S2(R1) Genotoxicity Testing and Data Interpretation for Pharmaceuticals Intended for Human Use

Administrative guidance, not enacted legislation. This captured publisher edition preserves the listed document and attachments; later developments are not independently consolidated.

Guidance for Industry
S2(R1) Genotoxicity Testing
and Data Interpretation for
Pharmaceuticals Intended for
Human Use
U.S. Department of Health and Human Services
Food and Drug Administration
Center for Drug Evaluation and Research (CDER)
Center for Biologics Evaluation and Research (CBER)
June 2012
ICH

Guidance for Industry
S2(R1) Genotoxicity Testing
and Data Interpretation for
Pharmaceuticals Intended for
Human Use
Additional copies are available from:
Office of Communications
Division of Drug Information, WO51, Room 2201
Center for Drug Evaluation and Research
Food and Drug Administration
10903 New Hampshire Ave., Silver Spring, MD 20993-0002
Phone: 301-796-3400; Fax: 301-847-8714
druginfo@fda.hhs.gov
http://www.fda.gov/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/default.htm
and/or
Office of Communication, Outreach and
Development, HFM-40
Center for Biologics Evaluation and Research
Food and Drug Administration
1401 Rockville Pike, Rockville, MD 20852-1448
http://www.fda.gov/BiologicsBloodVaccines/GuidanceComplianceRegulatoryInformation/Guidances/default.htm
(Tel) 800-835-4709 or 301-827-1800
U.S. Department of Health and Human Services
Food and Drug Administration
Center for Drug Evaluation and Research (CDER)
Center for Biologics Evaluation and Research (CBER)
June 2012
ICH

Contains Nonbinding Recommendations
TABLE OF CONTENTS
I. 	INTRODUCTION (1)....................................................................................................... 1
A. 	Objectives of the Guidance (1.1)................................................................................................... 1
B. 	Background (1.2)............................................................................................................................ 1
C. 	Scope of the Guidance (1.3)........................................................................................................... 2
D. 	General Principles (1.4) ................................................................................................................. 2
II. 	THE STANDARD TEST BATTERY FOR GENOTOXICITY (2) ............................. 2
A. 	Rationale (2.1) ................................................................................................................................ 2
B. 	Description of the Two Options for the Standard Battery (2.2) ................................................ 3
C. 	Modifications to the Test Battery (2.3)......................................................................................... 5
1. Exploratory Clinical Studies (2.3.1) ................................................................................................ 5
2. Testing Compounds That Are Toxic to Bacteria (2.3.2) .................................................................. 5
3. Compounds Bearing Structural Alerts for Genotoxic Activity (2.3.3) ............................................. 5
4. Limitations to the Use of In Vivo Tests (2.3.4)................................................................................. 6
D. 	Detection of Germ Cell Mutagens (2.4)........................................................................................ 6
III. 	RECOMMENDATIONS FOR IN VITRO TESTS (3).................................................. 6
A. 	Test Repetition and Interpretation (3.1) ...................................................................................... 6
B. 	Recommended Protocol for the Bacterial Mutation Assay (3.2) ............................................... 6
1. Selection of Top Dose Level (3.2.1) ................................................................................................. 6
2. Study Design/Test Protocol (3.2.2).................................................................................................. 7
C. 	Recommended Protocols for the Mammalian Cell Assays (3.3) ................................................ 7
1. Selection of Top Concentration (3.3.1)............................................................................................ 8
2. Study Design/Test Protocols (3.3.2) ................................................................................................ 8
3. Positive Controls (3.3.3) .................................................................................................................. 9
IV. 	RECOMMENDATIONS FOR IN VIVO TESTS (4) .................................................... 9
A. 	Tests for the Detection of Chromosome Damage In Vivo (4.1).................................................. 9
B. 	Other In Vivo Genotoxicity Tests (4.2) ........................................................................................ 9
C. 	Dose Selection for In Vivo Assays (4.3) ...................................................................................... 10
1. Short-term Studies (4.3.1) .............................................................................................................. 10
2. Multiple Administration Studies (4.3.2)......................................................................................... 10
3. Testing Compounds That Are Toxic for Blood or Marrow (4.3.3) ................................................ 11
D. 	Demonstration of Target Tissue Exposure for Negative In Vivo Test Results (4.4) .............. 11
1. When an In Vitro Genotoxicity Test Is Positive (or Not Done) (4.4.1).......................................... 12
2. When In Vitro Genotoxicity Tests Are Negative (4.4.2)................................................................. 12
E. 	Sampling Times for In Vivo Assays (4.5)................................................................................... 13

Contains Nonbinding Recommendations
F. 	Number of Animals Analyzed (4.6) ............................................................................................ 13
G. 	Use of Male/Female Rodents in In Vivo Genotoxicity Tests (4.7)............................................ 13
H. 	Route of Administration (4.8) ..................................................................................................... 13
I. 	Use of Positive Controls for In Vivo Studies (4.9) ..................................................................... 14
V. 	GUIDANCE ON EVALUATION OF TEST RESULTS AND ON FOLLOW-UP
TEST STATEGIES (5) ................................................................................................... 14
A. 	Assessment of Biological Relevance (5.1)................................................................................... 14
B. 	Evaluation of Results Obtained in In Vitro Tests (5.2) ............................................................ 15
1. Evaluation of Positive Results Obtained In Vitro in a Bacterial Mutation Assay (5.2.1).............. 15
2. Evaluation of Positive Results Obtained In Vitro in Mammalian Cell Assays (5.2.1)................... 15
3. Evaluation of In Vitro Negative Results (5.2.3) ............................................................................. 16
C. 	Evaluation of Results Obtained From In Vivo Tests (5.3) ....................................................... 16
D. 	Follow-up Strategies for Positive Results (5.4).......................................................................... 16
1. Follow-up to Findings In Vitro in Mammalian Cell Tests (5.4.1) ................................................. 16
2. Follow-up to a Positive In Vivo Micronucleus Assay (5.4.2) ........................................................ 18
E. 	Follow-up Genotoxicity Testing in Relation to Tumor Findings in a Carcinogenicity
Bioassay (5.5)................................................................................................................................ 18
VI. 	NOTES (6) ....................................................................................................................... 19
VII. 	GLOSSARY (7)............................................................................................................... 25
VIII. 	REFERENCES (8) .......................................................................................................... 28

Contains Nonbinding Recommendations
Guidance for Industry1
S2(R1) Genotoxicity Testing and Data Interpretation for
Pharmaceuticals Intended for Human Use
This guidance represents the Food and Drug Administration’s (FDA’s) current thinking on this topic. It
does not create or confer any rights for or on any person and does not operate to bind FDA or the public.
You can use an alternative approach if the approach satisfies the requirements of the applicable statutes
and regulations. If you want to discuss an alternative approach, contact the FDA staff responsible for
implementing this guidance. If you cannot identify the appropriate FDA staff, call the appropriate
number listed on the title page of this guidance.
I. 	INTRODUCTION (1)2
A. 	Objectives of the Guidance (1.1)
This guidance combines and replaces two ICH guidances, S2A Specific Aspects for Regulatory
Genotoxicity Tests for Pharmaceuticals (ICH S2A guidance) and S2B Genotoxicity: A Standard
Battery for Genotoxicity Testing of Pharmaceuticals (ICH S2B guidance). The purpose of the
revision is to provide guidance on optimizing the standard genetic toxicology battery for
prediction of potential human risks, and on interpreting results, with the goal of improving risk
characterization for carcinogenic effects that have their basis in changes in the genetic material.
The revised guidance describes internationally agreed-upon standards for follow-up testing and
interpretation of positive results in vitro and in vivo in the standard genetic toxicology battery,
including assessment of nonrelevant findings. This guidance is intended to apply only to
products being developed as human pharmaceuticals.
B. 	Background (1.2)
The recommendations from the latest Organization for Economic Co-operation and Development
(OECD) guidelines and the reports from the International Workshops on Genotoxicity Testing
(IWGT) have been considered where relevant. In certain cases, the recommendations in this
1 This guidance was developed within the Safety Implementation Working Group of the International Conference on
Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) and has been
subject to consultation by the regulatory parties, in accordance with the ICH process. This document has been
endorsed by the ICH Steering Committee at Step 4 of the ICH process, November 2011. At Step 4 of the process,
the final draft is recommended for adoption to the regulatory bodies of the European Union, Japan, and the United
States.
2 Arabic numbers reflect the organizational breakdown of the document endorsed by the ICH Steering Committee at
Step 4 of the ICH process, November 2011.
1

Contains Nonbinding Recommendations
guidance are different from the OECD or the IWGT recommendations, and are noted in the text.
The following notes for guidance should be applied in conjunction with other ICH guidances.
C. 	Scope of the Guidance (1.3)
The focus of this guidance is testing of new small molecule drug substances, and the guidance
does not apply to biologics. Advice on the timing of the studies relative to clinical development
is provided in the ICH guidance M3(R2) Nonclinical Safety Studies for the Conduct of Human
Clinical Trials and Marketing Authorization for Pharmaceuticals (ICH M3(R2) guidance).3 The
recommendations in the guidance should be applied in conjunction with other ICH guidances.
D. 	General Principles (1.4)
Genotoxicity tests can be defined as in vitro and in vivo tests designed to detect compounds that
induce genetic damage by various mechanisms. These tests enable hazard identification with
respect to damage to DNA and its fixation. Fixation of damage to DNA in the form of gene
mutations, larger scale chromosomal damage, or recombination is generally considered to be
essential for heritable effects and in the multistep process of malignancy, a complex process in
which genetic changes might possibly play only a part. Numerical chromosome changes have
also been associated with tumorigenesis and can indicate a potential for aneuploidy in germ cells.
Compounds that are positive in tests that detect such kinds of damage have the potential to be
human carcinogens and/or mutagens. Because the relationship between exposure to particular
chemicals and carcinogenesis is established for humans, although a similar relationship has been
difficult to prove for heritable diseases, genotoxicity tests have been used mainly for the
prediction of carcinogenicity. Nevertheless, because germ line mutations are clearly associated
with human disease, the suspicion that a compound might induce heritable effects is considered
to be just as serious as the suspicion that a compound might induce cancer. In addition, the
outcome of genotoxicity tests can be valuable for interpreting carcinogenicity studies.
II. 	THE STANDARD TEST BATTERY FOR GENOTOXICITY (2)
A. 	Rationale (2.1)
Registration of pharmaceuticals requires a comprehensive assessment of their genotoxic
potential. Extensive reviews have shown that many compounds that are mutagenic in the
bacterial reverse mutation (Ames) test are rodent carcinogens. Addition of in vitro mammalian
tests increases sensitivity for detection of rodent carcinogens and broadens the spectrum of
genetic events detected, but also decreases the specificity of prediction (i.e., increases the
incidence of positive results that do not correlate with rodent carcinogenicity). Nevertheless, a
battery approach is still reasonable because no single test is capable of detecting all genotoxic
mechanisms relevant in tumorigenesis.
3 We update guidances periodically. To make sure you have the most recent version of a guidance, check the FDA
Drugs guidance Web page at
http://www.fda.gov/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/default.htm.
2

