¦— j» United States Prevention, Pesticides EPA 740-C-09-002
Environmental Protection and Toxic Substances October 2009
%#Crri A9ency
Endocrine Disruptor
Screening Program
Test Guidelines
OPPTS 890.1100:
Amphibian
Metamorphosis (Frog)
Page 1
-------
NOTICE
This guideline is one of a series of test guidelines established by the Office of
Prevention, Pesticides and Toxic Substances (OPPTS), United States Environmental Protection
Agency for use in testing pesticides and chemical substances to develop data for submission to
the Agency under the Toxic Substances Control Act (TSCA) (15 U.S.C. 2601, et seq.), the
Federal Insecticide, Fungicide and Rodenticide Act (FIFRA) (7 U.S.C. 136, etseq.), and section
408 of the Federal Food, Drug and Cosmetic (FFDCA) (21 U.S.C. 346a).
The OPPTS test guidelines serve as a compendium of accepted scientific
methodologies and protocols that are intended to provide data to inform regulatory decisions
under TSCA, FIFRA, and/or FFDCA. This document provides guidance for conducting the test,
and is also used by EPA, the public, and the companies that are subject to data submission
requirements under TSCA, FIFRA and/or the FFDCA. As a guidance document, these
guidelines are not binding on either EPA or any outside parties, and the EPA may depart from
the guidelines where circumstances warrant and without prior notice. The procedures contained
in this guideline are strongly recommended for generating the data that are the subject of the
guideline, but EPA recognizes that departures may be appropriate in specific situations. You
may propose alternatives to the recommendations described in these guidelines, and the
Agency will assess them for appropriateness on a case-by-case basis.
For additional information about OPPTS harmonized test guidelines and to access the
guidelines electronically, please go to http://www.epa.gov/oppts and select "Test Methods &
Guidelines" on the left side navigation menu. You may also access the guidelines in
http://www.regulations.gov grouped by Series under Docket ID #s: EPA-HQ-OPPT-2009-0150
through EPA-HQ-OPPT-2009-0159, and EPA-HQ-OPPT-2009-0576.
-------
OPPTS 890.1100: Amphibian Metamorphosis (Frog).
(a) Scope.
(1) Applicability. This guideline is intended to meet testing requirements of
the Toxic Substances Control Act (TSCA) (15 U.S.C. 2601, et seq.), the
Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) (7 U.S.C.
136, et seq.), and the Federal Food, Drug, and Cosmetic Act (FFDCA) (21
U.S.C. 346a).
(2) Background. The Endocrine Disruptor Screening Program (EDSP)
reflects a two-tiered approach to implement the statutory testing
requirements of FFDCA section 408(p) (21 U.S.C. 346a). In general, EPA
intends to use the data collected under the EDSP, along with other
information, to determine if a pesticide chemical, or other substances, may
pose a risk to human health or the environment due to disruption of the
endocrine system.
This test guideline is intended to be used in conjunction with other
guidelines in the OPPTS 890 series that make up the full screening
battery under the EDSP to identify substances that have the potential to
interact with the estrogen, androgen, or thyroid hormone (Tier 1
"screening"). The determination will be made on a weight-of-evidence
basis taking into account data from the Tier 1 assays and other
scientifically relevant information available. The fact that a substance may
interact with a hormone system, however, does not mean that when the
substance is used, it will cause adverse effects in humans or ecological
systems.
Chemicals that go through Tier 1 screening and are found to have the
potential to interact with the estrogen, androgen, or thyroid hormone
systems will proceed to the next stage of the EDSP where EPA will
determine which, if any, of the Tier 2 tests are necessary based on the
available data. Tier 2 testing is designed to identify any adverse
endocrine-related effects caused by the substance, and establish a
quantitative relationship between the dose and that endocrine effect.
(3) Source. OPPTS developed this guideline through a process of
harmonization with Test Guideline 231 published by the Organization for
Economic Cooperation and Development (OECD) (Ref. 1).
(b) Purpose. The Amphibian Metamorphosis Assay (AMA) is a screening assay
intended to empirically identify substances which may interfere with the normal
function of the hypothalamic-pituitary-thyroid (HPT) axis. The AMA represents a
generalized vertebrate model to the extent that it is based on the conserved
structures and functions of the HPT axis. It is an important assay because
amphibian metamorphosis provides a well-studied, thyroid-dependent process
Page 3
-------
which responds to substances active within the HPT axis, and it is the only
existing assay that detects thyroid activity in an animal undergoing morphological
development.
Introduction. The need to develop and validate an assay capable of detecting
substances active in the thyroid system of vertebrate species originates from
concerns that environmental levels of chemicals may cause adverse effects in
both humans and wildlife. The Amphibian Metamorphosis Assay (AMA)
underwent an extensive validation program which included intra- and inter-
laboratory studies demonstrating the relevance and reliability of the assay (Ref.
2, Ref. 3). Subsequently, the validation of the assay was subject to peer-review
by a panel of independent experts (Ref. 4). This Test guideline is the outcome of
the experience gained during the validation studies for the detection of thyroid
active substances
General Experimental Design. The general experimental design entails
exposing stage 51 Xenopus laevis tadpoles to a minimum of three different
concentrations of a test chemical and a dilution water control (and solvent
control, if necessary) for 21 days. There are four replicates for each test
treatment. Larval density at test initiation is 20 tadpoles per test tank for all
treatment groups. The observational endpoints are hind limb length, snout to
vent length (SVL), developmental stage, wet weight, thyroid histology, and daily
observations of mortality.
Description of the Method
(1) Test Species. Xenopus laevis is routinely cultured in laboratories
worldwide and is easily obtainable through commercial suppliers.
Reproduction can be easily induced in this species throughout the year
using human chorionic gonadotropin (hCG) injections and the resultant
larvae can be routinely reared to selected developmental stages in large
numbers to permit the use of stage-specific test protocols. It is preferred
that larvae used in the assay are derived from in-house adults. As an
alternative although this is not the preferred procedure, eggs or embryos
may be shipped to the laboratory performing the test and allowed to
acclimate; the shipping of larval stages for use in the test is unacceptable.
(2) Equipment and Supplies. The following equipment and supplies are
needed for the conduct of this assay:
(i) Exposure system (see description below);
(ii) Glass or stainless steel aquaria (see description below);
(iii) Breeding tanks
(iv) Temperature controlling apparatus (e.g., heaters or coolers
(adjustable to 22° ± 1°C));
(v) Thermometer;
(vi) Binocular dissection microscope;
Page 4
-------
(vii) Digital camera with at least 4 megapixel resolution and micro
function;
(ix) Image digitizing software;
(x) Petri dish (e.g. 100 x 15 mm) or transparent plastic chamber of
comparable size;
(xi) Analytical balance capable of measuring to 3 decimal places (mg);
(xii) Dissolved oxygen meter;
(xiii) pH meter;
(xiv) Light intensity meter capable of measuring in lux units;
(xv) Miscellaneous laboratory glassware and tools;
(xvi) Adjustable pipetters (10 to 5,000 |jL) or assorted pipettes of
equivalent sizes;
(xvii) Test chemical in sufficient quantities to conduct the study,
preferably of one lot #; and
(xviii) Analytical instrumentation appropriate for the chemical on test or
contracted analytical services.
Chemical Testability. The AMA is based upon an aqueous exposure
protocol whereby test chemical is introduced into the test chambers via a
flow-through system. Flow-through methods however, introduce
constraints on the types of chemicals that can be tested, as determined by
the physicochemical properties of the compound. Therefore, prior to using
this protocol, baseline information about the chemical should be obtained
that is relevant to determining the testability, and the OECD Guidance
Document on Aquatic Toxicity Testing of Difficult Substances and Mixtures
should be consulted (Ref. 5). Characteristics which indicate that the
chemical may be difficult to test in aquatic systems include: high octanol
water partitioning coefficients (log KoW), high volatility, susceptibility to
hydrolysis, and susceptibility to photolysis under ambient laboratory
lighting conditions. Other factors may also be relevant to determining
testability and should be determined on a case by case basis. If a
successful test is not possible for the chemical using a flow-through test
system, a static renewal system may be employed. If neither system is
capable of accommodating the test chemical, then the default is to not test
it using this protocol.
Exposure System. A flow-through diluter system is preferred, when
possible, over a static renewal system. If physical and/or chemical
properties of any of the test substances are not amenable to a flow-
through diluter system, then an alternative exposure system (e.g., static-
renewal) can be employed.
The system components should have water-contact components of glass,
stainless steel, and/or Teflon®. However, suitable plastics can be utilized
if they do not compromise the study. Exposure tanks should be glass or
stainless steel aquaria, equipped with standpipes that result in an
approximate tank volume between 4.0 and 10.0 L and minimum water
Page 5
-------
depth of 10 to 15 cm. The system should be capable of supporting all
exposure concentrations and a control, with four replicates per treatment.
The flow rate to each tank should be held constant in consideration of both
the maintenance of biological conditions and chemical exposure (e.g. 25
mL/min). The treatment tanks should be randomly assigned to a position
in the exposure system in order to reduce potential positional effects,
including slight variations in temperature, light intensity, etc.
Fluorescent lighting should be used to provide a photoperiod of 12 hr light:
12 hr dark at an intensity that ranges from 600 to 2,000 lux (lumens/m2) at
the water surface.
Water temperature should be maintained at 22° ± 1°C, pH maintained
between 6.5 to 8.5, and the dissolved oxygen (DO) concentration > 3.5
mg/L (> 40% of the air saturation) in each test tank. As a minimum water
temperature, pH and dissolved oxygen should be measured weekly;
temperature should preferably be measured continuously in at least one
test vessel. Total hardness and alkalinity should be measured in the
control vessels and one vessel at the highest concentration at least once
per week.
Appendix 1 outlines the experimental conditions under which the protocol
should be executed. For further information on setting up flow-through
exposure systems and/or static renewal systems, please refer to the
ASTM Standard Guide for Conducting Acute Toxicity Tests on Test
Materials with Fishes, Macroinvertebrates, and Amphibians (Ref. 6) and
general aquatic toxicology texts.
