Abstract
Background: When mitral valve (MV) surgery is indicated, repair is preferred over replacement; however, this preference is not supported by evidence from clinical trials. Furthermore, the benefits of MV repair may not be universal for all etiologies of MV disease.
Methods and Results: This study identified a total of 18,428 patients who underwent MV repair (n=4,817) or MV replacement (n=13,611) during 2001–2018 from Taiwan’s National Health Insurance Research Database. These patients were classified into 4 etiologies: infective endocarditis (IE, n=2,678), rheumatic heart disease (RHD, n=4,524), ischemic mitral regurgitation (IMR, n=3,893), and degenerative mitral regurgitation (DMR, n=7,333). After propensity matching, all-cause mortality during follow-up was lower among patients receiving MV repair than among patients receiving MV replacement in the IE, IMR, and DMR groups (hazard ratio [HR]=0.72, 95% confidence interval [CI]: 0.55–0.93; HR=0.82, 95% CI: 0.73–0.92; and HR 0.73, 95% CI: 0.64–0.84, respectively). However, in the RHD group, the MV reoperation rate was higher after MV repair than after MV replacement (subdistribution HR=1.91, 95% CI: 1.02–3.55).
Conclusions: In comparison with MV replacement, MV repair was associated with a lower late mortality in patients with IE, IMR, and DMR, and a higher risk of reoperation in patients with RHD.
Mitral valve (MV) disease is the most common form of valvular heart disease in the United States.1 In severe cases, mitral regurgitation (MR) can be treated with surgical intervention to improve symptoms and prevent heart failure.2 When MV surgery is indicated, MV repair is recommended in preference to MV replacement if a successful and durable repair is possible.3,4 However, the evidence supporting this preference remains weak due to the lack of clinical trials.2–4 The relevant studies either had a small sample size, lacked long-term follow-up, or were relatively outdated. Furthermore, the benefits of MV repair might not be universal for all etiologies of MV disease4 (including degenerative MR [DMR], ischemic MR [IMR], rheumatic heart disease [RHD], and infective endocarditis [IE]).
Accordingly, we used data from a nationwide database with a maximum 18-year complete follow-up to compare the long-term outcomes of MV repair and replacement among the 4 etiologies. We used propensity score matching (PSM) and inverse probability of treatment weighting (IPTW) to minimize the baseline differences between patients who underwent MV repair and those who underwent MV replacement.
Methods
Data Source
In this retrospective nationwide cohort study, we used data from Taiwan’s National Health Insurance Research Database (NHIRD). The NHIRD contains medical and reimbursement records from Taiwan’s National Health Insurance (NHI) program. Taiwan’s universal obligatory NHI program was launched in March 1995 and covers >99.8% of the 24 million residents in Taiwan. The data stored in the NHIRD represent the real-world conditions in Taiwan (Supplementary Text 1).
All personal data in the NHIRD are deidentified and anonymized; therefore, individual informed consent was not required for the present study. This study was approved by the Institutional Review Board of Chang Gung Medical Foundation (number: 202102032B0; date of approval: December 8, 2021) and by the NHI Administration (the data holder).
Study Population
A total of 20,915 patients who received surgical MV repair or MV replacement between January 1, 2001, and December 31, 2018, were identified from the NHIRD (Figure 1A). If a patient underwent ≥2 MV surgeries during this period, the first MV surgery was selected as the index surgery. Whether the MV repair or replacement was performed was verified using a combination of NHI reimbursement codes, International Classification of Diseases (ICD; Ninth or 10th Revision), Clinical Modification procedural codes, and NHI supply codes to ascertain the surgery type, valve location, and valve type, respectively. This approach to identify MV repair and replacement has been adopted and validated in previous studies using data from the NHIRD.5,6
Patients were excluded from our analysis if their demographic data (e.g., age or sex) were missing (n=118), if they were aged <20 years (n=401), if they received both MV repair and MV replacement during the index hospitalization (n=136), if they had received valve surgery prior to the study start date (n=686), if they received multiple valve replacements using a combination of bioprosthetic and mechanical valves (n=106), or if they had MV disease etiology other than IE, RHD, IMR, or DMR (n=1,040). After exclusion, 18,428 patients were enrolled into the study cohort, comprising 4,817 patients who underwent MV repair and 13,611 patients who underwent MV replacement.
MV Disease Etiology
We used a specific algorithm to determine the etiology of each patient’s MV disease. First, if the patient was assigned an inpatient diagnosis of IE, RHD, IMR, or DMR during the index hospitalization, they were classified accordingly. If a patient was assigned ≥2 of the 4 diagnoses during the index hospitalization, we reclassified them with a single diagnosis, and the priority of classification was: (1) IE; (2) RHD; (3) IMR; and (4) DMR. If a patient had not been assigned 1 of the 4 diagnoses during the index hospitalization, we referred to their previous inpatient diagnoses, if available. If 0 of the 4 diagnoses were assigned during either the index hospitalization or previous hospitalizations, the patient was excluded from the analysis. We used data from the Chang Gung Memorial Hospital (CGMH) system as a comparator to validate the algorithm, and found that the accuracy rates were 93.2%, 94.9%, 87.5%, and 88.4% for the IE, RHD, IMR, and DMR groups, respectively (Supplementary Text 2).
The techniques used for MV repair varies according to the etiologies. The main pathologic changes of DMR are chordae elongation or rupture, and leaflet enlargement and prolapse. The repair methods for DMR include resection of the excessive leaflet, chordae transfer and artificial chord. In addition, mitral annuloplasty is also performed to restore and maintain the normal annular configuration. In IE surgery, the key point is to completely resect all infected tissue. Patch augmentation can be used to repair the defect.7,8 RHD causes commissural fusion, leaflet and chordae thickening, stiffness, and even calcification. Repair methods for RHD include commissurotomy, leaflet decalcification, and leaflet augmentation. However, MV repair for RHD is associated with a relatively low successful rate and high recurrence rate. Ischemic MR is the result of dilatation of the left ventricle and mitral annulus. Downsizing ring annuloplasty is a commonly used method.9
Covariates and Outcomes
The covariates in this study are listed in Supplementary Tables 1–4. For patients classified into the IE group, status of IE (active or old) and history of drug abuse were identified. We defined active IE as IE that was newly diagnosed during the index hospitalization, whereas old IE referred to inpatient diagnosis of IE before the index hospitalization.6,10 For patients classified into the RHD group, previous percutaneous transvenous mitral commissurotomy (PTMC) and type of MV disorder (MR or mitral stenosis [MS]) were also included in the analysis.11,12
The primary outcome for this study was all-cause mortality during follow-up. The Taiwan Death Registry in the Health and Welfare Data Science Center database contains information on the date, place, and cause of death. Secondary outcomes for this study were MV reoperation, IE during follow-up, stroke, major bleeding, re-admission due to heart failure, all-cause readmission, major adverse cardiac and cerebrovascular events (MACCEs, defined as any one of the following: all-cause mortality, stroke, re-admission due to heart failure, and MV reoperation) and composite valve complications (defined as any one of the following: IE, stroke, and major bleeding). We identified the occurrence of complications (i.e., stroke and major bleeding) during follow-up by using the principal diagnosis from the emergency department or hospitalization records. We identified MV reoperation using the same criteria as the index MV surgeries. We identified all-cause re-admission using inpatient reimbursements, and we identified re-admission due to heart failure using the principal inpatient diagnosis. The ICD codes are listed in Supplementary Table 5.
