Endocrine Disruptor Screening and Testing Advisory Committee Final Report

EDSTAC Final Report Chapter Three August 1998

Chapter Three

Conceptual Framework and Principles

EDSTAC Final Report Chapter Three August 1998

Table of Contents

I. Introduction...........................................................................................................................1

II. Description of Endocrine Disruptor....................................................................................3

III. Definition of Other Key Terms ..........................................................................................4

IV. Overview of the EDSTAC Conceptual Framework...........................................................5

V. Scope of the EDSTAC Conceptual Framework..................................................................6

VI. Where Endocrine Disruption Fits in the Broader Context ...............................................6

VII. General Principles to Guide the Development of the Endocrine Disruptor Screening

and Testing Program ................................................................................................................7

VIII. The EDSTAC Conceptual Framework ..........................................................................9

A. Obtain and Analyze Existing Information to Set Priorities ..................................................9

B. Tier 1 Screening to Detect Interactions With the Endocrine System....................................9

C. Tier 2 Testing to Determine and Characterize Endocrine Disruption.................................10

IX. Additional Components and Clarifications to the Original EDSTAC Conceptual

Framework..............................................................................................................................11

A. High Throughput Pre-Screening....................................................................................... 12

B. Alternative Means of Meeting Versus Bypassing Tier 1 Screening....................................13

1. Alternative Means to Meet T1S Information Requirements .......................................... 13

2. Bypassing T1S ............................................................................................................. 13

X. Compilation of Chapter Three Recommendations ........................................................... 14

XI. Literature Cited ................................................................................................................ 18

Figures

Figure 3.1 EDSTAC Conceputal Framework .........................................................................

EDSTAC Final Report Chapter Three August 1998

I. Introduction

In this chapter the EDSTAC presents the conceptual framework and principles that have served as

the foundation upon which all the other recommendations contained in this report have been built.

In addition to this introductory section, the chapter includes sections on: the description of

“endocrine disruptor;” definitions for several other key terms; the purpose and context for

endocrine disruptor screening and testing; an overview, scope, and general principles for the

framework; the conceptual framework itself; and a discussion of other important conceptual

agreements reached after the framework was originally developed.

The initial purpose of the EDSTAC Conceptual Framework was to lay the groundwork for future

EDSTAC recommendations and to inform, facilitate, and expedite the work of the EDSTAC

work groups. The Principles Work Group developed the original version of the framework in

early 1997, and the EDSTAC reached a tentative consensus agreement on the contents of this

chapter in May 1997. Subsequently, the document was used by the work groups and the

EDSTAC to guide their deliberations. At the time the EDSTAC agreed to the framework’s

content, members recognized the evolutionary nature of the document and agreed to revisit it, as

appropriate, throughout their deliberations.

During the EDSTAC’s deliberations, another concept was identified to be considered in the

context of the original Conceptual Framework, and other concepts that were already contained

within the framework were clarified. The new concept incorporated into the framework was the

use of “high throughput pre-screening.” The concepts further clarified relate to the scenarios

under which chemicals would be permitted to bypass Tier 1 Screening (T1S) and the

interconnectedness of these bypass scenarios with other elements of the Conceptual Framework.

These concepts are introduced at the end of the chapter and further elaborated upon in subsequent

chapters.

In its final version, the EDSTAC Conceptual Framework is intended to provide guidance to EPA

regarding development and implementation of its endocrine disruptor screening and testing

program, as well as future expansion, as appropriate, of the program. The EDSTAC believes the

Conceptual Framework it has developed, as well as the principles underlying the

recommendations it is providing to EPA for priority setting and the screening and testing program

itself, are applicable to the consideration of other hormone systems in addition to those involving

estrogen, androgen, and thyroid hormone. All of the recommendations contained herein are

premised on the principle of scientific validity.

The Conceptual Framework is summarized in the decision flowchart contained in Figure 3.1,

which shows how screens and tests are used to evaluate potential endocrine disruptors. The

structure of the Conceptual Framework was placed into tiers to illustrate how chemical

substances and mixtures can be sequentially sorted into groups that are increasingly likely to be

classified as endocrine disruptors, thereby warranting additional attention.

3 -1

EDSTAC Final Report Chapter Three August 1998

Figure 3.1

EDSTAC Conceptual Framework*

INITIAL SORTING

Obtain and Analyze Existing Data

SUFFICIENT DATA

VOLUNTARY

BYPASS

of Tier 1 Screening to

go to Tier 2 Testing

SET

PRIORITIES

For Tier 1 Screening

TIER 1 SCREENING

Detect Interaction With Endocrine System

[estrogen/androgen/thyroid]

TIER 2 TESTING

Determine and Characterize Endocrine

Disrupting Effects

HAZARD ASSESSMENT

Yes

SUFFICIENT

DATA

to go to

Hazard

Assessment

HOLD

No Further

Analysis Required

at This Time

* For a more detailed version of the initial sorting and priority setting components of this framework,

please see Figure 4.1.

