The long-term clinical impact of mitral prosthesis–patient mismatch (PPM) after mitral valve replacement (MVR) remains incompletely defined. We aimed to systematically assess the prevalence of mitral PPM and its association with long-term mortality, cardiac death, and heart failure-related hospitalizations. We conducted a systematic review and meta-analysis of observational studies reporting Kaplan–Meier time-to-event data in patients with and without mitral PPM. Sixteen studies, including 10,872 patients, were analyzed. The pooled prevalence of any mitral PPM was 43% (95% confidence interval [CI], 29% to 59%), with significant heterogeneity ( I ² = 100%). All-cause mortality was higher with PPM (hazard ratio [HR], 1.32; 95% CI, 1.20 to 1.45; p < 0.001; RMST difference, −3.35 years; 95% CI, −3.98 to −2.71; p < 0.001), as was cardiac mortality (HR, 1.96; 95% CI, 1.54 to 2.51; p < 0.001) and risk of heart failure hospitalization (HR, 2.82; 95% CI, 1.86 to 4.29; p < 0.001). Risk-adjusted analyses confirmed these associations for all-cause and cardiac mortality. Severity-stratified analyses demonstrated a gradient effect: moderate PPM showed a trend toward higher mortality (HR, 1.12; 95% CI, 0.98 to 1.27; p = 0.08), whereas severe PPM significantly increased mortality risk (HR, 1.36; 95% CI, 1.15 to 1.60; p < 0.001). Sensitivity analyses according to PPM quantification method (in vivo vs Doppler) and leave-one-out testing confirmed the robustness of the results. In conclusion, mitral PPM is common and independently associated with increased long-term all-cause mortality, cardiac death, and heart failure hospitalizations after MVR, with a dose-response relationship according to PPM severity. These findings highlight the importance of preventive strategies to avoid PPM when planning MVR.
Prosthesis-patient mismatch (PPM) occurs when the effective orifice area (EOA) of a prosthetic valve is too small relative to a patient’s body size, resulting in abnormally high postoperative transvalvular pressure gradients and impaired hemodynamics. While PPM has been extensively studied in the aortic position, its hemodynamic and clinical implications after mitral valve replacement (MVR) are less well characterized. In the aortic position, severe PPM occurs in 2% to 20% of patients and is associated with a 30% risk increase in mortality, equating to a 2-year reduction in life expectancy over 25 years of follow-up, and a 60% higher risk of heart failure rehospitalization. ,
By analogy to the aortic position, mitral PPM can be viewed as a form of residual mitral stenosis, resulting in persistently elevated transprosthetic gradients. The concept of mitral PPM is not new. In 1981, Rahimtoola and Murphy first reported a patient who remained symptomatic after MVR, with persistent pulmonary hypertension (PH) and progressive right-sided heart failure, due to a prosthesis that was too small relative to the patient’s body size. Subsequent studies by Dumesnil and colleagues demonstrated that the indexed EOA (EOAi) correlates with transvalvular gradients and pulmonary pressures, providing a physiologic basis for defining mitral PPM. More recently, Pibarot et al , further emphasized the importance of EOAi in guiding prosthesis selection to optimize hemodynamic performance and minimize the risk of PPM. In the mitral position, EOAi correlates less strongly with mean transprosthetic gradient than in the aortic valve, likely due to greater sensitivity to heart rate and diastolic filling, as a result, thresholds for mitral PPM differ from those of aortic prostheses, and its clinical impact may also differ. Nevertheless, evidence over the past two decades indicates that mitral PPM is independently associated with adverse outcomes. In light of the growing evidence of its prognostic significance, this study aimed to assess the impact of PPM on outcomes after MVR by means of a pooled analysis of Kaplan–Meier‐estimated reconstructed time‐to‐event data from studies comparing patients with and without PPM after MVR to evaluate its effect on patient‐relevant outcomes.