Contains Nonbinding Recommendations
The general features of a standard test battery are as follows:
i. 	Assessment of mutagenicity in a bacterial reverse gene mutation test. This test has
been shown to detect relevant genetic changes and the majority of genotoxic rodent
and human carcinogens.
ii. 	Genotoxicity should also be evaluated in mammalian cells in vitro and/or in vivo as
follows.
Several in vitro mammalian cell systems are widely used and can be considered sufficiently
validated: the in vitro metaphase chromosome aberration assay, the in vitro micronucleus assay
(note 1) and the mouse lymphoma L5178Y cell Tk (thymidine kinase) gene mutation assay
(MLA). These three assays are currently considered equally appropriate and therefore
interchangeable for measurement of chromosomal damage when used together with other
genotoxicity tests in a standard battery for testing of pharmaceuticals, if the test protocols
recommended in this guidance are used. In vivo test(s) are included in the test battery because
some agents are mutagenic in vivo but not in vitro (note 2) and because it is desirable to include
assays that account for such factors as absorption, distribution, metabolism, and excretion. The
choice of an analysis either of micronuclei in erythrocytes (in blood or bone marrow), or of
chromosome aberrations in metaphase cells in bone marrow, is currently included for this reason
(note 3). Lymphocytes cultured from treated animals can also be used for cytogenetic analysis,
although experience with such analyses is less widespread.
In vitro and in vivo tests that measure chromosomal aberrations in metaphase cells can detect a
wide spectrum of changes in chromosomal integrity. Breakage of chromatids or chromosomes
can result in micronucleus formation if an acentric fragment is produced; therefore, assays that
detect either chromosomal aberrations or micronuclei are considered appropriate for detecting
clastogens. Micronuclei can also result from lagging of one or more whole chromosome(s) at
anaphase and thus micronucleus tests have the potential to detect some aneuploidy inducers. The
MLA detects mutations in the Tk gene that result from both gene mutations and chromosome
damage. There is some evidence that MLA can also detect chromosome loss.
There are several additional in vivo assays that can be used in the battery or as follow-up tests to
develop weight of evidence in assessing results of in vitro or in vivo assays (see below).
Negative results in appropriate in vivo assays (usually two), with adequate justification for the
endpoints measured, and demonstration of exposure (see section IV.D (4.4)) are generally
considered sufficient to demonstrate absence of significant genotoxic risk.
B. 	Description of the Two Options for the Standard Battery (2.2)
The following two options for the standard battery are considered equally suitable (see note 4):
Option 1
i. 	A test for gene mutation in bacteria.
3

Contains Nonbinding Recommendations
ii. 	A cytogenetic test for chromosomal damage (the in vitro metaphase chromosome
aberration test or in vitro micronucleus test), or an in vitro mouse lymphoma Tk gene
mutation assay.
iii. An in vivo test for genotoxicity, generally a test for chromosomal damage using
rodent hematopoietic cells, either for micronuclei or for chromosomal aberrations in
metaphase cells.
Option 2
i. 	A test for gene mutation in bacteria.
ii. 	An in vivo assessment of genotoxicity with two different tissues, usually an assay for
micronuclei using rodent hematopoietic cells and a second in vivo assay. Typically,
this would be a DNA strand breakage assay in liver, unless otherwise justified (see
below; also section IV.B (4.2) and note 12).
There is more historical experience with Option 1, partly because it is based on the ICH S2A
guidance and the ICH S2B guidance. Nevertheless, the reasoning behind considering Options 1
and 2 equally suitable is as follows: When a positive result occurs in an in vitro mammalian cell
assay, clearly negative results in two well-conducted in vivo assays, in appropriate tissues and
with demonstrated adequate exposure, are considered sufficient evidence for lack of genotoxic
potential in vivo (see section V.D.1.i (5.4.1.1) below). Thus, a test strategy in which two in vivo
assays are conducted is the same strategy that should be used to follow up a positive result in
vitro (see note 4).
Under both standard battery options, either acute or repeat-dose study designs in vivo can be
used. In case of repeated administrations, attempts should be made to incorporate the
genotoxicity endpoints into toxicity studies, if scientifically justified. When more than one
endpoint is evaluated in vivo, it is preferable that they are incorporated into a single study. Often
sufficient information on the likely suitability of the doses for the repeat-dose toxicology study is
available before the study begins and can be used to determine whether an acute or an integrated
test would be suitable.
For compounds that give negative results, the completion of either option of the standard test
battery, performed and evaluated in accordance with current recommendations, will usually
provide sufficient assurance of the absence of genotoxic activity and no additional tests are
warranted. Compounds that give positive results in the standard test battery might, depending on
their therapeutic use, need to be tested more extensively (see section V (5)).
There are several in vivo assays that can be used as the second part of the in vivo assessment
under Option 2 (see section IV.B (4.2)), some of which can be integrated into repeat-dose
toxicology studies. The liver is typically the preferred tissue because of exposure and
metabolizing capacity, but choice of in vivo tissue and assay should be based on factors such as
any knowledge of the potential mechanism, of the metabolism in vivo, or of the exposed tissues
thought to be relevant.
4

Contains Nonbinding Recommendations
Information on numerical changes can be derived from the mammalian cell assays in vitro and
from the micronucleus assays in vitro or in vivo. Elements of the standard protocols that can
indicate such potential are elevations in the mitotic index, polyploidy induction, and
micronucleus evaluation. There is also experimental evidence that spindle poisons can be
detected in MLA. The preferred in vivo cytogenetic test under Option 2 is the micronucleus
assay, not a chromosome aberration assay, to include more direct capability for detection of
chromosome loss (potential for aneuploidy).
The suggested standard set of tests does not imply that other genotoxicity tests are generally
considered inadequate or inappropriate. Additional tests can be used for further investigation of
genotoxicity test results obtained in the standard battery (see sections IV.B (4.2) and V (5)).
Alternative species, including nonrodents, can also be used if indicated, and if sufficiently
validated.
Under conditions in which one or more tests in the standard battery cannot be employed for
technical reasons, alternative validated tests can serve as substitutes, provided sufficient
scientific justification is given.
C. 	Modifications to the Test Battery (2.3)
The following sections describe situations where modification of the standard test battery might
be advisable.
1. 	Exploratory Clinical Studies (2.3.1)
For certain exploratory clinical studies, fewer genotoxicity assays or different criteria for
justification of the maximum dose in vivo might apply (see ICH M3(R2) guidance).
2. 	Testing Compounds That Are Toxic to Bacteria (2.3.2)
In cases where compounds are highly toxic to bacteria (e.g., some antibiotics), the bacterial
reverse mutation (Ames) test should still be carried out, just as cytotoxic compounds are tested in
mammalian cells, because mutagenicity can occur at lower, less toxic concentrations. In such
cases, any one of the in vitro mammalian cell assays should also be done (i.e., Option 1 should
be followed).
3. 	Compounds Bearing Structural Alerts for Genotoxic Activity (2.3.3)
Structurally alerting compounds (note 5) are usually detectable in the standard test battery since
the majority of structural alerts are defined in relation to bacterial mutagenicity. A few chemical
classes are known to be more easily detected in mammalian cell chromosome damage assays
than bacterial mutation assays. Thus, negative results in either test battery with a compound that
has a structural alert is usually considered sufficient assurance of a lack of genotoxicity.
However, for compounds bearing certain specific structural alerts, modification to standard
protocols can be appropriate (note 5). The choice of additional test(s) or protocol
5

Contains Nonbinding Recommendations
modification(s) depends on the chemical nature, the known reactivity, and any metabolism data
on the structurally alerting compound in question.
4. 	Limitations to the Use of In Vivo Tests (2.3.4)
There are compounds for which many in vivo tests (typically in bone marrow, blood, or liver) do
not provide additional useful information. These include compounds for which data on
toxicokinetics or pharmacokinetics indicate that the compounds are not systemically absorbed
and therefore are not available to the target tissues. Examples of such compounds are some
radioimaging agents, aluminum-based antacids, some compounds given by inhalation, and some
dermally or other topically applied pharmaceuticals. In cases where a modification of the route
of administration does not provide sufficient target tissue exposure, and no suitable genotoxicity
assay is available in the most exposed tissue, it might be appropriate to base the evaluation only
on in vitro testing. In some cases, evaluation of genotoxic effects at the site of contact can be
warranted, although such assays have not yet been widely used (note 6).
D. 	Detection of Germ Cell Mutagens (2.4)
Results of comparative studies have shown that, in a qualitative sense, most germ cell mutagens
are likely to be detected as genotoxic in somatic cell tests so that negative results of in vivo
somatic cell genotoxicity tests generally indicate the absence of germ cell effects.
III. 	RECOMMENDATIONS FOR IN VITRO TESTS (3)
A. 	Test Repetition and Interpretation (3.1)
Reproducibility of experimental results is an essential component of research involving novel
methods or unexpected findings; however, the routine testing of drugs with standard, widely
used genotoxicity tests often does not call for replication. These tests are sufficiently well-
characterized and have sufficient internal controls that repetition of a clearly positive or negative
assay is not usually warranted. Ideally, it should be possible to declare test results clearly
negative or clearly positive. However, test results sometimes do not fit the predetermined
criteria for a positive or negative call and therefore are declared equivocal. The application of
statistical methods can aid in data interpretation; however, adequate biological interpretation is
of critical importance. An equivocal test that is repeated might result in (1) a clearly positive
outcome, and thus an overall positive result; (2) a negative outcome, so that the result is not
reproducible and overall negative, or (3) another equivocal result, with a final conclusion that
remains equivocal.
B. 	Recommended Protocol for the Bacterial Mutation Assay (3.2)
Advice on the protocols is given in the OECD guidelines (1997) and the IWGT report
(Gatehouse et al., 1994).
1. 	Selection of Top Dose Level (3.2.1)
6