(i) Water quality. Any water that is locally available {e.g. springwater
or charcoal-filtered tap water) and permits normal growth and
development of X. laevis tadpoles could be used. Because local
water quality can differ substantially from one area to another,
analysis of water quality should be undertaken, particularly, if
historical data on the utility of the water for raising Xenopus is not
available. Special attention should be given that the water is free of
copper, chlorine and chloramines, all of which are toxic to frogs and
tadpoles. It is further recommended to analyze the water
concerning background levels of fluoride, perchlorate and chlorate
(by-product of drinking water disinfection) as all of these anions are
substrates of the iodine transporter of the thyroid gland and
elevated levels of each of these anions may confound the study
outcome. Analysis should be performed before testing begins and
the testing water should normally be free from these anions.
(ii) Iodide Concentration in Test Water. In order for the thyroid gland
to synthesize TH, sufficient iodide needs to be available to the
larvae through a combination of aqueous and dietary sources.
Page 6
-------
Currently, there are no empirically derived guidelines for minimal
iodide concentrations. However, iodide availability may affect the
responsiveness of the thyroid system to thyroid active agents and is
known to modulate the basal activity of the thyroid gland, an aspect
that deserves attention when interpreting the results from thyroid
histopathology. Likewise, excessive iodide concentrations could
potentially mask the effects of thyroid active substances.
Therefore, measured aqueous iodide concentrations from the test
water should be reported. Based on the available data from the
validation studies, the protocol has been demonstrated to work well
when test water iodide (I") concentrations ranged between 0.5 and
10|jg/L. Ideally, the minimum iodide concentration in the test water
should be 0.5|jg/L. If the test water is reconstituted from deionized
water, iodine must be added at a minimum concentration of
0.5|jg/L. Iodide supplementation is only needed if the concentration
is naturally below 0.5 |jg/L in the test water; then, the iodide
supplementation should not exceed 2 |jg/L. Any supplementation
of the test water with iodine or other salts must be noted in the
report.
Holding of Animals.
(i) Adult Care and Breeding. Adult care and breeding is conducted
in accordance with standard guidelines (Ref. 7) and the reader is
directed to the standard guide for performing FETAX for more
detailed information. Such standard guidelines provide an example
of appropriate care and breeding methods, but strict adherence is
not required. To induce breeding, pairs (3-5) of adult females and
males are injected with human chorionic gonadotropin (hCG).
Female and male specimens are injected with approximately 800
IU-1000 III and 600 IU-800 III, respectively, of hCG dissolved in
0.6-0.9% saline solution. Breeding pairs are held in large tanks,
undisturbed and under static conditions to promote amplexus. The
bottom of each breeding tank should have a false bottom of
stainless steel or plastic mesh which permits the egg masses to fall
to the bottom of the tank. Frogs injected in the late afternoon will
usually deposit most of their eggs by mid morning of the next day.
After a sufficient quantity of eggs are released and fertilized, adults
should be removed from the breeding tanks.
(ii) Larval Care and Selection. After the adults are removed from the
breeding tanks, the eggs are collected and evaluated for viability
using a representative sub-set of the embryos from all breeding
tanks. The best individual spawn(s) (2-3 recommended to evaluate
the quality of the spawns) should be retained based upon embryo
viability and the presence of an adequate number (minimum of
1500) of embryos. All the organisms used in a study must originate
Page 7
-------
from a single spawning event (i.e., the spawns should not be co-
mixed).
The embryos are transferred into a large flat pan or dish and all
obvious dead or abnormal eggs (see definition in (Ref. 6)) are
removed using a pipette or eyedropper. The sound embryos from
each of the three spawns are transferred into three separate
hatching tanks.
Four days after being placed in the hatching tanks, the best spawn,
based on viability and hatching success, is selected and the larvae
are transferred into an appropriate number of rearing tanks at 22°
±1°C. In addition, some additional larvae are moved into extra
tanks for use as replacements in the event that mortalities occur in
the rearing tanks during the first week. This procedure maintains
consistent organism density and thereby reduces developmental
divergence within the cohort of a single spawn.
All rearing tanks should be siphoned clean daily. As a precaution,
vinyl or nitrile gloves are preferred to latex gloves. Mortalities
should be removed daily and replacement larvae should be added
back to maintain the organism density during the first week.
Feeding must occur at least twice per day.
During the pre-exposure phase, tadpoles are acclimated to the
conditions of the actual exposure phase including; the type of food,
temperature, light-dark cycle, and the culture medium. Therefore, it
is recommended that the same culture/dilution water be used
during the pre-exposure phase and the exposure phase. If a static
culture system is used for maintaining tadpoles during the pre-
exposure phase, the culture medium should be replaced completely
at least twice per week.
Crowding, caused by high larval densities during the pre-exposure
period, should be avoided because such effects could markedly
affect tadpole development during the subsequent testing phase.
Therefore, the rearing density should not exceed approximately
four tadpoles/L culture medium (static exposure system) or 10
tadpoles/L culture medium (with e.g. 50 mL/min flow rate in the pre-
exposure or culturing system). Under these conditions, tadpoles
should develop from stages 45/46 to stage 51 within twelve days.
Representative tadpoles of this stock population should be
inspected daily for developmental stage in order to estimate the
appropriate time point for initiation of exposure. Care should be
used to minimize stress and trauma to the tadpoles, especially
during movement, cleaning of aquaria, and manipulation of larvae.
Page 8
-------
Stressful conditions/activities should be avoided such as loud
and/or incessant noise, tapping on aquaria, vibrations in the
aquaria, excessive activity in the laboratory, and rapid changes in
environmental media (light availability, temperature, pH, DO, water
flow rates, etc.) If tadpoles do not develop to stage 51 within 17
days after fertilization, excessive stress should be considered as a
potential culprit.
(iii) Larval Culture and Feeding. Tadpoles are fed with Sera Micron®
(Sera GmbH, Heinsberg, Germany) throughout the pre-exposure
period (after NF stage 45/46) and during the entire test period of 21
days, or other diet that has demonstrated to allow equal
performance of tadpoles within the AMA when using Sera Micron®.
The feeding regime during the pre-exposure period should be
carefully adjusted to meet the demands of the developing tadpoles.
That is, small portions of food should be provided to the newly
hatched tadpoles several times per day (at least twice). Excess
food should be avoided i) to maintain water quality and ii) to prevent
the clogging of gill filters with food particles and detritus.
For Sera Micron®, the daily food rations should be increased along
with tadpole growth to approximately 30 mg/animal/day shortly
before test initiation. Sera Micron®, a commercially available
tadpole food that has been shown in the validation studies to
support proper growth and development of X. laevis tadpoles, is a
fine particulate that stays suspended in the water column for a long
period of time and is subject to washing out with the flow.
Therefore, the total daily amount of food should be divided into
smaller portions and fed at least twice daily.
For Sera Micron®, the feeding regime is outlined in Table 1.
Feeding rates must be recorded. Sera Micron® can be fed dry or
as a stock solution prepared in dilution water. Such a stock
solution should be freshly prepared every other day and stored at
4°C when not in use.
Table 1. Feeding Regime for X. laevis Tadpoles During the In-life Portion of the AMA in
Flow-through Conditions.
Study Day
Food ration (mg Sera Micron/animal/day)
0-4
30
5-7
40
8-10
50
11-14
70
15-21
80
Page 9
-------
Analytical Chemistry. Prior to conducting a study, the stability of the test
compound should be evaluated using existing information on its solubility,
degradability, and volatility. Test solutions from each replicate tank at
each concentration should be sampled for analytical chemistry analyses at
test initiation (day 0), and weekly during the test for a minimum of four
samples. It is recommended that each test concentration be analyzed
during system preparation, prior to test initiation, to verify system
performance. In addition, it is recommended that stock solutions be
analyzed when they are changed, especially if the volume of the stock
solution does not provide adequate amounts of chemical to span the
duration of routine sampling periods. In the case of chemicals which
cannot be detected at some or all of the concentrations used in a test,
stock solutions should be measured and system flow rates recorded to
calculate actual concentrations.
Chemical Delivery. The method used to introduce the test chemical to
the system can vary depending on its physicochemical properties. Water
soluble compounds can be dissolved in aliquots of test water at a
concentration which allows delivery at the target test concentration in a
flow-through system. Chemicals which are liquid at room temperature and
sparingly soluble in water can be introduced using liquid:liquid saturator
methods. (Ref. 8) Chemicals which are solid at room temperature and are
sparingly soluble in water can be introduced using glass wool column
saturators (Ref. 8).
The preference is to use a carrier-free test system, however different test
chemicals will possess varied physicochemical properties that will likely
require different approaches for preparation of chemical exposure water.
It is preferred that effort be made to avoid solvents or carriers because: i)
certain solvents themselves may result in toxicity and/or undesirable or
unexpected endocrinological responses, ii) testing chemicals above their
water solubility (as can frequently occur through the use of solvents) can
result in inaccurate determinations of effective concentrations, and iii) the
use of solvents in longer-term tests can result in a significant degree of
"biofilming" associated with microbial activity.
For difficult to test substances, a solvent may be employed as a last
resort, and the OECD Guidance Document on aquatic toxicity testing of
difficult substances and mixtures should be consulted (Ref. 5) to
determine the best method. The choice of solvent will be determined by
the chemical properties of the substance. Solvents which have been
found to be effective for aquatic toxicity testing include acetone, ethanol,
methanol, dimethyl formamide and triethylene glycol.
In case a solvent carrier is used, solvent concentrations should be below
the chronic No Observed Effect Concentration (NOEC); the OECD
Guidance Document recommends a maximum of 100|jl/L; a recent review
Page 10
-------
recommends that solvent concentrations as low as 20|jl/L of dilution water
be used (Ref. 13). If solvent carriers are used, appropriate solvent
controls (clean water) must be evaluated in addition to non-solvent
controls. If it is not possible to administer a chemical via the water, either
because of physicochemical characteristics (low solubility) or limited
chemical availability, introducing it via the diet may be considered.
Preliminary work has been conducted on dietary exposures; however, this
route of exposure is not commonly used. The choice of method should be
documented and analytically verified.
Selection of Test Concentrations
(i) Establishing the High Test Concentration. For the purposes of
this test, the high test concentration should be set by the solubility
limit of the test substance; the maximum tolerated concentration
(MTC) for acutely toxic chemicals; or 100 mg/L, whichever is
lowest.
The MTC is defined as the highest test concentration of the
chemical which results in less than 10% acute mortality. Using this
approach assumes that there are existing empirical acute mortality
data from which the MTC can be estimated. Estimating the MTC
can be inexact and typically requires some professional judgment.
