Diabetes, metformin use, and survival in esophageal cancer: a population-based cohort study.

Abstract

It is unclear how diabetes and metformin use is associated with survival of esophageal cancer. This population-based cohort study included new cases of esophageal cancer reported in Sweden from 2006 to 2018 with follow-up through 2019. Diabetes status and metformin use were analyzed in relation to all-cause and disease-specific mortality using multivariable Cox regression. The hazard ratios (HRs) with 95% confidence intervals (CIs) were adjusted for age, sex, calendar year, obesity, comorbidity, and use of nonsteroidal anti-inflammatory drugs or statins. For comparison reasons, 3 other antidiabetic medications were also analyzed (ie, sulfonylureas, insulin, and thiazolidinedione). Among 4851 esophageal cancer patients (8404 person-years), 4072 (84%) died during follow-up. Compared with esophageal cancer patients with diabetes but not using metformin, decreased all-cause mortality was indicated among nondiabetic patients (without metformin) (HR = 0.86, 95% CI = 0.77 to 0.96) and diabetic patients who used metformin (HR = 0.86, 95% CI = 0.75 to 1.00). The hazard ratios of all-cause mortality decreased with a higher daily dose of metformin (Ptrend = .04). The corresponding hazard ratios for disease-specific mortality were similar but slightly attenuated. The results were also similar in separate analyses of esophageal cancer patients with adenocarcinoma or squamous cell carcinoma, with tumor stage I-II or III-IV, and in those who had or had not undergone surgery. No associations with mortality outcomes were found for use of sulfonylureas, insulin, or thiazolidinedione. Diabetes was associated with an increased all-cause mortality, whereas metformin use was associated with decreased all-cause mortality among esophageal cancer patients. More research is needed to determine if metformin affects survival in esophageal cancer.