Abstract
BACKGROUND
Mitral valve replacement (MVR) is recognized for its significant operative risks. This risk is commonly quantified with The Society of Thoracic Surgeons (STS) Risk Calculator. The present study aimed to investigate the long-term (5-year) predictive capabilities of the STS predicted risk of mortality (PROM) in patients undergoing isolated MVR.
METHODS
Patients undergoing isolated MVR at a high-volume mitral reference center from 2010–2023 were stratified by STS PROM (low: <3%, intermediate: 3%–6%, and high/extreme: >6%). Primary endpoints were short-term (30-day) and long-term (5-year) mortality. Kaplan-Meier survival analysis was used to compare short- and long-term survival among risk stratified groups. Cox regression modeling was used to identify predictors of long-term survival.
RESULTS
Of 1444 of MVR patients, 855 patients underwent isolated MVR. The mean STS PROM for low, intermediate, and high/extreme-risk groups was 1.7%, 4.3%, and 12.7%, respectively. Thirty-day mortality was lower than predicted in the low (0.3%) and high/extreme risk (10.9%) groups. Five-year survival was 90.8%, 75.5%, and 56.4% respectively. Kaplan-Meier survival analysis at 5 years identified significant long-term survival probability differences among the PROM groups, even after accounting for perioperative risk (P < .001). Preoperative dialysis (hazard ratio [HR], 1.70; 95% CI 1.05–2.73; P = .03), hypertension (HR, 1.71; 95% CI, 1.13–2.59; P = .011), and urgent operative status (HR, 1.57; 95% CI, 1.09–2.26; P = .02) were among the characteristics associated with worse long-term survival.
CONCLUSIONS
Although predictive models may overestimate surgical risk at high-volume mitral centers, the STS PROM correlates with 5-year mortality in isolated MVR. These findings may provide more specific risk stratification criteria, identifying patients best suited for operative intervention.
Mitral valve replacement (MVR) is recognized for its significant operative risks, leading many high-risk patients to be excluded from surgical management.1,2 As transcatheter therapies advance, the importance of surgical risk stratification and patient selection has grown considerably. Surgeons may utilize various tools to evaluate potential surgical risk, most commonly The Society of Thoracic Surgeons (STS) Risk Calculator.3 Composed of a robust and evolving calculation of patient specific variables, the STS Risk Calculator allows for predictions of major morbidity, operative mortality, and short-term (30-day) outcomes.
While the STS Risk Calculator is well-established for predicting short-term mortality (predicted risk of mortality [PROM]), its association with long-term mortality remains less understood.4,5 Emerging studies have suggested that PROM may be a reliable independent predictor of long-term survival in cardiac surgery, perhaps even up to 14 years after surgery.6,7 Although foundational, these studies include a diverse range of cardiac procedures with varying operative risk and limited numbers of mitral operations. Further work is needed to examine patient- and hospital-specific factors that may influence survival rates in isolated mitral surgery.
Accurate short- and long-term risk stratification of patients needing MVR could aid in the identification of those more suitable for less-invasive transcatheter interventions. The present study aimed to investigate the long-term (5-year) predictive capabilities of the STS PROM in patients undergoing isolated MVR at a high-volume mitral reference center. Secondarily, we sought to identify independent predictors of long-term survival in these patients. Finally, we aimed to assess the short-term (30-day) predictive accuracy of PROM for isolated MVR.
PATIENTS AND METHODS
STUDY POPULATION.
This study was approved by the University of Michigan institutional review board (HUM00148119). Data from all patients who underwent MVR from 2010 to 2023 (N = 1444), at a single high-volume mitral reference center, were extracted from the STS institutional database. Isolated MVR was defined as MVR procedures with or without tricuspid valve repair, left atrial appendage exclusion and surgical ablation for atrial fibrillation, and/or atrial septal defect repair. Patients who underwent concomitant aortic valve and coronary artery bypass procedures were excluded from analysis. Considering the evolving definition of isolated MVR during this analysis period (including the addition of concomitant tricuspid valve repair), all patients who met the most current criteria for isolated MVR had a PROM calculation performed using all available variables with the STS Risk Calculator (version 2.0.5). A total of 855 patients were identified and subsequently stratified into tertiles by PROM. Low risk was defined as a PROM <3% (n = 319), intermediate 3%–6% (n = 276), and high/extreme >6% (n = 260).