Although the use of regression models may be the most technically
sound approach to estimating the MTC, a useful approximation of
the MTC can be derived from existing acute data by using 1/3 of
the acute LC50 value. However, acute toxicity data may be lacking
for the species on test. If species specific acute toxicity data are
not available, then a 96-hour LC50 test can be completed with
tadpoles that are representative {i.e., same stage) of those on test
in the AMA. Optionally, if data from other aquatic species are
available {e.g. LC50 studies in fish or other amphibian species), then
professional judgment may be used to estimate a likely MTC based
on inter-species extrapolation.
Alternatively, if the chemical is not acutely toxic and is soluble
above 100 mg/L, then 100 mg/L should be considered the highest
test concentration (HTC), as this concentration is typically
considered "practically non-toxic."
Although not the recommended procedure, static renewal methods
may be used where flow-through methods are inadequate to
achieve the MTC. If static renewal methods are used, then the
stability of the test chemical concentration must be documented
and remain within the performance criteria limits. Twenty-four hour
renewal periods are recommended. Renewal periods exceeding 72
hours are not acceptable. Additionally, water quality parameters
Page 11
-------
(e.g. DO, temperature, pH, etc.) must be measured at the end of
each renewal period, immediately prior to renewal.
(ii) Test Concentration Range. There is a required minimum of three
test concentrations and a clean water control (and vehicle control if
necessary). The minimum test concentration differential between
the highest and lowest should be about one order of magnitude.
The maximum dose separation is 0.1 and the minimum is 0.33.
Procedure.
(1) Test Initiation and Conduct.
(i) Day 0. The exposure should be initiated when a sufficient number
of tadpoles in the pre-exposure stock population have reached
developmental stage 51, according to Nieuwkoop and Faber (Ref
9), and which are less than or equal to 17 days of age post
fertilization. For selection of test animals, healthy and normal
looking tadpoles of the stock population should be pooled in a
single vessel containing an appropriate volume of dilution water.
For developmental stage determination, tadpoles should be
individually removed from the pooling tank using a small net or
strainer and transferred to a transparent measurement chamber
{e.g., 100 mm Petri dish) containing dilution water. For stage
determination, it is preferred not to use anesthesia, however one
may individually anesthetize the tadpoles using 100 mg/L tricaine
methanesulfonate {e.g. MS-222), appropriately buffered with
sodium bicarbonate (pH 7.0), before handling. If used,
methodology for appropriately using MS-222 for anesthesia should
be obtained from experienced laboratories and reported with the
test results. Animals should be carefully handled during this
transfer to minimize handling stress and to avoid any injury.
The developmental stage of the animals is determined using a
binocular dissection microscope. To reduce the ultimate variability
in developmental stage, it is important that this staging be
conducted as accurately as possible. According to Nieuwkoop and
Faber (Ref. 9), the primary developmental landmark for selecting
stage 51 organisms is hind limb morphology. The morphological
characteristics of the hind limbs should be examined under the
microscope. While the complete Nieuwkoop and Faber (Ref. 9)
guide should be consulted for comprehensive information on
staging tadpoles, one can reliably determine stage using prominent
morphological landmarks. The following table can be used to
simplify and standardize the staging process throughout the study
by identifying those prominent morphological landmarks associated
with different stages, assuming that development is normal.
Page 12
-------
Table 2. Prominent Morphological Staging Landmarks Based on N&F Guidance.
Prominent
Morphological
Landmarks
Developmental Stag
e
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
Hindlirnb
X
X
X
X
X
X
X
Forelimb
X
X
X
X
X
Craniofacial structure
X
X
X
X
Olfactory rierve
morphology
X
X
X
Tail iength
X
X
X
X
For test initiation, all tadpoles must be at stage 51. The most
prominent morphological staging landmark for that stage is hind
limb morphology, which is demonstrated in Figure 1.
Figure 1. Hind Limb Morphology of a Stage 51 Tadpole
In addition to the developmental stage selection, an optional size
selection of the experimental animals may be used. For this
purpose, the whole body length (not SVL) should be measured at
day 0 for a sub-sample of approximately 20 NF stage 51 tadpoles.
After calculation of the mean whole body length for this group of
animals, minimum and maximum limits for the whole body length of
Page 13
-------
experimental animals can be set by allowing a range of the mean
value ± 3 mm (mean values of whole body length range between
24.0 and 28.1 mm for stage 51 tadpoles). However, developmental
staging is the primary parameter in determining the readiness of
each test animal. Tadpoles exhibiting grossly visible malformations
or injuries should be excluded from the assay.
Tadpoles that meet the stage criteria described above are held in a
tank of clean culture water until the staging process is completed.
Once the staging is completed, the larvae are randomly distributed
to exposure treatment tanks until each tank contains 20 larvae.
Each treatment tank is then inspected for animals with abnormal
appearance (e.g., injuries, abnormal swimming behavior, etc.).
Overtly unhealthy looking tadpoles should be removed from the
treatment tanks and replaced with larvae newly selected from the
pooling tank.
Observations. For more in-depth information on test termination
procedures and processing of tadpoles, refer to the OECD
Guidance Document on Amphibian Thyroid Histology (Ref. 10).
Day 7 Measurements. On day 7, five randomly chosen tadpoles
per replicate are removed from each test tank. The random
procedure used must give each organism on test equal probability
of being selected. This can be achieved by using any randomizing
method but requires that each tadpole be netted. Tadpoles not
selected are returned to the tank of origin and the selected tadpoles
are humanely euthanized in 150 to 200 mg/L MS-222, appropriately
buffered with sodium bicarbonate to achieve pH 7.0. The
euthanized tadpoles are rinsed in water and blotted dry, followed by
body weight determination to the nearest milligram. Hind limb
length, snout to vent length, and developmental stage (using a
binocular dissection microscope) are determined for each tadpole.
Day 21 Measurements (Test Termination). At test termination
(day 21), the remaining tadpoles are removed from the test tanks
and humanely euthanized in 150 to 200 mg/L MS-222,
appropriately buffered with sodium bicarbonate to pH 7.0, as
above. Tadpoles are rinsed in water and blotted dry, followed by
body weight determination to the nearest milligram. Developmental
stage, SVL, hind limb length and wet body weight are measured for
each tadpole.
All larvae are placed in Davidson's fixative for 48 to 72 hours either
as whole body samples or as trimmed head tissue samples
containing the lower jaw for histological assessments. For
histopathology, a total of five tadpoles should be sampled from
Page 14
-------
each replicate tank. Since follicular cell height is stage dependent
(Ref. 11), the most appropriate sampling approach for histological
analyses is to use stage-matched individuals, whenever possible.
In order to select stage-matched individuals, all larvae must first be
staged prior to selection and subsequent processing for data
collection and preservation. This is necessary because normal
divergence in development will result in differential stage
distributions within each replicate tank.
Animals selected for histopathology (n=5 from each replicate)
should be matched to the median stage of the controls (pooled
replicates) whenever possible. If there are replicate tanks with
more than five larvae at the appropriate stage, then five larvae are
randomly selected.
If there are replicate tanks with fewer than five larvae at the
appropriate stage, then randomly selected individuals from the next
lower or upper developmental stage should be sampled to reach a
total sample size of five larvae per replicate. Preferably, the
decision to sample additional larvae from either the next lower or
upper developmental stage should be made based on an overall
evaluation of the stage distribution in the control and chemical
treatments. That is, if the chemical treatment is associated with a
retardation of development, than additional larvae should be
sampled from the next lower stage. In turn, if the chemical
treatment is associated with an acceleration of development, then
additional larvae should be sampled from the next upper stage.
In cases of severe alterations of tadpole development due to
treatment with a test chemical, there might be no overlap of the
stage distribution in the chemical treatments with the calculated
control median developmental stage. In only these cases, the
selection process should be modified by using a stage different
from the control median stage to achieve a stage-matched
sampling of larvae for thyroid histopathology. Furthermore, if
stages are indeterminate {i.e., asynchrony), then 5 tadpoles from
each replicate should be randomly chosen for histological analysis.
The rationale underlying sampling of any larvae that are not at a
stage equivalent to the control median developmental stage should
be reported.
Determination of Biological Endpoints. During the 21 day exposure
phase, measurement of primary endpoints is performed on days 7 and 21,
however daily observation of test animals is necessary. Table 3 provides
an overview of the measurement endpoints and the corresponding
observation time points. More detailed information for technical
procedures for measurement of apical endpoints and histological
Page 15
-------
assessments is available in the OECD Guidance Document on Amphibian
Thyroid Histopathology (Ref. 10).
Table 3. Observation Time Points for Primary Endpoints in the AMA.
Apical Endpoints
Daily
Day 7
Day 21
Mortality
•
Developmental Stage
•
•
Hind Limb Length
•
•
Snout-Vent Length
•
•
Wet Body Weight
•
•
Thyroid Gland Histology
•
Developmental stage, hind limb length, SVL and wet weight are the apical
endpoints of the AMA, and each is briefly discussed below. Further
technical information for collecting these data is available in the OECD
Guidance Document on Amphibian Thyroid Histopathology (Ref. 10)
including procedures for computer-assisted analysis which are
recommended for use.
(i) Developmental Stage. The developmental stage of X. laevis
tadpoles is determined using the staging criteria of Nieuwkoop and
Faber (Ref. 9). Developmental stage data are used to determine if
development is accelerated, asynchronous, delayed or unaffected.
Acceleration or delay of development is determined by making a
comparison between the median stage achieved by the control and
treated groups. Asynchronous development is reported when the
tissues examined are not malformed or abnormal, but the relative
timing of the morphogenesis or development of different tissues is
disrupted within a single tadpole.
(ii) Hind Limb Length. Differentiation and growth of the hind limbs
are under control of thyroid hormones and are major developmental
landmarks already used in the determination of developmental
stage. Hind limb development is used qualitatively in the
determination of developmental stage, but is considered here as a
quantitative endpoint. Therefore, hind limb length is measured as
an endpoint to detect effects on the thyroid axis (Figure 2). For
consistency, hind limb length is measured on the left hind limb.
