Clinical outcomes in patients with atrial fibrillation with or without concomitant diabetes after two years of edoxaban treatment: ETNA-AF-Europe registry

Abstract

Abstract Background/Introduction Concomitant diabetes mellitus (DM) in patients with atrial fibrillation (AF) has been associated with a higher risk of stroke/systemic embolic events (SEE), and higher mortality. Purpose To evaluate effectiveness and safety outcomes in AF patients with different types of concomitant DM vs those without concomitant DM after 2 years of edoxaban treatment in a sub-analysis of the prospective ETNA-AF-Europe registry. Methods Patients with AF who received edoxaban once daily were enrolled across 825 centres in 10 European countries. This sub-analysis is based on a data snapshot from 26th October 2020. Patients were grouped by DM status (with or without DM; insulin-treated or non-insulin-treated DM). Baseline characteristics are summarised descriptively. Kaplan-Meier (KM) survival curves of outcomes were constructed; annualised event rates (AERs; %/year) and comparisons between patients with or without DM by univariate analysis are presented. Comparisons of patients with insulin- vs non-insulin-treated DM were made by multivariate analysis adjusted for HbA1c. Results Patients' demographics and baseline characteristics are shown in Table 1. Of 13,133 patients with AF, 2885 (22.0%) had DM. Among patients with DM, 2756 (95.6%) received treatment: 605 (22.0%) with insulin and 2151 (78.0%) with non-insulin treatments. KM curves for effectiveness and safety outcomes are shown in Figure 1. AERs (%/year) of ischaemic stroke/transient ischaemic attack (TIA)/SEE were 0.86% in patients with AF and no DM, 0.87% with non-insulin-treated DM (p=0.9216 vs no DM) and 1.81% with insulin-treated DM (p=0.0022 vs no DM; p=0.0014 vs non-insulin-treated DM). AERs of myocardial infarction (MI) were 0.40%, 0.43% (p=0.7454 vs no DM), and 1.04% (p=0.0033 vs no DM) respectively; of major bleeding were 0.90%, 1.10% (p=0.2427 vs no DM), and 1.71% (p=0.0106 vs no DM), respectively; and of all-cause death were 3.36%, 5.02% (p<0.0001 vs no DM), and 8.91% (p<0.0001 vs no DM) respectively. In a multivariate analysis of DM patients adjusted for HbA1c (patients with AF and DM with HbA1c measured, n=1869), patients with AF and insulin-treated DM had a higher risk of stroke/TIA/SEE (HbA1c adjusted HR [95% CI]: 2.13 [1.12–4.05], p=0.0205) and all-cause death (1.83 [1.35–2.49], p=0.0001) vs those with non-insulin-treated DM; no significant difference in the risk of MI (2.10 [0.78–5.66], p=0.1448) or major bleeding (1.81 [0.95–3.44], p=0.0706) was detected between the groups. Conclusion In patients with AF and DM, risk of stroke/TIA/SEE, MI and major bleeding was confined to insulin-treated DM patients only. These patients also had poorer survival rates vs those without DM and those with non-insulin-treated DM, reinforcing previously published data showing a strong association between insulin treatment and poorer outcomes, and a mild/absent association between non-insulin treatments and poorer outcomes in patients with AF and DM. Funding Acknowledgement Type of funding sources: Private company. Main funding source(s): This research was funded by Daiichi Sankyo Europe.