Ecotoxicology and Environmental Safety 167 (2019) 513–519

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

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

Endpoint sensitivity in Amphibian Metamorphosis Assay T

ZhiChao Dang
National Institute for Public Health and the Environment (RIVM), A. van Leeuwenhoeklaan 9, 3720 BA Bilthoven, The Netherlands

A R T I C LE I N FO A B S T R A C T

Keywords: The Amphibian Metamorphosis Assay (AMA) is a screening test for detecting chemicals with thyroid activity.
HPT axis There is little experience in data interpretation and in using AMA data for screening, testing and identifying
Thyroid endocrine disruptors. To investigate the sensitivity of different endpoints of the AMA, the publically available
Endocrine disruptor data for 57 thyroid active and inactive chemicals were compiled and analyzed. Endpoints body weight and
Screening
length appeared as sensitive as apical thyroid responsive endpoints hind limb length (HLL) and developmental
Testing strategy
Risk assessment
stage (DS) for 12 thyroid active chemicals. The sensitivity of body weight, length and HLL was comparable,
which is higher than that of DS for 45 thyroid inactive chemicals. The decision logic of the AMA suggests that an
advanced development alone indicates thyroid activity. The analysis here showed that advanced development at
day 7 could indicate thyroid activity of a chemical. However, advanced development at day 21 may be influ-
enced by thyroid inactive chemicals. Among 39 thyroid inactive chemicals, which affected one or more end-
points, 33% and 77% induced changes in HLL and/or DS at day 7 and 21, respectively; only 10% influenced
thyroid histology. These results showed that apical thyroid responsive endpoints HLL and DS are influenced by
thyroid active chemicals as well as thyroid inactive chemical. Both endpoints should be combined with thyroid
histology for the identification of thyroid active chemicals. The use of the AMA in a testing strategy to identify
chemicals with thyroid activity is discussed.

1. Introduction
Identification of Endocrine Disrupting Chemicals (EDCs) is an im-
portant issue under several pieces of European Union (EU) legislation,
including the industrial chemicals regulation Registration, Evaluation,
Authorization and restriction of Chemicals, (REACH, EC, 1907/2006),
the Plant Protection Products Regulation (PPPR, EC, 1107/2009), and
the Biocidal Products Regulation (BPR, EC, 528/2012). Recently, the
European Commission (EC) adopted the criteria for identifying EDCs for
pesticides and biocides (EC, 2017), which were developed on the basis
of the WHO definition of EDCs (WHO, 2002). The identification of EDCs
is mainly based on the available experimental evidence, which should
address three elements: 1) chemical-induced adverse effects (adversity),
2) chemical-specific endocrine modes of action (MOAs), and 3) the
causal or plausible relationship (causality/plausibility) between ad-
verse effects and endocrine MOAs (EC, 2016, 2017). Guidance on
identification of EDCs was published by ECHA/EFSA in July 2018
(ECHA/EFSA, 2018).
Available tests for the detection of chemicals interfering with en-
docrine pathways are listed in the OECD conceptual framework (CF).
These tests are organized in five levels (OECD, 2012). Level 1 includes
existing data and non-testing information. Level 2 contains in vitro
assays providing data about selected endocrine MOAs. Levels 3–5 in-
clude in vivo assays providing data on adverse effects and/or on endo-
crine MOAs (OECD, 2012). Because of the limited availability of test
guidelines (TGs), current tests focus only on chemicals interfering with
the function of EATS (Estrogen, Androgen, Thyroid, and Steroidogen-
esis) pathways. In the CF, available tests for detecting the thyroid
pathway contains only in vivo amphibian and mammalian tests whereas
for EAS pathways in vitro tests are also available. Two amphibian as-
says, the Amphibian Metamorphosis Assay (AMA, TG231, OECD, 2009)
and the Larval Amphibian Growth & Development Assay (LAGDA,
TG241, OECD, 2015) are available for the detection of effects on the
thyroid pathway at levels 3 and 4, respectively (OECD, 2012). Both
tests have not yet been included in the standard information require-
ments of REACH, PPPR and BPR, but may be requested on a case-by-
case basis when there is a concern for a chemical with potential thyroid
activity (Dang et al., 2016).
An OECD test guideline for the AMA (OECD TG 231; EPA OCSPP
Guideline 890.1100) was adopted in 2009, which intends to identify
chemicals interfering with the thyroid pathway or the function of hy-
pothalamic-pituitary-thyroid (HPT) axis (OECD, 2009). The biological
basis for this assay is that amphibian metamorphosis is triggered by
thyroid hormone (TH) and controlled by the HPT axis (OECD, 2006;

