Scholarly article on topic 'Risk assessment of endocrine active chemicals: Identifying chemicals of regulatory concern'

Risk assessment of endocrine active chemicals: Identifying chemicals of regulatory concern Academic research paper on "Veterinary science"

CC BY-NC-ND
0
0
Share paper
OECD Field of science
Keywords
{"EU legislation" / "Endocrine disrupting properties" / Weight-of-evidence / Adversity / "Human relevance" / "Lead effect" / Specificity / Potency / "Population relevance"}

Abstract of research paper on Veterinary science, author of scientific article — Remi Bars, Ivana Fegert, Melanie Gross, Dick Lewis, Lennart Weltje, et al.

Abstract The European regulation on plant protection products (1107/2009) (EC, 2009a), the revisions to the biocides Directive (COM[2009]267) (EC, 2009b), and the regulation concerning chemicals (Regulation (EC) No. 1907/2006 ‘REACH’) (EC.2006) only support the marketing and use of chemical products on the basis that they do not induce endocrine disruption in humans or wildlife species. In the absence of agreed guidance on how to identify and evaluate endocrine activity and disruption within these pieces of legislation a European Centre for Ecotoxicology and Toxicology of Chemicals (ECETOC) task force was formed to provide scientific criteria that may be used within the context of these three legislative documents. The resulting ECETOC technical report (ECETOC, 2009a) and the associated workshop (ECETOC, 2009b) presented a science-based concept on how to identify endocrine activity and disrupting properties of chemicals for both human health and the environment. The synthesis of the technical report and the workshop report was published by the ECETOC task force (Bars et al., 2011a,b). Specific scientific criteria for the determination of endocrine activity and disrupting properties that integrate information from both regulatory (eco)toxicity studies and mechanistic/screening studies were proposed. These criteria combined the nature of the adverse effects detected in studies which give concern for endocrine toxicity with an understanding of the mode of action of toxicity so that adverse effects can be explained scientifically. A key element in the data evaluation is the consideration of all available information in a weight-of-evidence approach. However, to be able to discriminate chemicals with endocrine properties of low concern from those of higher concern (for regulatory purposes), the task force recognised that the concept needed further refinement. Following a discussion of the key factors at a second workshop of invited regulatory, academic and industry scientists (ECETOC, 2011), the task force developed further guidance, which is presented in this paper. For human health assessments these factors include the relevance to humans of the endocrine mechanism of toxicity, the specificity of the endocrine effects with respect to other potential toxic effects, the potency of the chemical to induce endocrine toxicity and consideration of exposure levels. For ecotoxicological assessments the key considerations include specificity and potency, but also extend to the consideration of population relevance and negligible exposure. It is intended that these complement and reinforce the approach originally described and previously published in this journal (Bars et al., 2011a,b).

Academic research paper on topic "Risk assessment of endocrine active chemicals: Identifying chemicals of regulatory concern"

Contents lists available at SciVerse ScienceDirect

Regulatory Toxicology and Pharmacology

journal homepage: www.elsevier.com/locate/yrtph

Workshop Report

Risk assessment of endocrine active chemicals: Identifying chemicals of regulatory concern

Remi Barsa, Ivana Fegertb, Melanie Grossc, Dick Lewis d, Lennart Weltjee, Arnd Weyersf, James R. Wheeler d, Malyka Galay-Burgos g'*

a Bayer CropScience, 355 rue Dostoïevski, F-06903 Sophia Antipolis, France b BASF SE, Carl-Bosch-Strasse 38, D-67056 Ludwigshafen, Germany

c WCA Environment Limited, Brunei House, Volunteer Way, Faringdon SN7 7YR, United Kingdom d Syngenta Product Safety, Jealott's Hill International Research Centre, Bracknell, Berkshire RG42 6EY, United Kingdom eBASF SE, Crop Protection - Ecotoxicology, Speyerer Strasse 2, D-67117 Limburgerhof, Germany fBayer CropScience, Building 6620, Environmental Safety - Ecotoxicology, D-40789 Monheim, Germany gECETOC, 4 Avenue E. Van Nieuwenhuyse, B-1160 Brussels, Belgium

ARTICLE INFO

ABSTRACT

Article history: Received 13 June 2012 Available online 23 June 2012

Keywords: EU legislation

Endocrine disrupting properties

Weight-of-evidence

Adversity

Human relevance

Lead effect

Specificity

Potency

Population relevance

The European regulation on plant protection products (1107/2009) (EC, 2009a), the revisions to the bio-cides Directive (C0M[2009]267) (EC, 2009b), and the regulation concerning chemicals (Regulation (EC) No. 1907/2006 'REACH') (EC.2006) only support the marketing and use of chemical products on the basis that they do not induce endocrine disruption in humans or wildlife species. In the absence of agreed guidance on how to identify and evaluate endocrine activity and disruption within these pieces of legislation a European Centre for Ecotoxicology and Toxicology of Chemicals (ECETOC) task force was formed to provide scientific criteria that may be used within the context of these three legislative documents. The resulting ECETOC technical report (ECETOC, 2009a) and the associated workshop (ECETOC, 2009b) presented a science-based concept on how to identify endocrine activity and disrupting properties of chemicals for both human health and the environment. The synthesis of the technical report and the workshop report was published by the ECETOC task force (Bars et al., 2011a,b). Specific scientific criteria for the determination of endocrine activity and disrupting properties that integrate information from both regulatory (eco)tox-icity studies and mechanistic/screening studies were proposed. These criteria combined the nature of the adverse effects detected in studies which give concern for endocrine toxicity with an understanding of the mode of action of toxicity so that adverse effects can be explained scientifically. A key element in the data evaluation is the consideration of all available information in a weight-of-evidence approach. However, to be able to discriminate chemicals with endocrine properties of low concern from those of higher concern (for regulatory purposes), the task force recognised that the concept needed further refinement. Following a discussion of the key factors at a second workshop of invited regulatory, academic and industry scientists (ECETOC, 2011), the task force developed further guidance, which is presented in this paper. For human health assessments these factors include the relevance to humans of the endocrine mechanism of toxicity, the specificity of the endocrine effects with respect to other potential toxic effects, the potency of the chemical to induce endocrine toxicity and consideration of exposure levels. For ecotoxicological assessments the key considerations include specificity and potency, but also extend to the consideration of population relevance and negligible exposure. It is intended that these complement and reinforce the approach originally described and previously published in this journal (Bars et al., 2011a,b).

© 2012 Elsevier Inc. All rights reserved.

Abbreviations: ACR, acute-to-chronic ratio; BfR, Bundesinstitut für Risikobewertung (German regulatory body); CLP, Classification, Labelling and Packaging; CRD, Chemicals Regulation Directorate (UK regulatory body); ECETOC, European Centre for Ecotoxicology and Toxicology of Chemicals; ED, endocrine disruptor; IPCS, International Programme on Chemical Safety; LOAEL, Lowest Observed Adverse Effect Level; MoA, mode of action; MoE, margin of exposure; NOAEL, No Observed Adverse Effect Level; NOEC, No Observed Effect Concentration; REACH, Registration, Evaluation, Authorisation and Restriction of Chemicals; STOT-RE, Specific Target Organ Toxicity-Repeated Exposure; TDI, Tolerable Daily Intake; WHO, World Health Organisation.

* Corresponding author. Fax: +32 2 675 3625.

E-mail address: malyka.galay-burgos@ecetoc.org (M. Galay-Burgos).

0273-2300/$ - see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.yrtph.2012.06.013

1. Introduction

Recent European legislation (Plant Protection Products Regulation 1107/2009; proposed new Biocidal Products Regulation COM[2009]267) (EC, 2009a,b): has created a hazard based approval criterion that only supports the marketing and use of chemicals on the basis that they do not induce endocrine disruption in humans or wildlife species. Substances with endocrine properties are also subject to authorisation under the European regulation on the Registration, Evaluation, Authorisation and Restriction of

Chemicals (REACH; 1907/2006) (EC, 2006). Hence, the regulatory consequences of identifying a substance as an endocrine disrupting chemical are severe. However, the fundamental scientific criteria, necessary to define endocrine disrupting properties, are not described in any of these legislative texts or accompanying guidance.

Consequently, in response to these legislative developments and in absence of regulatory criteria, the European Centre for Eco-toxicology and Toxicology of Chemicals (ECETOC) formed a task force to develop a science-based proposal on how to identify and assess chemicals with endocrine disrupting properties (ECETOC, 2009a1). ECETOC presented this proposal at a workshop of regulatory, academic and industry scientists (Barcelona; June 29-30, 2009) to evaluate the approach as a concept for identifying endocrine disrupting properties within a regulatory context (ECETOC, 2009b1). The proposed guidance was refined following input from the workshop, and was published by Bars et al. (2011a,b). The proposed scientific criteria integrated, in a weight of evidence approach, information from regulatory (eco)toxicity studies and mechanistic/ screening studies. These criteria combined evidence for adverse effects detected in apical whole-organism studies with an understanding of the mode of action (MoA) of endocrine toxicity. Briefly, the first part of the guidance consisted of flow charts describing data combinations of evidence (or absence thereof) that would lead to the determination that a substance had endocrine disrupting properties or not (the reader is referred to Bars et al., 2011a,b for details). In addition, since not all chemicals with endocrine disrupting properties are of equal hazard, an assessment of potency was also proposed as a second step to discriminate chemicals of high concern from those of lower concern (for regulatory purposes). However, the ECE-TOC task force recognised that this second part of the assessment needed further refinement.

