Early- and mid-term outcomes of reinterventions for aortic bioprosthesis failure

Background: The aim of this study was to evaluate early- and mid-term results of our actual practice embedding redo aortic valve replacement and transcatheter procedures for aortic bioprosthetic failure. Methods: Data for aortic valve reinterventions (redo surgical aortic valve replacement, isolated redo aortic valve replacement, and valve-in-valve transcatheter aortic valve implantation, transcatheter valve-in-valve procedure) were collected (2010–2019). Logistic regression analysis was performed to identify predictors favouring the choice of transcatheter against redo surgery. Cox analysis was used to study the association of preoperative variables with survival. Survival probabilities were calculated with Kaplan-Meier analysis and compared using a log-rank test. Results: A total of 125 patients were included (redo surgical aortic valve replacement: 84 patients, valve-in-valve transcatheter aortic valve implantation: 41 patients). Median age was 74 [63–80] years, 58% of the patients were male and the median logistic EuroSCORE was 15 [8–26] %. There was no early mortality. Eighteen patients (redo surgical aortic valve replacement: 15, valve-in-valve transcatheter aortic valve implantation: 3) sustained at least one postoperative complication. At pre-discharge transthoracic echocardiogram, valve-in-valve transcatheter aortic valve implantation had significantly higher trans-prosthetic gradients (mean gradient: valve-in-valve transcatheter aortic valve implantation 18 mmHg vs. redo surgical aortic valve replacement 14 mmHg, p < 0.001). Overall survival probabilities were 94% and 73% at 1 year and 5 years, respectively. Previous coronary artery bypass surgery operation and age were independently associated with lower survival probabilities during the follow-up. Conclusions: Redo surgical aortic valve replacement and valve-in-valve transcatheter aortic valve implantation are both safe and effective for aortic bioprosthetic failure. Further valve-in-valve data are needed to determine the haemodynamic performance of transcatheter prostheses and its impact on long-term outcomes