Abstract
Objective
Previous randomized controlled trials demonstrated comparable outcomes between posterior leaflet resections and neochord implantation in mitral valve (MV) repair. However, these studies were limited up to 1-year follow-up, and more recent evidence suggested that leaflet resections may offer superior long-term outcomes.
Methods
All patients who underwent MV repair with either resection or neochord implantation for posterior leaflet pathology between October 2011 and July 2024 were included. Propensity-score matching was used.
Results
A total of 457 patients underwent posterior leaflet reconstruction, with 334 (73.1%) requiring leaflet resection (resection group) and 123 (26.9%) receiving neochordae (neochord group). The median [interquartile range] follow-up for survival and reintervention was 5.93 [2.00, 9.09] and 5.48 [1.84, 9.02] years, respectively. Overall, the mean age was 63.9 years, and the mean ejection fraction was 58.9%. Robotic-assisted surgery was performed in 28.9% (n = 132). The 30-day mortality was 2.4% (n = 11). Propensity-score matching provided 119 patients in each group. Kaplan-Meier curves demonstrated similar survival at 5 years between these groups (93.1 ± 2.8% in the resection group and 89.6 ± 3.1% in the neochord group, P = .5). However, the neochord group had a greater cumulative incidence of reoperative MV surgery (8.0% vs 0.9% at 5 years in the resection group, P = .01).
Conclusions
Neochordae were implanted in 27.3% of patients undergoing MV repair. Neochord implantation was associated with a greater risk of MV reintervention in the long term. Careful patient selection and technical considerations are important when choosing the repair method.
Key Words: mitral valve repair, leaflet resection, neochord implantation, chord replacement, degenerative disease, posterior leaflet prolapse, mitral valve reintervention
Graphical abstract
Cumulative incidence of mitral valve reintervention between resection and neochord.
Central Message.
Neochord implantation may be associated with a greater risk of mitral valve reintervention and more-than-moderate mitral regurgitation after mitral valve repair for posterior leaflet prolapse.
Perspective.
Whether leaflet resection or neochord implantation yields better outcomes in mitral valve (MV) repair for posterior leaflet prolapse remains debated. Neochord implantation was associated with a greater risk of MV reintervention and more-than-moderate mitral regurgitation in the long term. Careful patient selection and technical considerations are important when choosing the repair method.
Mitral valve (MV) repair for posterior leaflet prolapse is commonly performed using either surgical resection of the leaflet or neochord implantation. Although leaflet resection remains the more frequently performed technique, neochord implantation has gained popularity because it preserves the natural anatomy of the leaflet. The selection of a technique is typically at the discretion of the surgeon, although certain anatomical variations may favor one approach over the other. Currently, there remains no widespread consensus or guideline recommending to use either technique. Previous studies have suggested no significant difference in outcomes between the “resect” and “respect” techniques regarding survival, reoperation rates, postoperative mitral stenosis (MS), and residual mitral regurgitation (MR).1, 2, 3, 4 Randomized controlled trials on this subject demonstrated similar outcomes between the 2 techniques, although their findings were limited by a lack of long-term follow-up.1,5 More recently, a Japanese observational study suggested a correlation between chordal replacement and recurrence of mild-to-severe MR at 1 year after surgery, and a Korean study found a similar association between the neochord technique and recurrence of moderate-to-severe MR at 5 years after surgery.6,7 In addition, a retrospective study from Duke University reported an increased rate of reintervention and severe MR at 10 years in patients who underwent chordal replacement.8 Conversely, other works have suggested an association between leaflet resection and increased MV mean gradient.9,10 In this retrospective study, we sought to compare the short- and long-term outcomes of patients who underwent MV repair for posterior leaflet prolapse using either leaflet resection or neochord implantation.
Methods
The study protocol was approved by the institutional review board of the University of Pittsburgh Medical Center on April 22, 2021: STUDY20070004. Informed consent was waived because of the nature of this study.
Data and Patient Selection
All patients who underwent MV repair via leaflet resection or chordal replacement between October 2011 and July 2024 were included in this study. Baseline patient data and intraoperative data were acquired via electronic health record. Patients with predominantly MS, Carpentier I or III, or anterior leaflet involvement were excluded. In addition, patients who required both resection and neochord implantation were excluded. Surgical techniques for resection and neochord implantation depended on each surgeon's preference.
