論文ID: CJ-24-0291
Background: There is concern about the durability of mitral valve repair (MVr) for mitral regurgitation (MR) in Marfan patients due to limited long-term data. Furthermore, a detailed time course of changes in cardiac function after MVr in Marfan patients has not been reported. We examined repair techniques, postoperative cardiac function, and outcomes of MVr in Marfan patients.
Methods and Results: We retrospectively reviewed 29 Marfan patients (mean [±SD] age 27.4±14.8 years) who underwent MVr at The University of Tokyo Hospital from 2010 to 2022. The mean follow-up period was 5.2±3.2 years. The causes of MR were isolated anterior leaflet prolapse in 25% of patients, isolated posterior leaflet prolapse in 11%, and bileaflet prolapse in 64%. Echocardiographic findings showed significant decreases in left ventricular (LV) diastolic and left atrial diameters 1 week after MVr. LV systolic diameter was significantly decreased 3 years after MVr, and LV ejection fraction initially declined before subsequently increasing. The in-hospital and 30-day mortality rates were 0%. At 5 years, the overall survival rate was 94% and the rate of freedom from MR was 84%.
Conclusions: The mid- to long-term outcomes after MVr in Marfan patients were satisfactory, supporting the durability of MVr in these patients. Postoperative cardiac reverse remodeling occurred in a phased manner in Marfan patients, similar to that in patients with degenerative MR.
Marfan syndrome is a genetic connective tissue disease that includes cardiovascular, ocular, and skeletal disorders.1 Mitral valve dysfunction occurs in 50% of patients with Marfan syndrome and is therefore one of the most common cardiac manifestations in these patients.2,3 The mitral valve lesions in patients with Marfan syndrome are often extensive and may include anterior leaflet or bileaflet prolapse, including Barlow’s syndrome, which makes mitral valve repair (MVr) challenging.4 Moreover, there are concerns about the durability of MVr in Marfan patients because of underlying connective tissue disorder. For these reasons, many institutes still tend to perform mitral valve replacement (MVR) for mitral regurgitation (MR) in Marfan patients, as evidenced by an MVr rate of 51% during 2010 and 2011 in a nationwide US database, which is less than the rate among patients with degenerative mitral valve disease.5,6
There have been several previous studies on the results of surgical treatment of MR in Marfan patients that have reported good mid- to long-term outcomes with high rates of MVr.4,7–10 However, these studies include relatively old data. Therefore, data regarding long-term outcomes of surgical treatment for MR in patients with Marfan syndrome in the current era are needed. Furthermore, there have been no previous reports describing the detailed time course of changes in cardiac function after mitral surgery in Marfan patients. Hence, the aim of the present study was to examine repair techniques, postoperative courses of cardiac function, and early- and mid- to long-term outcomes of MVr in Marfan patients.
This study was approved by the Research Ethics Committee of The University of Tokyo Hospital (Approval no. 3031-4; Approval date: 19 May, 2021). Because of the retrospective nature of the study, the requirement for informed consent was waived.
From January 2010 to December 2022, 30 patients with Marfan syndrome underwent mitral valve surgery at The University of Tokyo Hospital. One patient underwent MVR because of a history of Bentall procedure using a mechanical valve, which raised concerns about a long aortic cross-clamp time due to a concomitant total arch replacement. Excluding this 1 case of MVR, we retrospectively reviewed 29 patients with Marfan syndrome who underwent MVr. All patients were diagnosed with Marfan syndrome according to the revised Ghent criteria.11
Preoperative patient characteristics are presented in Table 1. The mean (±SD) follow-up period for the 29 patients who underwent MVr for MR was 5.2±3.2 years. Patients ranged in age from 5 to 60 years, with a mean age of 27.4±14.8 years; 3 patients were aged <12 years. Eight (28%) patients had had previous cardiac surgery (the David Procedure in 5, the Bentall procedure in 2, and combined Bentall procedure and total arch replacement in 1). A fibrillin 1 (FBN1) gene mutation was detected in 20 (69%) patients; 1 (3%) patient was negative for FBN1 gene mutations and no genetic testing was performed in the remaining 8 (28%) patients.