A considerable amount of work has also been undertaken by individual EU member states and organisations, which has generated approaches for determining endocrine disrupting properties that have significantly progressed current thinking in this area (ECETOC, 2009a; BfR, 2011). Recognising this, the ECETOC task force hosted a second workshop (Florence; May 9-10, 2011), the aim of which was to evaluate the emerging guidance produced by regulatory authorities and organisations, to identify areas of concordance and difference, to consolidate the common scientific themes, and provide a platform for constructive debate on areas of potential difference. The outcome of that workshop has been published in a separate report (ECETOC, 2011). This paper presents the revisions made to the ECETOC guidance including some contributions from that workshop, and represents the view of the ECE-TOC task force. The focus of this paper is to elaborate on key aspects of the second part of the ECETOC guidance, and it therefore complements the approach previously published in this journal (Bars et al., 2011a,b).

2. Refinements to the ECETOC proposal to identify EDCs of regulatory concern for human health

The criteria proposed by the ECETOC task force (ECETOC, 2009a; Bars et al., 2011a,b) were based on two requisite elements shared by the broadly accepted definitions for endocrine disrupting chemicals (e.g. Weybridge, 1996; EC, 1999; IPCS, 2002; Japanese Ministry of the Environment, 2005), i.e. that exogenous substances need to cause adverse effects in intact organisms and that the adverse effect is caused by an endocrine MoA. In the development of the original guidance Bars et al. (2011) adopted the Weybridge definition. However, since the IPCS Bars et al., 2011a definition is

1 ECETOC reports are available for free download at http://ecetoc.org/publications.

currently the most widely accepted definition and also takes populations into account, the IPCS definition has now also been adopted in this revision to the guidance.

The current primary toxicology test methods for detecting endocrine toxicity in mammals are the standard regulatory OECD studies (e.g. the rodent two-generation reproduction study (TG 416), the extended one-generation reproductive toxicity study (TG 443), the rodent chronic toxicity and oncogenicity studies (TG 451, TG 452, TG 453), and the recently enhanced 28 day toxic-ity study (TG 407)). Evidence for the MoA is best provided by (but not limited to) the recently validated in vitro and in vivo screening studies included in the US EPA Tier 1 endocrine test battery or levels 2-4 of the OECD conceptual framework for the testing and assessment of endocrine disrupting chemicals.

The first part of the ECETOC guidance considers five scenarios to guide the evaluation of available mammalian data to determine whether a substance has endocrine properties. Only one scenario (Scenario C; Bars et al., 2011) describes the data combination that would result in the conclusion that there is sufficient evidence of endocrine disruption. This data combination is met when adverse effects on endocrine relevant endpoints in apical or supporting non-apical in vivo studies are supported by mechanistic data from in vitro or in vivo studies, (i.e. the sequence of the biochemical and cellular events that underlies the adverse effect is described and understood, then conclusive proof of endocrine disruption can be considered as established). The other four scenarios (Scenarios A, B, D and E) describe data combinations from available studies that would result in the conclusion that there is no or insufficient evidence of endocrine disruption, and are discussed in Bars et al. (2011).

The principles of the WHO/IPCS conceptual framework for evaluating MoA for cancer and non-cancer endpoints (Boobis et al., 2006, 2008) should be applied for the weight-of-evidence evaluation of the available data. Briefly, the framework requires a description of the key toxicological events critical to the postulated MoA, followed by confirmation of a dose-response relationship, and a temporal association of the key events and the toxicological response. The strength, consistency and specificity of the effects then need to be determined, and the biological plausibility of the MoA and effects are evaluated. The framework also suggests that other MoAs should be considered as a part of the overall weight of the evidence. If, after applying this framework to the evaluation of the available data, it is established that there is sufficient evidence to determine a substance as an endocrine disrupter, it is then necessary to discriminate chemicals of high regulatory concern from those of lower regulatory concern. This is an important consideration because not all substances, for which there is evidence of endocrine disruption, represent the same hazard to humans. Therefore, they should not all be of equal regulatory concern and subject to the same severe regulatory consequences, such as hazard-based exclusion under the pesticides legislation and authorisation under REACH. This can be illustrated with the example of caffeine, for which relevant adverse effects were observed in in vivo studies and are supported by in vitro mechanistic data. Decreased numbers of copora lutea, implantations and foetuses for F1 females were observed following dosing with caffeine in an apical rat reproduction study (Bradford et al., 1983). In supporting in vivo studies effects such as decreased sperm motility and increase sperm density were recorded in a mouse reproduction study (Gulati et al., 1984) and an increased incidence of resorptions was observed in rat developmental toxicity studies (Bertrand et al., 1965,1970; Palm et al., 1978). In addition, an increased incidence of pituitary adenomas and mammary tumours were found in a 12 month rat chronic study (Yamagami et al., 1983) and in a 43 week mouse study (Welsch et al., 1988), respectively. Whilst no effects were found in an in vitro hER activation study, oestradiol

secretion was affected in an in vitro steroidogenesis (H295R) assay (Tinwell et al., 2011). The available data for caffeine fulfil Scenario C of the ECETOC framework, i.e. there is evidence from apical and supporting in vivo studies of adverse effects on endpoints that are relevant for the assessment of endocrine disruption, which is supported by mechanistic information from an in vitro study which provides a plausible MoA for the adverse effects observed in the in vivo studies. If caffeine were a chemical subject to REACH or the pesticide legislation, and in the absence of further consideration of additional factors such as potency, this chemical would face severe regulatory consequences due to its intrinsic endocrine disrupting properties. This example demonstrates that consideration of additional factors is required to discriminate chemicals of low concern from those that merit regulatory concern, otherwise many chemicals of low concern could potentially be subject to authorisation under REACH or phased out under the pesticide and biocide legislations. It is however relevant to note that even in the absence of potency considerations, caffeine, as a chemical, would fail any risk assessment approach (except for pharmaceuticals). The margin of exposure (MoE) for caffeine is from 15 to 35 between the adverse effects (mammary gland development and pituitary tumours) observed in cancer bioassays (50 mg/kg/day in mice; 120 mg/kg/day in rat) and the mean daily intake for a 70 kg human (240 mg/day). This example demonstrates the highly conservative approach that is taken for the regulation of pesticides and general chemicals where a margin of exposure of at least 100 is required between the No Observed Adverse Effect Level (NOAEL) from the most appropriate animal toxicity study and any human exposure.

A number of factors were proposed in the original ECETOC guidance to discriminate chemicals of low concern from those of higher concern and these are further developed and modified in this paper. A summary overview of these factors for human health assessment is provided in Fig. 1, and is described in more detail in the sections below.

Once it has been established that there is sufficient evidence for an endocrine disrupter of potential concern, the evaluation should proceed with an assessment of whether the endocrine MoA is relevant to humans (Section 2.1). The default assumption is to assume human relevance unless there is good scientific evidence to demonstrate otherwise. The evaluation of human relevance is followed by an assessment of specificity, which is required to determine whether the adverse effects observed occur at dose levels lower

than other forms of toxicity, e.g. neuro-, hepato-, or cardio-toxicity (Section 2.2). If the endocrine effect is not the most sensitive effect in the database, then the substance should not be considered of high regulatory concern for endocrine disruption and the substance should undergo risk assessment based on the most sensitive adverse effect. If the endocrine effect is the most sensitive effect, then the assessment proceeds to an evaluation of potency to define substances as being in one of two categories (Section 2.3). Category 2 (low concern) should undergo standard risk assessment. In addition, it is proposed that an exposure factor should be considered for Category 1 (high concern EDs). If the MoE is greater than 1000 then the Category 1 EDs should undergo normal risk assessment. Only if the MoE is less than 1000 should a substance be considered as an endocrine disrupter of regulatory concern to which the hazard based cut-off criterion should be applied. As a MoE of 100 is considered health protective for other adverse events then this represents a highly conservative approach.

2.1. Human relevance

The first step, after establishing that there is sufficient weight-of-evidence for potential endocrine disruption, is to assess whether the adverse effects observed in in vivo animal studies and the proposed endocrine MoA are relevant to humans. The current default assumption is to assume human relevance. However, this assumption can be re-evaluated if there are scientifically valid data to demonstrate non-relevance to humans. In the case of adverse endocrine mediated effects observed in experimental animals, the most well-known example for non-human relevance of animal data is the susceptibility of the rodent to disruption of thyroid function due to species-specific differences in synthesis, binding, metabolism/clearance and transport of the thyroid hormones. It has been well documented in the literature that certain hormon-ally-induced changes of the thyroid in rodents have little relevance to humans (Ames et al., 1987; Alison et al., 1994).

Other examples of MoA or adverse effects that are not relevant to humans exist. However, there are few well-documented cases where endocrine effects in experimental animals are recognised as not relevant to humans. Furthermore, it is often difficult to prove a negative, i.e. to prove non-relevance to humans. The IPCS frameworks for assessing the relevance of cancer and non-cancer MoA for humans (Boobis et al., 2006, 2008) should be used to evaluate the available evidence.