3 -2

EDSTAC Final Report Chapter Three August 1998

II. Description of Endocrine Disruptor

In any emerging scientific specialty area, numerous terms and definitions are used. As examples,

environmental estrogens, environmental hormones, endocrine disrupting chemicals, endocrine

modulators, endocrine disrupters, and endocrine disruptors may all be found in recent scientific

publications. Several definitions of endocrine disruptor have also been published (e.g., Kavlock

et al., 1996; European Commission, 1996; U.S. EPA, 1997). Tattersfield et al., 1997

distinguished between endocrine disruption (reduction or enhancement of hormone levels beyond

natural bounds) and endocrine modulation (adjustment within natural bounds), but viewed the

terms as interchangeable for their purposes. This ambiguity was reflected in the EDSTAC’s

deliberations.

In its initial discussions, the EDSTAC acknowledged the Kavlock et al., 1996 definition of an

endocrine disruptor as:

An exogenous agent which interferes with the synthesis, secretion, transport, binding,

action, or elimination of natural hormones in the body which are responsible for the

maintenance or homeostasis, reproduction, development or behavior.

Certain EDSTAC members were concerned that the Kavlock et al., 1996 definition was

formulated for research, and was too open ended for regulatory operations. In May 1997, the

EDSTAC developed the following working definition:

An exogenous substance that changes endocrine function and causes adverse effects at

the level of the organism, its progeny, and/or (sub)populations of organisms.

The working definition served a useful purpose in guiding the deliberations of the EDSTAC and

its work groups. However, over time it became clear that the EDSTAC was divided regarding

the acceptability of this working definition.

Two opposing viewpoints emerged. One view held that the definition must include the term

adverse, whereas the second view held that adverse was inappropriate and should be excluded

from the definition. Proponents for including adverse reasoned that a definition should distinguish

disruption from the wide range of hormone fluctuations necessary for normal physiological

adaptation. Proponents for excluding adverse reasoned that hormone function is so sensitive to

xenobiotic challenge, that any biochemical alteration during key developmental stages above

background may lead to serious, but subtle pathology later in life or in subsequent generations. In

addition, they argued that effects not adverse for an individual may be adverse at the population

level. Both sides acknowledged that clear delineation of adverse is at times subjective and may be

open to differences in interpretation. Toxicological effects occur along a continuum from subtle

biochemical events to gross pathology. The point at which an observable effect becomes truly

adverse is therefore a judgment that may differ among individual scientists.

3 -3

EDSTAC Final Report Chapter Three August 1998

The EDSTAC acknowledges that, at this time, knowledge and experience in endocrine disruptor

toxicology do not permit the simple categorization of all endocrine effects into adverse and non-

adverse. The capacity to make this distinction will improve as understanding of the assay systems

and long-term consequences of endocrine effects increases.

In order to achieve consensus, the EDSTAC agreed to the following general description to

express the range of members’ views within the context of the proposed screening and testing

program:

The EDSTAC describes an endocrine disruptor as an exogenous chemical substance or

mixture that alters the structure or function(s) of the endocrine system and causes adverse

effects at the level of the organism, its progeny, populations, or subpopulations of organisms,

based on scientific principles, data, weight-of-evidence, and the precautionary principle.

III. Definition of Other Key Terms

The EDSTAC agreed to utilize the following definitions in this report:

“Priority setting” is defined as the collection, evaluation, and analysis of existing relevant

information to determine whether, and in what relative order of priority, chemical substances or

mixtures will be subjected to screening, testing, or hazard assessment.

“Screening” is defined as the application of assays to determine whether a chemical substance or

mixture may interact with the endocrine system.

“Testing” is defined as a customized combination of tests and endpoints designed to determine

whether a chemical substance or mixture causes endocrine-mediated adverse effects and to

identify, characterize, and quantify these effects.

“Hazard assessment,” as used in this document, includes: (1) identification of the chemical

substances and mixtures that have endocrine disruption effects, which is often referred to as

“hazard identification,” and (2) establishment of the relationship between dose and effect, which is

often referred to as “dose-response assessment.”

“Chemical substances,” as used in this document, include naturally occurring and synthetic

chemicals and elements. “Mixtures” refers to commonly found combinations of chemical

substances, including those found in the environment.

The term “functional equivalency” is used at several critical junctures in the document. The

EDSTAC defines an assay, test, or endpoint as being “functionally equivalent” to a T1S or T2T

assay, test, or endpoint when it provides equivalent information for each endpoint being studied.

For purposes of the endocrine disruptor screening and testing program, EDSTAC-recommended

assays, tests, and endpoints must be validated and standardized prior to EPA’s use of functionally

3 -4

EDSTAC Final Report Chapter Three August 1998

equivalent information. As discussed elsewhere in this document, EPA should provide clear

guidance on the use of functionally equivalent assays, tests, and endpoints prior to the

implementation of the screening and testing program.

In general, the term “weight-of-evidence” is typically used to refer to a process by which trained

professionals judge the strengths and weaknesses of a collection of information to render an

overall conclusion that may not be evident from consideration of the individual data. Further

clarification of how “weight-of-evidence” principles will be applied to the EDSTP can be found in

Chapter Five, Section IV.