Preoperative Patient Characteristics
| No. patients | 29 |
| Follow-up period (years) | 5.2±3.2 |
| Age (years) | 27.4±14.8 |
| No. patients aged <12 years | 3 (10) |
| Male sex | 19 (66) |
| Height (cm) | 174±19.4 |
| Weight (kg) | 52.3±20.0 |
| Body surface area (m2) | 1.61±0.37 |
| Previous cardiac surgery | 8 (28) |
| Atrial fibrillation | 6 (21) |
| FBN1 mutation | |
| Positive | 20 (69) |
| Negative | 1 (3) |
| Unknown (no genetic testing) | 8 (28) |
Values are presented as the mean±SD or n (%). FBN1, fibrillin 1.
Follow up
All patients were followed up with transthoracic echocardiography (TTE) before surgery, 1 week postoperatively and annually thereafter. Transesophageal echocardiography was performed preoperatively and intraoperatively to plan and evaluate the MVr procedure. The results of TTE at 6 time points (i.e., before surgery, and 1 week, 6 months, and 1, 3, and 5 years after surgery) were reviewed. The dimensions of the left ventricle (LV) and left atrium (LA) were measured using a 2-dimensional parasternal long-axis view, and LV ejection fraction (LVEF) was measured using the modified Simpson method. The severity of MR was determined on the basis of quantitative assessment according to the guidelines of the American Society of Echocardiography as follows: 0, none or trivial; 1, mild; 2, moderate; 3, moderate-severe; 4, severe.12 When quantitative evaluation was difficult, the MR grade was evaluated in a semiquantitative way.
Operative TechniquesAll operations were performed via a median sternotomy. Our strategies of MVr for MR are detailed below.
Posterior Leaflet A triangular resection of excessive tissue was first performed. We used artificial chordae, if necessary. Height reduction by butterfly resection13 was also performed for a tall (>20 mm) posterior leaflet.
Anterior Leaflet We did not resect the anterior leaflet, and preferentially used artificial chordae reconstruction with double-armed polytetrafluoroethylene sutures.
Commissural Area Commissure stitches were placed to regulate regurgitation at commissural areas. In cases where a mitral valve at a commissural area was redundant, we added commissural closure stitches, even if there was no regurgitation in a commissural area, to prevent possible future regurgitation from the commissure.
Leaflet plication and cleft closure were also added in combination with the above techniques as appropriate.
Under-sized mitral annuloplasty (MAP) using a semirigid ring was performed basically, with the ring size determined on the basis of the intertrigonal distance or anterior leaflet area.
Statistical AnalysisData are presented as the mean±SD. Echocardiographic data before and after surgery were compared using paired t-tests. Overall survival and freedom from recurrent MR were analyzed using Kaplan-Meier analyses. Statistical significance was set at P<0.05. All analyses were performed using R software (R Foundation for Statistical Computing, Vienna, Austria; https://www.R-project.org/).
Intraoperative findings are presented in Table 2. One (3%) patient had atrial functional MR and 28 (97%) had mitral valve prolapse. Seven (25%) patients had isolated prolapse of the anterior leaflet, 3 (11%) had isolated prolapse of the posterior leaflet, and 18 (64%) had prolapse of both leaflets. Regarding horizontal direction, prolapse on the medial or both sides was present in 26 (93%) patients, whereas 2 (7%) patients had prolapse on the lateral side. Thirteen (45%) patients had valve features of Barlow disease, defined as a valve with prolapse of ≥2 of the 3 segments of each leaflet and the presence of excess tissue on both leaflets.14 The mean intertrigonal distance measured intraoperatively was 38.2±3.9 mm.
Intraoperative Findings
| Cause of MR | |
| Atrial functional MR | 1 (3) |
| Mitral valve prolapse | 28 (97) |
| Area of prolapse (longitudinal direction) | |
| Anterior leaflet only | 7 (25) |
| Posterior leaflet only | 3 (11) |
| Both leaflets | 18 (64) |
| Area of prolapse (horizontal direction) | |
| Medial side (A2, P2-A3, P3) | 15 (54) |
| Lateral side (A1, P1-A2, P2) | 2 (7) |
| Both sides | 11 (39) |
| Barlow valve | 13 (45) |
| Complicated with chordal rupture | 3 (10) |
| Intertrigonal distance (mm) | 38.2±3.9 |
Values are presented as the mean±SD or n (%). MR, mitral regurgitation.
