Valve-sparing root replacement versus composite valve graft root replacement: Analysis of more than 1500 patients from 2 aortic centers

Objectives: The long-term outcomes comparing valve-sparing root replacement, composite valve graft with bioprosthesis, and mechanical prosthesis have yet to be explored. We investigated the long-term survival and reintervention rates after 1 of 3 major aortic root replacements in patients with tricuspid aortic valves and patients with bicuspid aortic valves. Methods: A total of 1507 patients underwent valve-sparing root replacement (n = 700), composite valve graft with bioprosthesis (n = 703), or composite valve graft with mechanical prosthesis (n = 104) between 2004 and 2021 in 2 aortic centers, excluding those with dissection, endocarditis, stenosis, or prior aortic valve surgery. End points included mortality over time and cumulative incidence of aortic valve/proximal aorta reintervention. Multivariable Cox regression compared adjusted 12-year survival. Fine and Gray competing risk regression compared the risk and cumulative incidence of reintervention. Propensity score-matched subgroup analysis balanced the 2 major groups (composite valve graft with bioprosthesis and valve-sparing root replacement), and landmark analysis isolated outcomes beginning 4 years postoperatively. Results: On multivariable analysis, both composite valve graft with bioprosthesis (hazard ratio, 1.91, P = .001) and composite valve graft with mechanical prosthesis (hazard ratio, 2.62, P = .005) showed increased 12-year mortality risk versus valve-sparing root replacement. After propensity score matching, valve-sparing root replacement displayed improved 12-year survival versus composite valve graft with bioprosthesis (87.9% vs 78.8%, P = .033). Adjusted 12-year reintervention risk in patients receiving composite valve graft with bioprosthesis or composite valve graft with mechanical prosthesis versus valve-sparing root replacement was similar (composite valve graft with bioprosthesis subdistribution hazard ratio, 1.49, P = .170) (composite valve graft with mechanical prosthesis subdistribution hazard ratio, 0.28, P = .110), with a cumulative incidence of 7% in valve-sparing root replacement, 17% in composite valve graft with bioprosthesis, and 2% in composite valve graft with mechanical prosthesis (P = .420). Landmark analysis at 4 years showed an increased incidence of late reintervention in composite valve graft with bioprosthesis versus valve-sparing root replacement (P = .008). Conclusions: Valve-sparing root replacement, composite valve graft with mechanical prosthesis, and composite valve graft with bioprosthesis demonstrated excellent 12-year survival, with valve-sparing root replacement associated with better survival. All 3 groups have low incidence of reintervention, with valve-sparing root replacement showing decreased late postoperative need for reintervention compared with composite valve graft with bioprosthesis.