideration the toxicology and exposure databases for a
pesticide and the potential risks for the developing fetus, infant and child as well as
other populations? If not, why not?
Comment 1. Five commentors agreed that a weight-of-evidence (WOE) approach
was appropriate in making FQPA Safety Factor decisions, but recommended that
the Agency provide additional clarification and guidance regarding this concept and
its application (L015, 369, 372, 773, 775). One of these commentors (L015)
recommended that examples be provided to demonstrate how the toxicity and
exposure information can be integrated in arriving at the safety factor decision for a
pesticide. Additionally, the SAP supported a weight-of-evidence approach, since
experience has shown the WOE approach to be especially useful when addressing
complex issues such as those involved in this determination. They indicated that
the full range of data and evidence should be considered in making safety factor
decisions. The Panel did, however, express the need for definition of the term
"weight-of-evidence" and for more clearly defined descriptions of the conditions of
uncertainty that might lead to the application of safety factors of different
magnitudes. Two commentors did not believe that a weight-of-evidence approach
adequately satisfies the statutory language of FQPA (761/771, 778).
12
-------
Response. EPA continues to believe that a weight-of-evidence approach is
the most appropriate approach in determining the need for an FQPA Safety
Factor.
The weight-of-evidence approach to data evaluation is a general scientific
concept that promotes an integrative approach rather than item-by-item
consideration of all information pertinent to the question or issue being
addressed. It involves an assessment of the quality, adequacy, and consistency
of the data and an integration of the scientific data and conclusions, with
combined input from all relevant disciplines. The weight-of-evidence approach
recognizes that initial views of one kind of evidence may change significantly
when other information is brought into the interpretation. A key element of a
weight-of-evidence approach is that no single piece of information determines
the overall conclusion since all data are judged in combination.
The application of a weight-of-evidence approach is described in various
Agency documents that address the evaluation of linear and nonlinear toxicity
data that are used in assessing risk from chemical exposure. For example, an
online Integrated Risk Information System (IRIS) document (USEPA 1999m)
addresses the Reference Dose (RfD). A subsection (1.3.1.2 Weight-of-
Evidence Determination) states that WOE is the culmination of the hazard
identification step, and that it "summarizes the highlights of the information
gleaned from the principle and supportive studies...to determine the extent to
which a consistent, plausible picture of toxicity emerges." Similarly, the
Agency's draft Guidelines for Carcinogenicity Risk Assessment contains a
chapter (2.6 Weight-of-Evidence Evaluation for Potential Human
Carcinogenicity), which defines a WOE evaluation as "a collective evaluation of
all pertinent information so that the full impact of biological plausibility and
coherence is adequately considered" (US EPA 1999f). The IRIS background
document further provides a list of factors which could add to the weight-of-
evidence that the chemical poses a hazard to humans, while the draft cancer
guideline provides separate lists of factors (for human data, animal data, and
other data, as well as for the weight-of-the-totality-of-evidence) that can be used
to increase or decrease the weight. An example of the application of a weight-
of-evidence approach as it relates to a specific body of data can be found in the
OPP policy entitled The Use of Data on Choiinesterase Inhibition for Risk
Assessments of Organophosphorous and Carbamate Pesticides (US EPA
2000a).
13
-------
These background documents are useful as references that aid in the
general understanding of a weight-of-evidence approach to data evaluation.
However, none of these documents provides guidance that is directly applicable
to the WOE approach used in determining the FQPA Safety Factor(s). This is
because the FQPA-related weight-of-evidence approach includes consideration
not only of toxicity data, but of all the information on both hazard and exposure in
combination, to arrive at an integrated conclusion. This analysis involves close
scrutiny of all available data, the study results, and data interpretation; it also
incorporates carefully considered judgements regarding the completeness and
adequacy of the databases as well as the quality and reliability of the data.
It is not possible to define every combination and permutation of toxicity and
exposure information that might be evaluated in the course of reaching
conclusions regarding the need for the FQPA Safety Factor. However, the
revised Standard Operating Procedures for the Health Effects Division FQPA
Safety Factor Committee (USEPA 1999dJ will be used by OPP as a
companion tool in the implementation of the WOE assessment. This document
will provide guidance in the form of a list of questions which address in great
detail the completeness of both the exposure and hazard information and any
evidence of increased risks to infants and children. Illustrative examples appear
in the SOP in order to provide further guidance regarding the manner in which
the toxicity and exposure information can be integrated in arriving at the safety
factor decision for a pesticide.
In determining the FQPA Safety Factor, the weight-of-evidence, the level of
confidence, and residual uncertainties are used in interpreting the data that
Agency scientists have already assessed to be reliable (or, not, as the case
may be) and useful in the risk assessment. Given the complexity of this
process, however, EPA agrees with the commentors that the decision process
would benefit from the inclusion of examples and therefore is adding such
material to the SOP document. EPA believes that the WOE approach, as
described, is consistent with the language and intent of FQPA.
Comment 2. One commentor (775) suggested that EPA's criteria for determining
whether sufficient information exists for determining whether the FQPA factor
should be applied "indicate a move towards a checklist approach," notwithstanding
the expressed intent to use a WOE approach.
Response. The Agency reiterates its intent to use a weight-of-evidence
approach to its decisions about the FQPA Safety Factor. OPP does not
provide a prescriptive list or a "checklist" of criteria but rather the revised
guidance document generally describes factors and types of data that should be
14
-------
considered in a weight-of-evidence analysis for evaluating the adequacy of
information in characterizing risks to infants and children in order to guide the
risk assessor in addressing a variety of different situations that might be
encountered.
FIFRA Scientific Advisory Panel Comments
Comment 3. The SAP provided some assertions regarding the legal and policy
aspects of the use of the FQPA 10X factor, defining the 10X factor as a comfort or
policy factor, with minimal scientific basis.
Response. EPA does not agree with the assertion that there is minimal
scientific basis to the determination of the FQPA factor. The revised guidance
document clearly describes a process which is extensively informed by the
science, including the weight-of-evidence assessment of the toxicity and
exposure data for the pesticide under review. EPA considers statements by the
SAP that purport to interpret the meaning of statutory provisions to be beyond
the SAP's expertise.
Comment 4. The SAP recommended that the Agency consider published peer
reviewed reports in the "open" literature in the weight-of-evidence assessment, and
stated that a transparent characterization and usage of database uncertainty factors
is needed when data of these types are taken into account in the risk assessment
process. The Panel suggested that an SOP be developed for assessing available
peer-reviewed literature reports of toxicity studies that fall outside currently required
toxicology data sets. Additionally, it was recommended that the SOP might
address the acquisition, evaluation, and weighting of peer-reviewed animals
studies in the literature, as well as for human epidemiological data concerning the
potential for health effects to occur following inadvertent/incidental chemical
exposure.
Response. Current OPP practice includes the evaluation of information from
the peer-reviewed literature, when appropriate and available, and this
information is considered in the FQPA Safety Factor determination. The
revised SOP will include a question regarding this topic, and OPP expects that
the peer-reviewed literature will be searched and reviewed for pertinent
information on toxicity and exposure (including human data, which might consist
of case studies following accidental exposures and/or epidemiological studies)
as standard practice in the development of a risk assessment. However, OPP
recognizes that there is a lack of formalized procedures and process regarding
this issue. As with any study, the quality of data should be taken into account.
15
-------
An SOP is currently being developed in HED to address the use of peer
reviewed literature in pesticide risk assessment, and the recommendations
from the Panel will be helpful.
Comment 5. The SAP stated that OPP's decision logic is ambiguous, and that it
is impossible to understand the relation between available evidence, its
interpretation by experts, and the choice of a specific FQPA Safety Factor. It was
suggested that the decision logic should flow from answers to a series of questions,
two examples of which were provided by the Panel. In the first example, the
questions related to the availability of specific toxicity and exposure data; in the
second example, the questions related to what data or conservative default
assumptions would be needed to "relieve the 1OX safety factor."
The Panel concluded that the Agency should define assumptions that it will
adopt and apply in the absence of perfect information, recommending that the
Agency should approach the problem in the following sequence:
Judgment regarding data sufficiency and quality.
Judgment regarding application of conservative assumptions.
Judgment regarding application of additional safety factor.
If the Agency concludes that data are insufficient or of poor quality, it has two
options: (a) apply conservative default assumptions to estimate risk, or (b) do not
apply conservative assumptions, but instead apply an additional safety factor.
The Panel recommended that relationships between any assumptions applied
to uncertain information and the choice of specific safety factors should be explored
more fully in case studies that explore the most difficult regulatory situations, e.g.,
where there must be a choice regarding the management of a pesticide that is
registered for release to diverse environmental media, and for pesticides that act
via a common mechanism with other chemicals. These examples would need to
fully identify different sources of uncertainty, and then openly consider how its
assumptions account for this uncertainty. The relations between uncertainty, default
assumptions, and the choice of safety factors should be described.
Response. The revised SOP, which will be used as a tool in the process of
determining the FQPA Safety Factor for each individual chemical being
reviewed, will contain a series of questions regarding the nature and robustness
of the available toxicity and exposure data. These questions are being carefully
crafted to address, as completely as possible, all aspects of the information that
could potentially contribute to the derivation of the FQPA Safety Factor(s) for
that pesticide. EPA believes that this SOP will be more extensive than the
examples provided by the Panel, and that a change in approach to this aspect
16
-------
of the decision logic is not warranted.
The recommended approach to the problem of lack of sufficient information
on a chemical, and the logic sequence that the Panel provided, is consistent
with the approach OPP currently utilizes. The SOP will be expanded to include
basic examples to illustrate the decision logic used in the determination of the
FQPA Safety Factor. Every attempt was made to incorporate the information
recommended by the Panel; nevertheless it is noted that most of the detailed
logic that supports the default assumptions used to address recognized sources
of uncertainty are extensively addressed as generic issues in other peer-
reviewed OPP policy papers and documents, such as the residential exposure
SOPs (USEPA 1999g), the common mechanism of toxicity guidance (USEPA
1999h), the aggregate (USEPA 2001 b) and the cumulative risk assessment
guidance (USEPA 2002) documents. As a matter of practice in OPP, individual
chemical FQPA Safety Factor assessment documents make reference to
these.
Comment 6. The Panel encouraged the Agency to formally revisit and review the
core toxicology database every few years to ascertain if it is adequate, inadequate,
or contains redundant or useless requirements. The Agency was also encouraged
to examine its testing guidelines and where possible attempt to combine protocols
to save animal and financial resources. The Panel recognized that the Agency has
plans in this regard but wanted to further encourage and emphasize the need for
this action.
Response. OPP is moving forward with efforts that are addressing these
issues. Proposed revisions to 40 CFR 158, which includes the list of studies
included in the core toxicology database, are being developed; a draft of
Subpart W, which is a revised set of data requirements proposed specifically for
antimicrobial pesticides, is soon to be released for comment. It should be
noted that the Agency encourages the use of combined protocols for testing.
While this approach is useful for first pass testing, it is more difficult to attain in
the case of follow-up testing of reregistration/registration renewal. Thus, there
also remains a need for stand-alone protocols to accommodate these
situations. The Agency is, however, facilitating discussions (e.g., relative to
development of the Endocrine Disruptor Screening Program) that will result in
enhancement of endpoints in the current guideline studies, and, thus, more
efficient utilization of animals on test.
Comment 7. Several members of the Panel expressed the opinion that improved
17
-------
methods of neurotoxicity testing and validation of the assumptions regarding
children's exposure would ultimately make the weight-of-evidence approach a
stronger tool for risk assessment purposes. However, the Panel was concerned
that, at present, too many gaps in the available databases exist in order to be
confident in decisions made under this approach. Examples were provided of
gaps in the knowledge regarding exposure and toxicity information applicable to
children.
Response. OPP agrees with the Panel that, in some cases, there are gaps in
the understanding of both the toxicity of a pesticide to children or the exposure
of children to a pesticide. The Agency also agrees that additional data to
evaluate potential toxicity and exposure could be valuable. (Obviously, as
science and policy discussions advance, they will be incorporated into the
process, as appropriate.) The significance of the data gaps, however, will vary
from chemical to chemical and will depend, among other things, on the
assumptions that OPP makes in the absence of data. Thus, OPP believes that
its level of confidence in its decisions should take into account the full range of
information available about each chemical. The Agency does not agree with a
general assertion that the extent of its understanding of the risks of pesticides to
children is so limited that it lacks any confidence in its decisions.
Comment 8. The Panel provided comment on three specific issues that could
influence the weight-of-evidence analysis and FQPA Safety Factor determination:
First, the Panel expressed the opinion that the discussion of dose-response
slopes and their use in the interpretation of concern for the effects that may be seen
at lower doses, while statistically simple, is aimed in the right direction and
suggests improved methods of analysis for noncancer endpoints. However, they
also stated that the methods proposed provide only very limited evaluation of one
very conservative issue, the assumption that noncancer endpoints have thresholds.
Statistical methods exist for the evaluation of the shapes of dose-response curves
that can provide objective information that would be useful in evaluating this
hypothesis. While one can never get a definitive answer of whether a threshold
exists or not, one can estimate the appropriate concern for the possible lack of a
threshold. By applying methods which directly evaluate shape, this assumption can
be strengthened (less need for the 10X factor) or weakened (suggesting possible
need for the 10X factor).
Second, the Panel reminded OPP that NOAELs are not zero risk points; they
are points at which there is greater than a 5% chance that the control and
associated exposure group arise from the same distribution, and it is expected that
there is still some risk at the NOAEL. The Panel felt that it is important to take this
18
-------
issue into account when evaluating the need for the 10X safety factor. An example
provided by the Panel: a NOAEL for which the possible risk (e.g., upper 95% limit)
is 30% of the animals affected should have a very different bearing on the use of
10X safety factor than a NOAEL for which the possible risk is 1 %. Therefore,
failure to consider this issue in the evaluation could lead to substantial risks at
doses considered safe, an anticonservative risk assessment and the failure to
adequately protect the public when actions are based on such an assessment.
Third, the Panel commented on the dose-response nature of the data and the
relative potency of response. They indicated that there is some confusion as to
what kinds of data support the use of the 10X safety factor. For example, studies
providing a strong dose-response relationship (increasing severity with increasing
dose) creates greater concern for removal of the 10X safety factor. Yet, these
studies generally provide the strongest information for clear identification of a low-
risk exposure level and decrease the uncertainty in this estimate. Where dose-
response data are inconsistent, only available for insensitive endpoints, or from
studies of low statistical power, uncertainty is large and there is the possibility of
unacceptable residual risk remaining after the application of the standard factors.
Thus, the application of inconclusive dose-response information to 10X safety
factor decisions is unclear.
Response. EPA agrees that the shape of the dose response curve, the
characterization of the NOAEL, the dose-response nature of the data, and the
relative potency of response are pertinent factors in the determination of the
FQPA Safety Factor. These issues are relevant to the discussion on the weight-
of-evidence approach for making judgments about the degree of concern for
potential pre- and postnatal toxicity in humans (as summarized in the revised
guidance document). The discussion on the dose response nature of
experimental animal data has been expanded to address the Panel's concerns.
The key lines of evidence or criteria considered in determining the degree of
concern for potential pre- and postnatal toxicity are outlined and situations that
could raise or lower a concern for the young are discussed.
Comment 9. The Panel stated that the analysis of the information used to support
the addition of tests to the core list of studies for Tier 1 was incomplete, and went
on to say that a careful analysis focusing not on NOAELs but on correlations of
response patterns and magnitudes using more appropriate statistical tools would
provide a clearer interpretation and provide greater scientific support for any
eventual policy choice. Failure to do this analysis could leave serious gaps in the
database which could lead to improper application of the 10X factor. In addition,
because a NOAEL must be one of the administered doses, it is not clear that
evaluations of whether certain studies lead to lower NOAELs can be properly
19
-------
interpreted as providing more sensitive study endpoints. A more appropriate
analysis would use a standardized measure of risk, such as the ED05 or ED10 and
the bounds on this estimate.
Response. It is assumed that this comment is referring to the information
derived from the A Retrospective Analysis of Twelve Developmental
Neurotoxicity Studies Submitted to the USEPA Office of Prevention,
Pesticides, and Toxic Substances (OPPTS) (USEPA 1998c) and the role that
the findings from this analysis had in the conclusion to consider redefinition of
the core toxicity database. OPP recognizes the limitations in using study
NOAELs to analyze the sensitivity of the endpoints; this issue was raised in the
paper and during the HED presentation to the SAP in December 1998.
However, the findings from this analysis alone did not form the basis of the
proposal to move the neurotoxicity studies from the second to the first Tier of
testing. (A more detailed discussion of this topic is included in the response to
comments on Question 6 below.) Nevertheless, OPP agrees that the Panel's
comments on appropriate data analysis have merit and will be considered when
conducting future evaluation of more recently submitted DNT study data.
Given the scope of the evidence which OPP intends to consider, are there any other
types of information that OPP should consider in making its FQPA Safety Factor
determinations?
Comment 1. One commentor (776) raised a concern that studies in juvenile
animals do not represent the action of pesticides on early-developing systems,
such as neuronal migration and differentiation in the pre- and perinatal period of
development.
Response. This comment may reflect a misunderstanding on why studies in
adult animals are included in the core data set, as well as more traditional
developmental studies. In the OPP guidance document, the childhood exposure
period is defined as extending from preconception through adolescence (up to
18 to 21 years of age). Adult studies provide relevant information for this broad
childhood age range, when organs may be fully formed, yet functionally
immature. They also provide information on potential target organs that could
then be further evaluated in the reproduction or other developmental studies
(e.g., developmental immunotoxicity, developmental carcinogenesis, or
endocrine toxicity studies). EPA may also compare the results from studies with
adult animals to the effects observed in studies involving exposures to animals
earlier in the developmental process. However, the adult studies were never
intended to be used to substitute for studies with prenatal and/or early postnatal
20
-------
exposures. OPP believes that a database that includes quality prenatal
developmental, multigeneration reproductive, and adult neurotoxicity studies are
important for an evaluation of potential risks to children. Additionally, the
potential for effects on the developing nervous system should also be
considered. Thus, OPP plans to publish proposed revisions to its pesticide
data requirements regulation, 40 CFR 158, and expects to ask for comment on
a requirement for developmental neurotoxicity testing, which utilizes information
about each chemical and its toxicity to develop a rational, science-based
approach to the study design and testing strategy.
Comment 2. One commentor (773) recommended that, in the determination of the
need for the FQPA Safety Factor, OPP "should be aware of the large body of data
which led the 1993 NAS report to conclude that the existing 10X intraspecies
uncertainty factor is generally adequate to protect infants and children."
Response. The1993 NAS study on Pesticides in the Diets of Infants and
Children concluded that "quantitative differences in toxicity between children
and adults are usually less than a factor of approximately 10-fold" (p. 3) and yet
went on to recommend that an additional 10-fold uncertainty factor be applied to
the risk assessment if animal studies demonstrated evidence of developmental
toxicity, including postnatally (p. 9) (NRC 1993). Regardless of the content and
apparent inconsistencies in the NAS report, OPP is staying abreast of ongoing
research on age-related differences in response to toxicants. Pertinent data
that were available at the time that the draft policy document was written are
summarized in the 10X Task Force Toxicology Working Group report (US EPA
1999a). Since that time, a number of additional studies have been conducted,
almost exclusively with pharmaceuticals. Unfortunately, the studies conducted to
date do not fully resolve this issue, but do generally support the conclusion that
any quantitative differences that may exist between the young and healthy adults
are usually within an order of magnitude.
21
-------
FIFRA Scientific Advisory Panel Comments
Comment 3. The SAP noted that there are likely to be other types of data that will
have to be considered that may improve the process within the next decade.
Distinctions between particular types of toxicity data may become blurred.
Eventually, due to the increased use of molecular biology in Tier 1 screening tests,
such tools are likely to better predict toxicological responses in the future.
Response. The Agency acknowledges the promise of molecular based
approaches (e.g., genomics, proteomics, bioinformatics, animal transgenic
models) and the expanding role of mechanistic information in improving the
ability to project human health risks for environmental agents. There is
emphasis in the Agency's risk assessment approaches for such mechanistic
information (e.g., the EPA's draft revisions for Guidelines for Carcinogen Risk
Assessment (USEPA 1999f); and OPP's Guidance for Identifying Pesticide
Chemicals and Other Substances that Have a Common Mechanism of
Toxicity (USEPA 1999h)). Although there is increasing effort to develop the use
of gene profiles to predict adverse responses, the SAP rightly acknowledges
that this technology is at an early stage of development for use in standard
toxicology testing. The National Institute of Environmental Health Sciences has
established a National Center for Toxicogenomics to help advance this
technology. OPP is committed to keep abreast of new technologies and will
follow the progress of molecular approaches and will incorporate them into our
testing and risk assessment process as appropriate.
Comment 4. The SAP indicated that OPP should routinely obtain more (specific)
information on a given toxicity finding by triggering other more definitive studies.
Response. OPP agrees that, in some circumstances, requiring more definitive
studies on various toxicity findings may aid the regulatory process. Such
triggered studies should include scientifically-sound methods and should be
designed such that they could potentially provide key information relevant to the
risk assessment of the pesticide under review. Historically, OPP has required
the submission of such information on a case-by-case basis.
22
-------
ISSUE 3. Overall Conservatism of OPP Risk Assessment
Methodology
Do you agree with the view that the models and assumptions used by OPP in the risk
assessment process, together with reliable data available on specific pesticides and
other reliable, empirical data, typically do not understate risk? If not, under what
circumstances do you believe OPP's current approaches to assessing risks from
aggregate exposure to a single pesticide produce risk assessments that understate
the risks to infants and children?
Comment 1. Citing a number of specific deficiencies, three commentors (761,
771,776 and L21) argued that, overall, OPP's risk assessment methods were
generally inadequate. The commentors argued that: (1) the 1993 National
Academy of Sciences study, Pesticides in the Diets of Infants and Children, found
the Agency's historical practices deficient; (2) the public literature contains many
examples of biological variability that exceed the size of the uncertainty factors
historically used by EPA (10X interspecies and 10X intraspecies uncertainty
factors); (3) OPP lacks data on the magnitude of critical exposures, such as
inhalation of airborne indoor use pesticides and skin absorption of skin- applied
insect repellents; (4) OPP ignores information showing that children consume
significantly more drinking water for their size than adults; and (5) OPP does not
consider the potential for pesticides to bioconcentrate in milk. In light of these
flaws, the commentors argued that OPP should retain the FQPA Safety Factor in all
circumstances; one of these commentors further asserted that when the FQPA 1OX
Safety Factor is retained, the Agency's risk assessments are adequately health
protective for all identifiable populations.
The Scientific Advisory Panel noted there may be data on pharmaceuticals
indicating variability in children, and between children and adults, that would aid in
determining whether or not the uncertainty factors adequately account for variability
in the general population.
Response. The 1993 NAS Study criticized a number of aspects of the risk
assessment methods that OPP was using in the late 1980's, and suggested
improvements. OPP has extensively revised and strengthened its scientific
approaches to assessing pesticide risks, and thus, considers the conclusions of
the NAS no longer to be completely applicable. For example, OPP now
routinely evaluates the distribution of exposure to pesticide residues in food
through the use of probabilistic models that take advantage of updated food
consumption data and a vastly expanded pesticide residue database. OPP has
also developed new approaches to estimating the levels of pesticide in surface
and ground water that may be used for drinking water and incorporates any age-
23
-------
related differences that may exist with respect to drinking water intake. OPP
has improved its capability of estimating the exposure resulting from pesticide
use in and around residences and similar sites. OPP has also made enormous
strides in developing both its aggregate risk assessment and cumulative risk
assessment methodologies. Further, OPP is obtaining a large body of
additional data, particularly concerning residential pesticide use, that will
improve its residential risk assessments. While there is always room for future
improvement, OPP believes that its current methods are generally conservative.
OPP does not agree with the assertion that the traditional "10X intraspecies
uncertainty factor" is generally not sufficient to protect infants and children, who
may or may not be more sensitive than healthy adults to the adverse effects of a
pesticide. The available information in the scientific literature seems to support
the opposite view.
The 1993 NAS Report, for example, contains a variety of conclusions
concerning whether the 10X intraspecies uncertainty factor generally accounts
for the observed variability in sensitivity of children to the toxic effects of a
pesticide. In the chapter discussing perinatal and pediatric toxicity, the NAS
committee concluded:
Differences in toxicity between young, and mature animals may be in
either direction but are generally modest. The younger animal may be
more sensitive or may be less sensitive than the older animal to
comparable levels of exposure of toxic agents. The direction of these
differences appears to be compound specific as well as age specific
because toxicity may not vary linearly with age. In those instances where
such measures as LD50s are significantly different, the differences are
usually less than one order of magnitude and often substantially less
(pp. 105-106).
However, the NAS also stated in its discussion of uncertainty factors:
The other 10-fold intraspeaes uncertainty factor is meant to cover
variations within human populations, including genetic predisposition,
poor nutrition, disease status and age (Babich and Davis 1981). A
factor of 10 for intraspecies variation in susceptibility may be sufficient
for any one element of interpersonal difference but may not be sufficient
for multiple elements (p. 326, emphasis in the original).
24
-------
On the whole, OPP interprets these and other similar statements in the NAS
Report as meaning that for most chemicals, the very large majority of people,
including children, respond similarly, and that the traditional 10-fold intraspecies
uncertainty factor is sufficient to cover the variability in the human population. At
the same time, there are chemicals for which some humans may display a
greater range of variability and sometimes that variability appears age-related,
with children exhibiting a greater degree of sensitivity than adults.
The report of the Toxicology Working Group of the 10X Task Force supports
a similar conclusion. That report included an extensive discussion of the
literature on the general adequacy of the 10X intraspecies uncertainty factor.
The available data from the literature on variability includes data primarily on
pharmaceuticals. Various authors evaluated the intraspeaes uncertainty factor
using data from animal or human studies, as summarized by Dourson et al.
(1996). The report noted that most assessments looking at a large number of
chemical compounds had concluded that the traditional 10X factor appeared to
"cover" 90% or more of observed human variability.
The report of the Toxicology Working Group, however, also pointed out that
these assessments generally lacked information on the variation in sensitivity of
children, compared to adults. The limited literature specifically comparing the
sensitivity of children and adults seemed less conclusive, but suggested that
variability between children and adults may be somewhat greater than variability
among adults alone. For example, Renwick and Lazarus (1998) evaluated
variability in response in children and adults, and indicated that, generally,
infants and children do not represent a special subgroup from a kinetic point of
view as young children frequently eliminate drugs and other chemical agents
more readily than adults. They also showed, however, that marked differences
in kinetic parameters for some agents, for example, as much as a five-fold
reduction in clearance rates of theophylline in preterm infants versus adults,
might result in a number of children not being covered by half the 10-fold factor
that accounts for kinetic variability. Given that such differences between children
and adults (or children of various ages) may exist because of differences in
metabolic capacity or developmental stage of specific organ systems, these
authors suggested that the focus should be on delineating the differences in
sensitivity of developing organisms and on the variability in sensitivity compared
with adults, as well as differences in the sources and extent of exposure.
Renwick (1998) further evaluated the similarities and differences between
children and adults as well as between young and adult animal models and
humans, and reviewed comparative data on toxicokinetics. He suggested that
the 10-fold inter- and intraspecies factors may be sufficient if developmental
toxicity data are available. Because children tend to have a higher clearance
rate for many xenobiotics compared with adults, this may compensate to some
25
-------
extent for the potential increased sensitivity of the young during development.
The report of the Toxicology Working Group urged further research on the
general issue and cautioned strongly against making decisions to reduce the
10X intraspecies uncertainty factor unless there is a robust understanding of the
potential toxicity of the substance to the young, as well as adults. The Agency is
closely following the research being done on this topic and is providing funds for
several projects that address the issue of human variability. As additional
information becomes available on this topic, we will consider it so that OPP will
be better able to characterize the similarities and differences between children
and adults and determine the adequacy of the intraspecies uncertainty factor for
protecting children.
OPP notes that variability in a discrete biological function or process does
not necessarily or even presumptively translate into equivalent variability in
sensitivity to the toxic effects of a chemical. Because the occurrence of a toxic
effect in humans is often a complex process that involves a number of
pharmacokinetic steps as well as pharmacodynamics, human systems may
attenuate the effects of the chemical in a variety of ways, thus narrowing the
overall variability of human sensitivity. In other words, a chemical may affect
people differently or to different degrees, but scientifically, one would expect
such variation to be much smaller than the variability among the population with
regard to an isolated biological parameter, especially a parameter that may
have no biological relevance to the mechanism of chemical toxicity.
These overall conclusions clearly allow for the possibility that some
compounds may produce an effect at lower levels in children than adults, or at
lower levels in some adults than in other adults. But, EPA believes that the
examples, cited by the commentors, of interindividual variability in human
biological function greater than 10-fold are relatively rare exceptions, not the
rule. Therefore, the Agency believes that available information does not support
a categorical conclusion that the traditional 10X intraspecies uncertainty factor
is insufficiently protective.
The Agency recognizes, however, that in specific circumstances information
may indicate that humans will exhibit a greater range of sensitivity, and would
conclude, in those cases, that additional protective factors are needed in the
risk assessment. As discussed above, these decisions need to be made on a
case-by-case basis. (See also the discussion, below, of EPA's approach to
taking into account the potential impact of health conditions on sensitivity.) In
addition, the Agency agrees with and adopts the 10X Task Force's conclusions
on the need for further research, and for extreme caution in making decisions to
reduce the intraspecies uncertainty factor. Finally, the Agency believes that
26
-------
available literature underscores the need for careful consideration of the toxicity
information on each pesticide for indications that children may be more than ten
fold more sensitive than adults to its effects.
Contrary to the comments, OPP believes that it generally has a reliable
database on which to estimate exposure from use of skin-applied insect
repellents and airborne residues of indoor use pesticides. To the extent that
data are limited, OPP uses models that make conservative assumptions about
exposure. Therefore, OPP does not expect that its exposure assessments
understate exposure in this regard.
Contrary to the comment, OPP has always taken into account data on the
differences in drinking water consumption between children and adults. OPP's
ability to assess drinking water exposure will be even better now that OPP has
begun to use the very refined data in USDA's latest Continuing Study of Food
Intake by Individuals (1994-1996, 1998) on the sources and amounts of water in
the diets of people of different ages.
OPP also considers exposure to pesticides through milk. In addition to the
results of tests that specifically study the potential for bioconcentration, OPP
also has data on the presence of pesticide residues in milk and uses them,
together with data on milk consumption, to evaluate exposure. OPP notes that
data from USDA's Pesticide Data Program
(http://www.ams.usda.aov/science/pdp/index.htm) show that milk rarely, if ever,
contains measurable levels of a wide variety of pesticides.
In sum, OPP believes that it would not be scientifically credible to conclude
that retention of the FQPA Safety Factor is automatically necessary in all cases
to accommodate for age-related, intraindividual variability in sensitivity. Rather,
as discussed above in Issue 2, OPP thinks that each pesticide risk assessment
should be viewed on a case-by-case basis to determine whether the weight-of-
evidence indicates that the FQPA Safety Factor should be retained, reduced,
removed, or raised.
27
-------
Comment 2. Several commentors (773, 775 and L005) indicated that OPP's
current approach to risk assessment is inherently conservative and does not
understate potential risk. They noted that many, if not most, of the steps in the risk
estimation process appropriately are conservative and the overall result usually is
risk assessments that adequately protect public health.
Response. OPP agrees that its approach to risk assessment is generally
conservative, but in individual cases, OPP may not have sufficient information to
insure that its exposure assessment is conservative. While OPP believes such
circumstances are rare, each pesticide risk assessment should be evaluated
carefully to determine whether the risk assessment is adequately protective.
Therefore, OPP continues to believe that decisions about whether to retain the
FQPA Safety Factor should be made on a case-by-case basis, taking into
account the specific information available about the potential risks of the
pesticide. See responses in Issue 2, above.
FIFRA Scientific Advisory Panel Comments
Comment 3. Several members of the Panel expressed concern that it is difficult to
make the judgment, from the existing information as presented, that OPP's
proposed procedures for FQPA risk assessments are "sufficiently conservative and
do not understate the risks to infants and children." Such a judgment requires a
quantitative analysis of the likely residual risks that could remain after application of
OPP's procedures to chemicals that prove "positive" for developmental effects and
to chemicals for which the existing testing procedures fail to detect effects. For
positive chemicals, it was emphasized that the animal/human "uncertainty factor"
was, for the most part, a dosimetric adjustment factor that compensates for the fact
that humans tend to eliminate toxicants at a slower rate than experimental animals
(with middle values tending to be approximated by the ratio of human to animal
body weights to the 114 power-about four-fold in the case of rats and seven-fold in
the case of mice.) The generic average human/sensitive human factor of 10 fold
would need to encompass somewhat more than three standard deviations in a
possible lognormal distribution of human sensitivities in order to go from a 5% risk
level consistent with observations of a NOAEL and a one in one hundred thousand
or one in one million incidence of harm. Recent information on the spread of human
interindividual variability for mild effects in adults gives some grounds for
skepticism that a 10-fold factor will routinely encompass three standard deviations
of a human population distribution of thresholds.
28
-------
The proposal does not identify individuals that are inherently sensitive. For
example, there are a variety of multifactorial diseases for which certain chemical
agents could contribute to such conditions as Parkinson's disease, essential
hypertension, or noninsulin-dependent diabetes mellitus. Such individuals could
well be more sensitive to pesticide agents that lead to a number of secondary
disorders as apparently different as neurotoxicity (e.g., diabetic neuropathy) and
cancer (e.g., liver cancer). Therefore, there are conditions in which the current
process may not be sufficiently conservative because these sensitivities are not
likely to be tested for in the near future with new or established chemicals.
There is additional reason for concern for populations of children and
developing fetuses. In general, OPP's current approaches could fail to yield risk
assessments that are sufficiently conservative if one or more of the following
circumstances applies:
the battery of tests in rodents and other animals used does not effectively
measure a wide enough array of higher-level neurodevelopmental or other
developmental functions to detect important modes of action in people.
there is an insufficient allowance for human interindividual variability to cover
the diversity of human sensitivities, which in some cases may be
considerably broader than the diversity of sensitivities in experimental
animal populations.
there are deficiencies in estimating high end exposures for infants and
children.
the single-chemical risk assessment techniques fail to capture the
cumulative risks from chemicals with related or possibly interacting
mechanisms of toxicity.
The Panel suggested that it is important to test the degree of protection likely to
be afforded by OPP's risk assessment procedures by applying them on a
hypothetical basis to the observations that would be routinely produced by the
required pesticide testing protocols for an array of known "positive" developmental
toxicants. Such materials would include methyl mercury, lead, some specific
neuroactive noncoplanar PCB congeners, and an anticonvulsive agent with known
human developmental toxicity. After application of OPP's procedures for
determining reference doses to the test chemicals, quantitative risks could be
estimated at the reference dose (and possibly below) and the judgments could be
made of the advisability of retaining the FQPA uncertainty factor for such "positive"
compounds.
29
-------
Some Panel members expressed particular concern that pesticides that are
used in homes, daycare centers and schools, food production, and pesticides
contaminating water would be likely to lead to the greatest risk in underestimating
exposure from all sources and routes, as well as drive the risk relative to multiple
pesticides with similar modes of action. The limited exposure assessments are
well outlined in the Agency's background document. Less well acknowledged are
issues of short-term exposures at critical periods of development, including those
inside the uterus, as they relate to endocrine disrupting chemicals and
neurotoxicants. In the absence of improved knowledge about these exposures,
there should be a very conservative approach to the protection of the fetus and
child.
Models and assumptions employed require validation against empirical data
when such data exist and prospectively (with the planning of new studies) when they
do not. Considerable uncertainties surround exposure data, particularly in infants
and children, and suggest proactive and expanded acquisition of data for
validation.
Scenarios can likely be developed that would involve exposures to pesticides
that will predict risks greater than risks predicted by the current approaches. Panel
members differed whether such scenarios are considered. OPP must specify
some target percentiles of the expected population distribution of exposure for
routinely evaluating whether its standard procedures provide adequate protection
for relatively highly exposed people with an adequate degree of confidence.
On the other hand, some Panel members thought that the current approaches
are adequately conservative and, if properly applied, should be protective of infants
and children. It is the Panel's understanding that OPP will be taking into
consideration potential exposures from all sources and, specifically, exposures
during the entire span of human development. Further, according to OPP's
interpretation of the FQPA, the default FQPA 10X safety factor must be used in the
absence of reliable evidence justifying use of a different value. Use of the "risk cup"
approach, which takes into account the concept of cumulative risk, i.e., the potential
presence of residues of other pesticides with like mechanisms of action, adds to
the conservatism of OPP's approach.
30
-------
Response
Overall Response. The SAP's comments reflect different views on the question
of whether OPP's risk assessments "typically do not underestimate risk." Not
only does the range of the SAP's comments reflect the diversity of views
expressed by the public, but the Panel's specific comments contain two themes
that suggest why the members (and the public) may have reached different
conclusions on whether OPP risk assessments do not understate risk. First,
there appears to be an issue with regard to who should be considered in the
assessment. Should EPA endeavor to protect the most sensitive individual or
most highly exposed individual, or should the Agency focus on larger subgroups
of the population? Second, there appears to be an issue about how confident
OPP should be that the risks have been adequately assessed. Given the
absence of data to characterize completely the degree of conservatism in
OPP's risk assessments, it appears that members may have made different
assumptions about the conservatism of the general assessment methodology.
Both of these themes come together in the Panel's comments indicating various
circumstances in which OPP's risk assessment methodology might
underestimate risk for some people and pesticides.
Overall, OPP believes that the FQPA accords the Agency discretion in how
it interprets the statutory provisions to ensure that pesticides in food are "safe."
Therefore, OPP has considerable latitude in designing its risk assessment
methodology to determine whether a specific pesticide's exposure meets the
statutory safety standard. OPP has developed its risk assessment
methodology, consistent with EPA policies and practices, to generate risk
estimates which it believes are generally conservative, i.e., most, but not
necessarily all assessments will tend to describe the upper bound or high end of
the actual risk experienced by the exposed population, but the methodology may
understate potential risk for a very small percentage of chemicals or portion of
the exposed population.
Consequently, recognizing that there are many elements in a risk
assessment, and that these elements differ from assessment to assessment,
OPP believes that it is appropriate to review each assessment individually to
see how all of the elements-hazard identification, dose response evaluation,
and exposure assessment-tend to express potential risk. In the course of that
case-by-case review, OPP can determine whether it is necessary to retain an
FQPA Safety Factor or to make other adjustments to have reasonable
confidence that risks are not underestimated, especially for infants and children.
Statutory Framework for Risk Assessment. In designing its risk assessment
methodologies and making judgments about safety, OPP must follow the
31
-------
provisions of FQPA. While the FQPA establishes a stringent, protective
standard, that focuses particularly on children's health, OPP believes that the
Agency has considerable discretion under that standard in how to interpret the
statutory mandate. Specifically, FQPA requires OPP to determine whether
exposure to a pesticide in food, water and through other, nonoccupational
pathways is "safe," i.e., whether OPP can conclude that there is a "reasonable
certainty of no harm" from such exposure. In addition, the statute directs EPA, in
making the safety determination, to consider exposure to "major identifiable
subgroups."
Collectively, these statutory provisions allow OPP to exercise judgment
about how to implement the safety standard. Notably, FQPA does not require
OPP to look at the most sensitive individual or the most highly exposed
individual. Rather, the Agency may use its judgment to conduct the risk
assessment in a manner that looks broadly at the whole population, giving
particular consideration to the exposure of major identifiable subgroups.
Similarly, the statute does not address in detail how OPP should assess
exposure for the population in question. Finally, the statute does not specify
complete certitude, but only a reasonable certainty of no harm. In sum, EPA
interprets these provisions as allowing considerable discretion in designing its
risk assessments and judging whether a particular exposure meets the statutory
standard.
Specific Responses. The paragraphs below summarize and respond to the
specific Panel comments.
Risks to individuals who, statistics predict, might be extremely sensitive. Some
members of the SAP indicated that in a large population there is considerable
interindividual variability in sensitivity, and that a dose which would affect one out
of a million exposed individuals (or one in a hundred thousand) may be more
than 10 times lower than the dose which affects 5% of the population. (The SAP
comment appears to assume that a NOAEL established in animal studies
generally corresponds to the dose which affects no more than 5% of the test
group.) In a separate comment, the SAP encouraged OPP to establish "some
target percentiles of the exposed population distribution" for evaluating risk.