IV. Overview of the EDSTAC Conceptual Framework

The EDSTAC Conceptual Framework places activities in an ordered sequence. The elements of

this sequence include: (a) priority setting, which includes the sorting and prioritization of

chemical substances and mixtures for evaluation in screening and/or testing batteries;

(b) screening to detect chemical substances and mixtures capable of acting on endocrine systems;

and (c) testing to determine, characterize, and quantify the nature of the endocrine disrupting

properties of the chemical substances and mixtures identified by prior information and/or T1S.

The Conceptual Framework contains a series of decision points. At each of these points in the

process, all available information is evaluated to determine whether and how to proceed to the

next step(s). A “weight-of-evidence” approach is commonly used to make such a determination.

Three guiding principles should be adopted in the use of such a tiered decision-making system:

• This ordered sequence should not exclude the possibility that a chemical substance or mixture

could bypass one or more tiers when information warrants such a move (e.g., sufficient

prioritization data on endocrine disrupting properties may be available to initiate Tier 2

Testing (T2T) or hazard assessment).

• If information is not adequate to determine whether a chemical substance or mixture should

move to the next tier, there should be an active process for generating the information needed

to make such a decision.

• The criteria and default assumptions for deciding whether chemical substances or mixtures

move from one tier to the next, to the degree possible, should be developed in advance of

initiating a screening and testing strategy.

3 -5

EDSTAC Final Report Chapter Three August 1998

V. Scope of the EDSTAC Conceptual Framework

The Conceptual Framework is consistent with several central issues defined in the scope of

activity for EDSTAC:

• Screening and testing should be relevant to both human health and ecological effects.

• Screening and testing should initially emphasize identifying and characterizing effects that

enhance, mimic, or inhibit estrogen, androgen, and thyroid hormone-related processes. EPA

should consider tests that detect multiple hormone interactions, address endpoints in multiple

species, and predict long-term or delayed effects. EPA should periodically revisit the scope of

this strategy to permit inclusion of additional hormone systems, animals other than

vertebrates, other hormone-mediated effects, or new screens and tests as they become

available.

• Screening and testing should be capable of evaluating both chemical substances and common

mixtures.

VI. Where Endocrine Disruption Fits in the Broader Context

Many of the effects of endocrine disruption are manifested as disease processes that are already

recognized and addressed, to some degree, in current toxicological assessments. For example,

endocrine disruption may result in cancer, neurotoxicity, or reproductive or developmental

toxicity (i.e., infertility, birth defects, etc.). It is important to realize that these issues are

interconnected. For example, some cancers have their origin in prenatal life as do some

neurological problems. In addition, there may be some endocrine disruptive effects that may not

fit clearly into any of the three more well-recognized categories.

While considering the potential adverse human health and environmental effects due to endocrine

disruption, it should also be noted that the effects of human exposures to endocrine active

chemicals are not necessarily adverse. Knowledge of the functioning of human endocrine systems

has led to the development of numerous important medical applications of therapies that operate

through chemical modulation of endocrine systems. The applications represent the positive

effects of human exposures to endocrine active agents. Examples include birth control, adjunct

therapies for prostate and breast cancer, prevention of postmenopausal osteoporosis and heart

disease, treatment of hypothyroidism, and prevention or reversal of hair loss. In addition, the

consumption of a diet high in soy (and its phytoestrogens) is thought to contribute to low breast

cancer rates in some Asian populations. It must be kept in mind that endocrine active chemicals

with beneficial effects may still lead to adverse effects under circumstances of unintended,

inappropriate, or environmental exposures.

The scope of the program recommended by the EDSTAC addresses a small portion of all possible

3 -6

EDSTAC Final Report Chapter Three August 1998

hormonal effects. The scope includes disruption of estrogen, androgen, and thyroid hormonal

activities. This group of hormones includes those for which there exists the most data, and for

which standardized assays have been developed. These hormonal systems are a very limited part

of the potential universe of endocrine systems in all animals that may be affected by chemical

exposures. It is important, as the science evolves, for EPA to be creative in developing and

including new screens for additional modes of action or to use different modeling systems that will

improve the ability to detect endocrine-mediated environmental hazards, especially for non-

mammalian species.

While estrogen, androgen, and thyroid hormones are extremely important, and it is critical to

focus initial development of screens and tests to look for effects on them, the Committee wants it

to be clear that the scope of the EDSTP is quite limited. The EDSTAC has not devised a way to

test for all possible endocrine disruptive effects, nor is the Committee addressing the many non-

hormone-mediated causes of cancer, neurological toxicity, and toxicity to reproduction or to early

life stage developmental processes. When a chemical proceeds to the “hold box” of the

Conceptual Framework, it is because the chemical tested negative for the hormone systems

assessed, not for all hormonal effects or for other mechanisms that may cause these effects.

