Excess Reintervention With Mitroflow Prosthesis for Aortic Valve Replacement: Ten-Year Outcomes of a Randomized Trial

BACKGROUND Current bioprostheses are considered to have improved durability and better hemodynamic perfor-mance compared with previous designs, but there are limited comparative data on late outcomes. METHODS From 2009 through 2011, 300 adults with severe aortic valve stenosis undergoing aortic valve replacement (AVR) were randomly assigned to receive Edwards Magna, St Jude Epic, or Sorin Mitroflow bioprostheses (n = 100, n = 101, n = 99, respectively). Overall survival was analyzed using Kaplan-Meier and Cox proportional hazards methods, whereas competing risk analysis was used for all time-to-event outcomes. Serial echocardiographic data were fitted with longitudinal models stratified by implant valve size. RESULTS During median follow-up of 9.8 years (interquartile range, 8.7-10.2), 10-year survival was 50% for the Magna group, 42% for the Epic group, and 41% for the Mitroflow group (P = .415). Cumulative risk of stroke was 9% at 10 years, and rates were comparable for the three groups. Indexed aortic valve area and mean gradients were similar among the three groups receiving 19 mm and 21 mm valves, but in larger (23 mm or more) prostheses, gradients were lower (P < .001) and indexed aortic valve areas were higher in the Magna group (P < .001). The 10-year risk of endocarditis differed by group (P = .033), with higher incidence in the Mitroflow vs the Magna group (7% vs 0%, P = .019). Late risk of reinterventions in the Mitroflow group was 22%, compared with 0% in the Magna group (P < .001) and 5% in the Epic group (P = .008). CONCLUSIONS The Magna valve had the lowest gradients and largest indexed aortic valve area with larger implant sizes. The Mitroflow bioprosthesis is associated with an increased rate of reintervention and possible increased risk of infection compared with Magna and Epic valves.