32
-------
OPP agrees with the Panel that, for some types of effects, there may be
more than an order of magnitude difference between the level of a chemical
affecting 5% and 0.0001 % of exposed individuals. Such variability does not by
itself, however, mean that the complete risk assessment for any particular
pesticide understates its risk. It is equally important, for example, to consider
the components of the exposure assessment and the impact of applying other
uncertainty factors in judging the overall conservatism of the risk assessment.
As noted above, these judgments need to be made on a case-by-case basis.
Further, OPP agrees with the Panel that the judgment of how conservative a
risk assessment needs to be is, to some degree, a risk management
judgment-how confident is the decision-maker that the identified hazards will
not occur. While OPP has generally not established target percentiles for most
of the risk assessments it performs, it issued a science policy paper
recommending use of the 99.9th percentile of estimated acute exposure to
pesticides in food as the starting point in estimating the threshold of regulatory
concern (see Choosing a Percentile of Acute Dietary Exposure as a
Threshold of Regulatory Concern; USE PA 2000c).
Sensitivity of individuals with special health conditions. The SAP identified a
variety of "multifactorial" health conditions-Parkinson's disease, essential
hypertension, or noninsulin-dependent diabetes-that could increase an
individual's sensitivity to the toxic effects of a pesticide. Calling these people
"inherently sensitive," some of the Panel members indicated that such people
"could well be more sensitive" to the toxic effects of a pesticide.
Ordinarily, OPP does not take into account if and how a particular health
condition may contribute to the impact of pesticide exposure. OPP generally
assumes that the traditional intraspecies uncertainty factor (10X) is sufficient to
cover such possible variation in sensitivity. OPP also generally requires
information that would raise a reasonable concern about the possible link
between a specific health condition and increased sensitivity to a particular
pesticide. In cases where OPP has data to show that people with a particular
health condition are more susceptible to the toxic effects of a pesticide, OPP
will consider the information in its risk assessments and risk management
decisions. For example, OPP received information establishing that many
people were experiencing significant allergic reactions after having used
Allercare, a product marketed to control dust mites. OPP directed the
registrant of Allercare to recall its products, and to reformulate the pesticide to
remove the ingredient that caused the allergic response in users. As another
example, OPP received information raising concerns about the potential for a
biological pesticide, sp. Burkholder cepacia, to aggravate the health condition
of individuals with cystic fibrosis. OPP conducted a thorough risk assessment
33
-------
of this product, including a consultation with the SAP on this risk. In light of the
assessment, the company withdrew its application for registration.
Unevaluated toxic effects. Some Panel members indicated that the currently
required battery of animal toxicity studies does not effectively measure "a wide
enough array of higher-level neurodevelopmental or other developmental
functions to detect important modes of action in people." In connection with a
similar comment, some of the SAP members noted that there was a way to
evaluate more quantitatively whether risk assessment methods underestimated
risks: after performing additional tests to evaluate these effects, OPP would
need to compare the earlier (pretesting) risk assessments with assessments
that took such studies into account. In the same vein, another SAP comment
suggested conducting such a comparison with known neurotoxicants such as
methyl mercury, lead, PCB congeners, etc.
OPP has tailored its data requirements for assessing the potential risks of
particular pesticides to the characteristics and use patterns of the individual
pesticide. Additionally, OPP may impose additional data requirements on a
case-by-case basis through Data Call-in. As additional data are available, OPP
will consider the findings and include them in the risk assessment as
appropriate. Responses to Issues 6-10 describe the approach OPP uses to
call in neurotoxicity data and the policies for using Traditional uncertainty factors
and/or special FQPA Safety Factors in the absence of these data.
As to the suggestion by Panel members for conducting a series of
comparisons for risk assessments performed before and after the inclusion of
additional data or to a set of known neurotoxicants, this may be done in
conjunction with the ongoing evaluations of the developmental neurotoxicity
studies which have been submitted in response to the September 1999
chemical class DCI for all the cholinesterase-inhibiting organophosphorus
pesticides with established tolerances.
34
-------
Inadequately assessed hiah-end exposures for infants and children. The Panel
suggested that EPA should collect empirical data to validate the models on
which it bases exposure estimates.
OPP agrees that collection of empirical data to validate its exposure models
could reduce uncertainty and strengthen its risk assessments. The Agency has
made considerable progress in collecting data to validate and improve many of
its models. These efforts are described in more detail in the responses to
comments in Issues 13-16.
Unevaluated Exposure Scenarios. Some members of the Panel commented
that because it is possible to identify exposure scenarios that OPP has not
evaluated, it is possible to argue that OPP's risk assessment methodology may
underestimate risk for people who might be exposed in such scenarios.
OPP agrees that there are certain hypothetical exposure scenarios that are
not evaluated by current models. In Issue 15, OPP responds to public
comments identifying a number of potential exposure scenarios allegedly not
covered by current models, and discusses why OPP believes that these
exposures are not likely to be significant or that the current models are adequate
to estimate these exposures. The SAP did not identify any specific scenarios,
and therefore OPP cannot comment on whether there are additional potential
exposure scenarios not raised in the public comments. In addition, OPP has
added to, or modified, the Residential SOPs (US EPA 1999g) since the SAP
reviewed them in late 1998.
Critical exposure periods. Some members of the Panel expressed concern that
OPP needed to adopt a "very conservative approach" to evaluating pesticides if
exposure to the substance during a critical period of fetal development causes
developmental toxicity.
OPP believes that it has a conservative approach to evaluating pesticides
which produce toxic effects only if there is exposure during a critical period.
Basically, OPP estimates acute exposure to women of childbearing age and
assumes that a pregnant woman could receive an exposure at the high end of
the range during a critical phase of the development of her fetus. Even though
this approach most likely overstates the frequency of such exposures, OPP
believes that it is an appropriately conservative approach.
35
-------
Cumulative Risks Associated with Chemicals That Have a Common
Mechanism of Toxicity. The Panel commented that EPA's risk assessments of
individual pesticides may be insufficiently conservative if the pesticides share a
common mechanism of toxicity with other chemicals. OPP agrees that FQPA
requires the Agency to consider the cumulative risk of exposure to multiple
chemicals that have a common mechanism of toxicity, and that failure to
consider such potential risks could understate the potential risk. OPP, however,
does consider whether pesticides have a common mechanism of toxicity and, if
so, OPP evaluates the potential risks associated with exposure to them (see
Guidance for Identifying Pesticide Chemicals and Other Substances that
Have a Common Mechanism of Toxicity (US EPA 1999h),Guidance on
Cumulative Risk Assessment of Pesticide Chemicals That Have a Common
Mechanism of Toxicity (US EPA 2002).
In sum, OPP agrees with the SAP that, in certain circumstances, its general
approach to risk assessment may not be adequately conservative. Therefore,
for each pesticide risk assessment, OPP attempts to identify any special
circumstances that would necessitate a different approach. OPP risk
assessments document specific steps that OPP has taken to address unique
conditions involving the toxicity or exposure profile of a pesticide. In addition, at
the conclusion of the risk assessment for a food use pesticide, OPP performs a
weight-of-evidence assessment regarding whether to retain or modify the FQPA
Safety Factor. OPP believes these kinds of steps are appropriate to take into
account the possibility that the general approach to risk assessment should be
modified.
36
-------
ISSUE 4. Relationship of FQPA Safety Factor to Traditional
Uncertainty/Safety Factors
Do you agree with OPP's view that the FQPA Safety Factor should be applied in
addition to the interspecies and intraspecies uncertainty factors, but that the FQPA
Safety Factor should not be applied in a manner that results in "double-counting" of
uncertainties that are otherwise addressed in the toxicity and exposure assessments
through, for example, the database uncertainty factor or conservative exposure
models? If you disagree, why?
Comment 1. Commentor 778 (NRDC) asserted that Congress meant for OPP to
do more than continue its traditional use of uncertainty factors.
Response. OPP is doing more than continuing its traditional use of uncertainty
factors. As the Revised Guidance Document makes clear, the additional safety
factor provision both codifies and expands traditional use of safety factors.
NRDC does not explain how Congress' reference to the "completeness of the
toxicity database" can be read as excluding traditional use of safety/uncertainty
factors to address weaknesses in the toxicity database.
Comment 2. Commentor 778 (NRDC) pointed to the NAS 1993 "Kids' Study's"
critique of the then-current EPA data requirements, NRDC's 1998 study of such
data requirements, and EPA's proposed addition of several studies to EPA's Tier
1 toxicity testing requirements as grounds for retaining the additional 10X factor
across the board.
37
-------
Response. With respect to the 1993 NAS Report, it did include a number of
recommendations for changes in existing toxicity testing as well as suggestions
for new types of tests. During the time that this report was under development
between 1988 and 1993, the Agency developed and implemented the use of
several test guidelines for new study types (e.g., the developmental neurotoxicity
and the acute and subchronic adult neurotoxicity). Since that time, the Agency
has developed or upgraded several additional guidelines (e.g., upgrades to the
multigeneration reproductive toxicity and prenatal developmental toxicity
studies, enhanced evaluation of the immune system in the repeated dose
studies and a stand-alone, short-term immune system assay). However, even at
the time the Report was issued in 1993, the NAS did not take the view that any
perceived deficiencies in the existing database were of such consequence that
a full 10X safety factor was required across the board. For example, the NAS
wrote:
Although current uncertainty factors used to extrapolate toxicological data
to humans provided for 10-fold variations between species and within the
human population, additional protection for developmental toxicity may be
required, depending on the toxicant of interest and the amount of testing that
has been conducted, [emphasis added]
Similarly, the NAS report specified:
[T]he committee recommends that an uncertainty factor up to the 10-fold
factor traditionally used for fetal developmental toxicity should also be
considered for postnatal developmental toxicity and when data from
toxicity testing are incomplete.[emphasis added]
In both of these passages, the NAS committee clearly indicated that whether
an additional factor was needed due to inadequate testing required a case-
specific examination of the testing for the pesticide in question. Thus, even the
NAS, the organization that recommended the use of an additional 10-fold safety
factor, did not draw the conclusion advanced by NRDC.
EPA's Revised Data Requirements. NRDC argues that because EPA has
announced it plans to add additional studies to the Tier 1 toxicity testing
requirements, EPA has essentially concluded that the databases for all
pesticides are incomplete and an additional 10-fold safety factor should be
applied to all pesticides. NRDC appears essentially to be taking the position
that should EPA consider requiring additional toxicity testing for any pesticide
the statute automatically mandates application of an additional 10-fold
uncertainty factor. EPA does not believe that the statute imposes such a rigid
approach. Rather, the statute directs EPA to take into account the
38
-------
completeness of the database and to evaluate what existing, reliable data
indicate as to hazard and exposure. Moreover, as noted in the draft Policy
Paper, Congress did not address the situation of what action would be
appropriate where new toxicity testing requirements are imposed. Thus, EPA
has some flexibility with regard to situations involving new testing requirements.
The approach that OPP will take when data have been requested, but are
not yet available for assessment purposes is described in detail under Issue 7.
Comment 3. Commentors 369 (PMRA) and L015 (CDPR) agreed with OPP that it
should take steps to avoid the double-counting of factors.
Response. As stated in the revised guidance document and in this Response
to Comments document, Congress, in enacting the additional safety factor
provision, was not creating an entirely new component of pesticide risk
assessment. Rather, Congress was incorporating into the statute the existing
risk assessment process that already involved use of "uncertainty" or "safety"
factors to compensate for lacking toxicity data, while building upon that existing
process. Frequently, the additional safety factor has been referred to as if it
were a wholly new creation that operated independently from other traditional
uncertainty/safety factors. This shorthand description of the additional safety
factor increases the possibility that the additional factor could be applied in a
way that overlapped with traditional safety factors and essentially resulted in
double counting of factors designed to address database uncertainty and other
components of data integrity. Thus, it is incumbent upon OPP to clearly identify
why safety/uncertainty factors are needed so that OPP neither applies two
safety factors to address one concern nor fails to apply any safety factor to
address database uncertainty due to confusion as to what stage of the analysis
such factor should be applied. The revised guidance document includes the
principles and process of uncertainty/safety factor selection and application and
these are also discussed in greater detail in this document under Issues 7, 11,
and 12.
39
-------
Comment 4. Commentor 372 (AWWA) noted that the positions set forth by OPP
seem to be in line with their interests. It is important to note that the proposal for an
additional safety factor to account for the possibility that infants and young children
may in some instances not be protected with "reasonable certainty" by previous
uncertainty factors has some basis in recent scientific research. These previous
uncertainty factors accounted for known variations in human susceptibility and
sensitivity, but this variation was estimated primarily from data in adults where a
reliable database could be obtained. At the least, existing scientific evidence has
not been analyzed systematically to demonstrate that the existing uncertainty factors
are adequate in all cases. Until such an analysis is performed, it is reasonable for
the Agency to apply an additional safety factor in cases where the application of
traditional uncertainty factors cannot be shown with reasonable confidence to
include variability across the entire human population.
The OPP needs to develop and clearly state a procedure of scientific review
that will ensure the FQPA Safety Factor applied does not result in double counting
of sources of uncertainty. Still missing from the discussion, however, is an
explanation of how OPP will review the scientific merits of any proposed FQPA
Safety Factor applied in a specific case. The presumption would be a SAP review,
but this is not stated. Further, AWWA strongly encourages OPP to develop a
formal process for making uncertainty decisions based on a hierarchy of questions
that will give the regulatory manager the information needed to make decisions
based on sound science. The structure of the questions should be supported by
clearly defined Data Quality Objectives with data quality taking precedence over
arbitrary scheduling deadlines.
The commentor stated further that, in its opinion, OPP fails to make clear
several operational issues that could still result in double counting of uncertainty.
OPP needs to clarify how an additional database uncertainty factor and an FQPA
Safety Factor are related. Is the latter part of the former, or entirely separate? The
document is not clear on this point. It also suggests that the FQPA Safety Factor
might reflect more than an additional database uncertainty factor. This also leaves
the reader unclear whether additional considerations will be reflected in risk
characterization, whether some may be reflected in hazard characterization, and
how the two points of characterization are related.
40
-------
Response. OPP agrees that available data suggest that adults and children
may display a range of sensitivities to different chemicals, and that for some
chemicals, children may be more sensitive to the toxic effects of a chemical than
adults. Therefore OPP agrees that application of the historical uncertainty
factors may not always be sufficient to protect infants and children. As
discussed above, however, the scientific data on variability in sensitivity of
humans does not indicate the need for an additional uncertainty factor in most
cases. See Issue 3, Comment 1.
OPP agrees that it needs a systematic approach to making decisions about
the application of the FQPA Safety Factor. OPP accordingly is revising its
Standard Operating Procedure to guide decision-making on individual
chemicals. The SOP describes a very detailed process of collecting,
organizing, and analyzing information that assures that all information is
weighed appropriately. Although OPP does not think review of all its chemical-
specific decisions by the SAP would be necessary, OPP has sought SAP
advice on its guidance document and SOP, and would not hesitate to take
specific FQPA Safety Factor decisions to the SAP if OPP believed the SAP's
advice would be helpful.
OPP agrees that its approach to applying the FQPA Safety Factor should
avoid "double-counting" and that it should clearly distinguish between
considerations that are addressed by traditional uncertainty factors and those
that arise as a result of concerns unique to the FQPA. OPP has revised its
guidance document to make the differences clearer. The differences are also
discussed in the material presented below in Issues 5 to 8.
41
-------
Comment 5. Commentor 773 argued that OPP should not use the FQPA Safety
Factor to cover any or all uncertainty factors other than the interspecies and
intraspecies uncertainty factors (UFs). Rather, OPP should continue its prior
practice with respect to applying database uncertainty factors. OPP should use its
traditional approach to UFs for assessing risks to adults. But, because FQPA
contains the Safety Factor provision, OPP must analyze whether there is a basis for
reducing or removing the additional Safety Factor for children. However, this
analysis should focus on the "important differences" in risks to children and adults.
Specifically, the FQPA Safety Factor should be reserved for situations where: (1)
pesticides pose particular safety concerns for children because of the nature and
severity of the effects, the likelihood that children may be much more sensitive to the
adverse effects than are adults, such that there is a reasonable apprehension that
the traditional safety factors are insufficient to protect them; or (2) there is
substantial uncertainty about the aggregate exposure, properly defined (i.e.,
exposure from food, and where reliable data exist, other sources).
Response. OPP agrees that its decisions to reduce or remove the FQPA
Safety Factor should focus on whether there are important differences in risks to
children and adults. The Agency, however, understands the comment to
advocate a substantial narrowing of the scope of the FQPA Safety Factor that
excludes toxicity database uncertainty concerns from the FQPA Safety Factor.
Whether or not EPA agreed with such an approach as a policy matter, EPA
does not see how such an approach can be squared with the language of the
statute. The statute specifies that the FQPA Safety Factor is to "take into
account...the completeness of the data with respect to...toxicity to infants and
children." 21 U.S.C. §346a(b)(2)(C). Moreover, the NAS Report heavily
emphasized that one of the reasons for an additional safety factor was toxicity
database deficiencies.
Comment 6. Commentor 775 asserted that Congress intended that the FQPA
Factor be a risk management factor for safety, not a risk assessment uncertainty
factor. The policy document discusses at the length the Agency's belief that
Congress intended that the FQPA 1OX safety factor to function as an "uncertainty
factor" and that the factor would serve to codify and extend EPA's practice of
employing uncertainty factors in risk assessments. CMA believes OPP errs in its
interpretation of Congress' intent. The language of FQPA makes clear that the 10-
fold "margin of safety" is just that-a safety factor. OPP's discussion on the lack of
difference between an "uncertainty factor" and a "safety factor" blurs the distinction
between the purpose of and scientific rationales for applying such factors. An
uncertainty factor is typically applied during the hazard and risk assessments to
account for uncertainties and is incorporated into the resulting risk estimate. On the
other hand, a safety factor is applied in setting a safe level of exposure. Congress
42
-------
did not intend to interfere with the Agency's assessment of risks to children, but
essentially directed the Agency to err on the side of safety in setting tolerances
where it did not have sufficient data or there was clear evidence of pre- and
postnatal toxicity and to take into account "completeness of the data with respect to
exposure and toxicity to infants and children." Thus, CMA believes that the purpose
of the FQPA factor is to provide an additional margin of safety to protect infants and
children where warranted, and the factor should clearly be applied after the RfD is
set, not as part of the RfD-setting process.
It is CMA's belief that application of an FQPA Safety Factor, regardless of its
magnitude, must occur only after determination of the most scientifically supported
RfD. The intent of the FQPA is that the application of the FQPA Safety Factor is a
risk management policy decision that should be made when considering the overall
risk or lack of harm that could occur following exposure to pesticides. Therefore,
the decision to apply the FQPA Safety Factor would be made by the risk manager
after reviewing the risk characterization. CMA believes that the only, and not just
the "final decision on the FQPA factor be informed by the science presented in the
risk characterization and recommendation." (64 FR 37007)
Response. The main point of the comment seems to be that the FQPA Safety
Factor should not be used in setting the RfD but rather as some type of risk
management tool for determining whether pesticides that otherwise had
exposures within their RfD were nonetheless unsafe as to infants and children.
Therefore, the comment urges OPP to apply the FQPA Safety Factor after it has
derived the RfD.
OPP agrees that the final FQPA Safety Factor decision should be made
after the RfD is derived, but it takes that position for a different reason than the
one suggested by the comment. OPP sees value in keeping the RfD process
consistent across the Agency, and thus does not intend to use those aspects of
the FQPA Safety Factor that are unique to the policy concerns in the FQPA in
establishing pesticide RfDs; rather, those types of safety factors are only used in
establishing Population Adjusted Doses (PADs).
OPP does not agree with the comment that not differentiating between
safety factors and uncertainty factors in its FQPA Safety Factor policy "serves to
embed risk management decisions in the risk assessment process,
undermines the scientific rigor of the risk assessment process, introduces
unnecessary conservatism, and diminishes transparency." OPP has gone to
great lengths to emphasize the importance of full and careful explanation of the
basis for any additional safety/uncertainty factor used in assessing the risk
posed by a pesticide. This commitment to the adequate explanation of OPP
decisions insures that the concerns regarding diminished scientific rigor and
43
-------
transparency should not arise.
Neither does OPP agree that Congress intentionally used the term "margin
of safety" because it intended that it apply the additional margin of safety for
infants and children in a fundamentally different manner than uncertainty factors
are used. The comment claims that the term "margin of safety" is an
unambiguous term that has acquired a clear technical meaning. According to
the comment, an "MOS is a ratio between a hazard value (such as an RfD or,
more correctly, a NOAEL) and an exposure(s) likely to be experienced." An
uncertainty factor, the comment asserts, is a wholly separate concept used to
derive a hazard value such as a RfD. In fact, OPP's experience in the regulation
of pesticides has been that the derivation of hazard values such as RfDs and
computation of MOSs or MOEs (Margins of Exposure) are equivalent
processes. An RfD is calculated generally by dividing the appropriate NOAEL
by various uncertainty (safety) factors. The RfD is then regarded as the "safe"
dose and exposure is compared to the RfD to ensure that the RfD is not
exceeded. An MOS (or margin of exposure, as it is more commonly referred to
now) is calculated by dividing estimated or measured exposure into the
appropriate NOAEL. The resulting MOS/MOE value is judged acceptable or not
based on comparison to various uncertainty ("safety") factors. Importantly, OPP
has always assessed a specific pesticide risk using either a RfD approach or a
MOS/MOE approach. Moreover, for a given effect of concern for a pesticide,
the uncertainty ("safety") factors used would be precisely the same whether the
RfD or MOS/MOE approach were used. The two approaches are simply
different ways of quantitatively expressing precisely equivalent results. For
example, assume a risk assessment needs to be performed for a pesticide
where an uncertainty/safety factor of 100 is appropriate, and the NOAEL is 10
and exposure is 0.05. In these circumstances, the RfD would be 0.1 (10/100),
and exposure would only be 50 percent of the RfD. Under this situation, the
MOS/MOE would be 200 (10/0.05), a value twice as high as necessary for a
safety finding. Under either approach it is clear that exposure could double
before a safety concern was raised. OPP is unaware of any instance where it
has layered an MOS/MOE approach on top of a RfD or vice versa.
44
-------
As previously noted, the overlap between the concepts of: (1) margin of
safety or safety factors; and (2) uncertainty factors, is well-reflected in the
legislative history of the FQPA. In fact, contradicting the argument of CMA,
Congress referred to the traditional inter- and intraspecies uncertainty factors as
safety factors and the new children's safety factor as an uncertainty factor (see
H. Rept. 104-669, 41, 42; 1996).
CMA's argument that the statutory structure implies a clear distinction
between uncertainty factors (used for risk assessment) and safety factors (used
in risk management) is also unavailing. CMA argues that if Congress had
intended the FQPA Safety Factor to be an uncertainty factor it would have
included the factor in subsection (b)(2)(C)(i) which contains various risk
assessment considerations and not (b)(2)(C)(ii). EPA suspects little can be
read into placing the additional children's safety factor in a separate
provision-after all, it was one of the key pieces of the legislation. More
importantly, another provision of the statute demonstrates quite plainly that
Congress did not draft the FQPA so as to maintain the sharp distinction that
uncertainty factors were to be used only in risk assessment and safety factors
were not. Subsection (b)(2)(D)-a subsection devoted entirely to enumerating
factors to be considered in risk assessment-does not use the term "uncertainty
factor" but instead mentions "safety factors." Moreover, safety factors are
mentioned in a manner that clearly implies Congress was referring to the
traditional inter- and intraspecies uncertainty factors used by EPA in
extrapolating from animal data to human risk ("safety factors which...are
generally recognized as appropriate for the use of animal experimentation
data"). Thus, neither the statutory language in the children's safety factor
provision, the structure of the statute, nor the legislative history supports CMA's
argument.
Comment 7. Commentor 778 also raised the issue of terminology when it stated
its belief that OPP incorrectly calls the statutory presumption an "FQPA Safety
Factor;" in their view, it should be called an additional "margin of safety." By
referring to the FQPA provision's requirement as an additional "safety factor," the
comment claims that EPA is incorrectly trying to treat the FQPA provision as a
variation on the traditional uncertainty factors. This, in turn, provides excuses for not
retaining the FQPA margin of safety.
Response. Unlike the industry commentors who tried to develop a distinction
between safety and uncertainty factors, and RfD derivation and margin of
safety/margin of exposure approaches, NRDC claims there is a distinction
between a margin of safety approach and a safety factor approach. According
to this commentor, OPP uses the term "safety factor" so that it can treat the
45
-------
additional margin of safety provision "as a variation on the traditional uncertainty
factors." OPP explains elsewhere why it is appropriate to treat Congress'
reference to safety factors to extend to what OPP has generally referred to in
more recent times as uncertainty factors. This commentor, however, could
rightly note that Congress used neither the term "safety factor" nor "uncertainty
factor" in subsection (b)(2)(C)-the infants' and children's provision. Rather,
Congress used the term "margin of safety." Any attempt to distinguish between
a margin of safety approach and a safety factor approach, however,
immediately flounders because the two terms are inextricably linked. As
explained above in responding to Comment 6, OPP calculates a margin of
safety or margin of exposure by dividing the appropriate NOAEL by estimated
human exposure. The size of the resulting quotient represents the margin
between human exposure and the exposure that yielded no adverse effects in
experimental studies. The acceptability (i.e., safety) of the margin is evaluated
by comparing to the safety ("uncertainty") factor deemed necessary to address
such concerns as inter- and intraspecies variability (usually represented, in a
default situation, by two 10-fold factors). Hence, when Congress dictates that
the margin of safety should be 10-fold greater, it is saying that in addition to the
traditional intra- and interspecies 10-fold safety/uncertainty factors, another 10-
fold factor should be used. Because this additional 10-fold factor is to address
safety, it seems a very small leap to label it a safety factor.
Comment 8. Several comments were received which make note of lack of clarity
as to congressional intent when it crafted the language of the children's provisions
in FQPA and/or OPP's implementation of the statute with regard to the "additional
margin of safety."
Commentor 773 asserted that Congress was clearly misinformed about prior
Agency practice regarding the extent of use of the 10X factor and mistakenly
thought that the additional 10X was widely applied. Congress' misunderstanding is
traceable to inaccurate statements in the 1993 NAS Report on Pesticides in the
Diets of Infants and Children, which stated that EPA applied an additionaHOX
"whenever toxicity studies and metabolic/disposition studies have shown fetal
developmental effects." (This commentor also alleged that EPA further confused
the issue by a letter provided to the House Commerce Committee during the
debate over FQPA.) A Congressional report quoted that passage from the 1993
NAS Report to support inclusion of the 10X Safety Factor provision in FQPA and
directed EPA to interpret the provision in furtherance of the NAS statement.
Therefore, since Congress was trying to perpetuate what they understood to be
existing practice, OPP should interpret the 10X provision to continue prior practice,
as much as possible.
46
-------
Commentor L016 stated that OPP should seek to implement the additional
safety factor provision in a way that, as much as possible, continues the pre-FQPA
approach. OPP should not presume there is a congressional mandate to add a
new 10X factor in all or most cases. Congress clearly was misinformed about its
prior practices and thought that the additional safety factor was already widely used,
which was not the case.
Response. OPP agrees that special caution should be used in interpreting the
legislative history pertaining to the 1993 NAS Report in regard to the Report's
incorrect statement that OPP (and FDA's Center for Food Safety and Applied
Nutrition, for that matter) routinely applied an additional 10-fold uncertainty factor
where prenatal toxicity was identified. In fact, OPP has not blindly retained the
additional 10-fold safety factor if pre- or postnatal toxicity is identified. Rather, it
examines whether the pre- or postnatal toxicity shows that fetuses, infants, or
children are more susceptible to the pesticide than adults, and the pre- or
postnatal toxicity is particularly severe or has other characteristics that raise
concerns regarding the adequacy of traditional safety factors . This approach is
consistent with the concern reflected both in the statute and in the NAS Report
concerning the potential heightened sensitivity of infants and children to
pesticides.
Comment 9. Commentor 406 asserted that OPP has misinterpreted both the
statutory language and Congressional intent of FQPA. Under section
408(b)(2)(A)(l) of FFDCA, OPP is to establish safe tolerances. In addition, section
408(b)(2)(A)(ii) defines the safety standard as a reasonable certainty of no harm.
The latter term means exactly the existing regulatory practices of the FDA for food
additives. While Congress added several minor modifications to the old FFDCA
standard, the intent was for EPA to stay as close to FDA's practices under
reasonable certainty of no harm standard as possible. This commentor
recommends that OPP describe to Congress that regulatory toxicology practices in
effect at FDA before FQPA protected infants and children. Instead, OPP has
strayed far from this primary objective, acting as if the minor modifications in the
FQPA were the primary intent and has set off on a tangent of creating new,
unvalidated toxicology tests to measure pediatric endpoints.
OPP should cease inventing new regulatory practices and requiring new
toxicology tests that in effect create a new standard. This thrust of OPP's policy
development contradicts both statutory language and Congressional intent. If the
standard protected children in the past, no reason exists to change the standard
now. If OPP believes that children have higher exposures than adults, OPP should
adjust tolerances to reflect the higher exposures, not the interpretation of toxicology
data.
47
-------
In the context of animal testing required before 1996, a one-hundred fold safety
factor applied to the lowest no-effect level in a standard battery of animal tests
adequately protected the developing human fetus and human neonates. No good
reason exists to deviate from this practice. This commentor disagrees with the
Agency's interpretation of both congressional language and intent about an extra
10-fold safety factor. Both language and intent about an extra 10-fold safety factor
are ambiguous and sufficiently unclear that, in the face of a clear mandate to apply
the "reasonable certainty of no harm" standard, OPP should incorporate the
standard practice as policy. The standard practice will achieve safe levels for the
human fetus and neonate.
Response. OPP would agree that the "reasonable certainty of no harm"
standard applied by FDA since 1958 informs the content of the safety standard
in the FQPA. However, Congress also imposed numerous other considerations
and requirements on OPP in applying this safety standard. OPP's
implementation of the "reasonable certainty of no harm" standard should take
into account the whole of the congressional mandate included in the FQPA.
OPP has attempted to remain faithful to the language of the children's safety
factor provision in implementing decisions regarding this policy.
OPP is unsure what this commentor is referring to regarding "new,
unvalidated toxicology tests to measure pediatric endpoints." No new testing
protocols or guidelines have been developed directly as a consequence of the
children's protection provisions of FQPA. Two study types-the multigeneration
reproduction study and the prenatal developmental toxicity study-have been
upgraded. While these two updated guidelines were not issued until after
FQPA was passed in 1996, the work to improve their design began nearly a
decade earlier, as did the work to develop the acute and subchronic
neurotoxicity studies in adult animals and the developmental neurotoxicity study.
Comment 10. Commentor 778 argued that Congress clearly and knowingly
inserted a presumptive additional margin of safety in FQPA to address the
inadequacy of existing toxicity and exposure data for children. In their view, the 10X
policy reverses the presumption in FQPA which strongly favors retention of the 10X.
OPP erects too many hurdles to the decision to retain the 10X.
Response. OPP would agree that Congress included the additional 10-fold
factor as a default position or presumption. The commentor does not cite any
authority for the proposition that Congress intended this to be a "strong"
presumption. In deciding whether to use a factor different than the additional
10X, OPP must be guided by the language of the statute. That language
provides that OPP "may use a different margin of safety for the pesticide
48
-------
chemical residue only if, on the basis of reliable data, such margin will be safe
for infants and children." The reference to "reliable data" indicates that
Congress intended that any decision on changing the factor like other decisions
under the FFDCA, should not be based on speculation. See Issue 16 below, for
a discussion of the comments concerning the meaning of "reliable data."
However, since "reliability" is a criterion with broad applicability under FFDCA
section 408, it is difficult to argue that the requirement for reliable data creates a
"strong" presumption. More important to judging the strength of the presumption
is the standard Congress prescribed for when a different factor could be
chosen. What Congress mandated was that a different safety factor was
appropriate where the different factor would be "safe for infants and children." A
standard such as "safe for infants and children" does not narrowly limit OPP's
discretion but rather invokes the broad safety standard in section 408. Given
that Congress gave OPP such a broad standard to apply in making decisions
regarding changing the factor, OPP finds little support in the statutory language
for the commentor's contention regarding a "strong" presumption.
OPP does not agree that it has reversed the statutory default position of
including an additional safety factor. For the reasons explained in the draft
policy document and in the revised guidance document, OPP believes that,
when reliable data are available, it will be fully protective of human health to
make a individualized determination of an appropriate additional factor for
pesticides rather than simply rely on a default 10-fold factor. OPP also believes
its policy on making such determinations is fully consistent with the discretion
granted to determine that a different factor is appropriate.
Comment 11. Commentor 778 stated that OPP should revise its policy to assure
that the decision about the FQPA Safety Factor is made on the basis of "no harm to
children," rather than compliance with EPA's data requirements. By including the
word "potential" in FQPA ["take into account potential pre- and postnatal toxicity..."],
Congress did not intend to restrict the application of the additional 10X to
chemicals for which greater sensitivity had already been demonstrated. The OPP
policy takes the opposite approach, requiring there be some evidence of increased
susceptibility.
Response. OPP's approach does not limit application of the additional 10-fold
factor to situations where greater sensitivity or susceptibility has already been
demonstrated. In general, there can be three possible scenarios for a given
pesticide and potential sensitivity/susceptibility to infants and children: (1) there
are sufficient data showing infants and children have increased
sensitivity/susceptibility; (2) there are sufficient data showing no increased
sensitivity/susceptibility to infants and children; and (3) data are insufficient to
49
-------
demonstrate whether there is increased sensitivity/susceptibility to infants and
children. In scenarios #1 and #3, OPP may decide to retain the default 10X
children's safety factor or to assign a different safety factor greater than 1X.
Scenario #3 particularly relates to the circumstance where there is the
"potential" for increased sensitivity/susceptibility, and thus OPP is taking this
statutory term into account. See additional discussion in the response to
comments in Issue 7 (scenario #3) and Issue 11/12 (scenario #1), below.
Comment 12. Comments were received on the discussion in the proposed 10X
Policy with regard to whether or not there were any directional constraints when
making FQPA Safety Factor decisions.
Commentor 773 argued that several parts of the legislative history of the FQPA
Safety Factor provision indicate that the additional factor cannot exceed 10. First,
the House Committee report speaks of "an uncertainty factor up to the 10-fold factor
traditionally used by EPA." Second, Dr. Goldman, EPA Assistant Administrator,
wrote the House Commerce Committee that the proposed Safety Factor provision
would allow EPA to impose "an additional margin of safety of up to ten" if risks to
children required it. Third, the House floor debate referred to the flexibility given in
the bill for EPA "to apply a safety factor of less than 10-fold" if such a factor would
be safe for children. These specific passages, together with an absence of any
language suggesting the use of a safety factor larger than 10X, indicate that
Congress did not authorize the Agency to apply a factor greater than 10. Moreover,
the draft policy provision would reverse prior Agency policy articulated shortly after
enactment of FQPA that the FQPA Safety Factor can be no greater than 10X.
This commentor went on to say that if the database uncertainty factor is
considered part of the FQPA Safety Factor, the overall factor should not exceed ten
times the database uncertainty factor.
Response. OPP disagrees with the commentor's legal interpretation. The
statute, on its face, does not limit the additional factor for the protection of
infants and children to a maximum of 10X. To the contrary, the statutory
language specifically uses a term ("different" margin of safety) that does not
directionally constrain any decision in choosing the size of a factor replacing the
default value. Moreover, as the commentor admits, the legislative history
indicates the choice of the term "different" appears to have been a considered
choice by Congress. Although earlier drafts of the bill more closely used the
language from the 1993 NAS Report language recommending an additional
safety factor of "up to 10X," the final version of the bill amended this phrasing by
dropping the "up to" language and adding specific authority to assign a
"different" factor. Moreover, it is interesting to compare the sentence in the final
50
-------
bill granting authority to vary from the additional 10X factor with what appears to
be its predecessor from the Clinton Administration proposal of 1994 (S.2084).
The critical sentence from that bill specified:
Notwithstanding this requirement for an additional factor, the
Administrator may determine that a margin of safety for a pesticide chemical
residue is considered ample for children and infants if, on the basis of
reliable data, such margin will fully protect the public health (S.2084, 103d
Cong., 2d Sess. §3; 1994; section 408(b)(2)(C) as revised).
This provision from the earlier bill clearly did contain a directional indication
with regard to variation from the additional 10-fold factor. Thus, not only does
the statutory language specifically authorize OPP to adjust the factor in either
direction, but the history of that language suggests that was a conscious choice
by Congress.
The commentor points to other legislative history to suggest Congress
intended to bar OPP from exceeding 10X. First, it argues that when Congress
dropped the "up to" language from the bill that change was "accompanied by
agreement to preserve the 10X maximum by adding report language
incorporating the NAS report's "up to 10X" explanation." It, however, cites no
authority substantiating any such agreement. Moreover, although the NAS' "up
to 10X" language does appear in the House Commerce Committee Report, it is
simply as part of a long quote from the NAS Report (see H.Rept. 104-669, 43;
1993).
Second, the commentor cites to an EPA letter read into the Congressional
Record in which EPA stated:
[The children's safety factor] provision is consistent with current Agency
risk assessment practices. We have been actively working to implement
the NAS recommendations...In so doing, EPA scientists exercise their
best judgment, based on reliable data, to determine whether studies
accurately reflect the risk to children or if an additional margin of safety of
up to ten is required (letter from US EPA Assistant Administrator Lynn
Goldman to Chairman Thomas Bliley of the House Committee on
Commerce, July 23, 1996).
This letter merely states that OPP practice of applying a factor of up to 10X
is consistent with the statutory language. That is true. The plain language of the
statute gives OPP the authority to remove the additional factor, set a lower
additional factor, retain the additional factor, or set a higher additional factor.
Any action to exercise any of those four options would be consistent with the
statutory authorization.
51
-------
Third, the commentor cites a statement in the congressional debate to the
effect that the "bill does provide the additional flexibility to apply a safety factor of
less than 10-fold...." Again, this statement is consistent with a reading of the
statute as not imposing an upward directional constraint.
Fourth, the commentor claims that nothing in the legislative history mentions
using a factor greater than 10X. However, OPP does not believe that the plain
language of a statute can be countermanded simply because the legislative
history does not explicitly state that the language means what it says.
Finally, the commentor points to a document concerning an additional safety
factor for children prepared by OPP scientists for presentation to the FIFRA
Scientific Advisory Panel in October, 1996. That document was released
shortly after promulgation of the FQPA. In the document, OPP repeats the "up to
10X" formulation in the NAS Report. Although that document discusses the
FQPA children's safety factor provision, the original impetus for bringing the
issue to the SAP had been the NAS Report and the SAP paper had been in
production long before the passage of the FQPA. When the paper was revised
in light of SAP comments and further reflection on interpretation of the statutory
language, it was also amended to drop the description of the FQPA provision
as requiring "up to" an additional 10X and to include the term "different" margin
of safety (see USEPA 1997).
Accordingly, OPP concludes that nothing the commentor has cited in the
legislative history or from any OPP interpretations provides any authority for
overriding the plain language of the statute.
52
-------
Comment 13. Commentor 372 recommended that if a benchmark dose (BMD) is
used rather than a LOAEL, the uncertainty factor employed for use of a LOAEL
rather than a NOAEL (typically a 10) should be smaller, since the BMD presumably
provides a better estimate of the threshold than does the LOAEL.
Response. OPP agrees that a Benchmark Dose (BMD) may be a better
approach to derivation of an RfD or RfC, but the Agency guidance on the use of
BMD is not yet final. It is anticipated that the Agency's Benchmark Dose
Technical Guidance Document {US EPA 2000d) will be finalized in early 2002,
and when final, OPP will follow this guidance. This document, which was
developed under the auspices of a Risk Assessment Forum Technical Panel,
has received Agency review and is currently undergoing external review. It
includes a chapter on using the BMD in dose-response assessment, and
specifically addresses the application of uncertainty or adjustment factors to the
point of departure (POD), the magnitude of which should be determined case by
case.
53
-------
II. Toxicology Issues
FQPA directs EPA to apply "an additional tenfold margin of safety...to take into account
potential pre- and postnatal toxicity and completeness of the data with regard to exposure
and toxicity to infants and children." The statute further provides that EPA may "use a
different margin of safety...only if, on the basis of reliable data, such margin will be safe for
infants and children." Thus, the statute directs EPA, for each pesticide, to focus on three
points: (1) its potential pre-and postnatal toxicity to infants and children; (2) the
completeness of its exposure database; and (3) the completeness of its toxicity database.
