Assistant Professor Division of Pharmacoepidemiology and Pharmacoeconomics, Department of Medicine, Brigham and Women’s Hospital Boston, United States
Background: Patients with cirrhosis have increased risk of atrial fibrillation (AF) and both bleeding and thromboembolic events. While current guidelines in the general population recommend direct oral anticoagulants (DOACs) for patients with AF and elevated risk of stroke, data are scarce in cirrhosis. Moreover, there are no head-to-head data to guide individual DOAC selection, in patients with cirrhosis.
Objectives: To define the comparative effectiveness and safety of apixaban and rivaroxaban, the most commonly used DOACs in the US, in cirrhosis and AF.
Methods: Using administrative claims data (CMS Medicare), we identified all patients with cirrhosis and AF initiating apixaban or rivaroxaban, and no prior use of oral anticoagulants in the preceding 183 days, between 2013 and 2019 (cohort entry date [CED]= first dispensation). We required all patients to have both cirrhosis and non-valvular AF diagnoses in the 183 days before CED. Using 1:1 propensity score (PS) matching, we balanced 103 covariates among rivaroxaban and apixaban (reference) groups. Patients were followed in an as-treated approach for the primary composite outcome of ischemic stroke or systemic embolism, major bleeding (i.e., intracranial hemorrhage [ICH] or gastrointestinal bleed [GIB]) or all-cause mortality. We estimated database-specific, PS-matched rate differences (RD) using generalized linear models, and hazard ratios (HR) using Cox proportional hazards modeling.
Results: A total of 4,900 apixaban initiations and 2,139 rivaroxaban initiations met inclusion criteria and were 1:1 PS matched (n=1,904 pairs). Compared to apixaban initiators, rivaroxaban initiators had significantly higher rates of the primary composite outcome (RD, 59.2 [95%CI=27.0, 91.3]; HR 1.72 [95%CI=1.27-2.31]), including higher rates of major bleeding (RD, 59.8 [95%CI 30.6, 89.0]; HR 1.95 [95%CI=1.40-2.71]), particularly GI bleeding (RD, 41.7 [95%CI=17.8, 65.4]; HR 2.00, 95%CI=1.33-3.01]). Findings were consistent in patients with compensated cirrhosis (RD for primary outcome 42.4, 95%CI=5.9, 78.9), and among those using reduced-dose regimens (RD, 180.9, 95%CI=53.0, 308.8). No significant differences were observed for the secondary outcomes of ischemic stroke or systemic embolism, ICH, and all-cause mortality.
Conclusions: In patients with cirrhosis and non-valvular AF, rivaroxaban initiation as compared to apixaban was associated with higher rates of major bleeding, particularly GIB, and similar rates of ischemic stroke or systemic embolism, ICH, and all-cause mortality.
