Use of a claims-based active drug safety surveillance system to assess the risk of acute pancreatitis with exenatide or sitagliptin compared to metformin or glyburide

Objective: To estimate risk and relative risk (RR) of acute pancreatitis among patients using incretin-based diabetes therapies (exenatide or sitagliptin) compared to patients treated with agents with established safety profiles (metformin or glyburide). Research design and methods: The study population was derived from a large US commercial health insurance transaction database using an active drug safety surveillance system (i3 Aperio*). This analysis is based on data from June 2005 through June 2008. Cohorts of exenatide and sitagliptin initiators were each matched to an equal number of metformin or glyburide (met/gly) initiators using propensity scores to reduce confounding in the comparison of outcomes during follow-up. Patients with claims suggesting pancreatic disease in the 6 months prior to cohort entry were excluded. Main outcome measure: Claims for hospitalizations associated with a primary diagnosis of acute pancreatitis (ICD-9 577.0). Results: There were 27 996 exenatide initiators and 16 276 sitagliptin initiators and approximately equal numbers of matched comparators. During follow-up of up to 1 year, acute pancreatitis occurred among 0.13% of patients treated with exenatide and 0.12% of patients treated with sitagliptin. The risk of acute pancreatitis was comparable for initiators of exenatide (RR 1.0; 95% confidence interval (CI) 0.6-1.7) and sitagliptin (RR 1.0; 95% CI 0.5-2.0) relative to the comparison cohorts. Conclusions: These data do not provide evidence for an association of acute pancreatitis among initiators of exenatide or sitagliptin compared to met/gly initiators. These results are limited by the data available in an administrative, healthcare database.