OUTCOMES.
The primary endpoints of this study were short-term (30-day) and long-term (5-year) mortality, acquired through use of the institutional electronic medical record, the institutional STS database, and the Michigan Death Index.8 Secondary outcomes included identification of independent pre- and postoperative patient and hospital specific predictors of long-term survival.
STATISTICAL ANALYSIS.
Patient characteristics and postoperative complications were presented as counts (%) for categorical variables and median (interquartile range [IQR], 25th-75th percentile) for continuous variables. Differences among groups were evaluated using the χ2 test for categorical variables and the Kruskal-Wallis test for continuous variables. Kaplan-Meier survival analysis was used to compare short-term and long-term survival among groups stratified by risk, including a separate landmark analysis at 30 days. Finally, Cox proportional hazards regression models identified independent predictors of long-term survival. All statistical analysis were preformed using R (version 4.3.2; www.r-project.org) with level of significance set at α < 0.05.
RESULTS
PATIENT CHARACTERISTICS.
The median age of the entire cohort was 64 (IQR, 54–73) years, including 63.2% female and 81.6% White. The majority (75.5%) of operations was elective, while 23.0% were urgent status. History of previous myocardial infarction was seen in 30.5%, and previous diagnosis of heart failure (New York Heart Association class I-IV) was present in 57.3%. Overall, a cross-clamp was used in 71.5% of cases, with a median cross-clamp time of 97 (IQR, 74–122) minutes and a median cardiopulmonary bypass time of 112 (IQR, 84–149) minutes. A combined 67.6% of patients had moderate or worse mitral regurgitation prior to surgical intervention. Mean (SD) PROM of the entire population was 5.9% (6.54%). An extensive summary of patient characteristics can be seen in Table 1.
TABLE 1.
Preoperative Characteristic by Society of Thoracic Surgeons Risk Calculator Predicted Risk of Mortality (PROM)
| Variable | Overall | PROM <3% | PROM 3%-6% | PROM >6% | P Value |
|---|---|---|---|---|---|
|
| |||||
| N | 855 | 319 | 276 | 260 | |
| Age, y | 64 (54–73) | 56 (47–64) | 69 (59–75) | 70 (60–78) | <.001 |
| Cross-clamp time, min | 97 (74–122) | 94 (68–121) | 99 (78–125) | 94 (77–122) | .226 |
| Sex | |||||
| Male | 315 (36.8) | 122 (38.2) | 107 (38.8) | 86 (33.1) | .412 |
| Female | 540 (63.2) | 197 (61.8) | 169 (61.2) | 174 (66.9) | .318 |
| Racea | |||||
| White | 690 (81.6) | 252 (79.0) | 232 (84.1) | 206 (79.2) | .265 |
| Non-White | 165 (19.3) | 67 (21.0) | 44 (15.9) | 54 (20.8) | |
| Tobacco usea | 423 (49.4) | 156 (48.9) | 140 (50.7) | 127 (48.8) | .909 |
| Hypertension | 645 (75.4) | 188 (58.9) | 226 (81.9) | 231 (88.8) | <.001 |