Contains Nonbinding Recommendations
Maximum dose level
The maximum dose level recommended is 5000 micrograms (μg)/plate (or 5 microliters
(μL)/plate for liquid test substance) when not limited by solubility or cytotoxicity.
Limit of solubility
For bacterial cultures, precipitating doses are scored, provided precipitate does not interfere with
scoring, toxicity is not limiting the evaluation of the assay, and the top concentration does not
exceed 5000 μg/plate (or 5 μL/plate for liquid test substance). If no cytotoxicity is observed,
then the lowest precipitating dose should be used as the top dose scored. If dose-related
cytotoxicity or mutagenicity is noted, irrespective of solubility, the top dose scored should be
based on cytotoxicity as described below.
Limit of cytotoxicity
In the Ames test, the doses scored should show evidence of significant toxicity, but
without exceeding a top dose of 5000 μg/plate. Toxicity might be detected by a reduction in the
number of revertants, and/or clearing or diminution of the background lawn.
2. 	Study Design/Test Protocol (3.2.2)
The recommended set of bacterial strains (OECD) includes those that detect base substitution
and frameshift mutations as follows:
 	Salmonella typhimurium TA98
 	Salmonella typhimurium TA100
 	Salmonella typhimurium TA1535
 	Salmonella typhimurium TA1537 or TA97 or TA97a
 	Salmonella typhimurium TA102 or Escherichia coli WP2 uvrA or Escherichia coli
WP2 uvrA (pKM101)
One difference from the OECD and the IWGT recommendations is that based on experience
with testing pharmaceuticals, a single bacterial mutation (Ames) test is considered sufficient
when it is clearly negative or positive, and is carried out with a fully adequate protocol, including
all strains with and without metabolic activation, a suitable dose range that fulfills criteria for top
dose selection, and appropriate positive and negative controls. Also, for testing pharmaceuticals,
either the plate incorporation or the pre-incubation method is considered appropriate for this
single experiment (note 7). Equivocal or weak positive results might indicate that it would be
appropriate to repeat the test, possibly with a modified protocol such as appropriate spacing of
dose levels.
C. 	Recommended Protocols for the Mammalian Cell Assays (3.3)
Advice on the protocols is given in the OECD guidelines (1997) and the IWGT publications
(e.g., Kirsch-Volders et al., 2003; Moore et al., 2006); advice on interpreting MLA results is also
7

Contains Nonbinding Recommendations
given (Moore et al., 2006), including use of a global evaluation factor. Several differences from
these recommendations are noted here for testing pharmaceuticals, notably for selection of the
top concentration (see details below).
1. 	Selection of Top Concentration (3.3.1)
Maximum concentration
The maximum top concentration recommended is 1 millimolar (mM) or 0.5 milligram
(mg)/milliliter (mL), whichever is lower, when not limited by solubility in solvent or culture
medium or by cytotoxicity (note 8).
Limit of solubility
When solubility is limiting the ability to achieve the maximal concentration, the maximum
concentration, if not limited by cytotoxicity, should be the lowest concentration at which
minimal precipitate is visible in cultures, provided there is no interference with scoring.
Evaluation of precipitation can be done by naked eye or by methods such as light microscopy,
noting precipitate that persists or appears during culture (by the end of treatment).
Cytotoxicity
For in vitro cytogenetic assays for metaphase chromosome aberrations or for micronuclei,
cytotoxicity should not exceed a reduction of about 50 percent in cell growth (notes 9 and 10).
For the MLA, at the top dose there should be 80 to 90 percent cytotoxicity as measured by a
relative total growth (RTG) between 20 to 10 percent (note 9).
2. 	Study Design/Test Protocols (3.3.2)
For the cytogenetic evaluation of chromosomal damage in metaphase cells in vitro, the test
protocol should include the conduct of tests with and without metabolic activation, with
appropriate positive and negative controls. Treatment with the test articles should be for 3 to 6
hours with a sampling time approximately 1.5 normal cell cycles from the beginning of the
treatment. A continuous treatment without metabolic activation up to the sampling time of
approximately 1.5 normal cell cycles should be conducted in case of negative or equivocal
results for both short treatments, with and without metabolic activation. The same principles
apply to the in vitro micronucleus assay, except that the sampling time is typically 1.5 to 2
normal cell cycles from the beginning of treatment to allow cells to complete mitosis and enter
the next interphase. For both in vitro cytogenetic assays, there might be a need to modify the
protocol for certain types of chemicals that could be more readily detected by longer treatment,
delayed sampling times, or recovery periods (e.g., some nucleoside analogues and some
nitrosamines). In the metaphase aberration assay, information on the ploidy status should be
obtained by recording the incidence of polyploid (including endoreduplicated) metaphases as a
percentage of the number of metaphase cells. For MLA, the test protocol should include the
conduct of tests with and without metabolic activation, with appropriate positive and negative
controls, where the treatment with the test article is for 3 to 4 hours. A continuous treatment
8

Contains Nonbinding Recommendations
without metabolic activation for approximately 24 hours should be conducted in case of a
negative or equivocal result for both short treatments, with and without metabolic activation. A
standard MLA should include (1) the incorporation of positive controls that induce mainly small
colonies and (2) colony sizing for positive controls, solvent controls, and at least one positive test
compound concentration (should any exist), including the culture that gave the greatest mutant
frequency.
For mammalian cell assays in vitro, built-in confirmatory elements, such as those outlined above
(e.g., different treatment lengths, tests with and without metabolic activation), should be used.
Following such testing, further confirmatory testing in the case of clearly negative or positive
test results is not usually warranted. Equivocal or weak positive results might call for repeating
tests, possibly with a modified protocol such as appropriate spacing of the test concentrations.
3. 	Positive Controls (3.3.3)
Concurrent positive controls are important, but in vitro mammalian cell tests for genetic toxicity
are sufficiently standardized that use of positive controls can generally be confined to a positive
control with metabolic activation (when it is done concurrently with the non-activated test) to
demonstrate the activity of the metabolic activation system and the responsiveness of the test
system.
IV. 	RECOMMENDATIONS FOR IN VIVO TESTS (4)
A. 	Tests for the Detection of Chromosome Damage In Vivo (4.1)
Either the analysis of chromosomal aberrations or the measurement of micronucleated
polychromatic erythrocytes in bone marrow cells in vivo is considered appropriate for the
detection of clastogens. Both rats and mice are considered appropriate for use in the bone
marrow micronucleus test. Micronuclei can also be measured in immature (e.g., polychromatic)
erythrocytes in peripheral blood in the mouse, or in the newly formed reticulocytes in rat blood
(note 3). Likewise, immature erythrocytes can be used from any other species that has shown an
adequate sensitivity to detect clastogens/aneuploidy inducers in bone marrow or peripheral blood
(note 3). Systems for automated analysis (image analysis and flow cytometry) can be used if
appropriately validated (OECD, 1997; Hayashi et al., 2000; 2007). Chromosomal aberrations
can also be analyzed in peripheral lymphocytes cultured from treated rodents (note 11).
B. 	Other In Vivo Genotoxicity Tests (4.2)
The same in vivo tests described as the second test in the standard battery (Option 2) can be used
as follow-up tests to develop weight of evidence in assessing results of in vitro or in vivo assays
(notes 11 and 12). Although the type of effect seen in vitro and any knowledge of the
mechanism can help guide the choice of in vivo assay, investigation of chromosomal aberrations
or of gene mutations in endogenous genes is not feasible with standard methods in most tissues.
Although mutation can be measured in transgenes in rodents, this entails prolonged treatment
(e.g., 28 days) to allow for mutation expression, fixation, and accumulation, especially in tissues
with little cell division (see note 12). Thus the second in vivo assay will often evaluate a DNA
9

Contains Nonbinding Recommendations
damage endpoint as a surrogate. Assays with the most published experience and advice on
protocols include the DNA strand break assays, such as the single cell gel electrophoresis
(Comet) assay and alkaline elution assay, the in vivo transgenic mouse mutation assays and DNA
covalent binding assays (all of which can be applied in many tissues (note 12)), and the liver
unscheduled DNA synthesis (UDS) assay.
C. 	Dose Selection for In Vivo Assays (4.3)
Typically, three dose levels are analyzed (Hayashi et al., 2007).
1. 	Short-term Studies (4.3.1)
For short-term (usually 1 to 3 administrations) studies, the top dose recommended for
genotoxicity assays is a limit dose of 2000 mg/kilogram (kg), if this is tolerated, or a maximum
tolerated dose defined (for example, for the micronucleus assay (OECD)) as the dose producing
signs of toxicity such that higher dose levels, based on the same dosing regimen, would be
expected to produce lethality. Similar recommendations have been made for the Comet assay
(Hartmann et al., 2003) and transgenic mutation assay (Heddle et al., 2000). Suppression of
bone marrow red blood cell production should also be taken into account in dose selection.
Lower doses are generally spaced at approximately two- to three-fold intervals below this dose.
2. 	Multiple Administration Studies (4.3.2)
Option 1 Battery: When the in vivo genotoxicity test is integrated into a multiple administration
toxicology study, the doses are generally considered appropriate when the toxicology study
meets the criteria for an adequate study to support human clinical trials; this can differ from dose
selection criteria in the OECD guidelines for the in vivo micronucleus assay. This applies when
the in vitro mammalian cell test is negative (or nonrelevant positive (see section V (5))).
Follow-up studies or Option 2 battery: When carrying out follow-up studies to address any
indication of genotoxicity, or when using Option 2 with no in vitro mammalian cell assay,
several factors should be evaluated to determine whether the top dose is appropriate for
genotoxicity evaluation. Any one of the criteria listed below is considered sufficient to
demonstrate that the top dose in a toxicology study (typically in rats) is appropriate for
micronucleus analysis and for other genotoxicity evaluation:
i. 	Maximum feasible dose (MFD) based on physicochemical properties of the drug in
the vehicle (provided the MFD in that vehicle is similar to that achievable with acute
administration; note 13).
ii. 	Limit dose of 1000 mg/kg for studies of 14 days or longer, if this is tolerated.
iii. Maximal possible exposure demonstrated either by reaching a plateau/saturation in
exposure or by compound accumulation. In contrast, substantial reduction in
exposure to parent drug with time (e.g.,  50% reduction from initial exposure) can
disqualify the study (unless a blood sample taken in the first few days is available). If
10

