Dietary heterocyclic amines and microsatellite instability in colon adenocarcinomas

Microsatellite instability (MSI) is now an accepted and important pathway in colon tumorigenesis, occurring in 10-15% of sporadic colon cancers and almost all hereditary nonpolyposis colon cancers. Little is known about possible environmental influences on MSI status in colon cancer. We conducted an epidemiological study of 276 colon cancer cases in Los Angeles County that was designed to determine the population prevalence of MSI+ colon cancer and to identify environmental influences in the development of MSI+ tumors. Our results support the cigarette smoking and MSI+ association recently reported [Slattery,M.L., Curtin,K., Anderson,K. et al. (2000) J. Natl Cancer Inst., 92, 1831-1836]. Risk of MSI+ colon cancers increased with increasing dose (number of cigarettes per day) and increasing duration (years of smoking) of smoking. Compared with never-smokers, those who smoked 1-20 pack-years and > 20 pack-years showed odds ratios of 1.39 and 1.64, respectively (P for trend = 0.03). In addition, our results show, for the first time, that after adjustment for cigarette smoking habits and red meat intake, patients with MSI+ colon cancers had significantly higher dietary exposure to heterocyclic aromatic amines (HAA) as determined by two surrogates of high dietary HAA exposure: preference for well-done red meat and high frequencies of certain cooking practices (frying, barbecuing, broiling and using meat drippings). The risk of MSI+ colon cancer was increased 3-fold (smoking/red meat intake adjusted OR = 3.03, 95% CI = 1.06, 8.63) among patients who preferred to eat red meat that was very well-done and was increased >2-fold (smoking/red meat adjusted OR = 2.33, 95% CI = 1.00, 5.25) among those who frequently fried/barbecued/broiled their meats or used meat drippings. The risk of MSI+ colon cancer associated with cigarette smoking remained statistically significant after adjustment for high dietary HAA exposure. The significant association between cigarette smoking and dietary HAA with a specific subset of colon cancers may explain, at least in part, inconsistencies in results linking these two exposures to colon cancers. These results provide a potential mechanism of linking HAA exposure and cigarette smoking to a specific subset of colon cancers.