This chapter reviews the approaches used by the US Environmental Protection Agency (EPA) to answer questions about the adequacy of its toxicity-testing strategies and risk-assessment practices on the basis of its findings about nonmonotonic dose response (NMDR) relationships for endocrine disruptors in its draft State of the Science Evaluation: Nonmonotonic Dose Responses as They Apply to Estrogen, Androgen, and Thyroid Pathways and EPA Testing and Assessment Procedures (SOTS evaluation). The questions that EPA sought to answer are presented in Box 3-1. The committee evaluated whether EPA applied sound scientific principles, whether potential effects on toxicity-testing guidelines and risk-assessment practices were appropriately considered, and whether conclusions were adequately justified in response to Questions 2 and 3.
The answers to those questions depend on the scope of and answer to Question 1. As discussed at length in Chapter 2, the SOTS evaluation did not provide adequate analysis and support for its conclusions that NMDR curves were found more often in in vitro studies, at high doses, and for exposures of short duration. The committee found that EPA’s approach to evaluating whether NMDR curves exist for endocrine disruptors was not systematic, consistent, or transparent, especially with regard to the literature on estrogen and androgen effects. Thus, the same issues and problems discussed in Chapter 2 carry over to the review of the evidence in this chapter.
THE ENVIRONMENTAL PROTECTION AGENCY’S EVALUATION OF THE ADEQUACY OF TOXICITY TESTING
In trying to answer Question 2, EPA posed three more specific questions:
2a. Are there adverse effects with NMDRs that are not being identified by using the current chemical testing strategies?
2b. Are there NMDRs for adverse effects below the no observed adverse effect levels (NOAELS) or benchmark doses (BMD) derived from the current testing strategies?
2c. Do EPA chemical testing strategies detect relevant adverse effects for chemicals which produce NMDR curves for specific endpoints?
No description of “current chemical testing strategies” is provided, so it was unclear to the committee whether they included EPA’s guidelines for toxicity testing under the Toxic Substances Control Act (TSCA) or the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA), testing associated with the Endocrine Disruptor Screening Program (EDSP), testing associated with EPA’s ToxCast program, or the larger body of investigative toxicology and clinical or epidemiologic research. Each of those testing strategies has specific goals and applications that are important for identifying potential hazards, including those associated with endocrine disruptors, and for framing the mode of action (MOA) or weight of evidence approaches that ultimately affect risk assessments. For the purposes of the present review, the committee assumed that EPA’s questions and conclusions pertain to TSCA, FIFRA, and EDSP testing strategies.
EPA cited the work of a few European agencies, such as the European Food Safety Authority (EFSA 2013) and the Danish Centre on Endocrine Disrupters (Hass et al. 2013), that have analyzed the strengths and weaknesses of testing guidelines for addressing endocrine-disruptor toxicity and dose–response issues. Their approaches are relevant to the SOTS evaluation because they explicitly analyze the adequacy of specific testing methods. EPA should use a similar approach to analyze the strengths and weaknesses of its testing guidelines for addressing its questions about their adequacy for detecting adverse effects associated with NMDR curves.
BOX 3-1 Central Scientific Questions EPA Addressed in the SOTS Evaluation (EPA 2013)
• Question 1: a Do [NMDRs] exist for chemicals and if so under what conditions do they occur?
• Question 2: Do NMDRs capture adverse effects that are not captured using [EPA’s] current chemical testing strategies (i.e., false negatives)?
• Question 3: Do NMDRs provide key information that would alter EPA’s current weight of evidence conclusions and risk assessment determinations, either qualitatively or quantitatively?
aSee Chapter 2 for committee’s evaluation of how Question 1 was answered.
Failure to present the testing strategies under consideration and to develop a framework for analyzing them in the context of the NMDR evidence makes it difficult to understand how EPA drew its conclusions. As noted in Chapter 2, the SOTS evaluation should meet a higher standard of evaluation if it is to be a foundational synthesis of the literature to inform decisions about EPA’s testing strategies. In the sections below, the committee considers issues regarding each of the individual questions that EPA sought to answer.