Contains Nonbinding Recommendations
this is seen in one sex, generally the sex with reduced exposure would not be scored
at the end of the study, unless there is enhanced exposure to a metabolite of interest.
iv. Top dose is  50 percent of the top dose that would be used for acute administration,
i.e., close to the minimum lethal dose, if such acute data are available for other
reasons. (The top dose for acute administration micronucleus tests is currently
described in the OECD guidelines as the dose above which lethality would be
expected; similar guidance is given (e.g., Hartmann et al., 2003) for other in vivo
assays.)
Selection of a top dose based only on an exposure margin (multiple over clinical exposure)
without toxicity is not considered sufficient justification.
3. 	Testing Compounds That Are Toxic for Blood or Marrow (4.3.3)
Many compounds that induce aneuploidy, such as potent spindle poisons, are detectable in in
vivo micronucleus assays in bone marrow or blood only within a narrow range of doses
approaching toxic doses. This is also true for some clastogens. If toxicological data indicate
severe toxicity to the red blood cell lineage (e.g., marked suppression of PCEs (polychromatic
erythrocytes) or reticulocytes), doses scored should be spaced not more than about two fold
below the top, cytotoxic dose. If suitable doses are not included in a multiweek study, additional
data that could contribute to the detection of aneugens and some toxic clastogens could be
derived from any one of the following:
i. 	Early blood sampling (at 3 to 4 days) is advisable when there are marked increases in
toxicity with increasing treatment time. For example, when blood or bone marrow is
used for micronucleus measurement in a multiweek study (e.g., 28 days), and
reticulocytes are scored, marked hematotoxicity can affect the ability to detect
micronuclei (i.e., a dose that induces detectable increases in micronuclei after acute
treatment might be too toxic to analyze after multiple treatments (Hamada et al.,
2001)). The early sample can be used to provide assurance that clastogens and
potential aneugens are detected (but see notes 14 and 15).
ii. 	An in vitro mammalian cell micronucleus assay.
iii. An acute bone marrow micronucleus assay.
D. Demonstration of Target Tissue Exposure for Negative In Vivo Test Results (4.4)
In vivo tests have an important role in genotoxicity test strategies. The value of in vivo results is
directly related to the demonstration of adequate exposure of the target tissue to the test
compound. This is especially true for negative in vivo test results when in vitro test(s) have
shown convincing evidence of genotoxicity, or when no in vitro mammalian cell assay is used.
Evidence of adequate exposure could include toxicity in the tissue in question, or toxicokinetic
data as described in the following section.
11

Contains Nonbinding Recommendations
1. 	When an In Vitro Genotoxicity Test Is Positive (or Not Done) (4.4.1)
Assessments of in vivo exposure should be made at the top dose or other relevant doses using the
same species, strain, and dosing route used in the genotoxicity assay. When genotoxicity is
measured in toxicology assays, exposure information is generally available as part of the
toxicology assessment.
Demonstration of in vivo exposure should be made by any of the following measurements:
i. 	Cytotoxicity
a. 	For cytogenetic assays: By obtaining a significant change in the proportion of
immature erythrocytes among total erythrocytes in the tissue used (bone marrow
or blood) at the doses and sampling times used in the micronucleus test or by
measuring a significant reduction in mitotic index for the chromosomal aberration
assay.
b. 	For other in vivo genotoxicity assays: Toxicity in the liver or tissue being
assessed (e.g., by histopathological evaluation or blood biochemistry toxicity
indicators).
ii. 	Exposure
a. 	Measurement of drug-related material either in blood or plasma. The bone
marrow is a well-perfused tissue, and levels of drug-related materials in blood or
plasma are generally similar to those observed in bone marrow. The liver is
expected to be exposed for drugs with systemic exposure regardless of the route
of administration.
b. 	Direct measurement of drug-related material in target tissue, or autoradiographic
assessment of tissue exposure.
If systemic exposure is similar to or lower than expected clinical exposure, alternative strategies
might be called for such as:
(i) Use of a different route of administration;
(ii) Use of a different species with higher exposure;
(iii)Use of a different tissue or assay (see section II.C.4 (2.3.4), Limitations to the
Use of Standard In Vivo Tests).
When adequate exposure cannot be achieved (e.g., with compounds showing very poor target
tissue availability), conventional in vivo genotoxicity tests have little value.
2. 	When In Vitro Genotoxicity Tests Are Negative (4.4.2)
12

Contains Nonbinding Recommendations
If in vitro tests do not show genotoxic potential, in vivo (systemic) exposure can be assessed by
any of the methods above, or can be assumed from the results of standard absorption,
distribution, metabolism, and excretion (ADME) studies in rodents done for other purposes.
E. 	Sampling Times for In Vivo Assays (4.5)
Selection of the sampling time in the in vivo MN (micronucleus), chromosomal aberration, and
UDS test should follow OECD (1997).
When micronucleus analysis is integrated into multiweek studies, sampling of blood or bone
marrow can be done the day after the final administration (see recommendation for additional
blood sampling time in section III.C.3 (3.3.3) above).
For other genotoxicity assays, sampling time should be selected as appropriate for the endpoint
measured; for example, DNA damage/strand break measurements are usually made a few (e.g., 2
to 6) hours after the last administration for the multiple daily administration. In the case of
single administration, two sampling times should be used: a few hours and 24 hours after the
treatment.
In principle, studies of any length can be considered appropriate, provided the top dose/exposure
is adequate.
F. 	Number of Animals Analyzed (4.6)
The number of animals analyzed is determined by current recommendations for the micronucleus
assay (OECD) or other genotoxicity assays and generally does not include all the animals treated
for a toxicology study. Animals used for genotoxicity analyses should be randomly selected
from the group used for the toxicology study.
G. 	Use of Male/Female Rodents in In Vivo Genotoxicity Tests (4.7)
If sex-specific drugs are to be tested, then the assay can be done in the appropriate sex. In vivo
tests with the acute protocol can generally be carried out in only one sex. For acute tests, both
sexes should be considered only if any existing toxicity, metabolism, or exposure (Cmax (peak
concentration) or AUC (area under the plasma concentration curve)) data indicate a
toxicologically meaningful sex difference in the species being used. Otherwise, the use of males
alone is considered appropriate for acute genotoxicity tests. When the genotoxicity test is
integrated into a repeat-dose toxicology study in two sexes, samples can be collected from both
sexes, but a single sex can be scored if there is no substantial sex difference evident in
toxicity/metabolism. The dose levels for the sex(es) scored should meet the criteria for
appropriate dose levels (sections IV.C.2 (4.3.2) and IV.C.3 (4.3.3)).
Similar principles can be applied for other established in vivo genotoxicity tests.
H. 	Route of Administration (4.8)
13

Contains Nonbinding Recommendations
The route of administration is generally the expected clinical route (e.g., oral, intravenous, or
subcutaneous) but can be modified if appropriate to obtain systemic exposure (e.g., for topically
applied compounds (see section II.C.4 (2.3.4)).
I. 	Use of Positive Controls for In Vivo Studies (4.9)
For in vivo studies, it is considered sufficient to treat animals with a positive control only
periodically, and not concurrently with every assay, after a laboratory has established
competence in the use of the assay (note 16).
V. 	GUIDANCE ON EVALUATION OF TEST RESULTS AND ON FOLLOW-UP
TEST STATEGIES (5)
Comparative trials have shown conclusively that each in vitro test system generates both false
negative and false positive results in relation to predicting rodent carcinogenicity. Genotoxicity
test batteries (of in vitro and in vivo tests) detect carcinogens that are thought to act primarily via
a mechanism involving direct genetic damage, such as the majority of known human
carcinogens. Therefore, these batteries are not expected to detect nongenotoxic carcinogens.
Experimental conditions, such as the limited capability of the in vitro metabolic activation
systems, can lead to false negative results in in vitro tests. The test battery approach is designed
to reduce the risk of false negative results for compounds with genotoxic potential. On the other
hand a positive result in any assay for genotoxicity does not always mean that the test compound
poses a genotoxic/carcinogenic hazard to humans.
Although positive in vitro data could indicate intrinsic genotoxic properties of a drug,
appropriate in vivo data determine the biological significance of these in vitro signals in most
cases. Also, because there are several indirect mechanisms of genotoxicity that operate only
above certain concentrations, it is possible to establish a safe level (threshold) for classes of
drugs with evidence for such mechanisms (see section V.B (5.2) below; see also Müller and
Kasper, 2000; Scott et al., 1991; Thybaud et al., 2007).
A. 	Assessment of Biological Relevance (5.1)
The recommendations below assume that the test has been conducted using appropriate
conditions such as spacing of doses and levels of toxicity.
Small increases in apparent genotoxicity in vitro or in vivo should first be assessed for
reproducibility and biological significance. Examples of results that are not considered
biologically meaningful include:
i. 	Small increases that are statistically significant compared with the negative or solvent
control values but are within the confidence intervals of the appropriate historical
control values for the testing facility
ii. 	Weak/equivocal responses that are not reproducible
14

