Less-Invasive Aortic Valve Replacement: Trends and Outcomes From The Society of Thoracic Surgeons Database

This study compares outcomes of conventional and less-invasive (LI) approaches for aortic valve replacement (AVR) using The Society of Thoracic Surgeons database. Between 2011 and 2017, we identified 122,474 patients undergoing isolated primary AVR. Patients were categorized into 3 groups: (1) full sternotomy (FS) (n = 98,549; 78%), (2) partial sternotomy (PS) (n = 17,306; 15%), and (3) right thoracotomy (RT) (n = 6619; 7%). The rate of LI-AVR increased from 17% in 2011 to 23% in 2016 (P < .001). Femoral cannulation was used in 1.5% of FS, 5.4% of PS, and 71% of RT patients (P < .001). Full sternotomy patients were older and had higher rates of preoperative renal failure, atrial fibrillation, and stroke, and had a higher NYHA function class, lower ejection fraction, and higher STS risk score. Total operative, cardiopulmonary bypass, and cross-clamp time were longest in RT-AVR patients and shortest in those who had FS-AVR. Overall, unadjusted operative mortality was 1.9% (1.05% among low-risk patients) and was not different among the 3 groups (1.97% FS, 1.77% PS, and 1.90% RT; P = .4). The rate of postoperative stroke was 1.2% and was not different among the 3 groups (1.2% FS, 1.3% PS, and 1.1% RT; P = .3). After risk adjustment, these differences remained nonsignificant. After risk adjustment, prolonged ventilation and atrial fibrillation were less common in PS-AVR patients. The adjusted risk for blood transfusion was lower in RT-AVR patients, as was the incidence of renal failure. Femoral cannulation was not associated with increased risk for stroke or mortality after LI-AVR. Less-invasive AVR is associated with an operative mortality and postoperative stroke rate similar to that of FS. Less-invasive AVRs should serve as a benchmark for comparison between transcatheter aortic valve replacement and surgical AVR in low-risk patients.