Statistical Analysis
PSM was used to match each patient who underwent MV repair surgery with one patient who underwent MV replacement surgery to reduce possible selection bias when comparing outcomes. PSM was conducted separately for each of the 4 etiologies. The propensity score was the predicted probability that a patient would receive MV repair surgery given certain covariates, and this probability was calculated using a multivariate logistic regression model without considering interactions among the covariates. All of the covariates (listed in Table 1) were included in the regression model to calculate the propensity score; however, the follow-up year was replaced with the discharge date of the index hospitalization. After the propensity score was obtained, matching was processed using a greedy nearest neighbor algorithm without replacement, within a caliper of 0.2-fold the standard deviation of the logit of the propensity score and with random matching order. Matching quality was checked by using the absolute value of the standardized difference (STD) between the groups, with <0.1 indicating a negligible difference.
Table 1.
Baseline Characteristics of Patients Who Received MV Repair vs. Replacement Before and After Matching
| Variables |
Before matching |
After matching |
Repair
(n=4,817) |
Replacement
(n=13,611) |
STD |
Repair
(n=3,474) |
Replacement
(n=3,474) |
STD |
| Etiology |
| IE |
487 (10.1) |
2,191 (16.1) |
−0.18 |
467 (13.4) |
467 (13.4) |
<0.01 |
| RHD |
391 (8.1) |
4,133 (30.4) |
−0.59 |
372 (10.7) |
372 (10.7) |
<0.01 |
| IMR |
1,372 (28.5) |
2,521 (18.5) |
0.24 |
1,016 (29.2) |
1,016 (29.2) |
<0.01 |
| DMR |
2,567 (53.3) |
4,766 (35.0) |
0.37 |
1,619 (46.6) |
1,619 (46.6) |
<0.01 |
| Emergency surgery |
957 (16.6) |
3,377 (19.9) |
−0.09 |
764 (18.0) |
794 (18.6) |
−0.01 |
| Year of the index surgery |
| 2001–2003 |
434 (9.0) |
1,787 (13.1) |
−0.13 |
343 (9.9) |
355 (10.2) |
−0.01 |
| 2004–2006 |
551 (11.4) |
2,109 (15.5) |
−0.12 |
454 (13.1) |
423 (12.2) |
0.03 |
| 2007–2009 |
628 (13.0) |
2,490 (18.3) |
−0.15 |
482 (13.9) |
469 (13.5) |
0.01 |
| 2010–2012 |
781 (16.2) |
2,428 (17.8) |
−0.04 |
590 (17.0) |
613 (17.6) |
−0.02 |
| 2013–2015 |
1,034 (21.5) |
2,552 (18.7) |
0.07 |
706 (20.3) |
715 (20.6) |
−0.01 |
| 2016–2018 |
1,389 (28.8) |
2,245 (16.5) |
0.30 |
899 (25.9) |
899 (25.9) |
<0.01 |
| Age (years) |
58.9±14.2 |
60.8±13.4 |
−0.14 |
60.1±14.0 |
60.1±14.1 |
<0.01 |
| Age group (years) |
| <50 |
1,175 (24.4) |
2,728 (20.0) |
0.10 |
771 (22.2) |
745 (21.4) |
0.02 |
| 50–64 |
1,787 (37.1) |
5,054 (37.1) |
<0.01 |
1,272 (36.6) |
1,290 (37.1) |
−0.01 |
| 65–79 |
1,613 (33.5) |
4,983 (36.6) |
−0.07 |
1,238 (35.6) |
1,220 (35.1) |
0.01 |
| ≥80 |
242 (5.0) |
846 (6.2) |
−0.05 |
196 (5.6) |
222 (6.4) |
−0.03 |
| Male sex |
3,106 (64.5) |
7,016 (51.5) |
0.26 |
2,165 (62.3) |
2,170 (62.5) |
<0.01 |
| Urbanization level |
| Low |
608 (12.6) |
1,970 (14.5) |
−0.05 |
438 (12.6) |
464 (13.4) |
−0.02 |
| Moderate |
1,026 (21.3) |
3,436 (25.2) |
−0.09 |
791 (22.8) |
747 (21.5) |
0.03 |
| High |
2,151 (44.7) |
5,173 (38.0) |
0.14 |
1,498 (43.1) |
1,507 (43.4) |
−0.01 |
| Very high |
1,032 (21.4) |
3,032 (22.3) |
−0.02 |
747 (21.5) |
756 (21.8) |
−0.01 |
| Monthly income |
| Tertile 1 |
1,387 (28.8) |
4,585 (33.7) |
−0.11 |
1,071 (30.8) |
1,099 (31.6) |
−0.02 |
| Tertile 2 |
1,669 (34.6) |
5,094 (37.4) |
−0.06 |
1,238 (35.6) |
1,224 (35.2) |
0.01 |
| Tertile 3 |
1,761 (36.6) |
3,932 (28.9) |
0.16 |
1,165 (33.5) |
1,151 (33.1) |
0.01 |
| Comorbid conditions |
| Old stroke |
398 (8.3) |
1,549 (11.4) |
−0.10 |
322 (9.3) |
311 (9.0) |
0.01 |
| Diabetes mellitus |
1,136 (23.6) |
3,090 (22.7) |
0.02 |
853 (24.6) |
832 (23.9) |
0.01 |
| Hypertension |
2,646 (54.9) |
6,717 (49.3) |
0.11 |
1,896 (54.6) |
1,908 (54.9) |
−0.01 |
| Heart failure |
1,737 (36.1) |
5,547 (40.8) |
−0.10 |
1,343 (38.7) |
1,330 (38.3) |
0.01 |
| Coronary artery disease |
1,951 (40.5) |
5,419 (39.8) |
0.01 |
1,493 (43.0) |
1,499 (43.1) |
<0.01 |
| Atrial fibrillation |
1,476 (30.6) |
6,599 (48.5) |
−0.37 |
1,180 (34.0) |
1,190 (34.3) |
−0.01 |
| Chronic kidney disease |
915 (19.0) |
2,757 (20.3) |
−0.03 |
722 (20.8) |
748 (21.5) |
−0.02 |
| Dialysis |
210 (4.4) |
591 (4.3) |
<0.01 |
178 (5.1) |
178 (5.1) |
<0.01 |
| Chronic obstructive pulmonary disease |
506 (10.5) |
1,684 (12.4) |
−0.06 |
398 (11.5) |
381 (11.0) |
0.02 |
| Major bleeding history |
638 (13.2) |
2,152 (15.8) |
−0.07 |
499 (14.4) |