E-mail address: zhichao.dang@rivm.nl.

https://doi.org/10.1016/j.ecoenv.2018.10.028
Received 24 April 2018; Received in revised form 8 August 2018; Accepted 8 October 2018
Available online 26 October 2018
0147-6513/ © 2018 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/BY/4.0/).
Z. Dang
Coady et al., 2014). The hypothalamus produces thyrotropin-releasing
hormone (TRH), which stimulates the production of thyroid stimulating
hormone (TSH) in the anterior pituitary. TH is produced by the folli-
cular cells of the thyroid gland and is regulated by TSH in response to
feedback from circulating thyroid TH (EFSA, 2018). The AMA is a
screening assay which detects chemical interaction with thyroid-
mediated metamorphosis in Xenopus tadpoles. At Nieuwkoop and Faber
(NF) stage 51, premetamorphic tadpoles were exposed to at least three
concentrations of a test chemical plus controls. Following 21 days of
exposure, tadpoles are expected to develop through prometamorphosis
(at which time the thyroid gland matures) and potentially reach or
exceed metamorphic climax (NF stages 58–66). The observational
endpoints of this assay are hind limb length (HLL), snout to vent length
(SVL), developmental stage (DS) determined by using prominent mor-
phological landmarks, wet body weight, thyroid gland histology, and
daily observations of mortality. It is often considered that HLL, DS, and
thyroid gland histology are thyroid responsive endpoints and that
changes in these endpoints indicate a chemical interfering with the HPT
axis; whereas mortality, SVL and wet body weight serve as endpoints of
tadpole growth and health (OECD, 2009; Coady et al., 2014). Except
thyroid gland histology, the other endpoints are the apical endpoints of
the AMA (OECD, 2009). The apical endpoints are observed on study day
7 and day 21. It has been established that thyroid hormones (THs) in-
duce precocious development in premetamorphic tadpoles (OECD,
2006). Data on the advanced development indicated by two apical
endpoints HLL and DS, observed at day 7 are collected primarily to
indicate TH agonists. Data collected at day 21 can indicate both TH
agonists and antagonists (OECD, 2009).
Information on thyroid activity of a chemical is not only important
for identification of EDCs but also for triggering further testing. Under
REACH, for example, disturbance of the thyroid pathway is used as a
trigger to include the cohorts for developmental neurotoxicity and
immunotoxicity in the extended one generation toxicity study (ECHA,
2016; OECD, 2012a). According to the ECHA/EFSA guidance docu-
ment, AMA is considered as an important assay in the decision tree
indicating thyroid activity of a chemical "sufficiently investigated"
(ECHA/EFSA, 2018). There is limited experience in using the AMA data
for screening and testing potential EDCs in the regulatory field. Due to
the complexity of the endocrine regulation of growth and development
of tadpoles, many questions have been raised for the data interpretation
of this assay, including whether the AMA can detect all chemicals with
thyroid activity; whether thyroid responsive endpoints are specific for
chemicals with thyroid activity; whether changes in one of the thyroid
responsive endpoints should be considered as the thyroid concern for
further testing in the regulatory field. To answer these questions,
publicly available AMA data were compiled for a systematic analysis of
endpoint sensitivity of chemicals with and without potential to interact
with the HPT axis. This paper focuses on the use of these AMA data for
screening, testing, identification and risk assessment of EDCs.
2. Materials and methods
This paper used data from the OECD validation tests and the US EPA
EDSP (Endocrine Disruptor Screening Program) Tier 1 screening assays
for 52 chemicals (EPA, 2015; OECD, 2008). Publications were also
searched from PubMed by considering amphibian exposure stages, the
chemical exposure period for 21 days, the number of treatment con-
centrations (at least two exposure concentrations), the endpoints re-
ported (at least two endpoints including thyroid histology), mortality
(less than 10%), and data quality. Four papers have been found ac-
cording to the above selection criteria and the results of these papers
were included in the analysis. Results are compiled in Table 1. Con-
