Long-term effectiveness and safety of edoxaban in patients with atrial fibrillation: 4-year follow-up of more than 13,000 patients from the ETNA-AF-Europe study

Abstract

Abstract Background Direct, non-vitamin K antagonist oral anticoagulants (DOACs/NOACs) have become the first-choice therapy for stroke prevention in patients with atrial fibrillation (AF) at risk of stroke. Data on long-term effectiveness and safety of NOACs are scarce and not available from randomised clinical trials. Purpose We analysed 4-year outcome data in 13,632 European patients with AF treated with the NOAC, edoxaban. Methods This is the first report of the four-year follow-up information collected in the ETNA-AF-Europe (NCT02944019) study. It is a prospective registry, conducted in 825 centres enrolling edoxaban-treated patients in 10 European countries. Design and follow-up were agreed with the European Medicines Agency as part of the post-approval safety assessment of edoxaban. Key efficacy and safety outcomes were adjudicated. Results 13,164 patients (56.7% males) with AF treated with edoxaban (60 mg OD n=9617; 30 mg OD n=3042; missing dose n=505) were analysed: median (interquartile range [IQR]) age was 75 (68–80) years, weight 80 (70–90) kg, creatinine clearance (CrCl) 68.9 (52.7–87.9) mL/min, modified CHA2DS2-VASc and HAS-BLED scores 3 (2–4) and 2 (1–2), respectively in the overall population. Annualised all-cause and cardiovascular deaths in the overall population were 4.1%/year and 1.0%/year, respectively; higher in the edoxaban 30 mg vs 60 mg cohort (Figure 1). Annualised rates of stroke, transient ischaemic attack and systemic embolic events rates were <1% (0.6%/year, 0.3%/year and 0.1%/year, respectively); these proportions were similar between the two dose cohorts (Figure 1). The time-to-first event curves for all-cause death, stroke and major bleeding were almost linear throughout the 4-year follow-up (Figure 2), irrespective of dose cohort. Compared with patients receiving edoxaban 60 mg, those prescribed 30 mg were older (median [IQR] age: 80 [75–85] vs 73 [66–78] years), had greater frail proportion (27.0% vs 6.6%), a lower CrCl (46.1 [37.4–59.0] vs 75.9 [61.9–93.8] mL/min) and a higher CHA2DS2-VASc score (4 [3–5] vs 3 [2–4]). Annualised bleeding rate was 3.0%/year and higher in the 30 mg vs 60 mg cohort (Figure 1). Annualised major bleeding, intracranial haemorrhage (ICH) and major gastrointestinal (GI) bleeding rates were low (0.9%/year, 0.2%/year and 0.4%/year, respectively), with higher major bleeding and major GI bleeding rates in patients treated with edoxaban 30 mg OD (Figure 1). ICH rates were similar in both dose cohorts. Conclusion These findings illustrate the long-term effectiveness and safety of edoxaban. Treatment with edoxaban over 4 years in patients with AF is associated with a relatively low, linearly increasing rate of all-cause death and ischaemic stroke. Rates of bleeding and especially ICH rates were low.