Introduction: The purpose of this study is to evaluate whether cigarette smoking is associated with the CpG island methylator phenotype (CIMP) and/or V600E BRAF mutations in colon cancer. Cigarette smoking has been associated with microsatellite instability (MSI) in sporadic colon cancer in previous studies. Most microsatellite unstable colon cancers exhibit CIMP and harbor V600E BRAF mutations, as do a minority of microsatellite stable tumors. We hypothesized that the association of smoking with instability could be through induction of CIMP and/or BRAF mutations. If this were true, then one might expect to see relationships between cigarette smoking and CIMP and/or BRAF mutations in microsatellite stable tumors. Experimental procedures: A population-based sample of 898 individuals with colon cancer and 2397 controls had been previously evaluated for smoking history, CIMP, MSI and the BRAF V600E mutation. Smoking was defined as number of cigarettes per day: 0, 1-20, and >20. CIMP was defined as methylation of two or more of the following CpG islands: MINT1, MINT2, MINT31, p16, and hMLH1. Results: Smoking was associated with CIMP in microsatellite stable tumors as compared to controls without colon cancer, and a dose response relationship was seen: odds ratios for 1-20 and >20 cigarettes per day as compared to 0 per day were 1.25 (95% C.I. 0.89, 1.75) and 1.88 (95% C.I. 1.21, 2.91), respectively (p trend < 0.01). A comparison of CIMP and non-CIMP tumors also showed that smoking was associated with CIMP in stable cancers. Again, a dose response relationship was seen: odds ratios for 1-20 and >20 cigarettes per day were 1.33 (95% C.I. 0.92, 1.93) and 1.63 (95% C.I. 1.01, 2.64) respectively, (p trend 0.04). Smoking was also associated with the BRAF V600E mutation in microsatellite stable colon cancers as compared to controls, and a dose response relationship was seen: odds ratios for 1-20 and >20 cigarettes per day were 2.04 (95% C.I. 0.94, 4.41), and 3.37 (95% C.I. 1.32, 8.65), respectively (p trend 0.01). A comparison of BRAF mutated with BRAF wild-type tumors showed a nearly significant trend in the association of cigarette smoking with the BRAF V600E mutation in stable colon cancers: odds ratios for 1-20 and >20 cigarettes per day were 2.11 (95% C.I. 0.97, 4.59) and 2.47 (0.93, 6.51), respectively (p trend 0.05). A polytomous analysis showed that the association of cigarette smoking with colon cancer was limited to the minority of tumors (stable or unstable) which showed CIMP and/or BRAF mutations. Conclusions: The association of cigarette smoking with CIMP and the BRAF V600E mutation in microsatellite stable colon cancer implies a more general carcinogenic mechanism than that suggested by the previously identified association with microsatellite instability. Our results also suggest that previously identified associations between smoking and colon cancer are largely explained by this CIMP/BRAF association.
[Proc Amer Assoc Cancer Res, Volume 47, 2006]