Redo Surgical Aortic Valve Replacement vs Valve-in-Valve Transcatheter Aortic Valve Replacement for Degenerated Bioprosthetic Valves

Abstract

Background: Valve-in-valve transcatheter aortic valve replacement (ViV-TAVR) is associated with improved perioperative safety compared with redo surgical aortic valve replacement (redo-SAVR), but long-term outcomes remain uncertain. We therefore compare long-term outcomes of ViV-TAVR and redo-SAVR.

Methods: The study included 1:1 propensity score-matched Medicare beneficiaries with degenerated bioprosthetic valves admitted between September 29, 2011, and December 30, 2020, undergoing either redo-SAVR or ViV-TAVR. Exclusion criteria included endocarditis, other concomitant cardiac surgery, and aortic valve reintervention during the same admission. The primary outcome was 5-year survival. Composite secondary outcomes included major adverse cardiovascular events (30-day operative mortality, stroke, or acute myocardial infarction) and major valve event-free survival (congestive heart failure readmission, endocarditis, or aortic valve reintervention). Time-to-event analyses used Kaplan-Meier analysis and multivariable Cox proportional hazards modeling.

Results: Overall, 4699 patients, including 1775 redo-SAVR and 2924 ViV-TAVR patients, were identified. Redo-SAVR patients were younger (median [interquartile range], 72 [68-77] years vs 79 [73-84] years) with less congestive heart failure (39.6% vs 68.8%) and prior coronary artery bypass grafting (17.9% vs 32.0%; all P < .05). In the propensity score-matched cohorts of 1256 patients each, redo-SAVR had higher major adverse cardiovascular events (17.4% vs 13.1%; P = .003) but better major valve event-free (71 [62-79] months vs 43 [38-47] months; P < .001) and 5-year (62.3% vs 46.7%; P < .001) survival. After stratification by Charlson comorbidity index, the long-term survival benefit persisted in patients of lower (67.6% vs 54.9%; P = .001) and medium or higher risk (55.1% vs 36.7%; P < .001).

Conclusions: Redo-SAVR may have better long-term survival than ViV-TAVR despite greater perioperative morbidity. Clinical trial data are needed to fully inform clinical decision-making about degenerated bioprosthetic valve reintervention, particularly for patients with reasonable life expectancy.