Hind limb length is evaluated both at day 7 and at day 21 of the
test. On day 7, measuring hind limb length is straightforward, as
illustrated in Figure 2. However, measuring hind limb length on day
21 is more complicated due to bends in the limb. Therefore,
measurements of hind limb length at day 21 should originate at the
body wall and follow the midline of the limb through any angular
deviations. Changes in hind limb length at day 7, even if not
evident at day 21, are still considered significant for potential
Page 16
-------
thyroid activity. Length measurements are acquired from digital
photographs using image analysis software as described in the
OECD Guidance Document on Amphibian Thyroid Histopathology
(Ref. 10)
(iii) Body Length and Wet Weight. Determinations of snout to vent
length (SVL) (Figure 2) and wet weight are included In the test
protocol to assess possible effects of test substances on the growth
rate of tadpoles in comparison to the control group and are useful in
detecting generalized toxicity to the test compound. Because the
removal of adherent water for weight determinations can cause
stressful conditions for tadpoles and may cause skin damage,
these measurements are performed on the day 7 sub-sampled
tadpoles and ail remaining tadpoles at test termination (day 21).
For consistency, use the cranial aspect of the vent as the caudal
limit of the measurement.
Snout to vent length (SVL) is used to assess tadpole growth as
illustrated in Figure 2.
Figure 2. (A) Types of Body Length Measurements and (B) Hind Limb Length
Measurements forX laevis Tadpoles (1).
Ut.
tail length
hind limb
length
whole body length
(iv) Thyroid Gland Histology. While developmental stage and hind
limb length are important endpoints to evaluate exposure-related
changes in metamorphic development, developmental delay
cannot, by itself, be considered a diagnostic indicator of anti-
thyroidal activity. Some changes may only be observable by
routine histopathological analysis. Diagnostic criteria include
thyroid gland hypertrophy/atrophy, follicular cell hypertrophy,
follicular cell hyperplasia, and as additional qualitative criteria:
follicular lumen area, colloid quality and follicular cell height/shape.
Severity grading (4 grades) should be reported. Please refer to the
OECD Guidance Document on Amphibian Thyroid Histopathology
(Ref. 10) for information on morphologic and histological
Page 17
-------
preparation of samples, and for performing histologic analyses on
tissue samples. Further information regarding the normal thyroid
gland architecture during metamorphic development and more
photomicrographic examples of the core diagnostic criteria with
associated severity grades can be obtained in "Thyroid
Histopathology Assessments for the Amphibian Metamorphosis
Assay to Detect Thyroid-active Substances" in Toxicologic
Pathology (2009) (Ref. 15). Laboratories performing the assay for
the first time(s) should seek advice from experienced pathologists
for training purposes prior to undertaking histological analysis and
evaluation of the thyroid glands. Overt and significant changes in
apical endpoints indicating developmental acceleration or
asynchrony may preclude the necessity to perform
histopathological analysis of the thyroid glands. However, absence
of overt morphological changes or evidence of developmental delay
warrants histological analyses.
Mortality. All test tanks should be checked daily for dead tadpoles
and the numbers recorded for each tank. The date, concentration
and tank number for any observation of mortality should be
recorded. Dead animals should be removed from the test tank as
soon as observed. Mortality rates exceeding 10% may indicate
inappropriate test conditions or toxic effects of the test chemical.
Additional Observations. Cases of abnormal behavior and
grossly visible malformations and lesions should be recorded. The
date, concentration and tank number for any observation of
abnormal behavior, gross malformations, or lesions should be
recorded.
Normal behavior is characterized by the tadpoles being suspended
in the water column with tail elevated above the head, regular
rhythmic tail fin beating, periodic surfacing, operculating, and being
responsive to stimulus.
Abnormal behavior would include, for example, floating on the
surface, lying on the bottom of the tank, inverted or irregular
swimming, lack of surfacing activity, and being nonresponsive to
stimulus. In addition, gross differences in food consumption
between treatments should be recorded. Gross malformations and
lesions could include morphological abnormalities (e.g., limb
deformities), hemorrhagic lesions, bacterial or fungal infections, to
name a few. These determinations are qualitative and should be
considered akin to clinical signs of disease/stress and made in
comparison to control animals. If the occurrence or rate of
occurrence is greater in exposed tanks than in the controls, then
these should be considered as evidence for overt toxicity. Similar
Page 18
-------
lesions manifested across all groups, including the control group,
should be considered as evidence of inappropriate testing
conditions.
Data and Reporting.
(1) Data Collection. All data should be collected using electronic or manual
systems which conform to good laboratory practices (GLP). Study data
should include:
Test substance: Name and CAS number.
Characterization of the test substance: Physical-chemical
properties; information on stability and biodegradability.
Chemical observations and data: Method and frequency of
preparation of dilutions. Test chemical information includes actual
and nominal concentrations of the test chemical, and in some
cases, non-parent chemical, as appropriate. Test chemical
measurements may be required for stock solutions as well as for
test solutions.
Solvent (if other than water): Justification of the choice of solvent,
and characterization of solvent (nature, concentration used).
Test conditions:
Operational records: These consist of observations pertaining to
the functioning of the test system and the supporting environment
and infrastructure. Typical records include: ambient temperature,
test temperature, photoperiod, status of critical components of the
exposure system {e.g. pumps, cycle counters, pressures), flow
rates, water levels, stock bottle changes, and feeding records.
General water quality parameters include: pH, DO, conductivity,
total iodine, alkalinity, and hardness.
Deviations from the test method: This information should include
any information or narrative descriptions of deviations from the test
method.
Results:
Biological observations and data: These include daily observations
of mortality, food consumption, abnormal swimming behavior,
lethargy, loss of equilibrium, malformations, lesions, etc.
Observations and data collected at predetermined intervals include:
developmental stage, hind limb length, snout vent length, and wet
weight.
Page 19
-------
Statistical analytical techniques and justification of techniques used:
Results of the statistical analysis preferably in tabular form,
methods for determining whether outliers exist, and justification for
not using outliers.
Histological data: These include narrative descriptions, as well as
graded severity and incidence scores of specific observations, as
detailed in the histopathology guidance document.
Ad hoc observations: These observations should include narrative
descriptions of the study that do not fit into the previously described
categories.
(2) Data reporting. Appendix 2 contains daily data collection spreadsheets
that can be used as guidance for raw data entry and for calculations of
summary statistics. Additionally, reporting tables are provided that are
convenient for communicating summaries of endpoint data. Reporting
tables for histological assessments can be found in Appendix 2. Similar
electronic templates for the AMA can be accessed on EPA's OPPTS
Harmonized Test Guidelines website and in the public docket (Ref. 14).
(3) Guidance on Performance Criteria for Optimal Data Quality.
Generally, gross deviations from the test method will likely result in sub-
optimal data for interpretation or reporting. Therefore, the following criteria
in Table 4 have been developed as guidance for determining the quality of
the test performed, and the general performance of the control organisms.
Table 4. Performance Criteria for the AMA.
Criterion
Acceptable limits
Test concentrations
Maintained at < 20% CV (variability of measured test
concentration) over the 21 day test
Mortality in controls
<10% - mortality in any one replicate in the controls should
not exceed 2 tadpoles
Minimum median developmental stage of
controls at end of test
57
Spread of development stage in control
group
The 10tn and the 90tn percentile of the development stage
distribution should not differ by more than 4 stages
Dissolved Oxygen
> 40% air saturation*
PH
pH should be maintained between 6.5-8.5. The inter-
replicate/inter-treatment differentials should not exceed 0.5.
Water temperature
22° + 1°C - the inter-replicate/inter-treatment differentials
should not exceed 0.5 °C
Test concentrations (non-control) without
overt toxicity
>2
Replicate performance
< 2 replicates across the test can be compromised
Special conditions for use of a solvent
If a carrier solvent is used, both a solvent control and clean
water control must be used and results reported
Statistically significant differences between solvent control
and water control groups are treated specially. See below
for more information
Page 20
-------
Criterion
Acceptable limits
Special conditions for static renewal system
Representative chemical analyses before and after renewal
must be reported
Ammonia levels must be measured immediately prior to
renewal
All water quality parameters listed in Table 1 of Appendix 1
must be measured immediately prior to renewal
Renewal period must not exceed 72 hours
Appropriate feeding schedule (50% of the daily food ration
of Sera Micron ®)
*Aeration of water can be maintained through bubblers. It is recommended to set bubblers at
levels that do not create undue stress on the tadpoles.
(4) Test Validity. Achieving the following goals will likely deem a test to be
considered acceptable/valid:
(i) Valid experiment in a test determined to be negative for
thyroid activity:
~ For any given treatment (including controls), mortality should
not exceed 10%. For any given replicate, mortality should not
exceed three tadpoles, otherwise the replicate may be
considered compromised.
~ At least two treatment levels, with all four uncompromised
replicates, should be available for analysis.
~ At least two treatment levels without overt toxicity should be
available for analysis.
(ii) Valid experiment in a test determined to be positive for thyroid
activity:
~ Mortality of no more than two tadpoles/replicate in the
control group should occur.
(5) Decision logic for the conduct of the AMA. Decision logic was
developed for the AMA to provide logical assistance in the conduct and
interpretation of the results of the bioassay (see flow chart in Figure 3).
The decision logic, in essence, weighs the endpoints in that advanced
development, asynchronous development, and thyroid histopathology are
weighed heavily, while delayed development, snout-vent length and wet
body weight, parameters that can potentially be affected by general
toxicity, are weighted less heavily.
Page 21
-------
Figure 3. Decision Logic for the Conduct of the AMA.
NO
High Quality
Experiment
YES
YES
Likely Thyroid Active
NO
YES
Likely Thyroid Active
NO
YES
Likely Thyroid Active
NO
Likely
Thyroid Inactive:
Stop
Conduct AMA
Repeat Study
Advanced Development
Asynchronous Development
Remarkable Histological Effects
*Histology may be required by some regulatory authorities despite significant differences in
advanced and asynchronous development. The entity performing this test is encouraged to
consult the necessary authorities prior to the performing the test to determine which endpoints
are required.
(i) Advanced development (determined using developmental
stage, SVL and HLL). Advanced development is only known to
occur through effects which are thyroid hormone related. These
may be peripheral tissue effects such as direct interaction with the
thyroid hormone receptor (such as with T4) or effects which alter
circulating thyroid hormone levels. In either case, this is considered
sufficient evidence to indicate that the chemical has thyroid activity.
Advanced development is evaluated in one of two ways. First, the
general developmental stage can be evaluated using the
standardized approach detailed in Nieuwkoop and Faber (Ref. 9).
Second, specific morphological features may be quantified, such as
hind limb length, at both days 7 and 21, which is positively
associated with agonistic effects on the thyroid hormone receptor.
If statistically significant advances in development or hind limb
length occur, then the test indicates that the chemical is likely
thyroid active.
