Cancer incidence among patients treated with antidiabetic pharmacotherapy

SummaryBackgroundThe objective of this study was to quantify and compare the risks of colorectal, bladder, liver, pancreatic and melanoma cancers in users and non-users of various antidiabetic pharmacotherapies, including thiazolidinediones.MethodsWe conducted a retrospective cohort study among patients with type 2 diabetes using a large US population-based database. Pharmacy claims data were used to classify patients according to their use of antidiabetic pharmacotherapies. Using medical record review and claims-based adjudication, we identified newly diagnosed cases of colorectal, bladder, liver, pancreatic and melanoma cancers. We calculated cancer incidence rates in patients receiving antidiabetic pharmacotherapies and fit Poisson regression models to obtain rate ratios (RRs) and 95% confidence intervals (CIs) to compare the rate of each cancer among drug exposure categories. Rate ratios were adjusted for age, gender, and selected cancer risk factors.ResultsOverall, 813 confirmed study cancers occurred among 191,223 patients with diabetes during follow-up. Compared with non-use of pharmacotherapy, ever use of oral antidiabetic pharmacotherapies and insulin was generally not associated with an increased rate of colorectal, bladder, liver, pancreatic or melanoma cancer. Specifically, patients exposed to thiazolidinediones (as monotherapy or in combination) were not at greater risk of cancer than patients treated with other oral antidiabetics (colorectal: adjusted RR = 1.11, 95%CI = 0.86–1.43; bladder: adjusted RR = 1.05, 95%CI = 0.71–1.54; liver: adjusted RR = 0.76, 95%CI = 0.32–1.83; pancreas: adjusted RR = 0.67, 95%CI = 0.40–1.12; melanoma: adjusted RR = 0.67, 95%CI = 0.41–1.09).ConclusionsIn general, use of antidiabetic pharmacotherapies, including thiazolidinediones, was not associated with an increased rate of colorectal, bladder, liver, pancreatic or melanoma cancer.