In the Federal Register Notice announcing the availability of the draft policy document
and other related papers, the Agency posed a series of questions concerning the first and
third points of the FQPA Safety Factor provision relating to toxicology (see Toxicology
Issues: questions 5-10). This introduction provides a brief overview of OPP's approach to
applying the factor concerning the completeness of the toxicology database. The
Agency's response to the comments received on questions 5-10 posed by the Federal
Register Notice of Availability follow the introduction.
Completeness of the toxicity database. Since the statute directs OPP to consider the
completeness of the toxicity database for a pesticide, the draft guidance document
devoted considerable attention to discussing what constitutes a complete toxicity
database. Consistent with the basic approach of the overall guidance, OPP will make this
decision on a case-by-case basis. Nonetheless, as discussed in more detail below, OPP
has decided to apply in the revised guidance the following analysis, which draws on
Agency-wide policy, public comments, and insights gained through several years'
experience applying the FQPA Safety Factor provision. Because the FQPA safety factor
provision regarding the completeness of the toxicity database is similar to the traditional
database uncertainty factors used by the Agency to address inadequate or incomplete
data, when deriving RfDs, OPP intends to use standard Agency practice by applying
database uncertainty factors based on considerations of the available toxicity,
toxicodynamic and toxicokinetic data. The traditional database uncertainty factors are
intended to account for the potential for deriving an under protective RfD/RfC as a result of
an incomplete characterization of the chemical's toxicity. Relying on the Agency's existing
practice and policy of applying a database uncertainty factor will help ensure consistency
in approach and that the RfDs produced by OPP and other EPA program offices should
usually be the same for the same chemical.
The guidance document describes a set of factors considered in identifying the
significance of missing data. Mandatory and conditional data requirements for pesticides
are contained in 40 CFR 158. Further, OPP has long exercised its discretion to impose
data requirements on pesticides beyond those contained in 40 CFR 158, and therefore
also decides on a case-by-case basis if it needs additional special toxicity studies beyond
those routinely or conditionally required for particular pesticides in order to evaluate the
54
-------
potential hazards of the chemicals. All of these studies, whether routinely or conditionally
required or required due to the special characteristics of a pesticide or group of
pesticides, potentially bear on the risks posed to infants and children. Accordingly, the
"completeness" inquiry should be a broad one that takes into account all data deficiencies.
In other words, the risk assessor should consider the need for traditional uncertainty factors
not only when there are inadequacies or gaps in currently required studies on pesticides,
but also when other important data needed to evaluate potential risks to children are
missing or inadequate. This is similar to general Agency practice where the absence of
core or routine required studies or required triggered or special studies generally prompts
consideration of a traditional database uncertainty factor. While the Agency's general
practice has been, as a default, to apply a traditional UF of 3X, if one core study is
missing, or 10X, if more than one core study is missing, OPP intends to make this
determination on a case-by-case basis considering which studies are missing, in addition
to how many studies are missing. Having made a decision relating to the need for and
size of the traditional database UF when studies are missing, OPP will generally not
impose any special FQPA safety factor with respect to the absence of those same
studies OPP believes that the use of the traditional database uncertainty factor in the
manner described the policy statement should address any concerns about the
"completeness of the toxicity database" pertaining to missing studies.
The absence of the Development Neurotoxicity (DNT) study as well as any other study
important to evaluating effects in the young, will be considered under the traditional
database uncertainty factor analysis like any other missing study. In other words, when
data deficiencies exist, the risk assessor should consider the general, overall value of the
particular type of study to the risk assessment. Information about the potential adverse
effects of a chemical substance should take into consideration all relevant data, as well as
generally how likely those effects are to be the most sensitive toxic endpoint on which the
RfD or other hazard value is based. The analysis of data gaps should evaluate the overall
value of the missing study to the risk assessment process, including characterization of
effects on the young. Thus, in addition to identification of toxicity information that is lacking,
review of the available data may also provide information as to this potential. In deciding
to apply a database uncertainty factor to account for missing studies, the risk assessor
should evaluate how thorough the testing is with respect to life stage assessment, endpoint
assessment, route and duration of exposure. It should be emphasized that studies using
adult animal may help inform the judgment about potential effects in the young and the
need for additional studies. The size of any FQPA safety factor would depend on the
degree of concern about the pesticide's potential to affect the young and would take into
account the potency, severity, and persistence of the observed effects.
55
-------
ISSUE 5. Definition of Core Toxicity Database
Please comment on OPP's proposed criteria for defining the core toxicology
database.
Comment 1. Several commentors (369, 372) suggested that OPP's proposed
criteria for defining the core toxicology database could be more clearly presented.
There was some confusion on which of the recommendations of the Toxicology
Working Group were being adopted by OPP. One commentor (776) fully supported
OPP's proposal for the core toxicology database. Several others supported the
proposal with certain qualifications. One commentor (369) stressed that one
criterion for inclusion of a study in the core database should be that peer-reviewed
and publicly available guidelines of validated study types (accepted by the scientific
community) are available. Another commentor (773) agreed with the proposed
core database, but stated that a study should be considered part of the core data
set only if OPP has promulgated an appropriate regulation; a "Data Call-In" does
not substitute for rulemaking.
Response. In recognition of the commentor's confusion about how OPP
responded to the recommendations of the Toxicology Working Group
concerning the definition of a "core toxicology database," OPP has substantially
revised the Guidance Document to clarify its approach to evaluating the
adequacy of the toxicity. Please see Section III of the revised guidance
document; also see the introduction at the beginning of this section. It should be
noted that the revised guidance responds to the Agency's RAF Technical Panel
recent recommendation in its 2001 draft report that "The Technical Panel
agrees with the Toxicology Working Group of the 10X Task Force (USEPA
1999a) that an additional default child-specific factor beyond the interspecies,
intraspecies, and database deficiency uncertainty factors is not necessary, if
appropriate care has been taken in accounting for all deficiencies and
uncertainties in the database using the currently available uncertainty/variability
factors." The 1993 NRC report on Pesticides in the Diets of Infants and
Children (NRC 1993) indicated that additional protection for developmental
toxicity (essentially an additional 10-fold factor) may be required, depending on
the toxicant of interest and the amount of testing that has been conducted. The
recommendation by the RfD/RfC Technical Panel is viewed as consistent with
the NRC report on Pesticides in the Diets of Infants and Children (NRC 1993).
This is because use of the uncertainty factors in the manner recommended by
the Technical Panel addresses the NAS' concern regarding the completeness
of data pertaining to the safety of infants and children. In many respects, the
additional 10-fold factor for infants recommended by the NRC (NRC 1993) and
by FQPA is similar to the traditional database uncertainty factor used by the
56
-------
Agency to address inadequate or incomplete data (including data on pre- and
postnatal toxicity) during the risk assessment process. OPP, therefore, believes
that when data are inadequate to evaluate potential health risk to children or
other potentially sensitive subpopulations, then the traditional uncertainty factors
should be used to account for the uncertainty raised by incomplete data. The
FQPA factor should be considered in the risk characterization and used to
account for any residual concerns or uncertainties in either the hazard
characterization or exposure assessment. Therefore, OPP does not make the
distinction between core required data versus newly required data in the revised
guidance. See response above.
Comment 2. Many com mentors (L07, L29, L32, L033, 6-56, 58-191, 193-368,
370, 371, 373-400, 407-516, 518-522, 524-567, 569-614, 651-740, 742-760, 762-
769, 774, 777, 779 780, 782-797, 799-802, 806, 807, 811, 812, 815, 817, 818,
820-823, 825, 828, 829) disagreed with OPP's proposal due to animal welfare
concerns; increasing the number of toxicology studies in the core database would
increase the use of animals. A large number of individuals repeated this concern
through submission of a form letter.
Response. EPA is sensitive to public concerns about the use of animals in
chemical testing and observes certain guiding principles in all of its testing
programs. EPA is committed to avoiding duplicative testing, and to reducing
the numbers of animals used, refining existing study types to mitigate pain and
suffering and enhance the information gained from the existing guidelines, and
replacing animals in testing when scientifically valid alternatives exist. OPP has
taken every effort to include only the studies necessary to characterize the
potential hazards to children as required by FQPA in the proposed core
toxicology database. EPA encourages that toxicology studies be combined
when possible to help reduce the number of animals. In addition, EPA is
working within the framework of the Interagency Coordinating Committee for the
Validation of Alternative Methods (ICCVAM), and, internationally, with the
Organization for Economic Cooperation and Development (OECD) to ensure
the scientific acceptability of alternative methods.
Although the Agency agrees that the use of animals in testing should be
reduced, refined, and replaced whenever scientifically appropriate, as stated
elsewhere in these responses to comments, OPP does not believe that the
application of a 10-fold factor to the risk assessment for every chemical is an
adequate alternative to animal testing. Identification of potential hazard, dose-
response assessments, and evaluation of age-related differences in response
to toxicological insult, all of which are critical aspects of risk assessment and
FQPA factor determination, should include characterization through the use of
57
-------
an integrated animal systems, i.e., in vivo testing given the complexity of the
developing organism. This issue is discussed by Chengelis et al. (1995), who
concluded, following the review of a number of publications that examined this
topic, that mathematical models and simple cell culture systems cannot be used
in place of intact animal models to predict toxicities in humans, since "the final
expressions of toxicity in humans or animals are frequently the summation of
extensive and complex interactions at cellular and biochemical levels." This can
only be observed in a whole, intact animal system. Or, in the opinion of Chapin
and Heindel (1993), "...in vitro methods are important for addressing specific
mechanistic questions.... But without the demonstration of an in vivo effect at
some point in the investigation, in vitro data are like a headless chicken: active,
provocative, and attention-getting, but missing something quite important." OPP
also points out that the 1993 NAS report (Pesticides in the Diets of Infants and
Children) did not recommend that animal testing be replaced with a
presumptive 10-fold factor, but actually suggested that more extensive animal
testing be performed to evaluate potential toxicity to infants and children.
Therefore, the OPP position, as presented in the 10X policy paper, is consistent
with the NAS recommendations.
Comment 3. One commentor (369) requested clarification on whether OPP would
consider the availability of data on cholinesterase activity for compounds known to
inhibit this class of enzymes.
Response. For a cholinesterase-inhibiting pesticide, OPP thoroughly
evaluates all available cholinesterase activity data as part of the hazard
identification and dose-response assessment. These data, although not
specifically addressed in all OPPTS guidelines, have been required of
registrants since 1992, as specified in a Data Call-In (DCI) for adult
neurotoxicity studies that was issued at that time. The need for adequate
information on cholinesterase inhibition in young animals has been recognized
by OPP. In order to address this concern for the cholinesterase-inhibiting
organophosphorus pesticides, OPP has included a requirement for submission
of comparative blood and brain cholinesterase measurements in dams and
offspring in the September 10, 1999 DCI that was issued to require submission
of developmental neurotoxicity studies (as well as any outstanding adult
neurotoxicity data). OPP intends to include similar requirements in future DC Is
that will be issued, in accordance with the process described in the 10X policy
paper, for the other class of cholinesterase-inhibiting neurotoxic chemicals (e.g.,
the N-methyl carbamates). OPP also is encouraging the collection of enzyme
activities in peripheral tissues. For further details on how OPP interprets these
data in the risk assessment process, the reader is referred to the recently-
issued OPP science policy paper entitled The Use of Data on Cholinesterase
58
-------
Inhibition for Risk Assessments of Organophosphorus and Carbamate
Pesticide (US EPA 2000a).
Comment 4. One commentor (778) stated that the proposed core toxicology
database was inadequate to assess the potential hazards to children for several
reasons including the lack of toxicology studies for assessing all developing
systems, particularly the immune and endocrine systems, and the lack of toxicology
studies that assess the organism throughout its life span. In addition, comments
were made that potential developmental neurotoxicity should be assessed for all
chemicals.
Response. OPP agrees with the commentor that, as scientific understanding
of the potential toxicity of chemicals to humans grows, EPA needs to reexamine
its policies regarding the types of data that are useful for risk assessment. With
that in mind, in June 1999, the Agency's Risk Assessment Forum formed a
Reference Dose (RfD) Technical Panel to reevaluate the Agency's RfD
methodology. As part of this effort, the RfD Technical Panel is evaluating all of
the toxicology test guidelines and study designs the Agency most frequently
encounters in reference to the specific endpoints and life stages to determine
which elements are adequately assessed and which are not. When this analysis
is complete, the Agency will commence the updating of existing guidelines
and/or development of new test guidelines.
Clearly the fields of toxicology mentioned in the comment-immunotoxicity,
endocrine system effects, and developmental toxicity studies-are currently
some of the more active areas for research and development of new testing
methodologies. While there is a lot of new information, OPP does not think that,
at this time, the scientific foundation exists for making most of these types of
studies routine requirements. Nonetheless, EPA is moving to bring the
research effort in these fields to a stage that would warrant its use in a regulatory
context. Work is currently underway in the Agency to develop a testing protocol
for developmental immunotoxicity. The Developmental Immunotoxicity Working
Group, comprised of scientists from OPP, OPPT, ORD/NCEA, ORD/NHEERL,
and OCHP participated in a meeting on developmental immunotoxicity testing
held by ILSI/HESI in spring 2001 and continue to work with EPA's NHEERL
laboratories in researching the techniques that could be validated to predict
human hazard with respect to developmental immunotoxicity. Wth reference to
the evaluation of the endocrine system, several new endpoints which often have
a hormonally-mediated basis were incorporated into the updated two-
generation reproductive toxicity test guideline issued in 1998. In addition, EPA
is in the process of developing and validating other test guidelines to assess
impacts on the endocrine system as an element of the implementation of the
59
-------
FQPA-mandated Endocrine Disruptor Screening Program. Once these are
validated, these tests will also be incorporated into OPP's data requirements,
as appropriate. Until then, however, these studies are unlikely to be designated
as a routine data requirement. Regardless of whether a study is a routine data
requirement or not, when data gaps exist, in deciding to apply a database
uncertainty factor, the risk assessor should consider the overall value of the
particular type of study to the risk assessment. Information about the potential
adverse effects of a chemical substance should take into consideration all
relevant data, as well as generally how likely those effects are to be the most
sensitive toxic endpoint on which the RfD or other hazard value is based.
FIFRA Scientific Advisory Panel Comments
Comment 5. The SAP felt that the Agency was justified in including the evaluation
of the immune system as part of Tier 1. Guidelines for immunotoxicity testing
already exist with regard to chemicals (OPPTS 870.780) as well as for the
biochemical pest control agents (OPPTS 880 series). It is particularly significant
that a functional test for immunity be included in this data set. A test that challenges
the immune system to respond (such as the antibody response to sheep red blood
cells) is appropriate. The assay that utilizes this antigen has undergone extensive
validation. Furthermore, a considerable database exists with regard to pesticide
exposure on this response in experimental animals.
The SAP recommended that the Agency consider a flexible science-based
approach to the design and conduct of immunotoxicology studies by carefully
considering the results from the other tests proposed in Tier 1 that identify other
potential target organs and consideration of potential for recovery or transient
effects. It is cautioned that currently, predictive animal models for autoimmunity are
not well developed and the paucity of biological information on the developing
immune system represents a limitation in the identification and inclusion of such
endpoints into a testing protocol. The Agency should continue its efforts to develop
and validate protocols that are designed to evaluate the potential for chemically-
induced developmental immunotoxicity. The goal should be the creation of a
carefully designed developmental toxicity study that incorporates the evaluation of
functional immunity.
Response. OPP appreciates the Panel comments regarding the immediate
need for adult immunotoxicity screening. As the Panel noted, OPPTS already
has revised its repeated dose study guidelines to incorporate the evaluation of
additional endpoints as an immune system screen. In addition, OPP will
formally propose to include the in vitro screen as a Tier 1 requirement for certain
60
-------
categories of food-use pesticides when it proposes revisions to 40 CFR 158 in
the near future. The Agency agrees that a flexible, science-based approach to
study design is necessary. The OPPTS 870.7800 immunotoxicity guideline
states that "...the tests described in this guideline are intended to be used along
with data from routine toxicity testing, to provide more accurate information on
risk to the immune system. The tests in this guideline do not represent a
comprehensive assessment of immune function."
The need for the creation of a guideline for the assessment of developmental
immunotoxicity remains a priority in OPP, and work is underway in the Agency
to develop a testing protocol for developmental immunotoxicity (see previous
response to Comment 4.). Therefore, ongoing efforts within EPA continue to
explore updates in the science and in available testing methodologies, with the
expectation that guideline development will move forward as soon as feasible.
ISSUE 6. Approach to Calling in Neurotoxicity Data
After having considered the recommendations from the FIFRA Scientific Advisory
Panel and the Toxicology Working Group, OPP is beginning the process of calling in
data for three studies (the acute and subchronic neurotoxicity studies in adult
mammals and the developmental neurotoxicity study) for a subset of conventional
chemistry food-use pesticides known neurotoxicants. In addition, OPP will be
proposing to require the same set of studies for all conventional chemistry food-use
pesticides in the revision of the 40 CFR 158 regulations. Please comment on this
two-stage approach.
Comment 1. Three commentors (369, 776, 778) supported the OPP decision to
call in the data for the acute and subchronic neurotoxicity studies in adult mammals
and the developmental neurotoxicity study, and additionally concurred with the two-
stage approach as proposed in the 10X policy paper. Additionally, Commentor
773 agreed that it is reasonable, in principle, to require these types of studies for
compounds either known to be neurotoxic or structurally related to neurotoxic
compounds.
Response. Since the guidance document was presented to the SAP in May
1999, OPP has issued the first of the Data Call-Ins (DCIs), for 34
organophosphorus (OP) pesticides. The DCI (dated September 10, 1999) also
included a list of neurotoxic chemical classes for which future DCIs will be
generated. After the September 1999 DCI was issued, OPP and industry
scientists met to discuss topics related to study design and conduct, focusing on
specific requirements for the Developmental Neurotoxicity (DNT) study for the
61
-------
specific chemicals included in the DCI. OPP has received and reviewed the
draft DNT protocols for all the OPs that will undergo testing. OPP continues to
work with industry to facilitate the prompt initiation of the required DNT studies.
Comment 2. A number of commentors (593, 709, 712, 755, L07, L32) opposed
the addition of any animal testing to the standard battery of toxicity studies already
required to support the registration of pesticides. In addition to this generic
response, specific issues raised by commentors included opposition to the
addition of any new study without a full review of the existing 19 animal tests
required already (in accordance with guidelines drawn up by the NIH pursuant to the
1993 NIH Revitalization Act), to the proposed 21-day dermal toxicity study, and to
the addition of the DNT study that the commentors assert will kill 1,200 animals per
chemical (553, 709, 712, L32).
It was suggested by Commentor L32 that the 21-day dermal toxicity study
should not be used to extrapolate percutaneous absorption in humans, but that data
from human volunteers, from in vitro studies with human tissues and cells, and
PBPK models can be used to more accurately describe the ADME profile of a
specific chemical.
Commentors L32, 755, and L07 also stated the opinion that the addition of new
animal tests will not provide a "reasonable certainty" of the safety of chemicals nor a
"high level of confidence" that infants and children are adequately protected from
risk to pesticide exposure. Further, Commentor 755 suggests that such tests would
serve little purpose other than to provide dubious evidence to increase tolerance
levels; therefore, there would be no need for these tests to be conducted.
Commentor 775 asserts that interspecies toxicity factors may vary by factors far
greater than 10X, and that the safety of children would best be served by an
improvement of systems that take into account the metabolism and toxicokinetics of
the chemicals in human beings. Commentor L07 suggests that in lieu of animal
testing, a 10-fold safety factor should standard be added for all pesticides.
Response. Contrary to the comments, OPP is not proposing to add a
substantial number of new tests to its long-standing data requirements for
registration/reregistration of food use pesticides the evaluation of neurotoxicity
or other types of toxicity. As previously addressed in the Response to the
second comment under Issue 5, OPP is identifying, for the most part, key study
types from within the existing set of data already required and also is
encouraging the chemical sponsor to combine two or more stand-alone study
endpoint evaluations into single study designs, whenever feasible.
62
-------
That said, it is emphasized that the 10X policy paper recommends that
specific currently available toxicology studies be considered in evaluating the
completeness of the toxicology database, but did not propose the immediate
addition of any new, previously unavailable, animal study types to the database.
OPP continues to believe that consideration of the specified animal testing, as
described in the OPP 10X policy paper, will add significant value to the weight-
of-evidence evaluation that supports the determination of "a reasonable
certainty of no harm" to infants and children for any pesticide, and will increase
the overall level of confidence regarding this scientific judgement.
OPP does not think that the scientific basis has been adequately
established for using information from either //? vitro studies with human cells or
PBPK modeling in place of animal toxicity data. PBPK models are still under
development and are not yet standard methods. OPP would consider such data
on a case-by-case basis, the availability of such information would not lessen
the need for animal toxicity data.
In response to the commentor who suggested that, in lieu of animal testing,
the FQPA Safety Factor should be retained in all cases, OPP would respond by
noting that if there were no toxicity data generated to provide information upon
which a FQPA Safety Factor decision could be made, the legislative mandate
to make a "reasonable certainty of no harm" finding could not be met for most, if
not all, pesticide chemicals.
Comment 3. One commentor (788) suggested that any future developmental
neurotoxicity testing should be conducted concurrently during the course of a two-
generation reproductive toxicity study; such testing should only be required for new
pesticides and currently registered ones which do not already have a two-
generation reproduction study.
63
-------
Response. As noted elsewhere in this document, OPP encourages the
combining of testing, where practical. In this case, the Agency would find it
acceptable for a sponsor to conduct the developmental neurotoxicity (DNT)
study within the multigeneration reproduction study, but only if the combined
study includes evaluation of all of the elements cited in each individual
protocol. If it is apparent to the pesticide's sponsor that both the multigeneration
reproduction study and the DNT would be needed to support the initial or
continued registration of its chemical, the sponsor would be wise to consider the
possibility of conducting a combined study.
The Agency disagrees with the suggestion that the DNT should not be
required if a multigeneration reproduction study already exists. This study
design, on its own, does not include an adequate evaluation of nervous system
development and function.
Comment 4. Com mentors 772 and 773 disagreed that the neurotoxicity studies
should be conducted as Tier 1 studies, i.e., automatically required for all classes of
conventional chemical food-use compounds. Commentor 773 states that, until
issues regarding study design and interpretation are resolved, OPP should
proceed cautiously.
Commentor 772 provided additional extensive comments on this issue:
This commentor endorses a tiered testing approach so that additional animal use,
neurotoxicology testing resources, and EPA resources are focused on those
chemicals that actually represent a hazard. Neurotoxicological evaluations should
follow a tiered testing strategy in which the initial Tiers identify a hazard and
subsequent Tiers characterize any hazards that are identified.
Subchronic neurotoxicity tests in adult rodents should not be required for all
active ingredients. This study was designed to be, and has traditionally been, a
Tier II study. Conversely, the standard subchronic toxicology test has been a Tier I
requirement for many years. The latter test has been recently revised to include
formal neurobehavioral evaluations and enhanced clinical observations that provide
enhanced screening-level information to make an informed judgment about
neurotoxic potential. The adult neurotoxicity test battery could be triggered if the
screening data from the standard 90-day toxicology test suggest that a neurotoxic
hazard exists.
The acute neurotoxicity test in adult rodents should not be required for all active
ingredients. Instead, EPA should develop a first Tier acute toxicity test that
evaluates multiple toxicity endpoints. EPA should expand the requirements of the
64
-------
currently required acute oral toxicity study to include controls and additional dose
levels so that a reliable acute NOEL can be established. EPA recently provided
more detail in the description of clinical observations in the acute oral toxicity test
guidelines. These clinical observations can be conducted frequently which is an
important advantage over the endpoints required by the acute neurotoxicity study in
screening for acute effects. The acute neurotoxicity study could be triggered if
either the acute or subchronic repeated dose study suggests that a neurotoxic
hazard exists.
The developmental neurotoxicity (DNT) test should not be a requirement for all
pesticide active ingredients. Instead, EPA can enhance the standard reproduction
toxicity study to include endpoints that provide additional focus on developmental
neurotoxicity. EPA's proposal to require the DNT test as a first Tier test for all
pesticides is inadequately justified. EPA notes that Goldey et al. (1995) showed
that 50-70% of pesticides causing developmental neurotoxicity did not alter
traditional developmental landmarks and growth indices. Further, Goldey et al.
(1995) classified chemicals as a developmental neurotoxicant if they affect any of
the neurological domains, regardless of their mechanism or duration of action on
the nervous system. Goldey at al. (1995) then determined if any of these chemicals
had effects in the Chernoff/Kavlock assay which has limited in utero exposures and
limited data collection during the early postnatal period. This comparison
overstates effects as evidence of developmental neurotoxicity and underestimates
the screening ability of currently required developmental and reproductive toxicity
studies. The outcome of this comparison does not support the conclusion that all
compounds should therefore be subjected to the DNT battery. In fact, a major
conclusion from Goldey et al. (1995) is that "the importance of exposure periods
that span the development of the nervous system (which in rodents extends well into
the postnatal period) cannot be overemphasized in the hazard identification of
potential developmental neurotoxicants." An alternative interpretation of Goldey et
al. (1995) is that a standard reproduction toxicology study will more effectively
screen for potential developmental neurotoxicants than the Chernoff/Kavlock assay.
The commentor also disagrees with the EPA interpretation of A Retrospective
Analysis of Twelve Developmental Neurotoxicity Studies Submitted to the
USEPA Office of Prevention, Pesticides, and Toxic Substances (OPPTS), a
review of the results of DNT tests for 12 neurotoxicants (USEPA 1998c). EPA
partitioned the data for the 12 chemicals in various ways to support the need for
DNT tests for all chemicals. But EPA ignored the key outcome that the DNT tests
did not cause a single chemical on the list of 12 to be regulated at lower exposure
levels than the levels determine by other more traditional toxicology tests.
Response. OPP recognizes the need to improve and revise its data
requirements for pesticides. Since the promulgation of FQPA, a number of
65
-------
activities have been ongoing within and outside the Agency to evaluate the
types of testing approaches that would provide more efficient and thorough
evaluation of potential human risks, including children's risks. This includes
consideration of the need for new studies as well as the need to modify existing
guideline studies to provide a more comprehensive coverage of life stages, a
more systematic evaluation of toxicokinetics, and a more focused evaluation of
structural and functional toxicity in the young. For example, OPP plans to publish
proposed revisions to its pesticide data requirements regulation, 40CFR158.
and expects to ask for comment on a requirement for developmental
neurotoxicity testing, which utilizes information about each chemical and its
toxicity to develop a rational, science-based approach to the study design and
testing strategy. OPP further acknowledges that the scientific community is
developing, or in some cases already utilizes, other studies for evaluating the
young which are not required studies and for which there are no formal,
standardized test guidelines. There are ongoing activities within OPP and the
Agency to consider the need for other guidelines or studies important to
evaluate risk in infants and children, such as toxicokinetics in fetuses and/or
young animals, direct dosing of the offspring prior to weaning, enhanced
developmental neurotoxicity studies including specialized testing of sensory
and/or cognitive function, developmental immunotoxicity, and enhanced
evaluations of the potential to induce effects related to endocrine disruption.
These areas represent possible future revisions to current guidelines or
possible development and implementation of new guidelines.
OPP intends to seek comment on whether the acute and subchronic
neurotoxicity studies in adult animals be routine data requirements. These
studies are currently conditionally required ("triggered" by some special
characteristic of the pesticide or by potential use and exposure patterns or by
the results of the required studies) but are viewed as important data, given the
well established experience with these studies and their value in characterizing
neurotoxicants.
Issues raised by Commentor 772, regarding the interpretation of the
retrospective analysis of DNT studies submitted to OPPTS, are addressed in
depth in the OPP response to comments on Comment 9.2. The argument that
the DNT studies examined did not provide any information that was critical in the
regulation of the specific chemicals examined is not, in fact, correct.
Additionally, A Retrospective Analysis of Twelve Developmental Neurotoxicity
Studies Submitted to the USEPA Office of Prevention, Pesticides, and Toxic
Substances (OPPTS) (USEPA 1998c) provides extensive discussion
regarding the use of the DNT study in risk assessment, and affirming the value
of the study for this use.
66
-------
Comment 5. The commentor also recommended that an acute toxicity test be
developed which would allow a dose-response assessment and could potentially
provide endpoints for acute dietary risk assessment.
Response. There are efforts underway, both domestically and internationally,
to develop a single-exposure toxicity test guideline that would evaluate a wider
variety of endpoints than does the current acute neurotoxicity protocol. This test
would be used when the need for an acute dietary risk assessment was
identified. This determination would be made upon evaluation of toxicity
information gathered in the first Tier of mandatory testing for a food-use
pesticide. A draft guideline was prepared at the recent WHO/FAO Joint
Meeting on Pesticide Residues. This document, along with a commitment to
serve as sponsor for the development of an OECD guideline, was forwarded to
the manager of the OECD Test Guidelines Programme, for that organization's
consideration. These two items can be found on the WHO website at
(http://www.who.int/pcs/impr/jmpr.htm).
Comment 6. Commentor 772 endorses the use of the reproduction toxicity study
to screen for developmental neurotoxicity instead of OPP's proposal to required an
extensive DNT battery for all active ingredients for several reasons:
The reproduction study involves an extensive assessment of effects on
reproduction, behavior, and growth and development of the offspring. Such
information involving maternal and neonatal toxicity would add considerably to the
ability to interpret neurobehavioral changes in the developing animal.
Unlike the DNT study, the duration of exposure in the reproduction toxicity study
is most relevant to human exposure. Each chemical would be administered
continuously for an extended period of time, thereby providing the greatest
opportunity to detect possible effects on the developing nervous system by the
anticipated route of exposure.
67
-------
The reproduction toxicity study design already includes examinations of
selected F1 and F2 offspring at weaning (postnatal day 21) that are relevant to
neurotoxicity assessment: thorough necropsy examination of all P and F1 parental
animals; histopathology evaluation of reproductive organs, pituitary, adrenal glands
and target tissues; and endpoints such as weekly clinical observations, body weight
and food consumption measurements, clinical chemistry, and hematology.
A stand-alone DNT study uses a large number of animals. By obtaining
additional information from animals that are already being used in a reproduction
study, the proposal to enhance the protocol helps to reduce the number of animals
while not negatively impacting the ability to identify a developmental neurotoxicity
hazard. If screening-level developmental neurotoxicity information triggers further
investigation, the plan ensures that additional animal use is focused on those
chemicals that actually represent an hazard.
The commentor endorses the following specific enhancements to the
reproduction study:
Greater detail for clinical observations and functional tests of the offspring.
These enhancements would include observations that are performed while
the animal is outside the home cage, preferably in a standard arena, and the
use of rating scales for certain behavioral/neurological clinical signs.
Brain weight measurements on postnatal day 21 and day 60 (already
required).
Histological examination of the brain at postnatal day 21 and of the central
and peripheral nervous system at day 60.
A test of cognitive function to determine whether there is evidence of
potential effects on learning and memory.
A longer duration of exposure (from premating through PND 21) than that
used in the guideline developmental neurotoxicity study design. The longer
duration of exposure includes a period that involves continued development
of the nervous system (already required).
EPA should convene a workshop to discuss the details of a study design to
screen for developmental neurotoxicity. The participants should examine how best
to enhance the design of the reproduction toxicity study to create a first Tier screen
for neurotoxic potential.
The requirement for careful clinical examination during standard subchronic and
68
-------
chronic toxicity studies provides good guidance needed to screen animals for
neurobehavioral effects. However, the number of animals required and frequency of
evaluation should be amended so that the necessary toxicology evaluations can be
appropriately conducted in a manner that is not unnecessarily burdensome. EPA
requires a clinical examination be made at least once prior to the initiation of
treatment and once weekly during treatment in all animals. The commentor
embraces the detailed description of clinical examination, but believes that it is
more appropriate to conduct these evaluations monthly in the subchronic study and
quarterly in the chronic study. Of course, the health status of all rats would continue
to be monitored through the daily mortality checks and weekly cage-side
observations.
Response. OPP agrees with the commentor that the two-generation
reproduction study can be expanded to accommodate developmental
neurotoxicity testing. Since the developmental neurotoxicity study guideline was
finalized in 1991, the Agency has clearly indicated that, although the DNT
guideline describes a stand-alone study, a well-conducted combined
DNT/reproduction study would be considered adequate to meet regulatory
requirements. OPP sees no benefit in conducting a combined study if a valid
reproduction study already exists, and, in fact, would consider a combined study
in such a situation to be a waste of resources. The guideline itself contains
language that states this to be the case: "This protocol may be used as a
separate study, as a follow up to a standard prenatal developmental toxicity
and/or the (repeat dose) adult neurotoxicity study, or as part of a two-generation
reproduction study, with assessment of the offspring conducted on the second
(F2) generation." [OPPTS 870.6300 (c)]. Both the availability of a stand-alone
protocol and the authorization to conduct the study in a combined protocol, are
necessary to accommodate the testing needs for all pesticide chemicals,
whether related to registration or reregistration actions.
There are reasons that a combined two-generation
reproduction/developmental neurotoxicity study would be particularly useful, and
many of these considerations have been described in the comments. These
include evaluation of a population of offspring with maximized exposure duration
(i.e., that have been treated throughout pre- and postnatal life), assurance that
steady state levels of test substance in the animals have been achieved prior to
neurotoxicological testing, evaluation of neurobehavioral effects within the larger
context of assessments of maternal and neonatal toxicity offspring growth and
development, and better utilization of animals already on study.
69
-------
The proposal for enhancements to the reproduction study that are designed
to screen for developmental neurotoxicity includes functional observation battery
(FOB) assessments of pups, brain weight measurements at two time points,
qualitative neuropathology at two time points, and learning and memory testing.
The proposal does not include assessments of the ontogeny of motor activity or
auditory startle habituation (two particularly sensitive endpoints, as
demonstrated in USEPA 1998c), mention of a second time point for learning
and memory testing, or a quantitative evaluation of brain development in the
neuropathology examination. With the addition of these few missing items, the
commentor's proposed "enhanced reproduction study" would be equivalent to a
combined reproduction/DNT study.
The developmental neurotoxicity study protocol provides an assessment of a
wide selection of endpoints and time points in order to maximize the potential
for detection of effects on the developing nervous system. By reducing the
number of sensitive endpoints examined or by dropping the assessment of
ontogeny of nervous system development, as is proposed in the enhanced
reproduction study, the likelihood of detection of developmental neurotoxic
potential would be greatly reduced. Therefore, OPP considers the guideline
DNT study to be the representative screening assessment for developmental
neurotoxic potential, whether conducted as a stand-alone protocol or when
incorporated into a standard two-generation reproduction study. In addition,
OPP believes if the DNT is incorporated into the multigeneration reproductive
toxicity study, the combined study should incorporate all of the endpoints
required in the stand-alone DNT, as well as all of those required in a stand-alone
multigeneration reproduction study. Only under these circumstances would the
combined study to be considered an acceptable study.
Comment 7. Commentor L015 stated that, as part of the guideline development
process, EPA's criteria for interpretation of the results of the studies should be
developed and subjected to public comment.
Response. OPP scientists review all toxicology study data (including the
developmental neurotoxicity study) in detail, and interpret the findings utilizing
time-honored principles of basic scientific data analysis. These reviews are
conducted according to applicable peer-reviewed Agency risk assessment
guidance that specifically address the interpretation of data, e.g., the Guidelines
for Developmental Toxicity Risk Assessment (USEPA 1991) and the
Guidelines for Neurotoxicity Risk Assessment (USEPA 1998a). (Several
other peer-reviewed Agency guidelines also exist for other endpoints of toxicity.)
Further, to provide consistency in the reporting of developmental neurotoxicity
studies, OPP (in cooperation with Health Canada PMRA) has recently created a
standardized format that can be utilized in the generation of Data Evaluation
70
-------
Reports.
Regarding chemical-specific issues of data interpretation, OPP has
committed to reviewing the results of submitted developmental neurotoxicity
studies, within the context of each chemical toxicity database (including the adult
acute and subchronic studies, as well as the prenatal developmental studies
and the multigeneration reproduction study), in an external peer review process.
The first such review was conducted in December, 1998, when the FIFRA
Scientific Advisory Panel reviewed the Retrospective Analysis of Twelve
Developmental Neurotoxicity Studies Submitted to OPPTS. As more DNT
studies are submitted to OPPTS, subsequent reviews and analyses will be
conducted consistent with scientific principles. This process will be transparent
and will include an opportunity for public comment on HED's interpretation of the
study data and conclusions.
FIFRA Scientific Advisory Panel Comments
Comment 8. The Scientific Advisory Panel members agreed that the two-stage
approach for expanding the newly required test methods appears to be logical.
However, they expressed some concerns regarding what was seen as a collapse of
toxicology testing requirements into a single tier system. One member of the Panel
argued that such a system would not encompass all of the concerns that one would
have for purposes of dose-response assessment. Broad screening studies should
be used to identify hazard, and then a second tier of studies should be used to
specifically aim at establishing a dose-response relationship for endpoints most
useful for making a regulatory decision.
Response. The order of toxicological test conduct is not specified by the
Agency, but rather selected by the pesticide's sponsor. It is highly unlikely that a
sponsor/registrant will conduct all required toxicology studies simultaneously for
any specific pesticide chemical. Generally, the less-complex, shorter-duration
studies are conducted first. Therefore, information from other studies in the
database (e.g., acute, mutagenicity, subchronic, range-finding,
chronic/carcinogenicity, prenatal developmental, and/or reproduction studies)
will generally be available to contribute to study design and dose-setting
decisions for more complicated studies such as the DNT.
Comment 9. The SAP stated that, at present, the need for the developmental
neurotoxicity test seems to rest largely on the premise that it is at times the "most
sensitive" response from a dose-response perspective. The same argument could
be attributed to the endocrine system or even control of intermediary metabolism
71
-------
(e.g., cholesterol synthesis). Clearly, sensitivity arises from specificity in the
measurements one can identify to detect adverse effects and functional endpoints
with other organ systems. It was the opinion of the Panel that nervous system
evaluations come to the forefront because functional measurements are so much
richer than those evaluations applied to other organ systems. The opportunity to
refine the developmental neurotoxicity testing battery should not be missed. While
the endpoints in the current battery assess the integrative functioning of the sensory,
motor, and cognitive systems with supportive neuromorphology measurements, the
limited exposure via the mother/dam may not provide adequate or accurate levels
of exposure to the offspring to assess neurotoxicity. Aspects of the dosing
paradigm to consider are the extension of exposure to postnatal day 21 (consistent
with the OECD guidelines), direct administration of the compound to the offspring
after birth, and shorter intervals of exposure, including direct acute exposure during
development.
Response. OPP agrees with the Panel's assertion that the most sensitive
response in offspring could reside in a physical or functional endpoint in a tissue
or organ system which is not yet evaluated in depth within the context of existing
guideline toxicology studies. The availability of validated testing protocols for
the assessment of every possible biological endpoint, however, is not realistic.
(It is noted that several of the endpoints mentioned by the SAP are identified in
the 10X policy paper, for the development of further guidance.) The DNT study
is an exception, having been subjected to extensive scientific evaluation over a
decade ago, (as reported by Kimmel et al. 1990 in a special issue of
Neurotoxicology and Teratology, entitled Qualitative and Quantitative
Comparability of Human and Animal Developmental Neurotoxicity) and having
been peer reviewed by the FIFRA Scientific Advisory Panel, as well as through
a formal public comment process, prior to guideline finalization.
An analytical review of the developmental neurotoxicity studies that had been
received and evaluated by OPPTS from the time that the guideline was
developed until late 1998 was presented to the SAP in December 1998. Issues
relevant to potential guideline update and improvement were discussed in that
analysis and in the Panel members' responses to questions from OPP. Since
that time, OPP has, through a cooperative agreement, participated in a project
coordinated by ILSI Risk Sciences Institute to address many of these same
issues. Efforts to coordinate and harmonize with guideline development efforts
by the OECD, which also address some of these revision concerns, are also
ongoing.
While OPP has determined that DNT guideline revision activities should
proceed formally in the context of the OECD guideline development process,
OPP scientists have attempted to address some of the more immediate
72
-------
concerns in the Data Call-In process, particularly in areas where there appears
to be preliminary agreement that change/upgrading is needed. Specifically, the
DCI language that was issued for the organophosphorus (OP) pesticides on
September 10, 1999 has included modifications or additional requirements for
the conduct of the DNT study for this class of substances, including the following:
extention of the duration of dosing beyond postnatal day 10 to postnatal
day 21;
the need for a reliable estimate of the presence of test substance in the
milk, with a directive to consider other methods of exposure, including the
possibility that direct administration to offspring may be necessary;
an increase in the number of animals evaluated for neuropathology;
comparative evaluation of cholinesterase inhibition and behavior in
young organisms and adults.