Weight of evidence for defining ED of concern

Human relevance?

No No regulatory concern for ED

Standard risk assessment

Is the ED effect the most sensitive effect ? No No reg"latory

concern for El

Evaluate potency

1 1 J 1

I ED of high concern ED of low concern^]

MOE < 1000?

I ED of regulatory concern

Standard risk assessment

Fig. 1. Summary overview of refined ECETOC proposal to identify endocrine disrupters of regulatory concern for human health.

If human relevance is discounted based on the evidence, then the substance being evaluated is not of regulatory concern and should undergo standard risk assessment, rather than be subject to the hazard based cut-off criterion. If the observed effects and MoA are deemed to be relevant to humans, or if there are insufficient data to demonstrate non-relevance to humans, then the specificity of the effects needs to be evaluated (Fig. 1).

2.2. Lead toxic effect and specificity

The specificity of an adverse endocrine effect can be assessed by using the lead toxic effect approach, which considers the dose response relationship of all effects observed in the whole toxicity dossier available for a substance. The adverse effect that occurs at the lowest dose is considered the lead toxic effect. The lead toxic effect consequently describes the most sensitive toxicological response and drives the risk assessment of a substance. Any risk management measures based on the lead toxic effect will also be protective of other toxic effects (including endocrine effects) occurring at higher dose levels.

Within a study an endocrine mediated endpoint can be affected at lower, similar or higher doses than those which cause other types of toxicity. A substance should only be considered of regulatory concern, i.e. be subject to the cut-off criterion, when the endocrine mediated effect is the lead toxic effect and occurs at doses lower than those that cause other types of toxicity and a sufficient MoE cannot be established. In the case where the endocrine mediated effect is not the lead effect, i.e. the endocrine effect is observed at higher doses than those causing other toxicity, then the substance should not be considered an endocrine disrupting chemical of regulatory concern, and should not be subject to the cut-off criterion. Rather the substance should then proceed with a risk assessment based on the most sensitive (non-endocrine) lead effect. A factor of up to 10-fold degree of separation between the lead effect and an endocrine effect at higher doses may be considered sufficiently conservative for a substance not to be determined as an endocrine disrupter of regulatory concern. However, each substance should be evaluated on a case by case basis, taking into account the dose response and nature and severity of both the primary lead effect and the endocrine effects.

A hypothetical example describing this principle is depicted in Fig. 2. The available in vivo data for a hypothetical chemical ''Y'' provide evidence of adverse effects on endocrine relevant endpoints, i.e. uterine tumours were observed in a two-year rat carcin-ogenicity study and uterine glandular hyperplasia was noted at the one-year interim sacrifice of the same study. These findings were supported by mechanistic data, which provided a plausible endocrine mechanism for the adverse effects observed. Progesterone and testosterone secretion were affected in an in vitro H295R assay, progesterone and oestradiol were affected in an in vitro steroidogenesis assay using ovarian follicles, and progesterone and oestra-diol hormone levels were found to be perturbed in in vivo hormone studies following single and multiple doses. Based on the first half of the ECETOC guidance there is therefore sufficient evidence for endocrine disruption. The next step is to assess whether the endocrine MoA is relevant to humans in order to determine whether this substance should be considered of regulatory concern. In this hypothetical example human relevance has to be assumed, in the absence of data to the contrary, and the specificity of the endocrine effect is considered. The Lowest Observed Adverse Effect Levels (LOAELs) for the endocrine mediated effects are compared with the LOAELs for other toxic effects observed. In this example, a LOAEL of 35 mg/kg/d was determined for the uterine tumours. However, liver toxicity and liver adenomas were also observed with a LOAEL of 2 mg/kg/d. The endocrine mediated effect was therefore not the most sensitive effect, the degree of separation

between the effects was greater than 10-fold, and the risk assessment should therefore be performed with the NOAEL for the lead toxic effect (liver toxicity/carcinogenicity). It should be noted that the uterine tumors would not be ignored; they would result in an appropriate classification according to European Classification, Labelling and Packaging (CLP) criteria.

A fundamental principle of the lead toxic effect approach is that any risk management measures based on the lead toxic effect will also be protective of other toxic effects (including endocrine effects) occurring at higher dose levels. However, there are certain circumstances in the evaluation of endocrine effects that require careful consideration. When assessing lead effects, special consideration should be given to the (ir)reversibility of effects. For example, if a lead (non-endocrine) effect is found to be reversible, whereas the endocrine mediated effect observed at a higher dose is found to be irreversible, then regulation should be based on the irreversible endocrine mediated effect found at the higher dose, if more relevant. The irreversibility of effects which are manifest in later life stages, but caused due to exposure at critical time windows of development should also be given due consideration when assessing lead toxic effects.

2.3. Potency and exposure

Once it has been established that the endocrine effect is the lead toxic effect and that the nature of adversity, severity and reversibility of the effect has also been considered (ECETOC, 2002), the evaluation proceeds to an assessment of potency to discriminate endocrine disrupters of high (regulatory) concern from those of lower concern so that more potent chemicals will be regulated differently from less potent chemicals.

The potency of a substance is a factor of both the dose level at which adverse effects are caused and the duration required to cause the adverse effects. A substance with a lower NOAEL than another for the same endpoint can be considered intrinsically more potent. Further, a potent substance may cause an adverse effect after a short exposure duration, whereas a less potent substance may require a longer exposure duration to elicit the same effect. In a joint proposal the German Bundesinstitut für Risikobewertung (BfR) and the UK Chemicals Regulation Directorate (CRD) (Joint DE-UK position; BfR, 2011) recommend using an approach to potency that already exists in chemical legislation. In this a combination of dose level and exposure duration that cause specific target organ toxicity is considered. The Joint DE-UK position uses Specific Target Organ Toxicity-Repeated Exposure (STOT-RE) criteria to define potential endocrine disrupters as either endocrine disrupters of high or low regulatory concern. The STOT-RE criteria are discriminatory dose thresholds defined in the European CLP Regulations, and are used to determine whether substances should be identified by hazard classification and be assigned appropriate labelling. The Joint DE-UK position proposes that the dose thresholds for STOT-RE should be used to determine whether or not the hazardous property of''endocrine disruption" should be considered for regulatory purposes. The STOT-RE criteria are summarised in Table 1. The CLP regulations do not provide threshold values for chronic studies, but the Joint DE-UK position proposed chronic threshold values, which are half the sub-chronic values (by applying the sub-chronic to chronic assessment factor of two recommended in the REACH guidance). When a substance causes an adverse endocrine mediated effect at a dose level at or below the thresholds for the application of Category 1 STOT-RE hazard classification, the substance is considered as an endocrine disrupter of high regulatory concern (i.e. subject to regulatory prohibition or restriction depending on the relevant legislation). When a substance causes endocrine mediated effects at a dose level above the thresholds for Category 1 STOT-RE hazard classification, the substance should not be

_Study_|_Finding(s)_

_Multi-endpoint studies (apical, in vivo)_

2 year rat cancer bioassay | Uterine tumours _Supporting studies (non-apical in vivo)_

1 year interim sacrifice | Uterine glandular hyperplasia No indications in shorter term studies of uterine effects

Mechanistic data

Study Result

Estrogen receptor density Negative

In vitro steroidogenesis (H295R) Progesterone & testosterone secretion affected

In vitro steroidogenesis (ovarian follicles) Progesterone & estradiol secretion affected

In vivo hormone studies (single & multiple doses) Progesterone & estradiol plasma levels affected

Immature rat uterotrophic assay No effects

No adverse health effects giving concern to ED activity

ED activity giving concern to ED toxicity

Adverse effects giving concern for endocrine toxicity

Endocrine activity giving concern for endocrine toxicity

Adverse effects giving concern to ED toxicity

No Evidence of ED Activity

Sufficient evidence of ED of concern

Non-relevance of ED MOA to humans not demonstrated

Specificity Assessment

Comparison of LOAELs

Study ED LOAEL Lowest LOAEL (Liver effects)

Rat cancer bioassay 35 mg/kg/day 2 mg/kg/day

Endocrine effect is not the most sensitive end-point-

Perform risk assessment on lowest NOAEL (1 mg/kg/day)

Fig. 2. Lead toxic effect and specificity as discriminators in the assessment of endocrine disrupters of regulatory concern for human health for a theoretical example.

Table 1

Summary of dose thresholds for Category 1 and Category 2 classification for Specific Target Organ Toxicity-Repeated Exposure (STOT-RE).

