Aims/hypothesis: Incretin-based therapies have been suggested to increase the risk of pancreatitis, but the results of the available studies are controversial. Because results from prospective trials are limited by low statistical power, and because retrospective studies are often subject to bias, a pooled analysis of phase III clinical trials and two endpoint trials was performed.
Methods: Event numbers for acute pancreatitis and patient-years of exposure (PYOs) were obtained from representatives of the pharmaceutical companies, or by literature research. Data were pooled for glucagon-like peptide 1 (GLP-1) receptor agonists and dipeptidyl-peptidase 4 (DPP-4) inhibitors in comparison with their respective controls and expressed as exposure-adjusted incidence rates.
Results: A total of 38 cases of pancreatitis were reported in clinical trials with GLP-1 receptor agonists, comprising 17,775 PYOs. With the comparator treatment, nine events occurred in 5,863 PYOs. The pooled event rates were 2.1 and 1.5 per 1,000 PYOs, respectively, resulting in an OR of 1.39 (95% CI 0.67, 2.88). With DPP-4 inhibitors, 57 events were reported in 45,132 PYOs compared with 46 events in 38,883 PYOs with the comparator treatment. Pooled event rates were 1.3 and 1.2 per 1,000 PYOs, respectively, resulting in an OR of 1.07 (CI 0.72, 1.58).
Conclusions/interpretation: This analysis suggests a trend towards a slightly elevated risk of pancreatitis with GLP-1 receptor agonists. With DPP-4 inhibitors, no consistent trend was found. However, the incidence numbers of cases of pancreatitis were still very small, and the statistical power was limited. Future endpoint trials may help to provide a better estimate of the true risk of pancreatitis with incretin-based therapies.