|
|
||||||||
a Department of Surgery, Division of Cardiac Surgery, University of Michigan School of Medicine, Ann Arbor, Michigan
b Department of Pediatrics, Division of Pediatric Cardiology, University of Michigan School of Medicine, Ann Arbor, Michigan
c Cardiac Unit, Institute of Child Health, University College, London, United Kingdom
Accepted for publication October 21, 2009.
* Address correspondence to Dr Devaney, Department of Surgery, University of Michigan, 5143 Cardiovascular Center, 1500 E Medical Center Dr, Ann Arbor, MI 48109 (Email: edevaney{at}umich.edu).
Presented at the Fifty-fifth Annual Meeting of the Southern Thoracic Surgical Association, Austin, TX, Nov 5–8, 2008.
| Abstract |
|---|
|
|
|---|
Methods: A retrospective analysis was performed for all patients undergoing biventricular repair of atrioventricular septal defect with common atrioventricular valve and double-outlet right ventricle between 1991 and 2008. Patients with tetralogy of Fallot and common atrioventricular valve were excluded from analysis. Early and actuarial outcomes were evaluated using the
2 test for categorical variables and Wilcoxon rank sum for ordinal variables.
Results: The median age at operation was 16 months. Heterotaxy syndrome was present in 12 of the 16 patients (9 right isomerism and 3 left isomerism), and 6 had concurrent totally anomalous pulmonary venous connections. Primary repair was achieved in 6 patients, and 10 underwent one or more prior operations (most frequently a shunt, banding of the pulmonary trunk, or repair of the anomalous pulmonary venous connections). Enlargement of the ventricular septal defect by resection of the muscular outlet septum was required in 11 patients, in whom the ventricular septal defect emptied entirely or primarily to the inlet of the right ventricle. A conduit was placed from the right ventricle to the pulmonary arteries in 13. There was 1 death before discharge from hospital, 1 late death, and 2 episodes of heart block. Among survivors, follow-up was complete with a median follow-up of 66 months. No patient had late obstruction of the left ventricular outflow tract. The presence of heterotaxy with totally anomalous pulmonary venous connections was associated with combined mortality and significant morbidity (p = 0.008).
Conclusions: Although technically challenging, the surgical repair can be accomplished with acceptable early results. Heterotaxy syndrome, with concurrent anomalous connections of the pulmonary veins, represented the strongest identified risk factor for death or significant complication.
| Introduction |
|---|
|
|
|---|
|
| Patients and Methods |
|---|
|
|
|---|
2 for categorical variables and Wilcoxon rank sum for ordinal variables), and a p value less than 0.05 was used to determine significance.
Operative Procedures
All patients undergoing biventricular repair are placed on cardiopulmonary bypass with direct cannulation of the caval veins and moderate hypothermia. Deep hypothermia with low-flow cardiopulmonary bypass or brief circulatory arrest is utilized during the repair of totally anomalous pulmonary venous connection, when present, to improve exposure. After arrest of the heart, exposure of the common atrioventricular valve is achieved through either a right- or left-sided atriotomy, depending on anatomy. Repair of the defect is initially analogous to a two-patch repair of atrioventricular septal defect. A crescent shaped patch of polytetrafluoroethylene is used to close the inlet component of the interventricular communication, starting at the midpoint of the defect, and carrying the suture line inferiorly and posteriorly to the junction of the muscular ventricular septum with the annulus of the common atrioventricular valve (Fig 2). The anterior and superior portion of the patch is left unsutured. Marking stitches are placed in the patch superiorly where it would join the atrioventricular valvar annulus, and at the midpoint of the inlet component of the ventricular defect. Next, the corresponding portions of the superior and inferior bridging leaflets are secured to the crest of the patch, and the zone of apposition between the left ventricular components of the bridging leaflets is closed with fine interrupted sutures.
|
|
|
|
| Results |
|---|
|
|
|---|
|
|
Biventricular repair was successfully accomplished in all 16 patients. The median cardiopulmonary bypass time was 242 minutes, with a median cross-clamp time of 158 minutes. Enlargement of the interventricular communication was required in 11 patients. A conduit was placed from the right ventricle to the pulmonary arteries in 13 patients, and 1 other patient with distortion of the pulmonary arteries subsequent to banding had the right ventricular outflow tract patched with a monocusp valve. The conduit was selected on the basis of size and availability, with bovine jugular venous xenografts implanted in 7 cases, pulmonary allografts in 4, an aortic allograft in 1, and a porcine heterograft in 1. Heart block developed postoperatively in 2 patients, who required implantation of a pacemaker. Tachyarrhythmias were common postoperatively, with 8 patients having transient junctional ectopic tachycardia. There has been 1 early death, which was secondary to pulmonary vascular obstructive disease, and 1 late death at 3 months due to complications related to renal failure. In addition, 2 surviving patients experienced significant early morbidity. One patient required extracorporeal membrane oxygenation postoperatively because of severe hypoxemia. A second patient had a prolonged stay in intensive care due to dependence on a ventilator and hydrocephalus. At the time of discharge, 5 patients had moderate atrioventricular valvar insufficiency, this being mild or trivial in the remainder. No patients had atrioventricular valvar stenosis, significant residual interventricular communications, or obstruction of the left ventricular outflow tract.
