•Indirectly, by mechanistic association with micronucleus formation. fWith protocol modifications.

•Indirectly, by mechanistic association with micronucleus formation. fWith protocol modifications.

rium TA102, which detects such mutations within multiple copies of hisG genes or E. coli WP2 uvrA, which detects these mutations in the trpE gene, or the same E. coli strain possessing the plasmid pKM101. Because radiomimetics are selective for A-T sites, these strains detect most radiomimetics. Damage induced by cross-linking agents may be detected using repair proficient strains. As cross-linking agents are usually good clastogens, and a test for clastogenic events is part of the standard battery of pharmaceuticals, there is no need to include specific strains to detect such agents in the BRM.

2. Reasoning for the Mouse Lymphoma tk Assay

The necessity for a gene mutation test with mammalian cells in a standard battery for genotoxicity testing of pharmaceuticals was a major issue of discussion in the ICH EWG. The protocol for mammalian cell gene mutation assays as an aggregate was discussed by Aaron et al (7), and the respective OECD guideline was updated (6). However, among the assays described in this one guideline, the MLA has a unique place because of its apparent ability to detect several important types of genetic damage, the full range of which may not be detectable in other standard assays (8-11). Furthermore, there are unique aspects of the MLA, and the various methods by which it can be conducted, that warrant a separate OECD MLA guideline. It is noteworthy that certain types of damage detected by the MLA are important in oncogenes and tumor suppressor genes (Table 1).

In comparison, the HPRT (hypoxanthine guanine phosphoribosyl trans-ferase) assays using various cell lines detect only a limited spectrum of such genetic changes (Table 1). From Table 1, it appears also that it may be sufficient for the in vitro component of the standard genotoxicity test battery to consist of the BRM plus a single assay with mammalian cells, that is, either the in vitro CA (A chromosomal aberration test) or the MLA. This is in agreement with regulatory experience (12). It is accepted that the MLA detects a wide range of genetic effects and that the genetic damage manifest is expressed in living cells. However, for the use in a standard battery, ICH guidances specify that an in vitro CA and the MLA can be used interchangeably. In addition to the considerations given in Table 1, this conclusion was supported by the following studies:

1. A reinvestigation of nine previously designated mouse lymphoma tk uniquely positive rodent carcinogens reported by the NTP revealed that a majority of these compounds are clastogenic in vitro (structural aberrations or polyp-loidy) when tested according to updated protocol principles for the chromosomal aberration test (13).

2. Fifty-nine percent (20/34) of in vitro clastogens were detected in the MLA through the use of the standard 3 to 4 hours of treatment with test compound in the presence and absence of an exogenous metabolic activation system (14, 15).

3. Among the 41% (14/34) of compounds that were not detected, it was noted that they often induced chromosomal changes (chromosomal aberrations, polyploidy, or both) only when they were not removed from the treatment medium after the more routinely used short 3- to 6-hour treatment period (16, 17). Out of 15 such chemicals, including nucleoside analogues and spindle poisons, 11 compounds produced positive results in the MLA with a continuous treatment period of 24 hours in the absence of metabolic activation using the microwell methodology. These data led to the adoption of this latter treatment modification in the ICH S2B guidance.

Utilization of the ICH-recommended modified protocols for the MLA and the in vitro CA have thus provided essentially the same level of safety information within the ICH standard battery. Further protocol and data interpretation guidance can be found in the communications of two MLA workshops that were held under ICH auspicies (18, 19). Regarding the protocol detailed for the MLA in the ICH S2B guidance, there is a continuous effort to address the following issues (5): (a) the stipulation for a continuous treatment period of 24 hours may affect the specificity of the MLA for nongenotoxic compounds; (b) because the majority of the data included in the EWG analysis of the MLA were provided for the microwell version, it was not clear whether the same conclusions could be drawn for the agar plate version. The recommendation for the use of the microwell method that appeared in the ICH S2B guidance document, which was based on limited comparative data made available to the EWG at the time of guidance finalization, may need to be re-evaluated through an ICH maintenance process. Because it is presumed that both methods give valid results, either methodology, microwell or agar plate, should be acceptable.