Operative variables are presented in Table 3. The techniques used for MVr were as follows: implantation of artificial chordae in 26 (90%) patients, triangular resection in 15 (52%), butterfly resection in 1 (3%), leaflet plication in 7 (24%), cleft closure in 1 (3%), and commissure closure in 26 (90%). Cleft closure was performed in the patient who had a physiological indentation in the posterior leaflet and in whom the water test during surgery revealed regurgitation originating from this indentation. MAP using a semirigid ring was performed in 28 (97%) patients. The size of the ring ranged from 30 to 36 mm, with 82% of rings either 34 or 36 mm. Concomitant aortic procedures were performed in 7 (24%) patients, including the David Procedure in 6 (21%), and the Bentall procedure in 1 (3%). The mean cardiopulmonary bypass time was 230±97 min and the mean aortic cross-clamp time was 173±82 min.
Operative Variables
| Techniques of mitral valve repair | |
| Artificial chordae | 26 (90) |
| Triangular resection | 15 (52) |
| Butterfly resection | 1 (3) |
| Leaflet plication | 7 (24) |
| Cleft closure | 1 (3) |
| Commissure closure | 26 (90) |
| Mitral annuloplasty | |
| No mitral annuloplasty | 1 (3) |
| Physio ring II | 28 (97) |
| Ring size | |
| 30 mm | 1 (4) |
| 32 mm | 4 (14) |
| 34 mm | 12 (43) |
| 36 mm | 11 (39) |
| Concomitant procedure | |
| David | 6 (21) |
| Bentall | 1 (3) |
| Tricuspid annuloplasty | 5 (17) |
| Maze | 3 (10) |
| Left atrial appendage closure | 7 (24) |
| Left atrial plication | 2 (7) |
| Ravitch | 2 (7) |
| Patent fossa ovalis closure | 2 (7) |
| Intra-aortic balloon pump | 1 (3) |
| Peripheral VA-ECMO | 1 (3) |
Values are presented as n (%). VA-ECMO, veno-arterial extracorporeal membrane oxygenation.
Echocardiographic Findings
The time course of changes in the severity of MR are shown in Figure 1. Preoperative echocardiography revealed that the severity of MR was 2+ in 3 (10%) patients, 3+ in 10 (34%), and 4+ in 16 (55%). Two of 3 patients who had 2+ MR before surgery underwent concomitant aortic procedures. Postoperative echocardiography at 1 week revealed 0 or 1+ MR in all patients, but 3 (10%) patients had 2+ or 3+ MR at 1 year after surgery.
Time course of changes in the severity of mitral regurgitation (MR). Preoperatively (Pre), the severity of MR was 2+ in 3 (10%) patients, 3+ in 10 (34%), and 4+ in 16 (55%). Postoperative echocardiography at 1 week revealed 0 or 1+ MR in all patients; however, 1 year after surgery, 3 (10%) patients had 2+ or 3+ MR. MVr, mitral valve repair.
The time course of changes in echocardiographic data are shown in Figure 2. The LV diastolic diameter (LVDd) and LA diameter decreased significantly from baseline within the first week after surgery and continued to decrease gradually. LVEF decreased significantly from baseline at 1 week after surgery, but increased at 1 year after surgery, and returning to baseline levels 3 years after surgery. LV systolic diameter (LVDs) did not exhibit a significant decrease from the baseline up to 1 year after surgery, but was significantly lower than baseline at 3 years after surgery.
Time course of changes in the echocardiographic data. (A) Left ventricular diastolic diameter (LVDd) and (D) left atrial diameter (LAD) decreased significantly from baseline within the first week after surgery. (B) There was no show significant decrease in left ventricular systolic diameter (LVDs) from baseline up to 1 year after surgery, but LVDs was significantly lower than baseline at 3 years after surgery. (C) Although left ventricular ejection fraction (LVEF) decreased initially, it increased at 1 year after surgery. These findings indicate that cardiac reverse remodeling occurred in a phased manner. Data are the mean±SD. *P<0.05, **P<0.01, ***P<0.001 compared with preoperative echocardiography (Pre). MVr, mitral valve repair.