EPA has already developed toxicological screening and testing protocols to evaluate

carcinogenicity, developmental and reproductive toxicity, and neurotoxicity. Some of the

information developed from these screening and testing activities will be useful in evaluating the

endocrine disruption potential of chemical substances and mixtures. For example, results from

developmental toxicity testing could suggest the need to undertake T1S or T2T. Similarly, results

from screening and testing related to endocrine disruption could suggest the need for

neurotoxicity, developmental, or other toxicity screening or testing. The EDSTAC recommends

that EPA examine the interrelationships between these screening and testing protocols and take

advantage of potential opportunities to streamline protocols and ensure that the results of the

related screens and tests are taken advantage of in assessing the risk of endocrine disruption.

VII. General Principles to Guide the Development of the Endocrine

Disruptor Screening and Testing Program

Several principles have guided the development of the EDSTAC Conceptual Framework, and

should guide further development of specific processes to sort and prioritize, screen, and test

chemical substances and mixtures for endocrine disruption. These principles help ensure that the

strategy of screening and testing will serve the general purpose stated above, while recognizing

that societal resources must also be allocated to sources of environmental risk other than

endocrine disruption. Thus, the screening and testing strategy should:

• require the minimal number of screens and tests necessary to make sound decisions, thereby

reducing the time needed to make these decisions;

• examine existing screens and tests for their potential to predict, detect, and/or characterize

endocrine disruptors, ensuring that any modification to existing screens and tests does not

3 -7

EDSTAC Final Report Chapter Three August 1998

compromise their ability to predict other toxicity endpoints;

• systematically examine existing screening and testing data, not only for adverse endpoints in

high dose groups, but also for physiological changes in low dose groups;

• not detract from current and new efforts to assess the toxicity of chemical substances and

mixtures through mechanisms other than endocrine disruption;

• provide data that can be used for a broad range of management and regulatory programs in a

form that supports international harmonization of the data’s use;

• include periodic review of new scientific information;

• use a performance-based approach to the selection of screens, tests, and species, including the

use of more appropriate screens, tests, and species as they are developed and validated;

• be dynamic in order to stay current with the rapidly evolving science related to the endocrine

system; and

• be conducted at a minimal cost necessary to make the decisions within the EDSTAC

Conceptual Framework.

In addition to these nine broad principles, which place screening and testing for endocrine

disruption into a larger framework of environmental risk, there are several principles specific to

the screens and tests themselves:

• To make decisions within the EDSTAC Conceptual Framework, all screens and tests should

have well-defined endpoints.

• The use of animals should be reduced to the minimal level needed to obtain scientifically valid

results and interpretations.

• The results of screens and tests should support further research on effects of endocrine

disruptors on populations, communities, and ecosystems.

• In interpreting screening and testing results, a “weight-of-evidence” approach should be used,

but should be consistent with a principle of prudence in protecting human health and the

environment. In the case of T1S, this means that a relatively higher value is placed on

sensitivity as opposed to specificity. The goal is to minimize false negative results while also

ensuring that false positive results do not become so frequent that chemical substances cannot

be sorted meaningfully with respect to T2T.

• Screening and testing results should be reported in a format that facilitates database

development and analysis by a broad array of scientific, regulatory, and management

organizations.

• Decision criteria, such as those for statistical significance (e.g., necessary confidence intervals)

and biological plausibility, should be clearly defined.

1. If a necessary modification does compromise the existing toxicity assays, separate endocrine disruptor and

toxicity screens and/or tests should be conducted.

3 -8

EDSTAC Final Report Chapter Three August 1998

VIII. The EDSTAC Conceptual Framework

A. Obtain and Analyze Existing Information to Set Priorities

An ordered screening and testing strategy should begin with an effort to obtain and analyze

available information on new and existing chemical substances or mixtures. Information on toxic

and physiological effects, chemical structure activity relationships (SARs), use information,

product chemistry, exposure information, and legal mandates will be examined. Given limited

resources and capacity, as well as the potential magnitude of the task, it will be necessary to

develop a priority setting system to determine the relative order in which chemical substances and

mixtures will be subjected to T1S. An evaluation and analysis of this information will lead to one

of four possible determinations:

• polymers which will be placed into a “hold” status (with some exceptions) pending a review of

their monomers and oligomers;

• chemicals for which insufficient data exists to proceed to either T2T or hazard assessment and

will, therefore, need to be prioritized for T1S;

• chemicals for which sufficient data exists, or voluntary bypass of T1S, to go to T2T; and

• chemicals for which sufficient data exists to go to hazard assessment.

B. Tier 1 Screening to Detect Interactions With the Endocrine System

The purpose of T1S is to obtain a minimum, yet sufficient, set of valid and reliable data to detect

whether a chemical substance or mixture may interact with the endocrine system. Included in T1S

is a battery of assays designed to detect effects that enhance, mimic, or inhibit estrogen, androgen,

and thyroid hormone-related processes. In contrast to the more refined and detailed tests of Tier

2, the T1S assays should:

• be inexpensive, quick, and easy to perform;

• be validated and standardized as soon as possible, defining characteristics such as sensitivity

and specificity against a clearly defined standard, once it is identified;

• be more “sensitive” than they are “specific,” meaning they should have as their primary

objective the minimization of false negative or (Type II) errors, while permitting an as-of-yet

undetermined, but acceptable, level of false positive or (Type I) errors;

• capture multiple endpoints and reflect as many modes of endocrine action as possible;

• be broadly predictive across species, gender, and age; and

• yield data capable of being interpreted as either positive or negative for the purpose of

determining whether and how to conduct T2T.