Other mammalian cell gene mutation assays, such as those that monitor mutation at the HPRT locus, in addition to or as an alternative to the MLA, are advocated under ICH guidance in situations in which the BRM is not considered appropriate. Such situations can occur, for example, if the test compound is an antibacterial agent or interacts with the mammalian cell replication system.

3. Reasoning for the In Vitro Chromosomal Aberration Test (In Vitro CA)

A cytogenetic evaluation of chromosomal damage in vitro has traditionally been part of the screening for genotoxicity in all of the ICH regions. Its inclusion in the newly designed ICH standard battery was put into perspective when comparing its usefulness with the MLA (see above). The protocol for the in vitro CA has been redesigned in detail (20), and the recommendations in ICH guidances are in agreement with the recent OECD guideline update (6). However, two protocol aspects are still a matter of discussion: (a) the levels and method(s) of quantification of test compound-induced cytotoxicity for a valid test, and (b) the necessity to test compounds into the insoluble concentration range. The current recommendations, including those in ICH guidances and OECD guidelines in both areas, are thought by some to account for some positives of questionable relevance (12, 21, 22). Aid for interpretation of such results is given in the ICH S2A "Guidance on Specific Aspects of Regulatory Genotoxicity Tests." Further discussion regarding this topic can be found below in "Regulatory Interpretation of Genotox-icity Test Results." Alternatives to the in vitro chromosomal aberration assay that may be capable of providing comparable information, for example, the in vitro micronucleus test, are being evaluated (23-27).

4. Reasoning for the In Vivo Test Using Rodent Hematopoietic Cells

The ICH guidance S2B recommends the inclusion of an in vivo test with rodent hematopoietic cells in the standard battery. Historically, hematopoietic cells have been used for in vivo analysis of chromosomal damage, and a large database of study outcomes has been accumulated. In vivo tests such as the bone marrow mi-cronucleus test can be relatively insensitive (see below, "'Site of Contact' Evaluation of Genotoxicity," more than 80% of human carcinogens are positive in this assay (28). Compared to in vitro tests, in vivo tests such as the bone marrow mi-cronucleus test appear to respond only to a subset of genotoxins, that is, those in vitro genotoxins that can reach the target cells in this tissue at sufficient concentrations to induce chromosomal damage. Nevertheless, the ICH guidance S2A reinforces the importance of an in vivo test component in the standard battery because of such additional factors as absorption, distribution, metabolism, and excretion, which are only partially manifest in in vitro tests but are relevant to human use. This reasoning, together with the fact that in vivo clastogenesis is highly predictive of carcinogenesis (28-30), argues strongly for the inclusion of such an in vivo assay in the core test battery. For the standard battery, a test using rodent hematopoietic cells satisfies this requisite because the bone marrow is a well-perfused tissue, and levels of drug-related materials in bone marrow are often similar to those observed in blood or plasma. This was shown by direct comparisons of drug levels in the two compartments for a large series of different pharmaceuticals (31). The ICH guidances do not give any preference for the evaluation of micronuclei over the analysis of chromosomal aberrations using he-matopoietic cells. However, it is acknowledged that relative to the analysis of chromosomal aberrations in bone marrow cells, the analysis of micronuclei in erythrocytes is more expedient and convenient and that the latter is sensitive to the effects of some aneugenic compounds. Additionally, the presence of micro-nucleated cells can be assessed in the peripheral blood of mice (32, 33) and rats (34), which offers additional approaches for monitoring induction of micronu-clei. Should the sensitivity and specificity of the peripheral blood micronucleus test prove to be comparable to that of the bone marrow micronucleus test, using the former could provide some advantages over the latter. Some of these include (a) ease of sample collection, (b) the ability to examine multiple samples harvested from a given treated animal, (c) the ability to determine the attainment of steady-state levels of micronucleated cell induction (35), thereby optimizing for the time of bone marrow harvesting, and (d) avoiding the need to sacrifice treated animals unless bone marrow confirmation of the peripheral blood response is necessary.