Page 22
-------
The evaluation of test animals for the presence of accelerated
development relative to the control population will be based on
results of statistical analyses performed for the following four
endpoints:
hind limb length (normalized by SVL) on study day 7
hind limb length (normalized by SVL) on study day 21
developmental stage on study day 7
developmental stage on study day 21
Statistical analyses of hind limb length should be performed based
on measurements of the length of the left hind limb. Hind limb
length is normalized by taking the ratio of hind limb length to snout-
to-vent length of an individual. The means of the normalized values
for each treatment level are then compared. Acceleration of
development is indicated by a significant increase in the mean hind
limb length (normalized) in a chemical treatment group compared to
the control group on study day 7 and/or study day 21 (see
Appendix 3).
Statistical analyses of developmental stage should be performed
based on determination of developmental stages according to the
morphological criteria described by Nieuwkoop and Faber (Ref. 8).
Acceleration of development is indicated when the multi-quantal
analysis detects a significant increase of developmental stage
values in a chemical treatment group compared to the control group
on study day 7 and/or study day 21.
In the AMA test method, a significant effect on any of the four
endpoints mentioned above is regarded sufficient for a positive
detection of accelerated development. That is, significant effects
on hind limb length at a specific time point do not require
corroboration by significant effects on hind limb length at the
alternative time point, nor by significant effects on developmental
stage at this specific time point. In turn, significant effects on
developmental stage at a specific time point do not require
corroboration by significant effects on developmental stage at the
alternative time point, nor by significant effects on hind limb length
at this specific time point. The weight of evidence for accelerated
development will nevertheless increase if significant effects are
detected for more than one endpoint.
Asynchronous development (determined using developmental
stage criteria). Asynchronous development is characterized by
disruption of the relative timing of the morphogenesis or
development of different tissues within a single tadpole. The
inability to clearly establish the developmental stage of an organism
using the suite of morphological endpoints considered typical of any
Page 23
-------
given stage indicates that the tissues are developing
asynchronously through metamorphosis. Asynchronous
development is an indicator of thyroid activity. The only known
modes of action causing asynchronous development are through
effects of chemicals on peripheral thyroid hormone action and/or
thyroid hormone metabolism in developing tissues such as is
observed with deiodinase inhibitors.
The evaluation of test animals for the presence of asynchronous
development relative to the control population is based on gross
morphological assessment of test animals on study day 7 and study
day 21.
The description of normal development of X. laevis by Nieuwkoop
and Faber (Ref. 9) provides the framework for identifying a
sequential order of normal tissue remodelling. The term
"asynchronous development" refers specifically to those deviations
in tadpole gross morphological development that disallow the
definitive determination of a developmental stage according to the
criteria of Nieuwkoop and Faber (Ref. 9) because key
morphological landmarks show characteristics of different stages.
As implicated by the term "asynchronous development", only cases
showing deviations in the progress of remodelling of specific
tissues relative to the progress of remodelling of other tissues
should be considered. Some classical phenotypes include delay or
absence of fore limb emergence despite normal or advanced
development of hind limbs and tail tissues, or the precocious
resorption of gills relative to the stage of hind limb morphogenesis
and tail resorption. An animal will be recorded as showing
asynchronous development if it cannot be assigned to a stage
because it fails to meet a majority of the landmark developmental
criteria for a given Niewkoop and Faber stage (Ref. 9), or if there is
extreme delay or acceleration of one or more key features (e.g., tail
completely resorbed, but forelimbs not emerged). This assessment
is performed qualitatively and should examine the full suite of
landmark features listed by Nieuwkoop and Faber (Ref. 9).
However it is not necessary to record the developmental stage of
the various landmark features of animals being observed. Animals
recorded as showing asynchronous development are not assigned
to a Nieuwkoop and Faber (Ref. 9) development stage.
Thus, a central criterion for designating cases of abnormal
morphological development as "asynchronous development" is that
the relative timing of tissue remodelling and tissue morphogenesis
is disrupted whereas the morphology of affected tissues is not
overtly abnormal. One example to illustrate this interpretation of
Page 24
-------
gross morphological abnormalities is that retarded hind limb
morphogenesis relative to development of other tissues fulfills the
criterion of "asynchronous development" whereas cases showing
missing hind limbs, abnormal digits {e.g., ectrodactyly, Polydactyly),
or other overt limb malformations should not be considered as
"asynchronous development".
In this context, the major morphological landmarks that should be
evaluated for their coordinated metamorphic progress should
include hind limb morphogenesis, fore limb morphogenesis, fore
limb emergence, the stage of tail resorption (particularly the
resorption of the tail fin), and head morphology {e.g., gill size and
stage of gill resorption, lower jaw morphology, protrusion of
Meckel's cartilage).
Dependent on the mode of chemical action, different gross
morphological phenotypes can occur. Some classical phenotypes
include delay or absence of fore limb emergence despite normal or
advanced development of hind limbs and tail tissues, precocious
gill resorption relative to hind limb and tail remodelling.
Histopathology. If the chemical does not accelerate development
or cause asynchronous development, then histopathology of the
thyroid glands is evaluated using the available guidance document
(Ref. 10). Developmental retardation, in the absence of toxicity, is
a strong indicator of anti-thyroid activity, but the developmental
stage analysis is less sensitive and less diagnostic than the
histopathological analysis of the thyroid glands. Therefore,
conducting histopathological analyses of the thyroid glands is
required in this case. Effects on thyroid gland histology have been
demonstrated in the absence of developmental effects. If changes
in thyroid histopathology occur, then the chemical is considered to
be thyroid active. If no developmental delays or histological lesions
are observed in the thyroid glands, then the chemical is considered
to be thyroid inactive. The rationale for this decision is that the
thyroid glands are under the influence of TSH and any chemical
which alters circulating thyroid hormone sufficiently to alter TSH
secretion will result in histopathological changes in the thyroid
glands. Various modes and mechanisms of action can alter
circulating thyroid hormone. So, while thyroid hormone level is
indicative of a thyroid related effect, it is insufficient to determine
which mode or mechanism of action is related to the response.
Because this endpoint is not amenable to basic statistical
approaches, the determination of an effect associated with
exposure to a chemical shall be made through expert opinion by a
pathologist.
Page 25
-------
(iv) Delayed development (determined using developmental stage,
HLL, BW, SVL). Delayed development can occur through anti-
thyroidal mechanisms and through indirect toxicity. Mild
developmental delays coupled with overt signs of toxicity likely
indicate a non-specific toxic effect. Evaluation of non-thyroidal
toxicity is an essential element of the test to reduce the probability
of false positive outcomes. Excessive mortality is an obvious
indication that other toxic mechanisms are occurring. Similarly,
mild reductions in growth, as determined by wet weight and/or SVL
length, also suggest non-thyroidal toxicity. Apparent increases in
growth are commonly observed with compounds that negatively
affect normal development. Consequently, the presence of larger
animals does not necessarily indicate non-thyroidal toxicity.
However, growth should never be solely relied upon to determine
thyroid toxicity. Rather, growth, in conjunction with developmental
stage and thyroid histopathology, should be used to determine
thyroid activity. Other endpoints should also be considered in
determining overt toxicity including edema, hemorrhagic lesions,
lethargy, reduced food consumption, erratic/altered swimming
behavior, etc. If all test concentrations exhibit signs of overt
toxicity, the test compound must be re-evaluated at lower test
concentrations before determining whether the compound is
potentially thyroid active or thyroid inactive.
Statistically significant developmental delays, in absence of other
signs of overt toxicity, indicate that the chemical is thyroid active
(antagonistic). In the absence of strong statistical responses, this
outcome may be augmented with results from thyroid
histopathology.
Statistical analyses. Statistical analyses of the data should preferably
follow procedures described in the document Current Approaches in the
Statistical Analysis of Ecotoxicity Data: A Guidance to Application (Ref
12). For all continuous quantitative endpoints (HHL, SVL, wet weight)
consistent with a monotone dose-response, the Jonckheere-Terpstra test
should be applied in step-down manner to establish a significant treatment
effect.
For continuous endpoints that are not consistent with a monotone dose-
response, the data should be assessed for normality (preferably using the
Shapiro-Wilk or Anderson-Darling test) and variance homogeneity
(preferably using the Levene test). Both tests are performed on the
residuals from an ANOVA. Expert judgment can be used in lieu of these
formal tests for normality and variance homogeneity, though formal tests
are preferred. Where non-normality or variance heterogeneity is found, a
normalizing, variance stabilizing transformation should be sought. If the
data (perhaps after a transformation) are normally distributed with
Page 26
-------
homogeneous variance, a significant treatment effect is determined from
Dunnett's test. If the data (perhaps after a transformation) are normally
distributed with heterogeneous variance, a significant treatment effect is
determined from the Tamhane-Dunnett or T3 test or from the Mann-
Whitney-Wilcoxon U test. Where no normalizing transformation can be
found, a significant treatment effect is determined from the Mann-Whitney-
Wilcoxon U test using a Bonferroni-Holm adjustment to the p-values. The
Dunnett test is applied independently of any ANOVA F-test and the Mann-
Whitney test is applied independently of any overall Kruskall-Wallis test.
Significant mortality is not expected but should be assessed from the step-
down Cochran-Armitage test where the data are consistent with dose-
response monotonicity, and otherwise from Fisher's Exact test with a
Bonferroni-Holm adjustment.
A significant treatment effect for developmental stage is determined from
the step-down application of the Jonckheere-Terpstra test applied to the
replicate medians. Alternatively, and preferably, the multi-quantal
Jonckheere test from the 20th to the 80th percentile should be used for
effect determination, as it takes into account changes to the distribution
profile.
The appropriate unit of analysis is the replicate so the data consist of
replicate medians if the Jonckheere-Terpstra or Mann-Whitney U test is
used, or the replicate means if Dunnett's test is used. Dose-response
monotonicity can be assessed visually from the replicate and treatment
means or medians or from formal tests such as previously described (Ref.
12) With fewer than five replicates per treatment or control, the exact
permutation versions of the Jonckheere-Terpstra and Mann-Whitney tests
should be used if available. The statistical significance of all tests
indicated is judged at the 0.05 significance level.
Please see Figure 4 for a flow chart for performing statistical tests on
continuous data.
Page 27
-------
Figure 4. Flowchart for Statistical Approaches for Continuous Response Data.
Flow-Chart for Continuous Response
Are data consistent with monotone dose-response?