510 (14.7) |
−0.01 |
| Gastrointestinal bleeding history |
539 (11.2) |
1,777 (13.1) |
−0.06 |
419 (12.1) |
435 (12.5) |
−0.01 |
| Myocardial infarction history |
412 (8.6) |
914 (6.7) |
0.07 |
302 (8.7) |
311 (9.0) |
−0.01 |
| Hospital accreditation level |
| Medical center (teaching hospital) |
3,982 (82.7) |
9,201 (67.6) |
0.35 |
2,734 (78.7) |
2,704 (77.8) |
0.02 |
| Regional / district hospital |
835 (17.3) |
4,410 (32.4) |
−0.35 |
740 (21.3) |
770 (22.2) |
−0.02 |
| Cumulative hospital volume of valve surgery during 2001–2018 |
| 1st quartile |
624 (13.0) |
3,956 (29.1) |
−0.51 |
374 (10.8) |
353 (10.2) |
0.02 |
| 2nd quartile |
958 (19.9) |
3,299 (24.2) |
−0.33 |
694 (20.0) |
683 (19.7) |
0.01 |
| 3rd quartile |
1,289 (26.8) |
3,186 (23.4) |
0.07 |
1,196 (34.4) |
1,261 (36.3) |
−0.04 |
| 4th quartile |
1,946 (40.4) |
3,170 (23.3) |
0.67 |
1,210 (34.8) |
1,177 (33.9) |
0.02 |
| Cumulative hospital volume of MV repair surgery during 2001–2018 |
| 1st quartile |
383 (8.0) |
3,647 (26.8) |
−0.40 |
569 (16.4) |
565 (16.3) |
<0.01 |
| 2nd quartile |
760 (15.8) |
4,005 (29.4) |
−0.11 |
789 (22.7) |
797 (22.9) |
−0.01 |
| 3rd quartile |
1,477 (30.7) |
3,739 (27.5) |
0.08 |
882 (25.4) |
874 (25.2) |
0.01 |
| 4th quartile |
2,197 (45.6) |
2,220 (16.3) |
0.37 |
1,234 (35.5) |
1,238 (35.6) |
<0.01 |
| Additional valve surgery |
| AV repair |
37 (0.8) |
82 (0.6) |
0.02 |
26 (0.7) |
30 (0.9) |
−0.01 |
| AV replacement with mechanical prosthesis |
325 (6.7) |
1,545 (11.4) |
−0.16 |
292 (8.4) |
310 (8.9) |
−0.02 |
| AV replacement with bioprosthesis |
464 (9.6) |
1,014 (7.4) |
0.08 |
364 (10.5) |
380 (10.9) |
−0.01 |
| Tricuspid valve repair or replacement |
448 (9.3) |
2,425 (17.8) |
−0.25 |
413 (11.9) |
451 (13.0) |
−0.03 |
| Additional surgery |
| Coronary artery bypass surgery |
1,293 (26.8) |
2,370 (17.4) |
0.23 |
953 (27.4) |
941 (27.1) |
0.01 |
| Maze |
854 (17.7) |
2,708 (19.9) |
−0.06 |
642 (18.5) |
630 (18.1) |
0.01 |
| Aorta surgery |
104 (2.2) |
180 (1.3) |
0.06 |
70 (2.0) |
74 (2.1) |
−0.01 |
| Follow-up year |
5.3±4.6 |
5.9±4.9 |
−0.12 |
5.5±4.7 |
5.2±4.7 |
0.06 |
AV, aortic valve; DMR, degenerative mitral regurgitation; IE, infective endocarditis; IMR, ischemic mitral regurgitation; MR, mitral regurgitation; MS, mitral stenosis; MV, mitral valve; RHD, rheumatic heart disease; STD, standardized difference. Data are presented as frequency (percentage) or mean±standard deviation.
The risks of fatal (e.g., all-cause mortality) and non-fatal (e.g., MV reoperation) time-to-event outcomes between the MV repair and replacement groups were compared using the Cox proportional hazard model and Fine–Gray subdistribution hazard model, respectively. The details of the statistical analysis are provided in Supplementary Text 3.
Results
Patient Characteristics
The temporal trend of etiology among MV surgeries revealed that the number of MV surgeries due to IE, IMR, and DMR increased over time, whereas the number of MV surgeries due to RHD decreased over time (Figure 1B). The repair rate in the IE, IMR, and DMR groups increased over time, whereas the repair rate in the RHD group remained stable (Figure 1C). The completeness of follow-up in the present study was >99%.
Before matching, the etiology of MV disease was as follows: IE in 2,678 patients, RHD in 4,524 patients, IMR in 3,893 patients, and DMR in 7,333 patients. The proportions of repair surgeries in each group were 18.2% in the IE group, 8.6% in the RHD group, 35.2% in the IMR group, and 35.0% in the DMR group. The overall proportion of MV repair was 26.1%, but this proportion increased to 38.2% during 2016–2018 (Table 1). The proportion of male patients was higher in the repair group than in the replacement group (64.5% vs. 51.5%, STD=0.26). Regarding comorbidities, patients who underwent MV replacement were more likely to have atrial fibrillation than those who underwent MV repair. Regarding the hospital accreditation level, the majority of MV surgeries were conducted at medical centers; however, a higher proportion of repairs (82.7%) relative to replacements (67.6%, STD=0.35) was conducted at medical centers. Regarding cumulative volume of valve surgery and MV repair surgery at each hospital, preference for MV repair was positively correlated with the cumulative volume of surgeries performed. Among patients, the mean follow-up duration before matching was 5.3±4.6 and 5.9±4.9 years in the repair and replacement groups, respectively.
