Premature Structural Failure of Trifecta Bioprosthesis in Midterm Follow-up: A Single-Center Study.

BACKGROUND

A cluster of aortic bioprosthetic valve failures, most of which were Trifecta bioprostheses, was observed in Southampton General Hospital, Southampton, United Kingdom. This study was performed to assess whether the cluster represents a significant failure of this valve model or whether there is a selection bias that can explain the failure of these valves.

METHODS

This retrospective study evaluated all bioprosthetic aortic valve replacement operations performed between 2011 and 2016 inclusive in our center. The study compared the performance of the Trifecta valve (Abbott, Abbott Park, IL) with that of Perimount (Edwards Lifesciences, Irvine, CA), Perimount Magna Ease, and Mitroflow (LivaNova, London, United Kingdom) bioprostheses. In addition, the study analyzed patient-related and valve-related risk factors for early failure in the failed valves.

RESULTS

A total of 2807 bioprosthetic aortic valve replacements were performed. Of these, 836 were Trifecta valves, 1031 were Perimount, 449 were Perimount Magna Ease, and 351 were Mitroflow valves. A total of 24 Trifecta valves had premature structural failure, a number significantly higher than seen with Perimount or Perimount Magna Ease (no failure, P < .001 and P < .005, respectively) valves and the Mitroflow valve (1 failure, P < .05). There was no difference in the incidence of endocarditis or death. At the time of valve failure, 17 (71%) of the failed Trifecta valves had moderate or severe regurgitation, and the average peak gradient was 61 ± 29 mm Hg. The median failed prosthetic size was 23 mm. One failed valve had severe patient-prosthesis mismatch. The mean time to failure was 4.5 ± 1.7 years.

CONCLUSIONS

The Trifecta bioprosthesis has an increased incidence of early structural valve failure, which is significantly higher than that of Perimount, Perimount Magna Ease, or Mitroflow. No patient-related or valve-related cause for the failure could be identified.