Prediction of a New Smoking-Dependent Risk Score for Postsurgical Gastric Cancer-Specific Mortality: the Fiesta Study

We did post hoc analysis of the Fujian prospective investigation of cancer (FIESTA) database, aiming to identify smoking-dependent risk factors that differentially associated with gastric cancer-specific mortality, and further construct a smoking-dependent scoring system for risk assessment. Patients with gastric cancer received radical gastrectomy between 2000 and 2010, with the follow-up ending in 2015. 2779 patients had data on smoking status, including 2223 smokers and 556 never-smokers. Risk estimates were expressed as hazard ratio (HR) with 95% confidence interval (CI). The median follow-up time was 45.55 months. Smoking was not associated with postsurgical survival differences, whereas risk profiles differed between smokers and never-smokers. Particularly, the presence of the metabolic syndrome was significantly associated with increased gastric cancer-specific mortality in smokers (HR (95% CI), p: 2.73 (1.53-4.89), <0.001), but not in never-smokers. Based on significant factors of either smokers or never-smokers, we built a risk score, with 1 unit increment associating with 10% reduced risk of gastric cancer-specific mortality (HR (95% CI), p: 0.90 (0.88-0.91), <0.001). Our findings indicate there is a diverse risk profile between smokers and never-smokers and an additive interaction between smoking and the metabolic syndrome, and importantly we developed a scoring system with a strong predictive capability.