SETAC Focused Topic Meeting on Endocrine Disrupting Chemicals: Overview and Outcomes
Gerald Ankley, USEPA, Peter Matthiessen, Independent Consultant, United Kingdom, Mary Ann Ottinger, University of Houston and Glen Van Der Kraak, University of Guelph
The SETAC North America Focused Topic Meeting (FTM) on Endocrine Disruption was held 4–6 February 2014 at the USEPA campus in Research Triangle Park, NC. The meeting, which was co-chaired by Annegaaike Leopold (Wildlife International) and Holly Zahner (USFDA), was supported by more than 20 sponsors representing the private sector and government. More than 200 participants from 10 different countries representing four of the five SETAC geographic units attended. Moreover, five SETAC student members, supported by the meeting sponsors, attended the meeting.
This meeting was a follow-up to a similar workshop held 24–25 October 2012 in Brussels, which focused on research and regulatory issues for endocrine-disrupting chemicals (EDCs), with emphasis on the European perspective. The FTM provided an opportunity to explore EDC issues from a North American perspective, which differs somewhat from other parts of the world. Therefore, an important emphasis of the meeting concerned the status of the USEPA Endocrine Disruptor Screening Program (EDSP). The EDSP stems from a legislated mandate directing the USEPA to develop a screening and testing program for certain classes of EDCs, specifically those with the potential to cause adverse effects in humans or wildlife through direct perturbation of the hypothalamic–pituitary–gonadal and –thyroidal (HPG/T) axes. This is achieved through an initial screening of chemicals through 11 in vitro and in vivo assays designed to determine the presence of endocrine activity (Tier 1), followed by more intensive testing of active chemicals to determine if activity at Tier 1 translates into adverse effects, and through collecting data (e.g., dose-response relationships in full life cycle tests) suitable for conducting formal risk assessments (Tier 2).
The FTM was comprised of five sessions plus 30 poster presentations. This article provides a brief overview of the presentations and discussions associated with these five sessions, as well as major conclusions and recommendations.
Session 1—Setting the Stage
Glen Van Der Kraak (University of Guelph) opened the meeting with an overview of where the EDC issue currently stands in terms of science and policy. He noted that tremendous strides have been made in the almost 20 years since EDCs first became a focal point in the lay and scientific communities, for example, surveillance has increased, chemicals of concern are being defined, and tools and concepts for detecting and testing endocrine-active chemicals are being developed. He also noted that the field sits at a crossroads in terms of how risks should be identified and managed, highlighting the need for transparent, weight-of-evidence (WoE)-based approaches to determine when and where EDCs are actually problematic.
Mary Manibusan (USEPA) provided a synopsis of the EDSP, including description of legislative origins of the program and its evolution through the input of multi-stakeholder advisory groups. She described several recent (2013) reviews of the program by the USEPA Science Advisory Panel (SAP), focused on proposed WoE evaluations of submitted Tier 1 screening data, the status of Tier 2 tests, and options for prioritization of chemicals for screening and testing using computational and high throughput (HTP) tools. She noted that under current mandates, the EDSP program needs to consider around 10,000 chemicals relative to potential endocrine-disrupting activity and that this would be difficult to achieve without some sort of prioritization prior to Tier 1 and Tier 2 evaluations. The role of the adverse outcome pathway (AOP) concept as an integrating principle for developing prioritization and testing approaches for EDCs was stressed.