In the RHD group, patients who underwent MV repair were more likely to have MR, whereas patients who underwent MV replacement were more likely to have MS (proportion of MR in the repair group, 63.2%; in the replacement group, 33.0%, STD=0.63; proportion of MS in the replacement group, 57.1%; in the repair group, 29.2%, STD=−0.59). After matching, the differences between the repair and replacement groups were negligible in the overall cohort (Table 1) and in each etiology group (Supplementary Tables 1–4). Details about the numbers of different types of prosthetic valves implanted and the medications prescribed during follow-up are provided in Supplementary Tables 6 and 7, respectively.
Outcomes
After matching, the risks of late all-cause mortality (hazard ratio [HR]=0.80, 95% confidence interval [CI]: 0.74–0.87), MACCEs (HR=0.85, 95% CI: 0.79–0.91), and composite valve complications (subdistribution HR [SHR]=0.76, 95% CI: 0.67–0.85) were significantly lower in the repair group than in the replacement group in the overall cohort (Table 2, Figure 2).
Table 2.
Late Outcomes of Patients Who Received MV Repair vs. Replacement in the Propensity Score-Matched Cohort
| Outcome / Etiology |
Total
(n=6,948) |
Repair
(n=3,474) |
Replacement
(n=3,474) |
HR or SHR (95% CI)
of repair |
P value |
P for
interaction |
| All-cause mortality |
|
|
|
|
|
0.097 |
| IE |
233 (24.9) |
101 (21.6) |
132 (28.3) |
0.72 (0.55–0.93) |
0.013 |
|
| RHD |
256 (34.4) |
128 (34.4) |
128 (34.4) |
1.02 (0.80–1.30) |
0.903 |
|
| IMR |
1,078 (53.1) |
504 (49.6) |
574 (56.5) |
0.82 (0.73–0.92) |
<0.001 |
|
| DMR |
815 (25.2) |
352 (21.7) |
463 (28.6) |
0.73 (0.64–0.84) |
<0.001 |
|
| The overall cohort |
2,382 (34.3) |
1,085 (31.2) |
1,297 (37.3) |
0.80 (0.74–0.87) |
<0.001 |
|
| MV reoperation |
|
|
|
|
|
0.028 |
| IE |
34 (3.6) |
14 (3.0) |
20 (4.3) |
0.67 (0.35–1.32) |
0.248 |
|
| RHD |
46 (6.2) |
30 (8.1) |
16 (4.3) |
1.91 (1.02–3.55) |
0.042 |
|
| IMR |
28 (1.4) |
10 (1.0) |
18 (1.8) |
0.56 (0.26–1.21) |
0.142 |
|
| DMR |
127 (3.9) |
72 (4.4) |
55 (3.4) |
1.36 (0.96–1.92) |
0.081 |
|
| The overall cohort |
235 (3.4) |
126 (3.6) |
109 (3.1) |
1.18 (0.92–1.52) |
0.200 |
|
| IE |
|
|
|
|
|
0.740 |
| IE |
42 (4.5) |
19 (4.1) |
23 (4.9) |
0.82 (0.45–1.50) |
0.520 |
|
| RHD |
18 (2.4) |
8 (2.2) |
10 (2.7) |
0.80 (0.31–2.04) |
0.637 |
|
| IMR |
23 (1.1) |
9 (0.9) |
14 (1.4) |
0.64 (0.28–1.49) |
0.302 |
|
| DMR |
58 (1.8) |
20 (1.2) |
38 (2.3) |
0.53 (0.31–0.91) |
0.020 |
|
| The overall cohort |
141 (2.0) |
56 (1.6) |
85 (2.4) |
0.66 (0.47–0.93) |
0.016 |
|
| Stroke |
|
|
|
|
|
0.532 |
| IE |
57 (6.1) |
24 (5.1) |
33 (7.1) |
0.71 (0.42–1.20) |
0.200 |
|
| RHD |
89 (12.0) |
43 (11.6) |
46 (12.4) |
0.92 (0.61–1.39) |
0.688 |
|
| IMR |
168 (8.3) |
71 (7.0) |
97 (9.5) |
0.72 (0.53–0.98) |
0.035 |
|
| DMR |
266 (8.2) |
104 (6.4) |
162 (10.0) |
0.64 (0.50–0.82) |
<0.001 |
|
| The overall cohort |
580 (8.3) |
242 (7.0) |
338 (9.7) |
0.71 (0.60–0.84) |
<0.001 |
|
| Major bleeding |
|
|
|
|
|
0.568 |
| IE |
67 (7.2) |
29 (6.2) |
38 (8.1) |
0.75 (0.46–1.22) |
0.243 |
|
| RHD |
81 (10.9) |
40 (10.8) |
41 (11.0) |
0.98 (0.62–1.54) |
0.923 |
|
| IMR |
166 (8.2) |
67 (6.6) |
99 (9.7) |
0.67 (0.49–0.92) |
0.012 |
|
| DMR |
248 (7.7) |
104 (6.4) |
144 (8.9) |
0.73 (0.57–0.93) |
0.013 |
|
| The overall cohort |
562 (8.1) |
240 (6.9) |
322 (9.3) |
0.75 (0.63–0.88) |
0.001 |
|
| Re-admission for heart failure |
|
|
|
|
|
0.189 |
| IE |
36 (3.9) |
14 (3.0) |
22 (4.7) |
0.62 (0.31–1.22) |
0.166 |
|
| RHD |
96 (12.9) |
49 (13.2) |
47 (12.6) |
1.05 (0.72–1.54) |
0.801 |
|
| IMR |
307 (15.1) |
163 (16.0) |
144 (14.2) |
1.16 (0.93–1.45) |
0.187 |
|
| DMR |
332 (10.3) |
156 (9.6) |
176 (10.9) |
0.90 (0.72–1.11) |
0.314 |
|
| The overall cohort |
771 (11.1) |
382 (11.0) |
389 (11.2) |
0.99 (0.87–1.14) |
0.937 |
|
| Re-admission due to any cause |
|
|
|
|
|
0.563 |
| IE |
486 (52.0) |
235 (50.3) |
251 (53.7) |
0.94 (0.79–1.12) |
0.495 |
|
| RHD |
496 (66.7) |
244 (65.6) |
252 (67.7) |
0.90 (0.76–1.06) |
0.208 |
|
| IMR |
1,226 (60.3) |
609 (59.9) |
617 (60.7) |
0.996 (0.89–1.11) |
0.948 |
|
| DMR |
1,882 (58.1) |
909 (56.1) |
973 (60.1) |
0.91 (0.83–0.99) |
0.031 |
|
| The overall cohort |
4,090 (58.9) |
1,997 (57.5) |
2,093 (60.2) |
0.94 (0.88–0.996) |
0.038 |
|
| MACCE* |
|
|
|
|
|
0.042 |
| IE |
305 (32.7) |
135 (28.9) |
170 (36.4) |
0.73 (0.58–0.92) |
<0.018 |
|
| RHD |
339 (45.6) |
175 (47.0) |
164 (44.1) |
1.11 (0.90–1.37) |
0.344 |
|
| IMR |
1,226 (60.3) |
581 (57.2) |
645 (63.5) |
0.84 (0.76–0.94) |
<0.012 |
|
| DMR |
1,159 (35.8) |
531 (32.8) |
628 (38.8) |
0.82 (0.73–0.92) |
<0.001 |
|
| The overall cohort |
3,029 (43.6) |
1,422 (40.9) |
1,607 (46.3) |
0.85 (0.79–0.91) |
<0.001 |
|
| Composite valve complications# |
|
|
|
|
|
0.858 |
| IE |
134 (14.3) |
58 (12.4) |
76 (16.3) |
0.74 (0.52–1.04) |
0.084 |
|
| RHD |
163 (21.9) |
75 (20.2) |
88 (23.7) |
0.85 (0.62–1.15) |
0.290 |
|
| IMR |
302 (14.9) |
128 (12.6) |
174 (17.1) |
0.73 (0.58–0.91) |
0.005 |
|
| DMR |
475 (50.9) |
205 (43.9) |
270 (57.8) |
0.75 (0.63–0.90) |
0.002 |
|
| The overall cohort |
1,074 (15.5) |
466 (13.4) |
608 (17.5) |
0.76 (0.67–0.85) |
<0.001 |
|
CI, confidence interval; HR, hazard ratio; MACCE, major adverse cardiac and cerebrovascular events; SHR, subdistribution hazard ratio. Other abbreviations as in Table 1. *Any one of the following: all-cause mortality, stroke, re-admission due to heart failure, and MV reoperation. #Any one of the following: infective endocarditis, stroke, and major bleeding. Data are presented as number (percentage).