centrations were indicated by nominal concentrations reported in dif-
ferent studied. The molecular target of a chemical was also indicated
when information is available. It is noted that the majority of the stu-
dies compiled were performed according to the test guideline. The
514
Ecotoxicology and Environmental Safety 167 (2019) 513–519
number of animals in these studies was often 20 tadpoles.
Analysis for endpoint sensitivity is based on our previous approach
(Dang et al., 2011). In brief, responses of each endpoint at different
concentrations of the same study are divided into three effective cate-
gories:
1. Endpoint responses at the lowest observed effect concentration
(LOEC, marked with a red color in Table 1). This LOEC is different
from the lowest tested concentration when no effects have been
observed
2. Endpoint responses at concentrations above the LOEC, but equal to
or lower than the maximum tested concentration within the same
study (yellow color in Table 1). In the case of non-monotonic re-
sponses, endpoint responses at the lowest effective concentrations
were considered, which may fall into LOEC or this group. and
3. Absence of endpoint responses at all tested concentrations (white/
blank color in Table 1). Some studies did not report the effects on
the endpoints. These endpoints were indicated as NR (not reported)
in Table 1 and were not considered in the analysis of endpoint
sensitivity.
The endpoint HLL referred to the normalization of such an endpoint
to correct for the growth influence (OECD, 2009). Mortality of tadpoles
is considered as an endpoint in the AMA. This endpoint is not included
in the endpoint sensitivity analysis in this paper because all studies
selected had mortality less than 10%. Whether or not a chemical has the
potential to interact with the thyroid pathway is concluded on the basis
of the available experimental evidence. The conclusion of each che-
mical was adopted from the US EPA EDSP weight of evidence evalua-
tion (EPA, 2013, 2015) and the relevant papers. Conclusions regarding
sensitivity were drawn on the basis of the responses of each endpoint to
chemicals in the AMA. The number of observations for three effective
categories was plotted.
Meta-analysis was compared on the basis of responses of endpoints
of chemicals that were considered to have the potential to interact with
the HPT axis and that were not considered to have the potential to
interact with the HPT axis. An inverse-variance random-effects model
was applied to determine the standardised mean difference and 95%
confidence intervals (95% CI) of endpoints between thyroid active and
thyroid inactive chemicals. All statistical analyses were performed
using RevMan v5.3 software (The Nordic Cochrane Centre, The
Cochrane Collaboration, 2014). For all reported comparisons p-va-
lues < 0.05 were considered significant.
3. Results
Data were generated for 57 chemicals from 74 studies (Table 1). The
majority of chemicals were tested in African clawed frog Xenopus laevis.
Three chemicals (mefenacet, simetryn, and thiobencarb) were tested in
the western clawed frog Silurana tropicalis. Due to the limitation in the
number of studies available, it is not possible to compare the sensitivity
in different Amphibian species.
Among 57 chemicals, 12 chemicals are considered to have the po-
tential to interact with the HPT axis in amphibians (Table 1). These
include five chemicals used in the OECD validations, five chemicals
tested in the US EPA EDSP and two chemicals from the literature
(Table 1). These chemicals represent different mechanisms of action on
the thyroid system (Table 1). At day 7, the sensitivity of HLL and SVL
are comparable (Fig. 1, columns with colors), which is higher than that
of the endpoint DS. Compared to SVL, the sensitivity of body weight is
lower, which may be due to the lower number of observations for this
endpoint. When only LOECs are considered, the sensitivity of SVL, body
weight and DS is comparable, but is less than that of HLL (Fig. 1, red
color). At day 21, thyroid gland histology is the most sensitive end-
point, which is a bit more sensitive than the other 4 endpoints. Simi-
larly, when only LOECs are considered for thyroid active chemicals, the
Z. Dang Ecotoxicology and Environmental Safety 167 (2019) 513–519