| Diabetes | 253 (29.6) | 63 (19.7) | 84 (30.4) | 106 (40.8) | <.001 |
| Chronic liver disease | 60 (7.0) | 16 (5.0) | 15 (5.4) | 29 (11.2) | .007 |
| Chronic kidney disease | 54 (6.3) | 6 (1.9) | 13 (4.7) | 35 (13.5) | <.001 |
| Heart failurea | 489 (57.3) | 161 (50.8) | 165 (59.8) | 163 (62.7) | <.001 |
| Atrial fibrillationc | <.001 | ||||
| None | 314 (46.7) | 150 (58.4) | 89 (42.0) | 75 (36.9) | |
| Persistent | 182 (27.1) | 44 (17.1) | 72 (34.0) | 66 (32.5) | |
| Paroxysmal | 176 (26.2) | 63 (24.5) | 51 (24.1) | 62 (30.5) | |
| Previous MI | 261 (30.5) | 74 (23.2) | 83 (30.1) | 104 (40.0) | <.001 |
| MR severityb | .102 | ||||
| Trace | 70 (8.2) | 21 (6.6) | 20 (7.2) | 29 (11.2) | |
| Mild | 150 (17.5) | 43 (13.5) | 55 (19.9) | 52 (20.0) | |
| Moderate | 159 (18.6) | 58 (18.2) | 51 (18.5) | 50 (19.2) | |
| Severe | 419 (49.0) | 173 (54.2) | 134 (48.6) | 112 (43.1) | |
| Etiology | |||||
| Rheumatic | 220 (25.7) | 106 (33.2) | 65 (23.6) | 49 (18.8) | <.001 |
| Endocarditis | 83 (9.7) | 18 (5.6) | 27 (9.8) | 38 (14.6) | .001 |
| Degenerative | 95 (11.1) | 28 (8.8) | 37 (13.4) | 30 (11.5) | .194 |
| Functional | 28 (3.3) | 6 (1.9) | 10 (3.6) | 12 (4.6) | .171 |
| MAC | 36 (4.2) | 7 (2.2) | 14 (5.1) | 15 (5.8) | .071 |
| Mixed | 82 (9.6) | 21 (6.6) | 31 (11.2) | 30 (11.5) | .07 |
| Other | 353 (41.3) | 144 (45.1) | 109 (39.5) | 100 (38.5) | .204 |
| Unknown | 67 (7.8) | 22 (6.9) | 20 (7.2) | 25 (9.6) | .435 |
| Operative statusa | <.001 | ||||
| Elective | 644 (75.5) | 292 (92.1) | 210 (76.1) | 142 (54.6) | |
| Urgent | 196 (23.0) | 24 (7.6) | 66 (23.9) | 106 (40.8) | |
| Emergent | 12 (1.4) | 1 (0.3) | 0 (0.0) | 11 (4.2) | |
| Emergent salvage | 1 (0.1) | 0 (0.0) | 0 (0.0) | 1 (0.4) | |
| CSIa | .193 | ||||
| 1st | 402 (47.1) | 158 (49.8) | 131 (47.5) | 113 (43.5) | |
| 2nd | 343 (40.2) | 127 (40.1) | 114 (41.3) | 102 (39.2) | |
| 3rd | 80 (9.4) | 27 (8.5) | 22 (8.0) | 31 (11.9) | |
| 4th | 23 (2.7) | 5 (1.6) | 7 (2.5) | 11 (4.2) | |
| 5th | 5 (0.6) | 0 (0.0) | 2 (0.7) | 3 (1.2) | |
| Operative approach | .423 | ||||
| Sternotomy | 591 (69.1) | 223 (69.9) | 196 (71.0) | 174 (65.2) | |
| Thoracotomy | 250 (29.2) | 93 (29.2) | 76 (27.5) | 85 (31.8) | |
| Other | 14 (1.7) | 3 (0.9) | 4 (1.5) | 8 (3.0) | |
Degree of missingness ≤ 1%;
degree missingness ≤ 7.5%;
degree missingness ≤ 22%. Heart failure defined as New York Heart Association Class ≥III. Values reported as median (interquartile range: 25th-75th percentile) or n (%). P values based on χ2 test for categorical variables and Kruskal-Wallis test for continuous variables. CSI, cardiac surgery incidence (number of previous cardiac operations including current operation); MAC, mitral annulus calcification; MI, myocardial infarction; MR, mitral regurgitation; PROM, predicted risk of mortality.
