Interactions between detoxifying enzyme polymorphisms and susceptibility to cancer.

Many case-control studies on the role of detoxifying enzyme polymorphisms in susceptibility to cancer have identified significant associations, though few have identified effects with sufficient strength to be useful clinically. Odds ratios of 2-3 are the usual finding. Therefore, combinations of risk alleles are increasingly studied in the hope of identifying haplotypes with sufficient biological impact (odds ratio > 15) to warrant further study in a clinical setting. The study of interactions between different detoxifying enzyme loci should be based on biological sense, for example the classical view of two-step xenobiotic detoxification, overlapping substrate specificities, detoxification of molecules derived from the same pathological insult (e.g. detoxification of nitrosamines and polycyclic aromatic hydrocarbons in cigarette smoke) or simultaneous regulation of expression. The rationale behind these mechanisms is discussed. Considerable amounts of data have focused on the interaction between CYP1A1 and GSTM1, particularly in Japanese patients with lung cancer. In this regard there is accumulating evidence suggesting that the combination of GSTM1 null/CYP1A1 rare alleles, particularly in combination with smoking, confers highly significant increased risk. However, there is now some debate on the importance of these data in terms of chemical detoxificatlon, since certain studies suggest that the CYP1A1 polymorphisms that have been investigated do not influence function or expression of the gene. Indeed, the influence of CYP2D6 genetic variation is also disputed, suggesting that these polymorphisms may be acting more as linkage markers than by influencing chemical carcinogenesis themselves. What is clear from the many studies on interactions between detoxifying enzyme polymorphisms is the comparative lack of supporting data on synergism between these genes. Given the inevitable increase in the complexity of studies, a basic explanation of the statistical approaches to the assessment of interactions is included in this chapter. A more detailed statistical/epidemiological assessment is given elsewhere in the book. Since there is an increasing reluctance of journals to publish case-control studies describing the (usually moderate) influence of detoxifying enzyme polymorphisms, it is important that scientists understand what statistical approach is appropriate and address the issue from a novel and clinically significant angle. This is likely to involve multiple genes and subgroup analysis of some kind.