Adverse events with sodium-glucose co-transporter-2 inhibitors: A global analysis of international spontaneous reporting systems

BACKGROUND AND AIMS: We assessed post-marketing safety of sodium-glucose co-transporter-2 inhibitors (SGLT2-Is) by analyzing adverse events (AEs) reported in international pharmacovigilance databases. METHODS AND RESULTS: Eudravigilance, WHO-Vigibase (as of Feb 25, 2017) and the FDA Adverse Event Reporting System (FAERS, from 2004 to 2016 second quarter) were queried to extract AEs recording SGLT2-Is as suspect. Disproportionality analyses (case/non-case method) were performed in FAERS by calculating the reporting odds ratios (RORs) from System Organ Classes (SOCs) to Preferred Terms (PTs) (precise clinical entities). Potential signals were defined by statistically-significant ROR (lower limit of the 95% confidence interval - LL95%CI - >1) undetected by literature analysis (as of December 2016). SGLT2-Is were recorded in 7972, 19,775, 11,137 reports (Eudravigilance, WHO-Vigibase and FAERS, respectively); in FAERS, statistically significant ROR emerged for the following SOCs: "infections and infestations" (N = 2162; LL95%CI = 3.25), "metabolism and nutrition disorders" (2278; 1.36), "renal and urinary disorders" (1665; 2.31), "reproductive system and breast disorders" (471; 4.85), "skin and subcutaneous tissue disorders" (1136; 1.52). Skin toxicity emerged as potential signal (e.g., rash, photosensitivity, urticaria as PTs), both for SGLT2-Is as a class and as individual drugs. Severe adverse skin events (81 reports, 7% of the skin cases) mainly occurred in females aged 18-65 using SGLT2-Is as single antidiabetic regimen. CONCLUSION: Among antidiabetics, SGLT2-Is are associated with higher reporting of infections, metabolism, renal and reproductive AEs, corroborating clinical trial evidence. Their large reporting patterns and the unexpected signal of skin toxicity justify active vigilance by clinicians and "real-time" monitoring by pharmacovigilance experts.