To compare the long-term outcomes of mechanical versus bioprosthetic aortic valve replacement (AVR) in patients aged <50 years, we performed a study-level meta-analysis with reconstructed time-to-event data including studies published by December of 2023. The primary outcome was overall survival. Secondary outcomes included reoperation, major bleeding, and stroke. A total of 5 studies met our inclusion criteria, with a total of 4,245 patients (2,311 mechanical and 1,934 bioprosthetic). All studies were observational and the mean age of groups across the studies ranged from 38.2 to 43.0 years. The median follow-up time was 11.4 years (interquartile range 6.9 to 15.0). Bioprosthetic AVR was associated with reduced overall survival and higher risk of all-cause death (hazard ratio [HR] 1.170 95% confidence interval [CI] 1.002 to 1.364, p = 0.046), increased risk of reoperation over time (HR 2.581, 95% CI 2.102 to 3.168, p <0.001), decreased risk of major bleeding (HR 0.500, 95% CI 0.367 to 0.682, p <0.001), and decreased risk of stroke (HR 0.751, 95% C, 0.565 to 0.998, p = 0.049) compared with mechanical AVR in patients aged <50 years. In conclusion, for patients aged <50 years, bioprosthetic AVR is associated with increased mortality and risk of reoperation compared with mechanical valves. In contrast, mechanical AVR is associated with an increased risk of major bleeding events and stroke. These aspects should be carefully considered during the selection of valve type in this age group; however, we should keep in mind that the statistically significant differences in the risk of all-cause death and stroke might not be clinically relevant (because of marginal statistical significance).
Selecting between mechanical and bioprosthetic valves is a key consideration for patients undergoing surgical aortic valve replacement (AVR). In patients aged <50 years, recent American College of Cardiology/American Heart Association guidelines have favored mechanical valves because of their greater long-term durability and a perceived superior tolerance to long-term anticoagulation in younger patients. However, data on long-term outcomes, such as survival, reoperation, bleeding, and stroke, in this age group are limited. To date, few observational studies have been published with outcomes of mechanical and bioprosthetic AVR exclusively in patients aged <50 years. Mechanical valves have been the traditional choice in patients aged <50 years; however, their use has only constituted a slight majority in recent years while trending downward. The growing preference for bioprosthetic valves in younger patients is likely attributable to the growing availability and proved efficacy of valve-in-valve transcatheter AVR, , along with prioritization of patient preference against lifelong anticoagulation. With the shifting landscape of AVR in younger populations and a more nuanced decision-making process, understanding the outcomes between valve types is increasingly important. A previous meta-analysis by Tasoudis et al compared age-specific outcomes between bioprosthetic and mechanical AVR. This included a subgroup analysis of patients aged <50 years, with only 2 studies and explored overall survival as the sole outcome. Therefore, we performed an expanded meta-analysis using reconstructed Kaplan–Meier data from available retrospective observational studies to compare overall survival, reoperation, major bleeding, and stroke in patients aged <50 years who underwent surgical AVR with a mechanical or bioprosthetic valve.
Methods
This study followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines. All data supporting the findings of this study are available from the corresponding author upon reasonable request. Using the population, interventions, comparison, outcome, and study design strategy, studies were included based on the following criteria: (1) the population comprised patients aged <50 years who underwent surgical AVR. (2) There was a group that had a mechanical valve placed. (3) There was a second group that had a bioprosthetic valve placed. (4) The primary outcome studied was survival or all-cause mortality. The secondary outcomes included reoperation and/or major bleeding and/or stroke (outcomes are accompanied by the Kaplan–Meier curves in the original publications). (5) The study design was retrospective and monocentric/multicentric, with matched/unmatched populations.
The following sources were searched for articles meeting our inclusion criteria and published by December 31, 2023: PubMed/MEDLINE, EMBASE, and Web of Science. The exclusion criteria included studies with populations aged >50 years, studies with overlapping samples, and studies without Kaplan–Meier curves. There were no language restrictions.
The following steps were taken for study selection: (1) identification of titles and records through database search, (2) removal of duplicates, (3) screening and selection of abstracts, (4) assessment for eligibility through full-text articles, and (5) final inclusion in the study. Studies were selected by 2 independent reviewers. When there was a disagreement, a third reviewer decided to exclude or include the study. Ethical approval was not applicable for this study because it consisted of a systematic review and meta-analysis.
The Cochrane tool Risk Of Bias In Non-randomized Studies of Interventions was systematically used to assess included studies for risk of bias. A total of 2 independent reviewers made assessments and a third reviewer made the final decision if there was a disagreement.