Contains Nonbinding Recommendations
If either of the above conditions applies, the weight of evidence indicates a lack of genotoxic
potential, the test is considered negative or the findings not biologically relevant, and no further
testing is called for.
B. 	Evaluation of Results Obtained in In Vitro Tests (5.2)
In evaluating positive results, especially for the microbial mutagenicity test, the purity of the test
compound should be considered to determine whether the positive result could be attributable to
a contaminant.
1. 	Evaluation of Positive Results Obtained In Vitro in a Bacterial Mutation Assay
(5.2.1)
Because positive results in the Ames test are thought to indicate DNA reactivity, extensive
follow-up testing to assess the in vivo mutagenic and carcinogenic potential would be warranted
to assess the potential risk for treatment of patients, unless justified by appropriate risk-benefit
analysis.
There are some well-characterized examples of artifactual increases in colonies that are not truly
revertants. These increases can occur due to contamination with amino acids (i.e., providing
histidine for Salmonella typhimurium strains or tryptophan for Escherichia coli strains), so that
the bacterial reversion assay is not suitable for testing a peptide that is likely to degrade. Certain
cases exist where positive results in bacterial mutation assays might be shown not to indicate
genotoxic potential in vivo in humans, for example, when bacterial-specific metabolism occurs,
such as activation by bacterial nitroreductases.
2. 	Evaluation of Positive Results Obtained In Vitro in Mammalian Cell Assays
(5.2.1)
Recommendations for assessing weight of evidence and follow-up testing for positive
genotoxicity results are discussed in IWGT reports (e.g., Thybaud et al., 2007). In addition, the
scientific literature gives a number of conditions that can lead to a positive in vitro result of
questionable relevance. Therefore, any in vitro positive test result should be evaluated based on
an assessment of the weight of evidence as indicated below. This list is not exhaustive, but is
given as an aid to decision-making.
i. 	The conditions do not occur in vivo (pH; osmolality; precipitates).
(Note that the 1 mM limit avoids increases in osmolality, and that if the test
compound alters pH, it is advisable to adjust pH to the normal pH of untreated
cultures at the time of treatment).
ii. 	The effect occurs only at the most toxic concentrations.
-in the MLA increases, at ≥ 80 percent reduction in RTG
-for in vitro cytogenetic assays, when growth is suppressed by ≥ 50 percent
15

Contains Nonbinding Recommendations
If any of the above conditions apply, the weight of evidence indicates a lack of genotoxic
potential; the standard battery (Option 1) can be followed. Thus, a single in vivo test is
considered sufficient.
3. 	Evaluation of In Vitro Negative Results (5.2.3)
For in vitro negative results, further testing should be considered in special cases, such as the
following (the examples given are not exhaustive, but are given as an aid to decision-making):
the structure or known metabolism of the compound indicates that standard techniques for in
vitro metabolic activation (e.g., rodent liver S9) might be inadequate; the structure or known
activity of the compound indicates that the use of other test methods/systems might be
appropriate.
C. 	Evaluation of Results Obtained From In Vivo Tests (5.3)
In vivo tests have the advantage of taking into account absorption, distribution, and excretion,
which are not factors in in vitro tests, but are potentially relevant to human use. In addition,
metabolism is likely to be more relevant in vivo compared to the systems normally used in vitro.
If the in vivo and in vitro results do not agree, then the difference should be considered/explained
on a case-by-case basis (e.g., a difference in metabolism; rapid and efficient excretion of a
compound in vivo).
In vivo genotoxicity tests also have the potential to give misleading positive results that do not
indicate true genotoxicity. As examples:
(i) Increases in micronuclei can occur without administration of any genotoxic
agent, due to disturbance in erythropoiesis (Tweats et al., 2007, I).
(ii) DNA adduct data should be interpreted in the light of the known background
level of endogenous adducts.
(iii)Indirect, toxicity-related effects could influence the results of the DNA strand
break assays (e.g., alkaline elution and Comet assays).
Thus, it is important to take into account all the toxicological and hematological findings when
evaluating the genotoxicity data (note 15). Indirect effects related to toxicological changes could
have a safety margin and might not be clinically relevant.
D. 	Follow-up Strategies for Positive Results (5.4)
1. 	Follow-up to Findings In Vitro in Mammalian Cell Tests (5.4.1)
The following discussion assumes negative results in the Ames bacterial mutation assay.
i. 	Mechanistic/in vivo follow-up (5.4.1.1)
16

Contains Nonbinding Recommendations
When there is insufficient weight of evidence to indicate lack of relevance, recommended
follow-up for positive mammalian cell assays would be to provide experimental evidence, either
by additional in vitro studies (see item a below) or by carrying out two appropriate in vivo
assays (see item b below), as follows:
a. 	Mechanistic information that contributes to a weight of evidence for a lack of
relevant genotoxicity is often generated in vitro, for example evidence that a
test compound that induces chromosome aberrations or mutations in the MLA
is not a DNA damaging agent (e.g., other negative mutation/DNA damage
tests in addition to the Ames test; structural considerations), or evidence for an
indirect mechanism that might not be relevant in vivo or might have a
threshold (e.g., inhibition of DNA synthesis; reactive oxygen species
produced only at high concentrations) (Galloway et al., 1998; Scott et al.,
1991; Müller and Kasper, 2000). Similar studies can be used to follow up a
positive result in the in vitro micronucleus assay, or in this case, evidence can
include a known mechanism that indicates chromosome loss/aneuploidy, or
centromere staining experiments (note 17) that indicate chromosome loss.
Polyploidy is a common finding in chromosome aberration assays in vitro.
Although aneugens can induce polyploidy, polyploidy alone does not indicate
aneugenic potential and can simply indicate cell cycle perturbation; it is also
commonly associated with increasing cytotoxicity. If polyploidy, but no
structural chromosome breakage, is seen in an in vitro assay, generally a
negative in vivo micronucleus assay with assurance of appropriate exposure
would provide sufficient assurance of lack of potential for aneuploidy
induction.
If the above mechanistic information and weight of evidence supports the lack
of relevant genotoxicity, only a single in vivo test with appropriate evidence
of exposure is called for to establish the lack of genotoxic activity. This is
typically a cytogenetic assay, and the micronucleus assay in vivo is called for
when following up potential for chromosome loss.
If there is not sufficient weight of evidence or mechanistic information to rule
out relevant genotoxic potential, two in vivo tests are generally called for,
with appropriate endpoints and in appropriate tissues (usually two different
tissues), and with an emphasis on obtaining sufficient exposure in the in vivo
models.
Or
b. 	Two appropriate in vivo assays should be done, usually with different tissues,
and with supporting demonstration of exposure.
In summary, negative results in appropriate in vivo assays, with adequate justification for the
endpoints measured and demonstration of exposure (see section IV.D.1 (4.4.1)) are considered
sufficient to demonstrate absence of significant genotoxic risk.
17

Contains Nonbinding Recommendations
ii. 	Follow-up to an in vitro positive result that is dependent upon S9 activation
(5.4.1.2)
When positive results are seen only in the presence of the S9 activation system, it should first be
verified that metabolic activation is responsible and not some other difference in conditions (e.g.,
low or no serum in the S9 mix, compared with  10 percent serum in the non-activated
incubations). The follow-up strategy is then aimed at determining the relevance of the results in
vitro to conditions in vivo, and will generally focus on in vivo studies in liver (note 18).
2. 	Follow-up to a Positive In Vivo Micronucleus Assay (5.4.2)
If there is an increase in micronuclei in vivo, all the toxicological data should be evaluated to
determine whether a nongenotoxic effect could be the cause or a contributing factor (note 15). If
nonspecific effects of disturbed erythropoiesis or physiology (such as hypo/hyperthermia) are
suspected, an in vivo assay for chromosome aberrations might be more appropriate. If a real
increase is suspected, strategies should be used to demonstrate whether the increase is due to
chromosome loss or chromosome breakage (note 17). There is evidence that aneuploidy
induction (e.g., with spindle poisons) follows a nonlinear dose response. Thus, it might be
possible to determine that there is a threshold exposure below which chromosome loss is not
expected and to determine whether an appropriate safety margin exists compared with clinical
exposure.
In conclusion, the assessment of the genotoxic potential of a compound should take into account
the totality of the findings and acknowledge the intrinsic values and limitations of both in vitro
and in vivo tests.
E. 	Follow-up Genotoxicity Testing in Relation to Tumor Findings in a
Carcinogenicity Bioassay (5.5)
Additional genotoxicity testing in appropriate models can be conducted for compounds that were
negative in the standard test battery but which have shown increases in tumors in carcinogenicity
bioassay(s) with insufficient evidence to establish a nongenotoxic mechanism. To help
understand the mode of action, additional testing can include modified conditions for metabolic
activation in in vitro tests or can include in vivo tests measuring genetic damage in target organs
of tumor induction, such as DNA strand break assays (e.g., comet or alkaline elution assays),
liver UDS test, DNA covalent binding (e.g., by 32P-postlabeling), mutation induction in
transgenes, or molecular characterization of genetic changes in tumor-related genes (Kasper et
al., 2007).
18

Contains Nonbinding Recommendations
VI. 	NOTES (6)
Note 1. The in vitro micronucleus assay has been widely evaluated in international collaborative
studies (Kirsch-Volders et al., 2003), is validated by the European Center for the Validation of
Alternative Methods (ECVAM) (Corvi et al., 2008), and is the subject of an OECD guideline
487 (2010).
Note 2. There is a small but significant number of genotoxic carcinogens that are reliably
detected by the bone marrow tests for chromosomal damage but have yielded
negative/weak/conflicting results in the in vitro tests outlined in the standard battery options.
Carcinogens such as procarbazine, hydroquinone, urethane, and benzene fall into this category.
Some other examples from a survey of companies are described by Tweats et al., 2007, II.
Note 3. In principle, micronuclei in hematopoietic cells can be evaluated in bone marrow from
any species, and in blood from species that do not filter out circulating micronucleated
erythrocytes in the spleen. In laboratory mice, micronuclei can be measured in polychromatic
erythrocytes in blood, and mature (normochromatic) erythrocytes can be used when mice are
treated continuously for about 4 weeks or more. Although rats rapidly remove micronucleated
erythrocytes from the circulation, it has been established that micronucleus induction by a range
of clastogens and aneugens can be detected in rat blood reticulocytes (Wakata et al., 1998;
Hamada et al., 2001). Rat blood can be used for micronucleus analysis, provided methods are
used to ensure analysis of the newly formed reticulocytes (Hayashi et al., 2007; MacGregor et
al., 2006) and the sample size is sufficiently large to provide appropriate statistical sensitivity,
given the lower micronucleus levels in rat blood than in bone marrow (Kissling et al., 2007).
Whichever method is chosen, bone marrow or blood, automated or manual analysis, each
laboratory should determine the appropriate minimum sample size to ensure that scoring error is
maintained below the level of animal-to-animal variation.
Some experience is now available for micronucleus induction in the dog and rhesus monkey
(Harper et al., 2007; Hotchkiss et al., 2008). One example where such alternative species might
be useful would be in evaluation of a human metabolite that was not sufficiently represented in
rodents but was formed in the dog or monkey.
Note 4. Although the two options in the battery are equally suitable, specific knowledge about
an individual test compound can indicate that one option is preferable. For example, if systemic
exposure in animal models is equal to or less than anticipated clinical exposure, in vitro assays
should be employed: Option 1 (see also sections II.C.4 (2.3.4) and IV.D.1 (4.4.1)). On the other
hand, Option 2, including a test in liver, is recommended in cases where short-lived reactive
metabolites are expected to be generated in the liver.
Note 5. Certain structurally alerting molecular entities are recognized as being causally related
to the carcinogenic and/or mutagenic potential of chemicals. Examples of structural alerts
include alkylating electrophilic centers, unstable epoxides, aromatic amines, azo-structures, N
nitroso groups, and aromatic nitro-groups (Ashby and Paton, 1994). For some classes of
compounds with specific structural alerts, it is established that specific protocol
19