Table 1
Chemical induced changes in endpoints of 21-day Amphibian Metamorphosis Assay (Coady et al., 2010; Saka et al., 2013; Tietge et al., 2013).

Note, the lowest observed effect concentration (LOEC) was marked by the red color. Effects reported above the LOEC but equal or lower than the maximum tested
concentration within the same study was labeled by the yellow color. blank cell, no effects were observed at all tested concentrations. NR, not reported. HLL, hind
limb length; SVL, snout to vent length; DS, developmental stage; BW, body weight; Thy-his, thyroid histology. Results at Day 7 and 21 are indicated in combination
with endpoints. Chemicals in italics are thyroid active chemicals. 2-mercaptobenzothiazole and Iopanoic acid (IOP), inhibitor of iodothyronine deiodinases; 6-
propylthiouracil (PTU), inhibitor of TH synthesis; Sodium perchlorate, inhibitor of thyroidal iodide uptake; Thyroxine (T4), Thyroid receptor agonist. Conclusions on
a chemical with or without the potential to interact with the HPT (hypothalamic-pituitary-thyroid) axis were based on a weight of evidence evaluation of results of
AMA of the original reports or publications.

most sensitive endpoint is thyroid gland histology and the least sensi-
tive is the endpoint DS.
Among 57 chemicals, 45 chemicals are not considered to have the
potential to interact with the HPT axis in amphibians (Table 1). Of 45
chemicals, 6 chemicals (Benfluralin, Captan, Iprodione, Oxamyl, Me-
fenacet, and Thiobencarb) did not change any endpoints of AMA at all
tested concentrations; 39 chemicals induced changes in thyroid re-
sponsive endpoints. Among 39 chemicals, 33% (13 chemicals) and 77%
(30 chemicals) chemicals induced changes in HLL and/or DS at day 7
and 21, respectively. At day 7, only a decrease in HLL and/or a delayed
DS were observed for 13 chemicals (Ethoprop, Glyphosate, Metalaxyl,
Methomyl, Myclobutanil, Norflurazon, Phosmet, Pronamide, Pyriprox-
yfen, Simazine, Simetryn, Tebuconazole, Tetrachlorvinphos); whereas
increased (5 for HLL and 1 for DS) and decreased (20 for HLL and 10 for
DS) HLL and/or DS were observed at day 21 for 30 chemicals (17β-
estradiol, 2, 4 dichlorophenoxy acetic acid, Acephate, Acetone, Bifen-
thrin, Chlorpyrifos, Cyfluthrin, Diazinon, Dichlobenil, EPTC, Ethoprop,
Glyphosate, Flutolanil, Imidacloprid, Metalaxyl, Methomyl, Metola-
chlor, Myclobutanil, MGK-264, Norflurazon, o-Phenylphenol, PCNB,
Permethrin, Phosmet, Pronamide, Pyriproxyfen, Simazine, Simetryn,
Tebuconazole, Tetrachlorvinphos, and Trifluralin). Only 4 chemicals
(Bifenthrin, Simazine, Tebuconazole, and Tetrachlorvinphos) led to
changes in thyroid gland histology. According to the EPA weight of
evidence assessment, changes in thyroid gland histology induced by
these four chemicals were not associated with the incidence of follicular
cell hypertrophy or hyperplasia, which are expected to be more com-
monly affected in the presence of a thyroid-related developmental delay
(EPA, 2015). Among chemicals without the potential to interact with