These findings indicate that LV reverse remodeling occurred in a phased manner after MVr for MR in Marfan patients.
Early OutcomesEarly outcomes are presented in Table 4. The in-hospital and 30-day mortality rates were 0%. Three (10%) patients experienced postoperative bleeding requiring resternotomy. One of these patients underwent a concomitant David Procedure, and bleeding from the aortic root and the site of the coronary reconstruction was the cause of the resternotomy. The second patient had delayed sternal detachment, leading to bleeding from the sternum. The third patient had pinhole-like bleeding from the ascending aorta of unknown cause. Systolic anterior motion (SAM) occurred in 2 (7%) patients. One patient developed SAM intraoperatively during weaning from cardiopulmonary bypass that improved by reducing the catecholamine dose and increasing preload. The other patient did not develop SAM intraoperatively; rather, SAM was identified on echocardiography 1 day after surgery and improved with medical management. Neither of these patients has experienced SAM recurrence for >5 years postoperatively. One patient required an intra-aortic balloon pump and peripheral veno-arterial extracorporeal membrane oxygenation due to a deterioration in cardiac function for weaning from cardiopulmonary bypass. One patient who underwent a concomitant David Procedure had myocardial infarction and ventricular fibrillation due to stenosis of the reconstructed coronary artery.
Early Outcomes After Mitral Valve Repair
| 30-day mortality | 0 (0) |
| In-Hospital mortality | 0 (0) |
| Hospital stay (days) | |
| Mean±SD | 25.5±30.7 |
| Median | 18 |
| Complications | |
| Postoperative bleeding | 3 (10) |
| SAM of the mitral valve | 2 (7) |
| Intra-aortic balloon pump | 1 (3) |
| Peripheral VA-ECMO | 1 (3) |
| Myocardial infarction | 1 (3) |
| Ventricular fibrillation | 1 (3) |
| Tracheostomy | 1 (3) |
| Pneumothorax | 1 (3) |
Unless indicated otherwise, data are given as n (%). SAM, systolic anterior motion. Other abbreviations as in Tables 2,3.
Mid- to Long-Term Outcomes
The overall survival rate at 5 years was 94% (Figure 3A). One patient died due to heart failure following Type A aortic dissection 5 years after mitral valve surgery. The rate of freedom from recurrent MR (2+ or more) was 84% at 5 years (Figure 3B). MR recurred in 4 patients, 2 of whom them required reoperation 6 and 8 years after surgery. No other patients underwent reoperation during the follow-up period. Details of the 4 patients with recurrent MR are presented in the Supplementary Table.
Kaplan-Meier curves of (A) overall survival and (B) freedom from recurrent mitral regurgitation (MR; defined as 2+ or more). At 5 years, the overall survival rate was 94% and the rate of freedom from recurrent MR was 84%.
The main findings of the present study are that the mid- to long-term outcomes of MVr for MR in Marfan patients were satisfactory and that cardiac reverse remodeling occurred in a phased manner after MVr in Marfan patients.
It has been widely reported that, compared with MVR in degenerative MR disease, MVr has a better prognosis and preserves cardiac function.15–18 Helder et al. reported that MVr had better survival compared with MVR in Marfan patients, similar to outcomes in patients with degenerative MR disease.9 Previous studies on outcomes of mitral surgery for MR in Marfan patients have reported that at 5 years survival rates were 77–94% and rates of freedom from reoperation for MR were 80–96% for populations with mean age ranging from 27 to 41 years.4,7–9 It should be noted that these reports include relatively old data and that rates of MVr varied among studies. In the present study, at 5 years, the survival rate was 94% and the rate of freedom from reoperation for MR was 100%, with these rates comparing favorably with previous reports. This may be due to the following: the MVr rate in the present study was 100%; the MVr procedure for complex MR lesions has become more common and has been established in the current era; and the mean age of patients in the present study was 27 years, which is relatively younger than in previous reports. Our results, based on relatively recent data, indicate that the mid- to long-term outcomes after mitral surgery in Marfan patients were satisfactory, supporting the durability of MVr in Marfan patients.
