Comparative Safety and Efficacy of Rivaroxaban Versus Warfarin in Pediatric Patients Following the Fontan Procedure: A Retrospective Cohort Study

Background: The extracardiac Fontan procedure, a palliative intervention for single-ventricle physiology, is associated with significant thromboembolism and bleeding risks. Warfarin has been the standard anticoagulant, but its limitations have prompted exploration of nonvitamin K oral anticoagulants, such as rivaroxaban.

Objective: To compare the safety and efficacy of rivaroxaban vs. warfarin as postoperative anticoagulation in pediatric patients following the extracardiac Fontan procedure, focusing on thromboembolic events, bleeding complications, and treatment adherence.

Methods: This retrospective cohort study included 369 pediatric patients (aged 3–17 years) who underwent the extracardiac Fontan procedure at a single center from 2015 to 2022, selected from 412 cases reviewed, with 43 excluded due to incomplete follow-up or comorbidities. Patients received either warfarin (n = 177) or rivaroxaban (n = 192) for anticoagulation. Baseline characteristics, including age, sex, body weight, and pulmonary artery pressure, were comparable between groups. Anticoagulation was initiated on postoperative Day 1 per institutional protocol, excluding aspirin to standardize thromboembolism prevention due to high risk in nonanticoagulated patients. Warfarin was titrated to an international normalized ratio (INR) of 2.0–3.0, while rivaroxaban was dosed per European Medicines Agency guidelines. Outcomes included thromboembolic events (graft thrombosis, pulmonary embolism, and transient ischemic attacks [TIAs]), major and minor bleeding, mortality, and treatment discontinuation over a mean follow-up of 5 years. TIAs were defined per American Heart Association guidelines as transient neurological dysfunction lasting less than 24 h without infarction on neuroimaging. Statistical analyses used Fisher’s exact test, Kaplan–Meier survival analysis, and multivariable logistic regression, with p < 0.05 indicating significance; the study was designed to detect a 3% difference in major bleeding. The limited number of thromboembolic events resulted in wide confidence intervals, limiting precision in between-group comparisons.

Results: Thromboembolic events occurred in 5.1% (n = 9) of warfarin patients and 2.1% (n = 4) of rivaroxaban patients (OR = 2.5, 95% CI: 0.8–8.2). Major bleeding was significantly higher with warfarin (3.4%, n = 6; 4 intracranial and 2 gastrointestinal) than rivaroxaban (0.5%, n = 1; gastrointestinal; OR = 7.0, 95% CI: 1.2–40.8). Minor bleeding rates were 9.0% (warfarin) vs. 5.7% (rivaroxaban). Two warfarin-related deaths (intracranial hemorrhage and systemic embolism post-TIA) were recorded; none occurred with rivaroxaban. Treatment discontinuation was higher with warfarin (5.1% vs. 0.5%).

Conclusions: Rivaroxaban demonstrated a superior safety profile compared to warfarin, with significantly lower major bleeding rates, no associated mortality, and improved treatment adherence in pediatric Fontan patients. Its fixed-dose regimen simplifies management, although implementation requires consideration of cost and formulation access. Risk-stratified approaches and larger prospective trials are needed to optimize anticoagulation strategies.