The additional requirements, and the way in which these requirements will
need to be addressed in the protocols for these specific DNT studies, will
respond to the concerns raised by the Panel, at least for the OPs. Future DCIs,
for other chemical classes, are similarly expected to address these concerns,
along with any other chemical-specific considerations that are identified.
Comment 10. The SAP commented on the fact that the Agency is beginning the
process of calling in data for the developmental neurotoxicity study for a subset of
conventional chemistry food-use pesticides known to possess neurotoxic potential.
They stated that there is a certain logic in using known neurotoxic pesticides as the
initial test cases from which to gain knowledge and experience in the evaluation of
data from the newly required neurotoxicity studies. That is the case because there
should be a greater likelihood of at least some degree of neurotoxic effects
observed in tests of this subset of pesticides. However, there is also the likelihood
of bias from this data set of known neurotoxicants. Alternatively, the Agency should
consider that selecting a few pesticides from the universe of those that do not act
by neurotoxicity mechanisms could be instructive for comparison with
representative samples of the neurotoxicants (e.g., organophosphates, carbamates
and synthetic pyrethroids) in the developmental neurotoxicity studies. This would
allow the Agency to more accurately assess the sensitivity gained with the
developmental neurotoxicity data.
Response. OPP recognizes the challenges that will present themselves when
the DCI-initiated developmental neurotoxicity study data for the neurotoxic
73
-------
pesticides become available for review and analysis. Analysis of the results of
these studies alone will not be able to fully address the question of whether the
DNT study should be conducted routinely for all chemicals. Nevertheless, these
data, although biased towards the expectation of detection of an effect on the
neurological system, will aid in identifying differences in response between the
adult and developing animal. Therefore, the data will provide an answer to the
question of whether this specific, newly-added trigger for DNT testing was
predictive of differences in sensitivity. It is further noted that there have been a
number of pesticides that have been identified over the past decade as meeting
various criteria that would suggest that a DNT study should be performed.
Some of these chemicals are not considered to be neurotoxicants in the classic
sense (e.g., their principal mode of pesticidal action). Data from these
chemicals, which will also be called in through the DCI process, should provide
some basis for scientific comparison, but may still not definitively address the
question of whether the DNT study should be routinely conducted for all
chemicals. OPP plans to issue DCIs for some of these during the early phases
of the DCI process.
74
-------
ISSUE 7. Application of an FQPA Safety Factor Pending Receipt
of Newly-Required Studies
The OPP policy guidance indicates that one of the critical issues is whether or not to
apply an FQPA Safety Factor pending receipt of newly-required studies. There are a
variety of possible approaches. One possible approach would be to apply the FQPA
Safety Factor's database uncertainty component to gaps related to new core data
requirements only where there are specific concerns regarding the pesticide
pertaining to the data requirement. Alternatively, OPP could apply the default 10X
factor (or some other additional factor) whenever a new data requirement is added
and/or whenever a testing guideline is changed. Please explain how you think the
FQPA Safety Factor provision should be implemented when OPP makes such
changes. In commenting, please address whether OPP should apply the default
FQPA 10X factor, some different yet additional factor, or no factor at all in the
following circumstances:
A minor change to testing guidelines.
A major change to testing guidelines.
An addition of a new required test.
* An addition of a new required test to core requirements.
Comment 1. Several commentors (372, 773, 781, L016) supported the proposed
redefinition of the "core" toxicology database. Commentor 372 (American Water
Works Association) stated that the clear criteria OPP laid out for electing to require
a new methodology as part of the core toxicology database, represent sound
science. Specifically, Commentor 372 indicated strong support for what they
interpreted to be the OPP position, i.e., that new kinds of toxicological studies will
not be placed into the core toxicology database until it is clear that those studies
provide information that is essential in deciding whether infants and young children
are more susceptible/sensitive than the rest of the population. It was the opinion of
this commentor that if new studies (e.g., different kinds of neurotoxicity
assessments) are required as part of the core toxicology database, it will be
increasingly difficult for OPP to justify any numerical value for the FQPA Safety
Factor other than 10 until specific pesticides are reevaluated using the data from
the new studies. Additionally, Commentor 372 suggested that such broad
application of the 10-fold FQPA factor might be necessary even if the new studies
do not significantly improve the ability of the Agency to determine whether existing
Reference Doses (RfDs) are protective of infants and young children.
75
-------
One commentor (781) interpreted the draft 10X guidance document to state that
an FQPA Safety Factor would only be applied for missing data if those data were
considered to be part of the minimum core data requirements, and on that basis
disagreed with the use of the redefined core data set.
Commentor 776 recommended that EPA include the DNT study in the list of
core toxicology data that would lead to the application of a database uncertainty
factor if they were missing in the database.
Response. As discussed in responses to comments above, OPP does not
retain the distinction between core required data versus newly required data in
the revised guidance. The absence of any study important to evaluating effects
in the young will be considered under the traditional database uncertainty factor
like any other missing study. In other words, when data deficiencies exist, the
risk assessor should consider the general, overall value of the particular type of
study to the risk assessment. Information about the potential adverse effects of
a chemical substance should take into consideration all relevant data, as well as
generally how likely those effects are to be the most sensitive toxic endpoint on
which the RfD or other hazard value is based. The analysis of data gaps should
evaluate the overall value of the missing study to the risk assessment process,
including characterization of effects on the young. Thus, in addition to
identification of toxicity information that is lacking, review of the available data
may also provide information as to this potential. In deciding whether to retain
the default 1OX factor or to apply a database uncertainty factor to account for
missing studies, the risk assessor should evaluate how thorough the testing is
with respect to life stage assessment, endpoint assessment, route and duration
of exposure. It should be emphasized that studies using adult animals may help
inform the judgment about potential effects in the young and the need for
additional studies. The size of any FQPA safety factor would depend on the
degree of concern about the pesticide's potential to affect the young and would
take into account the potency, severity, and persistence of the observed effects.
Also see Section III of the Revised Guidance Document.
Notwithstanding commentor 781's characterization of the draft policy paper
regarding the application of the FQPA Safety Factor in situations where data
are missing, in the revised guidance document, the absence of developmental
neurotoxicity study would be treated like any other missing data.
76
-------
Furthermore, OPP recognizes the need to improve and revise its data
requirements for pesticides. Since the promulgation of FQPA, there have been
a number of activities ongoing within and outside the Agency to evaluate the
types of testing approaches that would provide more efficient and thorough
evaluation of potential human risks, including children's risks. This includes
consideration of the need for new studies as well as the need to modify existing
guideline studies to provide a more comprehensive coverage of life stages and
a more systematic evaluation of toxicokinetics. For example, OPP plans to
publish proposed revisions to its pesticide data requirements regulation, 40
CFR 158, and expects to ask for comment on including the developmental
neurotoxicity study as a standard requirement or using a logic- and information-
based approach that would aid in designing a mechanistic approach to evaluate
developmental neurotoxicity and/or other functional effects based on what is
known about the pesticide.
Comment 2. Retention of the full 10X FQPA factor in the absence of any single
piece of toxicity or exposure data that could reliably inform the Agency about
potential risks to infants and children was supported by several commentors (761,
771, 778, L021). The commentors stated specifically that the absence of a newly-
required study, or one that is required by Data Call-In, should in all cases
immediately lead to the retention of a 10-fold uncertainty factor. Commentor 778
rejected the EPA criterion that the absence of a newly-required study should lead to
retention of some database uncertainty factor only if the new data are so "key as to
warrant an additional uncertainty factor;" Commentor 771 stated that data
deficiencies should never be regarded as "minor," and Commentor 781 disagreed
with the use of a defined "core" toxicology data set in determining the need for an
FQPA factor. Reasons for retention of the full 10X FQPA factor (as cited in the
comments) included: (1) that application of the 10X in the absence of any data is a
statutory mandate (761, 771, 778, L021); (2) that the retention of an FQPA Safety
Factor would provide industry with an incentive for developing the data and
submitting missing studies to the Agency as quickly as possible (761, 771, 778);
and (3) that the need for a 10X factor is supported by comparisons of human and
rodent data for pharmaceuticals and specific neurotoxic pesticides (L021).
According to Commentor L021, there is at least a 6000-fold difference in response
to pharmaceuticals among humans. Additionally, Commentor L021 asserts that for
lead, PCBs, and mercury, neurotoxicity testing in rodents consistently
underestimates the dose that can affect the developing animal, and the
developmental neurotoxicity study in rodents can fail to identify effect levels in
humans, by a factor of greater than 10-fold.
77
-------
Response. The draft policy clearly indicated that OPP first assumes that the full
10X FQPA Safety Factor is applied for each food-use pesticide. A subsequent
weight-of-evidence analysis of the available toxicity and exposure data may
justify the use of a different (lesser or greater) factor for each chemical. OPP
believes that this approach is consistent with the statutory requirement of FQPA.
The retention of a 10-fold traditional database uncertainty factor for missing
toxicity studies or a safety factor for missing exposure data would generally
provide the regulated industry with an incentive to quickly conduct and submit
the missing studies. However, a desire to provide industry with an incentive to
quickly produce studies should not trump all other considerations, particularly
since OPP has adequate statutory mechanism to insure that data are provided
in a timely manner.
As previously addressed in the OPP response to comments on Issue 2, the
position that a uniform 10-fold safety factor is needed in every pesticide risk
assessment to protect infants and children, based upon pharmaceutical data
and selected neurotoxic pesticide data, is not an accurate representation of
emerging scientific information on this issue. Discussion on the adequacy of the
10-fold intraspecies factor to address age-related differences in
pharmacological and toxicological responses can be found in the 10X Task
Force Toxicology Working Group Report, which summarized information
available at the time it was drafted (USEPA 1999a). Since that time, additional
studies have been conducted to further evaluate this issue and appear to
reaffirm the adequacy of the 10-fold intraspecies factor, in the vast majority of
cases.
Commentor L021 cited a paper by Rice et al. (1996) to demonstrate that
studies in animals appear to underestimate the risk to humans for lead, methyl
mercury, and PCBs. In that paper, utilizing what information was available for
each of these three neurotoxic substances, the external (administered) dose in
animal studies was compared to the estimated intake level in humans and then
contrasted with the observed internal dose (i.e., body burden, as represented by
substance levels in various tissues) in humans. This commentor asserts that for
lead, PCBs, and mercury, neurotoxicity testing in rodents consistently
underestimates the dose that can affect the developing animal, and the
developmental neurotoxicity study in rodents can fail to identify effect levels in
humans, by a factor of greater than 10-fold.
78
-------
OPP does not think that the analysis by Rice et al. (1996) supports the
commentor's conclusion. First, the comparison does not appear appropriate.
A more precise approach might have been to compare internal dose (tissue
levels) in both animals and humans, and contrast this information with known
administered dose in animals and precise analytical measurements of exposure
in humans. (OPP notes that obtaining such data is not practical for the
regulation of most pesticides, since precise data on internal dose are seldom
available.) Second, it is also noted that the chemicals represented in this
analysis were selected because of their known potent neurotoxic potential to
humans, which is mediated and exacerbated in part by their specific chemical-
physical properties (i.e., long half life) and toxicokinetic disposition following
exposure (i.e., accumulation in the body). Therefore, OPP concludes that the
results of this analysis should not unduly influence policy development for the
regulation of pesticides (although if a similar chemical profile were observed in
any pesticide undergoing OPP review, it would be addressed appropriately in
the risk assessment).
Additionally, OPP believes that the above example from Rice et al. (1996) is
more of an exception than a typical example and it is inappropriate to use this
limited analysis of extremely potent neurotoxic chemicals to suggest that animal
testing cannot, as a general rule, be used to adequately predict human hazard
and dose response for the purposes of risk assessment. In fact, the opposite
position, that other organisms can serve as accurate predictive models of
toxicity in humans, is a key assumption in modern toxicological testing (see
standard toxicology textbooks such as Casarett and Doull's Toxicology
(Casarett and Doull 1995), Hayes' Principles and Methods of Toxicology
(Hayes 2000), and Regulatory Toxicology by Chengelis et al. 1995). Standard
risk assessment practices, which rely upon dose response information
generated in animal studies and known or anticipated exposure levels to
humans, are described in multiple Agency policy documents, such as the
guidelines for developmental (USEPA 1991), reproductive (USEPA 1996a),
and neurotoxicity (USEPA 1998a) risk assessment, and in numerous scientific
analyses and commentaries on the risk assessment process that exist in the
peer reviewed literature such as the 1983 and 1994 NAS/NRC reports (NRC
1983 and NRC 1994)
In summary, OPP believes that the best approach to the determination of an
appropriate FQPA Safety Factor for any pesticide will involve a case-by-case
weight-of-evidence analysis of all available toxicology and exposure data, rather
than a blanket application of a 10-fold FQPA factor for all pesticides. Such an
approach is scientifically defensible and consistent with current Agency
practices.
79
-------
Comment 3. There were several comments (772, 773, L011, L016) that focused
on the timing of the decision to retain an FQPA Safety Factor when new data
requirements were imposed (either as the result of redefining the core toxicology
database, or of issuing a Data Call-in). The opinion of these commentors was that
OPP should first communicate the need for the data requirement, and then allow a
sufficient amount of time for the registrant/sponsor to conduct the study before
imposing an additional database uncertainty factor or FQPA Safety Factor on the
basis of the completeness of the toxicity database.
Commentor 773 stated that there are legal and policy considerations that argue
for allowing pesticide registrants adequate time to conduct a newly required study
before OPP concludes that, in the absence of the study, the toxicology database is
"incomplete" and retains the FQPA Safety Factor. For a long time, the statutory
scheme has allowed EPA to impose data requirements and to apply sanctions if
the new data were not submitted in a timely fashion. The statutory scheme also
made it clear that companies were to be allowed adequate time to generate the
required data before the sanctions were applied. Nothing in the FQPA or its
legislative history indicates any Congressional intent to change this basic scheme.
Additionally, Commentor 773 predicts that serious market disruptions are likely
to occur if OPP takes any other approach. If the new study requirement
immediately triggers the additional FQPA Safety Factor, some portion of the use of
the pesticide may need to be eliminated. Yet once the study is completed, it may
appear that the additional FQPA Safety Factor was unnecessary, allowing the uses
to be reinstated.
Several commentors took the position that a decisive element in considering
whether the completeness of the data mandates retention of the additional safety
factor is whether pesticide manufacturers have been given adequate time to
produce any needed data. Commentor 773 writes:
OPP may determine that a study is needed to replace or supplement a
previously accepted study (e.g., to address recent revisions in the guidelines or a
recent reinterpretation of the earlier study), or that it wants a particular kind of study
not previously required for a category of pesticides (e.g., a developmental
neurotoxicity study). In such a situation, however, no additional safety factor should
be applied merely because the study has not been conducted, until the need for the
study has been announced and sufficient time to conduct and submit the study has
been provided.
80
-------
As a basis for this position, this commentor cites the FIFRA and FFDCA
provisions addressing regulatory requirements for data requirements and regulatory
procedures for requiring the submission of data in connection with existing
registrations and tolerances. The commentor allows OPP may retain the additional
safety factor for a pesticide if there is evidence suggesting that the pesticide may
cause adverse effects at lower levels than seen in the existing database.
Response. As explained in the Introduction appearing before Issue 5, OPP
proposed dividing the universe of toxicity data requirements into routinely
required (of which some are also core) studies and conditionally required
studies. In the revised guidance document, OPP does not make the distinction
between core required and conditionally required studies in considering the
need for a database uncertainty factors, as discussed above.
If the absence of any missing data raises concern, whether it is considered
to be part of or not to be part of the routine toxicology database, then either the
default 1OX factor would be retained or a traditional database uncertainty factor
would be applied using a weight-of-evidence approach. Thus, OPP disagrees
in part with the commentors: OPP does intend to apply a FQPA factor (either
the default value or a traditional database uncertainty factor) due to the absence
of a "supplementary" toxicity study if its absence raises concern about the
adequacy to evaluate potential effects in the young. OPP considers the FQPA
Safety Factor to encompass the traditional database uncertainty factors. As
explained in Section III of the guidance document, OPP has developed some
broad criteria to guide its general decision-making about the significance of
missing data. OPP would make the decision on a case-by-case basis.
If Agency scientists conclude, based upon available information on a
pesticide, that the absence of a certain study is critical to the assessment of
potential hazard to infants and children, then a traditional database uncertainty
factors would be applied (assuming reliable data allow OPP to conclude that
such factor is safe for infants and children), which would address the FQPA
safety factor. A DCI, addressing the specific study(ies) that are expected to
inform the scientific assessment, could be issued prior to or following the FQPA
factor determination. OPP believes that delaying the application of an
uncertainty or a safety factor to a risk assessment until identified critical missing
data are submitted to the Agency would not be consistent with the application of
the FQPA Safety Factor provision.
81
-------
OPP agrees with the commentor that its approach to applying the FQPA
Safety Factor should be consistent with the FIFRA and the FFDCA provisions
authorizing OPP to establish data requirements for pesticide registration
applications and tolerance petitions. Further, there are provisions in both
statutes mandating that OPP require registrants, pesticide manufacturers, or
others to keep databases on existing pesticides up to date. These Data Call-In
procedures specify that data submitters must be given sufficient time to submit
additional data before the sanctions in these provisions (which are designed to
ensure compliance with Data Call-ins) are applicable.
These Data Call-In provisions, however, do not address what inferences
OPP should make regarding pesticide safety while data are being produced.
OPP does not interpret the extreme sanction (suspension of registration or
tolerance) included in the Data Call-In provisions for not taking appropriate
steps to submit the data as barring OPP from making health protective
assumptions for risk assessment conducted while the data are outstanding.
Further, as noted in the draft policy, the children's safety provision also was
silent concerning how the completeness of data consideration should be
applied in circumstances involving new or revised data requirements.
OPP believes the primary guiding principle for interpreting the FQPA Safety
Factor provision should be the paramount importance of protecting the health of
infants and children. Further, the statutes' data collection provisions, including
the enforcement mechanisms in these provisions, do not, on their face, elevate
the concerns of pesticide manufacturers over the protection of infants and
children. Although whether pesticide manufacturers have had sufficient time to
produce data may have some relevance to the safety factor decision, this
consideration should not automatically trump the need to protect infants and
children. That does not mean that any data deficiency, no matter how small,
should be used to justify retaining the full 10-fold safety factor. Thus, OPP can
agree with IWG that "no additional safety factor should be applied merely
because [a] study has not been conducted...." As OPP noted in the draft policy
notice, taking such an expansive approach to implementation of the children's
safety factor provision may actually be counterproductive in terms of collecting
the best scientific data on pesticides. If any change in data requirements has
the effect of requiring that an additional 10-fold factor be applied to all
pesticides, OPP might very well be overly hesitant to require new studies in
order to keep the pesticide database up-to-date.
82
-------
Comment 4. A number of specific comments were received regarding the
appropriateness of retaining an uncertainty factor for studies that may be required
as a result of a redefinition of the core toxicology database or through a Data Call-
in. One commentor (L011) stated that retention of the FQPA factor based solely on
a change in data requirement(s) seems unfair and should not be necessary unless
additional data can show that the margins of safety are not sufficient. Several other
commentors focused their comments on the appropriateness of applying an FQPA
factor based upon the lack of adult and/or developmental neurotoxicity studies.
Commentor L015 stated that the absence of DNT and adult neurotoxicity studies
justifies an uncertainty factor, but the magnitude of this factor should be decided on
an individual basis. Commentor 773 disagreed with OPP's stated policy of using a
database uncertainty factor greater than one when a subchronic neurotoxicity study
has been triggered, but is absent. The commentor stated that OPP did not provide
an adequate explanation of this policy and that, in some cases, registrants have not
been notified that such studies are required. Commentor 772 stated that it is not
necessary to automatically add a database uncertainly factor when the DNT test is
missing. This position was based upon the outcome of the retrospective study in A
Retrospective Analysis of Twelve Developmental Neurotoxicity Studies
Submitted to the USEPA Office of Prevention, Pesticides, and Toxic Substances
(OPPTS) (USEPA 1998c), which demonstrated that the DNT studies received by
OPPTS have not caused a single chemical on the list of 12 to be regulated more
stringently than any other standard toxicology tests. The commentor concluded that
the traditional uncertainty factors of 100X appear to be sufficiently protective for
potential developmental effects, or at a minimum, that an automatic database
uncertainty factor for chronic risk assessment is not needed when a developmental
neurotoxicity study is missing.
Response. OPP believes that the decision to apply an FQPA factor cannot be
based on whether or not the margins of safety are sufficient (as proposed by
Commentor L011). This proposal is flawed in that it does not consider the
uncertainty that results from the absence of any study that is considered critical
in the assessment of hazard to infants and children. Sufficient margins of safety
cannot be determined unless hazard is fully characterized or unless the impact
of uncertainties has been addressed.
OPP agrees with Commentor L015, with the following clarification: that the
absence of the DNT and the two adult neurotoxicity studies may justify an
uncertainty factor, but that the magnitude of the factor should be determined on a
case-by-case basis for each chemical. (See comments and responses in Issue
8.)
83
-------
OPP has stated the intent to use a database uncertainty factor of greater
than one when a subchronic neurotoxicity study has been triggered, but is
absent. The fact that the study has been triggered indicates that the Agency has
determined that there is an issue of concern that could be addressed by the
subchronic neurotoxicity study for the specific pesticide under review. 40 CFR
158 already specifies that the acute and subchronic neurotoxicity study are
required for neurotoxic chemicals; therefore, sufficient notification of the
triggered requirement for this study has already been provided to registrants.
As with the subchronic study, OPP has concluded that the absence of a
triggered acute neurotoxicity study from the database does prompt concerns
relating to the "completeness of the toxicity database" in regard to the safety of
infants and children and will be taken into consideration in FQPA Safety Factor
decisions.
Issues raised by Commentor 772, regarding the interpretation of the
retrospective analysis of DNT studies submitted to OPPTS, are addressed in
depth in the OPP response to comments in Comment 9.2. The argument that
the DNT studies examined did not provide any information that was critical in the
regulation of the specific chemicals examined is not, in fact, correct.
Additionally, A Retrospective Analysis of Twelve Developmental Neurotoxicity
Studies Submitted to the USEPA Office of Prevention, Pesticides, and Toxic
Substances (OPPTS) (USEPA 1998c) provides extensive discussion
regarding the use of the DNT study in risk assessment, and affirming the value
of the study for this use.
Comment 5. Commentor 369 supported the general approach of applying an
additional safety factor in the absence of newly required studies, provided that a
specific concern for the endpoint addressed by the study in question has been
identified. This step would then apply regardless of whether a newly required study
is missing or whether changes have been made to an existing guideline. The factor
applied could be either the database uncertainty factor and/or an FQPA Safety
Factor for residual concerns about the adequacy of the database. Commentor 773
also indicated that this approach may be appropriate.
84
-------
As a related issue, Commentor 776 recommended that the Agency revisit the
organophosphate (OP) pesticides, since the 10X factor was not added to them
even though the developmental neurotoxicity studies were required.
Response. The approach described by Commentor 369 is, in general, that
which OPP uses to determine how the absence of new data requirements will
be considered in the FQPA Safety Factor evaluation. FQPA decisions that
result from this approach are informed by a weight-of-evidence assessment of
scientific data and an evaluation of residual uncertainties.
OPP agrees that the one particularly critical consideration for a safety factor
determination is what existing evidence tells us about the pesticide. Thus, when
additional data are needed on a pesticide, OPP examines whether that data
requirement is being imposed because of a general desire to upgrade
pesticide testing or because of some specific concern regarding the pesticide
that was identified in the testing that had already been completed. General
increases in the information needed to support pesticide registration or
reregistration (which could be comprised either of a generic increase in
required studies or of updates/revisions to existing testing guidelines) could be
initiated by an evolving understanding of the pharmacological or toxicological
sciences, or by the realization that the information provided is critical to an
adequate understanding of potential hazard. Specific concerns that could elicit
the request for additional information might include, for example, the observation
of an effect which is not fully characterized, or the observation of a pattern of
effects across the toxicological database that suggests that evaluation of
previously unexamined endpoints is necessary to fully characterize hazard and
dose-response. Such specific concerns could be identified for an individual
chemical, for a chemical class, or for multiple chemicals with related modes of
action.
Where the specific concern suggests that risk to infants and children may be
underestimated, the lack of the study would weigh heavily in favor of retaining
some or all of the FQPA factor. In such circumstances, there is a greater
chance that the needed data may identify a lower regulatory dose level than
observed in other studies in the chemical database, or demonstrate that young
animals are more susceptible than adults to effects of the pesticide, leading to a
more health protective approach to risk assessment (as opposed to
demonstrating that the additional factor was not necessary). Where additional
data requirements are being imposed more generically, it may be less likely that
there is a specific concern regarding the underestimation of risk to children, and
if so, there would be less need for the retention of an additional safety factor.
Comment 6. One commentor (773) stated that clarification is needed regarding
85
-------
whether the lack of a developmental neurotoxicity study would be the basis of the
decision to apply the FQPA Safety Factor.
Response. The presence or absence of a developmental neurotoxicity study
will be part of the weight-of-evidence determination for another component
(Residual Concerns) of the FQPA Safety Factor. A more extensive discussion
of this issue is presented in the Response to Comments 4 and 5 above.
ISSUE 8. Criteria Indicative of Increased Concern for Pre- and
Postnatal Hazards
In the absence of the results from any of the studies to be required through Data Call-
in Notices (i.e., the acute and subchronic neurotoxicity studies in adult mammals
and the developmental neurotoxicity study),
what information from existing studies on a specific chemical would increase or
decrease the concerns about the potential for prenatal and postnatal hazard, in
general,
and for neurotoxicity and developmental neurotoxicity, in particular?
which, if any, of the seven criteria discussed in Section V.A. 1.a., footnote 4 and
associated text of the OPP Guidance Document is appropriate for judging
whether there is increased concern about the potential for a pesticide to cause
developmental neurotoxicity?
* are there any other criteria which would be useful for informing this judgment?
Overview:
The draft policy contained a list of seven criteria that OPP proposed to apply on a
case-by-case basis to evaluate whether the database on a pesticide suggests reason
for concern that the chemical may cause developmental neurotoxicity. The criteria
were that the substance has been shown to:
cause central nervous system (CNS) malformations following prenatal exposure;
affect brain weight in offspring, which does not appear to be related to general
86
-------
growth retardation, following pre- and/or postnatal exposure; cause
neuropathology in developing or adult animals or neuropathy in humans;
cause persistent functional changes in the offspring which may be the result of
effects on the nervous system;
act to significantly alter hormonal responses associated with the development of
the nervous system, leading to significant developmental effects (e.g., effects on
sexual maturation);
act as a neurotoxicant in insects, unless other information about the chemical,
such as pharmacokinetic and pharmacodynamic data, demonstrate the
inappropriateness of such testing; or
cause evidence of adverse effects in tests of cognition, memory, and other
higher brain functions.
Comment 1. Most of the comments received on this issue focused on
characterizing an increased or decreased concern about potential neurotoxicity.
Those commentors (369, 773, 776, SAP) who offered their perspective on
neurotoxicity specifically expressed agreement, in principle, with the criteria that
OPP proposed as indicating an increased concern for neurotoxicity and
developmental neurotoxicity, and also to be used in "triggering" the requirement for
the developmental neurotoxicity study.
Two commentors (773, 776) provided suggestions for additional information
that could be considered that would lead to an increased as well as decreased
concern for developmental neurotoxicity. Commentor 773 reaffirmed the principle
that if existing data show neurotoxic effects, that would increase the concern for
other manifestations of neurotoxicity and/or developmental neurotoxicity. Evidence
of nondevelopmental effects known to have developmental consequences, such as
effects on thyroid function would also increase concern. On the other hand, this
commentor argued that the absence of data showing these effects would lessen the
concern for developmental neurotoxicity. Commentor 776 also noted that changes
in thyroid status, particularly hypothyroidism during pregnancy and its potential for
adverse impact on neurodevelopment of the fetus, would provide an indicator for
increased concern. They also suggested that noting brain weight as it relates to
body weight, generally, in studies would be helpful.
Commentors 369 and 773 agreed with criteria 1-5 and 7. Both commentors
questioned the specificity of criterion #6, relating to neurotoxicity in insects.
Commentor 773 asserted that this criterion was "inappropriately overbroad." In
87
-------
addition, the commentor suggested that a weight-of-evidence approach should
consider all available information and not rely on a single criterion to make
judgments about potential developmental neurotoxicity or level of concern.
Commentor 369 recommended that criterion #6 be changed to include evidence of
neurotoxicity in any species rather than simply insects. This commentor also
agreed with criterion 7 in principle, but expressed doubt about whether it was likely
to be useful given the paucity of validated test guidelines that address cognition,
memory and higher brain function.
FIFRA Scientific Advisory Panel Comments
The FIFRA SAP agreed with the seven criteria, but stated that clarification was
needed on how the criteria would be weighted for decision-making. The FIFRA
SAP also provided ten additional criteria that might be useful. The ten additional
criteria include that the substance has exhibited:
inhibition of cell division (e.g., colchicine);
specific toxicity/lethality for dividing cells (e.g., ionizing radiation);
changes in neuronal migration (e.g., methyl mercury);
neuroreceptor/neurotransmitter agonism or antagonism;
molecular resemblance or parent compounds or predictable metabolites to
known neurotoxins (e.g., gamma diketones such as 2,5-hexanedione,
certain nitriles/cyanide compounds, some metals and organometallic
compounds such as alkylmercury, lead, manganese, cholinesterase
inhibitors);
high lipophilicity conducive to concentration in lipid bilayers important for
neural functioning (e.g., PCBs);
identification of decreased biological factors in the adult that could present a
problem in the developing organism (e.g., decreased cholesterol with
carbon disulfide could be significant for the developing nervous system due
to its high demand for cholesterol);
88
-------
mode of action on the target species and its relationship to the human
system, whether directly or via an associated mechanism or human
homologue;
mutagenicity, clastogenicity, or carcinogenic responses may increase
concerns as well because of the implications that these effects have for low
dose extrapolation; and,
clear positive results from the two-generation reproduction studies and
prenatal developmental studies in the absence of maternal toxicity would
increase concern about pre- and postnatal hazards.
Response. As noted above, OPP is appreciative of the general support
expressed for the proposed criteria, and would like to take this opportunity to
clarify that the weight-of-evidence approach does include consideration of all
available information. While some might wish for more specific articulation in
how a weight-of-evidence approach might be applied, it clearly is case-specific.
It would be difficult, and, probably, not particularly useful to try to apply it in the
abstract. Examples of application of this decision logic can be found in OPP's
Hazard Identification Assessment Review Committee reports on individual
chemicals.
The first five criteria had been reviewed previously by the FIFRA SAP
(USEPA 1998d and USEPA 1998e) as indicators of developmental
neurotoxicity potential that would trigger a DNT study. Based on this review and
subsequent discussions, OPP proposed two additional criteria (6 and 7).
Criterion 6 ("acts as a neurotoxicant in insects") had been recommended for
inclusion by an earlier SAP and was proposed due to concerns for pesticides
that had been designed to target the nervous system, albeit in insects. This
criterion was informed by the knowledge that many of the traditional classes of
insecticides functioned via a neurotoxic mechanism that also was common to
mammalian species, including the human. OPP agrees that the criterion "acts
as a neurotoxicant in insects" is too narrow and has broadened this criterion to
include neurotoxicants in any species, with an obvious focus on mammalian
species (see Table 3 in the revised guidance document). With reference to
criterion 7 ("cause evidence of adverse effects in tests of cognition, memory,
and other higher brain functions"), OPP agrees that this criterion would not often
be used, at least in the near term, given the general paucity of data evaluating
these endpoints. However, OPP would like to point out that test guidelines do
exist that assess cognition and memory, and, when available, these data may
prove very useful in assessing potential concerns for developmental
neurotoxicity.
89
-------
OPP's analysis of the SAP's proposed additional criteria, which are now
proposed for use in determining a level of concern for developmental
neurotoxicity potential, reveals that several are not so much new criteria but
rather more detailed descriptions of types of information and specific modes of
action that can be combined within the context of OPP's proposed seven
criteria to provide more clarity. Some of them appear to be more suitable for
addressing pre- and postnatal hazard concerns, in general (SAP's #1, 2, 7, 8, 9,
10). In the revised guidance document, OPP has modified the list of criteria to
take the SAP's recommendations into account. The revised set, which
addresses concerns for developmental neurotoxicity specifically, not
developmental toxicity generally, includes that the substance (and/or a
metabolite/degradation product) demonstrates a potential to:
Cause treatment-related neurological effects in adult animal studies,
such as:
Clinical signs of neurotoxicity
Neuropathology
Functional or behavioral effects
Cause treatment-related neurological effects in developing animals,
following pre- and/or postnatal exposure, such as:
Nervous system malformations or neuropathology
Brain weight effects in offspring
Functional or behavioral changes in the offspring
Elicit a causative association between exposure and adverse
neurological effects in humans in epidemiological studies
Evoke a mechanism that is associated with adverse effects on the
development of the nervous system, such as:
Structure-activity relationship to known neurotoxicants
Altered neuroreceptor or neurotransmitter responses
Altered hormonal responses
OPP recognizes that there has been an apparent inconsistency in the past
with the practical application of the procedures when the developmental
neurotoxicity study was required for conventional food-use pesticides with
neurotoxic mechanisms (as noted by Commentor 776). The 1999 draft policy
stated that, in response to well-supported scientific recommendations, two
90
-------
criteria would be added to a previously-defined list of five criteria that are used,
either individually or as part of a weight-of-evidence assessment, establish
concern for developmental neurotoxic potential. When a specific cause for
concern is identified and the developmental neurotoxicity study has not yet been
submitted to the Agency, the scientific uncertainty regarding the risk
assessment for infants and children and the possibility that the findings from the
(absent) DNT study might result in a lower regulatory value (i.e., RfD or PAD),
would initiate retention of some or all of the FQPA Safety Factor. Although the
five previous triggering criteria were used in this manner, OPP delayed utilizing
the two new criteria in the same way, while awaiting public input on the issue.
OPP has not, by virtue of this delay, intended to suggest that these two
considerations were less critical or concern-worthy aspects of the weight-of-
evidence evaluation. Rather, OPP has both explicitly and implicitly recognized
the importance of the DNT study in hazard characterization, and has (as
described in the revised guidance document) begun the process of requiring the
DNT studies for neurotoxic pesticides via the DCI authority under FIFRA and the
FFDCA. The first DCI, requiring submission of the developmental neurotoxicity
study and the supporting adult acute and subchronic neurotoxicity studies for the
cholinesterase-inhibiting organophosphorus chemicals, was issued by OPP on
September 10, 1999. Data Call-Ins for these same three studies on other
classes of neurotoxic pesticides will be issued in the future.
A weight-of-evidence decision regarding the application of the database
uncertainty factor to account for the absence of any study including the absence
of a DNT should be used for consideration of the traditional database
uncertainty factor. In the case of a missing DNT, a number of factors (as
described in the revised policy) should be taken into account when making
judgements about the need for a traditional database uncertainty factor in the
absence of a DNT study for a given pesticide. The decision regarding the need
for a database uncertainty factor to address the absence of a DNT (or other
types of developmental toxicity studies) should be based on weighing all lines of
evidence for the chemical of interest, and combining the entire body of evidence
to make an informed judgment on the need for, and size of, the factor.
Judgement about the weight-of-evidence involves for example: considerations
of the quality and adequacy of available data; consistency of responses; the
multiplicity of observations in independent studies; and the severity, potency,
persistence and latency of effects induced by the agent in question.
Additional information bearing on the degree of concern about a pesticide's
potential for DNT may also be gained from: comparative pharmacokinetic and
metabolism studies; structure-activity relationship (SAR) analysis; and other
studies of an agent's physical and chemical properties. These considerations
or factors should not be scored mechanically by adding pluses and minuses;
91
-------
rather, they should be judged in combination. Simply because OPP has
required a DNT for a particular pesticide does not necessarily mean that a
database uncertainty factor is needed. However, if the available information
indicates that a DNT study is likely to identify a new hazard or effects at lower
dose levels of the pesticide that could significantly change the outcome of its
overall risk assessment, the database uncertainty factor should be considered.
Comment. One commentor (776) suggested that information on exposure patterns
may be useful to increase concern. They suggested that special attention should be
given to particular use patterns such as topical (skin) application and residential
uses or residues in drinking water-any situation where one might expect children
and pregnant women could be exposed. In contrast, another commentor (773)
suggested that exposure information may decrease concerns, that is, for those use
patterns that would make exposure of children unlikely.
Response. In both the draft and revised guidance documents, OPP states that
exposure information is taken into account in making decisions about the FQPA
Safety Factor. While the potential for exposure to children and pregnant women
is a consideration in the FQPA Safety Factor finding, Issue 8 relates to a
section of the guidance document that addresses issues only of hazard
assessment. While exposure is not a consideration at this point in the process,
OPP does take use information and activity patterns for children, among other
factors, into account when performing exposure assessments, which then are
combined with hazard assessments to produce risk assessments.
ISSUE 9. Expectations about the Results of DNT Studies
Please comment on whether you would expect that developmental neurotoxicity
studies would, for a substantial number of chemicals, identify effects that are not
detected in other studies and more fully characterize the potential risks of exposures
during development. In addition, please comment on the sensitivity of these tests
vis-a-vis other studies required and used for age-related comparisons for acute,
intermediate, or chronic RfD derivation (e.g., prenatal developmental toxicity or
multi-generation reproduction study, subchronic and chronic studies, etc.). Please
explain the basis of your opinion.
Comment 1. Commentors 773 and 776 agreed that the DNT study would identify
effects that are not detected in other studies, particularly since the study evaluates
endpoints (e.g., functional endpoints) that are not measured in other tests.
Commentor 773, however, expressed the opinion that this fact does not justify
92
-------
making the DNT study a routine Tier 1 requirement.
Response. OPP agrees that the DNT study will likely detect effects not
observed in other studies. A Retrospective Analysis of Twelve Developmental
Neurotoxicity Studies Submitted to the USEPA Office of Prevention,
Pesticides, and Toxic Substances (OPPTS) (USEPA 1998c) demonstrated
this to be the case, even within the small number of studies examined. OPP
plans to publish proposed revisions to its pesticide data requirements
regulation, 40 CFR 158, and expects to ask for comment on a requirement for
developmental neurotoxicity testing, which utilizes information about each
chemical and its toxicity to develop a rational, science-based approach to the
study design and testing strategy.
FIFRA Scientific Advisory Panel Comments
Comment 2. Commentor 369 stated that it is not yet possible to predict how the
results of these developmental neurotoxicity studies may impact the final risk
assessments. The impact will depend entirely on the nature and dose response of
the effects observed.
The SAP stated that consensus among the Panel members could not be
developed regarding whether NOAELs from developmental neurotoxicity studies
would be lower than from historically required studies. One member argued
strongly that the NOAELs or more appropriate benchmark doses identified by the
developmental neurotoxicity studies will be lower than those detected by the present
tests for a substantial number of pesticides. This member's prediction was based
on the fact that the effects of many teratogens (e.g., psychoactive compounds,
antiseizure medications, anticarcinogens, metals, radiation, retinoids, folate levels,
etc.) are already known to be detected at lower doses with these tests than with the
ones presently required. Another member did not accept the notion that the number
of chemicals with effects occurring at lower doses would be large, but agreed that
those identified would represent an important group. In addition, the analyses
already presented by the Agency (USEPA 1998c) indicate that NOAELs identified
by the use of developmental neurotoxicity testing are often not likely to be lower than
those characterized by prior testing methods. The Panel is aware that only one of
12 chemicals showed developmental neurotoxicity effects at lower doses than were
observed with the prior standard testing protocol. However, the Panel expressed
caution that the results from testing the 12 pesticides could not be applied to a
broader set of pesticides.
Response. OPP agrees with Commentor 369 in stating that it is not yet
93
-------
possible to predict exactly how the results of these developmental neurotoxicity
studies may impact any particular risk assessments.
The SAP appears to have misunderstood the concerns of OPP regarding
the analysis of the DNT studies, citing the fact that only one of the 12 pesticides
examined showed developmental neurotoxic effects at a lower dose than the
dose level at which effects were observed with current standard testing. The
OPP position on this issue is that even though only nine DNT studies on
pesticides were available for review and analysis (out of a total of 12 studies),
one of these nine DNT studies demonstrated developmental neurotoxic effects
at a lower dose than were observed with current standard testing for chronic
effects which may reduce the confidence that the traditional testing paradigm is
generally adequate for the assessment of potential toxicity to infants and
children.
