The porcine valve type predicts obstructive thrombosis beyond the first three postoperative months in bioprostheses in the aortic position

Abstract

BackgroundObstructive thrombosis of bioprosthetic valves is considered rare but may have dramatic consequences for the individual patient including repeat valve replacement, thrombolysis, or long-term anticoagulation. Whether the risk of obstructive thrombosis is dependent on the type of bioprosthesis (porcine versus bovine pericardial) is uncertain. Methods and resultsBetween 2007 and 2012 a total of 1751 patients received a single stented bioprosthesis in the aortic valve position, 749 (43%) were porcine and 1002 (57%) bovine. During a mean follow-up of 3.4±1.9years, obstructive thrombosis (identified by an increase in mean pressure gradient≥20mm Hg or a decrease in velocity ratio≥0.05 and confirmed by either ECG-gated computer tomography, a return to baseline of stenosis parameters under treatment with a vitamin K antagonist, or histology in case of reoperation) was diagnosed in 17 patients with a porcine (2.3%) and none with a bovine valve (p<0.001). The cumulative probability of developing an obstructive thrombosis was significantly higher in patients with a porcine valve (p<0.001 log-rank test). Adjusting for differences in baseline variables and stratification by the estimated propensity score showed that strata with a high probability of receiving a bovine valve had the highest number of obstructive thrombosis in porcine valves. These findings were further confirmed in a Poisson model and a competing risk model including all-cause mortality. Treatment of obstructive thrombosis with a vitamin K antagonist was safe and effective in 15/17 patients. ConclusionThe porcine valve type is an independent predictor of obstructive thrombosis in bioprostheses in the aortic position.