[Skip to Navigation]
Comment & Response
October 28, 2013

Glucagon-like Peptide 1–Based Drugs and Pancreatic Safety—Reply

Author Affiliations
  • 1Department of Medicine and the Center for Public Health and Human Rights, Johns Hopkins University Schools of Medicine and Public Health, Baltimore, Maryland
JAMA Intern Med. 2013;173(19):1843-1844. doi:10.1001/jamainternmed.2013.8128

In Reply Brodovicz et al misinterpret the results from previous studies as being inconsistent with our study.1 The meta-analysis of clinical trials of dipeptidyl peptidase-4 (DPP-4) inhibitors by Monami et al2 was underpowered through poor ascertainment or exclusion of subclinical levels of enzyme elevations. The authors did not report an optimal information size analysis, a necessary requirement for meta-analysis of drug safety that claim no risk.3 Despite these limitations, the results of the meta-analysis could not rule as much as a 73% increased risk of acute pancreatitis—a clinically significant excess hazard, which is not dissimilar to the statistically and clinically significant excess risk of more than 2-folds in our study. Similarly, the observational study by Garg et al4 could not rule out as much a 50% increased risk of acute pancreatitis with exenatide and 30% increased risk of acute pancreatitis with sitagliptin. These differences in magnitude are possible owing to differences in study design and populations, but the direction of effect has been consistently increased in all studies.

Add or change institution