Niklas Andersson (European Chemical Agency [ECHA]) discussed ECHA's role and activities as well as the current status of regulatory activities in the Europe Union (EU) relative to EDCs. ECHA is managing the implementation of REACH (Regulation [EC] No 1907/2006 on Registration, Evaluation, Authorisation and Restriction of Chemicals), CLP (Regulation [EC] No 1272/2008 on classification, labeling and packaging of substances and mixtures) and BPR (Regulation [EU] No 528/2012 on making available on the market and use of biocidal products). He highlighted that in addition to REACH and BPR, endocrine disrupting properties are also addressed in PPPR (Regulation [EC] No 1107/2009 on placing of plant protection products on the market). EDCs are addressed in the REACH regulation in the context of restriction and authorization, and in BPR and PPPR in the context of the approval of active substances and exclusion criteria. Under REACH, substances having endocrine disrupting properties may be identified, on a case-by-case basis, as substances of very high concern (SVHC) and added to the REACH Candidate List. Other substances on the Candidate List include chemicals known to cause cancer, mutations and toxicity to reproduction (CMR), and chemicals which are persistent, bioaccumulative and/or toxic in the environment or very persistent and very bioaccumulative (PBT/vPvB). Authorization requirements apply to SVHCs that are included in Annex XIV of REACH, with the objective of progressively replacing them with safer substances or technologies when economically and technically viable, reducing their use, and ultimately replacing them with safer alternatives. Risk assessment related to plant protection products and biocides encompasses endocrine disruption. Both regulations include exclusion criteria for approval of active substances, which means that the active substance is not approved if it has endocrine disrupting properties. However, depending on the regulation, derogations exist in case of, for example, prevention or control of serious danger to human health, animal health, negligible exposure or socio-economic consequences. The interim criteria for endocrine disrupters in BPR refer to the CLP criteria for Carc. 2 and Repr. 2. The European Commission has to decide on the criteria that will determine whether a material is an endocrine disruptor or not. Several working groups have been and are involved in establishing these criteria. Recently, there has been a delay due to an impact assessment that has to be done.
James Wheeler (Syngenta) discussed some of the repercussions of hazard versus risk-based approaches for EDCs, noting that the former could result in substantial economic impacts. Decreased options relative to plant protection products (pesticides) used to enhance food production are one such example. Regulatory needs from an industry perspective were discussed, including how the same data will likely be interpreted and have different consequences across geographies. Both Andersson and Wheeler noted that the regulatory situation in the EU is still is evolving and that at present it is not possible to predict exactly how EDCs will ultimately be addressed.
Session 2—Where We Have Been: Data Interpretation and Lessons Learned From Tier 1
Les Touart (USEPA) provided an overview of the 11 Tier 1 EDSP assays. The suite of tests is specifically designed to detect chemicals with the potential to perturb the HPG/T axes in vertebrates through mechanisms such as activation and antagonism of target nuclear hormone receptors and inhibition of hormone synthesis (www.epa.gov/endo). There are five in vitro assays focused on binding to and transactivation of the estrogen receptor, binding to the androgen receptor, and inhibition of sex steroid synthesis. There are six in vivo Tier 1 screens, four utilizing rats (Uterotrophic and Hershberger assays; male and female pubertal assays), one with the fathead minnow (fish short-term reproduction assay; FSTRA), and one with the amphibian Xenopus laevis (amphibian metamorphosis assay; AMA). Touart noted that, although each of the Tier 1 assays provides unique data, the suite was purposefully designed to result in some redundancy with respect to detecting endocrine pathways of concern.
Anne Gourmelon (Organisation for Economic Co-operation and Development [OECD]) described the critical role of the OECD in providing a forum for the international harmonization of tests used to regulate chemicals. OECD has contributed to the validation of many assays designed to detect EDCs, including several utilized by the EDSP. She also discussed a five-level conceptual framework developed through OECD that has been proposed for EDC screening and testing (OECD 2010). Briefly, level 1 consists of evaluation of existing data, levels 2 and 3 feature assays and activities analogous to Tier 1 of the EDSP, and levels 4 and 5—which are not yet well-defined in terms of application of discrete tests—are comparable to Tier 2 of the EDSP.
