Major bleeding risk with non‐vitamin K antagonist oral anticoagulant vs. aspirin in heart failure: network meta‐analysis

AIMS: Relative bleeding risks of different antithrombotic agents in heart failure (HF) patients is an important consideration in treatment decision making, making detailed comparative analysis desirable. The aim of this study was to conduct a network meta‐analysis to investigate the major bleeding risk for individual novel oral anticoagulants (NOACs) vs. aspirin among patients with HF. METHODS AND RESULTS: We searched Pubmed, EMBASE, Cochrane Collaboration Central Register of Controlled Clinical Trials, and Clinicaltrials.gov from 1966 to November 2019 to identify relevant randomized clinical trials. Studies comparing individual NOACs vs. aspirin were analysed using direct study‐level meta‐analysis. Studies comparing aspirin to warfarin and NOACs to warfarin were then additionally added using network (direct and indirect) study‐level meta‐analysis. Primary endpoint was major bleeding. Final analysis included nine trials with 34 367 participants, including one direct comparison trial (apixaban vs. aspirin) and eight indirect comparison trials against the shared warfarin comparator (four aspirin trials and one trial each of apixaban, dabigatran, rivaroxaban, and edoxaban). For apixaban, network meta‐analysis combing direct and indirect comparison showed that major bleeding risk might not be different between apixaban and aspirin (odds ratio, 1.18 [95% confidence interval, 0.38 to 3.65]) in HF patients. In contrast, indirect‐comparison meta‐analysis showed dabigatran, rivaroxaban, and edoxaban compared with aspirin might be associated with a higher risk of major bleeding in HF patients. CONCLUSIONS: In network meta‐analysis, apixaban might be associated with a comparable risk of major bleeding compared with aspirin in patients with HF, while other NOACs might be associated with a higher risk. However, such results were not strongly convincing because of lack of direct comparison in an original trial and small sample size of trials and participants. A clinical trial directly comparing apixaban vs. aspirin in patients with HF and sinus rhythm may be worth undertaking.