Regarding different etiologies, all-cause mortality during follow-up was significantly lower after MV repair than after MV replacement in the IE (HR=0.72, 95% CI: 0.55–0.93), IMR (HR=0.82, 95% CI: 0.73–0.92), and DMR (HR=0.73, 95% CI: 0.64–0.84) groups (Figure 3A,C,D), but it did not differ significantly in the RHD group (Figure 3B). The heterogeneity among the different etiology groups was not statistically significant (P for interaction=0.097; Table 2).
In the RHD group, the risk of MV reoperation was higher after MV repair than after MV replacement (SHR=1.91, 95% CI: 1.02–3.55). This effect was not observed in the IE, IMR, or DMR groups. The heterogeneity among the different etiology groups was significant (P for interaction=0.028; Table 2).
Regarding other late complications, MV repair was associated with lower risks of IE (SHR=0.66, 95% CI: 0.47–0.93), major bleeding (SHR=0.75, 95% CI: 0.63–0.88), all-cause readmission (SHR=0.94, 95% CI: 0.88–0.996), MACCEs (HR=0.85, 95% CI: 0.79–0.91), and composite valve complications (SHR=0.76, 95% CI: 0.67–0.85) in the overall population (Table 2).
Sensitivity and Subgroup Analyses
The sensitivity analysis using IPTW revealed that compared with MV replacement, MV repair was associated with a lower risk of all-cause mortality during follow-up in the overall population and in the IE, IMR, and DMR groups (Supplementary Table 8). In the RHD group, MV repair and MV replacement were associated with similar all-cause mortality during follow-up. The results of the sensitivity analysis were generally consistent with those of matching, which verified the robustness of the matching algorithm.
We divided patients into different subgroups according to age, atrial fibrillation, and cumulative volume of MV repair at the index hospital. In all of the subgroups, the heterogeneity for all-cause mortality (Supplementary Table 9), MACCEs (Supplementary Table 10), and composite valve complications (Supplementary Table 11) was non-significant (P value for interaction >0.05), indicating that the benefit of MV repair did not differ significantly among the subgroups.
In another subgroup analysis, we compared patients treated at high-volume hospitals with those treated at low-volume hospitals according to each of the 4 etiologies (Supplementary Table 12). We found that the heterogeneity between high-volume and low-volume hospitals was non-significant (P value for interaction >0.05), indicating that the benefit of MV repair did not differ significantly between high-volume and low-volume hospitals. We also noted that in the DMR group, the HRs of repair and the 95% CIs were <1 for both low-volume and high-volume hospitals.
Discussion
Principal Findings
The Taiwan NHIRD is a nationwide database containing the claims data of all residents of Taiwan, and it veritably demonstrates the real-world medical situation in Taiwan. Because the NHI program is obligatory and offers comprehensive quality of care with reasonable fees, reoperations and rehospitalizations are detected with minimal loss to follow-up. The large sample size of the NHIRD enables the use of state-of-the-art statistical methods to reduce confounding.
In this study using NHIRD data, MV repair was associated with a lower incidence of late mortality and re-operation than MV replacement in the overall cohort; however, the benefits of MV repair over replacement varied by etiology. For patients with IE, IMR, and DMR, MV repair was associated with a lower late mortality rate than MV replacement. Nevertheless, for patients with RHD, late mortality did not differ significantly between MV repair and replacement; furthermore, in the RHD group, repair was more likely than replacement to result in re-operation.
The overall MV repair rate in the present study appeared lower compared with other studies.13,14 This finding might be attributable to the following factors. First, in comparison to the US or Europe, Taiwan has a higher prevalence of IE and RHD and both etiologies were associated with a lower probability of repair.15 Second, the present study represents the real-world condition in Taiwan, and includes low-volume hospitals, which might have a lower rate of MV repair. However, we also noticed that the repair rate in the DMR and IMR groups increased over time (Figure 1C).
Degenerative Mitral Regurgitation
The American Heart Association guidelines recommend MV repair over MV replacement when the anatomical cause is degenerative disease.4 Two large studies have provided high-quality evidence supporting this preference. Lazam et al analyzed 1,922 patients (repair, n=410; replacement, n=205 after matching) with DMR and a flail leaflet, and they found that the 20-year mortality was lower after repair than after replacement.13 Hannan et al analyzed 2,259 patients (repair, n=450; replacement, n=450 after matching) with primary MR (excluding coronary artery disease, IE, and MS), and they found that 4-year mortality was lower after repair than after replacement.14 The present study demonstrated similar results, with the largest DMR cohort to date (repair, n=1,619; replacement, n=1,619 after matching), supporting the preference of MV repair over MV replacement for DMR.