The next three presentations dealt with various aspects of WoE analysis of data in the context of defining whether a chemical is an EDC. Keith Solomon (University of Guelph) provided an overview of WoE in the field of toxicology and highlighted the importance of utilizing scientifically credible, transparent approaches for conducting these types of analyses. He then provided a specific example of a WoE analysis for ecological impacts of an herbicide in which different facets of data from multiple studies were semi-quantitatively evaluated as to robustness and quality relative to their contribution to a cumulative final score. The next two presentations focused specifically upon use of the multiple lines of evidence emanating from Tier 1 screens to determine whether a chemical is endocrine-active and, as a consequence, may be a candidate for Tier 2 testing. Ellen Mihaich (Environmental and Regulatory Resources) described a formal framework (Borgert et al 2011), that employed a ranking scoring system for WoE analysis for the EDSP Tier 1 endpoints (Borgert et al 2014). An important aspect of that framework involves differential weighting of endpoints from the various Tier 1 assays relative to their diagnostic effectiveness in indicating perturbation of specific HPG/T pathways of concern. Amy Blankinship (USEPA) provided an overview of the conceptual basis of the WoE approach used by the USEPA to evaluate Tier 1 data, and she presented an example of application of the approach using submitted data for a chemical from the initial round of testing. The approach employs an integrated evaluation of the data using the guiding principles laid out in the 2011 USEPA EDSP WoE guidance document.
Gerald Ankley (USEPA) gave the final presentation of the session and described an analysis comparing the results of the Tier 1 in vivo assays with rats and fish using several model chemicals that impact different AOPs within the HPG axis (Ankley and Gray 2013). He noted the high degree of fidelity between the rat and fish assays with respect to identifying EDCs, and he discussed the possibility that, due to conservation of pathways across species, it might be possible to use two of the existing Tier 1 screens as “gatekeeper” assays for initial screening, the FSTRA and male pubertal assay.
Session 3—Where We Are Now: Tier 2 Testing
Les Touart provided an overview of USEPA efforts relative to the development and validation of Tier 2 tests with non-mammalian species, which was followed by a series of more detailed presentations by experts directly involved in each of four tests recently evaluated by the SAP. Kevin Flynn (USEPA) described the proposed Tier 2 fish test, a medaka extended one-generation–reproduction test (MEOGRT). The medaka test, developed and evaluated primarily by scientists from the U.S. and Japan, starts with F0 adults, proceeds through the entire F1 generation, and provides an option for an entire life cycle exposure of the F2 generation. The test, which features both pathway-specific (e.g., histology) and apical (growth, reproduction) endpoints, has been successfully evaluated using several chemicals expected to impact different HPG pathways. Sigmund Degitz (USEPA) described a Tier 2 amphibian test, the LAGDA (larval amphibian growth and development assay), designed to assess possible risks of both HPG- and HPT-active toxicants. The assay, which also was developed through a joint effort between the U.S. and Japan, is initiated with Nieuwkoop-Faber (NF) stage 8 embryos and proceeds through NF stage 62 (complete metamorphosis). As with the MEOGRT, the LAGDA considers a variety of both apical and more pathway-specific endpoints.
Tim Verslycke (Gradient) described a Tier 2 multi-generation invertebrate test with mysid shrimp. He noted that invertebrates such as insects and crustaceans comprise the great majority of animals on earth and that chemicals have been shown to disrupt development mediated by ecdysones, invertebrate hormones analogous to vertebrate steroid hormones. He also noted that it is unclear how results from Tier 1 screening assays, which focus on vertebrate hormones axes, would trigger use of the Tier 2 invertebrate test. Mary Ann Ottinger (University of Houston) discussed the status of the Tier 2 avian assay, a two-generation protocol with the Japanese quail. The assay features a number of core apical endpoints related to survival, growth, reproduction and, potentially, behavior as well as more pathway-specific endpoints like histology. She discussed unique life history and physiological characteristics of birds relative to other vertebrates in the context of the necessity of a Tier 2 test to forecast EDC risks.