STOT-RE Cat 1

STOT-RE Cat 2

Sub-acute and other short-term studies (e.g. developmental toxicity studies) Oral 30 mg/kg bw/day 300 mg/kg bw/day

Dermal 60 mg/kg bw/day 600 mg/kg bw/day

Inhalation (vapour) 0.6 mg/l/6 h/day 3 mg/l/6 h/day

Inhalation (dust/mist/fume) 0.06 mg/l/6 h/day 0.6 mg/l/6 h/day

Sub-chronic and other medium

studies) Oral Dermal

Inhalation (vapour) Inhalation (dust/mist/fume)

Chronic studiesa

Dermal

Inhalation (vapour) Inhalation (dust/mist/fume)

-term studies (e.g. 2-generation reproduction

10 mg/kg bw/day 20 mg/kg bw/day 0.2 mg/l/6 h/day 0.02 mg/l/6 h/day

5 mg/kg bw/day 10 mg/kg bw/day 0.1 mg/l/6 h/day 0.01 mg/l/6 h/day

100 mg/kg bw/day 200 mg/kg bw/day 1 mg/l/6 h/day 0.2 mg/l/6 h/day

50 mg/kg bw/day 100 mg/kg bw/day 0.5 mg/l/6 h/day 0.1 mg/l/6 h/day

a There are no guidance values in the CLP Regulations for chronic studies, but the Joint DE-UK position proposes that the guidance values for chronic studies should be half the subchronic study values (by applying the subchronic to chronic extrapolation assessment factor of two recommended in the REACH guidance on information requirements and chemical safety assessment, chapter R8).

considered as an endocrine disrupter of high regulatory concern (i.e. not subject to the severe regulatory consequences). Rather, these substances should be regulated through standard risk assessment and risk management methodologies.

The method proposed in the Joint DE-UK position has merit in being practical and the ECETOC task force considers it to be a reasonable approach. It is a pragmatic solution that in the absence of existing science-based potency factors is at least well founded in existing EU regulation. Therefore, a potency assessment based on STOT-RE criteria, such as that proposed by the Joint DE-UK position, could be used within the ECETOC framework to discriminate

endocrine disrupters of high regulatory concern from those of low regulatory concern. Endocrine disrupters causing effects at a dose level at or below the dose thresholds for the application of Category 1 STOT-RE hazard classification would be considered of high concern subject to regulatory action. Those causing effects above the dose threshold for Category 1 STOT-RE hazard classification would be considered of low concern, and would not be subject to regulatory action, but would proceed to standard risk assessment.

If a substance is considered to be of high concern, the proposed ECETOC approach proceeds to an assessment of exposure. A substance may be a highly potent endocrine disrupter, but if the exposure concentrations are so low that humans are not exposed to an effective dose, then the substance should not be considered of high regulatory concern. A very conservative and health protective MoE of 1000 is proposed for the exposure assessment. The NOAEL for endocrine disrupting effects is compared with the estimated exposure dose for humans, and if this results in a MoE greater than 1000, then the substance should not be considered of high regulatory concern and proceeds to risk assessment. However, if the MoE is less than 1000 then the substance should be determined as an endocrine disrupter of high regulatory concern, to which the cutoff criterion may be applied.

The use of potency and exposure as a discriminating factor in the assessment of endocrine disrupting properties can be illustrated with the example of the non-steroidal oestrogenic myco-toxin, zearalenone (Fig. 3). The available in vivo data and in vitro data provide sufficient evidence of endocrine disruption. In summary, pituitary adenomas were observed in chronic toxicity studies in the mouse (NTP, 1982), whilst pituitary, thyroid and adrenal weights were affected in a rat multi-generation study (Becci et al., 1982). Testicular and seminal vesicle atrophy, as well as hyperplasia of the prostate, mammary gland ducts and endome-trial tissue were found in a 90-day study with rats (NTP, 1982). In pigs, increased inter-oestrus interval, increased plasma progesterone levels, and prolonged maintenance of corpora lutea were found following a dosing period of 15 days (Edwards et al., 1987).

Zearalenone competitively binds to the oestrogen receptor in vitro (Kuiper et al., 1998), affects progesterone and testosterone secretion in in vitro steroidogenesis assays (Frizzell et al., 2011; Yang et al., 2007), and induces increased uterine weight in the utero-trophic assay (Christensen et al., 1965; Ueno et al., 1974). The MoA is considered relevant for humans, and the adverse endocrine effects were the lead toxic effects, occurring at doses lower than other forms of toxicity. The most sensitive NOAEL for adverse effects relevant for the assessment of endocrine disruption was 40 ig/kg bw/day in the 15-day pig study and the corresponding LOAEL was 200 ig/kg bw/dAY. The LOAEL of 200 ig/kg bw/day falls below the oral dose threshold (short-term studies: 30 mg/ kg bw/day) for application of Category 1 STOT-RE hazard classification and the substance is considered as an endocrine disrupter of high regulatory concern. The next step in the evaluation is to consider the MoE. The estimate for average daily intakes of zearale-none in European diets corresponds to approximately 0.03 ig/ kg bw/day (EC, 2000), resulting in an MoE greater than 1000 (NOAEL of 40 ig/kg bw/d divided by the estimated exposure dose of 0.03 ig/kg bw/day = 1333). It is proposed that substances with MoEs greater than 1000 undergo standard risk assessment. The EC Scientific Committee on Food has used exactly this approach to set a Tolerable Daily Intake (TDI) of 0.2 ig/kg bw/day for zearalenone, based on the NOAEL of 40 ig/kg bw/day and an assessment factor of 200 (EC, 2000).

3. Refinements to the ECETOC proposal to identify EDCs of regulatory concern in ecotoxicology

The assessment of endocrine disrupting effects in wildlife species is undertaken separately for aquatic (fish and amphibians) and terrestrial vertebrates (birds and mammals). Assessment for birds and mammals is particularly relevant to the regulation of

plant protection and biocidal products, with a tiered requirement for testing under REACH depending on tonnage bands, whilst the assessment for fish and amphibians is also relevant under a range of other EU regulatory directives (e.g. REACH, water framework directive). The first half of the ECETOC guidance outlines the testing and evidence required to determine whether a substance has endocrine disrupting properties in wildlife species (the reader is referred to Bars et al. 2011 for details). The guidance considers the results from targeted in vitro screens, targeted mechanistic in vivo screening assays, and apical (definitive) and supporting in vivo assays to establish evidence of an adverse population relevant effect with an understanding of the MoA underlying these effects. A holistic evaluation of all the data will be required to assess whether a substance should be regarded as an endocrine disrupter according to the IPCS definition. Therefore, a consistent weight-of-evidence evaluation is required to contend with the varying levels of significance and relevance of the various test types. This need has been re-affirmed in the publication of general weight-of-evidence considerations by the US EPA for evaluation of their Endocrine Disrupter Screening Program results (US EPA, 2011), as well as in the draft OECD Guidance Document on standardised test guidelines for evaluating chemicals for endocrine disruption (OECD, 2011). A similar framework to the WHO/IPCS MoA framework for the assessment of cancer and non-cancer endpoints, which is recommended for the human health assessments, is currently not available for ecotoxicological assessments. However, various guidance documents exist for weight-of-evidence evaluations specific to endocrine disruption (e.g. LRI-EMSG, 2000; Brown et al., 2001; Borgert et al. 2011). These existing methodologies should be evaluated and useful elements combined to assist with developing current regulatory requirements. Two common components of the existing weight of evidence methodologies are: (a) assessment of data quality,

Study 1 Finding(s)

Multi-endpoint studies (apical, in vivo)

Mouse cancer bioassay Pituitary adenomas

Rat multigeneration study Thyroid, pituitary & adrenal wt effects in pups

Supporting studies (non-apical, in vivo)

Rat & mouse: 90 day Endocrine/reproductive tissues affected

Pig: 15 day study Effects on estrus cycle, plasma progesterone, corpora lutea

Mechanistic data

Study Result

Estrogen receptor binding Positive

In vitro steroidogenesis (mouse Leydig cells) Testosterone secretion affected

In vitro steroidogenesis (pig ovarian granulosa cells) Progesterone secretion affected

Immature rat uterotrophic assay Positive

„ . . ED activity giving Adverse effects Endocrine activity Adverse effects ,

effects giving . . , No Evidence of ED

c0ncern t0 ED concern to ED giving concern for giving concern for giving concern to Actiwit

. . toxicity endocrine toxicity endocrine toxicity ED toxicity activity

Potency consideration

Study NOAEL Exposure duration

Pig study 40 pg/kg/day 15 days

Rat multigen 0.1 mg/kg/d 16 weeks

Fig. 3. Potency as a discriminator in the assessment of endocrine disrupters of regulatory concern for human health - illustrated with zearalenone.

e.g. Klimisch assessment (Klimisch et al., 1997), although these categories need to be modified and expanded to be more relevant for ecotoxicological assessments, and (b) a form of weighting based on the relevance of individual studies for the assessment of endocrine disruption.

In the evaluation of the available evidence for environmental assessments, specific consideration needs to be given to the population relevance of any adverse effect (Fig. 4). This is discussed in further detail in Section 3.1. The identification of endocrine disrupting properties is then followed by an assessment of endocrine specificity, to determine whether the adverse effects observed occur at dose/concentration levels equal to or lower than general toxicity. The assessment of specificity is conducted at two levels. The first level is to consider the specificity within a study or taxon. This forms part of the evaluation to determine whether a substance meets the agreed definition of an endocrine disrupter, i.e. causing an adverse effect, secondary (consequent) to changes in endocrine function (Fig. 4, Section 3.2). Once it has been identified that there is sufficient evidence as per the IPCS definition, the evaluation proceeds to the second level, where specificity is considered in relation to endpoints of other taxonomic groups in the same environmental compartment, which may drive the overall risk assessment (Fig. 4, Section 3.2). If the adverse effects are considered not specific at this stage, the substance proceeds to risk assessment based on the non-endocrine endpoint. However, if the adverse effects are specific, then the potency of the substance should be considered (Fig. 4, Section 3.3). The substance proceeds with a risk assessment based on the endocrine endpoint with an assessment factor based on potency, unless exposure is negligible and no risk assessment is required (Fig. 4, Section 3.4). The environmental assessment deviates at this stage from the assessment for human health, i.e. all substances proceed to a risk assessment and there is no initial hazard-based screening or exclusion. This is because there are no readily available threshold values for ecotoxicology in existing legislation (such as the CLP regulations) that are suitable. The CLP threshold values for classification of long-term aquatic hazards can be as high as 100 mg/L or as low as 0.01 mg/L.

