One-hundred-seventy patients were under the age of 60. Patient baseline and operative characteristics are summarized in Table 1; 64.1% of patients were male, and the mean age was 49 years (range, 24–59 years). Sixty-four patients (37.6%) were aged < 50 years and 25 (14.7%) were aged < 40 years. Aortic valve stenosis (n = 85, 50.0%) was the most frequent indication for surgery. The fundamental etiologies were as follows: bicuspid aortic valve (n = 74, 43.5%), degenerative (n = 51, 30.0%), infective endocarditis (n = 22, 12.9%), rheumatic (n = 17, 10.0%), and aortitis (n = 6, 3.5%).

A total of 170 patients underwent AVR, with 152 (89.4%) undergoing mechanical AVR and 18 (10.6%) receiving bioprosthetic AVR. The operative characteristics are listed in Table 1. The valve characteristics are listed in Online Resources 1 and 2. The mechanical valve types were ATS (Medtronic, Minneapolis, MN, USA) in 84 patients, On-X (On-X Life Technologies Inc., Austin, TX, USA) in 38 patients, St. Jude (St. Jude Medical Inc., St. Paul, MN, USA) in 23 patients, and CarboMedics (Sorin SpA, Milan, Italy) in 7 patients. The bioprosthetic valve types were INSPIRIS RESILIA (Edwards Lifesciences LLC) in 6 patients, Carpentier-Edwards Magna Ease (Edwards Lifesciences LLC, Irvine, CA, USA) in 10 patients, Mosaic bioprosthesis (Medtronic, Inc.) in 1 patient, and Trifecta (Abbott Vascular, Santa Clara, CA, USA) in 1 patient.

Regarding the surgical approach, all 170 patients underwent median sternotomy. The concomitant operations are listed in Table 2. Thirty patients (17.6%) required hemiarch aortic replacement and 28 (16.4%) underwent mitral valve surgery. The mean cardiopulmonary bypass time was 140.3 ± 59.6 min and mean ischemic time was 103.9 ± 44.8 min.

The early outcomes are summarized in Table 2. There was no operative mortality. The overall in-hospital mortality rate was 0.6% (n = 1). The patient with bicuspid valve died due to sepsis. Regarding postoperative complications, reoperation due to perivalvular leak was required in 1 patient (0.6%), and reoperation due to bleeding was required in 7 patients (4.1%). There was no incidence of stroke. Postoperative atrial fibrillation was reported in 27 patients (15.8%) and heart block in 1 patient (0.6%).

During the observation period, 10 late deaths occurred, including 7 cardiac-related deaths. The causes of late death are summarized in Table 3. The overall survival rate was 95.4 ± 1.7% at 5 years, 93.9 ± 2.3% at 10 years, 90.6 ± 3.8% at 15 years, and 73.2 ± 11.8% at 20 years (Fig. 2a). The results of Cox proportional hazards regression analysis for mortality after surgery is shown in Table 4. Hemodialysis was a prognostic factor of mortality (HR 0.034; P =  < 0.001).

Kaplan–Meier curve. a Kaplan–Meier curve of freedom from survival. b Kaplan–Meier curve of freedom from major bleeding. c Kaplan–Meier curve of freedom from thromboembolic events. d Kaplan–Meier curve of freedom from valve-related reoperation

In total, six major bleeding events occurred during the follow-up period, all of which occurred in patients with mechanical valves. Freedom from major bleeding was 96.4 ± 1.6% at 5, 10, and 15 years and 89.0 ± 7.3% at 20 years (Fig. 2b).

In total, six thromboembolic events occurred during the follow-up period, all of which occurred in patients with mechanical valves. Freedom from thromboembolic events was 98.7 ± 1.3% at 5 years, 97.3 ± 1.9% at 10 years, 90.5 ± 4.5% at 15 years, and 79.0 ± 11.3% at 20 years (Fig. 2c).

There were six valve-related reoperations during the follow-up period, two of which were performed in patients with a bioprosthetic valve (one case each of perivalvular leaks and SVD) and four in patients with a mechanical valve (two cases of pannus formations and two cases of endocarditis). The causes of valve-related reoperation are detailed in Table 5. Freedom from valve-related reoperation was 99.4 ± 0.6% at 5 years, 97.8 ± 1.7% at 10 and 15 years, and 63.9 ± 14.5% at 20 years (Fig. 2d).