Several key differences among groups existed. First, mean (SD) PROM for low, intermediate, and high/extreme risk groups was 1.7% (0.7%), 4.3% (0.8%), and 12.7% (8.3%), respectively. The low risk group was younger than the intermediate and high/extreme risk groups (low: 56 [IQR, 47–64] years, intermediate: 69 [IQR, 59–75] years, high/extreme: 70 [IQR, 60–78] years; P < .001). High risk patients were more likely to have urgent operation compared with the low and intermediate risk groups (low: 7.6%, intermediate: 23.9%, high/extreme: 40.8%; P < .001). As expected, there were several differences in the prevalence of comorbid disease. With each risk group, there was a higher burden of preoperative comorbidities, including hypertension (58.9% vs 81.9% vs 88.8%; P < .001), diabetes (19.7% vs 30.4 % vs 40.8%; P < .001), history of previous myocardial infarction (23.2% vs 30.1% vs 40.0%; P < .001), and heart failure (50.8% vs 59.8% vs 62.7%; P < .001).
No differences in sex, operative approach (sternotomy vs thoracotomy), number of previous cardiac operations, median cross-clamp time, or mitral regurgitation severity were identified. Interestingly, differences in the proportion of rheumatic mitral disease (33.2% vs 23.6% vs 18.8%; P < .001), as well as endocarditis (5.6% vs 9.8% vs 14.6%; P = .001) were seen. Notably, differences in the proportion of concomitant tricuspid valve repair were also observed across groups (27.9% vs 40.9% vs 41.9%; P < .001). A comprehensive list of isolated MVR procedure specifics can be found in Table 2.
TABLE 2.
Isolated Mitral Valve Replacement and Qualifying Procedure Types by PROM Group
| Procedure | Overall | PROM <3% | PROM 3%-6% | PROM >6% | P Value |
|---|---|---|---|---|---|
|
| |||||
| N | 855 | 319 | 276 | 260 | |
| MVR (only) | 390 (45.6) | 164 (51.4) | 113 (40.9) | 113 (43.5) | .027 |
| TVr | 311 (36.4) | 89 (27.9) | 113 (40.9) | 109 (41.9) | <.001 |
| AF | 253 (29.6) | 95 (29.8) | 86 (31.2) | 72 (27.7) | .677 |
| ASD | 59 (6.9) | 21 (6.6) | 21 (7.6) | 17 (6.5) | .853 |
| Multiple | 115 (13.5) | 38 (11.9) | 40 (14.5) | 37 (14.2) | .022 |
| All | 14 (1.6) | 4 (1.3) | 6 (2.2) | 4 (1.5) | .022 |
Values represent absolute counts of each procedure reported as n (%); P value based on χ2 test for categorical variables. AF, atrial fibrillation surgical procedure defined as left atrial appendage exclusion and ablation; ASD, atrial septal defect repair; MVR, mitral valve replacement with no other concomitant procedures; PROM, predicted risk of mortality; TVr, tricuspid valve repair.
POSTOPERATIVE OUTCOMES.
Following isolated MVR, the low and intermediate risk groups had significantly shorter total intensive care unit lengths of stay (hours) (low: 52 [IQR, 29–95], intermediate: 95 [IQR, 48–159], high/extreme: 121 [IQR, 72–203]; P < .001), as well as total hospital lengths of stay (days) (low: 8 [IQR, 5–13], intermediate: 12 [IQR, 7–21], high/extreme: 17 [IQR, 10–28]; P < .001), compared to the high/extreme risk group. Furthermore, the high/extreme risk group experienced significantly more postoperative complications, such as prolonged ventilation (low: 6.9% vs intermediate: 18.1% vs high/extreme: 36.5%; P < .001), postoperative pneumonia (low: 2.2% vs intermediate: 8.0% vs high/extreme: 12.3%; P < .001), and renal failure (low: 1.9% vs intermediate: 4.7% vs high/extreme: 13.5%; P < .001). Overall, the majority of patients (74.2%) was discharged home, although a higher proportion of intermediate and high/extreme risk patients were sent to extended care rehabilitation facilities. A summary of postoperative outcomes can be seen in Table 3.
TABLE 3.