Outcomes
Outcomes of interest included all-cause mortality, MV reintervention, and recurrence of more than moderate MR. MV reintervention included both surgical and transcatheter interventions and was captured for any indications, including infective endocarditis, if it occurred at one of our institutions. More-than-moderate degree of MR was identified any time after the index surgery and was chosen as an end point on the basis of the assumption that it is less likely to be influenced by echocardiogram technician or institutional variability and unlikely to regress once it occurs, whereas trace-to-moderate MR may be affected by patient's volume status or hemodynamics and might be subject to variability in interpretation. Follow-up data were obtained from the clinical warehouse at the University of Pittsburgh Medical Center, which contains all long-term survival data in patients undergoing cardiac surgery at this institution. The follow-up data were matched to the Social Security Death Index. Mechanisms of failure and type of reintervention were also collected if a patient's initial MV repair failed. Postoperative mean gradient was often undocumented and was not analyzed.
Statistics
Categorical variables were compared with the χ2 test or Fisher exact test where appropriate and presented as numbers and percentages. Continuous variables were compared with the Student t test, described as either mean ± standard deviations, or Mann-Whitney-Wilcoxon test, described as median with interquartile ranges, where appropriate. Missing data were less than 10% in all variables, and complete analyses were performed. To adjust the difference in the baseline characteristics between the groups, 1:1 nearest neighbor propensity-score mating (PSM) was applied, using a caliper of 0.2. The selected variables included annuloplasty, arrhythmia surgery, diabetes, and peripheral vascular disease on the basis of results of univariable analyses and clinical knowledge, given the fact they were risks either for morbidity and mortality in cardiac surgery, including MV surgery, or for MV repair failure. The balance between the groups were assessed by the Love plot. Standardized mean difference greater than 0.2 was considered not balanced. After the matching, postoperative outcomes were compared between the matched cohorts. Long-term survival was estimated with Kaplan-Meier curves. The log-rank test was used to compare the survival curves between the cohorts. Cumulative incidence curves were constructed for MV reintervention with death as the competing event, or for more-than-moderate MR with death or MV reintervention as the competing events. The Gray test was used to compare the incidences between cohorts. Several subgroup analyses were performed: (1) a Cox hazard regression analysis using mixed-effects model using surgeon as a random effect to account for surgeon-level effects, (2) a propensity-score calculation without diabetes and peripheral vascular disease because they may not be related to MV failure or reinterventions, and (3) a propensity-score calculation including surgical era (2011-2015, 2016-2020, and 2021-2024). All statistical analyses were conducted using R, version 4.1.3 (R Foundation for Statistical Computing).
Results
Baseline Characteristics
Of 457 patients included in the study, 334 (73.1%) had a leaflet resection (resection group) and 123 (26.9%) underwent chordal replacement (neochord group) (Table 1). Overall, mean age was 63.9 years old. Mean left ventricular ejection fraction was 58.9%. Resternotomy was performed in 2.2% (n = 10). A history of infective endocarditis was observed in 2.2% (n = 10). Moderate or greater tricuspid regurgitation was seen in 20.4% (n = 93). Before PSM, mean age among patients in the resection group (63.5 years old) did not differ from those in the neochord group (64.8 years old) (P = .30). Body surface area also did not differ between groups (1.98 m2 in the resection group vs 1.96 m2 in the neochord group, P = .52). Peripheral vascular disease was more common in the neochord group (8.1%, n = 10) relative to the resection group (3.3%, n = 11) (P = .04). There were no other baseline characteristics that differed between the groups. After PSM, all of the baseline characteristics were balanced between the 2 groups (Figure E1).