Individual patient data were reconstructed from published Kaplan–Meier graphs of all included studies using the “curve approach.” We used the 2-stage approach, as described by Liu et al, based on the R package “IPDfromKM” (version 1.2.3.0). For the first stage, raw coordinates (time, survival/event probability) were extracted from each subgroup in each of the respective Kaplan–Meier curves. In the second stage, the raw coordinates from the first stage and the numbers at risk at given time points were used to generate individual patient data (time to event or time to last follow-up for each individual patient). The individual patient data were then merged to create a final data set for each outcome.
Overall survival, freedom from reoperation, freedom from major bleeding, and freedom from stroke in patients aged <50 years were assessed visually using Kaplan–Meier estimates. Hazard ratios (HRs) with 95% confidence intervals (CIs) for the difference between the mechanical and bioprosthetic valve groups were calculated using a Cox frailty regression model. In this model, study groups (mechanical vs bioprosthetic valve) were included as a fixed effect. Between-study heterogeneity was assessed by inclusion of a γ frailty term, where individual studies modeled as a random effect using random intercepts. A likelihood ratio test was used to test the significance of the γ frailty term. A robust variance estimator was used to account for violations of the assumption of homoscedasticity. An HR >1 indicated a higher risk of an outcome in the bioprosthetic valve group compared with the mechanical valve group. The Grambsch–Therneau test and Schoenfield residual plots were used to assess the proportional hazards assumption. Flexible parametric survival models (Royston–Parmar models or generalized survival models) with B-splines were calculated to provide HRs with 95% CIs as a measure of association between intervention and outcome while allowing for a time-varying effect. All analyses were completed with R statistical software (version 4.2.3, Foundation for Statistical Computing, Vienna, Austria).
Results
After excluding duplicates and noneligible studies, 5 studies met our eligibility criteria ( Supplementary Figure 1 ). All studies were retrospective, nonrandomized, observational studies. A total of 4 studies included propensity score–matched cohorts. The study characteristics are listed in Supplementary Table 1 . A total of 4,245 patients were included (2,311 who received a mechanical valve and 1,934 who received a bioprosthetic valve). The median follow-up time was 11.4 years (interquartile range 6.9 to 15.0). The patients’ characteristics are listed in Supplementary Tables 2 and 3 . The mean ages of groups across studies were below the age of 50 years, ranging from 38.2 to 43.0 years. Female patients were well-represented. Supplementary Figure 2 shows the qualitative assessment of the studies with the Cochrane tool Risk Of Bias In Non-randomized Studies of Interventions, and we found moderate to serious risk of bias. The main concerns were regarding confounding because of the differences in baseline variables, selection bias, and missing data in certain studies.
Data for overall survival were pooled from all 5 included studies, including 4,245 patients (mechanical: 2,311 and bioprosthetic: 1,934). The pooled Kaplan–Meier curve is shown in Figure 1 . The overall survival at 16 years in the mechanical valve group was 78.9% (95% CI 76.1 to 81.8), and the overall survival in the bioprosthetic valve group was 78.2% (95% CI 76.0 to 80.5). As per Cox frailty regression, bioprosthetic AVR was associated with reduced survival compared with the mechanical AVR (HR 1.170, 95% CI 1.002 to 1.364, p = 0.046). There was no statistical evidence of violation of the proportional hazards assumption. There was significant statistical heterogeneity between studies (likelihood ratio test p <0.001).
Data for reoperation were available in 4 of the 5 studies, totaling 4,098 patients (mechanical: 2,235 and bioprosthetic: 1,863). The pooled Kaplan–Meier curve is shown in Figure 2 . Freedom from reoperation in the mechanical valve group was 90.5% (95% CI 88.8 to 92.2), and freedom from reoperation in the bioprosthetic valve group was 69.7% (95% CI 62.3 to 78.0) at 16 years. As per Cox frailty regression, bioprosthetic AVR was associated with an increased risk of reoperation compared with mechanical AVR (HR 2.581, 95% CI 2.102 to 3.168, p <0.001). There was significant statistical heterogeneity between studies (likelihood ratio test p <0.001).
There was statistical evidence of violation of the proportional hazards assumption, as seen in the Grambsch–Therneau test (p <0.001) and Schoenfeld residuals. Figure 3 presents an analysis of time-varying HRs for reoperation based on flexible parametric models with B-splines. This shows an initially similar risk of reoperation between groups, which becomes significantly in favor of the mechanical AVR group at 7.6 months.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