The final two talks of the session focused on (1) practical and logistic issues related to conducting the Tier 2 assays in contract labs, and (2) how data from the assays would or could be used for ecological risk assessment. Hank Krueger (Wildlife International) discussed challenges associated with running the fish, amphibian and mysid exposures, particularly from the standpoint of the generation of large-volume water stocks, ideally without use of solvent. He also noted the need for a different or enhanced expertise mix in terms of personnel compared to what was historically required for aquatic toxicity testing. Allen Olmstead (Bayer CropScience) discussed endpoints collected for Tier 2 tests, questioning the impact they will have on ecological risk assessments. He noted that endpoints such as measures of vitellogenin or gonad histology have not historically been applied to ecological risk assessments, and that there's a need to link these to population or individual level adverse outcomes.
Session 3 concluded with an open discussion concerning finalization of the Tier 2 assays and their ultimate application. The issue initially raised by Olmstead generated a significant amount of discussion, with many in the audience opining that it would be appropriate and timely to (re)consider how mechanistic data could be incorporated into ecological risk assessments. The critical role of AOPs as a basis for providing linkages across biological levels of organization was noted. Another discussion point raised in this session involved the desirability of involving OECD in final validation of the Tier 2 tests in a manner analogous to Tier 1 assays. It was noted that, with the anticipated completion of final protocols for some of the key Tier 2 tests (e.g., MEOGRT, LAGDA), now is the ideal time for this type of engagement.
Session 4—Where Are We With Hazard and Risk Assessment?
As noted above, regulatory assessments for dealing with EDCs have the potential to vary widely across the world. For exmaple, the EU and some other countries (e.g., Brazil) use a hazard-based approach, while other jurisdictions (e.g., US, Canada, Japan) employ risk-based analyses. Peter Matthiessen (independent consultant) provided an overview of the basis for these different basic approaches to regulating EDCs, noting that a departure from the more commonly used risk-oriented approach might be justifiable in two possible scenarios. First, when it is impossible to define a no-effect threshold for a chemical of concern. And second, when exposure for a chemical cannot reliably be determined. He felt that, for EDCs, current information does not clearly indicate that either scenario is routinely applicable, although he did note uncertainties that some in the scientific community have raised relative to undefined low-dose effects of EDCs.
Holly Zahner discussed different statutes and authorities in the U.S., Canada and Japan that support the regulation of EDCs. For example, EDCs in the U.S. are regulated for human health and ecological effects via the Food Quality Protection and Safe Drinking Water acts but have not been examined under the Clean Water or the Toxic Substances Control acts. Many regulatory activities for EDCs are under the purview of the USEPA, but EDCs contained in human and veterinary pharmaceuticals are under the purview of the FDA, which can require that specific types of tests be conducted to support ecological risk assessments.
The next two presentations in the session focused on case studies concerning the utilization of different data sources, including Tier 1 and Tier 1-type data, for assessing potential EDC-associated hazards and risks of three relatively data-rich chemicals: p-nonylphenol (NP), octylphenol (OP) and glyphosate. Katherine Coady (Dow Chemical Company) presented data showing that, although NP and OP clearly possess endocrine (estrogen agonist) activity both in vitro and in vivo, causes of toxicity of these relatively low-potency estrogens in a variety of species likely occur via biological mechanisms and pathways not directly related to endocrine function. This analysis suggests that apical effects observed, for example, in Tier 2 or Tier 2–type tests may not always be reflective of an endocrine mode of action, particularly in the case of industrial compounds that are not designed to have biological activity. Data were presented showing that the adverse effects associated with exposure to NP and OP occur at much higher concentrations than are detected in human biomonitoring studies, and concentrations of NP and OP in the environment are also below levels of concern. The importance of considering potency, critical effect, exposure and risk in the evaluation of endocrine activity and disruption were emphasized. Steven Levine (Monsanto Company) presented Tier 1 screening results and compared those results with functionally equivalent information from regulatory studies and the open literature for glyphosate. The two types of data were in good agreement, indicating no evidence of interaction with estrogen pathways, androgen pathways, steroidogenesis and disruption of the HPG/T axis by the herbicide.