Table 1.
Baseline characteristics
| Variable (%) | Before PSM |
After PSM |
||||
|---|---|---|---|---|---|---|
| Resection N = 334 | Neochord N = 123 | P value | Resection N = 119 | Neochord N = 119 | P value | |
| Age, y, mean ± SD | 63.5 ± 11.5 | 64.8 ± 10.3 | .30 | 65.0 ± 11.7 | 64.8 ± 9.92 | .89 |
| Male | 255 (76.3) | 85 (69.1) | .12 | 94 (79.0) | 83 (69.7) | .14 |
| Race | 1 | .40 | ||||
| White | 311 (95.4) | 117 (95.9) | 106 (93.0) | 113 (95.8) | ||
| Non-White | 15 (4.6) | 5 (4.1) | 8 (7.0) | 5 (4.2) | ||
| Body surface area, mean ± SD, m2 | 1.98 ± 0.24 | 1.96 ± 0.24 | .52 | 1.97 ± 0.22 | 1.96 ± 0.24 | .84 |
| Comorbidities | ||||||
| Hypertension | 219 (65.6) | 86 (69.9) | .43 | 87 (73.1) | 85 (71.4) | .89 |
| Dialysis | 3 (0.9) | 0 (0.0) | .57 | 1 (0.8) | 0 (0.0) | 1 |
| Diabetes | 32 (9.6) | 20 (16.3) | .07 | 18 (15.1) | 19 (16.0) | 1 |
| Cerebrovascular disease | 20 (6.0) | 12 (9.8) | .21 | 9 (7.6) | 11 (9.2) | .82 |
| Peripheral vascular disease | 11 (3.3) | 10 (8.1) | .04 | 8 (6.7) | 9 (7.6) | 1 |
| Chronic lung disease | 57 (17.1) | 18 (14.6) | .57 | 18 (15.1) | 18 (15.1) | 1 |
| History of endocarditis | 9 (2.7) | 1 (0.8) | .30 | 4 (3.4) | 1 (0.8) | .37 |
| Resternotomy | 5 (1.5) | 5 (4.1) | .14 | 1 (0.8) | 5 (4.2) | .21 |
| Surgery status | .16 | .73 | ||||
| Elective | 269 (80.5) | 108 (87.8) | 99 (83.2) | 104 (87.4) | ||
| Urgent | 62 (18.6) | 14 (11.4) | 19 (16.0) | 14 (11.8) | ||
| Emergent | 3 (0.9) | 1 (0.8) | 1 (0.8) | 1 (0.8) | ||
| Ejection fraction, %, mean ± SD | 58.9 ± 6.19 | 59.2 ± 8.34 | .65 | 58.8 ± 6.24 | 59.1 ± 8.46 | .78 |
| Mitral insufficiency | 1 | 1 | ||||
| None to mild | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | ||
| Moderate | 10 (3.0) | 3 (2.4) | 4 (3.4) | 3 (2.5) | ||
| Severe | 324 (97.0) | 120 (97.6) | 115 (96.6) | 116 (97.5) | ||
| Tricuspid insufficiency | .22 | .16 | ||||
| None | 4 (1.3) | 5 (4.6) | 0 (0.0) | 5 (4.8) | ||
| Trace | 95 (31.9) | 35 (32.1) | 35 (32.4) | 34 (32.4) | ||
| Mild | 128 (43.0) | 45 (41.3) | 49 (45.4) | 43 (41.0) | ||
| Moderate | 58 (19.5) | 16 (14.7) | 20 (18.5) | 16 (15.2) | ||
| Severe | 11 (3.7) | 8 (7.3) | 4 (3.7) | 7 (6.7) | ||
PSM, Propensity score matching; SD, standard deviation.
Figure E1.
Love plot for the distribution of standardized mean difference. Standardized mean difference (SMD) was plotted to each baseline variable before and after matching. Standardized mean difference greater than 0.2 was considered not balanced. SMD, Standardized mean difference.
Operative Characteristics
Overall, annuloplasty was performed in 97.2% (n = 444). Robotic-assisted surgery was conducted in 28.9% (n = 132). Concomitant surgery included arrhythmia surgery in 14.9% (n = 68), coronary artery bypass grafting in 15.3% (n = 70), and tricuspid valve repair or replacement in 8.1% (n = 37). Immediate postoperative echocardiogram demonstrated less-than-mild MR in 82.7% (n = 378) and mild MR in 12.6% (n = 56). In the resection group, triangular resection was most commonly performed (86.5%, n = 288/334) (Table 2). The median number of neochord used was 2.0 in the neochord group. Before PSM, annuloplasty was performed in 98.5% (n = 329) of the resection group compared with 93.5% (n = 115) of the neochord group (P = .01). The median cardiopulmonary bypass time was longer in the neochord group (155.0 minutes vs 133.0 minutes, P = .001) as well as crossclamp time (113.0 minutes in the neochord group vs 102.0 minutes in the resection group, P = .001). The median size of annuloplasty band or ring was larger in the neochord group (33.0 mm vs 32.0 mm in the resection group, P < .001).
Table 2.