Yes
No
Apply step-down Jonckheere-Terpstra test to
determine effects. With <5 reps per concentration,
use exact version of test if available.
Are data normally distributed
(possibly after transformation)?
Yes
No
r
Are variances homogeneous
(possibly after transformation)?
Yes
No
Use Mann-Whitney test with
Bonferroni-Holm adjustment to
determine effects. With <5 reps per
concentration, use exact version of
test if available.
Use Dunnett test to
determine effects.
Use Tamhane-Dunnett
(T3) test if available.
\Otherwise follow arrow.
Page 28
-------
Special Data Analysis Considerations.
(i) Use of compromised treatment levels. Several factors are
considered when determining whether a replicate or entire
treatment demonstrates overt toxicity and should be removed from
analysis. Overt toxicity is defined as >2 mortalities in any replicate
that can only be explained by toxicity rather than technical error.
Other signs of overt toxicity include hemorrhage, abnormal
behaviors, abnormal swimming patterns, anorexia, and any other
clinical signs of disease. For sub-lethal signs of toxicity, qualitative
evaluations may be necessary, and should always be made in
reference to the clean water control group.
(ii) Solvent controls. The use of a solvent should only be considered
as a last resort, when all other chemical delivery options have been
considered. If a solvent is used, then a clean water control must be
run in concert. At the termination of the test, an evaluation of the
potential effects of the solvent must be performed. This is done
through a statistical comparison of the solvent control group and
the clean water control group. If statistically significant differences
are detected in these endpoints between the clean water control
and solvent control groups, then best professional judgment should
be used to determine if the validity of the test is compromised.
(iii) Treatment groups achieving developmental stage 60 and
above. After stage 60, tadpoles show a reduction in size and
weight due to tissue resorption and reduction of absolute water
content. Thus, measurements of wet weight and SVL cannot
appropriately be used in statistical analyses for differences in
growth rates. Therefore, wet weight and length data from
organisms >NF60 must be censored and cannot be used in
analyses of replicate means or replicate medians. Two different
approaches could be used to analyze these growth-related
parameters.
One approach is to consider only tadpoles with developmental
stages lower or equal to stage 60 for the statistical analyses of wet
weight and/or SVL. This approach is believed to provide sufficiently
robust information about the severity of possible growth effects as
long as only a small proportion of test animals are removed from
the analyses (<20%). If an increased number of tadpoles show
development beyond stage 60 (>20%) in one or more nominal
concentration(s), then a two-factor ANOVA with a nested variance
structure should be undertaken on all tadpoles to assess growth
effects due to chemical treatments while taking into account the
Page 29
-------
effect of late stage development on growth. Appendix 3 provides
guidance on the two-factor AN OVA analysis of weight and length.
(h) References.
1. OECD (2009). Amphibian Metamorphosis Assay. OECD Guideline for the Testing
of Chemicals. No. 231. Paris, France.
2. OECD (2004). Report of the Validation of the Amphibian Metamorphosis Assay for
the Detection of Thyroid Active Substances: Phase 1 - Optimization of the Test
Protocol. Environmental Health and Safety Publications. Series on Testing and
Assessment. No. 77. Paris, France.
3. OECD (2007). Final Report of the Validation of the Amphibian Metamorphosis
Assay: Phase 2 - Multi-chemical Interlaboratory Study. Environmental Health and
Safety Publications. Series on Testing and Assessment. No. 76. Paris, France.
4. 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, France.
5. OECD (2000). Guidance Document on Aquatic Toxicity Testing of Difficult
Substances and Mixtures. Environmental Health and Safety Publications. Series
on Testing and Assessment. No. 23. Paris, France.
6. ASTM (2002). Standard Guide for Conducting Acute Toxicity Tests on Test
Materials with Fishes, Macroinvertebrates, and Amphibians. American Society for
Testing and Materials, ASTM E729-96(2002), Philadelphia, PA.
7. ASTM (2004). Standard Guide for Conducting the Frog Embryo Teratogenesis
Assay - Xenopus (FETAX). E 1439-98.
8. Kahl,M.D., Russom,C.L., DeFoe,D.L. & Hammermeister,D.E. (1999). Saturation
Units for Use in Aquatic Bioassays. Chemosphere 39, pp. 539-551.
9. Nieuwkoop,P.D. & Faber,J. (1994). Normal Table of Xenopus Laevis. Garland
Publishing, New York.
10. OECD (2007). Guidance Document on Amphibian Thyroid Histology.
Environmental Health and Safety Publications. Series on Testing and
Assessment. No. 82. Paris, France.
11. Dodd,M.H.I. & Dodd,J.M. (1976). Physiology of Amphibia. Lofts,B. (ed.),
Academic Press, New York, pp. 467-599.
Page 30
-------
12. OECD (2006). Current Approaches in the Statistical Analysis of Ecotoxicity Data:
A Guidance to Application. Environmental Health and Safety Publications. Series
on Testing and Assessment. No. 54. Paris, France.
13. Hutchinson TH, Shillabeer N, Winter MJ, Pickford DB (2006). Acute and Chronic
Effects of Carrier Solvents in Aquatic Organisms: A critical Review. Review.
Aquatic Toxicology, 76, pp.69-92.
14. EPA. OPPTS Harmonized Test Guidelines. Available on-line at:
http://www.epa.gov/oppts (select "Test Methods & Guidelines" on the left side
navigation menu). You may also access the guidelines in
httpV/www.regulations.gov grouped by Series under Docket ID #s: EPA-HQ-OPPT-
2009-0150 through EPA-HQ-OPPT-2009-0159, and EPA-HQ-OPPT-2009-0576.
15. Grim KC et at. (2009). Thyroid Histopathology Assessments for the Amphibian
Metamorphosis Assay to Detect Thyroid-active Substances. Toxicologic Pathology
37, pp 415-424.
Page 31
-------
Appendix 1
Table 1. Experimental Conditions for the 21-day Amphibian Metamorphosis Assay.
Test Animal
Xenopus laevis larvae
Initial Larval Stage
Nieuwkoop and Faber stage 51
Exposure Period
21 days
Larvae Selection Criteria
Developmental stage and total length (optional)
Test Concentrations
Minimum of 3 concentrations spanning approximately one
order of magnitude
Exposure Regime
Flow-through (preferred) and/or static-renewal
Test System Flow-Rate
25 mL/min (complete volume replacement ca. every 2.7 h)
Primary Endpoints / Determination Days
Mortality
Daily
Developmental Stage
D 7 and 21
Hind Limb Length
D 7 and 21
Snout-Vent Length
D 7 and 21
Wet Body Weight
D 7 and 21
Thyroid Histology
D 21
Dilution Water / Laboratory Control
Dechlorinated tap water (charcoal-filtered) or the equivalent
laboratory source
Larval Density
20 larvae / test vessel (5 / L)
Test Solution / Test Vessel
4-10 L (10-15 cm minimum water) / Glass or Stainless Steel
test vessel (e.g., 22.5 cm x 14 cm x 16.5 cm)
Replication
4 replicate test vessels / test concentration and control
Acceptable Mortality Rate in Controls
< 10% per replicate test vessel
Thyroid Fixation
Number
Fixed
All tadpoles (5/replicate are evaluated initially)
Region
Head or whole body
Fixation
Fluid
Davidson's fixative
Feeding
Food
Sera Micron® or equivalent
Amount /
Frequency
See Table 1 for feeding regime using Sera Micron
Lighting
Photoperiod
12 h Light: 12 h dark
Intensity
600 to 2000 lux (Measured at Water Surface)
Water Temperature
22° +1 °C
PH
6.5-8.5
Dissolved Oxygen (DO) Concentration
>3.5 mg/L (>40% Air Saturation)
Analytical Chemistry Sample Schedule
Once / Week (4 Sample Events / Test)
A- 1
-------
Appendix 2
Reporting Tables for Raw Data and Summary Data.
Table 1. General Test Chemical Information.
Reporting Organization:
Study Title:
Report Name:
Date of report:
Report filename:
Chemical Name:
CAS Number:
Purity
Stability
Octanol:Water Coefficient (Kow):
Chemical volatility
Other Chemical Constituents:
Solvent (if applicable):
Nominal Dose 1:
Nominal Dose 2:
Nominal Dose 3:
Nominal Dose 4:
Value
Units
Date
(day 0)
Date
(day 7)
Date
(day 14)
Date
(day 21)
Enter date (mm/dd/yy)
A - 2
-------
Table 2. Study Descriptors.
Chemical Name: 0
CAS #: 0
Assay: Amphibian Metamorphosis
Descriptor Value
Test Species
Source of larvae
Number Larvae/Replicate
Number Replicates/Dose
Larval stage (NF) at initial
dosing
Exposure system (flow-
through/static renewal)
Test-system flow rate (ml/min)
Dilution water/Laboratory
control source
Water contaminants
concentrations (flouride,
chlorates, etc.)
Test vessel (steel/glass)
Diet type
Diet amount
Diet frequency
Light/dark cycle (hours)
Light intensity (lux)
Iodine concentration (|jg/l_)
A- 3
-------
Table 3. Raw Test Conditions.
Chemical
Name: 0
CAS #: 0
0 0
0 0
0 0
0 0
Re
plicate Number
Re
plicate Number
Re
plicate Number
Re
plicate Number
Test Day
Date
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
0
01/00/00
7
01/07/00
14
01/14/00
21
01/21/00
^issolved^Ox^genJmcj/L^
0 0
0 0
0 0
0 0
Re
plicate Number
Re
plicate Number
Re
plicate Number
Re
plicate Number
Test Day
Date
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
0
01/00/00
7
01/07/00
14
01/14/00
21
01/21/00
pH
0 0
0 0
0 0
0 0
Re
plicate Number
Re
plicate Number
Re
plicate Number
Re
plicate Number
Test Day
Date
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
0
01/00/00
7
01/07/00
14
01/14/00
21
01/21/00
Total hardness
0 0
0 0
Re
plicate Number
Re
plicate Number
Test Day
Date
1
2
3
4
1
2
3
4
0
01/00/00
7
01/07/00
14
01/14/00
21
01/21/00
Alkalinity
0 0
0 0
Re
plicate Number
Re
plicate Number
Test Day
Date
1
2
3
4
1
2
3
4
0
01/00/00
7
01/07/00
14
01/14/00
21
01/21/00
A - 4
-------
Table 4. Mortality Data.