Postoperative Outcomes and Complications by PROM Group After Isolated Mitral Valve Replacements
| Variable | Overall | PROM <3% | PROM 3%-6% | PROM >6% | P Value |
|---|---|---|---|---|---|
|
| |||||
| N | 855 | 319 | 276 | 260 | |
| Total hospital length of staya, d | 11 (7–20) | 8 (5–13) | 12 (7–21) | 17 (10–28) | <.001 |
| Total ICU length of staya, h | 79 (45–149) | 52 (29–95) | 95 (48–159) | 121 (72–203) | <.001 |
| Prolonged mechanical ventilation | 167 (19.5) | 22 (6.9) | 50 (18.1) | 95 (36.5) | <.001 |
| Pneumonia | 61 (7.1) | 7 (2.2) | 22 (8.0) | 32 (12.3) | <.001 |
| Renal failure | 54 (6.3) | 6 (1.9) | 13 (4.7) | 35 (13.5) | <.001 |
| Hospital readmissionb | 165 (20.4) | 50 (15.8) | 53 (20.2) | 62 (27.0) | .028 |
| ICU readmissiona | 45 (5.3) | 13 (4.1) | 11 (4.0) | 21 (8.1) | .05 |
| Operative mortality | 61 (7.1) | 2 (0.6) | 21 (7.6) | 38 (14.6) | <.001 |
| Discharge locationb | <.001 | ||||
| Extended care facility | 195 (24.0) | 23 (7.3) | 68 (25.8) | 104 (45.0) | |
| Home | 602 (74.2) | 293 (92.7) | 191 (72.3) | 118 (51.1) | |
| Other | 14 (1.7) | 0 (0) | 5 (1.9) | 9 (3.9) | |
Degree of missingness ≤1%;
degree of missingness ≤5%. Prolonged mechanical ventilation defined by STS as postoperative pulmonary ventilation >24 hours; operative mortality defined as all deaths, regardless of cause, occurring during the hospitalization in which operation was performed, even after 30 days (including patients transferred to other acute care facilities), and all deaths occurring prior to 30 days. Values reported as median (interquartile range: 25th-75th percentile) or n (%). P values based on χ2 test for categorical variables and Kruskal-Wallis test for continuous variables. ICU, intensive care unit; PROM, predicted risk of mortality.
SHORT-TERM SURVIVAL.
The observed to expected 30-day mortality was lower in the low and high/extreme risk groups, respectively (0.3% vs 1.7%; 10.9% vs 12.7%). In the intermediate risk group, the observed 30-day mortality nearly matched expected (4.3% vs 4.4%). A summary of short- and long-term mortality rates are presented in Table 4. Kaplan-Meier survival analysis at 30 days revealed significant differences in survival probabilities among our PROM groups, particularly in the high/extreme risk group (P < .001) (Figure 1A). Notably, after 15 days, predicted survival differences of the low and intermediate risk groups became more pronounced.
TABLE 4.
Isolated Mitral Valve Replacement Observed Mortality by PROM Group
| Observed Mortality | ||||||
|---|---|---|---|---|---|---|
| Group | Mean PROM | 30 Day | 1 Year | 3 Year | 5 Year | 10 Year |
|
| ||||||
| PROM <3% | 1.70 (0.71) | 0.32 | 3.17 | 6.35 | 9.21 | 12.06 |
| PROM 3%-6% | 4.30 (0.83) | 4.40 | 14.65 | 21.25 | 24.54 | 30.04 |
| PROM >6% | 12.68 (8.27) | 10.89 | 27.63 | 37.74 | 43.58 | 49.03 |
Values are presented as percent. Mean PROM reported as mean (SD). PROM, predicted risk of mortality.
FIGURE 1.
(A) Short-term and (B) long-term Kaplan-Meier survival analysis by predicted risk of mortality groups.
LATE SURVIVAL.
Mean follow-up time was 2.26 years (95% CI, 1.96–2.55; P < .001), and did not differ among groups. Survival at 5 years was 90.8% in the low risk group, 75.5% in the intermediate risk group, and 56.4% in the high/extreme risk group. Additional information regarding proportion of known follow-up can be found in Supplemental Table 1. Kaplan-Meier survival analysis at 5 years identified significant long-term survival probability differences among the PROM groups (P < .001) (Figure 1B). Higher PROM was associated with worse long-term survival, even after adjusting for perioperative risk using a 30-day land-mark survival analysis (P < .001) (Supplemental Figure 1).
MULTIVARIABLE ANALYSIS.