Operative characteristics
| Variable (%) | Before PSM |
After PSM |
||||
|---|---|---|---|---|---|---|
| Resection N = 334 | Neochord N = 123 | P value | Resection N = 119 | Neochord N = 119 | P value | |
| Multiple prolapsed scallops (vs isolated) | 47 (14.1) | 29 (23.6) | .02 | 19 (16.0) | 29 (24.4) | .15 |
| Resection | ||||||
| Triangular resection | 288 (86.5) | N/A | 99 (83.2) | N/A | ||
| Quadrangular resection | 26 (7.8) | N/A | 11 (9.2) | N/A | ||
| Sliding plasty | 51 (15.3) | N/A | 18 (15.1) | N/A | ||
| Number of neochordae, median [IQR] | N/A | 2.00 [2.00, 4.00] | N/A | 2.00 [2.00, 4.00] | ||
| Right thoracotomy | 25 (7.5) | 7 (5.7) | .68 | 8 (6.7) | 7 (5.9) | 1 |
| Robotic-assisted | 99 (29.6) | 33 (26.8) | .64 | 35 (29.4) | 33 (27.7) | .89 |
| Concomitant surgery | ||||||
| CABG | 52 (15.6) | 18 (14.6) | .88 | 22 (18.5) | 18 (15.1) | .60 |
| Aortic valve replacement | 13 (3.9) | 8 (6.5) | .31 | 5 (4.2) | 8 (6.7) | .57 |
| TV repair or replacement | 29 (8.7) | 8 (6.5) | .56 | 10 (8.4) | 8 (6.7) | .81 |
| Arrhythmia surgery | 61 (18.3) | 7 (5.7) | .001 | 7 (5.9) | 7 (5.9) | 1 |
| Annuloplasty | 329 (98.5) | 115 (93.5) | .01 | 114 (95.8) | 115 (96.6) | 1 |
| Annuloplasty size, mm, median [IQR] | 32.0 [30.0, 34.0] | 33.0 [31.8, 36.0] | <.001 | 32.0 [30.0, 34.5] | 33.0 [31.5, 36.0] | .02 |
| Band (vs ring) | 108 (35.1) | 33 (30.8) | .48 | 36 (33.3) | 33 (30.8) | .77 |
| Immediate postoperative mitral regurgitation | ||||||
| None | 166 (51.1) | 48 (40.3) | .05 | 56 (48.7) | 48 (41.7) | .35 |
| Trace | 117 (36.0) | 47 (39.5) | .51 | 41 (35.7) | 45 (39.1) | .68 |
| Mild | 36 (11.1) | 20 (16.8) | .11 | 16 (13.9) | 18 (15.7) | .85 |
| Mild to moderate | 5 (1.5) | 2 (1.7) | 1 | 2 (1.7) | 2 (1.7) | 1 |
| Cardiopulmonary bypass time, min, median [IQR] | 133.0 [100.3, 168.8] | 155.0 [120.8, 187.0] | .001 | 142.0 [105.0, 168.0] | 155.0 [122.0, 187.0] | .01 |
| Crossclamp time, min, median [IQR] | 102.0 [74.0, 127.0] | 113.0 [87.0, 145.3] | .001 | 103.0 [79.0, 127.0] | 115.0 [88.0, 148.0] | .01 |
PSM, Propensity-score matching; N/A, not available; IQR, interquartile range; CABG, coronary artery bypass grafting; TV, tricuspid valve.
After PSM, the use of annuloplasty was balanced between the groups, and the rest of characteristics were also similar except for larger size of annuloplasty in the neochord group (33.0 mm vs 32.0 mm in the resection group, P = .02). The same operative trends persisted regarding bypass time (155.0 minutes in the neochord group vs 142.0 minutes in the resection group, P = .01) and crossclamp time (115.0 min in the neochord group vs 103.0 minutes in the resection group, P = .01).
Outcomes
Perioperative outcomes were comparable between groups after PSM (Table 3). The median [interquartile range] follow-up for survival, reintervention, and recurrence of MR was 6.36 [1.95, 9.13], 6.07 [1.60, 9.09], and 5.73 [1.45, 8.88] years after PSM, respectively. The 5-year completeness of follow-up was 56.3%. Estimated long-term survival was 93.1 ± 2.8% at 5 years in the resection group and 89.6 ± 3.1% at 5 years in the neochord group, respectively (P = .5, Figure 1). Although survival was equivalent, cumulative incidence of MV reintervention was significantly greater in the neochord group, with the estimated incidence of 8.0% at 5 years (vs 0.9% at 5 years in the resection group, P = .01, Figure 2). The use of neochord was independently associated with MV reintervention (hazard ratio, 3.39; 95% confidence interval, 1.14-10.1, P = .03) in the subgroup analysis with a mixed-effect model using surgeon as a random effect. Similarly, more-than-moderate MR was more frequently observed in the neochord group, with an incidence of 9.3% at 5 years (vs 0.9% at 5 years in the resection group, P = .001, Figure 3). The findings in subgroup analyses matching without diabetes and peripheral vascular disease (Figures E2 and E3) or with surgical era (Figures E4 and E5) remained comparable with the primary analysis.
Table 3.
Postoperative outcomes after PSM
| Variable (%) | Resection N = 119 | Neochord N = 119 | P value |
|---|---|---|---|
| 30-d mortality | 2 (1.7) | 5 (4.2) | .28 |
| Stroke | 1 (0.8) | 0 (0.0) | 1 |
| Return to operating room for bleeding | 4 (3.4) | 9 (7.6) | .25 |
| Prolonged ventilation | 5 (4.2) | 12 (10.1) | .13 |
| Renal failure | 0 (0.0) | 2 (1.7) | .50 |
| Sepsis | 1 (0.8) | 3 (2.5) | .62 |
| Surgical-site infection | 2 (1.7) | 0 (0.0) | .50 |
| Readmission | 2 (9.1) | 2 (11.8) | 1 |
| Postoperative length of stay, d, median [IQR] | 5.00 [4.00, 7.50] | 5.00 [4.00, 7.50] | .44 |
PSM, Propensity-score matching; IQR, interquartile range.
