|
|
||||||||
Ann Thorac Surg 2002;74:786-791
© 2002 The Society of Thoracic Surgeons
a Service de Cardiologie Pédiatrique, Hôpital Necker Enfants Malades, Paris, France
b Service de Chirurgie Cardiaque, Hôpital Necker Enfants Malades, Paris, France
c Service de Chirurgie Thoracique et Cardiovasculaire, Hôpital dEnfants de la Timone, Marseille, France
d Laboratoire de Biomathématiques, Université de la Méditerranée, Marseille, France
Accepted for publication April 30, 2002.
* Address reprint requests to Dr Fraisse, Service de Cardiologie A, Hôpital de la Timone, 13005 Marseille, France
e-mail: afraisse{at}mail.ap-hm.fr
| Abstract |
|---|
|
|
|---|
Methods. A retrospective review identified 21 patients operated upon with mitral valve cleft, abnormal ventriculo-arterial connections, and two well-developed ventricles. Eight patients had a ventricular outflow obstruction due to the mitral valve, whereas 2 had more than mild mitral regurgitation. One patient required initial mitral valve surgery. Eleven patients underwent biventricular repair, associated with mitral valve repair in 2 cases: arterial switch operation (n = 4), Senning operation (n = 3) associated with an arterial switch operation in one case, intraventricular repair (n = 3), and Rastelli-type extracardiac conduit repair (n = 1). Single-ventricle palliation was preferred in 10 patients with major mitral valve straddling (n = 5), outflow tract obstruction (n = 2), and noncommitted or multiple VSDs (n = 3).
Results. There were three hospital deaths, two of which occurred after biventricular repair and one after an early reoperation after a bidirectional cavopulmonary anastomosis. Postoperatively after biventricular repair, 1 patient required permanent pacemaker implantation and 3 patients were reoperated on for subaortic stenosis (n = 1) and mitral regurgitation (n = 2), with one late death. By multivariate analysis, patients with a double-outlet right ventricle were at greater risk of death (p = 0.04). After a mean follow-up period of 60.7 months (± 68.6 months), 16 patients are in New York Heart Association (NYHA) class I. One patient with a moderate mitral regurgitation on Doppler study is in NYHA class II.
Conclusions. The surgical management remains controversial in patients with abnormal ventriculo-arterial connections and mitral valve cleft. Biventricular repair may not always be feasible, especially in cases of complex intracardiac anatomy associated with mitral valve straddling. Single-ventricle palliation can be achieved in these patients, although it is unknown whether the long-term results are as good as those obtained with biventricular repair.
| Introduction |
|---|
|
|
|---|
| Material and methods |
|---|
|
|
|---|
Patient characteristics
A total of 21 patients with VA discordance, cleft MV, and two well-developed ventricles were diagnosed. Demographic and morphologic features are summarized in Table 1.
Sixteen patients presented as newborn. The mean age and weight at presentation were 12 months (± 48.7 months) and 5.7 kg (± 7.4 kg), respectively. All patients presented with cyanosis or tachypnea. Physical examination revealed a systolic murmur in 17 patients. The electrocardiogram (ECG) was in sinus rhythm in all patients. The chest roentgenogram showed a cardiomegaly with features of pulmonary congestion in 7 patients. The diagnosis of abnormal VA connection was a double-outlet right ventricle (DORV) in 4 patients and a transposition of the great arteries (TGA) in 15 cases. In the 2 remaining patients, an AV discordance (L-loop) was associated with a TGA. A conoventricular ventricular septal defect (VSD) was present in all the patients: subpulmonary (n = 14), subaortic (n = 5), or noncommitted (n = 2). In 5 cases, the VSD was restrictive. Two patients had additional muscular VSDs. A tricuspid valve straddling through an inlet VSD was found in 2 patients.
|
Echocardiography revealed in all 21 patients a cleft dividing the anterior (aortic) leaflet of MV into 2 with a mild MR through the cleft in 6 patients, a moderate to severe MR in 2, and no significant regurgitation in 13. In 10 cases with MV straddling, the cleft crossed the septal crest to insert into the right ventricle through tendinous cords (Fig 1). In 5 cases, the MV straddling was found to be major. Three patients had a subaortic obstruction, whereas 14 had an obstruction of the pulmonary outflow tract (Table 1). In 8 patients, this outflow obstruction was partially or totally due to the chordal attachments of the MV.
|
At the time of biventricular repair, the MV was repaired in two cases with MV straddling. In 1 patient with a DORV, major MV straddling, and subaortic obstruction due to the MV, the accessory MV chordae were divided in the right ventricle and the conal septum was excised before a left ventricle-to-aorta pericardial baffle was constructed. In another patient with TGA and minor MV straddling, the MV chordae adjacent to the cleft were divided in the right ventricle, attached onto the left-ventricular septal surface, and an arterial switch operation was performed.