Methods
This systematic review and meta-analysis were performed in adherence to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines ( Supplementary Table 1 ). The study protocol was prospectively registered (CRD420251041407). Ethical approval was not required for this study-level meta-analysis.
Search strategy
A comprehensive search of electronic databases (MEDLINE, Embase, Web of Science) was performed to identify studies published up to February 1, 2025. The search strategy was developed using the PECOS framework to address the research question: In patients undergoing surgical MVR, does PPM affect long-term overall survival compared with no PPM? We searched for the following terms: “mismatch OR PPM OR patient‐prosthesis mismatch OR prosthesis‐patient mismatch AND MVR OR mitral valve replacement OR surgical mitral valve replacement.” Additionally, reference lists of eligible studies and prior meta-analyses were screened to identify further relevant publications. Two investigators (X.J., M.P.S.) independently screened titles, abstracts, and full texts using prespecified inclusion and exclusion criteria. Studies were eligible if they:
-
•
Enrolled patients undergoing surgical MVR;
-
•
Reported Kaplan-Meier curves for overall survival or other patient-relevant outcomes stratified by PPM status (any, moderate, or severe);
-
•
Provided sufficient data for extraction.
Exclusion criteria included reviews, case reports, single-arm studies, studies including exclusively transcatheter MVRs, or studies with overlapping populations. When overlapping cohorts were identified, the most recent or most comprehensive study was included.
Data extraction, quality assessment, and statistical analysis
Patient characteristics and Kaplan–Meier curves were independently extracted by two reviewers (X.J., M.P.S.). The methodological quality and risk of bias of included studies were assessed independently using the Cochrane Risk of Bias tool and classified as low, moderate, or high risk of bias. Kaplan–Meier curves were digitized using Web Plot Digitizer software. Reconstructed time-to-event data and individual patient data were derived from the digitized Kaplan–Meier plots, along with the corresponding total number of patients, total events, and the number of patients at risk at various time intervals for each study arm. Hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated using a Cox frailty model with robust standard errors. Heterogeneity across studies was assessed by testing for an interaction between the study and treatment effect, incorporating a γ frailty term to account for between-study variability, with studies modeled as random effects. The significance of the variance parameter was evaluated using the likelihood ratio test. Proportional hazards were checked using the Grambsch–Therneau test and Schoenfeld residual diagnostic plots. The restricted mean survival time (RMST), interpretable as a quantification of loss of life expectancy, was calculated and the difference between both study arms was compared. As a further analysis, if more than two studies were available, conventional “two-step” meta-analysis using the DerSimonian Laird random-effects model was conducted, and forest plots were used to display the pooled estimates. Between-study heterogeneity was assessed through Cochran Q statistic and I 2. Publication bias was assessed via funnel plots and Egger’s test for each outcome of interest. Jackknife resampling analyses were performed, excluding each study in turn. Prespecified random-effects meta-regression analyses were also conducted to examine the impact of moderator variables on the outcomes. Significance testing was performed at the two-tailed 5% significance level. All analyses were completed with R Statistical Software (version 4.4.0, Foundation for Statistical Computing, Vienna, Austria).
Results
Study selection
Our systematic search identified 701 records from PubMed/MEDLINE, EMBASE, Scopus, and CCTR. After removing 384 duplicates, 317 abstracts were screened, of which 49 full-text articles were assessed for eligibility. Thirty-three studies were excluded due to absence of Kaplan-Meier curves, lack of long-term outcomes, overlapping cohorts, or outcomes not of interest. Ultimately, 16 studies were included in the meta-analysis ( Figure 1 ). ,,,,,,,,,,,,,,, All included studies were nonrandomized and retrospective observational ( Table 1 ). The studies collectively included 10,872 patients, with reported incidence of any PPM ranging from 8.8% to 85.9%. The pooled proportion using a random-effects model was 43% (95% CI, 29 to 59), with significant heterogeneity ( I ² = 100%, p < 0.01). Patients had a mean age ranging from 38 to 75 years, with variable female representation. Most studies included a mix of mechanical and bioprosthetic mitral valves, while a minority included exclusively mechanical or bioprosthetic valves. The pooled proportion using a random-effects model was 70% (95% CI, 56 to 83), with significant heterogeneity ( I ² = 100%, p < 0.01). Table 2 summarizes baseline patient characteristics, comorbidities, echocardiographic indices, and operative details. Supplementary Table 2 shows the qualitative assessment of the studies, and the overall internal validity was considered high risk of bias, mostly due to unmeasured confounding, and measurement of exposure (PPM).