Figure 1.
Long-term survival after mitral valve repair between via resection and via neochord implantation. The survival probability was plotted against years after surgery with 95% confidence intervals in patients who underwent mitral valve repair between by leaflet resection and by neochord implantation after propensity-score matching. The curves were compared with the log-rank test between the groups. The number of patients at risk is shown at the bottom.
Figure 2.
Long-term mitral valve reinterventions after mitral valve repair between via resection and via neochord implantation. The cumulative incidence of mitral valve reintervention was plotted against years after surgery with 95% confidence intervals in patients who underwent mitral valve repair and death as the competing event between by leaflet resection and by neochord implantation after propensity-score matching. The curves were compared with the Gray test between the groups. The number of patients at risk is shown at the bottom.
Figure 3.
Long-term recurrence of more-than-moderate mitral regurgitation after mitral valve repair between via resection and via neochord implantation. The cumulative incidence of more-than-moderate mitral regurgitation was plotted against years after surgery with 95% confidence intervals and death and mitral valve reintervention as the competing event in patients who underwent mitral valve repair between by leaflet resection and by neochord implantation after propensity-score matching. The curves were compared with the Gray test between the groups. The number of patients at risk is shown at the bottom.
Figure E2.
Long-term mitral valve reinterventions after mitral valve repair between via resection and via neochord implantation after propensity-score matching without diabetes and peripheral vascular disease. The cumulative incidence of mitral valve reintervention was plotted against years after surgery with 95% confidence intervals in patients who underwent mitral valve repair and death as the competing event between by leaflet resection and by neochord implantation after propensity-score matching without diabetes and peripheral vascular disease. The curves were compared with the Gray test between the groups. The number of patients at risk is shown at the bottom.
Figure E3.
Long-term recurrence of more-than-moderate mitral regurgitation after mitral valve repair between via resection and via neochord implantation after propensity-score matching without diabetes and peripheral vascular disease. The cumulative incidence of more-than-moderate mitral regurgitation was plotted against years after surgery with 95% confidence intervals and death and mitral valve reintervention as the competing event in patients who underwent mitral valve repair between by leaflet resection and by neochord implantation after propensity-score matching without diabetes and peripheral vascular disease. The curves were compared with the Gray test between the groups. The number of patients at risk is shown at the bottom.
Figure E4.
Long-term mitral valve reinterventions after mitral valve repair between via resection and via neochord implantation after propensity-score matching including surgical era. The year of surgery was divided into 3 eras: 2011-2015, 2016-2020, and 2021-2024. The cumulative incidence of mitral valve reintervention was plotted against years after surgery with 95% confidence intervals in patients who underwent mitral valve repair and death as the competing event between by leaflet resection and by neochord implantation after propensity- score matching including surgical era. The curves were compared with the Gray test between the groups. The number of patients at risk is shown at the bottom.
Figure E5.
Long-term recurrence of more-than-moderate mitral regurgitation after mitral valve repair between via resection and via neochord implantation after propensity-score matching including surgical era. The year of surgery was divided into 3 eras: 2011-2015, 2016-2020, and 2021-2024. The cumulative incidence of more-than-moderate mitral regurgitation was plotted against years after surgery with 95% confidence intervals and death and mitral valve reintervention as the competing event in patients who underwent mitral valve repair between by leaflet resection and by neochord implantation after propensity-score matching including surgical era. The curves were compared with the Gray test between the groups. The number of patients at risk is shown at the bottom.
MV Reintervention
A total of 16 patients necessitated MV reintervention during follow-up (Table 4). The median duration until MV reintervention was 1.47 years. Although redo MV repair (n = 6, 37.5%) or replacement (n = 8, 50%) was primarily performed, transcatheter MV reintervention, including edge-to-edge repair and transcatheter MV replacement, were also performed in 2 patients (12.5%). The mechanism of MV failure in postoperative leaflet resection patients who required reintervention was either failure of annuloplasty or recurrence of flail/prolapsed leaflets in equal frequency. MS was not observed as the mechanism of failure. In patients requiring MV reintervention after neochord implantation, 8 patients (80%) required 4 or more neochordae, and 3 patients (33.3%) experienced neochord-technique specific failure.
Table 4.
