This study demonstrated that the long-term outcomes of AVR in patients aged < 60 years were acceptable. The 15- and 20-year survival rates were 90.6 ± 3.8% and 73.2 ± 11.8%, respectively. Freedom from major bleeding events, valve-related reoperation, and thromboembolic events at 15 years was > 95%, 95%, and 90%, respectively.
Major bleeding remains the most devastating complication of AVR with mechanical valves. Reports comparing long-term outcomes after isolated AVR in patients aged 50–69 years are common, depending on whether they received a bioprosthetic or mechanical prosthetic valve [10–13]. Two retrospective studies of > 4000 patients aged 50–69 years reported a higher rate of major bleeding with a mechanical prosthesis [10, 11]. Few studies have compared the long-term outcomes of isolated AVR in patients aged < 60 years who underwent operation after the year 2000 [14, 15]. These studies reported that there was no difference between freedom from major bleeding and thromboembolic events at the 10-year follow-up in both groups in patients aged < 60 years [14, 15]. Wang et al. [15] reported that freedom from major bleeding events at 5 and 10 years was 98.1% and 96.9% in patients with bioprosthetic valves and 95.4% and 91.5% in patients with mechanical valves, respectively. Freedom from major bleeding events in this study was similar to that previously reported, although mechanical valve usage rate for AVR was 89.4% (152/170 patients).
The guidelines’ recommendation for anticoagulation of bileaflet mechanical valves in the aortic position is an international normalized ratio (INR) of 2.0–2.5 [16]. We infer that the lower incidence of major bleeding events in our study could be related to the strict INR control of 2.0–2.5. Additionally, the target INR was adjusted and controlled individually according to the risk of major bleeding or embolism. The Prospective Randomized On-X Valve Reduced Anticoagulation Clinical Trial (PROACT) reported that INR was safely maintained at 1.5–2.0 in high-risk patients, without differences in mortality or thromboembolic complications [17]. Recent data show that lower INR targets could reduce the rates of valve-related events.
With recent trends showing an increase in bioprosthetic valve use because of an active lifestyle and avoidance of lifelong anticoagulation [18], the risk of reoperation for SVD in patients with bioprosthetic valves has increased in younger patients [19]. Bourguignon et al. reported that freedom from SVD with bioprosthetic valve placement for patients aged < 60 years at 15 and 20 years was 66.8% and 37.2%, respectively. These patients also showed a higher risk of reoperation [20]. Conversely, Bouhout et al. reported that freedom from reoperation with a mechanical valve at 10 years was 94.1% in patients with a mean age of 53 years at the time of surgery [21]. In a propensity-matched cohort study by Christ et al. (year of operation: 1993–2002), freedom from reoperation with a mechanical valve after 20 years was 90.4% in patients aged < 60 years [22]. This study reported that 6 patients underwent reoperation. Five patients required valve-related reoperation more than 10 years after the initial surgery, four of which were mechanical valves: two for pannus formation and two for endocarditis. One patient with a bioprosthetic valve required reoperation for SVD 20 years after the initial surgery. Although reoperation after mechanical valve implantation is not commonly needed, prosthetic dysfunction remains a persistent concern. A recent annual report by the Japanese Association for Thoracic Surgery showed that the 30-day mortality for initial surgical AVR and redo-AVR was 1.9% and 3.4%, respectively [19]. Redo-AVR has approximately twice the surgical mortality rate as initial surgery. Fortunately, there have been no reports of mortality associated with redo-AVR.
The use of bioprosthetic valves has markedly increased in Japan [6]. Most patients who opt for prosthetic valves are informed about receiving or expect to receive ViV (SAVR in TAVR) in the future [23]. However, there are limited long-term data on ViV (SAVR in TAVR). A multicenter study with 1,006 ViV TAVR patients recently showed that these patients have an 8-year survival rate of only 38% [24]. Additionally, SAVR with a bioprosthetic valve may not be an optimal treatment choice unless there are no complications including additional treatment such as ViV for at least 15 years. Moreover, comparison with data such as that presented in this study is essential for the evaluation of long-term outcomes.
Our study has the inherent limitations of a single-institution retrospective study and was subject to selection bias. We could not compensate for the bias in patient selection by employing propensity matching to compare patients with similar backgrounds due to the small sample size.
Therefore, further multicenter clinical studies involving patients with similar backgrounds are warranted. Due to the small number of events, hemodialysis was a positive predictor through Cox proportional hazards regression analysis for mortality. Thus, long-term observation is warranted.