Study selection. Legend: PRISMA flow diagram illustrating the identification, screening, eligibility assessment, and inclusion of studies in the systematic review and meta-analysis.
Table 1
Characteristics of the studies evaluating outcomes of PPM following mitral valve replacement
| Study | Country | Period | Total sample— n | Mechanical— n | PPM— n | Any PPM definition | Adjustment for confounders |
|---|---|---|---|---|---|---|---|
| Lam et al | Canada | 1985–2005 | 884 | 657 (74.3%) | 280 (32%) | EOAi ≤1.25 | Multivariable analysis |
| Magne et al | Canada | 1986–2005 | 929 | 789 (84.9%) | 725 (78%) | EOAi ≤1.20 | Multivariable analysis |
| Jamieson et al | Canada | 1982–2002 | 2,440 | 1,083 (44.4%) | 2,095 (85.9%) | EOAi ≤1.20 | Multivariable analysis |
| Aziz et al | USA | 1992–2008 | 765 | 440 (58%) | 393 (51%) | EOAi ≤1.20 | Multivariable analysis |
| Bouchard et al | Canada | 1992–2005 | 714 | 714 (100%) | 63 (8.8%) | EOAi ≤1.30 | Multivariable analysis |
| Sakamoto et al | Japan | 1992–2005 | 84 | 75 (91%) | 25 (30%) | EOAi ≤1.20 | None |
| Matsuura et al | Japan | 1995–2008 | 212 | 131 (65.1%) | 125 (56%) | EOAi ≤1.20 | None |
| Shi et al | Multicenter | 2001–2009 | 1,006 | 622 (62%) | 665 (66.1%) | EOAi ≤1.20 | Propensity score and Multivariable analysis |
| Sato et al | Japan | 2000–2011 | 142 | 110 (77.5%) | 60 (42.3%) | EOAi ≤1.20 | Multivariable analysis |
| Borracci et al | Argentina | 2009–2013 | 136 | 78 (54.4%) | 96 (71%) | EOAi ≤1.20 | None |
| Hwang et al | Korea | 1992–2012 | 760 | 640 (84.2%) | 147 (19.3%) | EOAi ≤1.20 | Propensity score and Multivariable analysis |
| Ammannaya et al | India | 1990–2016 | 500 | 500 (100%) | 186 (37.2%) | EOAi ≤1.20 | Propensity score and Multivariable analysis |
| Lee et al | Korea | 2000–2013 | 445* | 362 (81.3%) | 165 (37.1%) | EOAi ≤1.20 | None |
| Akuffu et al | China | 2013–2015 | 1,067 | 868 (81.3%) | 189 (17.7%) | EOAi ≤1.20 | None |
| Tsubota et al | Japan | 2007–2015 | 660 | 189 (28.6%) | 248 (37.8%) | EOAi ≤1.20 | Multivariable analysis and Propensity Score Matching |
| Kitada et al | Japan | 2010–2018 | 128 | 0 (0.0%) | 34 (26.6%) | EOAi ≤1.20 | Multivariable analysis |
Summary of published studies assessing PPM after mitral valve replacement, including study country, period, total sample size, number of patients receiving mechanical prostheses, number and proportion of patients with PPM, definition of PPM, and adjustment for confounding variables.
*All patients in the study by Lee et al had rheumatic mitral disease.