3 -9

3 -

EDSTAC Final Report Chapter Three August 1998

Information gathered during T1S should be used to make initial judgments about areas of concern

and should direct and focus T2T. The interpretation of T1S results should be consistent with best

scientific judgment, formed on the basis of considerations such as “weight-of-evidence,”

consistency of the data set, and methodological strengths and limitations. Based on the evaluation

made during this phase, one of two decisions is possible:

1. Proceed to T2T – If the interpretation of results from the full battery of T1S assays is

determined to be “positive” (i.e., the screens produced evidence of interaction with the

endocrine system, within the scope of endocrine functions addressed by the program), the

chemical substance or mixture will enter T2T to characterize the nature of any endocrine

disrupting effects.

2. Hold Screening and Testing – If the interpretation of results from the full battery of T1S

assays is determined to be “negative” (i.e., the screens have not produced evidence of

interactions with the endocrine system, within the scope of endocrine functions addressed

by the program), and these results are not contravened by the “weight-of-evidence”

developed during the prioritization phase, no additional screening or testing is necessary

unless:

a) existing statutes require periodic review (e.g., FIFRA re-registration);

b) new statutory requirements mandate review;

c) new screens for endocrine disruption are incorporated into the strategy and it is

determined that these new screens may either generate significant new information or

they invalidate prior screens and therefore warrant the re-screening of chemical

substances and mixtures that have already been subjected to T1S; and/or

d) new information on the endocrine disrupting potential of the chemical substance or

mixture becomes available which warrants the re-screening of the chemical substance

or mixture.

C. Tier 2 Testing to Determine and Characterize Endocrine Disruption

The purpose of T2T is to determine and characterize the nature, likelihood, and dose-response

relationship of estrogen-, androgen-, and thyroid-related effects in humans and wildlife. Selection

of Tier 2 tests should be based upon T1S results and other relevant information. An underlying

principle of T2T is that it should provide information useful for human/ecological hazard

assessment. The T2T scheme should be flexible enough to allow for scientific judgment in the

selection of the most appropriate tests and endpoints, and costs should be practical. T2T should

be aimed at determining whether the chemical substance or mixture is an endocrine disruptor. In

addition, these tests should be designed to establish the relationship between different exposure

levels, timing and duration of exposure, and adverse effects, including developmental and

reproductive effects on the individual and its progeny.

3 -

EDSTAC Final Report Chapter Three August 1998

In comparison to the components of the T1S phase, T2T should ideally be both sensitive and

specific. In other words, it should be designed to minimize both false positive (Type I) and false

negative (Type II) errors. Additionally, this battery of tests should:

• include assessment of endpoints identified as relevant from T1S;

• include parental/offspring developmental endpoints (e.g., two-generation studies) in order to

adequately evaluate all life stages;

• include the life cycle of both viviparous (live birth) and oviparous (egg-laying) organisms;

• be conducted at a range of doses that allows full characterization of the adverse effects of the

chemical substance or mixture being tested;

• be conducted in accordance with Good Laboratory Practice (GLP) regulations to the degree

consistent with resources and the goal of timely decisions; and

• be validated, if need be, as soon as possible against a clearly defined standard.

Interpretation of results from T2T should reflect current scientific judgment, including

considerations such as “weight-of-evidence” and consistency of the data set.

A “negative” result in T2T should abrogate any additional endocrine screening and testing for that

particular chemical substance or mixture within the scope of endocrine functions defined by the

program unless:

a) existing statutes require periodic review (e.g., FIFRA re-registration);

b) new statutory requirements mandate review;

c) new screens or tests for endocrine disruption are incorporated into the EDSTAC

strategy which will generate significant new information, or invalidate prior screens

or tests upon which decisions have been made to stop screening and testing;

d) new information on the endocrine disrupting potential of the chemical substance or

mixture becomes available and it is determined that this new information warrants

additional testing; and/or

e) there are changes in the use and expected exposure patterns upon which the initial

selection of tests was made.

In the event of a “positive” outcome, the chemical substance or mixture will proceed to the hazard

assessment phase of decision-making, whereupon it may be decided that additional T2T is

required before a final determination of hazard can be made.

IX. Additional Components and Clarifications to the Original EDSTAC

Conceptual Framework

During the course of its deliberations, the EDSTAC identified an additional concept, the

incorporation of “high throughput pre-screening,” to be considered in the context of the original

Conceptual Framework. In addition, the EDSTAC clarified conditions under which a chemical

substance might be permitted to bypass T1S assays and, instead, go directly to T2T. These two

3 -

EDSTAC Final Report Chapter Three August 1998

issues are closely connected to other issues discussed later in the report, including testing at low

doses and the definitiveness of T2T.