515
Z. Dang
Fig. 1. Endpoint sensitivity of thyroid active chemicals in the Amphibian
Metamorphosis Assay. The number of observations with the lowest observed
effect concentration was indicated by LO; The number of observations at the
concentrations between the LOEC and the maximum tested concentration
within the same study was indicated by OE; The number of observations in
which no observed effects (NO) was reported at concentrations higher than the
maximum tested concentration was indicated by NO with the white (blank)
color. HLL, hind limb length; SVL, snout to vent length; DS, developmental
stage; BW, body weight; Thy-his, thyroid gland histology. Results at Day 7 and
21 are indicated in combination with endpoints. An increase (↑) and a decrease
(↓) in the endpoints of HLL, DS, SVL and body weight were indicated by dif-
ferent colors.
Fig. 2. Endpoint sensitivity of thyroid inactive chemicals in the Amphibian
Metamorphosis Assay. The number of observations with the lowest observed
effect concentration was indicated by LO; The number of observations at the
concentrations between the LOEC and the maximum tested concentration
within the same study was indicated by OE; The number of observations in
which no observed effects (NO) was reported at concentrations higher than the
maximum tested concentration was indicated by NO with the white (blank)
color. HLL, hind limb length; SVL, snout to vent length; DS, developmental
stage; BW, body weight; Thy-his, thyroid gland histology. Results at Day 7 and
21 are indicated in combination with endpoints. An increase (↑) and a decrease
(↓) in the endpoints of HLL, DS, SVL and body weight were indicated by dif-
ferent colors.
the HPT axis, seven chemicals (Chlorpyrifos, Ethoprop, Imidacloprid,
Myclobutanil, Phosmet, Pronamide, Pyriproxyfen) induced changes in
both HLL and DS (Table 1). The lack of effects on thyroid
516
Ecotoxicology and Environmental Safety 167 (2019) 513–519
Fig. 3. Endpoint sensitivity of thyroid active and inactive chemicals in the
Amphibian Metamorphosis Assay. The number of observations with the lowest
observed effect concentration was indicated by LO; The number of observations
at the concentrations between the LOEC and the maximum tested concentration
within the same study was indicated by OE; The number of observations in
which no observed effects (NO) was reported at concentrations higher than the
maximum tested concentration was indicated by NO with the white (blank)
color. HLL, hind limb length; SVL, snout to vent length; DS, developmental
stage; BW, body weight; Thy-his, thyroid gland histology. Results at Day 7 and
21 are indicated in combination with endpoints. An increase (↑) and a decrease
(↓) in the endpoints of HLL, DS, SVL and body weight were indicated by dif-
ferent colors.
histopathology along with a reduction in growth suggests develop-
mental delays induced by these chemicals may be due to the non-spe-
cific systemic toxicity and does not support a potential thyroid-related
effect (EPA, 2015). At day 7, the sensitivity of body weight and SVL is
higher than that of HLL and DS (Fig. 2). Interestingly, only a decrease in
HLL was observed at day 7. At day 21, endpoints HLL, SVL and body
weight are the most sensitive endpoints; endpoints thyroid gland his-
tology and DS are less sensitive (Fig. 2).
Fig. 3 shows the results of the endpoint sensitivity analysis for all 57
chemicals. At day 7, body weight and SVL are the most sensitive end-
points; whereas HLL is less sensitive and DS the least sensitive. When
only LOECs are considered, the sensitivity of HLL, SVL and body weight
is comparable, but is higher than that of the endpoint DS. At day 21, the
sensitivity of HLL, SVL and body weight is comparable and higher than
that of DS and thyroid gland histology. When only LOECs are con-
sidered, the most sensitive endpoint is HLL and the least sensitive is DS
(Fig. 3).
Meta-analysis showed that an increase in HLL at day 7 can indicate
chemicals with the potential to interact with the HPT axis (Fig. 4A);
whereas a decrease in HLL cannot distinguish chemicals with and
without thyroid activity (Fig. 4B). Changes in HLL at day 21 alone
cannot indicate chemicals with the potential to interact with the HPT
axis; combination with other endpoints e.g. DS and histology can in-
dicate thyroid activity of a chemical (Fig. 4C and D).
4. Discussion
This study showed that an increase in HLL at day 7 can indicate
chemicals with the potential to interact with the HPT axis (Fig. 1, red
color; Fig. 4A). This is because tadpoles at NF stage 51 are functionally
athyroid and THs induce precocious development (OECD, 2006; EFSA,
2018). These results corroborate the decision logic of the AMA (Fig. 4,
OECD, 2009), in which advanced development at day 7 could be used
as indication of thyroid activity. In contrast, a decrease in HLL at day 7
Z. Dang Ecotoxicology and Environmental Safety 167 (2019) 513–519

Fig. 4. Forest plot of meta-analysis of endpoint responses between chemicals with (indicated as TH) and without (indicated as non-TH) thyroid activity. HLL, hind
limb length; SVL, snout to vent length; DS, developmental stage; BW, body weight; Thy-his, thyroid gland histology. Results at Day 7 and 21 are indicated in
combination with endpoints. An increase and a decrease in the endpoints of HLL, DS, SVL and BW at day 7 was showed in A and B respectively; An increase and a
decrease in the endpoints of HLL, DS, SVL and BW at day 21 was showed in C and D respectively.

cannot indicate chemicals with the potential to interact with the HPT
axis because chemicals without the potential to interact with the HPT
axis can induce similar effects (Fig. 2). These results showed that the
direction (increase or decrease) of changes in the thyroid responsive
apical endpoints is important for the identification of chemicals with
the potential to interact with the HPT axis.
This study included 12 chemicals that are considered to have the
potential to interact with the HPT axis in amphibians. For these 12
chemicals, the sensitivity of two thyroid responsive apical endpoints,
HLL and DS, is comparable to that of growth apical endpoints, body
weight and SVL at day 21 (Fig. 1). For chemicals that are not considered
to have the potential to interact with the HPT axis, the sensitivity of
HLL is also comparable to that of SVL (Fig. 2). Overall, the sensitivity of
HLL is comparable to that of SVL, which is not related to whether a