It is inappropriate to assume that there would be many cases in which the
NOAEL from a DNT would be lower than one from a chronic study. After all, the
DNT reflects exposures that are not chronic in nature, but subchronic, at best. A
more appropriate comparison might be made with NOAELs from studies of
subchronic exposures such as a 28- or 90-day repeated dose study or the two-
generation reproduction study in which the offspring of the first generation are
exposed for a total of about 120 days.
Comment 3. One Panel member questioned the wisdom of moving tests of central
nervous system function into Tier 1, with no plans for testing the functions of other
organ systems. Another question regarding the proposed battery was whether it is
intended as a screen or as research. It was stressed that the results of the
developmental toxicity study must be usable for risk assessment.
Several members supported the idea that the Agency needs to improve and
refine the proposed neurotoxicity battery. Because new factors in development are
being discovered at a rapid rate, the Agency needs to be flexible, and the pace of
development, validation, acceptance, and implementation of new protocols needs
to be increased.
94
-------
Response. OPP does not believe that there is any justification to delay the
inclusion of the tests of adult central nervous system function in Tier 1, based
upon the lack of availability of testing guidelines that address specific functional
aspects of every other organ system. Such an approach would be neither
logical nor scientifically supportable. It is additionally noted by OPP that
guidelines already exist for at least some aspects of structure and function for
developing organ systems thought to be critical for risk assessment purposes
and most of these studies are required already in Tier 1 (per40CFR 158) for
the registration of conventional food-use pesticides. For at least one other
system, i.e., the reproductive system, an extensive evaluation of the effects on
function, including those that may arise from developmental exposure is routinely
conducted.
The use of the developmental neurotoxicity study to select endpoints for use
in risk assessment in OPP was addressed in some detail within the
retrospective analysis of twelve DNT studies cited above; this document was
reviewed by a FIFRA Scientific Advisory Panel in December, 1999 (USEPA
1998f). On a broader Agency-wide level, the use of the results of the DNT study
in risk assessment is also addressed in the Guidelines for Developmental
Toxicity Risk Assessment (USEPA 1991) and the Guidelines for Neurotoxicity
Risk Assessment (USEPA 1998a).
95
-------
ISSUE 10. Policy for Database Uncertainty Factors
OPP's guidance states that currently five studies (a multigeneration reproduction
study, prenatal developmental toxicity studies in two species, and chronic toxicity
studies in a rodent and non-rodent species) comprise the toxicity database
necessary to produce a high confidence RfD, and that some additional database
uncertainty factor will be imposed if the database on a pesticide lacks one or more of
these studies. OPP proposes to expand this core database to include the
subchronic neurotoxicity study. Eventually, OPP also includes the acute
neurotoxicity study in adult mammals and the developmental neurotoxicity study,
once these studies have met the criteria for inclusion in the core toxicity database.
Please comment on OPP's proposed approach to imposing a database uncertainty
factor of 3X if one key study is missing from the database and a factor of 10X if more
than one is missing.
Comment 1. One commentor (372) expressed the need for further clarification on
whether the adequacy of the database will be the primary consideration in selecting
the numerical value of the FQPA Safety Factor, and suggested that OPP include a
single section in the document that clearly outlines the issues that will be considered
in the decision-making.
Response. While the guidance document discusses the use of a database
uncertainty factor in cases where the toxicity database may not be adequate for
the derivation of high confidence reference values, the adequacy of the
toxicology database is NOT the sole focus for consideration in selecting the
numerical value of the FQPA Safety Factor. It is but one of three areas of
consideration, each given weight as appropriate to the pesticide under review.
As the guidance document states in a number of places, the FQPA Safety
Factor determination is premised on the application of a weight-of-evidence
approach which considers not just the adequacy of the data as it relates to
toxicity, but also the nature of the toxicity data for informing a judgment with
regard to the determination of the potential for pre- and postnatal and to the
adequacy of the database as it relates to the assessment of exposure.
Discussion of the weight-of-evidence approach can be found in the guidance
document chapters II, IV, and VI.
Comment 2. One commentor (773) agreed with OPP's proposal to use a 3X if
one study was missing and a 10X if more than one were missing. Another
commentor (372) agreed in principle, but suggested clarification as to why these
numerical values were chosen. Another commentor (369) agreed in principle, but
thought the choice was too prescriptive; a suggestion was made to provide more
96
-------
general guidance on how the database would be included in a more general
weight-of-evidence approach. Several commentors (761, 771) stated that the
factor should be at least 10X since OPP does not have "reliable" data to do
otherwise. One commentor (778) stated that the question is whether OPP has
complete and reliable data to assure no harm to children, not whether a sufficient
number of studies exist.
Response. OPP's proposed approach to imposing a database uncertainty
factor of 3X if one key study is missing from the database and a factor of 10X if
more than one is missing was based on EPA's longstanding RfD process which
identifies five uncertainty factors and one modifying factor that may be applied
to the NOAEL or BMD to derive an RfD (US EPA 1994). The maximum default
value for each of these factors is 10, although sometimes a different factor (most
often 3X) is used depending on the nature and quality of the information
available on the pesticide. Upon further reflection with an additional year of
experience in making FQPA Safety Factor determinations, OPP agrees with
Commentor 369 that the approach of applying a 3X if one study is missing and
a 10X if more than one study is missing is too prescriptive and is inappropriate
for a guidance document. OPP has revised the guidance document and will use
a weight-of-evidence approach in decisions regarding the size of the database
uncertainty factor, when it is deemed appropriate to impose one. OPP will base
the decision regarding the size of the database uncertainty factor on which
studies are missing in addition to how many studies are missing (see chapter
III).
With respect to the question of reliable data, it is OPP's position that the
reliable data requirement in the infants and children's provision does not
mandate that any specific kind of data be available, just that the data and
information that form the basis for the selection of a different safety factor must
be sufficiently sound such that OPP could routinely rely on such information in
taking regulatory action (also see response to comments under Issue 17).
Comment 3. One commentor (775) expressed concern over the use of the term
"high confidence RfD" and the potential implications for public perception. The
commentors stated that the term "high" or "low" confidence should be restricted to
the database as opposed to the RfD.
Response. OPP would like to clarify that the terms "high confidence RfD" and
"low confidence RfD" are terms that have been used by the Agency for nearly
two decades. A detailed description of the Agency RfD derivation process and
guiding principles can be found on the Agency IRIS website (see
http://www.epa.aov/iriswebp/iris/rfd.htm). Designation of high, medium, or low
97
-------
confidence as it applies to an RfD (or RfC) is based on a number of factors
including the adequacy of the core database, consistency of response, and
route of exposure. In the guidance document for the RfD Process, it indicates
that the EPA is attempting to standardize its approach to determining RfDs.
The RfD Work Group has developed a systematic approach to summarizing its
evaluations, conclusions, and reservations regarding RfDs in a 'cover sheet' of
a few pages in length. The cover sheet includes a statement on the confidence
(high, medium, or low) the evaluators have in the RfD. High confidence
indicates the judgment that the RfD is unlikely to change in the future because
there is consistency among the toxic responses observed in different sexes,
species, study designs, or in dose-response relationships, or that the reasons
for existing differences are well understood. High confidence is often given to
RfDs that are based on human data for the exposure route of concern, since in
such cases the problems of interspecies extrapolation have been avoided. Low
confidence indicates the judgment that the data supporting the RfD may be of
limited quality and/or quantity and that additional information could result in a
change in the RfD.
ISSUE 11/12. Applying the FQPA Safety Factor Clause Relating
to Potential Pre- and Postnatal Toxicity
OPP has decided to combine the comments and responses to the following two
questions in the same section because they both relate to the interpretation and
application of the clause in the FQPA Safety Factor provision directing EPA to take
into account "the potential pre- and postnatal toxicity of" a pesticide in deciding whether
infants and children are adequately protected. As summarized below and then
discussed in more detail in the Agency's responses to comments, OPP thinks that the
decisions about retaining the default 1OX factor and the need for, and size of, different
FQPA Safety Factor should reflect both the level of concern for pre- and postnatal
effects and the level of certainty about the hazard relative to other toxic effects.
98
-------
Issue 11. Criteria for Assessing Concerns about Potential Pre-
and Postnatal Effects
OPP is proposing to adopt the framework and its criteria/factors for assessing the
degree of concern about the potential for prenatal and postnatal effects as
recommended by the Toxicology Working Group. Please comment on the
appropriateness of the proposed criteria/factors for use in this assessment process,
and OPP's proposed approach for accommodating its concerns in the Reference
Dose derivation and FQPA Safety Factor decision processes, in the near term, and
in the longer term.
Issue 12. Interplay of Certainty about Hazard and Increased
Sensitivity of Infants or Children
When the hazard to infants and children is well-characterized, and the data show that
infants and/or children are more susceptible than adults, under what circumstances,
if any, should this information lead OPP to employ an additional safety factor?
Summary of OPP Approach to the Weighing of Potential Pre- and Postnatal
Toxicity
Concern for Potential Pre- and Postnatal Toxicity to Infants and Children. Because
the statute directs EPA, in making decisions about the FQPA Safety Factor, to
consider the "potential pre- and postnatal toxicity [of a pesticide] to infants and
children," the guidance document sets forth factors that would increase or decrease the
Agency's concern that a pesticide has the potential to cause toxicity to infants or
children that has not been adequately evaluated. Consistent with the basic approach of
the overall guidance, OPP will make this decision on a case-by-case basis, taking all
available, relevant information on toxicity and exposure into account to evaluate
whether the young are more sensitive or susceptible to the toxic effects of pesticides
than adults. OPP will evaluate whether the standard approach for determining an RfD
or RfC-using available data on toxicity and exposure and applying the traditional
uncertainty and modifying factors-provides assurance that infants and children will be
adequately protected, or alternatively whether the default FQPA Safety Factor should
be retained or a different FQPA safety factor adopted. As discussed in more detail
below, risk assessors should apply the following analysis which draws on Agency-wide
policy, public comments, and insights gained through several years' experience
applying the FQPA Safety Factor provision.
When the available database permits, risk assessors should determine the
99
-------
Reference dose(s) or RfD(s) for a pesticide by selecting an appropriate toxicity study
and applying traditional uncertainty or modifying factors to the highest dose level in that
study at which no adverse effects were observed (the NOAEL). As cited and explained
in the revised guidance document, the risk assessor should choose the appropriate
toxicity study taking into account established EPA policies and practices on the
determination of a RfD.
Since the determination of an RfD involves consideration of a number of issues that
OPP associates with increased concern for toxicity of a pesticide to infants and
children, the subsequent FQPA safety factor analysis relating to these issues should
focus on whether those issues have been adequately addressed in the RfD derivation
process. Specifically, risk assessors should evaluate whether the RfD process
adequately took into account any data indicating that young animals are more sensitive
to the toxic effects of a pesticide than mature animals or are susceptible to a different
array of toxic effects than mature animals. If younger animals appear to display
increased sensitivity or susceptibility, the risk assessor should determine the degree of
concern for potential toxicity, using the factors presented in the guidance document.
The risk assessor should not recommend retention of the default FQPA safety factor or
application of a different FQPA safety factor on a routine basis solely because some
kind of pre- or postnatal toxicity is observed or there is an apparent difference in
sensitivity.
After having determined the RfD(s) for a pesticide and the degree of concern for
observed pre-/postnatal toxicity (increased sensitivity or susceptibility) the risk
assessor should identify any residual concerns relating to the potential toxicity of the
pesticide to infants and children which will inform the final decision to retain the default
FQPA Safety Factor or adopt a different FQPA safety factor.
Comment 1. One commentor (369) questioned how the degree of concern for
children's health risks would be integrated into the selection of uncertainty factors in
establishing RfDs and PADs. Three commentors (773, L016, L005) suggested
that the appropriate consideration for degree of concern should be whether infants
and children are more than 10 times more sensitive than adults. One commentor
(776) questioned whether a high degree of concern for one of the criteria would be
enough to judge concern for pre- and postnatal toxicity. One commentor (773)
indicated that all criteria that relate to factors considered in the current uncertainty
factors should be eliminated, and asserted that the proposed criteria mandate
"double-counting" by focusing on factors that are already considered in the
traditional uncertainty factors.
100
-------
Response. The criteria for degree of concern for potential pre- and postnatal
toxicity in humans were developed by the Toxicology Working Group simply to
provide more detail on the issues that should be considered when evaluating the
available data to determine the overall degree of concern for such effects. Most
of these criteria already are considered in the development of the RfD and RfC,
but are not specifically detailed in the Agency's risk assessment guidelines for
developmental toxicity (or, any other endpoint of toxicity, for that matter).
Therefore, historically, they may not have been considered explicitly when a
hazard characterization is being developed for an environmental agent. As part
of a weight-of-evidence approach to evaluating the potential of pesticides to
produce pre- and postnatal toxicity, OPP will consider the information on all of
these issues before making a determination about the degree of concern for
potential pre- and postnatal toxicity in humans. If data are not available, for
example, on toxicokinetics, then assumptions must be made about whether, and
how, data indicating hazards for developmental effects in the species tested are
relevant to humans. Although OPP believes that the intraspecies (that is, human
to human) uncertainty factor is protective for any age-related differences in
sensitivity in most cases, no single criterion drives the decision about relative
sensitivity of the young versus adults; rather it comes from a consideration of all
of the criteria in a weight-of-evidence approach. This latter step-employment of
the modifying factor for this purpose-is not currently a component of the
Agency's "formal" procedure for deriving RfDs/RfCs. How to formalize this step
is being further addressed by EPA's RfD Technical Panel. Since there is no
formal Agency guidance, OPP has chosen to account for a high degree of
concern and other uncertainties not addressed by the inter- and interspecies
uncertainty factors by applying an FQPA Safety Factor greater than 1X. The
size of this component of the overall FQPA Safety Factor should take into
account any other factors that are applied in the RfD/RfC process, thus avoiding
double-counting. It is determined on a case-by-case basis, driven by existing
information (or, lack thereof). The FQPA Safety Factor and some of the
uncertainty factors applied in the derivation of the RfD/RfC are not mutually
exclusive, but are complementary in their attempts to account for all
uncertainties and the degree of concern.
101
-------
Comment 2. One commentor (776) questioned the criteria, particularly giving
effects of longer latency a low degree of concern, since the effects of developmental
toxicants may not manifest themselves until systems are fully mature and functional.
Response. The issue of latency of time to manifestation of effect was raised in
the criteria for degree of concern only in the case where organ systems have
matured to the point of being functional and similar types of effects are likely to
be seen in adults and the young, although to a different extent. In this case, if the
effects in adults are seen at a lower dose and occur sooner (shorter latency)
than in the young, this would constitute a lower degree of concern for children,
because this pattern of response would indicate lesser sensitivity in young than
adult animals. This has nothing to do with the long latency effects that may occur
as a result of early life exposure to a developing organ system that has not
matured and that do not occur as a result of adult exposure, e.g., effects that
may be seen in adulthood or old age that are the result of early nervous system
or other organ system impairment during development.
Comment 3. Three com mentors (369, 372, 773) stated the need for clarification
on why both a steep slope and a shallow slope would raise the degree of concern.
Response. OPP's initial policy considered both a steep slope and a shallow
slope to be of concern, but for different reasons. The discussion of steep and
shallow slope has been removed from the revised document. The discussion
focuses on the certainty surrounding the identification of the point of departure
(usually a NOAEL) that is used to derive the RfD. For example, when the dose-
response relationship is well-characterized, i.e., the NOAELs or BMD are
defined, there is a lower degree of concern than when the definition of the
NOAEL or BMD is poor; in the latter case, the degree of concern may increase.
There is also discussion on the degree of concern with respect to where in
the dose-response curve effects are seen. For example, the degree of concern
could decrease when developmental or adverse effects are seen only at higher
doses (e.g., approaching or greater than the maximum tolerated dose), or
observed only in the presence of severe or generalized (nonspecific) toxicity.
On the other hand, if developmental effects are seen at several doses including
those at lower doses than for adult toxicity the degree of concern could increase.
Comment 4. Two commentors (773, 372) questioned the statement that the
concern is higher if the increment between the NOAEL and LOAEL is large,
indicating that if anything, the larger the increment, the more conservative the
NOAEL, providing a greater level of assurance that the RfD is protective.
102
-------
Response. See response to Comment 3.
Comment 5. Two commentors (369 and 773) indicated that the criteria on pre-
and postnatal toxicity should include the element of relative severity of effect in
young versus adult animals, particularly for similar types of effects, as well as the
permanence and life-threatening nature of the effects. One commentor (773) also
suggested that the relevance of relative young to adult sensitivity data collected at
high dose levels for predicting potential differences at low doses should be included
in the criteria.
Response. OPP agrees that relative severity in young versus mature animals
where the effects are of a similar type should be included as a criterion.
Severity would also include consideration of permanence and life-threatening
nature, but these are not requirements for whether or not the effects should be
considered adverse. The issue of high to low dose relevance would be part of
the consideration of relevance of the experimental animal data to humans,
particularly as related to toxicokinetic and mechanism of action considerations.
Comment 6. One commentor (773) indicated that OPP should recognize that
human data relating to reproductive toxicity will rarely, if ever, be available and,
therefore, the lower degree of concern should focus on whether available
information supports a conclusion that "no effects occur at the level of potential
human exposure."
Response. OPP agrees with the assertion that there is a paucity of human
studies which investigate the potential for effects on reproduction. Nonetheless,
OPP does not agree with the statement that there should be a lower degree of
concern if the available information supports a conclusion that "no effects occur
at the level of potential human exposure." Absence of evidence does not
constitute evidence of absence of an effect. Proving no effect in humans is very
difficult, and even data showing no effects from limited human studies are
usually not adequate to indicate an absence of human effects, unless a wide
range of reproductive and developmental outcomes has been evaluated.
Comment 7. One commentor (773) asked for clarification of the difference
between the criteria for "relative potency of response" and "pre- and postnatal
toxicity in animal studies." There was also a suggestion that the criteria for "dose-
response nature of the experimental animal data" should be limited to data on
reproductive and developmental toxicity.
103
-------
Response. The criteria for pre- and postnatal toxicity in animal studies are part
of the hazard identification/characterization process and allow a qualitative
evaluation, whereas relative potency of response is more quantitative in nature,
having to do with the incidence and intensity of response over the dose-
response range, as well as the comparative doses in young and adult animals at
which effects are seen.
OPP disagrees that the data for dose-response evaluation should be limited
to reproductive and developmental toxicity data. Data from all available toxicity
studies, whether conducted with neonatal/juvenile or adult animals, should be
considered for several reasons. For example, adult data are important in
identifying potential target organs that may also be affected when exposures
occur in children whose major organ systems have already formed but are
functionally less mature than in adults. Because children include adolescents up
to 18-21 years of age, adult data also will provide important information about
potential target organs during this time period as well. Adult data also may
provide information on target organs to evaluate in the reproduction studies or
other developmental studies for similar target organ effects, e.g., developmental
immunotoxicity, developmental carcinogenesis, or endocrine toxicity studies. In
addition, adult data provide relative potency information when comparing or
estimating differences between children and adults.
FIFRA Scientific Advisory Panel Comments
Comment 8. The SAP commented that there needs to be some relatively well-
considered process for establishing factors for different outcomes in a severity-of-
effect determination.
Response. OPP has already incorporated limited consideration of the severity
of effects into the factors forjudging the level of concern for potential pre- and
postnatal toxicity. In general, "severe" effects are those which are enduring and
either life-threatening or involve serious impairment of an organism's ability to
function effectively. Less severe effects lack these characteristics. The
presence of severe effects contributes to an overall higher level of concern.
OPP, however, is reluctant to undertake any more specification of the weight
to be given to the type of toxic effect observed in animal studies for a
combination of reasons. First, while there is often similarity in response
between animals and humans, there are sufficient instances in which the
responses differ qualitatively, as well as quantitatively, that it may not be
sufficiently protective of public health to systematically give less weight to less
severe effects. Second, there is considerable lack of societal acceptance of the
104
-------
notion that humans should be assumed to tolerate exposures which have
caused adverse effects in animals, even if the effects are not particularly severe.
Finally, the range of toxic effects caused by chemical exposure is quite large,
and as a practical matter, it would be nearly impossible to gain general
agreement on how to assign different effects to a point along the continuum from
mild to severe.
Comment 9. The SAP endorsed the criteria developed for degree of concern, in
general, but suggested additional criteria for toxicokinetics and mechanism of
action, indicating that the lack of adequate data would add some degree of
uncertainty and should fall between the higher and lower extremes. They also
indicated that in cases where there are clear toxicokinetic differences between
humans and the experimental animal, the agent may not have been adequately
tested in a relevant species and may be indicative of important missing information.
Response. OPP agrees with the SAP that criteria should be added which
would describe a moderate degree of concern in the areas of toxicokinetics and
mechanism of action. The proposed table has been modified and the text of the
revised guidance document has been expanded to address these points.
Comment 10. One commentor (773) asked for clarification of whether the policy
means that: (1) the additional 10X would be applied to the NOAEL from a study in
which young animals showed more sensitivity to effects than adults, only if the
NOAEL was the lowest NOAEL from the available toxicity database; or (2) the
additional 10X would be applied to the lowest NOAEL from the available toxicity
database, if the database contained any study in which young animals showed
more sensitivity to effects than adults. The commentor stated that based on recent
decisions, it appears that OPP is using the second approach.
105
-------
One commentor (773) indicated that OPP's justification for applying an
additional 10X when children appear to be the most sensitive age group-'that OPP
wants a greater level of certainty that children and infants will be adequately
protected" -is an inadequate explanation for departing from its practice and policy
for the first year and a half after passage of the FQPA. OPP needs to explain why it
now concludes that the traditional 10X intraspecies factor is not sufficient to protect
the young, and why it departs from the approach used elsewhere in EPA, in other
federal agencies, and by regulatory organizations throughout the rest of the world.
This commentor and others (e.g., L005) argued that an additional FQPA safety
factor should be used only to compensate for the extent to which the expected
sensitivity difference exceeds the traditional 10X intraspecies factor, and that no
additional factor is needed in cases where the hazard to children is well-
characterized and is considered in the selection of the toxicity endpoint used in
calculating the RfD.
Commentor L005 further noted that an additional uncertainty factor could be
applied on a case-by-case basis, such as where there is inadequate toxicological
information or for those pesticide residues producing severe and irreversible
developmental or reproductive effects. The additional FQPA Safety Factor should
not be applied in all cases simply because a pesticide causes prenatal or postnatal
effects, or merely because the NOAEL for such effects is the lowest NOAEL.
Another commentor (776) indicated that the additional factor should be used in
all cases when the hazard to children is well characterized, and the data show that
infants and children are more susceptible than adults, and if the NOEL from the
most sensitive endpoint from developmental studies has not been used to effect
regulation.
Alternative options were presented by two additional commentors. Commentor
369 stated that its approach is to employ an additional safety factor when the data
set indicates that infants/children are more susceptible, and this is applied when the
study where the effect is noted is used in the risk assessment and setting of any
reference doses. They also indicated that when an endpoint from another study is
selected, the NOAEL selected may provide protection of infants and children, but if
not, an additional safety factor of equal or less magnitude may be considered to
ensure the required margin of safety for the protection of all subpopulations. One
commentor (L005) suggested that if developmental and reproductive toxicity
studies demonstrate that in utero or neonatal exposure results in severe adverse
effects while the same dose has no effect in the adult animal, then in the absence of
human data, an added safety factor to account for the possible increased sensitivity
of human infants and children may be warranted.
The SAP suggested that OPP's approach could be made more readily
106
-------
understandable by inclusion of one or more flow charts in the Agency's background
document. These should highlight decision points, the kinds of inputs considered at
each such point, and the possible alternative outcomes. Separate charts could, for
example, illustrate the past, current, and proposed approaches, and should illustrate
the entire process, including the incorporation of exposure data, leading to the final
regulatory outcome.
Response. In response to commentor (773), if a decision is made to retain the
default FQPA safety factor or apply a different FQPA safety factor, that factor
would be applied to the NOAEL (or BMD) selected as the basis for deriving the
RfD.
The timeline below outlines the approach to selecting the appropriate FQPA
Safety Factor for use in human health risk assessments and the major events
which induced changes in the approach used by OPP.
August 1996
The Food Quality Protection Act is passed unanimously by
Congress.
October 1996
March 1997
First draft of the 10X Policy/Guidance Paper is presented
to the FIFRA Science Advisory Panel.
10X Policy/Guidance Paper Revised Including Response
to SAP Comment
February 1998
March 1998
First organizational meeting of the Health Effects Division
(HED) FQPA Safety Factor Committee is formed and the
working Standard Operating Procedure (SOP) drafted.
Second draft of the 10X Policy/Guidance Paper is
presented to the FIFRA Science Advisory Panel.
May 1999 Third draft of the 10X Policy/Guidance Paper The Office of
Pesticide Programs' Policy on Determination of the
Appropriate FQPA Safety Factor(s) for Use in the
Tolerance-Setting Process (USEPA 1999c) is presented
to the FIFRA Science Advisory Panel along with the
revised SOP.
July 1999 Third draft of the 10X Policy Paper and revised SOP are
issued for Public Comment (64 FR 37001, Docket No.
37001).
Since the enactment of the FQPA in August 1996, there have been a
107
-------
number of changes in OPP's approach to implementation of the FQPA Safety
Factor provision. These changes resulted from the continued examination of the
law and the policy and their implications for OPP and other EPA offices, as well
as from the input of other agencies, the regulated community, and public and
environmental advocacy groups. In 1996, OPP began to implement the FQPA
Safety Factor provision. By the time of the March 1998 presentation to the
Scientific Advisory Panel, a number of conceptual and practical differences had
emerged, and OPP had made some changes to its application of the FQPA
Safety Factor.
First, there was a change in the timing of the determination of the need for
the FQPA Safety Factor within the overall risk assessment process. While the
FQPA Safety Factor was recommended and applied initially during the hazard
and dose response evaluation, now it was being determined during the last
phase of the risk assessment process-risk characterization-taking into account
both the hazard and exposure information on the chemical, as the law instructs
the Agency to do. In order to implement this process with efficiency and
consistency, the Office of Pesticide Programs' Health Effects Division (HED)
formed the FQPA Safety Factor Committee, which developed a Standard
Operating Procedure for making safety factor determinations.
Second, it was concluded that the FQPA Safety Factor-to the extent it went
beyond traditional uncertainty factors-should be determined after the calculation
of the RfD values, whereas previously it had sometimes been applied during the
calculation of the RfD. The concept of the Population Adjusted Dose (PAD) was
implemented in HED to accommodate this two-step process, and to
differentiate decisions under the FQPA from decisions by other Agency offices
and elsewhere that relied solely on uncertainty factors utilized historically in risk
assessment practices. Additionally, it was determined that the Safety Factor
decision would be made only for appropriate populations (infants and children
and women of bearing age) and exposure scenarios/durations, rather than the
population at large.
Third, it was made clear that the default 10X FQPA Safety Factor was
automatically included in each risk assessment, and that this factor could only
be replaced by a different FQPA Safety Factor if, after a full examination of both
the toxicity and exposure databases, it was found that reliable data
demonstrated such different factor to be safe for infants and children.
Fourth, the FQPA Safety Factor, which, to that point in time, generally had
been applied to accommodate uncertainty concerns alone, was now clearly
defined as addressing issues of both uncertainty and sensitivity/susceptibility.
The practical application of this concept had been molded by an OPP policy
108
-------
decision to retain the default FQPA 10-fold Safety Factor or apply some
different FQPA safety factor greater than 1X when any evidence of
sensitivity/susceptibility was observed in the hazard database, regardless of the
degree of concern. Admittedly, one might argue that this has led to some
instances of overprotection as an additional safety factor was applied to a study
NOAEL that adequately characterized the sensitivity (variability) of the effect in
the young.
The May 1999 draft OPP 10X policy document attempted to clarify these
concepts, while at the same time addressing the recommendations of the 10X
Task Force. It is apparent from comments received both from the SAP and the
public that OPP has not been fully successful in communicating its practices in a
clear and transparent manner. OPP believes that there is little value in
developing a comparison of past, present and future practices, as suggested by
the SAP, but it is incumbent on the Program to present clearly, and in as great
detail as practicable, a discussion of the general approach this guidance
describes. It should be noted that there are no plans to revisit past decisions,
simply for the purpose of determining, in the abstract, whether or not residual
overprotection may have occurred. OPP will attempt here, and through the
revised guidance document, to provide greater detail and transparency. The
revised Standard Operating Procedure (SOP) will also include several
examples of the application of the FQPA Safety Factor determination process
for specific chemicals.
Comment 11. Commentor 773 offered the suggestion that EPA implement the
additional Safety Factor provision in a manner that uses the concept of "trial
Population Adjusted Doses" that would then guide the eventual selection of the final
PADs for different age and gender groups and durations of exposure. In employing
this approach, the commentor would propose that OPP should apply any additional
FQPA Safety Factor only to the appropriate toxicity end point, i.e., the endpoint for
which available data suggest that traditional uncertainty factors would not be
adequately protective. Thus, the decision about the application of the FQPA Safety
Factor should depend on the specific risk scenario. The application of the FQPA
Safety Factor would depend on whether the reason is (1) a risk of developmental
toxicity, or (2) a risk of increased "susceptibility" [sic] to nondevelopmental effects.
Further, if the reason for an additional FQPA Safety Factor is an observed
developmental effect, OPP should produce a "trial PAD" by applying the safety
factor, along with other appropriate uncertainty and modifying factors, to the NOAEL
for that effect in the study in which it was seen. If the resulting PAD is the lowest for
a relevant subpopulation, it should be used as the scientific benchmark of safety.
Alternatively, if another toxicity study, relevant to the subpopulation, has a lower
109
-------
NOAEL, such that the application of traditional uncertainty and modifying factors
produces an RfD lower than the "trial PAD," then that RfD (without an further
reduction by an FQPA Safety Factor) should govern. For example, concern about a
developmental effect such as supernumerary ribs should not prompt application of
an FQPA Safety Factor to an RfD calculated from the NOAEL for a brain
cholinesterase inhibition (ChEl) effect in an acute neurotoxicity study. There is no
biological reason to think that the unrelated developmental effect makes the RfD
derived from the ChEl effect inadequate or that the ChEl effect makes the unrelated
developmental effect of more concern.
Commentor 773 goes on to argue that, in some cases it may be appropriate to
consider applying an FQPA Safety Factor based on concerns about potential
developmental toxicity, based on data from a test other than a developmental or
reproductive toxicity study. (For example, a chronic dosing study might show
effects on the thyroid function that could trigger developmental effects not seen in
other tests.) If so, OPP should construct a trial PAD appropriate for the
subpopulation and duration of exposure, and compare it to the RfD otherwise
calculated for the group. This step would resemble the approach described above.
Response. OPP agrees with the basic theoretical concept of "trial Population
Adjusted Doses" which could guide the eventual selection of the final PADs for
different age and gender groups and durations of exposure. A similar type of
comparison has been used informally by OPP when evaluating individual
pesticides. It is noted, however, that the commentor has omitted any
consideration of FQPA's mandate for adequacy of the exposure data in its
calculation of a final PAD. OPP also agrees with the commentor that the FQPA
Safety Factor decision should depend upon the specific risk scenario that is
being evaluated and has made its FQPA safety factor decisions considering
this information. For example, when a safety factor is required for risk to a
specific population subgroup (e.g., infants and children and/or women of
childbearing age) or particular duration of exposure.
However, since the Population Adjusted Dose is calculated as the
Reference Dose divided by the final FQPA safety factor required; and since the
final FQPA safety factor must take into consideration not only toxicity but
exposure considerations; it is not, in practice, feasible to perform a true Trial
PAD analysis for a specific population subgroup or duration of exposure at the
time of establishing RfDs and other toxicity endpoints. OPP believes that the
better approach is that described in the revised guidance document (see
Chapters II and VI), wherein OPP will: determine the RfD(s) for a pesticide
according to Agency policies and practices; evaluate whether the available data
indicate that young animals are more sensitive or susceptible than mature
animals; determine the degree of concern for potential toxicity or data gaps;
110
-------
identify any residual concerns or uncertainties relating to the potential toxicity of
the pesticide or relating to exposure to infants and children; and then base the
final decision to retain or modify the FQPA Safety Factor on the overall risk
characterization.
111
-------
III. Exposure Issues
ISSUE 13. Conservatism of Exposure Assessments for Residues
in Food
Subject to the qualifications expressed in the OPP policy document and the report
from the Exposure Working Group, OPP believes that each of the tiers for estimating
exposure to a pesticide through food, in almost all instances, will not underestimate
exposure to infants and children. Please comment on this conclusion, as it applies
to each of the tiers.
FIFRA Scientific Advisory Panel Comments
Comment 1. One commentor (773) indicated that the current tiered approach to
dietary risk assessment almost always overstated the actual dietary risk from
pesticide residues in food. This commentor also disagreed with the qualifications
on data and methodology included in the document. However, the SAP indicated
that current methods might underestimate possible dietary exposures, particularly
for infants and children and discusses a number of factors in the assessment that
would impact that likelihood including: age of data; sample size of age classes;
demographic stratification; accuracy of recipe files; use of percent crop treated
data; changes in marketing and processing; water; and pesticide residue data.
Response. OPP uniformly attempts to develop the best estimate of exposure
to pesticides in foods possible, using all available data and the best methods
possible. OPP encourages the continued development and improvement of
methodologies and tools to improve the ability to accurately estimate the
potential for exposure to pesticides in foods.
With regard to specific factors cited by the SAP as potentially impacting the
outcome of OPP's dietary risk assessments, OPP has already attempted to
address a number of these. With regard to the age of consumption data used in
the assessments, OPP has recently completed a cooperative effort with USDA
to translate the most recent data available from the Continuing Survey of Food
Intakes by Individuals. These data were collected during the period from 1994-
1996, with additionally sampling in 1998 to increase the numbers of children 12
years or younger in the survey. The additional data were collected specifically to
address the issue of limited numbers of children present in previous surveys and
to allow OPP to better meet its mandate under FQPA. The number of children
under six years of age were approximately doubled, greatly enhancing the
112
-------
confidence in estimating food consumption patterns for this age group. These
data are now being introduced into the OPP risk assessment process.
Demographic characteristics of respondents to the CSFII were collected in
sufficient detail to permit the evaluation of the differing consumption patterns by
age group, income level, region and season. However, due to the variety of
ethnic groups responding to the survey, the ability to evaluate differences in food
consumption based upon ethnic group is limited. In all but a few cases, too few
of any ethnic group are identified to permit evaluation of differences with any
degree of confidence. The SAP has raised the issue of the accuracy of recipe
files to convert consumption of foods to consumption of commodities. This
conversion is a potential source of uncertainty. However, USDA took great care
in researching the most appropriate decisions in developing recipes to reduce
to the extent possible the potential errors in the process. Prior to the translation
step, the recipe files and their underlying conventions were peered reviewed by
experts in the fields of food and nutrition and risk assessment to seek further
guidance on possible errors in the translation process.
The SAP indicated discomfort with the manner in which percent crop treated
data are used to adjust the food risk assessment. The percent crop treated
data used in OPP assessments reflects a synthesis of data from a variety of
sources, not a single data source to which the reader can be referred. OPP
takes this approach to offset strengths and weaknesses in each of the data
sources used and to permit a check of the results of each database based upon
its consistency with others. To this extent, the percent crop treated value used in
OPP risk assessment is in many ways strengthened relative to reliance on any
single source of information. When relying on monitoring data, OPP uses the
percent crop treat value in acute dietary risk assessments of individual
chemicals to correct for the number of nondetectable residue values reported for
the possibility that the commodity may not have been treated. All detectable
residues are retained as they were reported and only the proportion of
nondetects is adjusted. Therefore, the potential for significant impact on the
outcome of the assessment, particularly at the upper ends of the exposure
distribution, is minimal. When monitoring data are not available to estimate
exposure to pesticides in foods, OPP may use residue field trial data combined
with the percent of crop treated data to calculate anticipated residues in foods.
Residue field trial data reflect maximum application rates shortest preharvest
interval allowed on the label. Therefore, the upper end of pesticide residues
likely to be encountered. In addition, percent crop treated data used in the
assessment reflect the likely maximum values based upon agricultural use
statistics produced by public and private sources. As a result, underestimation
of exposure to pesticides in foods is unlikely. To the extent that regional
distributions of residues vary within a commodity, this factor can be addressed
by limiting the scope of the input data to focus on data reflecting the nature of
113
-------
this special case. However, in OPP's experience, this rarely occurs. Within
available monitoring data, little variation between collection points has been
observed. To further characterize any possible differences due to production
area, OPP has undertaken an analysis of available PDP data to definitively
evaluate any evidence of difference based upon region of production, domestic
vs foreign source and a variety of the factors.
The SAP suggested that the product form and processing should be
factored into the dietary risk assessment. Product form and processing are
carefully factored into the risk assessment process, particularly with regard to
the potential for concentrating or reducing the amounts of pesticide residues in
processed foods. These data are used to the extent that they are available and
their submission by the regulated community is encouraged by OPP.
The SAP indicated that drinking water and water in foods should be included
in the risk assessment. With regard to consumption figures for water used for
these purposes, OPP does now incorporate into its risk assessment process
consumption of water from CSFI11994-1996, 1998 using the Office of Water's
extraction of these data. These data were peer reviewed by the Science
Advisory Board and found to be acceptable. These consumption data are
sufficiently robust to permit a reliable estimate of exposure in water when used
in conjunction with reliable drinking water concentration estimates.
Comment 2. Commentor 405 questioned whether a sufficient number of individual
consumption days were available in the CSFII to permit estimation of the 99.9th
percentile of exposure.
Response. OPP has discussed the rationale for why this data set permits a
credible estimation of this percentile of exposure at length in a paper titled
Choosing a Percentile of Acute Dietary Exposure as a Threshold of
Regulatory Concern (US EPA 2000c). See also the response to Comment 1,
Issue 13, above.
Comment 3. Commentor 405 expressed support for OPP in its pursuit of the
capability to conduct chronic probabilistic assessments, reflecting variability in use
patterns and consumption overtime.
Response. While OPP believes that the current deterministic approach will not
underestimate chronic exposure to residues in food, it encourages the collection
of appropriate longitudinal consumption data and use/usage data to make
reliable chronic probabilistic assessment possible. The method of Nusser
114
-------
(1994) cited by the commentor was developed to extrapolate from short term
consumption of nutrients to longer term likely consumption patterns. Because of
the distribution of nutrients in foods, most nutrients are consumed almost daily,
making correction for likely daily intakes possible. However, food frequencies
are much less reliable in that a given food may be eaten only occasionally or
never reported during a given period of a consumption survey. This information
can not be construed to indicate that the food is never or rarely eaten, nor does
consumption on all days of the survey reflect necessarily that the food is eaten
every day. Because of this food frequency issue, the method of Nusser is not
readily adaptable from nutrients to foods.
Comment 4. Commentor 405 also recommended that guidance be developed for
extrapolation from field trial data on parent products to metabolites of concern.
Response. Because the Agency requires field trial data on the parent
compound and all of the identified metabolites of regulatory concern, this
extrapolation is not often required. Nonetheless, while no formal guidance is in
place, OPP does perform this extrapolation using available metabolism data in
lieu of empirical data for residue levels of the metabolite in those rare cases
where doing so is appropriate (i.e., when ratio of residue levels of parent and
metabolite remains constant as the residues decline).
Comment 5. Commentor L005 argues that exposures from residues in food and
exposures from other sources should not be combined because the estimates for
exposures from other sources fail to meet the criterion of "reliable."
Response. OPP has argued that screening level assessments for other routes
should not automatically be combined quantitatively with highly refined
assessments of residues in food. Rather, the screening level assessments
should be evaluated in tandem with the refined assessment of residues in food
to develop an understanding of the overall level of concern for a pesticide from
all appropriate routes and to identify likely areas for further concentration. OPP
is working aggressively to develop improved methods for estimating residential
exposure and drinking water concentration levels. An approach using data from
a pesticide with a similar use pattern as a surrogate for other, less well
documented pesticides is being developed for residential exposure
assessments. This approach is comparable to the use of surrogate data in
PHED to estimate some aspects of worker exposure. In the drinking water
arena, OPP is pursuing the use of modeling in conjunction with monitoring data
to develop a tiered approach which would result in more realistic estimates of
pesticide residue concentrations in drinking water.
115
-------
Comment 6. Commentor 778 indicated that OPP should retain the full 10X
uncertainty factor for cases in which limited information for food consumption are
available.
Response. As discussed in Choosing a Percentile of Acute Dietary
Exposure as a Threshold of Regulatory Concern (US EPA 2000c), OPP
believes that adequate consumption data are available to estimate pesticide
exposure to residues in food for all of the subgroups upon which we currently
regulate.