A. High Throughput Pre-Screening

During its deliberations the EDSTAC concluded that biological effects data were incomplete or

lacking for most chemical substances. In the absence of biological effects data, EPA would be left

with the choice of either raising or lowering the priority of a chemical based on the lack of effects

information.

To address this problem, the EDSTAC recommends that a subset of the in vitro assays

recommended for inclusion in the T1S battery should be conducted with the assistance of

automated technology, in advance of the priority setting step of the overall sorting and priority

setting process. The EDSTAC came to refer to this technology, which uses robotics and other

automated processes, and the role that it will play in the overall endocrine disruptor screening and

testing program, as “high throughput pre-screening” (HTPS).

HTPS results, by themselves, will not be sufficient to make a determination about whether a

chemical may interact with the endocrine system of an intact animal. Such determinations will

require additional screening and testing.

The primary purpose of the HTPS is to provide preliminary biological activity information for a

large number of chemicals in a relatively short period of time. This information will

simultaneously be of value for: (a) detecting hormonal activity (as a component of T1S); and (b)

providing at least some biological effects-related information for the estrogen, androgen, and

thyroid hormonal systems to assist in the effort to set priorities for T1S. The in vitro assays that

would be performed as part of the HTPS include transcriptional activation assays. Performance

of these assays would still be required as part of T1S for all chemicals that do not go through

HTPS.

The EDSTAC recommends that all chemicals currently produced in an amount equal to or greater

than 10,000 pounds per year (estimated to be about 15,000 chemicals) be subjected to HTPS.

Also, it is expected that all pesticides (i.e., both active ingredients and formulation inerts) will be

subjected to HTPS. Any chemicals subjected to the assays conducted in the HTPS step would

not be required to repeat the ER binding/transcriptional activation assay and the AR

binding/transcriptional activation assay as part of T1S. On the other hand, for any chemicals not

subjected to HTPS (e.g., production volumes less than 10,000 pounds per year), the assays in

T1S would result in information equivalent to that which would have been provided from HTPS.

The HTPS concept is explained in more detail in Chapter Four, Section V, and referred to often in

Chapter Five.

3 -

EDSTAC Final Report Chapter Three August 1998

B. Alternative Means of Meeting Versus Bypassing Tier 1 Screening

The EDSTAC expects the vast majority of chemicals included in the EDSTP to go through the

program in the logical, hierarchical manner in which the program was designed. Notwithstanding

this expectation, the EDSTAC recognizes there will be circumstances where it may be inefficient

to follow all steps of the EDSTP. For example, the EDSTAC Conceptual Framework allows

chemicals that have already been subjected to tests that are the “functional equivalents” of the

T2T two-generation reproductive toxicity tests, endocrine disruptor endpoints, taxa, and dosing

considerations recommended by the EDSTAC, to bypass both T1S and T2T and proceed directly

to hazard assessment.

The EDSTAC has identified two other circumstances where a chemical substance or mixture may

not be required to be evaluated in the assays included in the recommended T1S battery. Both of

these scenarios are discussed below and then in further detail in Chapters Four and Five.

1. Alternative Means to Meet T1S Information Requirements

The EDSTAC recommends that it should be permissible to complete the information requirements

of T1S through the submission of data that are “functionally equivalent” to the data that would be

generated from the recommended T1S battery. Further, functionally equivalent information could

be submitted for one or more of the recommended T1S assays or for the entire battery. The

EDSTAC believes it is helpful to distinguish this scenario, which is in essence an alternative

means of meeting the information requirements associated with T1S, from two other scenarios,

which are considered bypassing T1S.

2. Bypassing T1S

There are two scenarios in which the EDSTAC recommends that the owner of a chemical should

be permitted to bypass T1S. Each of these two scenarios has different implications for the

information requirements associated with completing T2T and hazard assessment following T2T.

a) Chemicals That Have Previously Been Subjected to Two-Generation Reproductive

Toxicity Tests

The first scenario includes those chemicals that have previously been subjected to mammalian and

wildlife developmental toxicity and/or reproductive toxicity testing, but such testing may not

include additional EAT endpoints for T2T, as specified in Chapter Five, Section VI. The

EDSTAC expects that food-use pesticides will fall into this category, given the requirements of

FIFRA and FQPA, as well as a small number of other types of pesticides and industrial chemicals.

The EDSTAC agrees that chemicals which meet this criterion for bypassing T1S would still be

subjected to the assays that will be part of the HTPS. In addition, chemicals which meet this

criterion will also be the most likely candidates for the alternative approaches for completing T2T,

as discussed in Chapter Five, Section V, C, 2.