517
Z. Dang
chemical has the potential to interact with the HPT axis (Fig. 3). This
study showed that more number of changes in the endpoints were ob-
served at day 21 than at day 7. Changes in the endpoints were often
consistent between observations at day 7 and day 21. There are, how-
ever, difference in some endpoints between observations at day 7 and
21. For example, an increase in HLL for carbofura was observed at day 7
but not at day 21, suggesting an importance of observation at day 7.
This study also showed that an increase in HLL and DS at day 21 was
observed for chemicals without the potential to interact with the HPT
axis (Fig. 2). These results suggest that advanced development at day 21
of the decision logic of the AMA should be combined with other thyroid
responsive endpoints for indication of a chemical with potential to in-
teract with the HPT axis. The results here suggest that it may not be
possible to distinguish chemicals with the potential to interact with the
HPT axis based exclusively on the apical thyroid responsive endpoints
HLL and DS at day 21. Besides apical thyroid responsive endpoints HLL
and DS, more evidence, e.g. thyroid gland histology or THs, is needed to
show a chemical with the potential to interact with the HPT axis.
This study showed that the direction of changes, i.e. increase or
decrease, of apical endpoints is important for identification of EDCs.
Normalization of HLL to correct the growth influence is important for
the direction of changes of this endpoint. For example, an increase in
HLL but a decrease in normalized HLL were observed in E2-exposed
tadpoles (OECD 2007). As observed in this study, a decrease but not an
increase was more often observed in apical endpoints (Fig. 3). This
decrease may result from not only chemicals with the potential to in-
teract with the HPT axis but also from e.g. systemic toxicity and stress
(EPA, 2015). Based on the EDC identification criteria and the ECHA/
EFSA guidance, non-specific secondary consequences of toxic effects
shall not be considered for the identification of EDC (EC, 2017; ECHA/
EFSA, 2018). The detailed analysis of thyroid gland histology may be
essential for elucidating whether a decrease in the apical endpoints
results from a chemical with the potential to interact with the HPT axis
or from systemic toxicity or stress (EFSA, 2018; Wegner et al., 2016).
As shown in this study, 33% and 77% chemicals without the po-
tential to interact with the HPT axis induced changes in one of the
thyroid responsive apical endpoints at day 7 and day 21, respectively. If
changes in one of the thyroid responsive endpoints were considered as
the thyroid concern, the percentage of chemicals without thyroid ac-
tivity would be too high. This study showed about 10% chemicals
without the potential to interact with the HPT axis changed thyroid
gland histology. Apparently, information on the thyroid gland histology
can decrease the number of false positives for screening chemicals.
Testing is often needed under the EU chemical regulations when
there is a thyroid concern. Under REACH, for example, testing is often
requested under substance evaluation. In principle, both in vitro and in
vivo testing could be requested. As the AMA is not designed to indicate a
molecular target of a chemical, if a molecular target is needed for
identifying an EDC or for elucidating the plausible link based on the
adverse outcome pathway (AOP), an in vitro assay that is designed to
interact with certain molecular targets (Murk et al., 2013) would be
suitable. According to the draft guidance on the identification of en-
docrine disruptors developed by ECHA and EFSA, the AMA is con-
sidered as a final test for "sufficiently investigated thyroid activity"
(ECHA/EFSA, 2018). This suggests that, if a thyroid concern raised by
e.g. in vitro test results, an AMA could be used to confirm this concern.
The ECHA/EFSA guidance document considers the results of the AMA
as an indication of thyroid activity of a chemical "sufficiently in-
vestigated" in the decision tree (ECHA/EFSA, 2018). Results of 12
chemicals with the potential to interact with the HPT axis suggest that
the AMA can be used for screening/testing for identification of EDCs.
Based on the current AMA database, however, it is not possible to
conclude whether the AMA is effective in screening/testing all chemi-
cals with the potential to interact with the HPT axis. According to the
results of EDSP tests, the papers by Pickford (2010) and by Wegner
et al. (2016), the AMA may have limitations in detecting all chemicals
