Abstract 4116975: Safety signals with SGLT2 inhibitor use in diabetic males with erythrocytosis, but not in females: Findings from the CANVAS Program and CREDENCE trial

Backgrounds: Mediation analyses have shown that an increase in hematocrit (Hct) levels significantly contributes to the cardiorenal benefits of SGLT2 inhibitors (SGLT2i). However, given the link between erythrocytosis and poor prognosis, concerns remain about whether SGLT2i-induced erythropoiesis in patients with erythrocytosis may increase thromboembolic events (TE) including myocardial infarction (MI). Methods: We conducted a pooled analysis utilizing individual participant data from the CANVAS Program and the CREDENCE trial, both aimed to evaluate the efficacy and safety of canagliflozin versus placebo in diabetic patients. The primary outcome was a composite of MI, stroke, or any other TE. Sex-specific Cox analyses were performed to assess effect modifications by baseline Hct levels, treated as categorical (anemia, normal, and erythrocytosis) or continuous variable. Results: Of the participants for whom baseline Hct values were available (98.5% [14,321/14,543]), 35% were female. Canagliflozin significantly increased the percentage of erythrocytosis from baseline to 1 year in males (5.6% to 16.9%) and females (1.9% to 5.2%), unlike placebo. Overall, canagliflozin did not alter the risk of the primary outcome for either male (HR 0.97; 95% CI, 0.86–1.10) or female (HR 0.95; 95% CI, 0.78–1.15). In males, baseline Hct levels modified treatment effect on the primary outcome, whether treated as categorical or continuous variables (Pinteraction <0.05), with benefits in anemic patients and harm in those with erythrocytosis (Figure). Meanwhile, no effect modifications were noted in females. In analyses for each component of the primary outcome, canagliflozin consistently reduced the risk of MI, stroke, and any TE in anemic male patients (HR 0.53, 0.49, and 0.71, respectively). However, mixed results were observed in those with erythrocytosis (HR 1.77, 0.62, and 1.11, respectively), suggesting detrimental effects on the primary outcome in these patients were primarily driven by MI. No heterogeneity by Hct was found for heart failure hospitalization or kidney outcome in both sexes. Conclusion: Canagliflozin may pose a safety concern regarding the risk of TE in diabetic male patients with erythrocytosis.