Predictors for major bleeding in elderly (75 years and over) patients with non-valvular atrial fibrillation at high risk of bleeding: sub-analysis of the ANAFIE Registry

Abstract Background In patients with atrial fibrillation (AF) receiving anticoagulant therapy, bleeding events are associated with reduced survival. Previous studies showed that bleeding events during anticoagulant therapy were more frequent in elderly AF patients than in younger patients. HAS-BLED score has been used to assess the risk of bleeding in AF patients. In patients at high bleeding risk (HAS-BLED score ≥3), we sought to identify other risk factors associated with major bleeding not included in HAS-BLED score in elderly non-valvular AF (NVAF) patients. Purpose The All Nippon Atrial Fibrillation In the Elderly (ANAFIE) Registry is a prospective, multicenter, observational study to collect real-world data on clinical status and prognosis in more than 30,000 Japanese patients (aged ≥75 y) with NVAF. This sub-analysis of the ANAFIE Registry assessed the 2-year outcomes and identified predictors for major bleeding in elderly NVAF patients with a high bleeding risk. Methods A total of 32,275 patients from the ANAFIE Registry were divided into 2 groups according to HAS-BLED score (≥3 [high-risk group] and ≤2 [reference group]). The annualized incidence rate, hazard ratio (HR) for clinical outcomes, and independent predictors for major bleeding were analyzed using Kaplan-Meier analysis and the Cox proportional-hazards model. Results A total of 6,826 patients constituted the high-risk group: mean age, 81.8 years old (75–80 years, 37.8%; 81–84 years, 33.9%; ≥85 years, 28.3%); male ratio, 72.2%; mean creatinine clearance (CrCL), 42.7 mL/min; history of major bleeding, 14.2%; presence of non-paroxysmal AF, 62.2%; mean total number of medicines used, 7.8. Anticoagulants were used in 91.2% (warfarin [WF], 29.9%; direct oral anticoagulants [DOACs], 61.2%). Proton-pump inhibitors (PPI) were administered in 46.5%. Compared to the reference group, the high-risk group had higher annualized incidence rates (/100 patient-year) of major bleeding (1.49 vs 0.97), intracranial hemorrhage (0.95 vs 0.70), gastrointestinal (GI) bleeding (2.63 vs 1.73), and all-cause mortality (5.50 vs 3.24). All-cause mortality more frequently occurred in patients aged ≥85 years compared to 75–79 years and those with CrCL <50 mL/min compared to CrCL ≥50 mL/min. In the high-risk group, DOAC subgroup had lower incidences of the above-mentioned outcomes other than GI bleeding than WF subgroup. The following relevant factors for major bleeding not included in HAS-BLED score were identified in the high-risk group: Body mass index (BMI) ≥25.0 kg/m2 (HR, 0.40), heart failure (HF) with reduced left ventricular ejection fraction (LVEF) (HR, 1.38), a fall within 1 year (HR, 2.29), and use of PPI (HR, 0.65). Conclusions Among elderly (≥75 years) Japanese NVAF patients in the high bleeding risk group (HAS-BLED score ≥3), HF with reduced LVEF, and a fall within 1 year were identified as independent predictors of major bleeding. BMI ≥25.0 kg/m2 and PPI use were protective for major bleeding. Funding Acknowledgement Type of funding sources: Private company. Main funding source(s): Daiichi Sankyo Co., Ltd.