Which SGLT2 inhibitor is safest to use?

The risk of cardiovascular events, such as heart failure (HF) and myocardial infarction (MI), are comparable across different sodium-glucose cotransporter 2 (SGLT2) inhibitors, according to a recent real-world Japan study.

“There was no significant difference in the risk of developing HF, MI, angina pectoris (AP), stroke, and atrial fibrillation (AF) among patients taking empagliflozin, dapagliflozin, canagliflozin, and other SGLT2 inhibitors,” the researchers said. “To the best of our knowledge, this is the first study to compare the wide-range cardiovascular outcomes of patients with diabetes mellitus (DM) treated with individual SGLT2 inhibitors using large-scale real-world data.”

Drawing from a nationwide claims database, the study included 25,315 DM patients (median age 52 years, 82.5 percent men) who had newly been prescribed with SGLT2 inhibitors, most common of which was empagliflozin (n=5,302). Dapagliflozin (n=4,681) and canagliflozin (n=4,411) were other common medications, while 10,921 patients were administered other types of SGLT2 inhibitors. The risks of different cardiovascular outcomes were compared across the different drugs.

Over a median follow-up of 814 days, the researchers documented 855 episodes of HF, 143 of MI, 815 of AP, 340 of stroke, and 139 of AF. Crude Cox regression analysis revealed no significant difference in the incidence of these cardiovascular outcomes when comparing across the different SGLT2 inhibitors. [Cardiovasc Diabetol 2022;21:67]

Completely controlling for confounders did not meaningfully alter the principal findings and revealed no significant difference in the risk of cardiovascular events. Compared to empagliflozin, for example, dapagliflozin conferred a comparable risk of HF (hazard ratio [HR], 1.02, 95 percent confidence interval [CI], 0.81–1.27).

Similarly, canagliflozin treatment led to a similar MI risk as empagliflozin (HR, 0.88, 95 percent CI, 0.51–1.52), while prescriptions of other SGLT2 inhibitors did not significantly aggravate the risk of stroke (HR, 0.98, 95 percent CI, 0.72–1.33).

Principal findings were likewise robust to various sensitivity analyses, focusing on patients who had taken the drug for ≥3 months and ≥9 months, and on those with available kidney function data. Multiple imputations for missing data or limiting analysis only to patients with type 2 DM also revealed no significant difference in cardiovascular outcomes among SGLT2 inhibitors.

“Our study is distinguishable from previous studies in that we compared the risk of developing specific cardiovascular outcomes in patients with DM between commercially available SGLT2 inhibitors using a large-scale real-world dataset in Japan,” the researchers said. “Simultaneously, our study raises additional research questions.”

These include the additional possible benefits of SGLT2 inhibitors on renal function, the value of these medications in secondary prevention settings (as opposed to primary prevention), and potential side effects and associated adverse events.

“To confirm our findings and address the aforementioned issues, conducting a randomized clinical trial is an ideal solution,” they said. “However, considering the required sample size, follow-up period, and costs, this would not be feasible in practice. Therefore, further analyses using epidemiological data are needed.”