Comment 7. Commentor 776 also suggested that an uncertainty factor be applied
when exposure data are missing, citing breast milk accumulation and farm area
well water as examples.
Response. In the case of farm area well water, OPP believes that the
conservative, upper bound estimates resulting from ground water modeling
would be protective of this exposure.
OPP is aware of the concern for exposure to infants in breast milk, but rarely has
adequate residue data on human breast milk to assess this potential source of
exposure. If data suggest that exposure by this route is not adequately estimated,
OPP will consider this issue in the judgement as to whether reliable data are
available to permit removal or reduction of the factor. In lieu of empirical data, the
Agency may use data from required ruminant studies, metabolism studies, and
information regarding the physical/chemical characteristics of the pesticide to
estimate residues in human breast milk.
FIFRA Scientific Advisory Panel Comments
Comment 8. The SAP indicated that OPP should endeavor to improve the
understanding of food consumption patterns for infants and children in order to
reduce the uncertainty of assessments for those age groups.
Response. OPP agrees that it is very important to have the fullest
understanding of children's food consumption patterns. To this end, OPP
actively supported the collection by USDA of supplemental food consumption
data for children and has incorporated these data into the risk assessment
process.
116
-------
Comment 9. The SAP also recommended improving the ability to understand the
nature and quantity of pesticides in food stuffs for which little monitoring or residue
field trial data are available.
Response. Again, OPP agrees and has implemented practices by which data
from similar chemical/crop combinations are used as surrogates for those with
less robust data sets. In addition, USDA's Pesticide Data Program (PDP)
continues to evolve, with new chemicals and commodities being added to the
sampling scheme (http://www.ams.usda.aov/science/pdp/index.htm). Usually,
good field trial data are available as well, given that they are required to be
generated in support of registration and/or reregistration.
Comment 10. The SAP questioned the manner in which percent crop treated data
are used to adjust the dietary risk assessment. The panel pointed out that these
data are commonly employed to adjust chronic exposure and risk estimates but are
not available for public review.
Response. OPP's pesticide use-related data come from a variety of sources
for both agricultural and nonagricultural pesticide uses. Sources of use-related
data fall into five general categories: (1) other government entities that produce
pesticide-use data (e.g., USDA's National Agricultural Statistics Service
(NASS) and the National Agricultural Pesticide Impact Assessment Program
(NAPIAP)); (2) data submitted by registrants, user organizations and other
interested parties; (3) proprietary data purchased from vendors whose business
is to obtain pesticide-use data; (4) the scientific literature; and (5) miscellaneous
Agency contracts and sources (For further details see The Role of Use-related
Information in Pesticide Risk Assessment and Risk Management, US EPA
2000e).
The percent crop treated data used in OPP assessments reflect a synthesis
of data from a variety of sources, not a single data source to which the reader
can be referred. OPP takes this approach to bolster the strengths and offset the
weaknesses there may be in any one of the data sources used and to permit a
check of the results of each database based upon its consistency with others.
Consequently, the percent crop treated values used in OPP risk assessments
are, in many ways, strengthened relative to those relying only on a single source
of information. The residue monitoring surveys such as the USDA Pesticide
Data Program (PDP) are used to corroborate the percent crop data since the
statistically representative sampling scheme represents commodities both
treated and untreated.
117
-------
Comment 11. The SAP asserted that regional distributions of residues vary within
a commodity.
Response. In fact, little variation between collection points has been observed
in the available monitoring data. OPP has begun an analysis of the available
PDP data to further characterize any possible differences due to production
area including domestic versus foreign sources and other factors. Preliminary
results indicate some differences between domestic vs. foreign sources, which
are taken into consideration when the Agency undertakes risk mitigation
measures.
Comment 12. The SAP and Commentor L005 indicated the product form and
processing should be factored into the dietary risk assessment.
Response. The consideration of product form and the effects of processing of
raw agricultural commodities has long been an element in the dietary risk
assessment process. Processing studies are required under 40 CFR 158.240.
These data provide information on how processing can affect the levels of
residues expected in final food forms. Processing and product form are
carefully factored into the risk assessment process, particularly with regard to
the potential for concentrating or reducing the amounts of pesticide residues in
processed foods.
Comment 13. The SAP considered the sampling design of federal surveys of
pesticide residues found in imported and domestically produced foods to be a
"significant source of uncertainty in dietary exposure estimates" since: (1) the
number of specific pesticide-food samples are small relative to the volume of food
in the marketplace; (2) processed foods receive limited attention from FDA and
USDA; (3) pesticides that require individualized tests are more rarely sampled than
pesticides detectable via multiresidue screens; (4) an increasing proportion of the
US food supply is imported; and (5) blending portions of crops selected from
different pieces of fruit, or from different crates or shipments will systematically
underestimate pesticide residue levels.
Response. OPP disagrees with the SAP on this point because: (1) monitoring
program sampling is statistically designed to represent residues throughout the
entire food supply; (2) processing studies are required as part of the data set for
registration and/or reregistration of each food use pesticide and the monitoring
programs sample many of the major processed foods (e.g., juices, frozen and
canned fruits and vegetables, etc.); (3) OPP only uses monitoring data if there is
a sufficient number of samples collected in a manner that represents the food
118
-------
supply adequately, and thus if there are not enough samples for a particular
pesticide/use combination in the monitoring database, OPP will rely on field trial
data; (4) monitoring programs collect samples at distribution centers
(warehouses or supermarkets) or ports of entry which contain both domestic
and imported commodities; and (5) OPP has developed a decompositing
method which allows an estimate of residues in single servings (e.g., one
orange or one apple). See Use of the Pesticide Data Program (PDP) in Acute
Dietary Assessment, USEPA1999i). Additionally, numerous single-serving
surveys have been conducted both in USDA's PDP and by industry task forces.
In conclusion, OPP notes that federal monitoring programs have been expanded
to include those foods most consumed by children including milk and grains.
Comment 14. SAP also raised the issue of accidental contamination of food but
agrees that it does not make sense to develop a national regulatory program
around such extremes. However, the Panel also noted that the program should
identify the minimum size of the group that might be impacted by usual consumption
of foodstuffs at some maximum level of probable contamination.
Response. OPP does not make decisions to grant, modify or revoke
tolerances, considering scenarios which may reflect the misuse of pesticides.
There, obviously, could be ways by which misuse could occur, and it would be
virtually impossible to address them in any sensible way in setting tolerance
levels. That is why there are regulatory monitoring systems in place (e.g., FDA
surveillance and compliance programs) designed to prevent contaminated food
from entering the channels-of-trade. Additional tolerance and pesticide misuse
enforcement is conducted by individual states. USDA PDP program conducts
monitoring for pesticide residues in major foods consumed by children and
reports violative residues to FDA. OPP understands that accidental
contamination and/or intentional misuse may occur but this would not likely affect
potential exposure to the population as a whole.
119
-------
ISSUE 14. Conservatism of Drinking Water Exposure
Assessments
OPP is developing a tiered approach to assessing the likelihood and magnitude of
contamination of drinking water and its sources by a pesticide. As an interim
approach, when direct assessment is not possible, is it reasonable and protective to
regard the estimates generated by OPP's current screening methodology as upper
bound pesticide concentrations for surface and ground water and to assume that this
concentration generally will not be exceeded in drinking water?
Comment 1. One commentor (778) stated that EPA does not have "reliable data"
to support using anything other than the additional 10X margin of safety when
estimating the exposure risks in drinking water since sufficient monitoring data
and/or adequate pesticide-specific data are usually lacking. To the contrary, two
other commentors (773 and 405) asserted that since the ecological fate models
grossly overestimate potential exposure from residues in drinking water and the
output represents highly conservative estimates (similar to Tier 1 or 2 analyses for
assessment of residues in food), an additional FQPA uncertainty factor is not
needed.
Response. The peer reviews conducted by the FIFRA Scientific Advisory
Panel in 1997 and 1998 agreed with OPP's view that the screening level
modeling outputs can in general be viewed as high-end estimates of potential
pesticide concentrations in drinking water. Further, these peer reviews
supported OPP's use of this approach for purposes of rapid, cost-effective
screening of pesticides for drinking water-related dietary concerns. An
additional safety factor to account for uncertainty around the accuracy of the
screening level drinking water estimates is therefore not typically warranted.
However, in cases where OPP is placing greater emphasis on what, in reality, is
a very limited amount of drinking water monitoring data for a compound and is
de-emphasizing the screening level model-based outputs in its drinking water
assessment, there is reason to consider the application of an additional safety
factor to account for uncertainty associated with using a limited monitoring data
set. (See also response to comment 4, below, for more discussion of why this
is important.)
120
-------
The long-term goal of OPP is to go beyond screening level drinking water
assessments, to go beyond the use of limited monitoring data with the
application of a safety factor, and to go beyond requiring pesticide-specific
drinking water monitoring data collection. The long term goal is for OPP to have
the capability to develop accurate predictions of pesticide concentrations in
drinking water resources at particular locations where a pesticide is used.
These predictions can then be combined with data on the distribution of water
consumption values across the population in order to produce distributions of
human exposure values associated with ingestion of drinking water from
different sources in different locations. The drinking water exposure
distributions can then either be combined with food-related exposures in some
probabilistic fashion, or, OPP could simply select a "reasonable worst case"
drinking water exposure value from a distribution and combine this value with
the food-related exposure values. In March 2000, OPP and the United States
Geological Survey (USGS) brought two regression-based models to the
Scientific Advisory Panel (SAP) which if realized would allow for the prediction
of pesticide levels at surface water-based community water system intakes.
The March SAP was very favorable in its review and comments and strongly
encouraged the program to continue down this path (USEPA 2000f).
Much progress has been made over the past year in refining the models
used for drinking water assessments. For example, in the case of surface
water, OPP has replaced the field pond scenario used in screening
assessments with an "index" reservoir based on an actual drinking water
reservoir in its Tier 2 screening assessments. The model has also been
adjusted for the percentage of the watershed feeding the reservoir that is
actually in agricultural production in order to more realistically reflect watershed-
scale use. OPP also plans to incorporate the "index" reservoir and percent crop
area in the Tier 1 surface water screening assessments. As indicated above, in
the longer term, OPP is working to develop and validate more sophisticated
approaches for estimating pesticide exposure in drinking water, including the
use of watershed-scale models and regression models which more accurately
capture basin-area processes and would be more appropriate for use in
quantitative human health risk assessments (For further details, the reader is
referred to the document entitled Estimating the Drinking Water Component of
a Dietary Exposure Assessment (US EPA 1999j).
121
-------
Comment 2. Two commentors (773 and 405) agreed that since the estimates of
pesticide concentrations in drinking water are high end or upper bound values, such
estimates should be used only for screening purposes and should not be combined
with estimates of food exposure to produce an aggregate exposure estimate.
Commentor 405 suggested that the approach for determining potential exposure
from drinking water should be consistent with determining potential exposure from
foods when using highly conservative assumptions (e.g., tolerance-level residues,
100% crop treated), the 99.9th percentile of exposure is not an appropriate
regulatory target. Instead, regulating at a lower percentile of exposure would be
highly protective of infants and children when the results of such screening models
are used.
Response. As previously stated (see Issue 13), under some circumstances it
may not be appropriate to quantitatively combine certain tiered assessments. It
is recommended that screening level assessments, for example, be evaluated
in tandem with the highly refined assessments to develop an understanding of
the overall level of concern for a pesticide from all appropriate routes and to
identify likely areas for further data development. OPP is pursuing the use of
improved modeling in conjunction with monitoring data to develop a tiered
approach which would result in more realistic estimates of pesticide
concentrations in drinking water.
Comment 3. Commentors L005 and 405 suggested that the consumption data on
drinking water (i.e., tap water) as well as commercial processing and bottled water
provided by the CSFII be considered as equally "reliable" as the consumption data
reported for any other foods in the surveys. Additionally, Commentor 405
suggested that drinking water be included in the USDA PDP and that finished
drinking water (i.e., not raw surface or ground water) be considered as any other
commodity reported in the USDA surveys.
Response. OPP has incorporated water consumption data from CSFI11994-
1996, 1998 into its risk assessment process using the Office of Water's
extraction of these data. These data were peer reviewed by EPA's Science
Advisory Board and found to be acceptable. These consumption data are
sufficiently robust to permit a reliable estimate of exposure in water when used
in conjunction with reliable drinking water residue concentration estimates.
122
-------
On June 6, 2000, OPP scientists gave a presentation to the Scientific
Advisory Panel concerning a national drinking water survey design for collecting
data appropriate for assessing chronic exposure. Based on the feedback from
the June 2000 SAP, it appears that there would be a number of limitations on
the use of a "PDP-type" approach to drinking water monitoring (i.e., the
collection of finished water samples using the basic statistical design framework
of the PDP) for drinking water. During OPP's consultation with the SAP, the
Agency discussed design options and a proposed survey design framework for
assessing annual average pesticide concentrations in surface-derived drinking
water. The population of interest in the survey for a particular pesticide was
defined as the community water systems (CWSs) using surface-derived
drinking water with the target pesticide used in the watershed. The SAP
generally endorsed the Agency's recommended design structure and
recommended that at least some CWS's in the study be monitored for three
years to assess year-to-year variability. The SAP also recommended that more
emphasis be placed on raw water samples than finished water samples
because of the complexity introduced in temporally paired sampling in a water
treatment facility. They urged the Agency to investigate the impact of treatment
on pesticide concentrations in finished water and take the approach of
scientifically determining the impact of different "units" of the treatment "train" on
different types of pesticides and to use these data to develop appropriate water
treatment factors to account for water treatment effects on pesticide removal
and transformation. Finally, the SAP recommended that the Agency initiate a
pilot scale study to test the framework design and agreed with adding the
census approach to the framework for very large systems (USEPA 2000g).
FIFRA Scientific Advisory Panel Comments
Comment 5. The SAP agreed in prior reviews that the water modeling approach
is sufficiently conservative but notes that the models appear to be most useful for
identifying those pesticides that are unlikely to reach water in appreciable
concentrations. Departure from the upper-bound estimates by virtue of examining
exposure in current databases must be done with caution. In many of the
databases used as a source of monitoring data, it is not possible to identify the
source of drinking water because many cities depend upon several sources. Bias
has also been introduced because negatives may simply come from areas where a
pesticide was not used. Information on drinking water samples seldom identifies
the treatment processes used in sampled drinking water systems which is
absolutely necessary to understand before monitoring results can be
generalized-even to other systems that use the same source water. Consequently,
it is probably not appropriate to rely on databases to determine the potential of a
pesticide to impact drinking water. It would seem that the way that this question can
123
-------
be answered with rigor is to design studies to specifically evaluate this question.
Dependence upon existing or even future databases that may be more
representative may, in fact, not represent the use patterns associated with a
particular pesticide, even though the database could be representative for the
country as a whole.
Another issue raised was that significant exposures to pesticides are likely to
be episodic. Large systems are unlikely to end up with significant exposures for
many reasons. Better water treatment and large volumes increase the likelihood of
dilution and other considerations. Small systems could be exposed to a spill
located close to source, encounter storm events that might introduce particulate
matter into the water, have high levels of local irrigation, etc., that all increase the
vulnerability of drinking water to pesticides.
Response. OPP believes its risk assessments would be strengthened by
additional monitoring data and in parallel, the development of improved
predictive models and approaches for producing more accurate estimates of
pesticide residue concentrations in drinking water. OPP agrees that relying on
limited monitoring data in lieu of current screening level model-based estimates
OR placing more emphasis on limited monitoring data over screening level
estimates should be done with caution, because there is a possibility that the
results of limited monitoring efforts would underestimate concentrations to which
a significant number of people could be exposed in drinking water. OPP
recognizes that, if it is unable to interpret limited monitoring data as to how well
it represents use of the pesticide, that it could misestimate actual exposures
when relying on these measurements in its risk assessment. This suggests the
need for OPP to consider including an additional safety factor for drinking water
exposure in those cases where OPP is relying on or placing more emphasis on
limited monitoring data as the basis for a determination that there is no
significant risk posed by exposure to a pesticide through the drinking water
route.
Efforts currently underway to collect additional monitoring data include
requesting monitoring and runoff studies on individual pesticides from their
sponsors, working with the U.S. Geological Survey (USGS) to obtain more
regional- and national-scale monitoring data on multiple pesticides in reservoirs
and other surface water-based sources, and exploring design considerations for
a national survey of pesticides in drinking water with various government
agencies and industry groups and associations.
124
-------
ISSUE 15. Conservatism of Residential Exposure Assessments
OPP is developing approaches to assess the likelihood and magnitude of exposure
to pesticides in residential and other non-occupational use scenarios. When direct
assessment is not possible, is it reasonable and protective to regard the estimates of
exposure for the major residential and other non-occupational exposure use
scenarios developed by OPP as upper bound estimates of the exposure received by
infants and children from such use?
Comment 1. One commentor (369) indicated that the level of certainty in the
residential SOP-based assessments, particularly with regard to the number of
replicates available in the Pesticide Handlers' Exposure Database (PHED), was
not adequate to permit characterization of whether estimates of exposure were
conservative or not with any certainty. In particular, the commentor pointed to the
departure from the NAFTA criteria for subsetting PHED data which require a
minimum of 15 replicates to be classified as grades A and B for estimation
purposes.
Response. OPP acknowledges that some of the data sets in PHED fall short
of this criterion. However, where user-controlled input variables are high end
values, it is unlikely that the SOPs will underestimate residential exposures.
OPP agrees that the risk assessor must characterize the amount and quality of
data supporting each estimate and highlight areas of concern or uncertainty for
the risk manager. OPP also encourages the production of additional exposure
data that may be used to bolster this valuable tool and to reduce the
uncertainties in the residential risk assessment process.
Comment 2. Commentor 369 also noted the existence of other models such as
CONSEXPO and TherDbase which may be useful in estimating residential
exposure. The commentor encourages the validation of these models.
Response. OPP is aware of these as well as other models that are currently
under development and joins the commentor in encouraging validation be
undertaken. The reader should note, however, that TherDbase is no longer a
supported product.
Comment 3. Commentor 369 also noted that some pathways (e.g., track-in from
turf) have not been quantitatively characterized in the SOPs and recommends that
these scenarios be researched to close this information gap.
125
-------
Response. OPP is looking at other sources of exposure for children that may
be associated with pesticides used in and around the home, including dust/dirt
containing pesticides tracked into the home and related routes of possible
exposure. These routes of exposure, available research, and methods for
assessing these kinds of exposures were presented to the SAP in September
1999. The Panel comments and conclusions can be found in the final report of
the meeting (USEPA 1999k).
FIFRA Scientific Advisory Panel Comments
Comment 4. The Scientific Advisory Panel commended thelOX Task Force
Exposure Working Group for the background document they developed,
acknowledging that it advocated a number of major steps forward in the exposure
assessment process that would help overcome some of the shortcomings in OPP's
Residential Exposure standard operating procedures (SOPs), including the
incorporation of probabilistic approaches, the recognition of narrowly defined age
groups relevant to specific exposure-related behaviors (i.e., prenatal, crawlers,
young toddlers, etc.), movement of pesticides across media (e.g., deposition on
nontarget surfaces), and recognition of the importance of receptor-based (as
opposed to source-based) exposure assessments that examine important
exposure issues from the perspective of how and where children spend time. The
Panel urged OPP to fully integrate the above steps into the exposure assessment
process for nondietary exposures. The SAP remarked that the production of this
document, nearly two years after the initial residential SOP protocols were
developed, is a reversal of the order in which these activities should have taken
place, meaning that it is difficult to answer the question put forward to the Panel
because the question implies that it is possible to judge or determine through
empirical or semiempirical techniques if the scenarios as articulated in the
document are reasonable and protective.
Response. When the Standard Operating Procedures (SOPs) for Residential
Exposure Assessments (USEPA 1999g) were originally developed they were
intended to serve as a screening level tool for use in deterministic risk
assessments and not to be used in probabilistic assessments. The best
available data were used at the time and, in many cases, the level of refinement
was limited by the data that were available. OPP is currently engaged in many
efforts to develop further improvements to the risk assessment process. These
include utilizing new data sources characterizing activity patterns and behaviors
relating to pesticide use in the home, incorporating data into risk assessments
as distributions rather than point estimates, and better defining scenarios
characteristics in the SOPs. The key factors for adding these refinements to our
process are the availability of appropriate data and also the development of
126
-------
appropriate risk assessment methodologies.
OPP has funded, through a cooperative agreement, the development of a
probabilistic model called Lifeline which incorporates residential exposures. It
is also working in various capacities with the developers of several other
probabilistic models that include a residential component. One such model is
Residential SHEDS-Pesticides model, a two-stage Monte Carlo model for
estimating children's exposure to pesticides via the dermal and nondietary
ingestion routes developed by EPA's Office of Research and Development
(ORD). Another model is Calendexfrom Novigen Sciences.
OPP assisted in the planning and execution of a technical workshop held
July 26-27, 2000 and sponsored by the Agency's Risk Assessment Forum
entitled Workshop on Issues Associated with Considering Developmental
Changes in Behavior and Anatomy When Assessing in Children
(http://www.epa.aov/ncea/raf). This workshop involved discussions about
available data and research to help characterize age-related changes in
behavior and exposure factors of children to help improve children's exposure
and risk assessment. Movement of pesticide residues across media and the
development of receptor-based exposure models are also key components to
future OPP efforts related to residential risk assessment. As new and better
data become available from academia, ORD, and industry, the treatment of
these issues in our risk assessment process will become more refined. OPP
will use mechanistic data, such as track-in information, indoor air emission
terms, and residential use data, to better understand multimedia issues. OPP
will also consider survey and videography information to understand how
behaviors contribute to children's exposure in order to develop receptor-based
approaches. Finally, OPP will use population-based monitoring data, such as
those being developed through the National Human Exposure Assessment
Study (NHEXAS), to establish "real-life" exposure levels with which to validate
the models.
Comment 5. The SAP also noted that the decision whether scenario-based
residential and other nonoccupational exposure assessments are sufficiently
conservative so as to not underestimate exposures hinges on several issues:
Whether the scenarios chosen are exhaustive, i.e., have included every
potential exposure scenario and have not overlooked cross media transfer.
127
-------
Whether measurement and assessment data and exposure factors are
accurately characterized.
Whether exposure factors based on data and default assumptions have
been chosen in a consistent manner and reflect within individual variability in
behaviors so that assessors know whether or not contact rates and durations
are truly upper bound.
The timing of exposures relative to one another, given that many pesticide
applications take place on a seasonal basis. It is possible that exposures by
more than one scenario (e.g., turf applications and wading pool exposures)
can take place within a day or days of each other.
Concerns were raised by the SAP and other commentors regarding the
inadequacy of the residential SOPs, particularly weaknesses in assumptions about
hand-to-mouth and object-to-mouth activities and ingestion of dust, soils, and turf.
According to these commentors, the goal should be to use the median values of
well articulated exposure distributions (e.g., body weights or surface areas, for
example) along with conservative but defensible upper-bound estimates where
chemical-specific data do not exist (e.g., 100% inhalation absorption).
Response. The SAP appears to be raising two points: the adequacy of the
measurement and assessment data and exposure factors defined in the
(revised) Residential SOPs and the ability of an assessor to properly assess
aggregate exposures which may occur in the real world. On the first point, in
order to ensure that the standard of "reasonable certainty of no harm" was
established in the risk assessments completed by OPP, the original Standard
Operating Procedures (SOPs) for Residential Exposure Assessments
(USEPA 1999g) were developed using a deterministic approach to exposure
assessment that intentionally produced bounding estimates. This approach is
based on conservative assumptions and, the Agency believes results in
exposure estimates for a single exposure pathway that are sufficiently
protective. The residential SOP's represent the best 'state-of-science' possible
at this time. OPP disagrees with the implication by the commentor that the
measurement and assessment data and exposure factors are not accurately
characterized by the Residential SOP's. Indeed, OPP has expanded
considerably the number and types of residential exposure scenarios included,
e.g., the number of hand-to-mouth contacts per hour has increased, and relied
upon conservative assumptions in the face of uncertainty (see Overview of
Issues Related to the Standard Operating Procedures for Residential
Exposure Assessment, USEPA 1999n).
As previously stated, other efforts within the Agency to further improve
128
-------
exposure estimates, including the development of the Residential Stochastic
Human Exposure and Dose Simulation Model for Pesticides (or Residential-
SHEDS model) by the Office of Research and Development. A publication
describing this model framework and presenting a case study application of the
model is on NIEHS' Environmental Health Perspectives web site
(http://ehpnet1.niehs.nih.aov/docs/2000/108p505-514zartarian/abstract.htmh.
and the manuscript published in the June issue of Environmental Health
Perspectives (Zartarian et al. 2000). A peer review of the SHEDS-Pesticides
model will be conducted in the near future.
The second point raised by the SAP is the ability of an assessor to
adequately perform and characterize an aggregate exposure which includes the
residential pathway. OPP points to the revised guidance for conducting
aggregate exposure risk assessments (US EPA 2001b) which articulates the
Agency's framework in which to evaluate aggregate exposure. The fundamental
approach of the guidance is a move toward population-based exposure and risk
assessment which is comprised of data representing "coherent individuals" in
the population. By "coherent individual," OPP refers to the use of various data
sets or professional judgment to represent individuals in the population which
are spatially and temporally linked to, and demographically consistent with, each
individual. This 'linking' of data sets, each data set possibly playing a part in
defining the total aggregate exposure to the individual in the population, or use
of professional judgement to define likely linkages, will ensure the use of all
relevant exposure scenarios, timing of exposures, and move closer to reflecting
the interindividual variability accurately. In this way, aggregate exposure
assessment can be used to analyze the (representative) total population. OPP
is confident that the use of such data sets which are based on an individual can
be developed using available data sources which adequately and accurately
assess the temporal, spatial and demographic characteristics of a
representative population within an aggregate assessment.
129
-------
ISSUE 16. Overall Conservatism of Exposure Assessments
In OPP's view, its aggregate exposure assessments generally do not underestimate
the exposure to infants and children because the aggregate exposure is calculated
by adding the high-end estimates of exposure to pesticides in food, to the high-end
estimates of exposure to pesticides both in water and as a consequence of pesticide
use in residential and similar settings. Please comment on this view.
Comment 1. Commentor 773 (pp. 7, 36, 56) states that the models OPP currently
uses to estimate drinking water and residential exposure are so conservative that
they are extremely unlikely to underestimate exposure, and that is a good reason for
not applying an additional FQPA uncertainty factor because of nonfood issues.
OPP should, however, change its approach to calculating aggregate exposure.
OPP should not incorporate deliberate overestimates of pesticide exposure into
assessments just for the purpose of avoiding the use of an additional 10X Safety
Factor.
This commentor (p.56) agrees that the current OPP estimates of exposures
from residential pesticide use are high end, upper bound values, but disagrees with
OPP that such estimates should be combined with estimates of food exposure to
produce an aggregate exposure estimate. Such estimates should be used only for
screening purposes.
Commentor 369 concurred that a simple additive approach to aggregate
exposure (i.e., summing high-end estimates for food, water and residential) would
not likely underestimate exposure to infants and children. However, as the Agency
moves toward probabilistic approaches, it will be important that the level of
confidence in the inputs is high; the methodological and data development currently
underway will eventually increase the level of certainty associated with these inputs.
Response. OPP bases its exposure estimates on a variety of models which
reflect the various pathways through which exposure could occur. These models
utilize the data described in 40 CFR: residue chemistry data (40 CFR 158.240)
used in dietary food risk assessment; environmental fate data (40 CFR
158.290) used in dietary drinking water risk assessment; and reentry protection
data (40 CFR 158.390) used in residential (nonoccupational) risk assessment.
Many other data sources, modeling techniques, and other resources are
available to supplement the guideline exposure studies submitted to the Agency
and together allow consideration of all potential exposure pathways.
Additional data sources for dietary food risk assessment may include:
130
-------
Food and Drug Administration (FDA) Regulatory monitoring data (crop
surveillance and compliance data); United States Department of Agriculture
(USDA) Pesticide Data Program monitoring data; State monitoring data; FDA
Total Diet Survey; and consumption data from the USDA Nationwide Food
Consumption Survey (NFCS) and Continuing Surveys of Food Intakes by
Individuals (CSFII).
Additional data sources for dietary drinking water risk assessment may
include monitoring data from: the United States Geological Survey (USGS)
National Water-Quality Assessment Program (NAWQA Program) and the
National Stream Quality Accounting Network (NASQAN); the Office of Water
STORET database; the OPP Pesticides in Groundwater Database; and the
National Pesticide Survey.
Additional resources for residential (nonoccupational) risk assessment
include: the Pesticide Handler Exposure Database (PHED); the Exposure
Factors Handbook; and the OPP Standard Operating Procedures (SOPs) for
Residential Exposure Assessments; chemical-specific use/usage data from
OPP Biological and Economic Analysis Division; and Center for Disease
Control (CDC) biomonitoring or epidemiology data.
Collectively and individually, these data sets have both strengths and
limitations. EPA believes that its approach to using these data is generally
conservative, i.e., unlikely to underestimate potential exposure. On a case-by-
case basis, however, EPA will examine a particular exposure assessment to
determine-in light of the particular data used and the methods of combining the
data-whether the assessment may understate exposure.
Comment 3. 778 (NRDC), p. 8. The absence of child-specific exposure data is
sufficient reason to apply the FQPA 10X margin of safety. It is not appropriate to
argue that incomplete, limited information or data may be used to develop an
estimate that does not understate exposure and therefore that the Agency may
ignore the statutory mandate to apply the 10X.
Response. See the response to comments under Issue 17 below.
Comment 4. 778 (NRDC), p. 3. The "completeness" of the exposure database
under FQPA is different from the adequacy of the database to support registration.
Response. See the response to comments under Issue 17 below.
131
-------
Comment 5. 781 (Farmworker Justice Fund; p.6) could not agree with the
Agency's blithe assumption that "it is not underestimating exposure to infants and
children or to the general population" and that "the FQPA Safety Factor would not
be needed to address uncertainties in the exposure database." The 1993 NAS
Report, Pesticides in the Diets of Infants and Children, raised many concerns
about the quality of EPA's pesticide residue data and problems with the reliance on
"average" (adult) exposures. Nor has the Agency accurately modeled children's
exposure to pesticides at home, in school, or through drinking water. Given the
deficiencies found by the NAS report, the scant actual data, and the untested nature
of EPA's assumptions, it cannot be said that the exposure data is complete for the
general population of infants and children.
Response. OPP exercises great care in ensuring that its estimates of
exposure to infants and children are health protective. A number positive steps
have been taken to improve the risk assessment process and the quality of
available data since the issuance of the 1993 NAS Report. OPP now has
extensive, age-specific food consumption data for infants and children. This is
combined with monitoring data designed to target pesticide residues in the
major children's foods. USDA's Pesticide Data Program generates nationally
representative data defining pesticide residues with great precision. Exposure
to pesticides in foods is estimated using the full range of the distributions of
food consumption and pesticide residues, including information specific to
childhood exposures. Water concentration estimates used in OPP risk
assessments currently reflect high end values, making underestimation of
exposure to pesticides in drinking water unlikely. OPP also includes a number
of conservative assumptions in its residential exposure assessments to ensure
that exposures in the home are not underestimated. These conservative
assumptions produce estimates of exposure that are sufficiently health
protective to encompass exposures to children in a variety of settings including
farming areas.
Comment 6. 781 (Farmworker Justice Fund), p. 6-7. The inadequacy of data
concerning exposures to the general population of infants and children is
compounded many fold for the children of farmworkers. These children face many
more exposures from drift, take-home exposure and direct exposure in the fields.
Since neither the EPA's actual data nor its assumptions adequately take into
account the exposures of farmworker children, the exposure database for virtually all
pesticides must be viewed as incomplete, necessitating at least the imposition of a
10X safety factor.
Response. See the response to comment 7 below.
132
-------
Comment 7. 781 (Farmworker Justice Fund). Farmworker children should be a
sentinel population in determining the adequacy of the exposure data. Farm
children experience a wider range of exposures to pesticides than most other
children (e.g., playing/walking in fields, spray drift, parents' contaminated clothing,
fruits/vegetables from the fields, dust in homes, spray drift at school, drinking water).
Children of rural residents may be viewed as a second significant subpopulation
because they share some of the same sources of exposure. In support, the
commentor, cited several exposure studies. Because of the significant body of
data demonstrating the extent of pre- and postnatal exposures of farmworker
children to pesticides, farmworker children should be deemed a sentinel population
by EPA for purposes of determining the adequacy of the exposure database for
purposes of applying the 10X safety factor.
Response. Again, OPP agrees that it is important to assess whether farm
worker children are currently at risk and to address that risk. However, OPP has
no information indicating that exposures to farm worker children differ
significantly from exposures to children in the population at large. OPP includes
a number of conservative assumptions in its residential exposure assessments
to ensure that exposures in the home are not underestimated. Thus, for
pesticides registered for residential uses, these conservative assumptions
produce estimates of exposure that generally encompass exposures to children
in a variety of settings, including farming areas. Since FQPA was passed, the
Agency has generally been examining the special vulnerabilities of children, has
adopted specific policies (such as the 10X policy) to protect children when
available information indicates such vulnerability, and has been developing
specific methods for estimating risks to children resulting from exposure to
registered pesticide uses. Specific to farm worker children, EPA is continuing
to seek and evaluate information as to whether certain activities of children,
such as playing on lawns, are analogous to exposures that farm children may
receive. EPA is also developing with USGS a project to gather data on
farmworker exposure to pesticides from their drinking water sources. The issue
of exposure to farm worker children was included in the presentation to the SAP
in September 1999. The Panel comments and conclusions can be found in the
final report of the meeting (USEPA 1999k).
133
-------
FIFRA Scientific Advisory Panel Comments
Comment 8. The SAP postulated that combining data of varying quality (i.e., food,
water, nondietary) with widely different confidence intervals affects the end result.
Deterministic approaches are not necessarily always more conservative than
assessments that use distributional approaches, especially when the data sets for
concentration, contact rates, and duration are robust. This is not a reasonable view
in light of the severe defects in the assessment of nondietary exposures of the fetus,
infants, and children. As a result, there is no confidence in the assessment of
aggregate exposure.
Response. OPP disagrees with the SAP that there is no confidence in the
assessment of aggregate exposure due to the combining of data of varying
quality. OPP agrees that aggregate exposure assessment may include data of
different quality, collected for different purposes, and that such data can be
treated both as distributional and as point estimates of exposure. However, the
goal of the aggregate assessment is to give as accurate and reasonable an
estimate of distributions of risks as possible, without underestimating the upper
tail. Others have commented on and expressed support for OPP's use of
different types of data sets for the purpose of aggregate assessment in the past.
For example, the I LSI panel and previous SAPs have stated that it is acceptable
to combine probabilistic and deterministic types of data in the food pathway.
OPP believes that a full understanding of the uncertainty and a careful
characterization of the results is vital when combining different types of data.
OPP agrees that sensitivity analyses are key to fully understanding the
combined data distributions. OPP also agrees that it is most appropriate to use
central tendency values when defining deterministic values. The definition and
description of data distributions is addressed in an accompanying OPP
Science Policy document entitled Guidance for Submission of Probabilistic
Human Health Exposure Assessments to the Office of Pesticide Programs
(US EPA 1998b) OPP is investigating the use of two-stage Monte Carlo
techniques for use in probabilistic assessment models which will aid the Agency
in quantifying the uncertainty and variability of the estimates.
134
-------
ISSUE 17. Reliable Data
Several commentors raised the issue of how the terms "reliable data" and "reliable
information" in the FQPA should be interpreted. Although these comments address the
application of the children's safety factor, the statutory interpretation issues raised by
these comments have implications for many other aspects of implementation of the
FQPA. Because this interpretational issue should not be viewed in isolation, OPP has
attempted to address below all of the comments provided on this issue. OPP first
discussed these issues in the Response to Comments document for the Aggregate
Exposure policy in December, 2001. The FQPA Implementation Working Group (IWG)
submitted further comments on these issues in a letter responding to the Aggregate
Exposure Policy Response to Comment document. IWG Letter to Stephen Johnson
and Robert Fabricant (January xx, 2002) [hereinafter cited as "IWG Letter"]. IWG
asserted that OPP mischaracterized its earlier comments on these issues and did not
respond adequately to them. This Response to Comments document addresses these
latest comments from IWG as well.
The two primary statutory provisions cited by commentors are the general definition
of safety in section 408(b)(2)(A)(ii) and the language in section 408(b)(2)(C)
addressing when it is appropriate for OPP to select an FQPA Safety Factor "different"
from the additional tenfold factor to protect infants and children. Section
408(b)(2)(A)(ii) states:
[T]he term "safe"...means that [EPA] has determined that there is a reasonable
certainty that no harm will result from aggregate exposure to the pesticide chemical
residue, including all anticipated dietary exposures and all other exposure for which
there is reliable information.
Section 408(b)(2)(C) provides that:
Notwithstanding such requirement for an additional margin of safety, [EPA] may use
a different margin of safety for the pesticide chemical residue only if, on the basis of
reliable data, such margin will be safe for infants and children.
Another relevant provision is subsection (b)(2)(D)(i) which provides that in making
tolerance decisions:
[EPA] shall consider, among other relevant factors, ...the validity, completeness,
and reliability of the available data from studies of the pesticide chemical and
pesticide chemical residue...."
135
-------
Industry Comments
Essentially, industry's view is that the "reliable information" language in section
408(b)(2)(A)(ii) is a screen, or gatekeeper, for both whether nondietary exposure
can be considered as part of a pesticide's aggregate exposure for the purpose of
making the safety determination and whether nondietary exposure can be
considered in judging the completeness of the exposure database for children's
safety factor purposes. According to IWG (whose views were repeated in
comments from others, e.g., Commentor L016 (Michigan Farm Bureau) [hereinafter
cited as IWG 10X Comments], Congress' intent in including the "reliable
information" language in section 408(b)(2)(A)(ii) was to ensure that data on
pesticide exposure, other than exposure through residues in food, would only be
taken into account in assessing aggregate exposure if there exist data providing "a
reasonable estimate of the actual, real-world level of exposure to the
pesticide...including] information on the distribution of the exposure, so that
probabilistic estimates of aggregate exposure can be made" (IWG 10X Comments
at 39). Moreover, IWG argues that the reliable information language in section
408(b)(2)(A)(ii) not only serves as a screen for what information is considered in
calculating aggregate exposure but also as a gatekeeper for what exposure
scenarios should be considered in determining, for children's safety factor
purposes, the completeness of the exposure database.
As stated by IWG, "[w]e do not think that Congress meant that when [OPP] is
assessing the acceptability of the risk from a well-defined exposure, it should have
to add a 10X factor to account for some other possible exposure for which there are
no reliable data" (IWG Roadmap at VIII-7). Thus, IWG asserts that if OPP does not
have reliable information on a nonfood exposure scenario, that exposure scenario
should be completely excluded from the frame of reference in making safety factor
decisions. In other words, IWG does not believe that the "reliable data" test for
assigning a different FQPA Safety Factor even comes into play as to nonfood
exposure scenarios lacking reliable information precisely because OPP does not
have reliable data on this exposure scenario. IWG states, however, "that in some
circumstances there might be enough quantitative information [on nondietary
exposure] to satisfy [the "reliable information" requirement],..but there still might be
some need for an additional uncertainty factor." IWG Letter at xx. Further, IWG
states that "[i]n the unlikely event that important information on exposure via food
residues is missing, and such information might show a significant risk to infants or
children that is not accounted for by normal use of probabilistic exposure
assessments in assessing food-based risk, then [OPP] has the authority to retain
an additional FQPA safety factor to address that uncertainty." (IWG 10X
Comments at 40).
136
-------
Given this legal interpretation, IWG criticizes OPP's approach of using models
to insure that exposure is not underestimated for drinking water and residential
exposure issue. OPP has relied on such models to conclude that it has "reliable
data" on exposure issues in making decisions on the children's safety factor. IWG
claims that this approach is unnecessary because "[m]odels designed to produce
conservative overestimates, and the overestimates that they generate, cannot be
considered 'reliable information' for purposes of the 'aggregate exposure'
computation..." (IWG 10X Comments at 39).