However, the possibility remains that the better outcomes after MV repair are attributable to surgeon experience and hospital volume and not to the repair procedure, per se. Multiple studies have reported that the proportion of MV repairs to replacements is significantly higher at high-volume hospitals or because of experienced surgeons,14,16–18 and that outcomes are better at such hospitals or with these surgeons.17,18 Although previous studies have enrolled large numbers of patients and have used matching and IPTW, selection bias of surgeons and hospitals were not included in the statistical adjustments.13,14 Hence, we included the cumulative volume of valve surgery and the cumulative volume of MV repair surgery for each index hospital as confounders in the matching and IPTW analyses. After matching, the differences in the cumulative volumes were all negligible (STD <0.1). We demonstrated that under balanced baseline characteristics, including hospital volume, and MV repair for DMR was associated with lower rates of mortality, IE, stroke, major bleeding, all-cause readmission, MACCEs, and composite valve complications compared with MV replacement. Furthermore, subgroup analyses revealed that HRs and 95% CIs for MV repair were <1 at both high-volume and low-volume hospitals, indicating that the benefits of repair over replacement are unrelated to the hospital volume (Supplementary Table 10). Previous studies have demonstrated that higher MV surgery volume leads to better outcomes, emphasizing the role of surgeon experience in beneficial surgical outcomes.17,18 By contrast, the findings of the present study suggest that, even if MV surgery is performed at a less-experienced hospital, MV repair for DMR should still be encouraged.
Infective Endocarditis
In patients with IE, better long-term outcomes for MV repair than for MV replacement have been demonstrated by single-center studies,19–21 meta-analyses,22,23 and large database studies with PSM.6,15 Similarly, we found significantly lower rates of late mortality and MACCEs after MV repair for patients with IE. We believe that for patients requiring MV surgery due to IE, a repair-first strategy is reasonable.
Rheumatic Heart Disease
In patients with RHD, MV repair was associated with lower risks of heart failure,24 hemorrhagic events, and thromboembolic events25 compared with MV replacement. Yakub et al reported comparable durability of MV repair for RHD and DMR.26 For RHD, when MV surgery was performed by highly experienced surgeons (>500 MV repairs and 200 rheumatic MV repairs per surgeon), MV repair was associated with a lower late mortality rate compared with MV replacement.27,28 Remenyi et al suggested that for RHD, MV repair should be attempted whenever possible.29 Despite the findings of Yakub et al, the durability of the MV repair for RHD remains a major concern. Meta-analyses and a nationwide population-based study found that MV repair was associated with a significantly higher risk of reoperation compared with MV replacement.12,30,31
The present study found that for patients with RHD, the risks of late mortality and valve-related complications for MV repair and replacement were not significantly different, but between the 2 options, MV repair was significantly more likely to result in MV reoperation. Accordingly, when MV surgery is required for patients with RHD, the choice of the procedure should be individualized and based on age, patient preference and compliance, degree of valve calcification and destruction, regional socioeconomic and sanitization status, and the surgeon’s level of experience.
Ischemic Mitral Regurgitation
The present study found that for patients with IMR, MV repair was associated with lower risks of all-cause mortality, stroke, major bleeding, MACCEs, and valve-related complications compared with MV replacement. Using a nationwide database and PSM, Deja et al found a lower risk of late mortality after MV repair than after replacement; however, they also found that the survival curves of MV repair and replacement converged with time, and they postulated that a higher prevalence of persistent or recurrent MR after MV repair might contribute to late mortality.32 Previous studies have found that MV repair for IMR was associated with numerous poor outcomes when compared with MV replacement, including higher rates of persistent or recurrent MR,33–36 serious adverse events related to heart failure, cardiovascular rehospitalization 2 years postoperatively,34 all-cause reoperation, and valve-related reoperation.37
Although MV repair prevents prosthetic valve-related complications,38 the effectiveness and durability of MV repair for IMR is questionable. However, many of the previous studies used mitral annuloplasty as the only repair procedure, but various subvalvular procedures have been developed as adjunct therapy (of mitral annuloplasty) to improve the outcomes of MV repair, and they have been reported to reduce MR persistence and recurrence while restoring left ventricular function.39–41 The benefits and the appropriate subvalvular techniques for different left ventricular geometry require further investigation.42
Study Limitations
The present study has several limitations. First, the present study is retrospective in nature, and selection bias is inevitable. Although we used PSM and IPTW to balance baseline characteristics, including the volume of valve surgery and MV repair surgery at each hospital, we could only adjust for known confounders, and unknown confounders might still exist. Outside of a randomized control trial, truly balancing both groups of patients in all aspects is impossible. Second, the diagnoses of MV etiologies were based on ICD codes, and coding errors are possible. Hence, we used the electronic medical records from the CGMH system as a comparator for validation and found that the accuracy rates of our diagnoses were between 87.5% and 94.9%. Third, the NHIRD lacks some important clinical information, including echocardiogram and laboratory reports; therefore, calculation of mortality risk using the European System for Cardiac Operative Risk Evaluation or the Society of Thoracic Surgeons Predicted Risk of Mortality score is not possible. Instead, we used the diagnoses of comorbidities, including heart failure, renal failure, and diabetes mellitus, to evaluate and balance operative risks between the 2 groups.
Conclusions
In the present study using Taiwan’s NHIRD, we found that compared with MV replacement, MV repair significantly reduced valve-related complications, including IE, stroke, and major bleeding. Relative to those who received MV replacement, late mortality was lower after MV repair in patients with DMR, IE, and IMR. Nevertheless, MV repair resulted in a high risk of reoperation in patients with RHD. On the basis of these findings, we believe that the etiology of MV disease is essential for the outcomes of MV repair vs. replacement. Accordingly, when MV surgery is required, the choice of repair or replacement should be individualized and based on the etiology of MV disease.
Acknowledgments
This study used data from the NHIRD provided by the NHI Administration, Ministry of Health and Welfare of Taiwan and managed by the National Health Research Institutes of Taiwan. However, the interpretations and conclusions contained in this paper represent only those of the authors. The authors thank the Maintenance Project of the Center for Big Data Analytics and Statistics at CGMH, Linkou Medical Center, Taoyuan City, for offering statistical consultation and data analysis. The authors also thank Alfred Hsing-Fen Lin and Zoe Ya-Jhu Syu for their assistance with statistical analysis.
Disclosures
The authors declare no conflicts of interest.
Sources of Funding
This work was supported by a grant from CGMH, Taiwan (CORPG3N0281, CORPG3M0371, CMRPG3L0101, CMRPG3L0102, CMRPG3L0103, CFRPG3M0011, and BMRPD95 [Shao-Wei Chen]). This work was also supported by the Ministry of Science and Technology grant (MOST-112-2314-B-182A-107 [Shao-Wei Chen]).