Chemical
Name: 0
CAS #: 0
O
o
O
O
O
O
O
O
Replicate Number
Replicate Number
Replicate Number
Replicate Number
Test Day
Date
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
0
01/00/00
1
01/01/00
2
01/02/00
3
01/03/00
4
01/04/00
5
01/05/00
6
01/06/00
7
01/07/00
8
01/08/00
9
01/09/00
10
01/10/00
11
01/11/00
12
01/12/00
13
01/13/00
14
01/14/00
15
01/15/00
16
01/16/00
17
01/17/00
18
01/18/00
19
01/19/00
20
01/20/00
21
01/21/00
Replicate
Count
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Treatment
Count
0
0
0
0
Percent
Mortality
#DIV/0!
#DIV/0!
#DIV/0!
#DIV/0!
A- 5
-------
Table 5. Day 7 Raw Endpoint Data.
Chemical Name: 0
CAS #: 0
Date (Day 7): 01/00/00
Nominal Replicate Individual Developmental
Concentration Number Identifier _^^ta2e__
Late
Stacje?_
Hindlimb
Length
(mm)
Snout to
Vent
Length
(mm)
Whole
Organism
Wet Weight
("Si
A - 6
-------
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
£
0
0
0
01/00/00
Whole
Organism
Wet Weight
(mg)
3
2
2
2
2
2
3
3
3
3
3
2
2
2
Replicate
Number
Individual
Identifier
Developmental
Stage
Late Stage?
Hindlimb
Length
(mm)
Snout to
Vent
Length
(mm)
A- 7
-------
Chemical Name:
CAS#:
Date (Day 7):
0
0
01/00/00
Whole
Organism
Wet Weight
(mg)
0 2
0 3
0 3
0 3
0 3
0 3
0
0
0
0
Nominal
Concentration
Replicate
Number
Individual
Identifier
Developmental
Stage
Late Stage?
Hindlimb
Length
(mm)
Snout to
Vent
Length
(mm)
A- 8
-------
Table 6. Day 7 Summary Endpoint Data.
CTMMiCU H4»«: 0
CAH * 0
L>j» T &unrr>fe> Data
1 3I.i;«
Hinal Line- L*Agtn
|HL'L!i
StIOUt Vsill Lt 'UU' IS-VL^t
«V«e 'ABIjnt
f
•
t
¦
¦f*
K"
£«&*
Hup.udLu
_
»
¦o
„
1
*
£
„
5
n
25
¦v
i
8
Hois »iui Qaia
UiiIU
taimtitt
i
1
t
J
t
*
#
bl
X
£
t
r
i
K
y.
0
0
1
|
WW
-
WW
0
WW
WW
WW
WW
C-
—m
WW
0
WW
u
m-
0
JW
0
4WV
JWIr
fWV
fWl
3
HJ
0
0
2
-
WW
0
-
i.
WW
G
WW
WW
WW
WW
c
—til
••
WW
0
WW
WW
WW
WW
:¦
m—
3
3
4W«
AWV
0
0
0
3
0
WW
•
-
c
WW
G
WW
WW
WW
WW
c
-rh
a
WW
Q
WW
WW
WW
WW
Cl
m—i
0
3
wv
*wv
4W«
»w«
D
".;j
0
e
4
WW
c
WW
0
WW
WW
WW
WW
0
-m
WW
0
WW
WW
WW
WW
3
¦SW*
0
i»w*
«w«
i»W(l
JWH
|Q
1 0
0
1
.
WW
9
•
0
WW
G
WW
WW
WW
WW
c
—m
CI
WW
0
WW
WW
WW
WW
"
m-
3
a
i"WM
HWf
JWV
sa
¦¦'A
0
D
2
WW
a
c
WW
c
WW
WW
WW
0'
-m
0
WW
0
WW
WW
WW
WW
0
m-
0
^w»i
0
0
0
D
3
WW
c
WW
c
WW
WW
WW
WW
c
-IT
0
WW
0
WW
WW
WW
WW
U
m-
3
0
fW«
i»WV
fW«
f»WM
j
0
0
4
-
WW-
-
•
0
WW
0
Q
-n,
::
WW
Q
WW
WW
WW
WW
0
m^-
3
0
4WV
(W(
<1WW
DM*
a
0
t>
1
WW
c
WW
o
WW
WW
WW
WW
G
-m
WW
0
WW
WW
WW
WW
Q
0
(*WV
PWV
|>WV
0
0
0
2
0
WW
t
0
c
WW
a
l_.ui
WW
WW
WW
G
¦ ¦¦.
0
WW
0
WW
WW
WW
WW
Cl
mrt
3
JWJ
0
*w«
wv
^wv
iWS
j
0
0
3
•
WW
..
.
a
WW
0
WW
WW
WW
WW
c
—m
WW
0
WW
WW
WW
WW
mm
3
|3W4l
3
*WV
*wv
in>(
^wv
a
0
D
4
WW
c
WW
0
WW
WW
WW
WW
C
••
0
WW
a
WW
WW
WW
WW
a
m-,
3
JWJ
0
JWt
¦awv
*w«
4W¥
3
ITU
0
D
1
WW
.
0
WW
0
WW
WW
WW
WW
G
—m
WW
a
WW
WW
WW
WW
0
ilWtl
|>WV
0
I 0
a
2
-
WW
•->
-
c
WW
c
WW
WW
WW
WW
c
-m
¦
WW
0
WW
WW
WW
WW
m-
3
<»Wt«
0
i«WM
JWt!
«WM
awt
>3
0
0
3
0
WW
•
0
c
WW
G
WW
WW
WW
WW
G
—m
0
WW
Q
WW
WW
WW
WW
a
m—
3
t»W«i
0
4WW
#w*
wv
JWV
3
0
0
*
WW
£
WW
G
WW
WW
WW
WW
c-
mtn
0
WW
Q
WW
WW
WW
WW
0
(T*S«
3
0
*wv
JWV
<>wv
3
OtOMi
Hon nai Dot* UiiIU
D 0
c a
d a
d a
D*v*luprnjrtl
M iuft
1 n s i
snum f 3 1 1
SWUM! WW i»h"> WW I'tiU WW
3NUM' WW WW WW WW WW
A- 9
-------
Table 7. Day 21 Raw Endpoint Data.
Chemical Name: 0
CAS #: 0
Date (Day 21): 01/00/00
Snout to Whole
Nominal Replicate Individual Developmental Hindlimb Vent Length Organism Wet Histology
Concentration Number Identifier Stage Late Stage? Length (mm) (mm) Weight (mg) (Y/N)
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
2
0
2
0
2
0
2
0
2
0
2
0
2
0
2
0
2
0
2
0
2
0
2
0
2
0
2
0
2
0
3
0
3
0
3
0
3
0
3
0
3
0
3
0
3
0
3
0
3
0
3
0
3
0
3
0
3
0
3
0
4
0
4
0
4
0
4
0
4
0
4
0
4
0
4
0
4
0
4
0
4
A- 10
-------
0
0_
_0
0
_0
_0
_0
_0
_0
_0
_0
0
0
0
0
0
0
0
0
0
_0
_0
_0
_0
_0
_0
_0
_0
0
_0
0
0
0
0
0
0
0
0
0_
0_
0
_0
_0
_0
_0
_0
_0
_0
_0
0
4
4_
4
2
2
2
2
2
2_
2_
2_
2_
2_
2_
2_
2_
2_
2
_3
£
1.
_3
_3
3
3
3
3
3
3
3
_3
_3
3
_4
_4
_4
_4
_4
_4
4_
_4
_4
_4
4
4
4
A- 11
-------
0_
0
0_
0
0_
0_
0
0_
0_
0
0_
0_
0
0_
0_
0
0_
0
0_
0_
0
0_
0_
0
0_
0_
0
0_
0_
0
0_
_0
0
G_
0_
0
0_
_0
0
G_
0_
0
0_
0_
0
0_
0_
0
0_
0_
0
0_
_0
0_
0_
0_
0
0_
_0
0
4
_2
_2
2
_2
_2
2
_2
_2
2
_2
_2
2
_2
_2
2
_3
_3
3
_3
_3
_3
_3
3
_3
_3
_3
_3
_3
3
_4
4
_4
_4
4_
4_
4_
4_
4_
_4
4_
4
A- 12
-------
0
J)
0_
0^
0_
0
0_
0^
0_
0
0^
0_
0_
0
0
0_
0_
0_
0
0
0_
0_
0_
0
0_
0_
0^
0_
0
0_
0_
0^
0_
0
0^
0_
0_
0
0
0_
0_
01
0
0
0_
0_
0_
0
0_
0_
0^
0_
0
0^
0_
0^
0_
0
0
_2
_2
_2
2
2
_2
_2
_2
2
_2
_2
_2
_2
2
£
_3
3^
_3
3
_3
_3
3^
_3
3
3^
£
3
_4
_4
_4
4
4
_4
_4
_4
4
_4
_4
_4
_4
4
4
A- 13
-------
Table 8. Day 21 Summary Endpoint Data.
Cfiimeati Man* &
CMS It &
LsT|i 21 Simnargr Osfci
Itapikata
Nonlnil Dcm Dow ¦Jnls liuubir
Lv#ilDpm«ribil Huge
Hind Unfa LanQlh |H.L|
Srreul Van! Langlh |SVLi
AW WdqM: |
±
1
1
JL
<
a
1
S
d-
J.
i.
J.
1
9
§
5
%
c
X
*
-
I
S
o
JL
1
r
J.
X
X
±
|
J-
j.
fe
B
A
X
•
if
8
o
X
J!
X
J.
±
JL
1
JL
i
?
s
c
:
<
a
X
s
•d
JL
£
r
! D
0
1
IMf
L»
o
.
c
ILT-VU
ifitatdi
tu
»;
irir
IM.JH
IL4V>:i
*uv*:i
MM
u
•uvc-
ICfk'.Vl
flUVCl
y
*
0
3
0
U
0
Ml*
.