Cox proportional hazards regression models identified several patient characteristics associated with worse long-term survival. Age (hazard ratio [HR], 1.02; 95% CI, 1.01–1.03; P = .001), preoperative dialysis (HR, 1.70; 95% CI, 1.05–2.73; P = .030), hypertension (HR, 1.71; 95% CI, 1.13–2.59; P = .011), urgent operative status (HR, 1.57; 95% CI, 1.09–2.26; P = .016), and tobacco use (HR, 1.47; 95% CI, 1.08–2.00; P = .015) were among the preoperative characteristics associated with worse long-term survival. Interestingly, heart failure, chronic liver disease, history of stroke, and emergent operative status were not associated with worse survival in our models. Postoperatively, hospital readmission (HR, 1.41; 95% CI, 1.03–1.92; P = .032) was associated with worse survival. A forest plot of patient and hospital specific characteristics as independent predictors of survival can be found in Figure 2 with a comprehensive list in Supplemental Table 2.
FIGURE 2.
Hazard ratios forest plot for patient- and hospital-specific characteristics as independent predictors of survival in isolated mitral valve replacement. Heart failure defined as NewYork Heart Association Class I-IV. (CAD, coronary artery disease; CVA, cerebrovascular accident; Hx, history; ICU, intensive care unit; LOS, length of stay; MI, myocardial infarction; Op, operative.)
COMMENT
The present study evaluated outcomes of isolated MVR as well as the association of STS PROM with operative mortality and late survival. First, we identified significant 5-year survival probability differences of three risk stratified PROM categories, even after accounting for perioperative risk. Higher PROM was associated with worse long-term survival. Second, we identified several patient- and hospital-specific risk factors associated with worse long-term survival, including hypertension, urgent operative status, preoperative dialysis, and hospital readmission. Third, we found that the short-term (30-day) predictive capability of PROM may overestimate observed mortality at a high-volume mitral center, particularly for low and high risk patients.
The emergence of transcatheter MVR (TMVR) has highlighted the growing importance of accurately stratifying surgical risk. Numerous clinical trials have explored the potential benefits of TMVR, with the most notable being improved early survival in higher-risk patients.9 However, despite this short-term mortality benefit, concerns persist regarding the durability and long-term outcomes of TMVR.10 There remains a patient population that could still benefit from surgical MVR, but an accurate method for identifying this group has yet to be established. Based on the notable differences in long-term survival probabilities observed with our defined PROM risk categories (low <3%, intermediate 3%–6%, high/extreme >6%), we propose that these groups could effectively classify long-term risk, pending further detailed validation. Clearly defined risk-stratified groups with significant survival differences could assist in determining candidacy for isolated MVR, facilitating more informed patient decision-making that considers both short- and long-term outcomes.
Previous studies have sought to elucidate the relationship between PROM and long-term survival in cardiac surgery. One of the most robust investigations explored this association among 24,222 qualifying PROM procedures at a single academic center from 1996 to 2009.6 The group found that among those who survived the first 30 days, each percentage-point increase in PROM score corresponded to a 9.6% increase (95% CI, 9.3–10.0; P < .001) in the instantaneous hazard of death. The authors suggest that this association continued long term, with the predictive ability of PROM persisting for up to 14 years postoperatively. That study, however, included all types of operations and did not separate MVR from mitral valve repair. Given the well-established differences in all-cause long-term morbidity and mortality between MVR and mitral valve repair, our study extends this research by confirming these findings specifically in patients undergoing isolated MVR.
Integral in this patient selection is the identification of patient- and hospital-specific predictive factors contributing to worse survival after mitral valve operations.7,11 In their investigation involving 1256 patients undergoing PROM qualifying procedures, Ben-David and associates7 identified several risk factors associated with diminished long-term survival. These included age over 65 years (HR, 1.6; 95% CI, 1.1–2.3; P = .02), preoperative dialysis (HR, 2.5; 95% CI, 1.2–4.9; P = .01), diabetes (HR, 1.5; 95% CI, 1.1–2.0; P = .02), and increased PROM. Notably, their study encompassed all PROM qualifying procedures at a single hospital, with MVR and mitral valve repair cases constituting only 10%–14% of all procedures. Our study further refines these findings by focusing on the highest-risk STS procedure—isolated MVR—at a high-volume mitral referral center. Our results reaffirm the association of factors like advanced age and preoperative dialysis with worse long-term survival, while also identifying additional predictive variables such as hypertension, tobacco use, previous myocardial infarction, and urgent operative status. Such predictors could aid in the identification of patients at risk for increased morbidity and mortality.
