Comparative thromboembolic risk and polypharmacy-related drug-drug interaction signals among antidiabetic medications: a large-scale FAERS pharmacovigilance study.

BACKGROUND

The relationship between antidiabetic drugs and thromboembolic events remains unclear. The FDA Adverse Event Reporting System (FAERS) data was used to systematically assess safety signals and polypharmacy-related drug-drug interactions of antidiabetic medications.

METHODS

By analyzing FAERS reports from 2004 to 2025 that included antidiabetic drugs, the Reporting Odds Ratio (ROR), the Proportional Reporting Ratio (PRR), the Information Component (IC), and the Empirical Bayes Geometric Mean (EBGM) were used to assess disproportionality. Factors such as the main suspected drug, event seriousness, age, sex, US origin, and the recent reporting period were examined in sensitivity analyses. Drug-drug interactions (DDIs) were evaluated using the Ω shrinkage measure and adjusted for false discovery rate.

RESULTS

We analyzed 81,280,515 FAERS reports from 2004-2025, of which 498,750 (0.61%) involved at least one of 30 antidiabetic drugs, revealing strong arterial thromboembolic signals for rosiglitazone (IC = 4.68), gliclazide (IC = 1.55), linagliptin (IC = 1.56), dapagliflozin (IC = 1.25), and several DPP-4 inhibitors. Only insulin degludec (IC = 0.73, IC025 = 0.61) and insulin aspart (IC = 0.28, IC025 = 0.21) showed nominal venous thromboembolic signals. For GLP-1 agonists, no thromboembolic signals were detected. These observations were confirmed by sensitivity analyses across all subgroups. Moreover, 257 significant drug-drug interactions were identified, particularly between insulin analogues and simvastatin, acetaminophen, or gabapentin.

CONCLUSION

DPP-4 inhibitors, SGLT2 inhibitors, and certain insulins showed positive arterial thromboembolic signals. No positive thromboembolic signals were detected for GLP-1 agonists, consistent with their favorable cardiovascular safety profile observed in randomized controlled trials. Numerous significant drug-drug interactions were also detected, particularly for insulin analogues with other drugs. Further research is necessary to understand the clinical importance of these interactions.