Efficacy and safety of edoxaban in patients with diabetes mellitus in the ENGAGE AF-TIMI 48 trial

Background: Diabetes mellitus is an independent risk factor for stroke and atrial fibrillation. Therefore, the risk/benefit profile of the oral factor Xa inhibitor edoxaban stratified by diabetes is of clinical interest.

Methods: 21,105 patients enrolled in ENGAGE AF-TIMI 48 were stratified into 2 pre-specified groups: without (N = 13,481) and with diabetes (N = 7,624).

Results: On average, patients with diabetes were younger, and had a higher body mass index, CHA₂DS₂-VASc score and baseline endogenous Factor Xa activity. After multivariate adjustments, patients with diabetes had a similar rate of stroke and systemic embolism compared to those without diabetes (adjusted hazard ratio (HR_adj) 1.08; 95% confidence interval (CI) 0.94-1.24; p = 0.28). However, the risk of major bleeding was significantly higher in patients with diabetes (HR_adj 1.28; 95% CI 1.14-1.44; p < 0.001). The treatment effect of edoxaban (vs warfarin) was not modified by diabetes (all p-interactions > 0.05), a finding supported by the preserved edoxaban concentrations and inhibition of Factor Xa regardless of diabetes. The HRs of stroke and systemic embolism in patients receiving the higher-dose edoxaban regimen vs warfarin were 0.93 and 0.84 (p-interaction = 0.54) in those with and without diabetes respectively. The higher-dose edoxaban regimen reduced major bleeding (by 19-21%) and cardiovascular death (by 7-17%) regardless of diabetes (p-interactions = 0.81 and 0.33 respectively).

Conclusion: Patients with diabetes in ENGAGE AF-TIMI 48 had higher bleeding risk, but after adjustment similar stroke risk, compared to those without diabetes. The higher-dose edoxaban regimen had similar efficacy compared to warfarin, while reducing bleeding and cardiovascular mortality, irrespective of diabetes.