Interestingly, a strong association between higher hospital volume of mitral valve surgeries and lower adjusted operative morbidity and mortality has been established.12–14 For example, Gammie and colleagues12 reported a robust, multicenter retrospective analysis of elective mitral operations conducted between 2000 and 2003. After adjustment for preoperative clinical risk, rates of operative mortality decreased with increasing mitral surgery volume, specifically with hospital volume of greater than 140 mitral surgeries per year (odds ratio, 0.48; 95% CI, 0.28–0.82; P = .0043). This volume-mortality relationship remained significant even when accounting for the proportion of MVR vs mitral valve repair procedures. In our study, we observed less mortality than expected in our low and high risk patients, at 30 days. Given the established correlation between hospital volume and reduced mortality in mitral surgery, it is possible that PROM calculations may overestimate mortality risks for patients undergoing isolated MVR at high-volume centers. At 30 days, notable differences in survival probability existed among our defined PROM categories, although they were most pronounced in high/extreme risk patients. Moreover, the high/extreme risk cohort had dismal long-term predicted survival (43.58% at 5 years), even in a high-volume mitral center, suggestive that surgical alternatives may have been appropriate.
This study should be interpreted in the context of key limitations. First, the authors recognize that the STS Risk Calculator is an evolving algorithm, with potential fluctuations in accuracy and variable composition throughout the study period. Modifiers were not used in this analysis. Values for PROM group stratification were selected based upon establishment of equal cohort size, with previous literature providing clues as to where clinically significant boundaries may exist.15 Although well-defined long-term survival differences were established, further detailed stratification could allow for more specific predicted survival. Additionally, this work only includes patients who were offered and received an operation. Those who were turned down as poor surgical candidates were not captured in our model. Further work could incorporate several cardiac surgery risk models for an accurate risk assessment in isolated MVR. Additional studies could also attempt longer survival analysis, especially with a more diverse patient population with longer patient follow-up, as our study consisted of a mean follow-up time of 2.26 years and was powered enough to provide statistical significance to only 5 years. Finally, it is important to recognize that the generalizability of our findings may only extend to other high-volume mitral valve reference centers, as volume and complexity of disease may influence outcomes, risk stratification, and surgical decision-making.
In conclusion, these findings suggest that the STS Risk Calculator short-term PROM correlates with 5-year mortality in isolated mitral valve replacement. Several patient- and hospital-specific risk factors were identified and associated with worse long-term survival. These findings collectively offer more refined risk stratification criteria to help identify patients who are candidates for isolated mitral valve replacement and those who may be better suited pursuing alternative therapies.
Supplementary Material
The Supplemental Material can be viewed in the online version of this article [https://doi.org/10.1016/j.athoracsur.2025.07.031] on http://www.annalsthoracicsurgery.org.
FUNDING SOURCES
Matthew Kazaleh reports financial support was provided by National Institute of Health (Surgeon Scientist Training in Cardiac Disease grant, 1T32HL166113-01).
Footnotes
DISCLOSURES
Gorav Ailawadi reports a relationship with Edwards Lifesciences Corporation that includes: consulting or advisory; with Medtronic that includes: consulting or advisory; with Abbott that includes: consulting or advisory; with Johnson & Johnson MedTech that includes: consulting or advisory; with Philips that includes: consulting or advisory; with Arthrex that includes: consulting or advisory; with Anteris Technologies Ltd that includes: consulting or advisory; with W. L. Gore & Associates Inc that includes: consulting or advisory. The other authors have no conflicts of interest to disclose.
Presented as an ePoster at the Seventy-first Annual Meeting of the Southern Thoracic Surgical Association, Austin, TX, Nov 7–10, 2024.
