480 Effect of Aspirin and Warfarin on Thromboembolic Events in Patients With Fontan Surgery

Abstract

Patients with univentricular hearts and Fontan palliation are at risk for thromboembolic complications. While ASA and oral anticoagulation therapies are the mainstay of prophylaxis, controversy exists as to the optimal strategy. We, therefore, examined the effect of prophylaxis with either ASA or warfarin on the time to thromboembolism (TE). All patients born before January 1, 1985, without a prior history of TE undergoing fontan surgery at Children's Hospital Boston between 1973 and 1991 surviving beyond 30 days post surgery and without a TE event within 2 weeks of their surgery were included. A total of 210 patients (49% male) underwent a first Fontan surgery at a median age of 8.5 years. Two thirds (63.8%) had either tricuspid atresia or a double-inlet left ventricle and 70.5% had a morphologic left ventricle. 102 patients (48.6%) had an RA to PA connection, 23 (11.0%) an RA to RV conduit, 81 (38.6%) a lateral tunnel, and 4 (1.9%) an extracardiac connection. No TE prophylaxis was prescribed in 50.0%, whereas 24.3% received ASA only, and 25.7% oral anticoagulation. Over a median follow-up of 14.5 years, 40 (19.1%) patients experienced their first TE event. Twenty-eight events occurred among patients without prophylaxis (26.7%), 5 (9.8%) in the ASA group, and 7 (13.0%) in the oral anticoagulation group. Freedom from TE or TE-related death, censoring for Fontan revision or conversion, is shown in the Kaplan-Meier plot in Figure 1. In multivariate Cox regression analyses, a lack of ASA or oral anticoagulation was associated with an adjusted HR of 8.1 [95% CI (3.4-19.3), p<0.001] compared to prophylaxis with either. Although both ASA and warfarin were individually superior to no prophylaxis, there was no significant advantage of one therapy over the other (p=0.891). A low cardiac index (HR 2.7, 95% CI 1.2-6.2), history of supraventricular tachycardia (HR 2.9, 95% CI 1.1-7.5), and a diagnosis of hypoplastic left heart (HR 55.9, 95% CI 5.2-599) were predictive of TE events. Secondary analyses involving alternative group assignment algorithms and varying the post-operative TE event exclusion period yielded similar results. In this retrospective cohort study, prophylaxis with either ASA or oral anticoagulation appeared to offer significant protection against TE events following Fontan palliation, with no difference between the two therapies. Reasons as to why oral anticoagulation did not outperform ASA remain speculative, including similar efficacy, suboptimal compliance/anticoagulation levels, and/or residual confounding, and warrant further study.