Age-adjusted risk factors are independently associated with an increased risk of major bleeding during the two-year follow-up of the ETNA-AF-Europe registry

Abstract

Abstract Background Non-vitamin K antagonist oral anticoagulants (NOACs) are a preferred treatment option over warfarin for anticoagulation in patients with atrial fibrillation (AF). Management decisions for thromboprophylaxis in AF need to balance the risk of stroke against the risk of bleeding. Various patient characteristics have been identified as independent risk factors for bleeding. A substantial number of bleeding events might be prevented if independent predictors of bleeding were identified. Purpose The present analysis aims at assessing age-adjusted risk predictors of major bleeding during two-year follow-up of unselected European patients with AF in the ETNA-AF-Europe registry. Methods ETNA-AF-Europe is a prospective, multi-centre, post-authorisation, observational study conducted in 825 centres enrolling patients treated with edoxaban once daily in 10 European countries. Wald Chi square tested the association between risk predictors and major bleeding after adjusting for age, given that age is a well-known, strong predictor of anticoagulation-related bleeding in patients with AF. Results Overall, 13,417 patients with AF (edoxaban 60 mg: n=10,248; edoxaban 30 mg: n=3169) completed the two-year follow-up. The mean age was 73.6±9.5 years, with ∼84% of the patients aged over 65 years. Mean CHA2DS2-VASc and HAS-BLED scores were 3.2 and 2.5, respectively. 438 (3.3%) patients had a history of bleeding events at baseline, of which 138 (1.0%) had a history of major bleeding event. Univariate analysis demonstrated that recalculated glomerular filtration rate (Cockcroft-Gault Equation) (GFR-CG) at baseline was the strongest age-adjusted predictor of major bleeding (Wald Chi-Square: 31.84; p<0.0001) (Figures 1 and 2), followed by history of major or clinically relevant non-major (CRNM) bleeding (24.08; p<0.0001), HAS-BLED score (21.10; p<0.0001), history of heart failure (derived) (16.59; p<0.0001), subjective frailty as assessed by physician (17.35; p=0.0002), history of major bleeding (14.14; p=0.0002), chronic obstructive pulmonary disease (COPD) (12.84; p=0.0003), CHA2DS2-VASc (12.14; p=0.0005), history of myocardial infarction (MI) (7.79; p=0.005), and left ventricular ejection fraction (LVEF) categorised by 40% (5.45; p=0.02). Conclusion Bleeding events on therapy with edoxaban can be predicted by quantifying kidney disease and capturing information on heart failure, frailty, prior bleeding, chronic obstructive lung disease and history of myocardial infarction. These data highlight the need for optimal management of anticoagulation therapy and close follow-up of patients with such risk profiles. Funding Acknowledgement Type of funding sources: Private company. Main funding source(s): Daiichi Sankyo Europe GmbH Figure 1. GFR-CG as a predictor of major bleedingFigure 2. Predictors of major bleeding