Stroke prevention with direct oral anticoagulants in high risk elderly atrial fibrillation patients at increased bleeding risk.

Background Elderly atrial fibrillation (AF) patients with risk factors of bleeding are often considered ineligible for standard oral anticoagulants (OACs). The ELDERCARE-AF trial recently showed that edoxaban 15mg/day was superior to placebo for preventing stroke or systemic embolism and did not result in a significantly higher incidence of major bleeding. Our aim was to investigate a real-world cohort of AF patients similar to the ELDERCARE-AF cohort, with regard to the impact of direct oral anticoagulant (DOAC) use compared to non-OAC use, in relation to clinical outcomes. Methods From January 1, 2012 to December 31, 2016, 15,183 AF patients aged ≥80 years (mean age 86.63 years [SD 4.79]; 48.7% male) with a CHADS2 score ≥2 who met the enrollment criteria (generally similar to ELDERCARE-AF) were identified from the Taiwan National Health Insurance Research Database. Patients were categorized into 2 groups according to their stroke prevention strategies, ie. without OACs (n = 9,084) and DOACs (n = 6,099). Patients receiving DOACs were further stratified into reduced-dose or full-dose regimen groups. Results Compared to the non-OAC group as a reference, DOAC use (whether as reduced dose or full dose) was associated with a lower risk of ischaemic stroke (adjusted hazard ratio [aHR] 0.77, 95% confidence interval [CI] 0.67-0.88) and all-cause mortality (aHR 0.39, 95%CI 0.37-0.42) while the risks of ICH and major bleeding were similar. The risks of composite outcomes of 'ischaemic stroke or mortality' (aHR 0.42, 95%CI 0.40-0.45) and 'ischaemic stroke or major bleeding or mortality' (aHR 0.49, 95%CI 0.46-0.52) were significantly lower with DOAC use. When compared to non-OAC as the reference groups, DOACs (whether reduced dose or full dose) showed a positive NCB. The results were generally consistent even after the propensity matching. Conclusions In routine clinical care, DOACs (whether reduced or full dose) were associated with a lower risk of ischemic stroke, mortality and the composite endpoint, when compared to non-OAC use in high risk elderly AF patients at increased bleeding risk. Our findings provide complimentary 'real world' data to support the generalizability of the results of ELDERCARE-AF trial to other DOACs in the daily clinical practice.