EOAi = indexed effective orifice area; MVR = mitral valve replacement; PPM = prosthesis-patient mismatch.
Table 2
Patient characteristics of the included studies
| Study | Age, years | Female, % | BSA, m 2 | HTN, % | DM, % | AF, % | CVD, % | Prior cardiac surgery, % | CKD, % | LVEF, % | MS, % | MR, % | Bypass time, minute | Cross-clamp time, minute |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Lam et al | 63 ± 12/65 ± 11 | 381 (73)/140 (27) | 1.7 ± 0.2/1.9 ± 0.2 | NA | NA | 248 (41)/76 (27) | NA | 112 (19)/33 (12) | NA | NA | 262 (44)/109 (39) | 157 (26)/91 (33) | NA | NA |
| Magne et al | 64.6 ± 10.6/62.5 ± 11.5 | 165 (80.9)/401 (55.3) | 1.54 ± 0.16/1.75 ± 0.17 | 80 (39.2)/277 (32.2) | 22 (10.8)/147 (20.3) | 103 (50.5)/269 (37.1) | NA | NA | 23 (11.3)/146 (20.1) | 59 ± 12/58.2 ± 13.3 | 110 (53.9)/349 (48.1) | 94 (46.1)/360 (49.7) | 99 ± 45/114 ± 46 | 77 ± 28/82 ± 30 |
| Jamieson et al | 62.5 ± 14.3/64.1 ± 11.6 | 274 (79.4)/971 (46.3) | 1.53 ± 0.12/1.85 ± 0.22 | NA | NA | 122 (35.4)/578 (27.6) | NA | NA | NA | NA | NA | NA | NA | NA |
| Aziz et al | 60 ± 15/63 ± 14 | NA | 1.76 ± 0.20/1.95 ± 0.25 | NA | 57 (15)/120 (30.5) | NA | NA | NA | 47 (13)/64 (16.3) | 47 ± 13/47 ± 13 | NA | NA | NA | NA |
| Bouchard et al | 60.4 ± 9.8/56.9 ± 12.1 | 408 (74.7)/96 (56.5) | NA | NA | NA | 242 (44.4)/64 (38.1) | NA | NA | NA | NA | NA | NA | 98.6 ± 39.7/120.2 ± 62.3 | 71.4 ± 27.7/87.4 ± 38.3 |
| Sakamoto et al | 57.0 ± 10.8/60.4 ± 10.9 | 7 (28)/41 (70) | 1.65 ± 0.15/1.49 ± 0.14 | 5 (20)/15 (25) | 1 (4)/6 (10) | 11 (44)/42 (71) | 7 (28)/15 (25) | 6 (24)/11 (19) | 0 (0)/2 (2) | 64.4 ± 13.1/64.8 ± 10.6 | 4 (16)/24 (41) | 19 (76)/25 (42) | 217 ± 38/219 ± 37 | 147 ± 34/149 ± 32 |
| Matsuura et al | 64.0 ± 8.6/62.5 ± 10.6 | 67 (77)/51 (41) | 1.40 ± 0.11/1.62 ± 0.15 | 12 (13.7)/25 (20) | 4 (4.6)/13 (10.4) | 42 (48.3)/45 (36) | 10 (11.4)/10 (8) | NA | NA | 58.7 ± 14.3/59.1 ± 13.5 | NA | NA | 104 ± 43/110 ± 46 | 74.7 ± 30.8/80.1 ± 33.6 |
| Shi et al | 64 ± 15/65 ± 13 | 198 (58)/306 (46) | 1.7 ± 0.2/1.8 ± 0.2 | 173 (51)/392 (58.9) | 39 (11)/114 (17.1) | 145 (43)/260 (39) | 42 (12)/73 (11) | NA | NA | NA | 110 (32)/180 (27) | 289 (85)/582 (87.5) | 143 ± 66/147 ± 67 | 108 ± 50/111 ± 49 |