Contains Nonbinding Recommendations
modifications/additional tests are important for optimum detection of genotoxicity (e.g.,
molecules containing an azo-group, glycosides, compounds such as nitroimidazoles requiring
nitroreduction for activation, compounds such as phenacetin requiring a different rodent S9 for
metabolic activation).
Note 6. There is some experience with in vivo assays for micronucleus induction in skin and
colon (Hayashi et al., 2007), and DNA damage assays in these tissues can also be an appropriate
substitute.
Note 7. A few chemicals are more easily detectable either with plate-incorporation or with pre-
incubation methods, though differences are typically quantitative rather than qualitative
(Gatehouse et al., 1994). Experience in the pharmaceutical industry where drugs have been
tested in both protocols has not resulted in different results for the two methods, and, in the
IWGT report (Gatehouse et al., 1994), the examples of chemical classes listed as more easily
detectable in the pre-incubation protocol are generally not pharmaceuticals and are positive in in
vivo genotoxicity tests in liver. These include short chain aliphatic nitrosamines, divalent
metals, aldehydes (e.g., formaldehyde, crotonaldehyde), azo dyes (e.g., butter yellow),
pyrrolizidine alkaloids, allyl compounds (allyl isothiocyanate, allyl chloride), and nitro
(aromatic, aliphatic) compounds.
Note 8. The rationale for a maximum concentration of 1 mM for in vitro mammalian cell assays
includes the following: The test battery includes the Ames test and an in vivo assay. This
battery optimizes the detection of genotoxic carcinogens without relying on any individual assay
alone. 	There is a very low likelihood of compounds of concern (DNA damaging carcinogens)
that are not detected in Ames test or in vivo genotoxicity assay, but are detectable in an in vitro
mammalian assay only above 1 mM. Second, a limit of 1 mM maintains the element of hazard
identification, being higher than clinical exposures to known pharmaceuticals, including those
that concentrate in tissues (Goodman & Gilman, 2001), and is also higher than the levels
generally achievable in preclinical studies in vivo. Certain drugs are known to require quite high
clinical exposures for therapeutic effect, e.g., nucleoside analogs and some antibiotics. Although
comparison of potency with existing drugs can be of interest to sponsors, perhaps even above the
1 mM limit, it is ultimately the in vivo tests that determine relevance for human safety. For
pharmaceuticals with unusually low molecular weight (e.g., less than 200) higher test
concentrations should be considered.
Note 9. Although some genotoxic carcinogens are not detectable in in vitro genotoxicity assays
unless the concentrations tested induce some degree of cytotoxicity, DNA damaging agents are
generally detectable with only moderate levels of toxicity (Greenwood et al., 2004). As
cytotoxicity increases, mechanisms other than direct DNA damage by a compound or its
metabolites can lead to positive results that are related to cytotoxicity and not genotoxicity. Such
indirect induction of DNA damage secondary to damage to non-DNA targets is more likely to
occur above a certain concentration threshold. The disruption of cellular processes is not
expected to occur at lower, pharmacologically relevant concentrations.
In cytogenetic assays, even weak clastogens that are known to be carcinogens are positive
without exceeding a 50 percent reduction in cell counts. On the other hand, compounds that are
20

Contains Nonbinding Recommendations
not DNA damaging, mutagenic, or carcinogenic can induce chromosome breakage at toxic
concentrations. For both in vitro cytogenetic assays, the chromosome aberration assay and the in
vitro micronucleus assay, a limit of about 50 percent growth reduction is considered appropriate.
For cytogenetic assays in cell lines, measurement of cell population growth over time (by
measuring the change in cell number during culture relative to control, e.g., by the method
referred to as population doubling (PD (see note 10)), has been shown to be a useful measure of
cytotoxicity, as it is known that cell numbers can underestimate toxicity. For lymphocyte
cultures, an inhibition of proliferation not exceeding about 50 percent is considered sufficient;
this can be measured by mitotic index (MI) for metaphase aberration assays and by an index
based on cytokinesis block for in vitro micronucleus assays. In addition, for the in vitro
micronucleus assay, since micronuclei are scored in the interphase subsequent to a mitotic
division, it is important to verify that cells have progressed through the cell cycle. This can be
done by use of cytochalasin B to allow nuclear division but not cell division, so that micronuclei
can be scored in binucleate cells (the preferred method for lymphocytes). For cell lines, other
methods to demonstrate cell proliferation, including cell population growth over time (PD) as
described above, can be used (Kirsch-Volders et al., 2003).
For MLA, appropriate sensitivity is achieved by limiting the top concentration to one with close
to 20 percent Relative Total Growth (RTG) (10 to 20%) both for soft agar and for microwell
methods (Moore et al., 2002). Reviews of published data using the current criteria found very
few chemicals that were positive in MLA only at concentrations with less than 20 percent RTG
and that were rodent carcinogens, and convincing evidence of genotoxic carcinogenesis for this
category is lacking. The consensus is that caution is appropriate in interpreting results when
increases in mutation are seen only below 20 percent RTG, and a result would not be considered
positive if the increase in mutant fraction occurred only at  10 percent RTG.
In conclusion, caution is appropriate in interpreting positive results obtained as reduction in
growth/survival approaches or exceeds 50 percent for cytogenetics assays or 80 percent for
MLA. It is acknowledged that the evaluation of cells treated at these levels of cytotoxicity/clonal
survival can result in greater sensitivity but bears an increased risk of nonrelevant positive
results. The battery approach for genotoxicity is designed to ensure appropriate sensitivity
without relying on single in vitro mammalian cell tests at high cytotoxicity.
To obtain an appropriate toxicity range, a preliminary range-finding assay over a broad range of
concentrations is useful, but in the genotoxicity assay it is often critical to use multiple
concentrations that are spaced quite closely (less than two-fold dilutions). Extra concentrations
can be tested but not all concentrations need be evaluated for genotoxicity. It is not intended that
multiple experiments be carried out to reach exactly 50 percent reduction in growth, for example,
or exactly 80 percent reduction in RTG.
Note 10. For in vitro cytogenetic assays, it is appropriate to use a measure of relative cell growth
to assess toxicity because cell counts can underestimate toxicity (Greenwood et al., 2004). Using
calculated population doublings (see glossary) to estimate the 50 percent growth reduction level,
it was demonstrated that the frequency of positive results with compounds that are not mutagenic
21

Contains Nonbinding Recommendations
or carcinogenic is reduced, while agents that act via direct interaction with DNA are reliably
positive.
Note 11. In certain cases, it can be useful to examine chromosome aberrations at metaphase in
lymphocytes cultured from test animals after one or more administrations of test compound, just
as bone marrow metaphase cells can be used. Because circulating lymphocytes are not
replicating, agents that require replication for their genotoxic effect (e.g., some nucleoside
analogs) are not expected to be detected in this cell type. 	Because some lymphocytes are
relatively long-lived, in principle there is the potential for accumulation of unrepaired DNA
damage in vivo that would give rise to aberrations when the cells are stimulated to divide in
vitro. The in vivo lymphocyte assay can be useful in following up indications of clastogenicity,
but in general another tissue such as liver is a more informative supplement to the micronucleus
assay in hematopoietic cells because exposure to drug and metabolite(s) is often higher in liver.
Note 12. The inclusion of a second in vivo assay in the battery is to provide assurance of lack of
genotoxicity by use of a tissue that is well exposed to a drug and/or its metabolites; a small
number of carcinogens that are considered genotoxic gave positive results in a test in liver but
were negative in a cytogenetics assay in vivo in bone marrow. These examples likely reflect a
lack of appropriate metabolic activity or lack of reactive intermediates delivered to the
hematopoietic cells of the bone marrow.
Assays for DNA strand breaks, DNA adducts, and mutations in transgenes have the advantage
that they can be applied in many tissues. Internationally agreed protocols are not yet in place for
all the in vivo assays, although considerable experience and published data and protocol
recommendations exist for DNA strand break assays (Comet and alkaline elution assays), DNA
adduct (covalent binding) measurements, and transgenic rodent mutation assays, in addition to
the UDS assay. For a compound that is positive in vitro in the MLA and induces predominantly
large colonies, and is also shown not to induce chromosome breakage in an in vitro metaphase
assay, an in vivo assay for mutation, such as a transgenic mouse mutation assay, should be
considered in preference to a DNA strand break assay. The UDS assay is considered useful
mainly for compounds that induce bulky DNA adducts or are positive in the Ames test. Because
cytotoxicity induces DNA strand breakage, careful cytotoxicity assessment is needed to avoid
confounding the results of DNA strand break assays. This has been well-characterized for the in
vitro alkaline elution test (Storer et al., 1996) but not yet fully validated for the Comet assay. In
principle, the DNA strand break assays can be used in repeat-dose toxicology assays with
appropriate dose levels and sampling times.
Because liver of mature animals is not a highly mitotic tissue, often a non-cytogenetic endpoint
is used for the second assay; but when dividing hepatocytes are present, such as after partial
hepatectomy, or in young rats (Hayashi et al., 2007), micronucleus analysis in liver is possible,
and detects known genotoxic compounds.
Note 13. For common vehicles like aqueous methyl cellulose, this would usually be appropriate,
but for vehicles such as Tween 80, the volume that can be administered could be as much as 30
fold lower than that given acutely.
22