Earl Gray (USEPA) discussed non-monotonic dose response relationships (NMDRs) relative to assessing EDC risks, an issue that has been identified as a concern relative to possible low-dose effects of endocrine-active chemicals. Partially in support of a USEPA position paper on the topic, which is currently being reviewed by the National Academy of Sciences in the US, Gray conducted an extensive analysis of extant mammalian-oriented literature concerning responses to a number of known estrogens, androgens, anti-androgens and inhibitors of steroid synthesis. He concluded:
- NMDRs are not uncommon in in vitro systems due to variables such as overt toxicity of putative EDCs to cells at high test concentrations
- Many NMDRs reported in vivo have not been adequately documented (e.g., statistically) or independently replicated
- Valid in vivo NMDRs can occur in studies with EDCs but typically not in treatments that would be considered low dose
- Little evidence exists supporting the proposal that NMDRs/low dose effects have resulted in inadequate assessments of long-term risks of EDCs
In the final talk, Christopher Borgert (CEHT, University of Florida) presented a modeling-based study to define whether EDCs would be expected to have defined thresholds for eliciting effects (Borgert et al. 2013). His analysis suggested that normal endocrine signaling relies on the existence of thresholds and that safe levels of exposure can be inferred based on consideration of affinity, efficacy, potency and mass action of a chemical against the background and variability of normal endocrine function. This suggests that defining an endocrine hazard requires a consideration of potency and thresholds, and that risk-based approaches to regulating EDCs can be appropriate.
Session 4 concluded with an open discussion concerning the issue of hazard and risk as a basis for regulating EDCs. The main recommendation emanating from this discussion involved the desirability of conducting an open-forum scientific workshop (e.g., using a SETAC Pellston model) designed to directly consider the issue using, for example, a series of case studies with several data-rich EDCs.
Session 5—Where Do We Go From Here: Challenges and Future of EDC Testing
The final session focused on new considerations relative to EDC risks and evolving science in the context of prioritization methods for testing. Gerald LeBlanc (North Carolina State University) presented an overview of an OECD report in which several additional important endocrine-type systems were identified beyond the estrogenic, androgenic, thyroid and steroidogenic pathways that are the primary focus of endocrine screening and testing programs (LeBlanc GA et al. 2012). Signaling pathways of specific concern are associated with the peroxisome proliferator-activated receptors (PPAR) and vitamin D receptors and retinoid receptors. Perturbation of these pathways may be involved in various diseases in humans (e.g., obesity, diabetes, metabolic syndrome) and could affect energy metabolism (growth, reproduction) in wildlife. In a complementary talk Seth Kullman (North Carolina State University) then described newer in vitro and in vivo approaches and endpoints for assessing chemicals that might impact these and other endocrine signaling pathways likely not captured by current Tier 1 screening methods. He noted the importance of utilization of the AOP framework to provide transparent linkages between outputs of screening assays and relevant apical outcomes that result from pathway perturbation.
The next two presentations focused on tools that could enhance prioritization of the large lists of chemicals of potential concern to the EDSP prior to Tier 1 screening. Kevin Crofton (USEPA) discussed the application of in vitro high-throughput (HTP) assays in conjunction with rapid methods to estimate exposure and dose to help prioritize the 10,000+ chemicals expected to be assessed through the EDSP. He described the USEPA ToxCast program which, as part of a much larger suite of assays, includes several assays relevant to the HPG/T axes, particularly in terms of chemical interactions with the nuclear hormone receptors. To date, HTP information has been generated for more than 8,500 chemicals, many of which have been identified as targets for the EDSP. This information, in conjunction with measured or predicted exposure data, provides a logical basis for prioritizing chemicals of most concern. Patricia Schmieder (USEPA) presented an in silico tool designed to predict the potential of a chemical to bind to the estrogen receptor, thereby potentially causing toxicity through defined AOPs. The Estrogen Receptor Expert System was developed using internationally established (OECD) validation principles, and it is well suited to rapidlyassess large chemical datasets to identify those with the greatest potential to bind to the receptor (USEPA 2013).