In the present study, phased cardiac reverse remodeling after MVr was demonstrated in Marfan patients. The LVDd and LA diameter were significantly smaller and LVEF was significantly lower 1 week after surgery than at the baseline. Thereafter, the LVDd and LA diameter continued to decrease gradually, whereas LVEF increased before returning to baseline levels 3 years after surgery. Conversely, LVDs started to decrease mildly at the time of discharge, decreasing to a significantly lower level than at baseline 3 years after surgery. These courses of postoperative cardiac reverse remodeling are similar to those in patients with degenerative MR undergoing MVr.19–23 Our results indicate that LA and LV reverse remodeling after MVr for MR in Marfan patients occurred in a similar manner to that seen in patients with degenerative MR.
Previous reports have shown that surgical intervention for MR before the onset of LVEF reduction and LV enlargement improves the prognosis of postoperative cardiac function.20,21,24 In the present study, 3 (10%) patients underwent mitral surgery for 2+ MR, and 10 (34%) underwent surgery for 3+ MR. Pyeritz and Wappel reported that mitral valve dysfunction was progressive in nearly half the patients with Marfan syndrome over a 4-year follow-up period.2 Furthermore, calcification of the mitral annulus is often seen in Marfan patients in the chronic phase, making MVr difficult.4,7,8 Based on these previous findings and our results, we consider that MVr alone for 3 or 4+ MR and MVr with an aortic root procedure for 2+ or greater MR are reasonable in Marfan patients, which is in agreement with a previous report.7 However, there is a concern that the aortic cross-clamp time may be prolonged in cases of complex mitral valve lesions concomitant with valve-sparing aortic root surgery. A decision should be made whether to perform MVr and aortic root surgery simultaneously or whether 2-stage surgeries should be performed based on an overall consideration of an individual patient’s status.
Our results indicated a high proportion of bileaflet and anterior leaflet prolapse (64% and 25%, respectively) in Marfan patients compared with myxomatous mitral valve disease, which is consistent with a previous report.4 Bhudia et al. demonstrated that the mitral valve leaflet in Marfan patients was significantly longer and thinner than in patients with myxomatous disease.4 In patients with excessive leaflet tissue, like Marfan patients, an excessively small semirigid annuloplasty ring compared with the area of the anterior leaflet should be avoided to prevent SAM.25,26 Two cases of SAM were seen in the present study, both of which improved with medical treatment with no recurrence. The proportion of commissural closure was high in our study due to our performing commissural closure to prevent possible future regurgitation when the mitral valve in a commissural area was redundant. Further investigations are needed to determine whether this technique is effective in preventing the recurrence of MR.
In our study, 2 patients required reoperation for recurrent MR, one with active infective endocarditis and the other with anterior leaflet prolapse. The latter was a patient in whom an annuloplasty ring was not attached due to time constraints despite enlargement of the mitral annulus (intertrigonal distance 36 mm). The intertrigonal distance in this study was markedly enlarged, with a mean of 38.2 mm, suggesting that MAP is an indispensable technique for MR in Marfan patients to avoid recurrence.
The limitations of this study include its retrospective nature and the fact that it was a single-center study with a small number of patients. Approximately 40% of patients had a follow-up period of <5 years, which could influence long-term outcomes. Some echocardiographic data, especially at 6 months and 5 years after surgery, were missing, which could influence the assessment of cardiac reverse remodeling.
The mid- to long-term outcomes after MVr for MR in Marfan patients were satisfactory, supporting the durability of MVr in Marfan patients. Postoperative cardiac reverse remodeling occurred in a phased manner in Marfan patients similar to that seen in degenerative MR patients.
This study did not receive any specific funding.
M.O. is a member of Circulation Journal’s Editorial Team. The remaining authors have no conflicts of interest to declare.
This study was approved by the Research Ethics Committee of The University of Tokyo Hospital (Approval no. 3031-4; Approval date: 19 May, 2021).
The deidentified participant data will not be shared.
Please find supplementary file(s);
https://doi.org/10.1253/circj.CJ-24-0291
