Objective
Current American Heart Association/American College of Cardiology guidelines recommend anticoagulation and antiplatelet therapy during the first 90 postoperative days; however, there is wide variability in the administration of antithrombotic therapy after bioprosthetic aortic valve replacement. We sought to determine whether early antithrombotic therapy was necessary in patients undergoing isolated bioprosthetic aortic valve implantation and who were discharged in normal sinus rhythm.
Methods
From December 2001 to October 2008, 1131 patients underwent isolated bioprosthetic aortic valve implantation at Brigham and Women's Hospital. After exclusion of patients who underwent concomitant operations (n = 138, 12%), patients who were anticoagulated preoperatively (n = 4, 0.4%), and patients who experienced postoperative refractory atrial fibrillation requiring anticoagulation at discharge (n = 128, 11%), our study base consisted of 861 patients. Patients were followed for 90 days postoperatively for the occurrence of thromboembolism, including stroke, transient ischemic attack, or peripheral thromboembolic events and bleeding complications.
Results
Of the 861 patients included in this study, 133 (15%) were anticoagulated with warfarin sodium (AC+) postoperatively and 728 (85%) were not (AC−). Patients who received postoperative anticoagulation were older; had a higher incidence of hypertension, cerebrovascular accident, and pulmonary vascular disease; and were more symptomatic at presentation. The 90-day risk of thromboembolism (cerebrovascular accident, transient ischemic attack, or peripheral thromboembolism) after surgery was 5% (n = 6) in those who were anticoagulated and 5% (n = 39) in those who were not (P = .67). Independent predictors of thromboembolism were found to be increasing age (odds ratio, 1.03; P = .03), female gender (odds ratio, 2.23; P = .005), short stature (odds ratio, 0.97; P = .002), smoking status (P = .05), New York Heart Association III/IV (odds ratio 1.77, P = .04), and a 19-mm bioprosthetic aortic valve prosthesis (odds ratio, 2.22; P = .03). Evaluation of each predictor with postoperative acetylsalicylic acid+ and AC+ interaction terms revealed that female patients (odds ratio, 0.75; P = .03 AC+; odds ratio, 0.66; P = .02 acetylsalicylic acid+) and patients with a 19-mm bioprosthetic aortic valve (odds ratio, 0.65; P = .02 AC+; odds ratio, 0.36; P = .01 acetylsalicylic acid+) had a reduction in the incidence of thromboembolism when administered acetylsalicylic acid or warfarin sodium. Patients who were in New York Heart Association III/IV also had a reduction of thromboembolism when given vitamin K antagonist (odds ratio, 0.73; P = .04); a similar trend was observed in patients given acetylsalicylic acid (odds ratio, 0.34; P = .06).
Conclusion
Early anticoagulation after isolated bioprosthetic aortic valve replacement in patients in normal sinus rhythm does not seem to reduce the risk of thromboembolism except in high-risk groups. Current recommendations should be revisited, because the only patients who may benefit from anticoagulation are female, those who are highly symptomatic, and those with a small aortic prosthesis.