•S-IV.IJI
#L-rwdi
m vrtl
u
—
—
—
rrrr
irir
1
mjMi
1
•UV«CI
IUV.CI
f
—
TITI
Till
:
MM
ij
»i -
ICfk'.VI
AUIVC
LI
—
—
ITU
ITU
I '¦>
0
4
IHW
Li
0
ffH
rjftVtll
tofwm
KJ Vtf
ij
irir
MJNl
c&v.ci
«L1V.*CI
ini
MM
IJ
nivc
ILft'.VI
ALKVCI
u
ITiJ
D
1
•ft
IMf
H
D
rn
•SftVU
mfwdi
•;
irir
INJH'
-
•wv*:i
iWV.«CI
mi
MM
¦ i -. -
IL'ffc'.WI
*1' . •
:¦
ITU
D
0
1
<0
fiW
D
0
G
fHftVUl
l-IV.UI
tu
u
IT IT
j
INJMl
0
ILWC"
ILIV^I
¦:
tlTI
c
MM
0
»i.« .¦
I Lf»'.*]"
iDivei
D
ITU
B
0
3
>&
IHI
0
~
:
_
IL'IVrOl
•LlflMJI
tu
u
IT IT
¦5
INJNl
0
•lwci
IUVJCI
c
TITI
c
MM
u
It.' V ft
IDfk'.llI
iHVCI
Q
ITU
0
4
MM
:>
0
l^7U>
m v..-
u
mm
Irir
WJH
•^V-J
ww
IWJJI
iWVCI
Li
ra
D
1
IMf
Li
d
rm
ILTV.U
I^IV.UI
•CI V.-'
i:
ITIT
'j
1-jJM
ILW"CI
•DtWCI
ii
IUVC
ILiOJI
iUVCI
LI
.
.
iru
0
$
d
II#
U
rn
•urv.u'
IHi
UftTUI
WW*-j
u
—
—
irir
gfeiM
Hm
IL4V.ICI
iL4V.»CI
""
u
iuvr.4
•WVA1I
flUVCI
U
—
ER
i d
0
4
IHI
IJ
ij
cr^ji
Ml*
fZWUI
fU V.I'
ij
—
—
ITIT
mjNj
MM*
IUVJCI
iwv.ti
—
TITI
1111
i.
IttVO
Illfk'.VI
•I -' . '
L»
—
—
"¦U
miJL
4>
0
1
<6
IW
:¦
0
TT
cc
i
•2rV.U
INV
lUIV.UI
IU V .-j
u
irir
I
MJMi
1
HHP
•dv»:i
iWVKI
-
1ITI
c
¦:
»! ¦ v -
ILffc'.-Ui
ittvci
Li
i-u
I 0
D
2
ft
IHW
Li
D
i
rTT!
i
i-ivaji
At#*
I'Jfk'.VI
ID ViD
u
ITIT
:¦
HM-IHM
:<
HIW
•o«i
•LW*CI
¦
1111
u
•uvrj
ID(VA]I
ilMVCi
;¦
ITU
*
0
3
6
IHI
D
0
i
WW
i
i-rvXr
MM
1-fCAl
IU
u
irir
:¦
H..W
:•
HIH
1 Mill
IWVKI
iUVCI
¦:
1111
mm
u
«UVft4
ILfVAJi
IUVC
"
ITU
I 0
B
4
IHI
LI
LI
UjhJ
cfvjji
•I#
I^UVjJ!
IU V.-'
IJ
irir
UlMM
ILWO
iLKCI
ImmI
¦ •
IEKHI
ILW.VI
iUVC!
L'
DivttofJlrtn
1il Stag* HLL BVL WW
i
Dot
¦ UriU J
1
9
I
s
1
G
a
awuu
•IMVJCI
IIM
iMIf
mi
•OVd
C
a
ItXV.CI
inn
IIMI
Mil
«xvfa
C
3
#NUM;
IIM
IMNI
Mil
AUDI
c
3
#MUP/
-------
9L - V
io/Aia#
io/Aia#
io/Aia#
iO/Aia#
io/Aia#
0
io/Aia#
io/Aia#
iO/Aia#
iO/Aia#
io/Aia#
0
io/Aia#
m/Aia#
io/Aia#
io/Aia#
m/Aia#
0
io/Ata#
io/Aia#
io/Aia#
io/Aia#
iO/Aia#
0
8LU|1 J3A0
sBEiSAV
\z *ea
n tea
l tea
o tea
asoa
| Bill 111 Of-J
io/Aia#
io/Aia#
io/Aia#
io/Aia#
0
0 0
u
OO/IS/IO
io/Aia#
so/Aia#
io/Aia#
io/Aia#
0
0 0
IZ
OO/LS/LO
io/Aia#
io/Aia#
io/Aia#
io/Aia#
0
0 0
\z
00/12/10
io/Aia#
io/Aia#
io/Aia#
io/Aia#
0
0 0
1z
OO/LS/IO
io/Aia#
io/Aia#
io/Aia#
io/Aia#
0
0 0
VI
00/l7t/l0
io/Aia#
io/Aia#
io/Aia#
iOMia#
0
0 0
VI
OO/H/IO
io/Aia#
io/Aia#
io/Aia#
io/Aia#
0
0 0
VI
OO/H/IO
io/Aia#
io/Aia#
io/Aia#
io/Aia#
0
0 0
VI
OO/H/IO
io/Aia#
iO/Aia#
io/Aia#
io/Aia#
0
0 0
I
OO/ZO/LO
io/Aia#
io/Aia#
io/Aia#
iOMia#
0
0 0
I
00/90/10
io/Aia#
io/Aia#
io/Aia*
io/Aia#
~
0 0
I
OO/ZO/l 0
io/Aia#
io/Aia#
iO/AIQ#
iO/Ald#
0
0 0
I
OO/ZO/LO
io/Aia#
iO/Aia#
io/Aia#
io/Aia#
0
0 0
0
00/00/10
io/Aia#
io/Aia#
io,'Aia#
io/Aia#
0
0 0
0
00/00/10
io/Aia#
io/Aia#
iO/Aia#
iOMia#
0
0 0
0
00/00/10
io/Aia#
io/Aia#
m/Aia#
io/Aia#
~
0 0
0
00/00/LO
(UG3M) UOHEJJU3DUOO
0 LI | SO Q | E?LI 1LUOJJ JO %
AO
as
ssoo
|L'nK'V uesw
£
Z
SljUfl 3SOQ
3SOQ
leiniuof.j
rtea }S3_L
fie a
jsqiuriN ajeojidsy
0 ¦# SV3
0 :3uieu |G0|iU3i|3
B}BQ AjJSILUSLjQ '6 9|qBl
-------
Total:
Dose
Animal ID -
replicate 2
Dose
Animal ID -
replicate 1
Thyroid gland
hypertrophy
Thyroid gland
atrophy
Follicular cell
hypertrophy
Follicular cell
hyperplasia
Total:
Dose
Animal ID -
replicate 2
Dose
Animal ID -
replicate 1
Thyroid gland
hypertrophy
Thyroid gland
atrophy
Follicular cell
hypertrophy
Follicular cell
hyperplasia
Total:
Control Animal
ID - replicate 2
Control Animal
ID - replicate 1
Thyroid gland
hypertrophy
Thyroid gland
atrophy
Follicular cell
hypertrophy
Follicular cell
hyperplasia
D
03
0)
CT
CD
I
go'
o
T3
Q3_
0 °
S" °
CD (Q
1 *<
S' 73
— CD
T3
O
=3
(Q
Total:
Dose
Animal ID -
replicate 2
Dose
Animal ID -
replicate 1
Thyroid gland
hypertrophy
Thyroid gland
atrophy
Follicular cell
hypertrophy
Follicular cell
hyperplasia
0)
a;
CD
V)
a
o
—s
CD
o
CD
Tl 13.
03 Q)
-------
Total:
Dose
Animal ID -
replicate 2
Dose
Animal ID -
replicate 1
Follicular lumen
area increase
Follicular lumen
area decrease
Total:
Dose
Animal ID -
replicate 2
Dose
Animal ID -
replicate 1
Follicular lumen
area increase
Follicular lumen
area decrease
Total:
Control Animal
ID - replicate 2
Control Animal
ID - replicate 1
Follicular lumen
area increase
Follicular lumen
area decrease
Total:
Dose
Animal ID -
replicate 2
Dose
Animal ID -
replicate 1
Follicular lumen
area increase
Follicular lumen
area decrease
-------
Dose
ID-
te 2
Dose
ID-
te 1
Animal
replica
Animal
replica
Dose
ID-
te 2
Dose
ID-
te 1
Animal
replica
Animal
replica
Dose
ID-
te 2
Dose
ID-
te 1
Animal
replica
Animal
replica
Control Animal
ID - replicate 2
Control Animal
ID - replicate 1
"0
03
o
3"
CD
3
O
03
0)
a;
CD
~ =r N>
0)
—s
—s
CD
<
CD
D
CD
V)
O
o
=3
V)
03
<
CD
Q.
CD
in
o
-a'
o'
I
go'
o
T3
CD
E?
O
O
eg
o'
0)
=3
Q.
=3'
CQ
V)
-------
APPENDIX 3
Alternative Analysis of Weight and Length in the Case of Late Stage Development
Exceeding 20% of Tadpoles in One Or More Concentration(s)
If an increased number of tadpoles show development beyond stage 60 (>20%) in one
or more nominal concentration(s), then a two-factor AN OVA with a nested variance
structure should be undertaken on all tadpoles to assess growth effects due to chemical
treatments while taking into account the effect of late stage development on growth.
The proposal is to use all data but take into account the effect of late stage
development. This can be done with a two-factor ANOVA with a nested variance
structure. Define LateStage-Yes' for an animal if its developmental stage is 61 or
greater. Otherwise, define LateStage-No'. Then a two-factor ANOVA with
concentration and LateStage and their interaction can be done, with Rep(Conc) a
random factor and Tadpole(Rep) another random effect. This still treats the rep as the
unit of analysis and gives essentially the same results as a weighted analysis of
rep*latestage means, weighted by the number of animals per mean. If the data violate
the normality or variance homogeneity requirements of ANOVA, then a normalized
rank-order transform can be done to remove that objection.
In addition to the standard ANOVA F-tests for the effects of CONC, LateStage, and their
interactions, the interaction F-test can be "sliced" into two additional ANOVA F-tests,
one on the mean responses across concentrations for LateStage='No' and another on
the mean responses across concentrations for LateStage-Yes.' Further comparisons
of treatment means against controls are done within each level of LateStage. A trend-
type analysis can be done using appropriate contrasts or simple pairwise comparisons
can be done if there is evidence of non-monotone dose-response within a level of the
LateStage variable. A Bonferroni-Holm adjustment to the p-values is made only if the
corresponding F-slice is not significant. This can be done in SAS and, presumably,
other statistical software packages. Complications can arise when there are no late
stage animals in some concentrations, but these situations can be handled in a straight-
forward fashion.
A- 19
------- |