3 -

EDSTAC Final Report Chapter Three August 1998

b) Chemicals for Which There is No Prior Toxicity Testing

The second scenario includes those chemicals for which the owner of the chemical has decided to

voluntarily go to T2T without having completed the full T1S battery or any prior two-generation

reproductive toxicity testing. These chemicals must be evaluated in the HTPS assays. In

addition, chemicals that bypass T1S under this second scenario must be evaluated in all the tests

of the T2T battery (i.e., the mammalian and non-mammalian multi-generation tests with all the

recommended endpoints), consistent with the principles governing T2T, which are set forth in

Chapter Five, Section V, C.

X. Compilation of Chapter Three Recommendations

1. The EDSTAC developed a tiered Conceptual Framework that formed the basis for its

screening and testing strategy and all subsequent recommendations. The tiered framework

consists of the following three major activities:

• Priority setting includes the sorting and prioritization of chemical substances and mixtures

based on existing information. The existing information would be used to sort chemicals into

four categories. An evaluation and analysis of this information will lead to sorting chemicals

into one of four categories:

• Polymers, which are placed into a “hold” status (with some exceptions) pending a review

of their monomers and oligomers.

• Chemicals for which there is insufficient data regarding endocrine disruption and will

therefore need to be prioritized for Tier 1 Screening.

• Chemicals for which sufficient data exists to proceed to Tier 2 Testing.

• Chemicals for which sufficient data exists to go to hazard assessment.

• Tier 1 Screening (T1S) to detect chemical substances and mixtures capable of acting on

endocrine systems.

• Tier 2 Testing (T2T) to determine, characterize, and quantify the nature of the endocrine

disrupting properties of the chemical substances and mixtures identified by prior information

and/or T1S.

2. The EDSTAC recommended the adoption of several principles to guide the use of the

Conceptual Framework.

• A chemical may bypass one or more tiers when warranted by appropriate information

(e.g., sufficient prioritization data on endocrine disrupting properties to initiate T2T or

hazard assessment).

• If information is inadequate to determine if a chemical should move to the next tier, an

active process should be developed for generating the needed information to make such a

3 -

EDSTAC Final Report Chapter Three August 1998

decision.

• Criteria and default assumptions for deciding whether chemicals move from one tier to the

next should be developed in advance of initiating screening and testing.

3. Within the context of the Conceptual Framework, the EDSTAC recommends that the overall

scope of the screening and testing strategy should:

• be relevant to both human health and ecological effects;

• initially emphasize identifying and characterizing effects that enhance, mimic, or inhibit

estrogenic, androgenic, and thyroid hormone-related processes;

• consider tests that detect multiple hormone interactions, address endpoints in multiple

species, and predict long-term or delayed effects;

• be periodically revisited to permit inclusion of additional hormone-mediated effects or new

screens and tests as they become available;

• be capable of evaluating the endocrine disrupting properties of chemical substances and

common mixtures; and

• allow determination of possible additive, synergistic, or antagonistic effects caused by

interactions among the components of mixtures.

4. The EDSTAC recommends nine broad principles to guide the implementation of the

endocrine disruptor screening and testing program. The screening and testing program

should:

• require the minimal number of screens and tests necessary to make sound decisions,

thereby reducing the time needed to make these decisions;

• examine existing screens and tests for their potential to predict, detect, and/or characterize

endocrine disruptors, ensuring that any modification to existing screens and tests does not

compromise their ability to predict other toxicity endpoints;

• systematically examine existing screening and testing data not only for adverse endpoints

in high dose groups, but also for physiological changes in low dose groups;

• not detract from current and new efforts to assess the toxicity of compounds and mixtures

through mechanisms other than endocrine disruption;

• provide data that can be used for a broad range of management and regulatory programs

in a form that supports international harmonization of their use;

• include periodic review of new scientific information;

• use a performance-based approach to the selection of screens, tests, and species, including

the use of more appropriate screens, tests, and species as they are developed and

validated;

• be dynamic in order to stay current with the rapidly evolving science related to the

endocrine system; and

• be conducted at the minimum cost necessary to make the decisions within the EDSTAC

Conceptual Framework.

5. The EDSTAC also recommended six guiding principles specific to the screens and tests

3 -

EDSTAC Final Report Chapter Three August 1998

themselves.

• To facilitate making decisions within the EDSTAC Conceptual Framework, all screens

and tests should have well-defined endpoints.

• The use of animals should be reduced to the minimum level needed to obtain scientifically

valid results and interpretations.

• The results of screens and tests should support further research on effects of endocrine

disruptors on populations, communities, and ecosystems.

• In interpreting screening and testing results, a “weight-of-evidence” approach should be

used, but should be consistent with a principle of prudence in protecting human health and

the environment. In the case of T1S, this means that the strategy will err on the side of

false positive identifications rather than false negatives.

• Screening and testing results should be reported in a format that facilitates database

development and analysis by a broad array of scientific, regulatory, and management

organizations.

• Decision criteria, such as those for determining statistical significance (e.g., necessary

confidence intervals) and biological plausibility, should be clearly defined.

