Epidemiological studies of the relationship between carcinogenicity and DNA damage

The new methods for studying DNA damage in vivo offer cancer epidemiologists both a new study endpoint and a refinement of exposure measurement. Up to the present, they have not been applied in studies that satisfy standard epidemiological criteria. In particular, more attention must be paid to the selection of study samples and to ensuring adequate numbers of study subjects. There are substantial problems in studying the link between DNA damage and carcinogenesis at the individual level, including the long latency and the multifactorial etiology of most cancers. Acute leukaemia induced by alkylating agent therapy represents one of the few human situations in which such studies are feasible, since patients are exposed at carefully measured and recorded doses, the risk is high and specific, and follow-up is good. A considerable saving in resources could be achieved by storing samples from a large number of treated patients and carrying out analyses of DNA damage only in those from people who develop leukaemia and from suitably chosen matched controls.