An increased risk of lower limb amputations (LLA) has been suspected with the use of sodium-glucose cotransporter type 2 inhibitors (SGLT2is) in the CANVAS program with canagliflozin and in pharmacovigilance reports with all SGLT2is. Even if reassuring observations were reported in several large prospective placebo-controlled cardiovascular outcome trials, real-life conditions in more frailty patients might be associated with a higher risk.

This work analyses the incidence of LLA events in retrospective observational studies that compared SGLT2i users with patients treated with dipeptidyl peptidase-4 inhibitors (DPP-4is), a pharmacological class with an excellent safety profile. A meta-analysis of 12 comparative cohort studies (9 of them using a propensity score matching) worldwide has been performed.

The relative risk of LLA tends to be slightly lower in SGLT2i users (1228 LLA events/711159 patients) versus DPP-4i users: 2167 LLA events/1121914 patients, with a hazard ratio 0.91, 95% CI 0.85-0.98, p=0.01). However, a high between-study heterogeneity was observed (I² = 79%, P<0.00001), which could not be explained by differences across countries, between studies with/without propensity score matching, between cohorts treated with/without canagliflozin or between patients with/without peripheral artery disease. The incidence rate expressed as a number of LLA events per 1000 patient.years was almost similar among SGLT2i users and DPP-4i users (2.48±1.45 versus 2.67±3.09, p=0.849).

Physicians should not fear an increased risk of LLA with SGLT2is compared with DPP-4is in daily clinical practice, even if caution may be advised in some patients exposed to special conditions.