Ten-year outcomes of valve-replacing versus valve-sparing aortic root replacement in patients with Marfan syndrome: An observational multicenter study.

Abstract

We compared 10-year outcomes of aortic valve-replacing (AVR) and aortic valve-sparing (AVS) root replacement in patients with Marfan syndrome (MFS). Patients who met strict Ghent diagnostic criteria for MFS and who underwent either AVR or AVS between March 1, 2005, and December 31, 2010, were enrolled in a 3-year prospective observational study. The follow-up, which was clinical and echocardiographic, was later extended to 20 years. Time-to-event analyses were performed with and without inverse probability weighting to adjust for baseline differences; comparison between groups assessed major adverse valve-related (MAVRE) and other events. Of 316 patients enrolled, 77 underwent AVR and 239 underwent AVS; 214 gave reconsent for 20-year follow-up. For survivors, median durations for clinical and echocardiographic follow-up were 12.1 years (interquartile range, 4.6-15.1 y) and 10.1 years (interquartile range, 3.0-14.1 y), respectively. Ten-year adjusted survival rates for the AVR and AVS groups were similar (77.9% ± 4.4% vs 88.4% ± 2.6%; P = .2), as were adjusted rates of MAVRE (36.1% ± 4.2% vs 35.0% ± 3.3%; P = .5). Inverse probability weighting-adjusted competing-risk modeling associated AVS with greater cumulative incidences of structural valve deterioration or nonstructural valve dysfunction (24.9% vs 4.4%; P = .01) and aortic regurgitation ≥2+ (26.5% vs none; P < .001); similar analysis associated AVR with a greater cumulative incidence of bleeding (23.2% vs 2.7%; P < .001). Ten years postoperatively, the accumulation of structural valve deterioration in patients who underwent AVS and bleeding events in patients who underwent AVR balance out in terms of MAVRE. We plan to follow this cohort for 20 years after aortic root replacement, which may reveal late differences in reintervention rates.