Question 2a: Are There Adverse Effects with NMDRs That Are Not Being Identified by Using the Current Chemical Testing Strategies?
In response to this question, the SOTS evaluation states that EPA’s toxicity-testing strategy is not intended to identify all possible adverse outcomes of chemical exposure but instead is intended to provide broad coverage of sensitive end points that are relevant to human health. It states that estrogen-, androgen-, and thyroid-responsive effects are evaluated as part of testing in EDSP and that multigeneration studies include end points that are sensitive to androgen and estrogen action. It concludes that “current testing strategies are unlikely to mischaracterize, as a consequence of NMDR, a chemical that has the potential for adverse perturbations of the estrogen, androgen or thyroid pathways.” However, the SOTS evaluation provides no analysis to support that conclusion.
Answering Question 2a in a defensible and transparent manner requires a comparison of the end points evaluated in current testing strategies that are thought to be sensitive to estrogen, androgen, or thyroid effects with the end points associated with NMDR curves. The MOAs associated with those effects should then be evaluated in the context of additional supportive data, if available, to help in understanding the sequences of key events that ultimately lead to the adverse effects. As the SOTS evaluation noted, not all measured end points are themselves adverse or necessarily key events associated with an adverse response, so careful consideration is needed. Early development is an especially sensitive window of susceptibility, and perturbations that are adaptive in the average adult may give rise to an adverse response in the fetus, for example. The analysis should give careful consideration to the suite of end points assessed in reproductive and developmental toxicity and related tests, given the importance of hormones in fetal, perinatal, and adolescent development. A more rigorous approach to analyzing the evidence would help in framing the data to determine whether current testing strategies are qualitatively identifying hazards irrespective of whether an NMDR curve is present and whether follow-on investigative studies would be needed to differentiate key events associated with adverse events from events that are considered adaptive.
Question 2b. Are there NMDRs for Adverse Effects Below the No Observed Adverse Effect Levels or Benchmark Doses Derived From the Current Testing Strategies?
In response to this question, the SOTS evaluation states that “for estrogen, androgen or thyroid MoA[s] that provide adequate information to make an assessment, our evaluation shows that there is not sufficient evidence of NMDRs for adverse effects below the NOAELs or BMD derived from the current testing strategies.” However, the evaluation provides no analysis of EPA’s testing protocols for their adequacy to detect NMDR curves for adverse responses at doses below NOAELs or BMDs derived from traditional toxicology-testing protocols. Without such an analysis of the strengths and weaknesses associated with testing protocols, EPA’s conclusions are not well supported.
Answering Question 2b in a more rigorous manner will require an analysis of protocols for their ability to detect adverse effects associated with perturbations of the estrogen, androgen, and thyroid pathways and for their ability to detect NMDR curves. EPA should pay particular attention to issues of study design—such as the number of dose groups, dose spacing, numbers of animals tested, end points evaluated, and statistical power—in considering whether current testing strategies have the ability to detect NMDR curves for adverse responses. For example, toxicity studies that test the minimum of three dose groups and a control group are inadequate for identifying complex dose–response relationships. That limitation is acknowledged in the SOTS evaluation, as evidenced by EPA’s focus on studies that use four or preferably more dose groups. As discussed in Chapter 2, EPA was left with a relatively small number of studies from which to draw definitive conclusions on NMDR curves for adverse effects associated with a broad range of exposures and multiple potential outcomes. That may well reflect true inadequacies in the literature and even in EPA’s testing protocols, but a more rigorous analysis of the literature is needed to support any conclusions, including a finding that the available data are inadequate to reach a conclusion.
Question 2c. Do EPA Chemical Testing Strategies Detect Relevant Adverse Effects for Chemicals Which Produce NMDR for Specific Endpoints?