518
Ecotoxicology and Environmental Safety 167 (2019) 513–519
with the potential to interact with the HPT axis. Some chemicals with
the potential to interact with the HPT axis in mammals cannot be de-
tected in the AMA. Apparently, the current approach for a decision
making on the basis of the AMA results, stated in the draft guidance
(ECHA/EFSA, 2018), may not be sufficient. It is therefore suggested
that both the AMA and mammalian tests should be considered for
"sufficiently investigated thyroid activity".
AMA is a screening assay providing not only mechanistic informa-
tion on thyroid activity but also data potentially relevant to risk as-
sessment. There are, however, some limitations in the use of this assay
(Wheeler et al., 2014). Compared to the draft test guideline of the Xe-
nopus Embryonic Thyroid Assay (XETA), currently under development
in the OECD, AMA takes longer exposure, with more cost. It would be
important to compare these two assays in detecting chemicals with
thyroid activity to explore the possibility whether AMA can be replaced
by XETA for detecting thyroid activity.
In the EU discussions, two important issues have been raised for the
use of AMA: 1) whether an AMA or a LAGDA should be performed first
and whether a positive response in AMA should be followed by LAGDA;
2) should amphibian tests be included for aquatic risk assessment.
Compared to the AMA of short exposure with three test concentrations,
the LAGDA has a longer exposure with at least five test concentrations.
This test design allows the LAGDA to provide, besides endocrine MOAs,
dose response information that can be used to derive a reliable NOEC/
ECx. As both tests intend to indicate chemicals with the potential to
interact with the HPT axis, the AMA may be sufficient for identification
of EDC and no further testing with LADGA is needed. The LAGDA may
then only be needed for the purpose of risk assessment. It is important
to note that LAGDA has been validated by using a limited number of
chemicals because of long exposure, high number of animals and high
cost. There is little experience in using this assay for the regulatory
purpose. The question whether chronic toxicity to amphibians is cov-
ered by the surrogate species fish has been addressed by Weltje et al.
(2013). The authors concluded that fish and amphibian toxicity data
are highly correlated and that fish are more sensitive than amphibians
(Weltje et al., 2013). In their paper, the authors also showed that am-
phibians were more sensitive than fish for one of the thyroid active
chemical, sodium perchlorate. The authors concluded that “substance
that specifically interfere with biochemical pathways involved in am-
phibian metamorphosis may not be detected when using fish as surro-
gates”. Overall, there are still some uncertainties on the comparison of
fish and amphibian chronic toxicity data. It is important to further
explore whether amphibians are more sensitive to thyroid active che-
micals than fish.
Acknowledgements
The author thanks internal reviewers, Joop de Knecht, Els Smit,
Lidka Maslankiewicz, and Theo Vermeire for their critical reading and
suggestions. Richard Dang from Leiden University Medical Center is
acknowledged for the statistical analysis. This project was financially
supported by Strategisch Programma RIVM (project number S/124003/
01) and the REACH methodology (project number M/260025/15/SM)
from the Dutch Ministry of Infrastructure and Water Management.
Disclaimers/Competing interests declaration
None.
References
Coady, K., Marino, T., Thomas, J., Currie, R., Hancock, G., Crofoot, J., McNalley, L.,
McFadden, L., Geter, D., Klecka, G., 2010. Evaluation of the amphibian metamor-
phosis assay: exposure to the goitrogen methimazole and the endogenous thyroid
hormone L-thyroxine. Environ. Toxicol. Chem. 29, 869–880.
Coady, K.K., Lehman, C.M., Currie, R.J., Marino, T.A., 2014. Challenges and approaches
to conducting and interpreting the amphibian metamorphosis assay and the fish
Z. Dang
short-term reproduction assay. Birth Defects Res. B Dev. Reprod. Toxicol. 101, 80–89.
Dang, Z.C., Smit, E., van Vlaardingen, P., Moermond, C., Bodar, C., 2016. Endocrine
Disrupting Chemicals Within EU Legal Frameworks: Environmental Perspectives. pp.
1–40 (2016-0145).
Dang, Z.C., Li, K., Yin, H.W., Hakkert, B., Vermeire, T., 2011. Endpoint sensitivity in fish
endocrine disruption assays: regulatory implications. Toxicol. Let. 202, 36–46.
EC, 2006. Regulation (EC) No 1907/2006, concerning the Regulation, Evaluation,