One promising application of the in vivo micronucleus assay and a potentially significant advantage is its integration into the general toxicology screen (30, 35-39). The endpoint is compatible with longer term (e.g., 14-, 28-, and 90-day) general toxicology testing (which uses repeated exposures over the duration of the study) and can be monitored over the course of such studies. The demonstration of attaining steady-state levels of micronucleus induction (35) in bone marrow after repeated treatments at lower doses that are comparable to the peak levels achieved after acute exposure to higher doses supports the practicability of such an approach. In addition, the use of peripheral blood (37) makes it feasible to monitor for the temporal manifestation of micronuclei during the course of longer term toxicology studies performed in rodents. Some advantages offered by the concomitant assessment of such genotoxic effects over the course of general toxicology studies include (a) examining endpoints such as micronucleus induction in animals undergoing repeated exposure, (b) precluding the need to conduct an independent test for chromosome damage, and (c) limiting the unnecessary use of additional animals.

5. Other Test Models in Genetic Toxicology

Regulatory genotoxicity testing is focused on widely used, sufficiently standardized, and well-characterized test methods. Such test methods constitute the standard battery of genotoxicity tests for pharmaceuticals according to ICH guidances. However, genetic toxicology offers a wide variety of test methods in addition to those named for the standard battery of tests. Because most of these methods are not very widely used and not sufficiently standardized and characterized, such tests may be used to give supplementary information but, under normal circumstances, are not used to replace the standard methods. Specific circumstances in which additional tests may yield valuable information are described throughout the text and in the notes to the ICH guidances. One area in which such tests are important is the follow-up evaluation of in vitro positive results in in vivo test approaches in addition to the test for chromosomal damage in rodent bone marrow.

From time to time, consideration is given in the scientific community to implementing new test procedures that necessarily involve an appraisal of their status of use, degree of standardization, and potential for regulatory acceptance. One such effort is the International Workshop on Genotoxicity Test Procedures (IWGTP; Washington, DC/US, March 1999). In this workshop, the following four test methods were reviewed and discussed: (1) the in vitro micronucleus test; (2) the Comet assay, either in vitro or in vivo; (3) in vivo transgenic mutation assays; and (4) test systems to detect photochemical genotoxicity. It is understood that these methods either provide more efficient use of resources for genetic toxicology testing, for example, the in vitro micronucleus test, or could fill information gaps in regulatory decision making in many cases.

B. Regulatory Interpretation of Genotoxicity Test Results

Experience with testing, data collection for regulatory submission, and regulatory review largely influenced the recommendations given in both ICH genotoxicity guidances. In this regard, ICH guidances should facilitate appropriate interpretation of genotoxicity test results. Some issues of regulatory interpretation are discussed below, on the basis of experience with submissions to the Federal Institute for Drugs and Medical Devices (BfArM) of Germany.

1. Experience with Submissions for New Pharmaceuticals

Table 2 gives an overview of the number of tests that have been submitted in standard battery genotoxicity test categories for 335 new pharmaceutical entities within the 8 years from 1990-1997. "New" in this context means "active ingredient not previously available on the German pharmaceutical market." In the period analyzed, mutagenicity testing of pharmaceuticals was performed in accordance with the EU guideline (40) still valid at the time, encompassing four tests from the following categories:

• Test for gene mutation in bacteria

Table 2. Results of Standard Battery Genotoxicity Tests for New Pharmaceuticals


No. of compounds tested

No. of compounds with positive test results

% Compounds with positive test results

Bacterial reverse mutation

0 0

Post a comment