The final talk in the session focused on the issue of hypothesized epigenetic effects of EDCs. These types of effects, which appear to be heritable from an exposed generation to the next (unexposed) generation through novel chemical modifications of DNA, may not be captured using current screening and testing protocols for EDCs. Miriam Jacobs (Public Health England) provided an overview of the epigenetics issue relative to EDCs and talked about how current test guidelines and test methods could be optimized to detect these effects (Greally and Jacobs 2013).
The final discussion acknowledged the dichotomy that exists in the scientific community regarding the emphasis on hazard-based and risk-based approaches that are being used to assess the effects of EDCs. One of the proposed outcomes of the FTM was a workshop, with SETAC as one of the organizers, which would use selected case studies to compare and contrast assessments based on hazard and risk assessment criteria. Meeting participants strongly expressed the position that EDC assessments should be risk based and incorporate both hazard and exposure assessments. In considering hazard, there was recognition that the endocrine system is dynamic and that changes in endocrine responses need not necessarily reflect an adverse outcome. Rather it was important to consider whether changes in endocrine function were linked to apical endpoints such as growth, reproduction and development. Equally, when assessing the actions of EDCs, it is important to consider exposure and the assessments should be conducted at concentrations that reflect environmentally relevant concentrations. Participants noted the considerable progress that had been made in the development, validation and in some cases application of Tier 1 screens and Tier 2 tests to assess the possible impacts of EDCs. However, there was ongoing concern on the utility of the test battery to be applied to the huge task of screening and testing of the many chemicals for which information on their endocrine activity is lacking. There was strong support to continue the research on streamlining the test battery, and to explore means of reducing the numbers of tests (and subsequently vertebrate animals) required to screen for possible EDCs.
Ankley GT and Gray LE. 2013. Cross-species conservation of endocrine pathways: A critical analysis of Tier 1 fish and rat screening assays with 12 model chemicals. Environmental Toxicology and Chemistry. 32(5): 1084-1087.
Borgert CJ, Mihaich EM, Ortego LS, Bentley KS, Holmes CM, Levine SL, Becker RA. 2011. Hypothesis-driven weight of evidence framework for evaluating data within the USEPA’s Endocrine Disruptor Screening Program. Regul Toxicol Pharmacol. 61(2): 185-191.
Borgert C, Baker SP, Matthews JC. 2013. Potency matters: Thresholds govern endocrine activity. Regul. Toxicol. Pharmacol. 67(1): 83-88.
Borgert CJ, Stuchal, LD, Mihaich EM, Becker RA, Bentley KS, Brausch JM, Coady K, Geter DR, Gordon E, Guiney PD, Hess F, Holmes CM, LeBaron MJ, Levine S, Marty S, Mukhi S, Neal BH, Ortego LS, Saltmiras DA, Snajdr S, Staveley J, Tobia A. 2014. Relevance Weighting of Tier 1 Endocrine Screening Endpoints by Rank Order. Birth Defects Research, Part B: Developmental and Reproductive Toxicology. 101(1): 90-113.
Greally JM, Jacobs MN. 2013. In vitro and in vivo testing methods of epigenomic endpoints for evaluating endocrine disruptors. ALTEX 30(4): 445-471.
LeBlanc GA, Norris DO, Kloas W, Kullman SW, Baldwin WS, Greally JM. 2012. State-of-the-science on novel in vitro and in vivo screening and testing methods and endpoints for evaluating endocrine disruptors. OECD monograph 178; ENV/JM/MONO. Paris (FR).
Organisation for Economic Co-operation and Development. 2010. Workshop Report on OECD Countries Activities Regarding Testing, Assessment and Management of Endocrine Disrupters. Paris (FR).
USEPA Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) Science Advisory Panel (SAP). 2013a. Prioritization of the Endocrine Disruptor Screening Program Universe of Chemicals for an Estrogen Receptor Adverse Outcome Pathway Using Computational Toxicology Tool. FIFRA SAP Meeting; Arlington (VA).
Authors’ contact information: firstname.lastname@example.org, email@example.com, firstname.lastname@example.org and email@example.com
Return to the Globe