EPA’s response to Question 2c appears to be that current testing assays in the EDSP’s screening battery detect disruption of estrogen, androgen, and thyroid pathways and that standard multigenerational tests “have measures that are sensitive to disruption of the estrogen and androgen signaling pathways” (p. 13). However, as discussed in relation to Questions 2a and 2b, the SOTS evaluation provides no analysis to support these answers. A rigorous comparison of the end points evaluated in EPA’s current testing strategy with the expected outcomes of exposures to chemicals that affect the estrogen, androgen, and thyroid pathways is needed. EPA could then identify the outcomes for which NMDR curves are
plausible and consider the extent to which they would be captured by the other end points. The analysis could then be supported by a review of the human physiologic, toxicologic, and epidemiologic literature to provide context for toxicity assessments conducted in laboratory animals.
It is unclear whether “relevant adverse effects” refers to ones that would be important for the purposes of hazard identification or ones that would lead to further testing requirements or investigative research to establish adversity. Furthermore, not all end points are considered equal in a risk assessment; some effects are given more weight than others. For example, cardiac malformations can provide stronger evidence of a developmental hazard than would modulation of a hormone concentration. Both are measurable end points, but they have different considerations in hazard identification. The SOTS evaluation does not include an analysis that considered both the extent of end-point coverage and the quality of the response.
Question 3 of the SOTS evaluation was: Do NMDRs provide key information that would alter EPA’s current weight of evidence conclusions and risk assessment determinations, either qualitatively or quantitatively? A conclusion was drawn that “data from studies in which NMDRs are identified may be biologically relevant and as such should be evaluated in context with the totality of the available scientific data in weight of evidence (WoE) conclusions and risk assessment determinations. … NMDRs can have impact on both qualitative and quantitative risk assessments, but cannot be considered in isolation from other data for the chemical and biological response being considered” (p. 14).
Although risk-assessment guidelines specific to endocrine disruptors have not been developed, the committee recognizes that endocrine disruption has been treated as a specific MOA in guideline categories, such as those for developmental toxicity (EPA 1991), reproductive toxicity (EPA 1996), cancer (EPA 2005a,b), neurotoxicity (EPA 1998a), mutagenicity (EPA 1986), and ecologic effects (EPA 1998b). EPA acknowledged that many prior reviews and expert panels concluded that there is a need to address specifically how endocrine disruptor–related dose–response relationships affect current risk-assessment practices. However, EPA did not analyze how evidence of NMDR curves would be considered in its current risk-assessment practices and drew only the general conclusion that NMDR curves, if deemed important and relevant, would have an effect. Thus, the SOTS evaluation fails to provide the analysis needed to answer Question 3 in a defensible manner.
To address Question 3 directly and respond to the challenges that have been raised, the committee recommends that the draft SOTS evaluation be revised to include a specific analysis and presentation of how NMDR relation-
ships, to the extent that evidence supports their existence for adverse effects, would be addressed in current risk-assessment guidelines and practices and under what circumstances they would be found inadequate. A key consideration will be an examination of how NMDR curves bear on the WOE and MOA analyses that drive the risk-assessment process. Even if the current literature is found not to provide compelling evidence of NMDR curves of health relevance in the exposure range of regulatory concern, an assessment of how such evidence would alter risk assessment is needed to answer Question 3. For example, consideration could be given to whether cross-species extrapolations should be performed differently from current practices. Those types of analyses are needed to determine whether NMDR curves provide key information that would alter EPA’s current WOE conclusions and risk-assessment determinations.
Noticeable omissions in the draft SOTS evaluation are consideration of human studies—including the physiologic, clinical, and epidemiologic evidence—and discussion of how evidence of NMDR curves in ranges relevant to humans would be incorporated into risk assessments. Furthermore, little attention is paid to potential issues associated with, for example, cross-species, lifestage, sex, route-to-route, and in vitro–to–in vivo extrapolations that help to frame the MOA and dose–response analyses that are important for developing human health risk assessments. Those issues can be even more problematic in extrapolating from aquatic or ecologic studies to the broader ecosystem or human health. Clarifying how such issues could be addressed in human and environmental health assessments would be an important contribution to the SOTS evaluation.