Authorisation and Restriction of Chemicals (REACH), establishing a European
Chemicals Agency., in 1907/2006, E. Union, Editor. 2006.
EC, 2009. 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 Union L309:1-50.
EC, 2012. Regulation (EU) No 528/2012 of the European Parliament and of the Council of
22 May 2012 concerning the making available on the market and use of biocidal
products. Official Journal of the European Union L167/1.
EC, 2016. Communication from the Commission to the European Parliament and the
Council on endocrine disruptors and the draft Commission acts setting out scientific
criteria for their determination in the context of the EU legislation on plant protection
products and biocidal products. Brussels: COM(2016) 350 final. 〈http://ec.europa.
eu/health/endocrine_disruptors/docs/com_2016_350_en.pdf〉.
EC, 2017. Commission delegated regulation of setting out scientific criteria for the de-
termination of endocrine-disrupting properties pursuant to Regulation (EU) No 528/
2012.
ECHA, 2016. Guidance on Information Requirements and Chemical Safety Assessment
Chapter R.7b: Endpoint specific guidance.
ECHA/EFSA, 2018. Guidance for the identification of endocrine disruptors in the context
of Regulations (EU) No 528/2012 and (EC) No 1107/2009. 〈https://efsa.
onlinelibrary.wiley.com/doi/epdf/〉10.2903/j.efsa.2018.5311).
EFSA Panel on Plant Protection Products and their Residues, 2018. Scientific Opinion on
the state of the science on pesticide risk assessment for amphibians and reptiles. EFSA
J. 16 (2), 5125. https://doi.org/10.2903/j.efsa.2018.5125. (5125, 301 pp).
EPA, 2013. Tier 1 battery review, sections on AMA: 〈https://www.regulations.gov/
document?D=EPA-HQ-OPP-2013–0075–0003〉.
EPA, 2015. Endocrine Disruptor Screening Program Tier 1 Assessments, 〈https://www.
epa.gov/endocrine-disruption/endocrine-disruptor-screening-program-tier-1-
screening-determinations-and〉 (accessed 13 Jan 2017).
Murk, A.J., Rijntjes, E., Blaauboer, B.J., Clewell, R., Crofton, K.M., Dingemans, M.M.,
519
Ecotoxicology and Environmental Safety 167 (2019) 513–519
Furlow, J.D., Kavlock, R., Kohrle, J., Opitz, R., Traas, T., Visser, T.J., Xia, M., Gutleb,
A.C., 2013. Mechanism-based testing strategy using in vitro approaches for-
identification of thyroid hormone disrupting chemicals. Toxicol. In Vitro 27,
1320–1346.
OECD, 2006. Detailed Review Paper on Thyroid Hormone Disruption Assays. OECD Series
on Testing and Assessment. OECD, Paris.
OECD, 2008. Report of the Validation Peer Review for the Amphibian Metamorphosis
Assay and Agreement of the Working Group of the National Coordinators of the Test
Guidelines Programme on the Follow-up of this Report. Environmental Health and
Safety Publications. Series on Testing and Assessment. No. 92. Paris.
OECD, 2009. Test No. 231: Amphibian Metamorphosis Assay.
OECD, 2012. Guidance Document on Standardised Test Guidelines for Evaluating
Chemicals for Endocrine Disruption. Organisation for Economic Cooperation and
Development. Paris, France: Series on Testing and Assessment No. 150. Report nr.
ENV_JM_MONO(2012)22, 524 p.
OECD, 2012a. Test No. 443: Extended One-Generation Reproductive Toxicity Study.
OECD, 2015. Test No. 241: The Larval Amphibian Growth and Development Assay
(LAGDA).
Pickford, D.B., 2010. Screening chemicals for thyroid-disrupting activity: a critical com-
parison of mammalian and amphibian models. Crit. Rev. Toxicol. 40, 845–892.
Saka, M., Tada, N., Kamata, Y., 2013. Application of an amphibian (Silurana tropicalis)
metamorphosis assay to the testing of the chronic toxicity of three rice paddy her-
bicides: simetryn, mefenacet, and thiobencarb. Ecotoxicol. Environ. Saf. 92, 135–143.
Tietge, J.E., Degitz, S.J., Haselman, J.T., Butterworth, B.C., Korte, J.J., Kosian, P.A.,
Lindberg-Livingston, A.J., Burgess, E.M., Blackshear, P.E., Hornung, M.W., 2013.
Inhibition of the thyroid hormone pathway in Xenopus laevis by 2-mercapto-
benzothiazole. Aquat. Toxicol. 126, 128–136.
Wegner, S., Browne, P., Dix, D., 2016. Identifying reference chemicals for thyroid
bioactivity screening. Reprod. Toxicol. 65, 402–413.
Weltje, L., Simpson, P., Gross, M., Crane, M., Wheeler, J., 2013. Comparative acute and
chronic sensitivity of fish and amphibians: a critical review of data. Environ. Toxicol.
Chem. 32, 984–994.
Wheeler, J., Weltje, L., Green, R., 2014. Mind the gap: concerns using endpoints from
endocrine screening assays in risk assessment. Regul. Toxicol. Pharmacol. 69,
289–295.
WHO, 2002. Global Assessment of the State-of-the-science of Endocrine Disruptors. World
Health Organisation, Geneva, Switzerland, pp. 180.