IRB Information
This study was approved by the Institutional Review Board of Chang Gung Medical Foundation (number: 202102032B0; date of approval: December 8, 2021).
Data Availability Statement
The dataset used in this study is held by the Taiwan Ministry of Health and Welfare (MOHW). To apply for the access of the dataset, researchers can visit the website of the National health Informatics Project of the MOHW (https://dep.mohw.gove.tw/dos/np-2497-133.html) or contact the National Health Insurance Research Database (nhird@nhri.edu.tw).
Supplementary Files
Please find supplementary file(s);
https://doi.org/10.1253/circj.CJ-23-0640
References
- 1.
Nkomo VT, Gardin JM, Skelton TN, Gottdiener JS, Scott CG, Enriquez-Sarano M. Burden of valvular heart diseases: A population-based study. Lancet 2006; 368: 1005–1011, doi:10.1016/s0140-6736(06)69208-8.
- 2.
Enriquez-Sarano M, Akins CW, Vahanian A. Mitral regurgitation. Lancet 2009; 373: 1382–1394, doi:10.1016/s0140-6736(09)60692-9.
- 3.
Vahanian A, Beyersdorf F, Praz F, Milojevic M, Baldus S, Bauersachs J, et al. 2021 ESC/EACTS Guidelines for the management of valvular heart disease. Eur Heart J 2022; 43: 561–632, doi:10.1093/eurheartj/ehab395.
- 4.
Otto CM, Nishimura RA, Bonow RO, Carabello BA, Erwin JP 3rd, Gentile F, et al. 2020 ACC/AHA Guideline for the management of patients with valvular heart disease: A report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation 2021; 143: e72–e227, doi:10.1161/CIR.0000000000000923.
- 5.
Kostyunin AE, Yuzhalin AE, Rezvova MA, Ovcharenko EA, Glushkova TV, Kutikhin AG. Degeneration of bioprosthetic heart valves: Update 2020. J Am Heart Assoc 2020; 9: e018506, doi:10.1161/JAHA.120.018506.
- 6.
Lee HA, Cheng YT, Wu VC, Chou AH, Chu PH, Tsai FC, et al. Nationwide cohort study of mitral valve repair versus replacement for infective endocarditis. J Thorac Cardiovasc Surg 2018; 156: 1473–1483.e2, doi:10.1016/j.jtcvs.2018.04.064.
- 7.
Fukunaga N, Tamura N. Mitral valve repair with glutaraldehyde-treated autologous pericardium for active infective endocarditis. Gen Thorac Cardiovasc Surg 2022; 70: 683–684, doi:10.1007/s11748-022-01817-3.
- 8.
Kinoshita R, Arai H, Yashima M, Hachimaru T, Fujiwara T, Tamura K, et al. Mid-term results of mitral valve reconstruction with autologous pericardial patch for active infective endocarditis with extensive leaflet destruction. Gen Thorac Cardiovasc Surg 2022; 70: 694–704, doi:10.1007/s11748-022-01776-9.
- 9.
Matsuura K, Kumamaru H, Matsumiya G, Motomura N. Late outcome of coronary artery bypass grafting with or without mitral repair for moderate or moderate-severe ischemic mitral regurgitation. Gen Thorac Cardiovasc Surg 2023; 71: 543–551, doi:10.1007/s11748-023-01925-8.
- 10.
Lee HA, Chou AH, Wu VC, Chan YS, Cheng YT, Chang CH, et al. Nationwide cohort study of tricuspid valve repair versus replacement for infective endocarditis. Eur J Cardiothorac Surg 2021; 59: 878–886, doi:10.1093/ejcts/ezaa390.
- 11.
Chen CY, Chan YH, Wu VC, Liu KS, Cheng YT, Chu PH, et al. Bioprosthetic versus mechanical mitral valve replacements in patients with rheumatic heart disease. J Thorac Cardiovasc Surg 2023; 165: 1050–1060.e8, doi:10.1016/j.jtcvs.2021.03.033.
- 12.
Chen SW, Chen CY, Chien-Chia Wu V, Chou AH, Cheng YT, Chang SH, et al. Mitral valve repair versus replacement in patients with rheumatic heart disease. J Thorac Cardiovasc Surg 2022; 164: 57–67.e11, doi:10.1016/j.jtcvs.2020.07.117.
- 13.
Lazam S, Vanoverschelde JL, Tribouilloy C, Grigioni F, Suri RM, Avierinos JF, 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, doi:10.1161/CIRCULATIONAHA.116.023340.
- 14.
Hannan EL, Samadashvili Z, Smith CR, Lahey SJ, Gold JP, Jordan D, et al. Mitral valve repair versus replacement for patients with preserved left ventricular function without heart failure symptoms. J Thorac Cardiovasc Surg 2019; 157: 1432–1439.e2, doi:10.1016/j.jtcvs.2018.08.091.
- 15.
Toyoda N, Itagaki S, Egorova NN, Tannous H, Anyanwu AC, El-Eshmawi A, et al. Real-world outcomes of surgery for native mitral valve endocarditis. J Thorac Cardiovasc Surg 2017; 154: 1906–1912.e9, doi:10.1016/j.jtcvs.2017.07.077.
- 16.
LaPar DJ, Ailawadi G, Isbell JM, Crosby IK, Kern JA, Rich JB, et al. Mitral valve repair rates correlate with surgeon and institutional experience. J Thorac Cardiovasc Surg 2014; 148: 995–1003; discussion 1003–1004, doi:10.1016/j.jtcvs.2014.06.039.
- 17.
Chikwe J, Toyoda N, Anyanwu AC, Itagaki S, Egorova NN, Boateng P, et al. Relation of mitral valve surgery volume to repair rate, durability, and survival. J Am Coll Cardiol 2017; S0735-1097(17)30677-0, doi:10.1016/j.jacc.2017.02.026.
- 18.
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, doi:10.1161/CIRCULATIONAHA.106.634436.
- 19.
Solari S, De Kerchove L, Tamer S, Aphram G, Baert J, Borsellino S, et al. Active infective mitral valve endocarditis: Is a repair-oriented surgery safe and durable? Eur J Cardiothorac Surg 2019; 55: 256–262, doi:10.1093/ejcts/ezy242.
- 20.
Defauw RJ, Tomsic A, van Brakel TJ, Marsan NA, Klautz RJM, Palmen M. A structured approach to native mitral valve infective endocarditis: Is repair better than replacement? Eur J Cardiothorac Surg 2020; 58: 544–550, doi:10.1093/ejcts/ezaa079.
- 21.
Ruttmann E, Legit C, Poelzl G, Mueller S, Chevtchik O, Cottogni M, et al. Mitral valve repair provides improved outcome over replacement in active infective endocarditis. J Thorac Cardiovasc Surg 2005; 130: 765–771, doi:10.1016/j.jtcvs.2005.03.016.
