Mechanistic and genetic studies of radiation tumorigenesis in the mouse--implications for low dose risk estimation.

Radiation cancer risk estimates remain firmly based upon epidemiological data. Experimental validation of the fundamental aspects of these risk estimates relies on animal studies. In particular, animal model systems for radiation carcinogenesis can provide data for mechanistic modelling approaches to risk estimation. The accuracy and validity of risk estimation models developed will depend upon the extent of our understanding of the process of radiation carcinogenesis. The study of 'spontaneous' tumours in humans continues to provide a sound context in which to consider the mechanisms of radiation carcinogenesis. Several mouse radiation carcinogenesis systems are considered here with particular reference to the nature of the initiating event and the influence of genetic susceptibility on radiation-induced cancer.