Ten patients had various single ventricle palliations, including seven cases with MV straddling. In five of the seven cases, the MV straddling was major and precluded biventricular repair by itself. In one of them, the amount of straddling was initially considered as minor on multiple echocardiographic views (Fig 1), and an intraventricular repair was planned. However, intraoperatively, the MV straddling was major and it was decided to perform a bidirectional cavopulmonary anastomosis. In 2 patients with minor MV straddling and subpulmonary obstruction due to the MV, single-ventricle palliation was preferred because it was not believed that an unobstructed left ventricle-to-aortic connection could be established. In the 3 remaining patients, the reason for choosing single-ventricle palliation was a noncommitted VSD in two cases and the presence of multiple apical VSDs in one case.
A bidirectional cavopulmonary anastomosis was performed in 7 patients, after a right modified Blalock-Taussig shunt in 1 patient, and after coarctation repair with a pulmonary artery banding in another patient. One patient had an additional left modified Blalock-Taussig shunt 4 years after a bidirectional cavopulmonary anastomosis, because of moderate cyanosis. Two patients had only systemic-to-pulmonary artery shunts, including a right modified Blalock-Taussig shunt followed by a left modified Blalock-Taussig shunt in 1 patient, and a left modified Blalock-Taussig shunt alone in another patient. One patient had a pulmonary artery banding. The mean age at last palliative surgery was 27.6 months (± 44.6 months) No patient was scheduled for a modified Fontan procedure after the last palliative surgery.
Statistical analysis
Group statistics are expressed as mean ± SD. To identify the risk factors for single-ventricle palliation and death, we used Fishers exact test for univariate analysis and logistic regression for multivariate analysis. Survival was estimated by the Kaplan-Meier method for qualitative variables and compared with log-rank test. Cox proportional Hazard model was used for quantitative variables. Cox proportional hazard regression was used to estimate the risk ratio for death in multivariate analysis. A p less than 0.05 was considered significant.
| Results |
|---|
|
|
|---|
One late death occurred in a patient with TGA and minor MV straddling, 4 months after an arterial switch operation with MV chordae reinsertion was performed. Postoperatively, he had congestive heart failure due to severe MR with an anterior MV prolapse by echocardiography. He was reoperated on 15 days after the biventricular repair. Rupture of chordae reinsertion was found together with a prolapse of the lateral part of the anterior leaflet. Mitral valvuloplasty was attempted with suture of the prolapsed part of the anterior leaflet to the adjacent posterior leaflet, and a partial annuloplasty. Postoperatively, after transient improvement, he again experienced heart failure with worsened MR. Intraoperatively during a second reoperation, a recurrence of the anterior leaflet prolapse was found, with elongated chordae. An Alfieri-type repair was attempted with anterolateral commissure plication annuloplasty. Although the MV appeared to be competent after saline solution injection in the left ventricle, the patient could not be weaned from bypass in the setting of severe left-ventricular dysfunction.
There were 2 other reoperations after biventricular repair. Five years after an arterial switch operation, a patient underwent MV cleft suture associated with mitral annuloplasty for severe MR. Resection of conal septum for residual subaortic stenosis was performed in another patient who had a left ventricle-to-aortic valve intracardiac baffle. Additionally, 1 patient underwent pacemaker implantation for postoperative complete heart block, after an atrial switch operation with VSD closure
The estimates of survival are 73% and 90% in patients with biventricular repair and single-ventricle palliation, respectively (p = 0.4). In multivariate analysis, there was no clinical or morphologic feature associated with single-ventricle palliation, although there was a borderline relationship for MV straddling (p = 0.06). Patients with a DORV were at greater risk of death (p = 0.04, Table 2). After a mean follow-up period of 60.7 months (± 68.6 months), 16 patients are all in New York Heart Association (NYHA) class I. One patient with a moderate MR by color-Doppler is in NYHA class II. Four other patients have a mild residual MR, whereas there is no MR in the remaining patients. No left-ventricular outflow tract obstruction due to the MV chordae occurred in the patients who did not have obstruction at the time of diagnosis.
|
| Comment |
|---|
|
|
|---|
Echocardiographic considerations
Although MV cleft is a rare anomaly, its frequency in patients with abnormal VA connections emphasizes the importance of systematic echocardiographic examination of the MV from multiple acoustic windows and views in these patients. In our study, the MV cleft was diagnosed by cross-sectional echocardiography, aided by color Doppler for assessment of MV function. However, in one case, the cleft was not initially seen and was only diagnosed 5 years after the arterial switch operation in the setting of congestive heart failure with severe MR. In another case, the amount of straddling was considered as minor on multiple echocardiographic views (Fig 1). However, intraoperatively, the MV straddling was major and associated with a muscular subpulmonary stenosis. A bidirectional cavopulmonary anastomosis was performed instead of an intraventricular repair.