Characteristics of patients who required mitral reinterventions
| Patient | Preoperative prolapse | Resection | Number of chords used | Annuloplasty | Duration until reintervention, y | Mechanism of failure | Method of reintervention |
|---|---|---|---|---|---|---|---|
| 1 | P2 | Quadrangular with sliding plasty | N/A | Yes | 3.4 | P1 and P2 prolapse | Redo-repair |
| 2 | P2 | Triangular | N/A | Yes | 1.5 | Failure of annuloplasty | Redo-repair |
| 3 | More than 1 scallop | Triangular | N/A | Yes | 7 mo | Failure of annuloplasty | Resternotomy replacement |
| 4 | P2 | Triangular | N/A | Yes | 3.4 | P1 and P2 prolapse | Resternotomy replacement |
| 5 | P3 | Quadrangular | N/A | Yes | 10 mo | Failure of annuloplasty | Redo-repair |
| 6 | P2 | Triangular | N/A | Yes | 1.2 | P3 flail | TEER |
| 7 | More than 1 scallop | N/A | 4 | Yes | 9.8 | Posterior leaflet prolapse | TMVR |
| 8 | P3 | N/A | 4 | Yes | 7.4 | Degenerative | Resternotomy replacement |
| 9 | More than 1 scallop | N/A | 6 | Yes | 3 d | Papillary muscle rupture | Resternotomy replacement |
| 10 | P2 | N/A | 4 | No | 2 mo | P1 prolapse and cleft between P1 and P2 | Redo-repair |
| 11 | P3 | N/A | 2 | No | 3.5 | P1 prolapse | Resternotomy replacement |
| 12 | P2 | N/A | 4 | Yes | 1.2 | Posterior leaflet prolapse | Redo-repair |
| 13 | P2 | N/A | 4 | No | 2.3 | Torn neochords | Resternotomy replacement |
| 14 | P1 | N/A | 6 | Yes | 9 mo | Degenerative | Redo-repair |
| 15 | P2 | N/A | 8 | Yes | 4.5 | Torn neochords | Resternotomy replacement |
| 16 | More than 1 scallop | N/A | 2 | Yes | 1.4 | P1 and P2 flail | Resternotomy replacement |
N/A, Not available; TEER, transcatheter edge-to-edge repair; TMVR, transcatheter mitral valve replacement.
Discussion
This retrospective analysis of 457 patients who underwent MV repair for posterior leaflet prolapse either via leaflet resection or neochord implantation examined differences in patient characteristics and outcomes. This study demonstrated that (1) long-term survival was comparable between the groups, (2) patients who underwent neochord implantation had greater rates of MV reintervention, and (3) recurrence of more-than-moderate MR at 5 years compared with those with a leaflet resection-based MV repair. In addition, context on reintervention method is provided, providing insights into mechanisms of failure for each surgical technique and differences in actual reintervention.
Long-term survival between both groups was comparable in the present study. This is consistent with the prevailing thought that the long-term survivability of both techniques is extremely positive, to the point of restoring complete life expectancy.11 Although only a few studies reported 10-year survival comparing these techniques, they estimated 10-year survival rates ranging from 79% to 96% after MV repair using each technique.8,12 Although long-term survival is generally not a major concern in MV repair, maximizing a patient's freedom from long-term complications is essential in order to reduce the personal, medical, and economic burden associated with MV reintervention or recurrent MV disease. Notably, the operative mortality for redo MV surgery was 6.6% nationwide, which was more than twice that of primary MV surgery, and redo MV surgery is better to be avoided.13
Patients who received neochord implantation experienced a significantly greater rate of MV reintervention, consistent with findings from a previous study.8 In contrast, other studies with a 10-year follow-up reported no significant difference in reintervention rates.12,14 The discrepancies between these studies may be attributed to differences in cohorts, 2 studies including both anterior and posterior leaflet prolapse, whereas the current study focused exclusively on posterior leaflet prolapse, and the other study included patients with posterior leaflet prolapse and dilated left ventricle. However, variations in institutional and surgeon experiences, including operative techniques and decision-making processes, likely affects outcomes, and the inconsistency of findings may be multifactorial. A detailed analysis of failure mechanisms for each technique within individual institutions could provide valuable insights into optimizing long-term outcomes.
Interestingly, the mechanism of failure in one half of the patients who underwent MV reintervention after repair in the present study was failure as a result of annuloplasty-related issues such as dehiscence, none of which occurred in the neochord group. Furthermore, only one third of the reinterventions after neochord implantation were directly related to technique- or chord-specific failure. One possible explanation for failure after neochord implantation was left ventricular remodeling, which might lead to a mismatch between leaflet height and neochord length over time.7,12 Alternatively, further degenerative changes after repair may be a potential reason. Degenerative changes of MV in the initial operation likely involves not only the prolapsed leaflet but also other structures such as adjacent leaflets or chordae. Resection of prolapsed leaflets, which reforms or reinforces the surrounding geometry, might offer longer durability by mitigating repair failure or reintervention than neochord implantation, which restores the natural appearance but leaves potentially abnormal tissue intact.
