Objectives: Cigarette smoking has been linked to somatic genetic and epigenetic aberrations, including CpG island methylator phenotype (CIMP)-high, microsatellite instability (MSI)-high and BRAF mutation. These molecular features have been associated with synchronous primary colorectal cancers (CRCs). Thus, we examined the hypothesis that smoking might be associated with the risk of synchronous CRCs.
Methods: Within the Health Professionals Follow-up Study and Nurses' Health Study, we examined the relationship of smoking and incidence of CRC according to tumor synchronicity, using duplication-method Cox proportional hazards regression analysis.
Results: We confirmed 1,981 solitary CRC and 45 synchronous CRC cases during follow-up of 134,305 individuals. CRC risk associated with smoking differed significantly by tumor synchronicity status (Pheterogeneity<0.001). When comparing current smokers with never smokers, multivariable hazard ratios (HR) were 5.27 (95% confidence interval (CI), 2.08-13.40) for synchronous CRCs and 0.97 (95% CI, 0.83-1.14) for solitary CRC. Similarly, differential associations were observed when examining cumulative pack-years smoked (Pheterogeneity=0.006). Smoking cessation for ≥10 years relative to current smoking might reduce the risk of synchronous CRCs (multivariable HR=0.42; 95% CI, 0.19-0.95), but not solitary CRC (multivariable HR=1.10; 95% CI, 0.94-1.29; Pheterogeneity=0.001). Comparing current and former smokers with never smokers, multivariable HRs for synchronous CRCs were significantly higher than those of solitary CRC positive for either CIMP-high, MSI-high, or BRAF mutation (Pheterogeneity=0.002).
Conclusions: Smoking is associated with an elevated risk of synchronous CRCs. Our data support a model where smoking contributes to an etiologic field effect that favors these somatic molecular alterations and the development of multiple primary tumors.