Contains Nonbinding Recommendations
Note 14. Caution is appropriate if the toxicological study design includes additional blood
sampling, e.g., for measurement of exposure. Such bleeding could perturb the results of
micronucleus analysis since erythropoiesis stimulated by bleeding can lead to increases in
micronucleated erythrocytes.
Note 15. Increases in micronuclei can occur without administration of any genotoxic agent, due
to disturbance in erythropoiesis (such as regenerative anemia; extramedullary hematopoiesis),
stress, and hypo- and hyperthermia (reviewed by Tweats et al., 2007, I). In blood, changes in
spleen function that affect clearance of micronucleated cells from the blood could lead to small
increases in circulating micronucleated red blood cells.
Note 16. Positive controls for either short-term or repeat-dose genotoxicity studies:
For micronucleus (and other cytogenetic) assays, the purpose of the positive control is to verify
that the individuals scoring the slides can reliably detect increases in micronuclei. This can be
accomplished by use of samples from periodic studies (every few months) of small groups of
animals (one sex) given acute treatment with a positive control. For manual scoring, such slides
can be included in coded slides scored from each study. Positive control slides should not be
obvious to readers based on their staining properties or micronucleus frequency. For automated
scoring, appropriate quality control samples should be used with each assay.
For other in vivo genotoxicity assays, the purpose of positive controls is to demonstrate reliable
detection of an increase in DNA damage/mutagenicity using the assay in the chosen species,
tissue, and protocol. After a laboratory has demonstrated that it can consistently detect
appropriate positive control compounds in multiple independent experiments, carrying out
positive control experiments periodically is generally sufficient provided experimental
conditions are not changed. However, currently it is considered that for the Comet assay,
concurrent positive controls are advisable.
Note 17. Determination of whether micronucleus induction is due primarily to chromosome loss
or to chromosome breakage could include staining micronuclei in vitro or in vivo to determine
whether centromeres are present, e.g., using fluorescent in situ hybridization (FISH) with probes
for DNA sequences in the centromeric region, or a labeled antibody to kinetochore proteins. If
the majority of induced micronuclei are centromere positive, this suggests chromosome loss.
(Note that even potent tubule poisons like colchicine and vinblastine do not produce 100%
kinetochore positive micronuclei, but more typically 70 to 80 percent, and are accepted as
primarily aneugens for assessing risk). An alternative approach is to carry out an in vitro or in
vivo assay for metaphase structural aberrations; if negative, this would imply that micronucleus
induction is related to chromosome loss.
Note 18. Standard induced S9 mix has higher activation capacity than human S9, and lacks
phase two detoxification capability unless specific cofactors are supplied. Also, nonspecific
activation can occur in vitro with high test substrate concentrations (see Kirkland et al., 2007).
Genotoxicity testing with human S9 or other human-relevant activation systems can be helpful.
Analysis of the metabolite profile in the genotoxicity test incubations for comparison with
known metabolite profiles in preclinical species (in uninduced microsomes or hepatocytes, or in
vivo) or in preparations from humans can also help determine the relevance of test results (Ku et
23

Contains Nonbinding Recommendations
al., 2007), and follow-up studies will usually focus on in vivo testing in liver. A compound that
gives positive results in vitro with S9 might not induce genotoxicity in vivo because the
metabolite is not formed, is formed in very small quantities, or is metabolically detoxified or
rapidly excreted, indicating a lack of risk in vivo.
24

Contains Nonbinding Recommendations
VII. 	GLOSSARY (7)
Alkaline elution assay: See DNA strand break assay.
Aneuploidy: Numerical deviation of the modal number of chromosomes in a cell or organism.
Base substitution: The substitution of one or more base(s) for another in the nucleotide
sequence. This can lead to an altered protein.
Cell proliferation: The ability of cells to divide and to form daughter cells.
Centromere/kinetochore: Structures in chromosomes essential for association of sister
chromatids and for attachment of spindle fibers that move daughter chromosomes to the poles
and ensure inclusion in daughter nuclei.
Clastogen: An agent that produces structural breakage of chromosomes, usually detectable by
light microscopy.
Cloning efficiency: The efficiency of single cells to form clones. It is usually measured after
seeding low numbers of cells in a suitable environment.
Comet assay: See DNA strand break assay.
Culture confluency: A quantification of the cell density in a culture by visual inspection.
Cytogenetic evaluation: Chromosome structure analysis in mitosis or meiosis by light
microscopy or micronucleus analysis.
DNA adduct: Product of covalent binding of a chemical to DNA.
DNA repair: Reconstitution of the original DNA sequence after DNA damage.
DNA strand breaks: Single or double strand scissions in the DNA.
DNA strand break assay: Alkaline treatment that converts certain types of DNA lesions into
strand breaks that can be detected by the alkaline elution technique, measuring migration rate
through a filter, or by the single cell gel electrophoresis or Comet test (in which cells embedded
in a thin layer of gel on a microscope slide are subjected to electric current, causing shorter
pieces of DNA to migrate out of the nucleus into a Comet tail). The extent of DNA migration is
measured visually under the microscope on stained cells.
Frameshift mutation: A mutation (change in the genetic code) in which one base or two
adjacent bases are added to (inserted in) or deleted from the nucleotide sequence of a gene. This
can lead to an altered or truncated protein.
25

Contains Nonbinding Recommendations
Gene mutation: A detectable permanent change within a single gene or its regulating sequences.
The changes can be point mutations, insertions, or deletions.
Genetic endpoint: The precise type or class of genetic change investigated (e.g., gene
mutations, chromosomal aberrations, DNA strand breaks, DNA repair, DNA adduct formation).
Genotoxicity: A broad term that refers to any deleterious change in the genetic material
regardless of the mechanism by which the change is induced.
Micronucleus: Particle in a cell that contains nuclear DNA; it might contain a whole
chromosome(s) or a broken centric or acentric part(s) of a chromosome(s).
Mitotic index: Percentage of cells in the different stages of mitosis amongst the cells not in
mitosis (interphase) in a preparation (slide).
Numerical chromosome changes: Chromosome numbers different from the original haploid or
diploid set of chromosomes; for cell lines, chromosome numbers different from the modal
chromosome set.
Plasmid: Genetic element in addition to the normal bacterial genome. A plasmid might be
inserted into the host chromosome or form an extra-chromosomal element.
Point mutations: Changes in the genetic codes, usually confined to a single DNA base pair.
Polychromatic erythrocyte: An immature erythrocyte in an intermediate stage of development
that still contains ribosomes and, as such, can be distinguished from mature normochromatic
erythrocytes (lacking ribosomes) by stains selective for RNA.
Polyploidy: Numerical deviation of the modal number of chromosomes in a cell, with
approximately whole multiples of the haploid number. Endoreduplication is a morphological
form of polyploidy in which chromosome pairs are associated at metaphase as
diplochromosomes.
Population doubling or culture growth: This can be calculated in different ways; one example
of an appropriate formula is: Population doublings (PDs) = the log of the ratio of the final count
(N) to the starting (baseline) count (Xo), divided by the log of 2. 	That is: PD = [log(N  Xo)] 
log 2.
Recombination: Breakage and balanced or unbalanced rejoining of DNA.
RTG (relative total growth): This measure of cytotoxicity takes the relative suspension growth
(based on cell loss and cell growth from the beginning of treatment to the second day post
treatment) and multiplies it by the relative plating efficiency at the time of cloning for mutant
quantization.
Single cell gel electrophoresis assay: Comet assay. 	See DNA strand break assay.
26

Contains Nonbinding Recommendations
Survival (in the context of mutagenicity testing): Proportion of living cells among dead cells,
usually determined by staining or colony counting methods after a certain treatment interval.
Transgene: An exogenous or foreign gene inserted into the host genome, either into somatic
cells or germ line cells.
Unscheduled DNA synthesis (UDS): DNA synthesis that occurs at some stage in the cell cycle
other than S-phase in response to DNA damage. It is usually associated with DNA excision
repair.
27

Contains Nonbinding Recommendations
VIII. 	REFERENCES (8)
Ashby, J., D. Paton, “The influence of chemical structure on the extent and sites of
carcinogenesis for 522 rodent carcinogens and 55 different human carcinogen exposures,”
Mutation Research, 286:3-74, 1994.
Corvi, R., S. Albertini, T. Hartung, S. Hoffmann, D. Maurici, S. Pfuhler, J. van Benthem, P.
Vanparys, “ECVAM retrospective validation of the in vitro micronucleus test (MNT),”
Mutagenesis, 23:271-283, 2008.
Galloway, S.M., J.E. Miller, M.J. Armstrong, C.L. Bean, T.R. Skopek, W.W. Nichols, “DNA
synthesis inhibition as an indirect mechanism of chromosome aberrations: Comparison of DNA-
reactive and non-DNA-reactive clastogens,” Mutation Research, 400:169-186, 1998.
Gatehouse, D., S. Haworth, T. Cebula, E. Gocke, L. Kier, T. Matsushima, C. Melcion, T. Nohmi,
T. Ohta, S. Venitt, E. Zeiger, “Report from the working group on bacterial mutation assays:
International Workshop on Standardisation of Genotoxicity Test Procedures,” Mutation
Research, 312:217-233, 1994.
Goodman & Gilman’s The Pharmacological Basis of Therapeutics. J.G. Hardman, L.E. Limbird,
A. G. Gilman (Eds.), McGraw-Hill Professional, New York; 10th edition (August 13, 2001).
Greenwood, S.K., R.B. Hill, J.T. Sun, M.J. Armstrong, T.E. Johnson, J.P. Gara, S.M. Galloway,
“Population doubling: A simple and more accurate estimation of cell growth suppression in the
in vitro assay for chromosomal aberrations that reduces irrelevant positive results,”
Environmental and Molecular Mutagenesis, 43:36-44, 2004.
Hamada S., S. Sutou, T. Morita, A. Wakata, S. Asanami, S. Hosoya et al., “Evaluation of the
rodent micronucleus assay by a 28-day treatment protocol: summary of the 13th collaborative
study by the collaborative study group for the micronucleus test (CSGMT)/Environmental
Mutagen Society of Japan (JEMS)–Mammalian Mutagenicity Study Group (MMS),”
Environmental and Molecular Mutagenesis, 37:93-110, 2001.
Harper, S.B, S.D. Dertiner, M.E. Bishop, A.M. Lynch, M. Lorenzo, M. Saylor, J.T. MacGregor.
Flow cytometric analysis of micronuclei in peripheral recticulocytes. III. An efficient method of
monitoring chromosomal damage in the beagle dog. Toxicological Sciences, 100:406-414, 2007.
Hartmann A., E. Agurell, C. Beevers, S. Brendler-Schwaab, B. Burlinson, P. Clay, A. Collins, A.
Smith, G. Speit, V. Thybaud, R.R. Tice, “Recommendations for conducting the in vivo alkaline
Comet assay,” Mutagenesis, 18:45-51, 2003.
Hayashi, M., J.T. MacGregor, D.G. Gatehouse, I. Adler, D.H. Blakey, S.D. Dertinger, G.
Krishna, T. Morita, A. Russo, S. Sutou, “In vivo rodent erythrocyte micronucleus assay. II. Some
aspects of protocol design including repeated treatments, integration with toxicity testing, and
automated scoring,” Environmental and Molecular Mutagenesis, 35:234-252, 2000.
28