This study's finding that leaflet resection was also associated with a greater recurrence rate of more-than-moderate MR aligns with recent literature.7,8 The exact reasons for this increased recurrence is not well understood and likely differs on a case-by-case basis, although potential failure mechanisms have been speculated, as discussed earlier. It is important to consider that the neochord technique is often used in patients with more complex MV pathologies, potentially playing a role in increased MR recurrence rate. The complexity of valve pathology or technical difficulty in neochord implantation may be reflected in the longer cardiopulmonary bypass and crossclamp time in the neochord group compared with the resection group in the current study. In fact, 80% of patients who required MV reintervention had received 4 or more neochordae during their initial MV repair. However, the present study attempted to account for underlying pathology by excluding anterior leaflet involvement and ensuring a balanced comparison of prolapsed leaflet pathology, whether it was isolated or involving multiple leaflets, between 2 groups. Importantly, better clinical outcomes with neochord implantation for posterior leaflet prolapse have rarely been reported compared with resection, although physiologic parameters may be better and larger annuloplasty size may be achieved with neochord implantation as shown in the present study.15, 16, 17 Given the fact that some evidence suggested decreased longevity of repair using the neochord technique, vigilance long-term postoperative monitoring remains critical. Although such complex pathologies still indicate the use of neochord, the findings of this study suggest that use of the neochord technique should be approached with increased caution because of its long-term implications.
This study was limited by its institutional nature, analyzing patients largely from a similar region and context with small cohort of surgeons. As such, consideration of these findings in the broader scientific context before generalized to other settings. Because this study was conducted in a large academic medical center, similar results may not be applicable to smaller hospitals that more heavily specialize in one technique or have a patient population with comorbidities that may disproportionately impact outcomes for one of these techniques. Although PSM was used to mitigate bias, there may be some implicit selection bias. The heterogeneity of MV failure, both on the scale of the valve itself and on that of the holistic patient, makes this difficult to effectively control. In addition, the findings were derived from PSM-matched cohorts and may not be applicable to the entire cohort. Survival curves were statistically estimated, and the follow-up duration was limited. Further, follow-up for echocardiogram was restricted because of the retrospective nature of the study, as there was no predefined interval.
In conclusion, there is no consensus on the long-term impacts of leaflet resection or neochord implantation in MV repair. Long-term survivals between patients who received these operations are not appreciably different, but the neochord group experienced greater incidence of MV reintervention and recurrence of more-than-moderate MR compared with the resection group. The characteristics of failure and mechanism of reintervention differed between groups and may provide further insights in each technique.
Webcast
You can watch a Webcast of this AATS meeting presentation by going to: https://www.aats.org/resources/posterior-leaflet-reconstructi-9735.
Audio
You can listen to the discussion audio of this article by going to the supplementary material section below.
Conflict of Interest Statement
Dr Sultan receives institutional research support from Abbott, AtriCure, Artivion, W. L. Gore & Associates, Inc, Edwards, Medtronic, and Terumo Aortic. Dr Kaczorowski is a consultant for Abiomed. None of these are related to this manuscript. All other authors reported no conflicts of interest.
The Journal policy requires editors and reviewers to disclose conflicts of interest and to decline handling or reviewing manuscripts for which they may have a conflict of interest. The editors and reviewers of this article have no conflicts of interest.
Footnotes
Dr Ibrahim Sultan is an Associate Editor. The peer-review process for this paper was handled by Dr Leonard Girardi.
Dr Ogami and E. Chetkof contributed equally to this article.
Appendix E1
Supplementary Data
This is an audio recording of the presentation of the associated abstract and discussion. The abstract presentation ends at 6:14 and the discussion begins at 6:17. There is no visual component as this is an audio recording.