6. The EDSTAC recommends that T1S provide the minimum, yet valid and reliable, data to

detect interactions with the endocrine system. In contrast to Tier 2 tests, T1S assays should:

• be inexpensive, quick, and easy to perform;

• be validated and standardized as soon as possible, defining characteristics such as

sensitivity and specificity against a “gold standard,” once it is identified;

• be more “sensitive” than they are “specific,” meaning that they should have as their

primary objective the minimization of false negative or (Type II) errors, while permitting

an as-of-yet undetermined, but acceptable, level of false positive or (Type I) errors;

• capture multiple endpoints and reflect as many modes of endocrine action as possible;

• be broadly predictive across species, gender, and age; and

• yield data capable of being interpreted as either positive or negative for the purpose of

determining whether and how to conduct T2T.

7. The EDSTAC recommends that T1S be used to make initial judgments about areas of concern

in order to direct the focus of T2T. The interpretation of T1S results should be consistent with

best scientific judgment, formed on the basis of considerations such as “weight-of-evidence,”

consistency of the data set, and methodological strengths and limitations.

8. The EDSTAC recommends that T2T be based upon T1S results and other relevant

information. An underlying principle of T2T is that it should provide information useful for

human and ecological hazard assessment. The T2T scheme should be flexible enough to allow

for scientific judgment in the selection of the most appropriate tests and endpoints, and costs

should be reasonable. Tests should be aimed at determining whether the chemical substance or

mixture is an endocrine disruptor and whether the effects are a result of primary or secondary

disturbances of endocrine function. In addition, these tests should be designed to establish the

relationship between different exposure levels, timing and duration of exposure, and adverse

3 -

EDSTAC Final Report Chapter Three August 1998

effects, including developmental and reproductive effects on the individual and its progeny.

9. In contrast to T1S, the EDSTAC recommends that T2T should be both sensitive and specific,

and designed to minimize false positive (Type I) and false negative (Type II) errors.

Additionally, this battery of tests should:

• include assessment of endpoints identified as relevant from Tier 1 screens;

• include parental/offspring developmental endpoints (e.g., two-generation studies) in order

to adequately evaluate all life stages;

• include the life cycle of both viviparous (live birth) and oviparous (egg-laying) organisms;

• be conducted at a range of doses that allow full characterization of the adverse effects of

the chemical substance or mixture being tested;

• be conducted in accordance with Good Laboratory Practice (GLP) regulations to the

degree consistent with resources and the goal of timely decisions; and

• be validated, if need be, as soon as possible against a clearly defined standard.

10. The EDSTAC recommends that a subset of the T1S in vitro assays be conducted with the

assistance of automated technology to provide biological effects information to assist in the

overall sorting and priority setting process. Because of the role this technology will play in

the overall EDSTP, the EDSTAC refers to it as “high throughput pre-screening” (HTPS).

The EDSTAC recommends that all chemicals currently produced in an amount equal to or

greater than 10,000 pounds per year (estimated to be about 15,000 chemicals) be subjected

to HTPS. Also, it is expected that all pesticides (i.e., both active ingredients and formulation

inerts) will be subjected to HTPS. Any chemicals subjected to the assays conducted in the

HTPS step would not be required to repeat the ER binding/transcriptional activation assay

and the AR binding/transcriptional activation assay as part of T1S. On the other hand, for

any chemicals not subjected to HTPS (e.g., production volumes less than 10,000 pounds per

year), the assays in T1S would result in information equivalent to that which would have

been provided from HTPS.

11. The EDSTAC recommends that the vast majority of chemicals go through priority setting,

T1S, and T2T in a sequential manner. However, the EDSTAC also recognizes there may be

individual cases in which T1S is bypassed. Three situations were identified where a chemical

may bypass T1S, each with different implications for information requirements in T2T.

• Alternative means to meet T1S information requirements through the generation of data

which are “functionally equivalent” to data derived from the recommended screening

battery.

• Bypassing T1S for chemicals (e.g., food-use pesticides) that have previously been

subjected to two-generation reproductive toxicity tests. Such chemicals should still be

subjected to high throughput pre-screening assays.

• Bypassing T1S for chemicals for which there is no prior toxicology testing but the owner

has voluntarily decided to proceed directly to testing. Such chemicals must be evaluated

in the high throughput pre-screening assays, and all of the tests in the T2T battery.

3 -

EDSTAC Final Report Chapter Three August 1998

XI. Literature Cited

European Commission, “European Workshop on the Impact of Endocrine Disrupters on Human

Health and the Environment,” Environment and Climate Research Programme, DG XII,

European Commission, Report EUR 17549, 1997.

Kavlock, R.J. et al., “Research needs for risk assessment of health and environmental effects of

endocrine disruptors: A review of the U. S. EPA-sponsored workshop,” Environmental

Health Perspectives, 104, 1996, pp. 715-740.

Tattersfield, L. et al. (Ed.). “SETAC-Europe/OECD/EC Expert Workshop on Endocrine

Modulators and Wildlife: Assessment and Testing,” SETAC-Europe, Brussels, 1997.

U. S. EPA, “Special Report on Environmental Endocrine Disruption: An Effects Assessment and

Analysis. Office of Research and Development,” EPA/630/R-96/012, 1997.