Relationships between in vitro genetic toxicity and carcinogenicity studies in animals

In vitro genetic toxicity assays currently in use cannot be used to unambiguously define all potential carcinogens. The relationship between the various in vitro endpoints (mutation, cytogenetic effects, transformation) and the patterns of tumorigenicity is quite complex. Additional data, particularly for noncarcinogens, are needed to better understand the inconsistencies and to define the limits of predictability. For example, we must empirically define the concordance and discordance between in vivo and in vitro systems related to chemical classes or structures. At this time, since there are no nontrivial generalizations and no general predictivity is possible, we must use judgment in applying short-term test results. There is a clear need to identify as soon as possible the minimum number of in vitro systems that can be used, nonredundantly, to discriminate between carcinogens and noncarcinogens. However, we must continue to address the sources of discordance in order to understand the ways in which short-term test results can and cannot be used. Many trans-sex/trans-species carcinogens show the capacity to induce multiple endpoints of genetic toxicity and this class of rodent carcinogens may have the greatest implications for potential human health effects. A much more selective use of short-term test results to identify potential carcinogens such as these will achieve many of the goals for which these tests were developed.