References
- 1.Chan V., Mazer C.D., Ali F.M., et al. Randomized, controlled trial comparing mitral valve repair with leaflet resection versus leaflet preservation on functional mitral stenosis: the CAMRA CardioLink-2 study. Circulation. 2020;142(14):1342–1350. doi: 10.1161/CIRCULATIONAHA.120.046853. [DOI] [PubMed] [Google Scholar]
- 2.Tomšič A., Holubec T., Sandoval E., et al. Mitral valve repair with resection and non-resection techniques in Barlow’s disease: a multi-center study. Int J Cardiol. 2024;413 doi: 10.1016/j.ijcard.2024.132387. [DOI] [PubMed] [Google Scholar]
- 3.Caldonazo T., Sá M.P., Jacquemyn X., et al. Respect versus resect approaches for mitral valve repair: a meta-analysis of reconstructed time-to-event data. Am J Cardiol. 2024;213:5–11. doi: 10.1016/j.amjcard.2023.12.010. [DOI] [PubMed] [Google Scholar]
- 4.Ma J., Liu J., Wei P., et al. Quadrangular resection versus chordal replacement for degenerative posterior mitral leaflet prolapse. Ann Transl Med. 2021;9(1):60. doi: 10.21037/atm-20-7475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Falk V., Seeburger J., Czesla M., et al. How does the use of polytetrafluoroethylene neochordae for posterior mitral valve prolapse (loop technique) compare with leaflet resection? A prospective randomized trial. J Thorac Cardiovasc Surg. 2008;136(5):1200–1206. doi: 10.1016/j.jtcvs.2008.07.028. [DOI] [PubMed] [Google Scholar]
- 6.Kakuta T., Fukushima S., Minami K., et al. What is the optimal mitral valve repair for isolated posterior leaflet prolapse to achieve long term durability? J Am Heart Assoc. 2023;12(11) doi: 10.1161/JAHA.122.028607. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kwon Y., Kim H.J., Kim J.B., et al. Mitral valve repair with leaflet resection versus preservation for degenerative posterior leaflet prolapse. J Thorac Cardiovasc Surg. 2025;169(6):1704–1713.e5. doi: 10.1016/j.jtcvs.2024.06.027. [DOI] [PubMed] [Google Scholar]
- 8.Choi J.W., Bishawi M., Zwischenberger B., Gaca J., Carr K., Glower D.D. Outcomes of leaflet resection vs chordal replacement for degenerative mitral regurgitation. Ann Thorac Surg. 2024;118(5):1161–1166. doi: 10.1016/j.athoracsur.2024.06.036. [DOI] [PubMed] [Google Scholar]
- 9.Sawazaki M., Tomari S., Zaikokuji K., Imaeda Y. Controversy in mitral valve repair, resection or chordal replacement? Gen Thorac Cardiovasc Surg. 2014;62(10):581–585. doi: 10.1007/s11748-014-0459-x. [DOI] [PubMed] [Google Scholar]
- 10.Sakaguchi T., Hiraoka A., Totsugawa T., et al. Clinical impact of the repair technique for posterior mitral leaflet prolapse: resect or respect? J Card Surg. 2021;36(3):971–977. doi: 10.1111/jocs.15312. [DOI] [PubMed] [Google Scholar]
- 11.Watt T.M.F., Brescia A.A., Murray S.L., et al. Degenerative mitral valve repair restores life expectancy. Ann Thorac Surg. 2020;109(3):794–801. doi: 10.1016/j.athoracsur.2019.07.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Del Forno B., Tavana K., Ruffo C., et al. Neochordae implantation versus leaflet resection in mitral valve posterior leaflet prolapse and dilated left ventricle: a propensity score matching comparison with long-term follow-up. Eur J Cardiothorac Surg. 2023;64(4) doi: 10.1093/ejcts/ezad274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kilic A., Acker M.A., Gleason T.G., et al. Clinical outcomes of mitral valve reoperations in the United States: an analysis of the society of thoracic surgeons national database. Ann Thorac Surg. 2019;107(3):754–759. doi: 10.1016/j.athoracsur.2018.08.083. [DOI] [PubMed] [Google Scholar]
- 14.Pfannmueller B., Misfeld M., Verevkin A., et al. Loop neochord versus leaflet resection techniques for minimally invasive mitral valve repair: long-term results. Eur J Cardiothorac Surg. 2021;59(1):180–186. doi: 10.1093/ejcts/ezaa255. [DOI] [PubMed] [Google Scholar]
- 15.Lange R., Guenther T., Noebauer C., et al. Chordal replacement versus quadrangular resection for repair of isolated posterior mitral leaflet prolapse. Ann Thorac Surg. 2010;89(4):1163–1170. doi: 10.1016/j.athoracsur.2009.12.057. [DOI] [PubMed] [Google Scholar]
- 16.Imasaka K., Tayama E., Tomita Y. Left ventricular performance early after repair for posterior mitral leaflet prolapse: chordal replacement versus leaflet resection. J Thorac Cardiovasc Surg. 2015;150(3):538–545. doi: 10.1016/j.jtcvs.2015.06.022. [DOI] [PubMed] [Google Scholar]
- 17.Mesana T.G., Lam B.K., Chan V., Chen K., Ruel M., Chan K. Clinical evaluation of functional mitral stenosis after mitral valve repair for degenerative disease: potential effect on surgical strategy. J Thorac Cardiovasc Surg. 2013;146(6):1418–1425. doi: 10.1016/j.jtcvs.2013.08.011. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
This is an audio recording of the presentation of the associated abstract and discussion. The abstract presentation ends at 6:14 and the discussion begins at 6:17. There is no visual component as this is an audio recording.