3.1. Adversity and population relevance

There are specific differences between the fields of human health and environmental risk assessment that require a different approach to be taken for the assessment of endocrine disrupting effects in wildlife species. In contrast to the human health assessment, the protection goal of environmental assessments is the protection of populations rather than individuals (Suter et al., 1993; EFSA, 2010a). This difference in protection goals between human and environmental risk assessments is particularly important to the determination of endocrine disrupting properties of chemicals. It requires specific consideration of adversity in relation to the population relevance of endpoints measured in ecotoxicological studies.

An adverse effect has been defined as follows: ''A change in the morphology, physiology, growth, development, reproduction, or life span of an organism, system, or (sub)population that results in an impairment of functional capacity, an impairment of the capacity to compensate for additional stress, or an increase in susceptibility to other influences" (IPCS, 2004). This IPCS definition of an adverse effect includes consideration of a population level effect. Since the protection level is set at the population level for environmental assessments, for an effect to be considered adverse it should have the potential to impact at the population level. In order to determine that a substance has endocrine disrupting properties, population relevant endpoints need to be associated with specific endocrine mechanisms of action (Bars et al., 2011a). However, many non-specific endpoints (e.g. growth) may also be of population significance. Therefore, for the assessment of endocrine disruption, the endpoints of interest are adverse population level effects that are a consequence of disturbance of the endocrine system in definitive (apical) tests.

Population relevant effects are those that affect population growth or dynamics. For example: age at first reproduction, size of a reproductive event, frequency of reproductive events, duration of reproductive period, viability of young and sex ratio. Clearly, some of these effects are population relevant and also diagnostic of endocrine modulation (e.g. sex ratio in fish in the absence of gender-dependent mortality). However, some effects are known

Fig. 4. Summary overview of the refined ECETOC proposal to identify endocrine disrupters of regulatory concern for wildlife species.

to be responsive (and even sensitive) to, but not necessarily diagnostic of, endocrine modulation (e.g. fecundity, which can be affected by general toxicity). In such circumstances supporting information within a test (co-occurring diagnostic endpoints, e.g. vitellogenin in the case of fish2) or information from other in vivo testing tiers (diagnostic endpoints from screening assays at similar doses/concentrations) will be required to link the population relevant effect to an endocrine mechanism. This principle could be applied to all taxonomic groups being assessed (fish, amphibians, birds and mammals) with the exception of invertebrates for which there are no widely accepted mechanistic endpoints.

In practical terms, it is useful to consider the population relevance of the different endpoints. As the science and our level of understanding evolve, further linkages to population impact may become established. Currently, there is a general acceptance that mechanistic endpoints in isolation do not necessarily imply adverse effects and so are employed as "signposts" (Hutchinson et al., 2006) in the testing and interpretation logic rather than directly in risk assessment (Knacker et al., 2010). Thus a consideration of endpoints is necessary and this should constantly be re-evaluated to allow for future improvements in our understanding. For example, using approaches such as those described by the Adverse Outcome Pathway concept (Ankley et al., 2010) may help establish robust linkages between mechanistic endpoints and adverse effects.

Regulatory tests that measure endocrine-mediated and non-endocrine-mediated population relevant effects are essentially those at level 5 (potentially also level 4 on an endpoint by endpoint basis) of the OECD's Conceptual Framework for the Testing and Assessment of Endocrine Disrupting Chemicals. For fish and amphibians, these are the fish sexual development test (OECD TG 234), fish life-cycle tests (e.g. OPPTS 850-1500 or medaka multigeneration test currently under development) and the amphibian growth and development test (currently under development). For birds, they are the chronic reproduction test (OECD TG 206) and the avian two generation test (currently under development). For wild mammals, they are the two generation rodent test (OECD TG 416), the extended one generation reproduction study (OECD TG 443) and other tests that may also be useful in the determination of apical population level effects (OECD TG 414, TG 408 and TG 407). The endocrine diagnostic endpoints may come from these studies or from additional in vivo screening assays (levels 3 and 4 of the conceptual framework). As described previously (ECETOC, 2009a; Bars et al., 2011a) for mechanistic endpoints to be considered reliable they should be observed in the absence of systemic toxicity.

Clearly there will be a hierarchy and weight-of-evidence element in interpreting findings for particular substances. Certain clusters of endpoints will be more significant than a single endpoint in isolation, unless the magnitude of change in that finding is very compelling. Therefore, expert judgement will be required to make the link between endocrine-mediated endpoints and population relevant effects. This will inevitably need to incorporate thinking around what constitutes a biologically meaningful effect as opposed to merely a statistically significant effect. For example, very low changes in response variables (e.g. 3%) can be statistically significant (depending on the number of animals/replicates, homogeneity of starting populations, measurability etc.) but may be biologically irrelevant for population maintenance or even individual performance. In studies with many endpoints, e.g. the multi-generation tests, the probability of detecting an "effect" as a type I error

2 This should be considered preliminary since the amphibian metamorphosis assay is a screening test and is not optimised to assess apical effects for use in risk assessment (e.g. large spacing factor between only three test concentrations and apical measures on relatively few individuals).

increases. Therefore, interrogation of the data for biological relevance becomes increasingly important with the number of endpoints measured. Thus, for environmental risk assessments the concept of NOAEL may be a useful tool to help distinguish the magnitude of effect, though it should be recognised that this does not always address population relevance within a study. Another equally important consideration is the question of what constitutes co-occurrence of mechanistic symptoms and population relevant effects, i.e. can they be found at the same concentration or dose suggesting a causal relationship? Further research is required to address these issues.

3.2. Specificity

As described in Section 2.2 for the human health assessment, the specificity of an endocrine effect can be assessed by using the lead toxic effect approach. The lead toxic effect describes the most sensitive ecotoxicological response and drives the risk assessment of a substance. Any risk management measures based on the lead toxic effect will also be protective of other toxic effects (including endocrine effects) occurring at higher dose levels. For environmental assessments the lead toxic effect can be considered within a study/species (or taxon) or amongst different studies/taxa within an environmental compartment.

3.2.1. Specificity within a study

Within a study an endocrine mediated endpoint can be affected at a lower, the same or higher dose/concentration than that causing general toxicity. This should be considered in the first part of the original ECETOC assessment (ECETOC, 2009a), when linking mechanistic and apical in vivo studies to decide whether a substance is a potential endocrine disrupter of regulatory concern or not (Fig. 4).

A substance should only be considered of regulatory concern when the endocrine mediated effect is the lead effect and occurs at concentrations lower than those that cause other significant tox-icity. In the case where the endocrine mediated effect is not the lead effect, i.e. the endocrine effect is observed at a higher concentration than that causing other toxicity, then the substance should be considered of low concern. The substance should then proceed with a risk assessment based on the most sensitive (non-endocrine) endpoint.

A hypothetical example of a potential aromatase inhibiting chemical tested in a fish full life-cycle study can be used to demonstrate the principle. For example, if the chemical induces a sex ratio shift as the most sensitive or lead effect, the substance would be considered of regulatory concern and proceed to the next step in the assessment. However, if the lead effects are early lifestage mortality and decreased growth, with a decrease in reproduction (correlated with a decrease in vitellogenin) observed at higher concentrations, then the substance should proceed to a risk assessment based on the non-endocrine lead effect. This rationale would also apply if the endocrine mediated endpoints occurred at the same concentration as the other systemic effects.

It is possible to quantify this effect by dividing the endocrine-mediated NOEC by the non-endocrine NOEC from chronic apical tests with fish and amphibians. The resulting ratio can be used to indicate if the endocrine endpoints are specific and whether they drive the risk assessment. If the ratio is less than 1, the endocrine mediated effects would be considered specific, as they occur at a concentration lower than that causing other general toxicity, and the substance would proceed to the next step in the evaluation. If the resulting ratio is greater than 1, the endocrine mediated effects would not be considered specific. The substance would proceed to a risk assessment, which would have a built-in margin of safety for the endocrine effects occurring at higher concentrations

than the lead toxic effect. If the ratio equals or is close to one then the substance is causing endocrine effects and general systemic toxicity at the same or similar concentrations. In this case the endocrine effects would not be considered specific. The same approach to quantify specificity is followed for birds and mammals using the endocrine-mediated NO(A)EL and the lowest non-endocrine-mediated NO(A)EL from chronic tests (but only when dosed via the diet; gavage studies are not applicable (EFSA, 2010b)).