There are several building blocks to the IWG's legal interpretation. First, IWG
asserts that the phrase in subsection (b)(2)(A)(ii) "for which there is reliable
information" applies not to the requirement to consider "all anticipated dietary
exposures" but only to consideration of "all other exposures..." Thus, according to
IWG, subsection (b)(2)(A)(ii) imposes a reliability test on nondietary exposures but
no such test on dietary exposures. IWG claims that Congress chose not to impose
a "reliability criterion" on dietary exposure information because Congress was
aware of the quality of the data on such exposure already in OPP files and the
reliability of government residue monitoring programs. Second, IWG contends that
the term "dietary exposure" only extends to pesticide residues in food and not
residues in drinking water. Thus, IWG argues that the reliability criterion attaches to
exposure to pesticide residues in drinking water. Although IWG admits that "water
is often thought of as a component of the diet," IWG asserts that the language of the
statute and a pre-FQPA action by OPP and FDA suggest that drinking water is not
included in the term dietary exposure. (IWG 10X Comments at 38-39. The
statutory language IWG cites is section 408(b)(2)(D)(vi) that describes aggregate
exposure as "including dietary exposure under the tolerance and all other tolerances
in effect for the pesticide chemical residue, and exposure from other
nonoccupational sources." The OPP/FDA action noted is the joint agency
interpretation following passage of the Safe Drinking Water Act that the term "food"
in the FFDCA does not include drinking water.
IWG does not contend that this legal interpretation is compelled by the statute;
however, it does assert that its interpretation is a permissible one. It claims that "[i]t
would be extremely arbitrary for OPP to proceed to use worst-case model
estimates without stating publicly whether it has the legal authority to do so, and
whether it would be good policy" (IWG 10X Comments at 40).
137
-------
NRDC Comments
NRDC takes a dramatically different approach to the terms "reliable data" and
"reliable information" as they are used in subsections 408(b)(2)(C) and
408(b)(2)(A)(ii), respectively. NRDC argues that OPP has inappropriately merged
the concepts of reliable data and reliable information. Although NRDC does not
explain how it would define either of these terms, it does make clear that it believes
"reliable information" sweeps more broadly than "reliable data." Importantly, as to
the children's safety factor, NRDC asserts that exposure estimates based on
models are not data. According to NRDC such model estimates are
information-information that must be considered in calculating aggregate
exposure-but not data sufficient to address concerns about the completeness of
the exposure database and not reliable data sufficient to justify choosing a different
safety factor than the additional tenfold children's safety factor.
In support of this argument NRDC points out that in two places, section 408(b)
refers to both data and information in a single provision. See §408(b)(2)(E)(i) and
(b)(2)(F).
Response. The IWG's and NRDC's approaches to the terms "reliable data"
and "reliable information," exposure estimates from models, and the children's
safety factor could not be more polar. IWG claims OPP's existing models for
drinking water and residential exposure may not be considered as reliable
information and, therefore, IWG would not include such model estimates of
exposure in aggregate exposure. Further, IWG believes that exposures
excluded from consideration under aggregate exposure are irrelevant to the
children's safety factor decision and, thus, in their view, the inability of a model
to yield reliable information for an exposure scenario would not necessitate
retention of an additional tenfold safety factor due to incompleteness of the
exposure database. In contrast, NRDC argues that model estimates are
reliable information but not reliable data. Thus, NRDC would include model
estimates in calculations of aggregate exposure and would conclude that, if
OPP is using a model to estimate exposure, reliable data do not exist to permit
removal of the additional tenfold children's safety factor.
OPP views both of these positions as extreme and cannot agree with either
one. Each of the points raised by the commentors is discussed fully below.
138
-------
The Reliability Requirement and Aggregate Exposure. OPP believes that
IWG misreads the statutory requirements pertaining to reliable information on
aggregate exposure. The IWG argues that Congress, by inserting the reliable
information requirement in that provision, was erecting a special standard of
reliability applicable to nondietary exposures whereas no reliability requirement
was applied to dietary exposures. (See IWG 10X Comments at 38 ("Regarding
dietary exposure, the statute does not impose a 'reliable information'
requirement.")) Not only does this special standard have a role as a gatekeeper
for what nondietary exposures may be included in aggregate exposure, but,
according to IWG, the "reliable information" requirement has a substantive
content as well. IWG asserts that the "reliable information" requirement
mandates that nondietary exposure data must demonstrate not just the
magnitude of exposure but the magnitude of "actual, real-world exposure levels"
including "information on the distribution of the exposure, so that probabilistic
estimates of aggregate exposure can be made." (IWG 1OX Comments at 39).
IWG's theory that the "reliable information" requirement serves as
gatekeeper for aggregate exposure is not necessarily problematic by itself.
However, when this gatekeeper argument is coupled with IWG's interpretation of
the scope of the "reliable information" requirement (i.e., the "reliable information"
requirement has a specific substantive scope addressing the magnitude of
exposure), IWG's approach becomes implausible. There is no basis in the
language of the statute for giving an evidentiary type of requirement-reliability-a
substantive content. Moreover, IWG's interpretation is difficult to square with the
structure of the FQPA and particularly the reliability requirements in subsections
(b)(2)(D)(i) and (b)(2)(C). Finally, IWG's approach runs counter to the clearly-
expressed congressional intent to provide additional protection to infants and
children when there are concerns about the completeness of the exposure
database. These issues are discussed in more detail below.
To What Exposure Data Does the "Reliable Information" Requirement Apply?
OPP thinks that the better interpretation of subsection (b)(2)(A)(ii) is that the
"reliable information" requirement applies only to nondietary exposure and not
dietary exposure. IWG believes strongly that this is so. IWG Letter at 3
(describing as "untenable" the notion that the reliable information requirement
applies to dietary exposure); IWG 10x Comments at 38 (Regarding dietary
exposure, the statute does not impose a "reliable information" requirement.).
OPP also thinks that the statutory language is sufficiently ambiguous that phrase
in subsection (b)(2)(A)(ii) referencing "reliable information" could be interpreted
as applying to both dietary and nondietary exposure and thus reinforces the
reliability consideration in subsection (b)(2)(D)(i). Nonetheless, OPP prefers the
first interpretation to the second because it appears to be the more natural
construction of the language in subsection (b)(2)(A)(ii). Further, as the
139
-------
discussion below will show, OPP's interpretation of the meaning of the "reliable
information" requirement supports the reasonableness of the first interpretation
because it allows this requirement to mesh in an understandable fashion with
the reliability mandate in subsection (b)(2)(D)(i) and the "reliable data"
requirement in the children's safety factor provision.
Under OPP's preferred interpretation, OPP can, in theory, agree with the
gatekeeper argument advanced by IWG regarding the "reliable information"
language in subsection (b)(2)(A)(ii)-namely, if the "reliable information"
requirement is not met for nondietary exposure, OPP should not assume that
such exposure will occur in its estimation of aggregate exposure. However, as
discussed below, because OPP disagrees with IWG concerning the
"substance" of the reliable information test, OPP differs with IWG on how this
test should be implemented.
What Qualifies as Reliable Information? OPP believes that the key to
interpreting the meaning of the "reliable information" requirement are the plain
language of that provision and the other statutory provisions bearing on the
issue of reliability. The common meaning of "reliable information" is information
that is trustworthy, or in the scientific sense, information that is reproducible.
However, both the construction of subsection (b)(2)(A)(ii) as well as other
provisions in the statute addressing reliability convince OPP that the "reliable
information" requirement in subsection (b)(2)(A)(ii) does not simply impose a
general trustworthiness standard to all aspects of nondietary exposure
information.
As noted above, the better interpretation of subsection (b)(2)(A)(ii) is that the
"reliable information" requirement applies only to nondietary exposure and not
dietary exposure. Ordinarily, one would presume that Congress would be
concerned about the reliability of any exposure data, whether dietary or
nondietary, forming the basis of an OPP safety finding. This differential
application of the "reliable information" requirement, therefore, must reflect a
congressional conclusion that, at least as to some aspect of exposure data, a
demonstration of reliability was important when such data concerned nondietary
exposure but was unnecessary regarding dietary exposure.
OPP used this principle as a guide in interpreting the "reliable information"
requirement in subsection (b)(2)(A)(ii). In other words, OPP sought an
interpretation of the "reliable information" requirement that focused on reliability
issues that are likely to be unique to nondietary exposure. In OPP's judgment,
the reliability question that is unique to nondietary exposure in the pesticide
tolerance context is whether exposure is occurring by the nondietary route.
Because the FFDCA concerns tolerances, or maximum residue levels for
140
-------
residues in food, it can be presumed that any tolerance action involves exposure
through the diet. See 21 U.S.C. §346a(a) (describing the scope of section 408
as extending to "any pesticide chemical residue in or on a food In such
circumstances, no reliable information is required to show that exposure can
occur by the dietary route. On the other hand, that a tolerance permits residues
in food does not mean that there is human exposure through nondietary
channels (e.g.. oral exposure to pesticides of young children from hand-to-mouth
behavior in areas in or around homes where pesticides have been used). For
example, many pesticides registered for use in agriculture have no permitted
uses in residential settings that could lead to nondietary exposure. Thus, it is
logical to demand reliable information that nondietary exposure can occur by the
nondietary route prior to considering such exposure in making a safety
determination. Accordingly, OPP interprets the "reliable information"
requirement as being directed primarily at identifying whether exposure occurs
by a certain pathway or route (i.e., nondietary). Such an interpretation meshes
well with the statutory language because the provision in question specifically
includes the reliable information requirement in the context of mentioning the
routes of dietary and nondietary exposure.
OPP's interpretation also harmonizes the "reliable information" requirement
in subsection (b)(2)(A)(ii) with the mandate in subsection (b)(2)(D)(i) for OPP to
consider the "reliability" of data in making safety determinations. Although at
first blush there appears to be some tension between subsection (b)(2)(A)(ii)
imposing a reliability requirement only on nondietary data and
subsection(b)(2)(D)(i)'s direction to consider the reliability of data generally, any
such tension is relieved by limiting the reliability inquiry in subsection (b)(2)(A)(ii)
to an issue-whether exposure is occurring by the route in question-for which no
reliability inquiry is necessary as to dietary exposure based on the regulatory
action involved-i.e., setting maximum pesticide residue levels in food. As is
appropriate, questions regarding the magnitude and distribution of dietary
exposure still would fall under subsection (b)(2)(D)(i)'s reliability inquiry.
IWG takes a quite different approach. It asserts that, because OPP has
specific and reliable information on the actual residue levels and distribution of
those residues in food, the reliable information requirement must have been
intended to assure that OPP had a similar type of reliable information on
nondietary exposure. In IWG's words, "[f]or information to be considered
reliable..., it must provide a reasonable estimate of the actual, real-world level of
exposure to the pesticide ...[including] information on the distribution of the
exposure, so that probabilistic estimates of aggregate exposure can be made."
IWG Letter at xx; IWG 10X Comments at 39. IWG does not believe that this very
specific and substantive construction of the term "reliable information" as it
applies nondietary exposure implies that OPP is without power to investigate
141
-------
the reliability of data on the levels and distribution of dietary exposure. Rather,
IWG asserts that OPP can demand reliable information on dietary exposure
based on the mandate to consider data reliability in subsection (b)(2)(D)(i) and
OPP's data regulations and guidelines. Further, IWG states that "missing" data
pertaining to infants and children can be addressed through retention of the
additional safety factor for the protection of infants and children, as provided in
subsection (b)(2)(c).
In OPP's view, IWG does not provide an adequate explanation for its
infusion of the term "reliable information" with a specific, substantive content.
First, it is a considerable leap to interpret a general statutory term regarding
essentially an evidentiary issue ("reliable information") as imposing a
substantive requirement of the type of information that must be submitted
(information on the "actual, real-world level" and "distribution" of residues). OPP
is reluctant to preclude consideration of "reliable information" concerning
nondietary exposure that does not fit IWG's specific substantive construction of
that term without reference to some language in the statute or legislative history
that supports IWG's interpretation. IWG provides no such support.
Second, OPP questions the accuracy of factual premise to IWG's
argument-that OPP has data on "actual, real-world level" and "distribution" on all
pesticides used on food. Actually, there is a fair amount of variability in the
quality of data OPP has on pesticide residues in food. Although OPP does
have very specific and detailed information concerning the "actual, real-world"
residue levels and distribution of those residues for many agricultural pesticides,
this is certainly not true as to all pesticides used on food. For many pesticides,
OPP may have little or no data on actual residues from monitoring of food in
distribution channels (especially as to new pesticides) but instead OPP must
depend on residue data from residue field trials. Residue levels measured in
such trials are higher, often significantly so, than the actual residue levels to
which consumers are exposed and such studies may provide limited information
concerning the distribution of residue levels.1 IWG's factual premise is also
weakened by its concession that it is perfectly appropriate-even required-that
OPP evaluate the reliability of data pertaining to pesticide residues in food.
IWG Letter at 4. After all, IWG's interpretation of the "reliable information"
requirement is based on the proposition that the statute imposed no "reliable
information" requirement on dietary exposure because "there was no general
need for concern about whether available information on exposure from food
residues is reliable." IWG Comments at 4.
1IWG does not address how its "reliable information" test-information comparable to that on dietary
exposure - can be consistently applied in the face of this wide variation with regard to exposure information
on food.
142
-------
Third, OPP does not believe that IWG has provided an adequate explanation
of how the "reliable information" requirement in subsection (b)(2)(A)(ii) can
comfortably be integrated with the mandate to assess data reliability in
subsection (b)(2)(D)(i). Given IWG's definition of the "reliable information"
requirement as demanding reliable information on the actual level and
distribution of residues, it seems inconsistent to assert, on one hand, that
"[Regarding dietary exposure, the statute does not impose a 'reliable
information' requirement," IWG 10X Comments at 38, and to argue, on the other
hand, that the reliability mandate in subsection (b)(2)(D)(i) and OPP regulations
and guidelines mandate that OPP test the reliability of dietary information on the
level and distribution of pesticide residues.2
Further, IWG does not address the considerable interpretational problem
posed by its construction of subsection (b)(2)(D)(i). If, in fact, the statute does
impose an unqualified reliability test on the level and distribution of dietary
exposure information through subsection (b)(2)(D)(i), that same provision would
impose a similar test on nondietary information rendering the "reliable
information" requirement in subsection (b)(2)(A)(ii) wholly superfluous. Similarly,
OPP believes that all of IWG's attempts to describe alternate general sources of
authority to investigate the reliability of dietary exposure data tend to undermine
its attempt to give the "reliable information" requirement a specific, substantive
content because these other sources are based in the end on OPP's broad
authority to require the submission of data or authority not limited to data on
residues in food.
Finally, OPP does not accept the IWG view that Congress did not impose a
"reliable information" requirement on dietary exposure information because it
intended that a lack of reliable information on such exposures be addressed
through the retention of the additional ten-fold factor for the protection of infants
and children. The IWG approach unreasonably restricts OPP's ability to deal
with issues relating to the reliability of dietary exposure data. Specifically, under
IWG's approach, OPP's only recourse (unless IWG's theory that subsection
(b)(2)(D)(i) provides a redundant reliability requirement is accepted), when
dietary exposure data is judged unreliable, becomes retention of the additional
2ln fact, some statements by IWG seem to leave little room for the subsection (b)(2)(D)(i) reliability
mandate as to dietary exposure: "If [OPP] is correct [in interpreting the term "dietary" as including drinking
water], then drinking water exposure always must be determined and the aggregate of food and drinking
water exposure must be tested against the FQPA safety criterion in making decisions on food tolerance,
without regard to the reliability of the information on drinking water exposure." IWG, Comments on
Guidance for Performing Aggregate Exposure and Risk Assessments (February 10, 2000) (emphasis
added).
143
-------
10x factor for the protection of infants and children.3 In contrast, OPP's
interpretation provides OPP with the flexibility to reject all or part of dietary
exposure data with reliability problems and deny a tolerance petition as
unsupported or to use all or part of such data in conjunction with an additional
factor for the protection of infants and children. Because the additional
children's safety factor is designed to provide increased protection for infants
and children in response to concerns regarding, among other things, the
completeness of the exposure database, OPP believes the children's safety
factor should be implemented in a manner that provides additional protections
for infants and children and not as a substitute for standard administrative
practices such as demanding that reliable information support safety
determinations on tolerances. Further, OPP is reluctant to rely on a provision
that is premised on a concern about the completeness of the exposure
database as supporting an interpretation of another portion of the statute that
would bar OPP from taking into account certain potential pesticide exposures
unless a very specific and stringent reliability test was met.
Application of the Reliability Requirement in Subsection (b)(2)(A)(ii). As
indicated, OPP's preferred interpretation of the reliability requirement in
subsection (b)(2)(A)(ii) is that is directs OPP to consider whether there is
trustworthy and reproducible information on whether there is exposure occurring
by the nondietary pathway in assessing the aggregate risk imposed by a
pesticide. If OPP concludes there is no reliable information showing exposure
by a nondietary pathway, OPP will not assume that there is nondietary exposure
to the pesticide. If OPP finds that reliable information does show exposure by a
nondietary route/pathway, OPP must take such exposure into account in
assessing the aggregate risk posed by the pesticide whether or not OPP is
able to quantify with precision the level of such nondietary exposure.
As to reliable data bearing on whether exposure occurs by a given
route/pathway, OPP believes that information can reliably demonstrate
exposure by a given route/ pathway even if OPP does not have data
documenting the magnitude of exposure levels to humans. For example, OPP
has a large body of data showing that pesticide exposure can occur when there
is residential use (e.g., insecticides are applied as a crack or crevice spray in a
dwelling or other occupied structure, applied to the lawn, etc.). Further, OPP
has compiled extensive data detailing the physical properties and
3This argument becomes more problematic when it is considered that IWG also argues that the
children's safety factor cannot exceed 10X. IWG 10X Comments at 10. Thus, if OPP had only unreliable
data on dietary exposure and OPP had reason to believe that such data was likely to understate exposure
by greater than tenfold, OPP would still be limited to no more than use of a 10X safety factor to protect
infants and children.
144
-------
characteristics of those pesticides that potentially may result in human exposure
under this use scenario. Thus, where the physical properties and characteristics
of a specific pesticide, when considered in light of the generic data OPP has on
pesticide exposure in nonoccupational residential settings, show that it is likely
that the presence of that pesticide will result in human exposure if used in under
a given scenario, OPP would have reliable information showing such nondietary
exposure. (See also the additional discussion on models and the reliability
requirements below.)
IWG expresses the concern that OPP's interpretation "creates the real and
scary possibility that OPP will feel free to conclude, without any good reason,
that high levels of nonfood residues might occur, and that OPP will then feel free
to include those speculative, high-level residue values in aggregate exposure
calculations." IWG Letter at 4. OPP does not see the basis for such a
conclusion. OPP made clear in the Aggregate Exposure Response to
Comments document that the "reliable information" requirement mandated that
OPP have accurate and reproducible information that nondietary exposure was
occurring prior to taking such exposure into account in aggregate exposure.
Nothing in this interpretation would permit OPP to make unreasonable
conclusions about the level of nondietary exposure in tolerance actions.
Similarly, OPP does not believe that any aspect of its approach to the children's
safety factor provision would sanction OPP reliance on unreasonable or
speculative exposure assessments.
The Reliability Requirement in Subsection (b)(2)(A)(ii) and the Children's
Safety Factor. Once the limited nature of the reliability requirement in
subsection (b)(2)(A)(ii) is recognized, OPP can agree with IWG that if OPP
does not have reliable information showing that exposure is occurring by a
nondietary route/pathway, OPP generally should be able to conclude that as to
nondietary exposure OPP has sufficient reliable data to ensure the safety of
infants and children in regard to the tolerance in question. Although OPP is not
certain that the gatekeeper role the "reliable information" requirement plays for
aggregate exposure applies with equal effect in section (b)(2)(C), as a practical
matter, under OPP's approach it is unlikely to make a significant difference.
OPP's major difference with IWG in regard to the children's safety factor
provision turns, once again, on what constitutes "reliable information" showing
that exposure is occurring by a nondietary route/pathway. IWG asserts that such
a showing cannot be made absent a database demonstrating actual, real-world
exposure levels and the distribution of those levels sufficient to conduct a
probabilistic risk assessment. As explained above, OPP believes that the
reliable information threshold is crossed once it can be shown that exposure is
more likely to occur than not by the route/pathway in question, whether or not the
145
-------
information can precisely define the level or distribution of exposure. The effect
of IWG's interpretation of the "reliable information" requirement combined with
its assertion that this requirement serves as a strict gatekeeper to what
exposure scenarios can be considered in the "reliable data" inquiry in the
children's safety factor is that, for all practical purposes, the children factor
provision's concern with the completeness of the exposure database is
obviated for nondietary exposure. Under the IWG's approach, it is difficult to
imagine a scenario that would justify retaining an additional safety factor due to
concerns regarding the database on nondietary exposure. If the database
contains "reliable information" on the nondietary exposure-meaning in IWG's
words "a reasonable estimate of the actual, real-world level of exposure to the
pesticide...including] information on the distribution of the exposure, so that
probabilistic estimates of aggregate exposure can be made"-it would seem
that the nondietary exposure database is complete and therefore no additional
safety factor is warranted with regard to nondietary exposure. If the database
does not reliably address the magnitude or duration of exposure, then "reliable
information" requirement would bar consideration of nondietary exposure as a
potential concern bearing on the completeness of the database with regard to
the need for a additional safety factor. Instead, the IWG approach would
suggest that "reliable data" support removing the additional safety factor. In
effect, under the IWG approach, the absence of "reliable information" on the
magnitude or distribution of nondietary exposure necessarily means that there is
"reliable data" on nondietary exposure such that the additional safety factor for
the protection of children is unnecessary, and can be removed, so far as
concerns any risk to infants and children from nondietary exposure.
In its latest comments on this issue, IWG takes exception to this
characterization of its position. IWG claims that this conclusion was neither
explicit nor implicit in its comments. According to IWG, "[w]e said-clearly-that if
there is not reliable information on a no-food exposure route, that route should
not be included in aggregate exposure and thus should not be regulated under
section 408 until better information was available." Jd- In its letter, IWG now
allows that if the "reliable information" test is met for nondietary exposure (i.e.,
there is reliable information on the actual level and distribution of exposures),
there could be circumstances where "there still might be some need for an
additional uncertainty factor." IWG asserts, however, that its prior comments
were not addressing this issue but "were pointing out our very urgent concern
that OPP's approach would allow, and even require, the inclusion of gross
overestimates of nonfood exposure in tolerance reassessments...in order to
avoid the possibility of having to use an exposure uncertainty factor." Jd-
After carefully rereading all of IWG's comments on OPP's aggregate
exposure and children's safety factor policies, IWG's Roadmap paper, and
146
-------
IWG's recent letter, OPP believes that it has fairly characterized IWG's
approach to "reliable information," aggregate exposure and the children's safety
factor provision. IWG may not have used the same words OPP did in
describing IWG's position but the clear effect of IWG's interpretation of "reliable
information" on aggregate exposure leads precisely to the result described
above.
IWG is quite clear that information bearing on nondietary exposure which
cannot meet IWG's interpretation of the "reliable information" requirement
(because it does not provide information on the "actual level" or "distribution" of
pesticide residues) is excluded from the FQPA's "aggregate exposure" concept
used to judge the safety of pesticide tolerances. IWG Comments at 39. IWG is
equally clear that nondietary exposure information excluded from aggregate
exposure should not be relied upon as a justification for retention of the
additional 10X safety factor for the protection of infants and children. IWG
Comments at 39-40. Moreover, according to IWG, the inquiry under the
Children's Safety Factor provision as to whether there are "reliable data" to
remove the additional safety factor should focus on "whether there is sufficient
reliable data about toxicity, about dietary exposure, and about the other kinds of
exposure that qualify for inclusion in the aggregate exposure assessment..."
IWG Roadmap at VIII-7 (emphasis added). Thus, the "reliable data" inquiry in
the Children's Safety Factor Provision, according to IWG, is applied to a
narrower set of data-the set of data defined by the "aggregate exposure"
concept and its integral test of "reliable information" for nondietary exposure
data-than the "reliable information" requirement.
In other words, IWG's position seems, essentially, to be that the issue of
nondietary exposure does not even appear on the radar screen in applying the
children's safety factor if information on nondietary exposure is judged unreliable
when OPP assesses aggregate exposure. OPP grants that IWG's approach
does not directly maintain that "unreliable information" under subsection
(b)(2)(A)(ii) is "reliable data" under subsection (b)(2)(c). But, by making
aggregate exposure and the reliable information requirement a screening
mechanism for the scope of the reliable data requirement in the children's safety
provision, IWG's approach would compel OPP to conclude that it has "reliable
data" for decisions involving the children's safety factor in situations where OPP
has concluded that it does not have "reliable information" on nondietary
exposure. In these circumstances, whether or not OPP could make the "reliable
data" finding would depend entirely on issues unrelated to nondietary exposure.
Thus, according to IWG, OPP's conclusion on the unreliability of the nondietary
exposure data necessarily means that it has "reliable data" to make a safety
finding for children as far as nondietary exposure is concerned. In essence,
OPP would be concluding that it has "reliable data" showing there is no
147
-------
nondietary exposure. Importantly, because of the high hurdle at which IWG has
pegged the "reliable information" requirement, an OPP conclusion that it has
reliable data to show no nondietary exposure is compelled in circumstances
where OPP does not have data showing actual real-world exposure levels and
the distribution of those levels but does have data definitely showing such
exposure is occurring.
The first and foremost legal difficulty with this strict linkage of the scope of
the "reliable data" inquiry under the children's safety factor provision to how IWG
has defined "aggregate exposure" is that, as explained earlier, the statute
contains no language defining "reliable information" in the demanding,
substantive manner put forward by IWG. Further, IWG's approach to aggregate
exposure and the "reliable information" requirement is at odds with Congress'
expressly-stated concerns regarding exposure in the children's safety factor
provision. According to IWG, where there is not reliable information on
nondietary exposure "that route [of exposure] should not be included in
aggregate exposure and thus should not be regulated under §408 [which
includes the children's safety factor provision] until better information is
available." IWG Letter at 6 (emphasis added); accord IWG 10X Comments at
40. Congress explicitly specified, however, that the children's safety factor is to
be added to "to take into account...completeness of the data with respect to
exposure...to infants and children." 21 U.S.C. §346a(b)(2)(C). The conflict is
clear: IWG's gatekeeper approach to aggregate exposure and the children's
safety factor provision based on its strict reading of the "reliable information"
requirement cancels, to a large degree, the congressional intent to take a more
protective approach when exposure data are missing or inadequate. Under
IWG's approach, there are likely to be numerous situations where nondietary
exposure data will be excluded from aggregate exposure on reliability grounds
but the children's safety factor will be deemed unnecessary despite the fact that
there is not complete data on nondietary exposure.
For all of the above reasons, OPP believes that: (1) it has characterized
IWG's position accurately; and (2) IWG's position on the "reliable information"
requirement and the interaction of this requirement and "aggregate exposure"
and the children's safety factor is unreasonable.
Models and Reliable information/Data. OPP objects to the IWG's claim that
OPP's models for drinking water and residential exposure will not produce
"reliable information" and NRDC's similar conclusion regarding models
generally and "reliable data." After all, any exposure estimate is a model of
some sort. It is a false dichotomy to suggest that models and data (or
information) are opposite extremes. Rather, models, as "users" of both
empirical data and assumptions based upon empirical data and informed by
148
-------
scientific judgment, allow scientists to generalize from a less than perfect data
set (and data are never perfect). For example, short of measuring the pesticide
residues in every sip of water and every bite of food as it is being consumed,
OPP must model or estimate exposure values for residues in drinking water and
food. The need for models exists whether the exposure estimate is based on
monitoring values in drinking water and food, residue values from field studies,
or data on a pesticide's properties and characteristics which are used to predict
anticipated residue levels in water and food. Monitoring data may produce a
more realistic and reliable estimate of exposure, but the reliability of any method
of estimating exposure will have to be evaluated based on what data the method
relies upon.
The IWG is more concerned with a particular drinking water model (the farm
pond model) and the residential exposure SOPs than models generally. IWG
asserts that these models cannot produce reliable information because they are
designed to overestimate exposure. In contrast, IWG notes that it "has
supported the use of models in estimating drinking water and residential
exposure in its comments on those science policy papers, although it has asked
that the models and their outputs be reasonably likely to allow decent estimates
of actual exposure if they are to be used for purposes other than screening."
IWG Letter at 5. Similarly, NRDC has expressed concern with the accuracy of
OPP models, particularly with regard to nondietary exposure.
OPP is aware of the criticisms that have been leveled at these screening
level models by IWG and continues to take steps to improve the drinking water
modeling techniques used in FQPA risk assessments (US EPA 1999j and US
EPA 1999g). As with IWG's criticisms, OPP has taken steps to address the
inadequacies identified by NRDC (USEPA 1999k). As indicated in Aggregate
Exposure Response to Comment document, OPP will continue to take under
consideration any concerns raised regarding whether its exposure models
overstate exposure. Whether any particular model, when used to estimate
exposure for a particular pesticide, produces estimates of sufficient quality for
those estimates to be quantitatively considered in an estimate of aggregate
exposure involves a case-by-case inquiry. OPP would reiterate, however, that
the inability of a model to produce a fully reliable quantitative estimate of "actual
real-world" exposure levels and the distribution of those levels would not mean
that safety evaluations can disregard reliable information showing such
exposure does, or will, occur.
The Diet Does Not Include Drinking Water Argument. Given the common,
everyday meaning of the term diet as including both food and water, IWG would
have to find some fairly explicit statutory language to support its claim that
"dietary exposure" does not include exposure form drinking water. This IWG
149
-------
cannot do.
IWG cites to subsection (b)(2)(D)(vi) describing aggregate exposure as
"including dietary exposure under the tolerance and all other tolerances in effect
for the pesticide chemical residue, and exposure from other nonoccupational
sources" as supporting its position. IWG argues that this language "surely is
susceptible of the reading that 'dietary exposure' includes precisely and only
those exposures 'under...tolerances.'" IWG Letter at 6. IWG asserts that if the
term dietary is read as including food and water, the inclusion of the term
"dietary" in this phrase is superfluous. OPP does not find this argument
convincing. Although perhaps this language can be read in the manner IWG
describes, it is not a compelled reading. Another reading is that the language is
meant to capture only that aspect of dietary exposure that occurs
"under...tolerances." IWG's protest that OPP's interpretation of dietary as
applying to both food and water renders the term dietary as meaningless in this
provision can be equally leveled at IWG's interpretation. IWG states that
"drinking water exposures definitely are not 'under...tolerances,"' and thus
appears to take the position that the only exposure under a tolerance can be an
exposure in food. IWG Letter at 6. If that is the case and the term "dietary"
means food, the inclusion of the term "dietary" in the phrase "dietary exposure
under the tolerance" adds nothing to the meaning of the phrase. In the final
analysis, the statutory language in subsection (b)(2)(D)(vi) is too opaque to
justify abandoning the commonsense reading of the term "dietary."
150
-------
IWG next argues that, because OPP has decided that it will not write
tolerances for pesticides in drinking water because drinking water is not food,
"drinking water exposures definitely are not 'under...tolerance,' and thus at least
arguably are not 'dietary' within the meaning of the statute." IWG Letter at 6.
Because this argument is based on IWG'S prior contention that the language
addressing dietary exposure under tolerances in subsection (b)(2)(D)(vi) must
be read in a certain fashion, it is not convincing for the reasons stated above.
IWG further asserts that it would have been reasonable for Congress to treat
similarly exposure to pesticides in drinking water and exposures to pesticides
from residential pesticide applications, rather than grouping drinking water
exposures with exposures to pesticides in food. OPP does not disagree with
the proposition the Congress could have reasonably chosen such an approach;
however, as noted, there is no explicit language in the statute drawing this
distinction and OPP is not convinced that anything in the statute suggests this
approach.
Also unpersuasive is the IWG's argument that because OPP and FDA have
treated drinking water as not a "food" under the FFDCA, drinking water is not
part of the diet. This argument fails to recognize that the question is not whether
water is food, but whether water is part of the diet. Furthermore, OPP and FDA
decided to interpret the term "food" as not encompassing drinking water based
on their conclusion that Congress' passage of the Safe Drinking Water Act was
an implied repeal of OPP and FDA's tolerance setting authority over pesticides
in drinking water under the FFDCA. See 44 FR 42775 (July 20, 1979).
However, here, there has been no action by Congress that would suggest that
the term "dietary" should be read in other than its dictionary sense.
Finally, IWG requests that OPP acknowledge that its interpretation of the
term "dietary" is not legally compelled and explain its policy basis for that
interpretation. Whether or not OPP's interpretation of "dietary" is legally
compelled, OPP would note that whether drinking water is part of dietary
exposure or not holds great significance only if IWG's interpretation of the
"reliable information" requirement is accepted. Because, as explained above,
OPP views IWG's interpretation of the "reliable information" requirement as
unreasonable, OPP sees no reason to go beyond commonsense interpretation
of the term "dietary."
The Difference Between Information and Data. Although NRDC claims the
statute draws a clear distinction between "data" and "information," NRDC does
not explain or elaborate on that distinction other than to state that, in the context
of drinking water exposure, data means "monitoring data" and not exposure
estimates from models. NRDC does not address the fact that OPP's drinking
151
-------
water models are based both on generic environmental and pesticide data and
empirical data on a pesticide's specific properties and characteristics.
OPP would note that the dictionary defines data and information by cross-
referencing between these terms and thus information is defined as data and
data is defined as information. See, e.g., Webster's New World Dictionary (2d
College Ed. 1976). Given this overlap, it seems unlikely Congress intended
OPP to make critical regulatory decisions by dissecting the fine distinctions
between the terms "data" and "information." In any event, even if the term "data"
is regarded somehow as only capturing some type of information derived from a
scientific study, OPP believes its models are based on information meeting this
description.
In sum, OPP disagrees with the major policy implications that both the IWG
and NRDC ascribe to the terms "reliable data" and "reliable information" based
on either a rather hyper-technical reading of the statute or little more than mere
speculation. OPP has been unable to find any legislative history, and the
commentors have cited none, that supports the notion that the use of the term
"reliable information" or "reliable data," or the use of the term "data" instead of
"information" and vice-versa, were intended to have far reaching policy
significance. OPP believes Congress' inclusion of the terms "reliable data" and
"reliable information" had a much more prosaic purpose-Congress merely
wanted to reconfirm that reliability is a necessary criterion for any data or
information, or model based on data or information, used in risk assessment
under the FFDCA.
Scope. OPP believes that there exist two more reasonable interpretations
regarding the scope of the reliability requirement in subsection (b)(2)(A)(ii).
Both of these interpretations are more consistent with the general principles of
administrative law and practice and with the other language of the statute. They
take into account both the "reliable information" requirement in subsection
(b)(2)(A)(ii) and the requirement in subsection (b)(2)(D)(i) that OPP consider
"the validity, completeness, and reliability of the available data from studies of
the pesticide chemical and pesticide chemical residue."
152
-------
The first, and OPP's preferred interpretation, is that the reliable information
requirement in subsection (b)(2)(A)(ii) is directed primarily at identifying whether
exposure occurs by a certain pathway or route (i.e., nondietary) and that the
reliability consideration in subsection (b)(2)(D)(i) more broadly insures that
exposure estimates (addressing the magnitude and distribution of exposure)
are reliable, whether that exposure is dietary or nondietary. Two reasons
support this interpretation. First, the "reliable information" requirement in
subsection (b)(2)(A)(ii) is in a clause specifically discussing routes/pathways of
exposure ("dietary exposure and all other exposures"). Second, as discussed
above, reading the reliable information requirement more broadly contradicts
the direct command of subsection (b)(2)(D)(i) (consider the reliability of data) by
implying that reliability is not a pertinent consideration as to dietary exposure
data, and does so in a manner that appears to condone arbitrary agency
decision-making. On the other hand, the more narrow (route-specific)
construction of the reliability requirement in subsection (b)(2)(A)(ii) can logically
be squared with subsection (b)(2)(D)(i). It makes sense for Congress not to
have imposed a reliability requirement on the question of whether exposure
occurs by the dietary route/pathway. After all, this statutory section addresses
setting maximum levels for pesticide residues in food, an important part of the
diet. Setting a tolerance level for a pesticide residue in food presupposes that
there will be some exposure to the pesticide through the dietary route/pathway.
For these exposure issues under section 408, reliability considerations apply
principally, if not entirely, to the question of amount of exposure.
Either of these two interpretations is more reasonable than IWG's
interpretation because they do not impute to Congress an intent to authorize
arbitrary action by an administrative agency (i.e., the agency may rely on
unreliable data). In the absence of a clearer statutory pronouncement, or at least
some support in the legislative history, OPP is unwilling to endorse an approach
that presumes such congressional intent. OPP prefers the first interpretation to
the second because it appears to be the more natural construction of the
language in subsection (b)(2)(A)(ii) and because it gives some separate
purpose for the inclusion of the reliability language in subsection (b)(2)(A)(ii).
That purpose is to direct OPP to examine whether some trustworthy information
is available to show that exposure would occur (or is occurring) by the
nondietary pathway.
The Substance of the Reliable Information Requirement. The common
meaning of "reliable information" is information that is trustworthy, or in the
scientific sense, information that is reproducible. Accordingly, OPP believes
that the reliable information requirement in subsection (b)(2)(A)(ii) simply is
designed to ensure that information considered by OPP is trustworthy and
reproducible. (OPP sees a similar role for the reliability consideration in
153
-------
subsection (b)(2)(D)(i).) IWG's argument that OPP should depart from this plain
meaning of the term "reliable information" and impute a more substantive role
for the reliability requirement is unpersuasive.
IWG argues that the general language in the subsection (b)(2)(A)(ii)
establishing a reliability criterion for nondietary exposure imposes more than
some type of reproducibility test. According to IWG, the reliable information
requirement substantively defines what the information must show and the
specificity of the information itself. IWG states that "[f]or information to be
considered reliable..., it must provide a reasonable estimate of the actual, real-
world level of exposure to the pesticide...[including] information on the
distribution of the exposure, so that probabilistic estimates of aggregate
exposure can be made." IWG's logic is as follows: (1) Congress has imposed
no reliability requirement on dietary exposure data; (2) IWG claims this was
because OPP often has data on the actual, real-world levels of pesticide
residues on food including data on the distributions of those residue levels; and,
thus, (3) there is a "strong implication" that for information on the nondietary
route/pathway to be reliable it must be comparable to the information OPP has
on dietary exposure.
Each of the three steps in this argument, however, is faulty. The first
premise-that there is no reliability requirement pertaining to dietary exposure
data-has already been shown to be untenable if asserted broadly (i.e., not just
applying to whether there is exposure by a given route), as the IWG comment
does. As noted, it is illogical to suggest that Congress removed any constraint
regarding the need for reliable information on dietary exposure data pertaining
to the magnitude and distribution of exposure. Second, IWG's claim that the
lack of a reliability requirement as to dietary exposure data is due to the nature
of the data that OPP collects on dietary exposure is nothing more than
speculation. IWG cites no authority to support this proposition. Moreover, as
noted above, there is an alternative and logical reason appearing on the face of
the statute as to why Congress might not have imposed a reliability criterion on
exposure through the dietary route- this statutory section addresses setting
maximum levels for pesticide residues in food. Given this explanation based on
the statutory structure there is no need to speculate concerning other
motivations. Finally, even if the first two steps of IWG's argument are correct
(that there is no reliability requirement pertaining to the magnitude and
distribution of dietary exposure and that dropping that requirement is due to the
quality of OPP's exposure data on food), it does not follow that data on
nondietary exposure must be comparable to food exposure data collected by
OPP. At most, there would be an implication that one type of data-data on
actual real-world levels of pesticides including information on the distribution of
residue levels-would be considered reliable. It would not preclude other data
154
-------
from meeting the reliability requirement.4
ISSUE 18. Policy or Rule
Overview. OPP requested comments on how this policy could be structured so as
to provide meaningful guidance without at the same time imposing binding
requirements on either the government or outside parties. OPP received a few
comments on this issue relating to inclusion of data requirements in regulations and the
need to define a "core" data set by regulation. OPP will also take this opportunity to
respond to a petition from pesticide manufacturer and grower groups requesting,
among other things, rulemaking that "lay[s] out the circumstances in which an additional
safety factor will be imposed" and specifies "the data that EPA may require in order to
decide whether, in establishing tolerances, an additional safety factor is needed to
protect infants and children." Petition for Rulemaking to Develop Policies and
Procedures for Implementing the Food Quality Protection Act of 1996 32 (May 22,
1998). The petition asserts that a series of EPA documents has "effectively
established guidelines for implementation of the tenfold factor...." Jd. One of the
documents cited is the draft version of 10X policy paper. The pesticide manufacturer
and grower groups have also filed a lawsuit against EPA that claims that EPA "has
consistently used its FQPA safety factor policy in a binding and unvarying manner...."
