N-nitrosamine (NA) impurities in pharmaceuticals represent "cohort of concern" compounds under ICH M7(R2), due to their mutagenic/carcinogenic potential, involving cytochrome P450 (CYP)-mediated metabolic activation. Increasing interest in mammalian cell-based genotoxicity/mutagenicity assays prompted our assessment of the in vitro alkaline comet assay regarding its predictive power for NAs. Here, precision-cut liver slices (PCLiS), primary human hepatocytes (PHH), primary rat hepatocytes (PRH), and HepG2 cells with rat or hamster S9-mix were investigated as in vitro model systems. Metabolic competence was characterized beforehand. For performance evaluation, a panel of known-mutagenic N-nitroso-dimethylamine (NDMA), N-nitroso-diethanolamine, N-nitroso-methylaniline, S-N-nitroso-nornicotine, N-methyl-N-nitroso-2-propanamine and reported non-mutagenic (methyl-t-butylnitrosamine, N-nitrosoproline) was tested, together with Nitrosamine Drug Substance-Related Impurities N-nitrosodesloratadine, N-nitrosofolic acid, N-nitrosofluoxetine (NFluo) at a concentration range of 0.005-10 mM. After 2 h (PCLiS, PHH and PRH) or 4 h (HepG2), NDMA concentration-dependently induced DNA strand breaks in all in vitro models. Sensitivity/specificity of the various liver cell models for prediction of carcinogenic NAs were 100%/50% (HepG2 with hamster S9-mix), 50%/100% (PHH, PRH), and 50%/50% (HepG2 with rat S9-mix), respectively. Benchmark dose modeling indicated a higher relative in vitro comet assay response for NFluo compared to NDMA in all cell systems. In conclusion, the in vitro comet assay represents a sensitive and/or specific tool for complementing regulatory in vitro tests in prediction of mutagenic NAs. However, further optimization work is needed, using expanded training sets of compounds and thorough validation of liver cell models, before the in vitro comet assay could be incorporated in the standard battery for genotoxicity testing.
Djuari et al. (Tue,) studied this question.