Confidence in bioprosthetic durability without anticoagulation has led to a contemporary trend of offering tissue valves to younger patients. Close monitoring of mechanical valve patients at lower international normalized ratio (INR) thresholds may reduce anticoagulation morbidity. We prospectively compared results of bioprostheses (BP) and a bileaflet mechanical prosthesis (MP) monitored at low INR thresholds.Methods
Patients received an isolated Carpentier-Edwards bovine or Medtronic porcine BP or the On-X MP. INR targets were 2.0 for MP recipients undergoing aortic valve replacement and 2.5 for mitral valve replacement, using point-of-care home monitoring. Operations consecutively performed between September 2003 and August 2007 were propensity matched using logistic regression by preoperative covariates of age, sex, valve position, New York Heart Association class, ejection fraction, atrial fibrillation, and creatinine.Results
Of 469 operations consecutively performed, 172 patients were matched for analysis. Mean age was 56.2 ± 9.6 years (range, 24 to 72 years). Median follow-up was 4.0 years (total follow-up, 667.0 patient-years). No late bleeding events occurred. The thromboembolic complication rates per patient-year were 0.77% for MP and 0.78% for BP (p = 0.67). There were 9 BP deaths vs 4 MP (2.35% vs 1.41%/patient-year; BP hazard ratio, 0.60; 95% confidence interval, 0.13 to 2.15). Postoperative linearized mortality benefit of MP was observed as early as 5 years, reaching significance at 7.5 years (p = 0.04).Conclusions
Patients aged 65 years or younger with MP and closely monitored anticoagulation display noninferiority to BP from bleeding and thromboembolic complications. MP valves begin to confer mortality benefit over BP as early as 7.5 years postoperatively. Because this predates the timeline of typical structural valve degeneration, equipoise is suggested when choosing a BP in a young patient with future expectations of valve-in-valve intervention.