Results of an international round robin study with the Ames fluctuation test

The German mirror committee of ISO TC147 SC 5 WG 9 ‘Genotoxicity’ organised an international round robin study on the Ames fluctuation test (ISO DIS 11350). Nineteen laboratories from seven countries participated in the study. The laboratories had to test four coded water samples. Sample one, a river water sample, was used as a negative control water sample and for spiking with strain-specific mixtures of mutagens (samples two and three). Sample four was an effluent from an industrial waste water plant. Validity criteria were fulfilled in 86 to 100 % of all test conditions (bacterial strain / }; S9-mix). The overall test sensitivity was 100 %. The test specificity ranged from 80 to 100 % depending on the test condition. For statistical evaluation an arcsinsquare- root transformation was applied to generate a metric data set. An ANOVA was performed to identify samples with a significantly higher numbers of revertants. A threshold value derived from a large set of negative controls was used in a final step to exclude results of individual tests with unexpected low numbers of spontaneous revertants. The lowest ineffective dilution values (LIDs) were then determined by this new approach. The variability of the LID values among the laboratories amounted in most cases to no more than two dilution steps around the median. The data produced in this round robin study proved a sufficient reproducibility of the test procedure and provide validity data for the finalization of the ISO standardization process.