Contains Nonbinding Recommendations
Hayashi, M., J.T. MacGregor, D.G. Gatehouse, D.H. Blakey, S.D. Dertinger, L. Abramsson-
Zetterberg, G. Krishna, T. Morita, A. Russo, N. Asano, H. Suzuki, W. Ohyama, D. Gibson, “In
vivo erythrocyte micronucleus assay. III. Validation and regulatory acceptance of automated
scoring and the use of rat peripheral blood reticulocytes, with discussion of non-hematopoietic
target cells and a single dose-level limit test,” Mutation Research, 627:10-30, 2007.
Heddle, J.A., S. Dean, T. Nohmi, M. Boerrigter, D. Casciano, G.R. Douglas, B.W. Glickman,
N.J. Gorelick, J.C. Mirsalis, H-J Martus, T.R. Skopek, V. Thybaud, K.R. Tindall, N. Yajima, “In
vivo transgenic mutation assays,” Environmental and Molecular Mutagenesis, 35:253-259, 2000.
Hotchkiss, C.E. , M.E. Bishop, S.D. Dertinger, W. Slikker, M.M. Moore, J.T. MacGregor, “Flow
cytometric analysis of micronuclei in peripheral blood reticulocytes IV: An index of
chromosomal damage in the rhesus monkey (Macaca mulatta),” Toxicological Sciences,
102:352-358, 2008.
Kasper, P., Y. Uno, R. Mauthe, N. Asano, G. Douglas, E. Matthews, M. Moore, L. Müller, M.
Nakajima, T. Singer, G. Speit, “Follow-up testing of rodent carcinogens not positive in the
standard genotoxicity testing battery: IWGT workgroup report,” Mutation Research, 627:106
116, 2007.
Kenelly, J.C., R. Waters, J. Ashby, P.A. Lefevre, B. Burlinson, D.J. Benford, S.W. Dean, I deG.
Mitchell, “In vivo rat liver UDS assay.” In Supplementary Mutagenicity Tests, UKEMS
Recommended Procedures, D.J. Kirkland and M. Fox (Eds.), Cambridge University Press, pp.
52-77, 1993.
Kirkland, D.J., S. Pfuhler, D. Tweats, M. Aardema, R. Corvi, F. Darroudi, A. Elhajouji, H. Glatt,
P. Hastwell, M. Hayashi, P. Kasper, S. Kirchner, A. Lynch, D. Marzin, D. Maurici, J-R.
Meunier, L. Müller, G. Nohynek, J. Parry, E. Parry, V. Thybaud, R. Tice, J. van Benthem, P.
Vanparys, P. White, “How to reduce false positive results when undertaking in vitro genotoxicity
testing and thus avoid unnecessary follow-up animal tests: Report of the ECVAM workshop,”
Mutation Research, 628:31-55, 2007.
.
Kirsch-Volders, M., T. Sofuni, M. Aardema, S. Albertini, D. Eastmond, M. Fenech, M. Ishidate,
S. Kirchner, E. Lorge, T. Morita, H. Norppa, J. Surralles, A. Vanhauwaert, A. Wakata, “Report
from the in vitro micronucleus assay working group,” Mutation Research, 540:153-163, 2003.
Kissling, GE., S.D. Dertinger, M. Hayashi, J.T. MacGregor., “Sensitivity of the erythrocyte
micronucleus assay: Dependence on number of cells scored and inter-animal variability,”
Mutation Research, 634:235-240, 2007.
Ku, W.W., A. Bigger, G. Brambilla, H. Glatt, E. Gocke, P.J. Guzzie, A. Hakura, M. Honma, H-J.
Martus, R.S. Obach, S. Roberts, “Strategy for genotoxicity testing—Metabolic considerations,”
Mutation Research, 627:59-77, 2007.
MacGregor J.T., M.E. Bishop, J.P. McNamee, M. Hayashi, N. Asano, A. Wakata, M. Makajima,
J. Saito, A. Aidoo, M.M. Moore, S.D. Dertinger, “Flow cytometric analysis of micronuclei in
29

Contains Nonbinding Recommendations
peripheral blood reticulocytes: II. An efficient method of monitoring chromosomal damage in
the rat,” Toxicological Sciences, 94:92-107, 2006.
Moore, M.M. , M. Honma, J. Clements, K. Harrington-Brock, T. Awogi, G. Bolcsfoldi, M.
Cifone, D. Collard, M. Fellows, K. Flanders, B. Gollapudi, P. Jenkinson,P. Kirby, S. Kirchner, J.
Kraycer, S. McEnaney, W. Muster, B. Myhr, M. O’Donovan, J. Oliver, M-C. Ouldelhkim, K.
Pant, R. Preston, C. Riach, R. San, H. Shimada, L.F. Stankowski, Jr. “Mouse Lymphoma
Thymidine Kinase Locus Gene Mutation Assay: Follow-Up International Workshop on
Genotoxicity Test Procedures, New Orleans, Louisiana, April 2000,” Environmental Molecular
Mutagenesis, 40:292-299, 2002.
Moore, M.M., M. Honma, J. Clements, G. Bolcsfoldi, B. Burlinson, M. Cifone, J. Clarke, R.
Delongchamp, R. Durward, M. Fellows, B. Gollapudi, S. Hou, P. Jenkinson, M. Lloyd, J.
Majeska, B. Myhr, M. O'Donovan, T. Omori, C. Riach, R. San, L. F. Stankowski, Jr., A.K.
Thakur, F. Van Goethem, S. Wakuri, I. Yoshimura, “Mouse Lymphoma Thymidine Kinase Gene
Mutation Assay: Follow-up Meeting of the International Workshop on Genotoxicity Testing–
Aberdeen, Scotland, 2003–Assay Acceptance Criteria, Positive Controls, and Data Evaluation,”
Environmental and Molecular Mutagenesis, 47:1-5, 2006.
Müller, L, P. Kasper, “Human biological relevance and the use of threshold arguments in
regulatory genotoxicity assessment: Experience with pharmaceuticals.” Mutation Research, 464:
9-34, 2000.
OECD Guidelines for Genetic Toxicology (1997), www.oecd.org/dataoecd.
Scott, D., S.M. Galloway, R.R. Marshall, M. Ishidate, Jr., D. Brusick, J. Ashby, B.C. Myhr,
“Genotoxicity under extreme culture conditions. A report from ICPEMC Task Group 9,”
Mutation Research, 257:147-204, 1991.
Storer, R.D., T.W. McKelvey, A.R. Kraynak, M.C. Elia, J.E. Barnum, L.S. Harmon, W.W.
Nichols, J.G. DeLuca, “Revalidation of the in vitro alkaline elution/rat hepatocyte assay for
DNA damage: Improved criteria for assessment of cytotoxicity and genotoxicity and results for
81 compounds,” Mutation Research, 368:59-101, 1996.
Suzuki, H., N. Ikeda, K. Kobayashi, Y. Terashima, Y. Shimada, T. Suzuki, T. Hagiwara, S.
Hatakeyama, K. Nagaoka, J. Yoshida, Y. Saito, J. Tanaka, M. Hayashi, “Evaluation of liver and
peripheral blood micronucleus assays with 9 chemicals using young rats. A study by the
Collaborative Study Group for the Micronucleus Test (CSGMT)/Japanese Environmental
Mutagen Society (JEMS) - Mammalian Mutagenicity Study Group (MMS),” Mutation Research,
583:133-145, 2005.
Thybaud, V., M. Aardema, J. Clements, K. Dearfield, S. Galloway, M. Hayashi, D. Jacobson-
Kram, D. Kirkland, J. T. MacGregor, D. Marzin, W. Ohyama, M. Schuler, H. Suzuki, E. Zeiger,
“Strategy for genotoxicity testing: Hazard identification and risk assessment in relation to in
vitro testing,” Mutation Research, 627:41-58, 2007.
30

Contains Nonbinding Recommendations
Tweats, D. J., D. Blakey, R.H. Heflich, A. Jacobs, S.D. Jacobsen, T. Morita, T. Nohmi, M.R.
O'Donovan, Y.F. Sasaki, T. Sofuni, R. Tice, “Report of the IWGT working group on strategies
and interpretation of regulatory in vivo tests. I. Increases in micronucleated bone marrow cells in
rodents that do not indicate genotoxic hazards,” Mutation Research, 627:78-91, 2007, I.
Tweats, D.J., D. Blakey, R.H. Heflich, A. Jacobs, S.D. Jacobsen, T. Morita, T. Nohmi, M.R.
O'Donovan, Y.F. Sasaki, T. Sofuni, R. Tice, “Report of the IWGT working group on
strategy/interpretation of regulatory in vivo tests. II. Identification of in vivo-only positive
compounds in the bone marrow micronucleus test,” Mutation Research, 627:92-105, 2007, II.
Wakata, A., Y. Miyamae, S. Sato, T. Suzuki, T. Morita, N. Asano, T. Awogi, K. Kondo, M.
Hayashi, “Evaluation of the rat micronucleus test with bone marrow and peripheral blood:
Summary of the 9th Collaborative Study by CSGMT/JEMS • MMS,” Environmental and
Molecular Mutagenesis, 32:84-100, 1998.
31