Reproducibility of carcinogenic potencies estimated in different rodent species. Methylene chloride: a useful example

The aim of this study was to investigate the reproducibility of carcinogenic potencies estimated by multistage model (and its linearized form) fitting to dose-response relationships referring to different rodent species and strains, different sexes, target organs and severity of prognosis. Methylene chloride represented a useful example since there were 11 single dose-response relationships available for the same administration route (inhalation) and for the above-mentioned variables. The multistage mathematical model and the 'linearized multistage' model (EPA's conservative procedure), which are the mainly adopted models for regulatory purposes, were used for the dose-response fitting of single trends and other different cumulative trends. The reproducibility of response frequencies at equal exposure doses was also analyzed. The mathematical function best describing each trend showed the constant presence of a linear component (for low doses) and, in the majority of cases, a quadratic second order component. Carcinogenic potencies were fairly consistent for the various parameters. A good interspecies reproducibility of carcinogenic potencies was observed. A high intraspecies reproducibility was also observed, even though obtained from different sexes, as well as varying degrees of severity of prognosis and, in the case of rats, different strains. The fluctuation of the estimated carcinogenic potencies was lower in mice than in rats.