Thus, the committee recommends that the answer to Question 3 be justified with specific reference to EPA’s risk-assessment guidelines. As noted earlier, EPA could use examples from other organizations to guide its own analyses. For example, the Danish Centre on Endocrine Disrupters (Hass et al. 2013) provides specific examples of the strengths and weaknesses of European testing guidelines, establishment of dose–response relationships (BMDs) or NOAELs, and statistical considerations and uncertainties for use in risk assessment. EPA should consider and expand on such analyses to establish the bridge between the state of the science for NMDR curves associated with estrogen, androgen, and thyroid pathways and how they could be implemented in risk assessments in a transparent manner that is ultimately useful to the regulatory community. In that way, the final SOTS evaluation would provide a more coherent and complete response to the questions that are posed.
As EPA considers this recommendation, it should also consider reiterating and clarifying its operating definitions. In addition to the definitions of low-dose effect, resilience, and adverse effect discussed in Chapter 2, definitions of WOE and MOA in a risk-assessment context should be clarified. Clear definitions are especially important so that future debates can be based on common understandings of the strengths and weaknesses of the science that underpins methods and results.
The committee found that the SOTS evaluation provided little documentation of the analyses performed to answer questions about toxicity-testing strategies and risk-assessment practices. EPA stated in the Executive Summary of the SOTS evaluation that it did not intend to review specific testing methods or risk assessments, but this decision needs to be reconciled with the central scientific questions that are posed. Because the SOTS evaluation is to be a foundational synthesis of the literature on NMDR curves for the estrogen, androgen, and thyroid pathways, a higher standard of evaluation is required. Developing at least a simple framework that identifies the testing methods and evaluates their ability to detect endocrine-mediated adverse effects and NMDR relationships is necessary to answer, fully and transparently, questions about the ability of current testing strategies to characterize hazards related to endocrine disruptors appropriately. EPA’s conclusions about the adequacy of toxicity testing should be rooted in specific reference to and demonstrations with its own testing guidelines in a systematic manner for the estrogen, androgen, and thyroid pathways even if it eventually decides that the available data are inadequate to reach a conclusion. Similarly, it should devise a framework for evaluating the qualitative and quantitative implications of NMDR curves for WOE conclusions and risk-assessment determinations. In that way, all the factors (including issues, strengths, and weaknesses) associated with establishing dose–response relationships, characterizing MOAs, and understanding route-to-route, in vitro to in vivo, and cross-species extrapolations can be presented openly and transparently, assumptions can be clearly stated, and uncertainties can be acknowledged.
The scope of Questions 2 and 3 should be better defined and tailored to the scope of Question 1 (see Chapter 2). Important elements include the following:
• EPA should specify the toxicity-testing strategies under consideration and evaluate them for their ability to detect NMDR relationships mediated by the estrogen, androgen, and thyroid hormone pathways. The tests should be evaluated in terms of the design criteria that would be necessary to detect such relationships and the adverse effects of interest. The evaluation should include careful consideration of reproductive and developmental toxicity and related tests, given the importance of hormones in fetal, perinatal, and adolescent development.
• As noted in Chapter 2, resilience, or adaptation, is a controversial topic and is not adequately addressed by EPA in the SOTS evaluation. EPA should distinguish between adverse effects and adaptive effects. Because effects that are adaptive in some people are adverse in others, consideration should be given to potential windows of susceptibility (for example, during fetal development),
particularly sensitive populations (for example, those with pre-existing health conditions), and other factors (such as multiple chemical exposures) in making these determinations.
• EPA’s SOTS evaluation should be expanded to indicate how NMDR relationships for estrogen, androgen, and thyroid pathways would be addressed under EPA’s current risk-assessment guidelines and practices, including statistical considerations, uncertainty analyses, lifestage or susceptibility issues, and MOAs.
