EURL ECVAM Genotoxicity and Carcinogenicity Consolidated Database of Ames Positive Chemicals

Positive results in the Ames test correlate well with carcinogenic potential, at least in rodents. However, the correlation is not perfect because mutagenesis is not carcinogenesis, and mutations are only one of many stages in tumour development. Also, situations can be envisaged where the mutagenic response may be specific to the bacteria or the test protocol, for example bacterial-specific metabolism, exceeding a detoxification threshold or the induction of oxidative damage to which bacteria may be more sensitive than mammalian cells in vitro or tissues in vivo. Since most chemicals are also tested for genotoxicity in mammalian cells, the pattern of mammalian cell results (positive, negative) may help identify whether Ames-positive results predict carcinogenic or in vivo mutagenic activity. A workshop was therefore organised and sponsored by the EU Reference Laboratory for Alternatives to Animal Testing (EURL ECVAM) to investigate this further. Participants presented data on results from other genotoxicity tests with Ames-positive compounds. Data came from published, regulatory agency and industry sources. The question was posed whether negative results in mammalian cell tests were often associated with absence of carcinogenic or in vivo genotoxic activity. In the limited time available, the presented data were combined and an initial analysis suggested that the association of negative in vitro mammalian cell test results with lack of in vivo genotoxic or carcinogenic activity could have some significance. Possible reasons why a positive Ames test may not be associated with in vivo activity and what additional investigations/tests might contribute to a more robust evaluation were discussed. Because a considerable overlap was identified among the different databases presented, it was recommended that a consolidated database should be built, with overlapping chemicals removed, so that a more robust analysis of the predictive capacity for potential carcinogenic and in vivo genotoxic activity could be derived from the patterns of mammalian cell test results obtained for Ames-positive compounds.

A Workshop sponsored by EURL ECVAM was held in Ispra, Italy in 2013 to address the question of whether the in vitro mammalian cell genotoxicity test results could complement and mitigate the implications of a positive Ames test response for the prediction of in vivo genotoxicity and carcinogenicity, and if patterns of results could be identified. Databases of Ames-positive chemicals that were tested for in vivo genotoxicity and/or carcinogenicity were collected from different sources and analysed individually (Kirkland et al., xxxx, this issue). Because there were overlaps and inconsistent test results among chemicals in the different databases, a combined database which eliminated the overlaps and evaluated the inconsistencies would be preferable for addressing the above question. A database of >700 Ames-positive chemicals also tested in vivo was compiled, and the results in in vitro mammalian cell tests were analysed. Because the database was limited to Ames-positive chemicals, the majority (>85%) of carcinogens and in vivo genotoxins were positive when tested in both in vitro gene mutation and genotoxicity/clastogenicity tests. However, about half (>45%) of chemicals that were not carcinogenic or genotoxic in vivo also gave the same patterns of mammalian cell results. Although the different frequencies were statistically significant, positive results in 2 in vitro mammalian cell tests did not, per se, add to the predictivity of the positive Ames test. By contrast, negative results in both in vitro mammalian cell tests were rare for Ames-positive carcinogens and in vivo genotoxins but, were significantly more frequent for Ames-positive chemicals that are not carcinogenic or genotoxic in vivo. Thus, in the case of an Ames-positive chemical, negative results in 2 in vitro mammalian cell tests covering both mutation and clastogenicity/aneugenicity endpoints should be considered as indicative of absence of in vivo genotoxic or carcinogenic potential.