Comparison of the Outcome of Porcine Bioprosthetic Versus Mechanical Prosthetic Replacement of the Tricuspid Valve in the Ebstein Anomaly

Our objective was to determine the relative merits of using a bioprosthetic porcine valve (BPV) versus a mechanical valve (MechV) when tricuspid valve (TV) replacement is required in patients with Ebstein anomaly. From 1972 to 2006, 333 patients received a BPV and 45 received a MechV. Patient records were reviewed, vital status ascertained, and all patients not known to be deceased were mailed a medical questionnaire or contacted by telephone. Early mortality was not statistically higher for patients who had a MechV (11%) than for those who had a BPV (5%) inserted in the TV position (p = 0.173). The only independent preoperative predictor of operative mortality was moderate to severe left ventricular dysfunction (odds ratio 3.1, p = 0.03); 20-year survival was better in patients who had a BPV (75%) than for those who had a MechV (43%, p = 0.003). On multivariate analysis, after adjusting for ablation of accessory pathways, sinus rhythm at dismissal, and concomitant repair of pulmonary valve stenosis, a BPV remained a predictor of late survival (hazard ratio 0.42, p = 0.004). Survival free of reoperation on the TV at 20 years postoperatively was similar for patients who had a MechV (49%) compared with those who had a BPV (42%) inserted (p = 0.941). A greater percentage of patients who had a MechV reported endocarditis (12% vs 2%), bleeding requiring hospitalization (6% vs 3%), and thrombosis (12% vs 6%); however, none of these differences were statistically significant. In conclusion, a BPV in the tricuspid position was an independent predictor of improved survival. This may be related to the higher incidence of bleeding and thrombotic complications in the patients with MechVs or may be related to differences between the 2 groups. A BPV may offer superior late survival when compared with a MechV when TV replacement is required in patients with Ebstein anomaly, but patient selection must be individualized. (c) 2009 Elsevier Inc. (Am J Cardiol 2009;103:555-561)