• As recommended in Chapter 2, EPA should reconsider its decision to exclude human evidence. Human physiologic, clinical, and epidemiologic literature may be useful in identifying end points, dose–response relationships, and MOAs that are relevant to the estrogen, androgen, and thyroid pathways. If such data are included, they could be used to highlight species differences in chemical interactions with respect to the three hormone pathways, including pharmacokinetic and pharmacodynamics processes that are important considerations in conducting risk assessments.
EFSA (European Food Safety Authority). 2013. Scientific opinion on the hazard assessment of endocrine disruptors: Scientific criteria for identification of endocrine disruptors and appropriateness of existing test methods for assessing effects mediated by these substances on human health and the environment. EFSA J. 11(3):3121.
EPA (U.S. Environmental Protection Agency). 1986. Guidelines for Mutagenicity Risk Assessment. EPA/630/R-98/003. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. September 1986 [online]. Available: http://www.epa.gov/raf/publications/pdfs/MUTAGEN2.PDF [accessed Sept. 19, 2013].
EPA (U.S. Environmental Protection Agency). 1991. Guidelines for Developmental Toxicity Risk Assessment. EPA/600/FR-91/001. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. December 1991 [online]. Available: http://www.epa.gov/raf/publications/pdfs/DEVTOX.PDF [accessed Sept. 19, 2013].
EPA (U.S. Environmental Protection Agency). 1996. Guidelines for Reproductive Toxicity Risk Assessment. EPA/630/R-96/009. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. October 1996 [online]. Available: http://www.epa.gov/raf/publications/pdfs/REPRO51.PDF [accessed Sept. 19, 2013].
EPA (U.S. Environmental Protection Agency). 1998a. Guidelines for Neurotoxicity Risk Assessment. EPA/630/R-95/001F. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. April 1998 [online]. Available: http://www.epa.gov/raf/publications/pdfs/NEUROTOX.PDF [accessed Sept. 19, 2013].
EPA (U.S. Environmental Protection Agency). 1998b. Guidelines for Ecological Risk Assessment. EPA/630/R-95/002F. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. April 1998 [online]. Available: http://www.epa.gov/raf/publications/pdfs/ECOTXTBX.PDF [accessed Sept. 19, 2013].
EPA (U.S. Environmental Protection Agency). 2005a. Guidelines for Carcinogen Risk Assessment. EPA/630/P-03/001F. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. March 2005 [online]. Available: http://www.epa.
gov/raf/publications/pdfs/CANCER_GUIDELINES_FINAL_3-25-05.PDF [accessed Sept. 19, 2013].
EPA (U.S. Environmental Protection Agency). 2005b. Supplemental Guidance for Assessing Susceptibility for Early-Life Exposures to Carcinogens. EPA/630/R-03/003F. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. March 2005 [online]. Available: http://www.epa.gov/raf/publications/pdfs/childrens_supplement_final.pdf [accessed Sept. 19, 2013].
EPA (U.S. Environmental Protection Agency). 2013. State of the Science Evaluation: Nonmonotonic Dose Responses as They Apply to Estrogen, Androgen, and Thyroid Pathways and EPA Testing and Assessment Procedures. U.S. Environmental Protection Agency. June 2013 [online]. Available: http://epa.gov/ncct/download_files/edr/NMDR.pdf [accessed Aug. 12, 2013].
Hass, U., S. Christiansen, M. Axelstad, K. Dreisig Sorensen, and J. Boberg. 2013. Input for the REACH-review in 2013 on Endocrine Disrupters. Final report. Center on Endocrine Disrupters, Division of Toxicology and Risk Assessment, National Food Institute, Technical University of Denmark. March 21, 2013 [online]. Available: http://www.mst.dk/NR/rdonlyres/B865F94A-54C0-43DF-AB0F-F07A593E1FD0/0/ReachreviewrapportFINAL21March.pdf [accessed July 15, 2013].