The acute-to-chronic ratio (ACR) may also be used as an indicator to determine whether the endocrine mediated effects are specific and warrant further evaluation within the framework or whether they would be covered through a standard risk assessment. The ACR is defined as the LC50 or LD50 from a short-term acute study, divided by the NO(A)EC or NO(A)EL from a long-term chronic study. A small ACR is associated with substances with a general MoA (e.g. narcotics) because the MoA remains the same in acute and chronic exposures. This means that the effects measured in a chronic study are a smaller magnitude of the same type of response (e.g. NOEC for survival versus LC50). If another MoA becomes active, then the ACR is generally larger. For example, ACRs for 17p-oestradiol, ethinyloestradiol and methyltestosterone in fish are reported as 390,000; 150,000 and >1,000,000, respectively (Hutchinson et al., 2003). In contrast, the average ACR for eleven non-polar narcotics was reported as 2.58 (Roex et al., 2000) and an extensive analysis of a large database of substances demonstrated that the ACR was commonly in the range of 4-50, with a slightly broader range of 1-70 covering around 90% of cases (ECE-TOC, 2003). The fish ACR for endocrine effects is calculated by taking the 96-h LC50 and dividing it by the lowest NOEC for an adverse endocrine mediated effect from the apical study (i.e. fish lifecyle study or fish sexual development test). For amphibians there is currently no standardised acute test to derive an LC50 from. Further, the apical growth and development test is still under development by the OECD. Only the amphibian screening assay (OECD 231) is available from which a preliminary3 NOEC can be derived. This preliminary amphibian NOEC could be compared with the fish 96-h LC50 (which is acceptable since acute fish and amphibian toxicity data are highly correlated (Aldrich, 2009)). If the resulting ACR is low, then endocrine specificity of the substance in question is unlikely. Knacker et al. (2010) proposed an ACR of 20 for fish as an indicator of specific effects. Furthermore, if the ACR is 610 then the effects would already be covered by the assessment factor applied in a standard risk assessment.

The ACR principle described above is, in principle, also applicable to birds and mammals. However, in contrast to the aquatic field, the ACR concept is hardly used in terrestrial vertebrate assessments. The ACR for endocrine effects is calculated by dividing the LD50 from the acute avian or mammalian study by the lowest NO(A)EL for the adverse endocrine-mediated effects obtained in the available reproduction study (and in future potentially the bird two-generation study, which is currently under development). Of course both endpoints need to be expressed in corresponding units, i.e. mg/kg bw/day. For mammals, the rat and for birds, mallard or quail will be the likely species available for comparisons. It should also be noted that only chronic studies in which animals are exposed via the diet are applicable; gavage dosing is considered inappropriate for the derivation of population-relevant adverse effects (EFSA, 2010b). There is no standard factor (such as the value of 10 for aquatic organisms) against which to compare the avian or mammal ACR, so that a judgement on the specificity of the effects is not trivial. Therefore, it would be useful to develop an avian and

3 This should be considered preliminary since the amphibian metamorphosis assay is a screening test and is not optimised to assess apical effects for use in risk assessment (e.g. large spacing factor between only three test concentrations and apical measures on relatively few individuals).

mammalian database from which reference values can be derived. In the bird and mammal risk assessment for plant protection products, there is only a factor of two between acute and chronic Toxicity Exposure Ratio (TER) values, but in contrast to the aquatic risk assessment exposure values are not the same.

3.2.2. Specificity across taxa

Once a population relevant endocrine effect has been confirmed as the lead effect within a taxon, then sufficient evidence of endocrine disrupting properties has been demonstrated according to the IPCS definition. The next step in the environmental evaluation is to consider the specificity and lead effect across taxa. Although endocrine effects may be observed as the most sensitive lead effect within one study or organism (generally the taxonomic group), this effect may be accounted for in a risk assessment by more sensitive non-endocrine endpoints observed in other taxonomic groups. Taking a holistic overview of the compartment, the (aquatic or terrestrial) risk assessment allows for a margin of safety that may sufficiently cover endocrine specific effects observed in one of the taxonomic groups considered. Any risk management measures based on the lead toxic effect in one taxon will also be protective of other toxic effects (including endocrine effects) occurring at higher concentration/dose levels in another taxon. For example, one can envisage a scenario where the assessment is driven by one taxonomic group due to a herbicidal MoA of a chemical (Fig. 5). This herbicide may cause oestrogenic effects in fish at relatively high concentrations. However, in the example presented, the most sensitive lead effect is the toxicity to macrophytes at a concentration of the herbicide 30-fold lower than that causing endocrine mediated effects in fish. In this case the herbicide should not be considered of regulatory concern and should proceed to a risk assessment based on the lower endpoint on primary producers (which of course will have an assessment factor of 10 applied). The endocrine effect on fish would not be considered relevant under conditions of safe use described by the risk assessment and the safety margin provided.

3.3. Potency

Once it has been established that the endocrine effect of a substance is the lead effect, both within and across taxa, the evaluation proceeds to an assessment of potency to discriminate endocrine disrupters of regulatory concern from those of low concern. All substances proceed with a risk assessment using assessment factors based on potency (Fig. 4). The size of potency based assessment factors to be used is an area for further development. Potency in the context of endocrine disruption is a measure to estimate at which concentration or dose a chemical can induce an endocrine mediated adverse (population relevant) effect. Hence, this aspect is strongly linked to specificity but aims to be more quantitative. Therefore, the endocrine-mediated NOEC/NO(A)EL needs to be compared with other endpoints. This can be done by calculating various measures, such as assessing the magnitude of the ACR, comparing the potency of the substance to a reference compound (e.g. the natural ligand of interest, such as 17p-oestra-diol for oestrogen receptor mediated effects), consideration of the duration of exposure that is required for an adverse effect to be induced, as well as the number of species in which the adverse effect is demonstrated. These measures are considered in more detail below.

3.3.1. Acute-to-chronic (endocrine) ratio

The use of ACRs has already been discussed in relation to the assessment of specificity in Section 3.2 above. In addition to providing information on the likely specificity of a substance (indicated by a large ACR), the magnitude of the ACR will also provide

o a. -a

1000 -,

NOEC/ECx

"endocrine endpoint"

endpoint used in RA

<r / o*^

Znj* o v<i>

Fig. 5. Endpoints for a hypothetical substance (herbicide). In this case, a factor of 30 exists between the endocrine-mediated fish NOEC and the macrophyte endpoint that drives the risk assessment. The endocrine-mediated endpoint in fish is nonspecific and the substance is not an endocrine disruptor of regulatory concern.

an indication of the potency of a substance. For example, the ACR for a weak endocrine disrupter may be greater than the general ACR range for non-specific acting chemicals (ACR range of 1-70), but it may not be of the magnitude of the endogenous hormones (ACR range of 150,000 to >1,000,000; Hutchinson et al., 2003).

3.3.2. Comparison with a reference compound

A comparison of the endocrine mediated NOEC/NO(A)EL of a chemical with a known endocrine mechanism of action can be made with that of a standard reference compound, e.g. 17p-oestra-diol, testosterone or fadrozole, assuming that these compounds have a potency of one. Then the potency (e.g. oestrogenic, andro-genic or aromatase inhibiting) of a compound can be calculated as the ratio of NOECs/NO(A)ELs. To be able to do this agreed endpoints need to be available for all relevant test systems and reference compounds covering the necessary mechanisms of action. Data generated to validate the (OECD) tests or assays, from which the NOECs/NO(A)ELs are derived from, may be used for this purpose. Finally, categories could be agreed to differentiate potent from non-potent endocrine active chemicals.

3.3.3. Other potency considerations - duration

When similar endpoints can be derived from various studies (e.g. fish reproduction in the fish short term reproduction assay and fish life-cycle test) a comparison can be made if effects occur already in the short-term study or only after longer exposure in the life-cycle study at comparable dose levels. It should be borne in mind that due to the differences in test designs there may be distinct differences in statistical power and resolution of effect levels (spacing factors between treatments) of the tests, hampering

comparisons. Further, in long-term tests the same endpoint is often measured several times and the earlier measurements can be compared with the later measurements (alternatively on a higher level, effects in the first generation can be compared with those in the second generation). As exposure of wildlife species under natural conditions is often rather periodic than continuous, effects that are invoked only after long-term continuous exposure are less likely to occur and hence are of lower concern (i.e. the compound is of lower potency). Further, the occurrence of long-term effects after realistic exposures can be substantiated in specially designed pulse-exposure studies. For example, Knudsen et al. (2011) investigated the uptake and biomarker responses (vitellogenin induction) in fish following pulsed exposure to 17p-oestradiol. Such studies may assist in differentiating chemicals into more and less potent ones. A differentiation should be made in such an evaluation between adverse population-relevant endpoints and diagnostic (bio)markers.

3.3.4. Other potency considerations - number of species affected

When endocrine-mediated effects occur in multiple species, the potency is considered higher than when the effect occurs in only one species. Effects should be related to a single endocrine MoA. In the toxicology database various mammal species may be available (e.g. rat, mouse, rabbit, dog) while in ecotoxicology fish, bird and possibly amphibian studies are available. The differences in exposure route - birds and mammals with oral exposure (including potential metabolism) and fish and amphibians with whole body (water) exposure, thereby largely bypassing metabolism, should be considered when comparing species.