We use mainly as diagnostic criteria the direction of the cleft toward the aortic root with attachment of the cleft edges to or through the ventricular septum by chordae tendinae, associated with a normal size and position of the left-ventricular papillary muscles and a normal ratio (
1) between the inlet and outlet chamber of the left ventricle [1, 4, 5]. In young patients, the subcostal view demonstrates the MV and its support apparatus, as well as simultaneous cross-sectional images of the ventricles and their outflow tract (Fig 1A). The VSD relation to the semilunar valves and the potential for a baffle repair from the left ventricle to the aorta can be seen from this view. The apical four-chamber view is useful to determine the extent of a MV overriding [6]. The parasternal short-axis view helps to visualize the direction of the cleft and the location of the papillary muscles (Fig 1B). Transesophageal echocardiography may be helpful in assessing intraoperatively the success of the MV repair but only gives little additional morphologic information, as noted by others [4]. The potential benefit of three-dimensional echocardiography remains undetermined and deserves further study.
Surgical considerations
As it is for other forms of VA discordance with complex intracardiac anatomy, the surgical management of patients with DORV or TGA with MV cleft remains controversial. Previous studies have demonstrated that some associated anomalies in VA discordance are risk factors for death or reoperation after biventricular repair, like the presence of a noncommitted VSD [11], multiple VSDs [12], common AV canal [13], aortic arch obstruction [12], and congenital MV anomaly [8]. Nevertheless, biventricular repair always seems preferable to a single-ventricle palliation [8, 14, 15], because the analysis of the long-term results of the Fontan operation demonstrates a late decline in functional status [15]. In our institutions, a partial cavo-pulmonary connection, with or without an additional systemic-to-pulmonary artery shunt, is sometimes considered as a definitive palliative procedure. This strategy seems to avoid the serious complications of Fontan-like circulation and provides mild cyanosis with little functional impairment [16]. However, single-ventricle repair with a modified Fontan procedure has been advocated by others because of a lower early mortality than biventricular repair for patients with complex forms of DORV, including hypoplastic right ventricle, AV discordance, straddling AV valves, noncommitted or multiple VSDs, crisscross AV relations, AV septal defect, pulmonary atresia, and anomalies of pulmonary artery branches [12, 17]. In our patients, the difference of mortality between single-ventricle palliation and biventricular repair did not reach significance (p = 0.4). The result of our multivariate analysis showing that a DORV was significantly associated with mortality (p = 0.04) may reflect only a more complex anatomy than TGA and does not preclude any type of repair. In the 2 patients with a DORV who subsequently died, 1 had single ventricle palliation whereas the other underwent biventricular repair.
Technically, as all the patients had two ventricles of normal size, discussion as to whether a biventricular repair is feasible in these patients will revolve around the surgical possibilities to perform an arterial switch operation or to baffle the LV outflow through the VSD to the closest semilunar valve without distorting the chordal apparatus of the MV and without obstructing LV outflow [14]. Although it does not preclude biventricular repair [14], MV straddling represents a somewhat more severe lesion than a simple cleft. Techniques have been described for repairing when the degree of MV straddling is relatively minor [14]. Frequently, enlargement of the VSD is necessary to avoid subaortic obstruction from the chordal apparatus crossing the VSD. In patients with DORV, the VSD patch is usually placed around the chordal insertion site. In TGA with MV cleft and straddling, the chordae can be retracted flush with the edge of the VSD and they are then fixed in that position by suturing of the VSD patch onto the crest of the septum. There were 7 patients with MV straddling in the single-ventricle group versus 3 patients in the biventricular group. Our patients with MV straddling had a tendency to undergo single-ventricle palliation rather than biventricular repair (p = 0.06). Also, the 2 patients with MV straddling who had MV surgery at the time of biventricular repair subsequently died, 1 at the time of biventricular repair (failure to wean from bypass) and the other during reoperation for severe residual MR. Single-ventricle palliation could have probably been achieved in these two cases.
Study limitations
This study has the inherent limitations of its retrospective nature. The number of patients having abnormal VA connections associated with a cleft MV may be greater than we found, as this anomaly is not always recognized, especially in the absence of MR or outflow obstruction. The surgical management of the patients was decided on a case-by-case basis in which they were not randomly allocated to single versus biventricular repair. Hence, it is not possible to know whether some of the patients who underwent a single-ventricle palliation would have survived a biventricular repair.
We conclude that biventricular repair in patients with abnormal VA connection and MV cleft is not always feasible, especially in cases of complex intracardiac anatomy: DORV with MV straddling, left-ventricular outflow tract obstruction, and noncommitted or multiple VSD. Single-ventricle palliation can be achieved with good short- and intermediate-term results in these patients. However, it is unknown whether the long-term survival and functional status of single-ventricle palliation in patients with two well-developed ventricles are as good as those obtained with a biventricular repair.
| References |
|---|
|
|
|---|
ur 1998;91:569-573.
This article has been cited by other articles:
![]() |
S. Abadir, V. Fouilloux, D. Metras, O. Ghez, B. Kreitmann, and A. Fraisse Isolated Cleft of the Mitral Valve: Distinctive Features and Surgical Management Ann. Thorac. Surg., September 1, 2009; 88(3): 839 - 843. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Van Praagh, D. Porras, G. Oppido, T. Geva, and R. Van Praagh Cleft mitral valve without ostium primum defect: anatomic data and surgical considerations based on 41 cases Ann. Thorac. Surg., June 1, 2003; 75(6): 1752 - 1762. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |