A critique of methods to measure cytotoxicity in mammalian cell genotoxicity assays.

Various methods have been used to estimate cytotoxicity in mammalian cell genotoxicity assays since their introduction more than four decades ago, and although there is no agreement on whether any single method is optimal, there is now a better appreciation of their limitations. Methods based on aspects of cellular function are inevitably inaccurate unless some estimate of cell number is included, and those using some measure of cell proliferation give different results depending on the mathematical model used. Although it would be desirable, it is not possible to provide a universal measure of cytotoxicity because the phenomenon is so complex. There is some flexibility in the limits of cytotoxicity proposed in regulatory guidelines, and it can be argued these could be even less precise. Also, to make valid comparisons of the performance of different test systems, novel or established, it would seem essential to use similar measures of cytotoxicity.