| Sato et al | 61.2 ± 8.3/61.8 ± 11.4 | 66 (80.5)/27 (45) | NA | 25 (30.5)/19 (31.7) | 12 (14.6)/11 (18.3) | 82 (100)/60 (100) | 15 (18.3)/14 (23.3) | 4 (4.9)/6 (10.0) | 1 (1.2)/2 (3.3) | NA | 33 (40.2)/27 (45.0) | 15 (18.3)/7 (11.7) | NA | NA |
| Borracci et al | 65.1 ± 12.9/67.6 ± 11.3 | 26 (65)/48 (50) | 1.74 ± 0.17/1.87 ± 0.18 | NA | NA | 11 (28)/38 (40) | NA | NA | NA | NA | NA | NA | NA | NA |
| Hwang et al | 50.4 ± 11.9/53.0 ± 13.9 | 450 (73)/58 (39) | 1.52 ± 0.14/1.71 ± 0.19 | 43 (7.0)/20 (13.6) | 40 (6.5)/17 (11.6) | 268 (43.7)/93 (63.3) | 81 (13.2)/17 (11.6) | NA | 5 (0.8)/4 (2.7) | NA | NA | NA | NA | NA |
| Ammannaya et al | 38.5 ± 16.2/39.9 ± 16.6 | 192 (61)/119 (64) | 1.52 ± 0.14/1.55 ± 0.15 | 54 (17.2)/29 (15.6) | 48 (15.3)/31 (11.3) | 98 (31.2)/31 (35.5) | NA | 71 (22.6)/46 (24.7) | NA | NA | 165 (52.5)/103 (55.4) | 92 (29.3)/53 (28.5) | 98.0 ± 33.0/99.1 ± 34.2 | 71.1 ± 23.8/73.0 ± 25.6 |
| Lee et al | 53.28 ± 10.82/55.96 ± 12.36 | 69 (24.6)/52 (31.5) | 1.55 ± 0.15/1.61 ± 0.14 | NA | NA | NA | NA | NA | NA | 60.71 ± 9.52/62.70 ± 9.28 | NA | NA | NA | NA |
| Akuffu et al | 54 (46–61)/63 (55–67) | 588 (67)/100 (52.9) | 1.55 ± 0.16/1.64 ± 0.17 | 145 (16.5)/47 (24.9) | 80 (9.2)/22 (11.6) | 441 (50.2)/92 (48.7) | 28 (3.2)/6 (3.2) | 46 (5.2)/8 (4.2) | NA | 62.09 ± 8.38/62.23 ± 8.95 | 658 (74.9)/128 (67.7) | 382 (43.5)/88 (46.5) | 83 (70–93)/83 (70–89) | 50 (41–63)/45 (40–55) |
| Tsubota et al | 69.4/71.1 | 269 (65.9)/95 (37.9) | 1.5/1.6 | 173 (42.4)/129 (52) | 66 (16.2)/61 (24.6) | 213 (52.2)/81 (32.7) | 38 (9.3)/17 (6.9) | 86 (21.1)/43 (17.3) | NA | 59.2 ± 12.5/56.9 ± 13.7 | 71 (17.4)/45 (18.1) | 250 (61.3)/169 (68.1) | 135 ± 55.5/169.2 ± 79.1 | 100.5 ± 43.2/123 ± 51.6 |
| Kitada et al | 74.9 ± 7.7/76.2 ± 5.3 | 50 (53.2)/10 (29.4) | 1.48 ± 0.17/1.60 ± 0.18 | 49 (52.1)/23 (67.6) | 14 (14.9)/6 (17.7) | 34 (36.2)/12 (35.3) | 21 (22.3)/14 (41.2) | 18 (19.2)/4 (11.8) | 9 (9.6)/5 (14.7) | 57.9 ± 13.1/58.2 ± 12.6 | NA | 34 (100)/78 (83.0) | 168 ± 59/183 ± 54 | 137 ± 50/150 ± 42 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