- 22.
He K, Song J, Luo H, Su H, Liang W, Bian L, et al. Valve replacement or repair in native mitral valve infective endocarditis-Which is better? A meta-analysis and systematic review. J Card Surg 2022; 37: 1004–1015, doi:10.1111/jocs.16227.
- 23.
Feringa HH, Shaw LJ, Poldermans D, Hoeks S, van der Wall EE, Dion RA, et al. Mitral valve repair and replacement in endocarditis: A systematic review of literature. Ann Thorac Surg 2007; 83: 564–570, doi:10.1016/j.athoracsur.2006.09.023.
- 24.
Jiao Y, Luo T, Zhang H, Han J, Li Y, Jia Y, et al. Repair versus replacement of mitral valves in cases of severe rheumatic mitral stenosis: Mid-term clinical outcomes. J Thorac Dis 2019; 11: 3951–3961, doi:10.21037/jtd.2019.08.101.
- 25.
Kim WK, Kim HJ, Kim JB, Jung SH, Choo SJ, Chung CH, et al. Clinical outcomes in 1731 patients undergoing mitral valve surgery for rheumatic valve disease. Heart 2018; 104: 841–848, doi:10.1136/heartjnl-2017-312249.
- 26.
Yakub MA, Dillon J, Krishna Moorthy PS, Pau KK, Nordin MN. Is rheumatic aetiology a predictor of poor outcome in the current era of mitral valve repair?: Contemporary long-term results of mitral valve repair in rheumatic heart disease. Eur J Cardiothorac Surg 2013; 44: 673–681, doi:10.1093/ejcts/ezt093.
- 27.
Fu J, Li Y, Zhang H, Han J, Jiao Y, Du J, et al. Outcomes of mitral valve repair compared with replacement for patients with rheumatic heart disease. J Thorac Cardiovasc Surg 2021; 162: 72–82.e7, doi:10.1016/j.jtcvs.2020.01.053.
- 28.
Yau TM, El-Ghoneimi YA, Armstrong S, Ivanov J, David TE. Mitral valve repair and replacement for rheumatic disease. J Thorac Cardiovasc Surg 2000; 119: 53–60, doi:10.1016/s0022-5223(00)70217-0.
- 29.
Remenyi B, ElGuindy A, Smith SC, Yacoub M, Holmes DR. Valvular aspects of rheumatic heart disease. Lancet 2016; 387: 1335–1346, doi:10.1016/s0140-6736(16)00547-x.
- 30.
Jiang Y, Wang C, Li G, Chen S. Clinical outcomes following surgical mitral valve repair or replacement in patients with rheumatic heart disease: A meta-analysis. Ann Transl Med 2021; 9: 204, doi:10.21037/atm-20-3542.
- 31.
Wang Z, Zhou C, Gu H, Zheng Z, Hu S. Mitral valve repair versus replacement in patients with rheumatic heart disease. J Heart Valve Dis 2013; 22: 333–339.
- 32.
Deja MA, Malinowski M, Widenka K, Stozynski N, Bartus K, Kapelak B, et al. Repair or replacement for secondary mitral regurgitation: Results from Polish National Registry. Ann Thorac Surg 2022; 113: 146–156, doi:10.1016/j.athoracsur.2020.12.059.
- 33.
Magne J, Girerd N, Senechal M, Mathieu P, Dagenais F, Dumesnil JG, et al. Mitral repair versus replacement for ischemic mitral regurgitation: Comparison of short-term and long-term survival. Circulation 2009; 120: S104–S111, doi:10.1161/CIRCULATIONAHA.108.843995.
- 34.
Goldstein D, Moskowitz AJ, Gelijns AC, Ailawadi G, Parides MK, Perrault LP, et al. Two-year outcomes of surgical treatment of severe ischemic mitral regurgitation. N Engl J Med 2016; 374: 344–353, doi:10.1056/NEJMoa1512913.
- 35.
Acker MA, Parides MK, Perrault LP, Moskowitz AJ, Gelijns AC, Voisine P, et al. Mitral-valve repair versus replacement for severe ischemic mitral regurgitation. N Engl J Med 2014; 370: 23–32, doi:10.1056/NEJMoa1312808.
- 36.
Gamal MA, El-Fiky MM, Gamea MM, Ali I. Mitral valve repair versus replacement in severe ischemic mitral regurgitation systematic review and meta-analysis. J Card Surg 2022; 37: 1591–1598, doi:10.1111/jocs.16477.
- 37.
Lorusso R, Gelsomino S, Vizzardi E, D’Aloia A, De Cicco G, Luca F, et al. Mitral valve repair or replacement for ischemic mitral regurgitation? The Italian Study on the Treatment of Ischemic Mitral Regurgitation (ISTIMIR). J Thorac Cardiovasc Surg 2013; 145: 128–139; 137–138, doi:10.1016/j.jtcvs.2012.09.042.
- 38.
Yoshikawa Y, Okada Y, Okita Y, Yaku H, Kobayashi J, Uesugi H, et al. Long-term outcomes of the mosaic mitral porcine bioprosthesis in Japan: Results from the Japan Mosaic Valve Long-Term Multicenter Study. Circ J 2022; 86: 449–457, doi:10.1253/circj.CJ-21-0528.
- 39.
Nappi F, Lusini M, Spadaccio C, Nenna A, Covino E, Acar C, et al. Papillary muscle approximation versus restrictive annuloplasty alone for severe ischemic mitral regurgitation. J Am Coll Cardiol 2016; 67: 2334–2346, doi:10.1016/j.jacc.2016.03.478.
- 40.
Calafiore AM, Refaie R, Iaco AL, Asif M, Al Shurafa HS, Al-Amri H, et al. Chordal cutting in ischemic mitral regurgitation: A propensity-matched study. J Thorac Cardiovasc Surg 2014; 148: 41–46, doi:10.1016/j.jtcvs.2013.07.036.
- 41.
Fattouch K, Castrovinci S, Murana G, Dioguardi P, Guccione F, Nasso G, et al. Papillary muscle relocation and mitral annuloplasty in ischemic mitral valve regurgitation: Midterm results. J Thorac Cardiovasc Surg 2014; 148: 1947–1950, doi:10.1016/j.jtcvs.2014.02.047.
- 42.
Servito T, Elbatarny M, Yanagawa B, Dokollari A, Bisleri G. Surgical repair of ischemic mitral regurgitation: One ring does not fit all. Curr Opin Cardiol 2021; 36: 154–162, doi:10.1097/HCO.0000000000000827.