3.4. Negligible exposure

According to the new Eurpean Union regulation for plant protection products (1107/2009) (EC, 2009a) a substance, which is considered to have endocrine disrupting properties, can be approved if exposure of wildlife species to that substance under realistic proposed conditions of use is negligible. A similar derogation is found in the proposed revision of the regulation for biocidal products (2009/0076 (COD)4 (EC, 2009b). In the proposed biocide regulation exposure is considered negligible for both humans and the environment ''in particular where the product is used in closed systems or strictly controlled conditions". Whilst negligible exposure is also defined for human health assessments in the new plant protections products regulation, there are currently no specified criteria for ''negligible exposure of wildlife species" to plant protection products. Based on the wording in the regulation it is evident that negligible exposure must fall somewhere between ''no exposure" (i.e. nominal concentrations of 0, or less than the limit of detection/limit of quantification) and a concentration representing an acceptable or low risk. Passing the usual risk assessment trigger is not sufficient. A differentiation is made between acceptable exposure that leads to the determination of acceptable risk in the assessment and negligible exposure.

Consideration of negligible exposure should focus on exposure of the organism group for which endocrine disruption has been demonstrated in earlier stages of the assessment. Exposure of other organism groups (e.g. in a different environmental compartment) at higher levels is not relevant in this context. This links with the consideration of specificity amongst taxa discussed in Section 3.2.

Negligible exposure can be defined by consideration of the application type or through the use of assessment factors during the risk assessment. Qualitative information on application methods can be used to define negligible exposure, since application

http://register.consilium.europa.eu/pdf/en/11/st05/st05032-re02.en11.pdf.

scenarios exist in which exposure would be expected to be negligible, e.g. application methods that avoid exposure of non-target wildlife species. The use of baits and feeders are examples that avoid such exposures.

Negligible exposure could also be assessed within the risk assessment for a substance. Thus, for substances with potential endocrine disrupting properties the margin of safety should be considered. For example, if there is a large margin between the NO(A)EC/NO(A)EL for endocrine effects and the exposure value (i.e. a high toxicity exposure ratio), then this may be considered as negligible risk of potential effects. Alternatively, it is possible to use additional assessment factors for endocrine endpoints to derive ''negligible exposure'' from ''acceptable exposure/low risk''. It is evident that clear criteria for negligible exposure of non-target organisms to be used under the plant protection products legislation needs to be developed and agreed.

4. Summary and conclusions

The guidance proposed by ECETOC (Bars et al., 2011a) provides a structured, science based framework to evaluate results from a variety of apical, mechanistic and screening toxicity studies. It integrates knowledge of adverse effects and MoA from these studies to reach a conclusion regarding the endocrine disrupting properties of substances, in accordance with the IPCS and other related definitions. This paper proposes refinements mainly to the second part of the original ECETOC guidance, in order to discriminate between chemicals of low concern from those of higher concern (for regulatory purposes). The concepts of lead toxic effect and potency, which also take into account the nature, severity and reversibility of the adverse effects, are proposed in order to bring the hazard-based cut-off criterion for endocrine disrupters more in line with scientific principles, so that potent endocrine disrupters will be regulated more stringently than the less potent. The revised guidance presents a consistent approach, and shares some elements with guidance proposed jointly by the German BfR and UK CRD, and the German Umweltbundesamt.

For human health assessments there are empirical well-defined threshold values in the CLP regulations that can be used to discriminate between those chemicals of high regulatory concern subject to strict regulatory action from those of lower regulatory concern. However, it should be noted that the additional criteria such as potency are not proposed to define whether a substance is an endocrine disrupter or not. Substances of lower regulatory concern would still be regarded as endocrine disrupters, but rather than being subject to hazard-based prohibition such substances would undergo risk assessment. For environmental assessments there are no similar thresholds in the CLP regulations that are suitable for potency discrimination of endocrine disrupting properties. Therefore, for environmental assessments all substances proceed to risk assessment with assessment factors based on potency.

For human health assessments the debate should now move from developing the concept to running many case studies through the revised framework. For ecotoxicological assessments further development is needed for the criteria proposed to allow them to be fully operational in a regulatory context. ECETOC hopes that the criteria proposed in this paper will contribute to the ongoing development of regulatory guidance under the relevant legislations.

Conflict of interest statement

The authors declare that there are no conflicts of interest. Melanie Gross' involvement in the production of the manuscript was supported by ECETOC.

Acknowledgment

The ECETOC task force is grateful for the expert input received from the invited participants at the workshop held in Florence on May 9-10, 2011, which the task force has used to refine its original guidance.

References

Aldrich, A., 2009. Empfindlichkeit von Amphibien gegenüber Pflanzenschutzmittel. AGRARForschung 16, 466-471.

Alison, R.H., Capen, C.C., Prentice, D.E., 1994. Neoplastic lesions of questionable significance to humans. Toxicol. Pathol. 22,179-186.

Ames, B.N., Magaw, R., Gold, L.S., 1987. Ranking possible carcinogenic hazards. Science 236, 271-280.

Ankley, G.T., Bennett, R.S., Erickson, R.J., Joff, D.J., Hornung, M.W., Johnson, R.D., Mount, D.R., Nichols, J.W., Russom, C.L., Schmieder, P.K., Serrano, J.A., Tietge, J.E., Villeneuve, D.L., 2010. Adverse outcome pathways: A conceptualframework to support ecotoxicology research and risk assessment. Environ. Toxicol. Chem. 29, 730-741.

Bars, R., Broeckaert, F., Fegert, I., Gross, M., Hallmark, N., Kedwards, T., Lewis, D., O'Hagan, S., Panter, H.G., Weltje, L., Weyers, A., Wheeler, J.R., Galay-Burgos, M., 2011a. Science based guidance for the assessment of endocrine disrupting properties of chemicals. Regul. Toxicol. Pharmacol. 59, 37-46.

Bars, R., Broeckaert, F., Fegert, I., Gross, M., Hallmark, N., Kedwards, T., Lewis, D., O'Hagan, S., Panter, H.G., Weltje, L., Weyers, A., Wheeler, J.R., Galay-Burgos, M., 2011b. Corrigendum to "Science based guidance for the assessment of endocrine disrupting properties of chemicals'' [Regul. Toxicol. Pharmacol. 59 (2011) 37-46].. Regul. Toxicol. Pharmacol. 60, 411-412.

Becci, P.J., Johnson, W.D., Hess, F.G., Gallo, M.A., Parent, R.A., 1982. Combined two-generation reproduction-teratogenesis study of zearalenone in the rat. J. Appl. Toxicol. 2, 201-206.

Bertrand, M., Schwam, E., Frandon, A., Vague, A., Alary, J., 1965. Sur un effet tératogène systématique et spécifique de la caféine chez les rongeurs. C.R. Soc. Biol. 159, 2199-2202.

Bertrand, M., Girod, J., Rigaud, M.E., 1970. Electrodactylie provoquée par la caféine chez les rongeurs, rôle des facteurs spécifiques. C.R. Soc. Biol. 164,1488-1489.

BfR (Bundesinstitut für Risikobewertung), 2011. Final Joint DE-UK position. Regulatory definition of an endocrine disrupter in relation to potential threat to human health. Available at: <http://www.bfr.bund.de/cm/349/regulatory_ definition_of_an_endocrine_disrupter_in_relation_to_potential_threat_to_ human_health.pdf> (accessed 08.06.11.).

Boobis, A.R., Cohen, S.M., Dellarco, V., McGregor, D., Meek, M.E., Vickers, C., Willcocks, D., Farland, W., 2006. IPCS framework for analyzing the relevance of a cancer mode of action for humans. Crit. Rev. Toxicol. 36, 781-792.

Boobis, A.R., Doe, J.E., Heinrich-Hirsch, B., Meek, M.E., Munn, S., Ruchirawat, M., Schlatter, J., Seed, J., Vickers, C., 2008. IPCS framework for analyzing the relevance of a noncancer mode of action for humans. Crit. Rev. Toxicol. 38, 8796.

Borgert, C.J., Mihaich, E.M., Quill, T.F., Marty, M.S., Levine, S.L., Becker, R.A., 2011. Evaluation of EPA's Tier 1 endocrine screening battery and recommendations for improving the interpretation of screening results. Regul. Toxicol. Pharmacol. 59, 397-411.

Bradford, J.C., Caldwell, J.A., Barbolt, T.A., Drobeck, H.P., 1983. Chronic administration of caffeine to two generations of rats. Teratology 27, 32A.

Brown, R.P., Greer, R.D., Mihaich, E.M., Guiney, P.D., 2001. A critical review of the scientific literature on potential endocrine-mediated effects in fish and wildlife. Ecotoxicol. Environ. Saf. 49,17-25.

Christensen, C.M., Nelson, G.H., Mirocha, C.J., 1965. Effect on the white rat uterus of a toxic substance isolated from Fusarium. Appl. Microbiol. 13, 653-659.

EC, 1999. European Commission endocrine disrupters definitions homepage. Available at: <http://ec.europa.eu/environment/endocrine/definitions/endodis_ en.htm> (accessed 8.10.09.).

EC, 2000. Health and Consumer Protection Directorate-General. Opinion of the Scientific Committee on Food on fusarium toxins Part 2: Zearalenone. Directorate C - Scientific Opinions SCF/CS/CNTM/MYC/22 Rev 3 Final.

