Post-authorisation Safety Study of Pioglitazone Use and Safety Endpoints of Interest in Denmark After Direct Healthcare Professional Communication

INTRODUCTION: A Direct Healthcare Professional Communication (DHPC) sent in Denmark on 11 August 2011 provided information on new pioglitazone labelling and guidance on monitoring treatment effectiveness. We describe pioglitazone use in Denmark after the DHPC, estimate the incidence of heart failure (HF), quantify pioglitazone cessation following a diagnosis of bladder cancer (BC) or uninvestigated macroscopic haematuria, and describe glycated haemoglobin (HbA1c) values. METHODS: This was a cohort study. From Danish population-based registries, cohorts of type 2 diabetes mellitus incident or prevalent users of pioglitazone or insulin in 2011-2015 were created. Patient characteristics, treatment patterns, laboratory results (available for a regional subset of the population), and incidence rates of HF and BC were estimated. RESULTS: There were 80 pioglitazone and 17,699 insulin incident users, 140 pioglitazone and 13,183 insulin prevalent users. There were no new BC cases among incident pioglitazone users, and < 5 new BC cases among prevalent pioglitazone users. Pioglitazone was rarely the first-line treatment. History of haematuria was documented in < 5 incident and 11 prevalent pioglitazone users. During follow-up, there were < 5 HF cases among 77 incident pioglitazone users and < 5 among 133 prevalent pioglitazone users without a history of HF. Median HbA1c at index date was 7.8% and 8.8% in incident pioglitazone and insulin cohorts, and 7.5% and 7.6% in prevalent pioglitazone and insulin cohorts, respectively. During follow-up of up to 4.4 years, 28.8% incident and 20.7% prevalent pioglitazone users discontinued pioglitazone. CONCLUSIONS: Numbers of pioglitazone users in Denmark were low and decreased over time. Risks of BC or HF were low and risk estimates imprecise.