Chemoprevention of second cancers.

BACKGROUND "Second cancers" can be thought of in two general categories: (a) those occurring as a consequence of cancer treatment and (b) primary cancers that are thought to develop largely as a consequence of prior lifestyle habits (e.g., chronic smoking, drinking, sun exposures), genetic susceptibility, or interactions of the two. Because there has been limited work on chemoprevention of treatment-related secondary cancers, this minireview will focus on chemoprevention of second cancers with lifestyle/genetic origins. METHODS/RESULTS Trials aimed at preventing second cancers in patients with tobacco-related cancers (head and neck, lung), skin cancers, breast cancer, and colorectal adenomatous polyps have been completed with some success. However, one finding that has emerged is that, across several cancer sites, subgroups are found with differential response to the chemopreventive agent. For example, smoking status, alcohol consumption, nutritional status, and host tumor characteristics seem to modify chemopreventive efficacy. Stratum-specific (subgroup) findings may occur by chance, requiring a need for supportive evidence from observational epidemiologic studies of the agent (where available), mechanistic studies, or results of other related trials. CONCLUSIONS Although chemoprevention of second cancers has been realized, it has become increasingly apparent that not all benefit equally. The finding of subgroup effects in completed trials results in the need to consider such subgroup effects in the design of future trials, by either restricting enrollment to particular subgroups (e.g., never or former smokers), or by increasing sample size requirements to allow for variation in response in subgroups in a statistically powerful way.