Fourth Amended Complaint, p. 33 American Farm Bureau Federation, et al. v. EPA.
Case No. 1:99CV01405 RCL (D.D.C.). Therefore, these groups assert this alleged
use of the policy is unlawful in that EPA is treating the policy as if it is a legislative rule
under the Administrative Procedures Act (APA) without complying with the APA
procedures for promulgating such rules.
Comments. IWG urged OPP to "propose and finalize 40 CFR 158 amendments to
specify Tier I toxicology and exposure data requirements, as well as higher-tier
requirements." IWG comment at 48. IWG asserted that "interested parties need a
single, readily available, authoritative source for this information...." Jd- IWG also
argued that the policy's definition of a core data set must be set out in regulations.
According to IWG, the "Administrative Procedure Act requires that test
requirements of general applicability be promulgated by rulemaking." Jd- at 47.
4IWG does not address the difficult interpretation raised by such an approach concerning
how OPP is to decide what nondietary data is comparable to dietary exposure data. Despite
IWG's claims to the contrary, OPP has several gradations of data on actual, real-world pesticide
residue levels in food. For some pesticides, OPP has full-blown studies from retail markets; in
other cases, it may have varying amounts of monitoring data; and, in many cases, it may have
only data from the crop field trials. The amount of distributional data OPP has on pesticide
residues in food is also variable.
155
-------
Petition. The pesticide manufacturer/grower petition requested that the Agency
undertake rulemaking on a number of topics including implementation of the
children's safety factor and the data requirements relevant to making safety factor
decisions. The petition urged that specifying data requirements relevant to the
safety factor decision was important because it gave pesticide registrants notice as
to what data needed to be prepared to avoid EPA finding the gaps in the toxicology
database necessitated retention of the children's safety factor.
The petition also lists various generic policy and legal reasons for issuing rules
regarding FQPA implementation. The policy reasons include: (1) a rule provides
greater transparency because the Notice and Comment process will provide formal
notification of EPA's views; (2) rulemaking will give all parties a chance to
participate in the development of policy not just those invited to Agency advisory
committees; (3) in a rulemaking EPA must respond to public comments on the
public record and must provide a concise statement of the basis and purpose for
the rule; (4) a rule provides certainty and stability because rules are subject to
judicial review and legal issues can be resolved once and for all; (5) the advisory
committee process and SAP review of policies has not adequately provided for
public participation; and (6) rulemaking on individual tolerances has not been an
adequate substitute for generic rulemakings. The legal reasons listed in the petition
include: (1) that FQPA policies 'impose obligations' and have 'significant effects
on private interests' and thus are, in fact, legislative rules requiring Notice and
Comment procedures; and (2) the FQPA "requires EPA to use Notice and
Comment rulemaking to establish general requirements or procedures for
implementing the key provisions of the FQPA." Pet. at 15
Legal Challenge. In the course of the American Farm Bureau Federation lawsuit,
the pesticide manufacturer and grower industry plaintiffs have cited portions of the
draft Safety Factor Policy that they consider to impose binding requirements.
Specifically, the plaintiffs state:
The Safety Factor Policy "describe[s] the policies employed by the Office of
Pesticide Programs in making a determination regarding the FQPA Safety Factor
when developing aggregate risk assessments and regulatory decisions for single
active ingredient pesticides." Safety Factor Policy at 10 (emphasis added). The
Policy goes on to "describe the factors/issues related to exposure assessment and
the completeness of the exposure database that must be considered when making
an FQPA Safety Factor finding." Jd- at 45 (emphasis added); see also jd- at 45-53.
EPA's default position is that a database uncertainty factor will always be applied
when the toxicology database lacks one or more of [certain] types of studies. Jd- at
55 (emphasis added). Under the Safety Factor Policy, "the absence of detailed
and specific exposure data would require the application of an additional safety
factor unless OPP can determine that the available data and its assessment
156
-------
methodologies give a high degree of confidence that exposure to infants and
children is not underestimated." Jd- at 8 (emphasis added).
Notice and comment are also required because the two Science Policies make
significant changes in prior EPA practice and policies. It is axiomatic that an
agency's change in existing policy constitutes a legislative rule requiring Notice and
Comment, [cites omitted]
The Safety Factor Policy likewise marks a dramatic change in EPA policy. As the
Policy itself indicates, it "for the first time, addresses the question of how additional
safety factors should be applied in situation where a toxicology database is
considered incomplete given changes in data requirements." FQPA Safety Factor
Policy at 19. [emphasis added by plaintiffs]. The Policy further states that it
effectuates "several changes in [EPA's] approach to the assessment of the
completeness of the toxicology database" and acknowledges that "application of
these criteria leads OPP to expand immediately the scope of the core database it
has historically considered." Jd- at 29-30.
EPA's assertion that "the plain language of the policies makes clear that EPA does
not intend to bind itself is demonstrably false. The policies themselves contain no
such indication. The general disclaimer EPA cites is found only in the Federal
Register Notices, not the policy papers themselves.
Response. The comments, petition, and lawsuit present two overall concerns.
First, commenters urge OPP to update its 40 CFR 158 data requirements.
OPP agrees that these regulations should be amended and plans to propose
amendments to the rule soon. Second, the pesticide manufacturing industry and
growers have argued that OPP must promulgate its children's safety factor
policy as a rule. The remainder of this response addresses this question.
157
-------
After considering the comments, the petition, and the arguments raised in
the lawsuit, OPP has decided to issue the Safety Factor Policy as an
interpretive rule and nonbinding policy guidance, not as a binding legislative
rule. Accordingly, OPP denies the petition from the pesticide manufacturers
and growers to the extent it sought rulemaking regarding this policy.
Several of the matters addressed in the policy concern legal questions.
Because these legal issues were resolved through interpretation of the statute,
OPP believes it is authorized under the APA to speak to these questions
through the means of an interpretive rule rather than a legislative rule. See
General Motors Corp. v. Ruckelshaus. 742 F.2d 1561, 1565 (D.C. Cir. 1984)
(en banc), cert, denied, 471 U.S. 1074 (1985). Examples of the interpretive
issues in the policy include questions such as what safety factors are the
children's safety factor intended to be in addition to and what constitutes reliable
information or data.
The bulk of the policy document does not involve interpretational issues but
rather describes OPP's views regarding the science and policy considerations
that bear on its decisions regarding the children's safety factor. OPP describes
in great detail the considerations that pertain to each of the three reasons the
statute enunciates for application of an additional 10-fold safety factor. For the
most part, these considerations involve science judgments of high complexity.
Certainly, final decisions which involve a weighing of all three factors require
OPP decision-makers to take into account a wide array of considerations.
Because safety factor decisions involve such a complex of factors, many related
to difficult science assessment issues, OPP has decided that this policy should
not be cast in the form of a binding legislative rule. There is simply no adequate
way to capture the multifaceted considerations involved in safety factor
decisions in a rigid rule format. Moreover, because the science issues that form
the basis for so many of the considerations are rapidly evolving as more and
more data becomes available, any rule-like formulation would either quickly be
outdated or would necessarily be so heavily caveated as to serve little purpose.
Nonetheless, OPP does not believe that decisions on the children's safety factor
should be made in a vacuum. Adopting this policy serves the important goal of
helping to insure consistency in OPP decision-making by regularizing, to a large
degree, the process and considerations that will be used in decision-making.
At the same time, using a policy, as opposed to a rule, to accomplish this goal,
leaves OPP with the flexibility to react to rapidly changing scientific
developments.
So that there is no misunderstanding regarding the nonbinding nature of the
policy, OPP has included in the introduction to the policy a clear explanation of
the manner in which the policy will be used and of the opportunity for affected
158
-------
parties to assert their views regarding the proper considerations in safety factor
decision-making.
This guidance document provides OPP's current thinking on application of
the provision in FFDCA section 408(b)(2)(C), regarding an additional safety
factor for the protection of infants and children. As such, it is intended to provide
guidance to OPP risk assessors to facilitate consistent implementation by OPP
of the children's safety factor provision and to increase understanding of OPP
actions by regulated entities and the public. Importantly, this guidance
document is a policy statement and not a legislative rule and thus is not binding
on OPP or on outside parties. It does not predetermine any pesticide-specific
decision regarding the children's safety factor. OPP remains free to take
actions that vary from the guidance provided in the document. For example,
OPP may deviate from the document based on developments in science or risk
assessment methodologies or changes in policy approach. Any such action
would be accompanied by an explanation for OPP's decision. Similarly, the
regulated community and the public retain the right to object both to the manner
in which the guidance document is applied to specific pesticides as well as to
the policy considerations underlying the guidance document. Such objections
could address any factual, scientific, policy, or legal conclusions or
interpretations in the guidance document. If such objections are persuasive,
OPP will be guided by them in the specific decision at hand and also modify the
policy, as appropriate.
To facilitate consistent decision-making, OPP staff should consider this
guidance document in all actions involving the additional children's safety factor.
OPP staff are cautioned, however, that, because this document is a guidance
policy and not a binding rule, they must consider the merit of all contentions from
outside parties regarding application of the children's safety factor to specific
pesticides. Should staff believe, for whatever reason, that action at variance
from this guidance document should be taken, that recommendation should be
flagged so that it can receive the full consideration of OPP decision-makers.
This position is consistent with the manner in which the Agency generally
approaches complex risk assessment issues and has resolved questions
regarding other science policies under the FQPA. Thus, EPA's views on major
risk assessment topics have been issued as policy guidances not binding rules.
See e.g., Guidelines for Carcinogen Risk Assessment (USEPA 1986; 51 FR
33992); Guidelines for Reproductive Toxicity Risk Assessment (USEPA
1996a; 61 FR 56274); Guidelines for Exposure Assessment {USEPA 1992;
57 £R 22888); and Proposed Guidelines for Carcinogen Risk Assessment
(USEPA 1996b; 61 FR 17960). Similarly, OPP's FQPA policy addressing the
selection of the population percentile used in calculating the threshold of
159
-------
regulatory concern in acute risk assessments was issued as a policy not a rule.
(See EPA 2000c). In their petition, the pesticide manufacturers and growers
cited to one EPA proposed rule that included "models and assumptions for
estimating public exposure" concerning certain air emission standards. See 59
Fed. Reg. 15504 (April 1, 1994). However, OPP would note that when that rule
was finalized, the portions addressing risk assessment were omitted. 61 Fed.
Reg. 68384 (December 27, 1998).
OPP found none of the arguments set forth in the rulemaking petition from
the pesticide manufacturers and growers to be persuasive. Each of those
arguments is addressed in turn below.
Transparency. The petition argued that a rule would provide greater
transparency because there would be formal notification of all parties
concerning the rulemaking. However, this formal notification concern was met
by the procedure OPP followed in developing this policy. OPP published Notice
of the draft policy in the Federal Register (64 FR 37001). That Notice provided
a concise summary of the policy and requested public comment on the policy.
Further, OPP put a full copy of the policy on its Internet Web site and generally
made copies available to the public.
Public Participation. The petition argued that a rulemaking would allow all
affected parties to participate not just advisory committee members. That
concern, however, has also been met by OPP's public comment process. As
noted, OPP received over 800 comments on this draft policy.
Response to Comments. The petition expressed a concern that without a
requirement to respond to comments and to provide a statement of the basis
and purpose for the policy, OPP would not in fact produce such documents.
OPP, however, believes that its policy document clearly articulates the basis
and purpose of the policy and that this Response to Comments document has
adequately addressed all significant comments.
Judicial Review. The petition argued that a rule provides certainty and stability
because unlike a policy document it would be subject to judicial review.
Generally, policy statements are not reviewed as ripe for review until they have
been applied to a concrete regulatory action. Similarly, generic rules are often
found unripe on the same grounds. On occasion, courts will review a generic
rule in the absence of a concrete application of the rule where a challenge to the
rule presents purely legal questions and there would be hardship to the
challenger in delaying review. This policy does include several legal
interpretations of the FQPA. However, OPP believes that many of these
interpretations are of the variety that judicial review of the interpretation would
160
-------
benefit from application of the interpretation in a concrete context (e.g., OPP's
determination of whether specific data constitutes "reliable data"), and, thus,
whether the interpretations are included in the policy as interpretive rules or
promulgated as binding legislative rules, is likely to have little effect on their
reviewability. Moreover, to the extent the policy contains legal interpretations,
those legal interpretations are arguably subject to judicial review if ripe.
Codifying the interpretations does not necessarily affect their ripeness. Thus,
this consideration does not appear to strongly support issuance of the policy as
a rule.
Advisory Committee Process and SAP Review. The petition claimed that
Agency attempts to get outside input into its policies through various advisory
committees and the FIFRA SAP have been inadequate. OPP believes the
advisory committee process and SAP review have provided important input.
However, to the extent these processes have provided only a limited forum for
public participation, the Notice and Comment process for the policy has
addressed any such concern.
Individual Tolerance Rulemakings. The petition argued that OPP has not
opened its policies up for comment in rulemakings addressing individual
tolerances. The petition also implies that application of OPP policies in the
context of such tolerance actions is not subject to judicial review. Pet. at 24.
Although OPP has not specifically requested comments on its policies in
tolerance actions, such comments would certainly be appropriate to the extent
the policy formed part of the basis for OPP's decision. Moreover, the petition is
clearly incorrect if it is suggesting that the lack of an explicit request for comment
on policies underlying a specific tolerance decision somehow insulates the
policy's application from administrative and judicial review.
Similarly, OPP found none of the legal reasons contained in the lawsuit filed
by associations representing pesticide manufacturers and growers to have
merit.
Policies Impose Obligations. The industry associations argued that FQPA
policies generally and the Children's Safety Factor Policy specifically impose
obligations and have significant effects on regulated parties and thus these
policies are binding rules and must be promulgated following Administrative
Procedure Act (APA) requirements. OPP has attempted to make clear that the
Children's Safety Factor Policy does not impose binding obligations on either
regulated parties or the government both in the policy document and in this
response to comments. Further, OPP does not believe that the policy itself has
significant effects on regulated parties because it does not impose any rights or
obligations. Rather, the policy provides nonbinding guidance on one aspect of
161
-------
pesticide risk assessment. It is true that the considerations in the policy when
taken into account in an individual risk assessment for a pesticide may affect
the ultimate decision on that pesticide. In this regard, this policy is similar to
other policies and numerous decisions not covered by policies that can affect
the outcome of a risk assessment. Yet, OPP has made clear that it is not the
Children's Safety Factor Policy that decides what safety factor is appropriate
for a given pesticide. The considerations discussed in the policy remain open
to question from interested parties. Moreover, OPP does not believe that any
policy or other decision that may have a substantial affect on a risk assessment
must, because of that potential effect, be promulgated as a rule. If this were
true, OPP could only issue risk assessment policies addressing insignificant
matters. Courts addressing disputes regarding whether agencies have treated
policies as rules have not envisioned such a narrow role for policy statements.
For example, in Community Nutrition Institute v. Young. 818 F.2d 943 (D.C. Cir.
1987), the DC Circuit noted:
Our holding today [that FDA's action levels have
been treated by FDA as binding] in no way indicates
that agencies develop written guidelines to aid their
exercise of discretion only at the peril of having a court
transmogrify those guidelines into binding norms. We
recognize that such guidelines have the not
inconsiderable benefits of apprizing the regulated
community of the agency's intentions as well as
informing the exercise of discretion by agents and
officers in the field. It is beyond question that many
such statements are nonbinding in nature and would
thus be characterized by a court as interpretative rules
or policy statements.
Jd- at 949. Finally, it is worth noting, that a risk assessment, is not itself a
regulatory action. Although a risk assessment may serve as the basis for a
tolerance rulemaking under the FFDCA, a risk assessment does not have
formal legal consequences. A guidance for conducting some aspect of a risk
assessment is even further removed from regulatory consequences.
162
-------
FQPA Requirement for Rulemaking. The petition claimed that section
408(e)(1 )(C) requires that general procedures for implementing section 408
must be promulgated as rules. The language of section 408(e)(1)(C), however,
is clearly permissive-"EPA may issue a regulation..." (emphasis added). This
language authorizes OPP to establish rules for "general procedures and
requirements to implement this section;" it does not mandate such rules.
Binding Language in the Children's Safety Factor Policy. OPP has made
several changes in the document in response to arguments in the litigation
regarding binding language. First, and most importantly, OPP agrees with the
pesticide and grower associations that OPP should include in the policy itself,
as opposed to the Federal Register Notice soliciting comment on the draft
policy, an explicit statement regarding the nonbinding nature of the policy. The
policy now contains a detailed caution to agency personnel in this regard and an
explicit invitation to affected parties to comment on individual decisions.
Second, the associations cited two instances where inappropriate mandatory
language was used. The first of these appeared to make it mandatory that OPP
staff consider certain factors in making the safety factor decision. DRAFT SFP
at 45 ("the factors/issues...that must be considered when making a FQPA
Safety Factor Finding"). OPP believes it is important, that for consistency's
sake, agency reviewers, in general, take into account similar considerations in
making safety factor determinations. Yet, OPP recognizes that not every listed
consideration will apply in every instance and, in some cases, factors not
mentioned in the policy should be examined. Further, OPP would reiterate that
since this is a policy, it does not carve into stone the listed considerations as
definitively relevant to the safety factor determination. Interested parties remain
free to argue to the Agency that a consideration mentioned in the policy is
inappropriate in general for safety factor determinations. OPP has amended
the language of the policy in an attempt to capture these concepts. The second
mandatory statement pertained to the use of a safety factor where core
toxicology data are missing. The draft policy stated that the "default position" is
that a factor "will always" be applied in those circumstances. Jd- at 55. This
framing of a default position could be misleading and so OPP has amended the
language to make clear that the default position in these circumstances is, in
fact, a true default position-i.e., an initial position that is subject to change after
consideration of case-specific factors.
Two other statements cited by the associations, OPP concluded, were not
problematic. First, the associations seemed troubled because the draft policy
stated that it described policies that "employed" by the OPP in making safety
factor decisions. OPP believes that it is important for consistency's sake that
OPP staff make safety factor decisions in the context of the safety factor policy.
Because the policy guides discretion without narrowly cabining it, OPP believes
163
-------
the policy is not a rule, and that it can require its staff to not abandon the context
of the policy without providing an adequate explanation for the alternate
approach. Second, the associations cited as problematic the following
statement: "the absence of detailed and specific exposure data would require
the application of an additional safety factor unless OPP can determine that the
available data and its assessment methodologies give a high degree of
confidence that exposure to infants and children is not underestimated." Jd- at 8
(emphasis added by associations). OPP believes that this is a fair
interpretation of the legal standard in section 408(b)(2)(C) and, as such, does
not run afoul of any APA constraints on legislative rulemaking.
Significant Change in Policy Requires a Legislative Rule. In their lawsuit, the
pesticide and grower associations also contend that because the Children's
Safety Factor Policy contains some variations from earlier policies on this
question, it must be promulgated as a legislative rule. OPP disagrees. The
essence of a policy statement is that it is not binding on the agency; thus, the
agency remains free to act in variance with the policy so long as it explains its
change in course. See Svncor Int'l Corp. v. Shalala. 127 F.3d 90, 94 (D.C. Cir.
1997) ("The agency retains the discretion and the authority to change its
position-even abruptly-in any specific case because a change in its policy
does not effect the legal norm."). OPP's earlier positions on application of the
children's safety factor are contained in nonbinding guidance not legislative
rules, so it is fully appropriate for OPP to revise its policy positions by use of a
policy statement, not a legislative rule.
Categorization of Data as "Core," "Routinely Required," or "Newly Required."
One commenter objected to OPP's use of categories for describing data,
stating that such categories could only by established by rule. These categories
have been dropped from the revised document.
164
-------
ISSUE 19. Additional Comments
A few comments were received which did not fit into any of the Issue categories.
They are captured here.
Determination of NOELS versus NOAELs
Comment 1. Several commentors (761, 771, 778) expressed the opinion that the
FQPA mandated the use of the NOEL, rather than the NOAEL, in establishing
reference doses; the House Commerce Committee Report, H. R. 104-669, Part 2
p. 9 (July 23, 1996) was cited in support of this position. One further assumption
leading to this conclusion was that nonadverse effects may occur at lower dose
levels than adverse effects; therefore, the NOEL will likely provide the most
sensitive endpoint and dose for pesticide regulation.
Response. OPP does not believe that the FQPA mandates that OPP, in
evaluating animal studies, use a NOEL instead of a NOAEL. The statute does
not mention either term. The legislative history does at one point use the term
NOEL but that legislative history does not indicate that Congress intentionally
used the term NOEL because it did not think it appropriate for OPP to consider
the NOAEL. H. Rept. 104-669, 104th Cong., 2d Sess. 41 (1996). In fact,
Congress appears to have assumed NOELs are NOAELs. For example, in
defining "threshold effect" Congress stated that this "is an effect for which the
Administrator is able to identify a level at which the pesticide chemical residue
will not cause or contribute to any known or anticipated harm to human health."
Jd- (emphasis added). If Congress had intended that threshold effects be based
on NOELs rather than NOAELs, it would not have used the word "harm" in
defining the effect.
Congress seems to have used the term NOEL because it was common
usage for OPP at that time FQPA was passed. However, prior to 1998, in
OPP's discussion of the hazard identification process of evaluating pesticide
toxicity, the term NOEL was used to describe the dose level at which no
significant adverse effects were noted. OPP's terminology was not consistent
with the rest of the Agency, as illustrated in EPA's Integrated Risk Information
System (IRIS). This system included more hazard terms than OPP generally
employed, including NOAEL, LOAEL, and FEL (Frank Effect Level). On
September 2, 1998, this apparent semantic inconsistency was eliminated by
HED Standard Operating Procedure (SOP) 98.3 which indicated that OPP
would commence using the terms NOAEL and LOAEL in their scientific reviews
and documents. It also stated, "In a practical sense, the terms NOEL and
NOAEL have been used interchangeably in OPP. As a general rule, OPP would
165
-------
consider as appropriate for hazard identification and risk assessment only
those effects which are adverse or potentially adverse. This inclusion of the
term NOAEL should not change any of our hazard endpoints for regulation but
add to the quality of the risk assessment."
Blue Ribbon Panel on 10X and "Reliable Data"
Comment 2. One commentor (778) stated that EPA should convene a "blue-
ribbon" panel under the auspices of the Children's Health Protection Advisory
Committee to assist in determining when there are "reliable data" on which to base
a decision to use anything other than the additional 10X margin of safety mandated
by FQPA.
Response. OPP agrees with the commentor that independent, external,
scientific peer review of its policy on the FQPA Safety Factor provision can add
significant value. But, the Agency does not agree with the particular
recommendation that an external peer review group separate from the FIFRA
SAP should be convened to consider this policy. The Agency has already
obtained extensive external peer review from its legislatively-mandated advisory
group. OPP believes that it is appropriate to continue using this established
advisory group which provided the earlier reviews.
OPP has consulted substantively with the FIFRA Scientific Advisory Panel
on four occasions about its interpretation and application of the FQPA Safety
Factor provision: October 29, 1996; March 25, 1998; July 30, 1998 (update
only); December 8, 1998; and May 25, 1999. The most recent SAP consultation
focused on drafts of the documents made available for public comment by the
Federal Register Notice of Availability published on July 8, 1999. The SAP's
comments appear in its final report on the meeting and are available at:
http://www.epa.aov/scipolv/sap/1999/mav/final.pdf). (These comments are also
addressed in this comment and response document.) The Agency also notes
that, shortly after publication of the Federal Register Notice of Availability, OPP
provided the Science subcommittee of the Children's Health Protection
Advisory Committee (CHPAC) an extensive briefing on the policy and
supporting documents. Neither the CHPAC nor any of the individual members
of the Science subcommittee submitted written comments on the FQPA Safety
Factor Policy or accompanying documents.
The Agency believes that it is appropriate to continue to use its SAP to
address new and unresolved science issues relating to the implementation of
the FQPA Safety Factor provision. Future consultations can build on the strong
foundation of knowledge and understanding achieved through previous
interactions.
166
-------
REFERENCES
Babich and Davis. 1981. "Food Tolerances and Action Levels: Do They Adequately
Protect Children"? Biosciences 31:429-438.
Klaassen, C.D.; M.O. Amdur; J. Doull (eds). 1995. Casarett and Doull's Toxicology: The
Basic Science of Poisons; 5th Edition. MacGraw-Hill: New York.
Chapin, R.E. and J.J. Heindel (eds). 1993. Methods in Toxicology: Male Reproductive
Toxicology, Vol. 3. Academic Press, Inc.: New York, p. 2.
Chengelis, C.P.; J.F. Holson; and SC. Gad (eds.). 1995. Regulatory Toxicology. Raven
Press: New York; p. 6.
Dourson, M.L.; S.P. Felter; and D. Robinson. 1996. "Evolution of science-based
uncertainty factors in noncancer risk assessment." Regul. Toxicol. Pharmacol.24:108-
120.
Goldey, E.S.; H.A. Tilson; and K.M. Crofton. 1995. "Implications of the use of neonatal
birth weight, growth, viability, and survivial data for predicting developmental neurotoxicity:
A survey of the literature." Neurotox. Teratol., 17:313-332.
Hayes, W.A. 2000. Principles and Methods of Toxicology. Taylor & Francis, Inc.
Kimmel C.K., et al. 1990. "Workshop on the qualitative and quantitative comparability of
human and animal developmental neurotoxicity: summary and implications." Neurotoxicol
Teratol. 1990 May-Jun;12(3):285-92.
National Research Council (NRC). 1983. Risk Assessment in the Federal Government:
Managing the Process, Committee on the Institutional Means for Assessment of Risks to
Public Health, Commission on Life Sciences, NRC, National Academy Press:
Washington, DC.
National Research Council (NRC). 1993. Pesticides in the Diets of Infants and Children;
Committee on Pesticides in the Diets of Infants and Children; Board on Agriculture and
Board on Environmental Studies and Toxicology, Commission on Life Sciences, National
Research Council, National Academy Press: Washington, DC. Available:
http://www. nap, edu/cataloa/2126. htm I
National Research Council (NRC). 1994. Science and Judgment in Risk Assessment,
Committee on Risk Assessment of Hazardous Air Pollutants, Commission on Life
Sciences, National Research Council, National Academy Press: Washington, DC.
167
-------
Nusser, S. M. 1994. Estimating Usual Intake Distributions for 24-hour Food Intake Data
Containing Many Zero Intakes; Proceedings of the American Statistical Association
Annual Meeting, August 13-14, 1994.
Renwick, A.G. 1998. "Toxicokinetics in Infants and Children in Relation to the ADI and
TDI." FoodAdditiv. Contamin. 15 (Suppl.): 17-35.
Renwick, A.G. and N.R. Lazarus. 1998. "Human variability and noncancer risk
assessment-an analysis of the default uncertainty factor." Regul. Toxicol. Pharmacol.
27:3-20.
Rice, D.C., et al. 1996. "Lessons for Neurotoxicology from Selected Model Compounds:
SGOMSEC Joint Report." Environmental Health Perspectives Supplement. 104(2): 105-
15.
U.S. Environmental Protection Agency. 1986. "Guidelines for Carcinogen Risk
Assessment;" September, 1986. EPA/630/R-00/004. National Center for Environmental
Assessment, Office of Research & Development, Washington, DC.
51 FR 33992. Available: http://www.epa.aov/ncea/raf/car2sab/auidelines 1986.pdf
U.S. Environmental Protection Agency. 1991. "Guidelines for Developmental Toxicity
Risk Assessment;" December 1991. EPA/600/FR-1/001. National Center for
Environmental Assessment, Office of Research & Development, Washington, DC.
56 FR 63798. Available: http://www.epa.aov/ncea/raf/pdfs/devtox.pdf
U.S. Environmental Protection Agency. 1992. "Guidelines for Exposure Assessment;"
May 29, 1992. EPA/600Z-92/001. National Center for Environmental Assessment, Office
of Research & Development, Washington, DC. 57 FR 22888. Available:
http://www.epa.aov/NCEA/exposure.htm
U.S. Environmental Protection Agency. 1994. "Methods for Derivation of Inhalation
Reference Concentrations and Application of Inhalation Dosimetry;" October 1994.
EPA/600/8-90/066F. Office of Research and Development. Washington, DC. Available:
http://www. epa.aov/clariton/clhtm l/pubtitle. htm I
U.S. Environmental Protection Agency. 1996a. "Guidelines for Reproductive Toxicity Risk
Assessment;" October 1996. EPA/630/R-96/009. National Center for Environmental
Assessment, Office of Research & Development, Washington, DC.
61 FR 56273. Available: http://www.epa.aov/nceawww1/raf/pdfs/repro51.pdf
168
-------
U.S. Environmental Protection Agency. 1996b. "Proposed Guidelines for Carcinogen
Risk Assessment;" April 1996. EPA/600/P-92/003C. National Center for Environmental
Assessment, Office of Research & Development, Washington, DC. 61 FR17960.
Available: http://www.epa.aov/NCEA/raf/pdfs/propcra 1996.pdf
U.S. Environmental Protection Agency. 1997. Memorandum from James N. Rowe to
Margaret Stasikowski. "Revised 10X Policy Paper with SAP Response/Comments;"
March 10, 1997. Office of Pesticide Programs, Office of Prevention, Pesticides, and
Toxic Substances. Washington, DC.
U.S. Environmental Protection Agency. 1998a. "Guidelines for Neurotoxicity Risk
Assessment;" April 1998. EPA/630/R-95/001F. National Center for Environmental
Assessment, Office of Research & Development, Washington, DC. 60 FR 26926.
Available: http://www.epa.aov/ncea/nurotox.htm
U.S. Environmental Protection Agency. 1998b. Draft Document "Guidance for
Submission of Probabilistic Human Health Exposure Assessments to the Office of
Pesticide Programs;" November 4, 1998. Office of Pesticide Programs, Office of
Prevention, Pesticides, and Toxic Substances, Washington, D.C. Available:
http://www.epa.gOv/fedrastr/EPA-PEST/1998/November/Dav-05/p29665.htm
U.S. Environmental Protection Agency. 1998c. Draft Document. "A Retrospective
Analysis of Twelve Developmental Neurotoxicity Studies Submitted to the USEPA Office
of Prevention, Pesticides, and Toxic Substances (OPPTS);" November 12, 1998. Office
of Pesticide Programs, Office of Prevention, Pesticides, and Toxic Substances.
Washington, DC. Available: http://www.epa.aov/scipolv/sap/1998/december/neuro.pdf
U.S. Environmental Protection Agency. 1998d. Final Report. "Scientific Advisory Panel
(SAP) March 1998 Meeting: A Set of Scientific Issues Being Considered by the Agency in
Connection with the Use of FQPA 10x Safety Factor to Address Special Sensitivity of
Infants and Children to Pesticides;" no date. FIFRA Scientific Advisory Panel, Office of
Science Coordination and Policy, Office of Prevention, Pesticides and Toxic Substances.
Washington, DC. Available: http://www.epa.gOv/scipolv/sap/1998/march/fqpa 10x.htm
U.S. Environmental Protection Agency. 1998e. HED Status Update to the Report of the
Scientific Advisory Panel on the FQPA Safety Factor for Infants and Children. July 30,
1998. Available: http://www.epa.oov/scipolv/sap/1998/julv/session5.htm
169
-------
U.S. Environmental Protection Agency. 1998f. Final Report of the December 8, 1998
FIFRA Scientific Advisory Panel Meeting. "Il-A Set of Scientific Issues Being Considered
by the Environmental Protection Agency Regarding: A Retrospective Analysis of
Developmental Neurotoxicity Studies;" January 22, 1999. SAP Report 99-01B. FIFRA
Scientific Advisory Panel, Office of Science Coordination and Policy, Office of Prevention,
Pesticides and Toxic Substances. Washington, DC. Available:
http://www.epa.gOv/scipolv/sap/1998/december/final.htm
U.S. Environmental Protection Agency. 1999a. Draft Document. "Toxicology Data
Requirements For Assessing Risks of Pesticide Exposure to Children's Health;" April 29,
1999. Report of the Toxicology Working Group of the Agency 10X Task Force.
Washington, DC. Available: http://www.epa.gOv/scipolv/sap/1999/mav/10xtx428.pdf
U.S. Environmental Protection Agency. 1999b. Draft Document. "Exposure Data
Requirements for Assessing Risks from Pesticide Exposure of Children;" March 8, 1999.
Washington, DC. Available: http://www.epa.goV/scipolv/sap/1999/mav/10xdoca3.pdf
U.S. Environmental Protection Agency. 1999c. Draft Document. "The Office of Pesticide
Programs' Policy on Determination of the Appropriate FQPA Safety Factor(s) for Use in
the Tolerance-Setting Process;" May 10, 1999. 64 FR48617. Office of Pesticide
Programs, Office of Prevention, Pesticides, and Toxic Substances. Washington, DC.
Available: http://www.epa.gOv/scipolv/sap/1999/mav/1 Oxpoli.pdf
U.S. Environmental Protection Agency. 1999d. Draft Document. "Standard Operating
Procedures for the Health Effects Division FQPA Safety Factor Committee;" April 26,
1999. Office of Pesticide Programs, Office of Prevention, Pesticides, and Toxic
Substances. Washington, DC. Available:
http://www.epa.g0v/scip0lv/sap/l 999/mav/sop hed.htm
U.S. Environmental Protection Agency. 1999f. Review Draft. "Guidelines for Carcinogen
Risk Assessment;" July 1999. NCEA-F-0644. National Center for Environmental
Assessment, Office of Research & Development, Washington, DC. Available:
http://www.epa.gov/ncea/raf/pdfs/cancer ols.pdf
U.S. Environmental Protection Agency. 1999g. Draft Document. "Standard Operating
Procedures (SOPs) for Residential Exposure Assessments;" December 19, 1997. Office
of Pesticide Programs, Office of Prevention, Pesticides, and Toxic Substances,
Washington, DC. Available: http://www.epa.gov/pesticides/trac/science/trac6a05.pdf
170
-------
U.S. Environmental Protection Agency. 1999h. "Guidance for Identifying Pesticide
Chemicals and Other Substances That Have a Common Mechanism of Toxicity;"
January 29, 1999. 64 FR 5796. Office of Pesticide Programs, Office of Prevention,
Pesticides, and Toxic Substances. Washington, DC. Available:
http://www.epa.aov/fedrastr/EPA-PEST/1999/Februarv/Dav-05/p2781.htm
U.S. Environmental Protection Agency. 1999i. Draft Document. "Use of the Pesticide
Data Program (PDP) in Acute Risk Assessment;" May 5, 1999. 64 FR 28485. Office of
Pesticide Programs, Office of Prevention, Pesticides, and Toxic Substances.
Washington, DC. Available:
http://www.epa.aov/fedrastr/EPA-PEST/1999/Mav/Dav-26/p13034.htm
U.S. Environmental Protection Agency. 1999j. "Estimating the Drinking Water
Component of A Dietary Exposure Assessment;" November 2, 1999. 64 FR 61346.
Office of Pesticide Programs, Office of Prevention, Pesticides, and Toxic Substances.
Washington, DC. Available:
http://www.epa.gOv/fedrastr/EPA-PEST/1999/November/Dav-10/p29451 .htm
U.S. Environmental Protection Agency. 1999k. Report from Session I of the FIFRA
Scientific Advisory Panel Meeting of September 21, 1999; November 18, 1999. SAP
Report 99-05D. FIFRA Scientific Advisory Panel, Office of Science Coordination and
Policy, Office of Prevention, Pesticides and Toxic Substances. Washington, DC.
Available: http://www.epa.gOv/scipolv/sap/1999/index.htm#september
U.S. Environmental Protection Agency. 19991. Report from Session I of the FIFRA
Scientific Advisory Panel Meeting of May 25, 1999. Sets of Scientific Issues Being
Considered by the Environmental Protection Agency Regarding: Session l-Office of
Pesticide Programs Policy for the Use of the FQPA 10x Safety Factor. Report 99-03.
FIFRA Scientific Advisory Panel, Office of Science Coordination and Policy, Office of
Prevention, Pesticides and Toxic Substances. Washington, DC. Available:
http://www.epa.gov/scipolv/sap/1999/mav/final.pdf
U.S. Environmental Protection Agency. 1999m. Background Document 1 A. "Reference
Dose (RfD): Description and Use in Health Risk Assessments;" March 15, 1993. Online.
National Center for Environmental Assessment, Office of Research and Development.
Washington, DC. Available: http://www.epa.gov/iriswebp/iris/rfd.htm
U.S. Environmental Protection Agency. 1999n. Background Paper for the FIFRA
Scientific Advisory Panel. "Overview of Issues Related to the Standard Operating
Procedures For Residential Exposure Assessment;" August 5, 1999. Office of Pesticide
Programs, Office of Prevention, Pesticides, and Toxic Substances. Washington, DC.
Available: http://www.epa.gov/scipolv/sap/1999/september/resid.pdf
171
-------
U.S. Environmental Protection Agency. 2000a. Office of Pesticide Programs
Science Policy. "The Use of Data on Cholinesterase Inhibition for Risk Assessments of
Organophosphorous and Carbamate Pesticides;" August 18, 2000. 65 £R 54521. Office
of Pesticide Programs, Office of Prevention, Pesticides, and Toxic Substances.
Washington, DC. Available:
http://www.epa.aov/fedrastr/EPA-PEST/2000/September/Dav-08/p22820.htm
U.S. Environmental Protection Agency. 2000c. "Choosing a Percentile of Acute Dietary
Exposure as a Threshold of Regulatory Concern;" March 16, 2000. 65 FR 15330. Office
of Pesticide Programs, Office of Prevention, Pesticides, and Toxic Substances.
Washington, DC. Available:
http://www.epa.aov/fedrastr/EPA-PEST/2000/March/Dav-22/p7126.htm
U.S. Environmental Protection Agency. 2000d. External Review Draft. "Benchmark Dose
Technical Guidance Document;" October 2000. EPA/630/R-00/001. National Center for
Environmental Assessment, Office of Research & Development, Washington, DC.
Available: http://www.epa.aov/ncea/bnchmrk/bmds peer.htm
U.S. Environmental Protection Agency. 2000e. "The Role of Use-Related Information in
Pesticide Risk Assessment and Risk Management." August 21, 2000. Office of Pesticide
Programs, Office of Prevention, Pesticides, and Toxic Substances, Washington, D.C.
Available: http://www.epa.aov/fedrastr/EPA-PEST/2000/Qctober/Dav-11Zp25933.htm
U.S. Environmental Protection Agency. 2000f. Background Paper for the March 3, 2000
FIFRA Scientific Advisory Panel Meeting. "Development and Use of Distributions of
Pesticide Concentrations in Drinking Water for FQPA Exposure Assessments;" February
11, 2000. FIFRA Scientific Advisory Panel, Office of Science Coordination and Policy,
Office of Prevention, Pesticides and Toxic Substances. Washington, DC. Available:
http://www.epa.aov/scipolv/sap/2000/februarv/revised.pdf
U.S. Environmental Protection Agency. 2000g. Final Report of the June 6, 2000 FIFRA
Scientific Advisory Panel Meeting. "Sets of Scientific Issues Being Considered by the
Environmental Protection Agency Regarding: Session I - Consultation: National Drinking
Water Survey Design for Assessing Chronic Exposure;" September 15, 2000. SAP
Report 2000-03. FIFRA Scientific Advisory Panel, Office of Science Coordination and
Policy, Office of Prevention, Pesticides and Toxic Substances. Washington, DC.
Available: http://www.epa.aov/scipolv/sap/2000/iune/finwateronlv.pdf
172
-------
U.S. Environmental Protection Agency. 2001b. "General Principles For Performing
Aggregate Exposure And Risk Assessments;" November 28, 2001. 66 FR 59428. Office
of Pesticide Programs, Office of Prevention, Pesticides, and Toxic Substances.
Washington, DC. Available:
http://www.epa.aov/fedrastr/EPA-PEST/2001/November/Dav-28/p29386.htm
U.S. Environmental Protection Agency. 2002. "Guidance on Cumulative Risk Assessment
of Pesticide Chemicals That Have a Common Mechanism of Toxicity;" January 14, 2002.
67 FR 2210. Office of Pesticide Programs, Office of Prevention, Pesticides, and Toxic
Substances. Washington, DC. Available:
http://www.epa.aov/fedrastr/EPA-PEST/2002/Januarv/Dav-16/p959.htm
Zartarian et al. 2000. "A Modeling Framework for Estimating Children's Residential
Exposure and Dose to Chlorpyrifos Via Dermal Residue Contact and Nondietary
Ingestion." Environmental Health Perspectives, Volume108, Number 6. Available:
http://ehpnet1.niehs.nih.aov/docs/2000/108p505-514zartarian/abstract.html
173
------- |