Masterstudiengang "Drug Regulatory Affairs"


Strategies for Carcinogenicity Testing of Human Pharmaceuticals ***

Dr. Anja Mecklenburg (Abschlußjahr: 2003)

Detecting the carcinogenic potential, i.e. the potential to induce tumor development, is a major issue in the preclinical development of pharmaceuticals, since such tests cannot be performed in humans. Tumors develop due to multiple sequential genetic alterations, and carcinogens can induce tumor development via genotoxic or non-genotoxic mechanisms. The differentation of these different carcinogenic mechanisms is important with regard to the detecion of a carcinogenic potential. For the detection of genotoxic substances, a standard battery of in vivo and in vitro tests exist. However, several compounds produce equivocal results in these assays, which does not necessarily imply that they are carcinogenic. For compounds with non-genotoxic mechanisms of carcinogenicity, no such standard battery exists. Traditionally, authorities based their risk assessment on two long-term carcinogenicity studies in two rodent species, usually in rats and mice. However, multiple surveys have revealed that generally little additional knowledge is gained from the long-term mouse bioassay. In order to improve the strategy for carcinogenicity assessment, new alternative models have been developed, employing genetically modified mice. These transgenic models are supposed to respond to carcinogenic substances within a relatively short period of time, without revealing the geriatric pathology that is usually seen in the 2-year rodent bioassay. In 1996, the ILSI-HESI "Alternatives to Carcinogenicity Testing (ACT) Initiative" was started as a collaborative effort by governmental, academic and industrial institutions, in order to evaluate some of the proposed alternative models for carcinogenicity testing. Results from this initiative provide evidence that these alternative models are useful for assessing the carcinogenic potential of a compound, when interpreted in combination with results from a 2-year bioassay. This is currently reflected by the ICH guideline S1B, which allows conducting a short- or medium-term study in a transgenic mouse model instead of the second long-term study. However, the ACT initiative also demonstrates that there are specific limitations of the alternative carcinogenicity models, and regulatory authorities consider these limitations when evaluating results from alternative studies in the risk assessment of pharmaceuticals. Therefore, integration of the new models into the strategy of carcinogenicity testing should be based on rational decisions, which have to be presented to the authority when applying for a marketing authorization. Authorities from Europe, the USA and Japan consider the p53+/- and the Tg-rasH2 model useful in providing data for regulatory purposes, although the FDA considers the p53+/- model acceptable only for clearly or equivocally genotoxic agents. The Tg.AC model is considered helpful in evaluating products intended for dermal application.

Own experiences gained from carcinogenicity studies in the p53+/- mouse with compounds that had shown some borderline results in the genotoxicity battery are presented. These studies demonstrate that indeed the transgenic models are able to provide additional mechanistic information, which finally allows an improved evaluation of the carcinogenic potential of a compound.

Although both pharmaceutical industry and regulatory authorities have meanwhile gained some experience with alternative models for carcinogenicity testing, the interpretation of results obtained from these studies is still difficult and represents the major limitation of these models, but comparable limitations have been seen with the long-term assays as well. However, more experience will lead to a greater acceptance and a greater use of these alternative assays.

The alternative transgenic mouse models can help to improve risk assessment for humans. These assays are at least as good as the mouse long-term assay, and probably present additional value because they allow some reflection on mechanisms of tumor development. Therefore the transgenic mouse models are adequate as contributor for the "weight of evidence". However, further data are needed. Against the background of 40 years of experience with the traditional 2-year rodent bioassay, the limited experience with the alternative mouse models do not immediately result in a clear recommendation for replacement of the 2-year assays. The present knowledge shall stimulate toxicologists in industry and regulatory agencies to move forward to more experience, as with a greater number of alternative assays, the evaluation and understanding progresses.

Pages: 45