EC, 2006. Regulation (EC) No 1907/2006 of the European Parliament and of the Council of 18th December 2006 concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) establishing a European Chemicals Agency, amending Directive 1999/45/EC and repealing Council Regulation (EEC) No 793/93 and Commission Regulation (EC) No 1488/94 as well as Council Directive 76/769/EEC and Commission Directives 91/155/EEC, 93/67/EEC, 93/105/EC and 2000/21/EC. Official Journal of the European Union L 396/1.

EC, 2009a. Regulation (EC) No 1107/2009 of the European Parliament and of the Council of 21 October 2009 concerning the placing of plant protection products on the market and repealing Council Directives 79/117/eEc and 91/414/EEC. Official Journal of the European Community, L 309/1.

EC, 2009b. Regulation (COM 2009/267) of the European Parliament and of the Council of 12th June 2009 replacing Directive 98/8/EC concerning the placing of biocidal products on the market. Official Journal of the European Community, L 276.

ECETOC, 2002. Recognition of, and differentiation between, adverse and non-adverse effects in toxicology studies. Brussels, European Centre for the Ecotoxicology and Toxicology of Chemicals.

ECETOC, 2003. Aquatic hazard assessment II. Technical Report No. 91. Brussels. ISSN-0773-8072-91.

ECETOC, 2009a. Guidance on identifying endocrine disrupting effects. Technical Report No. 106. Brussels. ISSN-0773-8072-106. Available at: <http://bit.ly/ ecetoc-tr106>.

ECETOC, 2009b. Workshop: Guidance on identifying endocrine disrupting effects, 29-30 June 2009, Barcelona. Workshop Report No. 16. Brussels. Available at: <http://bit.ly/ecetoc-wr16>.

ECETOC, 2011. Workshop: Risk assessment of endocrine disrupting chemicals, 9-10 May 2011, Florence. Workshop Report No. 21. Brussels. Available at: <http:// bit.ly/ecetoc-wr21 >.

Edwards, S., Cantley, T.C., Rottinghaus, G.E., Osweiler, G.D., Day, B.N., 1987. The effects of zearalenone on reproduction in swine I. The relationship between ingested zearalenone dose and anestrus in non-pregnant, sexually mature gilts. Theriogenology 28, 43-49.

EFSA (European Food Safety Authority), 2010a. "Scientific Opinion on the development of specific protection goal options for environmental risk assessment of pesticides, in particular in relation to the revision of the Guidance Documents on Aquatic and Terrestrial Ecotoxicology (SANCO/3268/ 2001 and SANCO/10329/2002)1.'' EFSA J. 8, 1-55.

EFSA (European Food Safety Authority), 2010b. ''Guidance of EFSA: Risk Assessment for Birds and Mammals (Question No EFSA-Q-2009-00223)." EFSA J. 7, 1438.

Frizzell, C., Ndossi, D., Verhaegen, S., Dahl, E., Eriksen, G., S0rlie, M., Ropstad, E., Muller, M., Elliott, C.T., Connolly, L., 2011. Endocrine disrupting effects of zearalenone, alpha- and beta-zearalenol at the level of nuclear receptor binding and steroidogenesis. Toxicol. Lett. 206, 210-217.

Gulati, D.K., Russell, V.S., Hommel, L., Poonacha, K.B., Sabharwal, P.S., Lamb, J.C. IV., 1984. Caffeine: reproduction and fertility assessment in CD-1 mice when administered in drinking water. National Toxicology Program. Technical Report NTP-85-097.

Hutchinson, T.H., Ankley, G.T., Segner, H., Tyler, C.R., 2006. Screening and testing for endocrine disruption in fish - biomarkers as signposts not traffic lights in risk assessment. Environ. Health Perspect. 114,106-114.

Hutchinson, T.H., Yokota, H., Hagino, S., Ozato, K., 2003. Development of fish tests for endocrine disruptors. Pure Appl. Chem. 75, 2343-2353.

IPCS (International Programme on Chemical Safety), 2002. Global assessment of the state-of-the-science of endocrine disruptors. WHO/IPCS/EDC 02.2. Available at: <http://www.who.int/ipcs/publications/new_issues/endocrine_disruptors/en/> (accessed 8.10.09.).

IPCS (International Programme on Chemical Safety), 2004. IPCS Risk assessment terminology. Geneva, World Health Organisation. Available at: <http:// www.who.int/ipcs/methods/harmonization/areas/ ipcsterminologyparts1and2.pdf> (accessed 25.11.11.).

Japanese Ministry of the Environment, 2005. MOE's Perspectives on Endocrine Disrupting Effects of Substances -ExTEND 2005. Available at: <http:// www.env.go.jp/en/chemi/ed/extend2005_full.pdf> (accessed 8.10.09.).

Klimisch, H.J., Andreae, M., Tillmann, U., 1997. A systematic approach for evaluating the quality of experimental toxicological and ecotoxicological data. Regul. Toxicol. Pharmacol. 25, 1-5.

Knacker, T., Boettcher, M., Frische, T., Rufli, H., Eigeler, T.M., Stolzenberg, H.C., Zok, S., Braunbeck, T., Schäfers, C., 2010. Environmental effect assessment for sexual

endocrine-disrupting chemicals: fish testing strategy. Integr. Environ. Assess. Manag. 6, 653-662.

Knudsen, J.J.G., Holbech, H., Madsen, S.S., Bjerregaard, P., 2011. Uptake of 17ß-estradiol and biomarker responses in brown trout (Salmo trutta) exposed to pulses. Environ. Pollut. 159, 3374-3380.

Kuiper, G.G., Lemmen, J.G., Carlsson, B., Corton, J.C., Safe, S.H., van der Saag, P.T., van der Burg, B., Gustafsson, J.A., 1998. Interaction of estrogenic chemicals and phytoestrogens with oestrogenic receptor beta. Endocrinology 139, 42524263.

LRI-EMSG, 2000. Towards the establishment of a weight of evidence approach to prioritising action in relation to endocrine disruption: Working paper as a basis for discussion. CEFIC Long Range Research Initiative - Endocrine Modulators Steering Group, Brussels.

NTP, 1982. National Toxicology Program Technical Report on the Carcinogenesis Bioassay of Zearalenone in F344/N rats and B6C3F1 mice. NIH Publication No. 83-1791, US Department of Health and Human Services, Research Triangle Park, NC, USA.

OECD, 2011. Draft Guidance Document on Standardised Test Guidelines for Evaluating Chemicals for Endocrine Disruption. Version 11 (May 2011). Available at: <http://www.oecd.org/document/12/0,3343,en_2649_34377_ 1898188_1_1_1_1,00.html> (accessed 1.9.11.).

Palm, P.E., Arnold, E.P., Rachwall, P.C., Leyczek, J.C., Teague, K.W., Kensler, C.J., 1978. Evaluation of the teratogenic potential of fresh-brewed coffee and caffeine in the rat. Toxicol. Appl. Pharmacol. 44, 1-16.

Roex, E.W.M., van Gestel, C.A.M., van Wezel, A.P., van Straalen, N.M., 2000. Ratios between acute aquatic toxicity and effects on population growth rates in relation to toxicant mode of action. Environ. Toxicol. Chem. 19, 685-693.

Suter II, G. W, Barnthouse, L.W., Bartell, S.M., Mill, T., Paterson, S., 1993. Ecological Risk Assessment. CRC Press, Boca Raton, Florida.

Tinwell H., Colombel S., Bars R., 2011. Evaluation of substances routinely taken in everyday life in the H295R steroidogenesis assay. Poster (abstract no. 2351). US Society of Toxicology, Washington March 2011.

Ueno, Y., Shimada, N., Yagasaki, S., Enomoto, M., 1974. Toxicological approaches to the metabolites of Fusaria VII. Effects of zearalenone on the uteri of mice and rats. Chem. Pharm. Bull. (Tokyo) 22, 2830-2835.

US EPA, 2011. Endocrine Disruptor Screening Program. Weight-of-evidence: Evaluation results of EDSP Tier 1 screening to identify the need for Tier 2 testing. September 14,2011. Available at: <http://www.epa.gov/endo/> (accessed 25.11.11.).

Yamagami, T., Handa, H., Takeuchi, J., Munemitsu, H., Aoki, M., Kato, Y., 1983. Rat pituitary adenoma and hyperplasia induced by caffeine administration. Surg. Neurol. 20, 323-331.

Yang, J., Zhang, Y., Wang, Y., Cui, S., 2007. Toxic effects of zearalenone and a-zearalenol on the regulation of steroidogenesis and testosterone production in mouse Leydig cells. Toxicol. in vitro 21, 558-565.

Welsch, C.W., Dehoog, J.V., O'Connor, D.H., 1988. Influence of caffeine consumption on carcinomatous and normal mammary gland development in mice. Cancer Res. 48, 2078-2082.

Weybridge, 1996. European Workshop on the impact of endocrine disrupters on human health and wildlife. 2-4 December 1996, Weybridge, UK. Report of Proceedings EUR 17549 Copenhagen, Denmark: European Commission DG XII, April 16, 1997). Available from: European Environment Agency, Kongens Nytorv 6, DK-1050 Copenhagen K, Denmark.