REFERENCES
- 1.Lazam S, Vanoverschelde JL, Tribouilloy C, et al. Twenty-year outcome after mitral repair versus replacement for severe degenerative mitral regurgitation: analysis of a large, prospective, multicenter, international registry. Circulation. 2017;135:410–422. 10.1161/CIRCULATIONAHA.116.023340 [DOI] [PubMed] [Google Scholar]
- 2.Mick SL, Keshavamurthy S, Gillinov AM. Mitral valve repair versus replacement. Ann Cardiothorac Surg. 2015;4:230–237. 10.3978/j.issn.2225-319X.2015.03.01 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.The Society of Thoracic Surgeons. STS risk calculator (version 2.0.5). https://www.sts.org/resources/acsd-operative-risk-calculator
- 4.O’Brien SM, Shahian DM, Filardo G, et al. The Society of Thoracic Surgeons 2008 cardiac surgery risk models: part 2—isolated valve surgery. Ann Thorac Surg. 2009;88(1 suppl):S23–S42. 10.1016/j.athoracsur.2009.05.056 [DOI] [PubMed] [Google Scholar]
- 5.Sinha S, Dimagli A, Dixon L, et al. Systematic review and meta-analysis of mortality risk prediction models in adult cardiac surgery. Interact Cardiovasc Thorac Surg. 2021;33:673–686. 10.1093/icvts/ivab151 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Puskas JD, Kilgo PD, Thourani VH, et al. The Society of Thoracic Surgeons 30-day predicted risk of mortality score also predicts long-term survival. Ann Thorac Surg. 2012;93:26–33 [discussion: 33–35]. 10.1016/j.athoracsur.2011.07.086 [DOI] [PubMed] [Google Scholar]
- 7.Ben-David EI, Blumenfeld O, Shapira-Daniels A, Shapira OM. Validation of the Society of Thoracic Surgeons predicted risk of mortality score for long-term survival after cardiac surgery in Israel. J Cardiothorac Surg. 2022;17:68. 10.1186/s13019-022-01809-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Michigan Department of Health and Human Services. Michigan Death Index, 2010–2023. Accessed November 2024. Published 2024. https://www.michigan.gov/mdhhs/
- 9.Zahr F, Song HK, Chadderdon SM, et al. 30-day outcomes following transfemoral transseptal transcatheter mitral valve replacement: intrepid TMVR early feasibility study results. JACC Cardiovasc Interv. 2022;15:80–89. 10.1016/j.jcin.2021.10.018 [DOI] [PubMed] [Google Scholar]
- 10.Simard T, Lloyd J, Crestanello J, et al. Five-year outcomes of transcatheter mitral valve implantation and redo surgery for mitral prosthesis degeneration. Catheter Cardiovasc Interv. 2022;99:1659–1665. 10.1002/ccd.30059 [DOI] [PubMed] [Google Scholar]
- 11.Mori M, Waldron C, Ragnarsson S, Krane M, Geirsson A. The high-risk features among patients undergoing mitral valve operation for ischemic mitral regurgitation: the 3-strike score. JTCVS Open. 2024;18:52–63. 10.1016/j.xjon.2024.02.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Gammie JS, O’Brien SM, Griffith BP, Ferguson TB, Peterson ED. Influence of hospital procedural volume on care process and mortality for patients undergoing elective surgery for mitral regurgitation. Circulation. 2007;115:881–887. 10.1161/CIRCULATIONAHA.106.634436 [DOI] [PubMed] [Google Scholar]
- 13.Akmaz B, van Kuijk SMJ, Sardari Nia P. Association between individual surgeon volume and outcome in mitral valve surgery: a systematic review. J Thorac Dis. 2021;13:4500–4510. 10.21037/jtd-21-578 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Badhwar V, Vemulapalli S, Mack MA, et al. Volume-outcome association of mitral valve surgery in the United States. JAMA Cardiol. 2020;5:1092–1101. 10.1001/jamacardio.2020.2221 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Serruys PW, Modolo R, Reardon M, et al. One-year outcomes of patients with severe aortic stenosis and an STS PROM of less than three percent in the SURTAVI trial. EuroIntervention. 2018;14:877–883. 10.4244/EIJ-D-18-00460 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.