At a median follow-up of 66 months, with a range from 1 month to 17 years, all discharged patients were clinically well, in the first or second classes of the New York Heart Association functional class system. Reoperation has been necessary in 6 patients. The patient with the monocusp patch ultimately required insertion of a pulmonary valve 8 months after biventricular repair. This patient also underwent concomitant mitral and tricuspid valvoplasty. An additional patient has required replacement of the left atrioventricular valve, and implantation of a pacemaker 3 months after repair. Replacement of the conduit has been performed in 2 patients to date. Another patient had partial dehiscence of the patch closing the interatrial defect, which required reoperation. An unusual patient, who had previously undergone a modified Fontan procedure before biventricular repair, required reoperation for placement of an absorbable band on the left pulmonary artery owing to the presence of significant left-sided pulmonary arteriovenous malformations. A subsequent catheterization in this patient revealed resolution of the arteriovenous malformations and no pulmonary artery stenosis. At follow-up, 3 patients had mild-to-moderate left or right atrioventricular valvar insufficiency, or both, and the remainder had mild or less insufficiency. No patient had late obstruction of the left ventricular outflow tract.
A series of clinical characteristics were evaluated to determine risk factors for poor outcome. The only significant risk factor for mortality and significant morbidity was the presence of totally anomalous pulmonary venous connection (p = 0.008). Nonsignificant factors included age at operation, prior procedure, heterotaxy syndrome without anomalous pulmonary venous connections, form of isomerism, and presence of dextrocardia.
| Comment |
|---|
|
|
|---|
The repair differs from that for tetralogy of Fallot associated with common atrioventricular valve primarily because of the exclusively right ventricular origin of the aorta, with its valve supported by a muscular infundibulum. That creates the difficulty of redirecting the outflow from the left ventricle through the interventricular communication to the aortic valve. Even when the communication opens toward the outlet regions, it is frequently difficult to avoid obstruction of the newly created channel from the left ventricle without enlarging the interventricular communication. In these cases, some groups have recommended functionally univentricular palliation. Recently, the outcomes have been described for nearly 100 patients with complex double-outlet right ventricle, either associated with heterotaxy or a common atrioventricular valve [2]. In this series, only 8 patients underwent biventricular repair, whereas 88 were managed using the functionally univentricular strategy. The authors identified atrioventricular valvar regurgitation, pulmonary venous obstruction, and neonatal presentation as risk factors for mortality. The decision to proceed with univentricular palliation for atrioventricular septal defect with common atrioventricular valve and double-outlet right ventricle must be tempered by the understanding that many patients may be poor candidates for the Fontan procedure owing to pulmonary artery stenosis, pulmonary hypertension, pulmonary venous obstruction, atrioventricular valve insufficiency, or other factors. In fact, 1 patient in our series had previously undergone Fontan palliation but was referred for conversion to biventricular repair because of chronic hypoxemia, caused by unilateral pulmonary arteriovenous malformations that developed as a consequence of diversion of hepatic venous blood away from that lung by the existing cavopulmonary connection.
We identified heterotaxy in 12 of our patients. This diagnosis was generally made by the surgeon or cardiologist, using a number of criteria including splenic anatomy, bronchopulmonary anatomy, venoatrial connections, and morphology of the atrial appendages. The most accurate method of determining the isomerism, which is the best means of stratifying heterotaxy, is by analysis of the pectinate musculature of the atrial appendage, but this cannot generally be determined by echocardiography, and is not always assessed by the surgeon in the operating room [11–13]. Both forms of isomerism are associated with a spectrum of complex cardiac malformations, and the complete surgical repair or palliation of these defects is accompanied by an extremely high mortality rate [14, 15]. Additionally, right isomerism is always linked to anomalous pulmonary venous connection, even if the pulmonary veins return directly to the heart, as such connections cannot be anatomically normal. The repair is often complicated by the development of pulmonary venous stenosis, which carries a poor prognosis [16]. We also found that the anomalous pulmonary venous connections associated with the heterotaxy syndrome proved to be a risk factor for poor outcome.
In our series, early and late obstruction within the newly created subaortic outflow tract were avoided by aggressive enlargement of the interventricular communication, which was required in most cases. Intraoperative echocardiography and direct intraventricular pressure measurement were utilized to assess the repair and adequacy of the left ventricular outflow tract. Transection of the aorta to examine the left ventricular outflow tract was not used in this series but may be useful in selected difficult cases. As illustrated in Figure 4, resection of the muscular outlet septum and subaortic infundibular musculature creates a wide communication with the left ventricle. In those patients with left-hand topology, or L-looping, the pathway of the conduction system would be expected to lie in the septum that is being resected, resulting in unavoidable complete heart block, although this was seen in only 1 patient in this series. A ventriculotomy is generally required to provide adequate visualization for this maneuver and, owing to the frequent occurrence of pulmonary stenosis in these patients, the right ventriculotomy is also necessary for placement of a conduit. Others reporting biventricular repair for this combination have similarly emphasizes the need for enlargement of the interventricular communication into the muscular outlet septum [10, 17].
The technical complexity of complete repair in these patients is frequently further complicated by the need to place a complex patch to separate the systemic and pulmonary venous return. In the presence of heterotaxy syndrome, bilateral superior caval veins, separate drainage of the hepatic and inferior caval veins, and anomalous pulmonary venous connections all serve to make septation of the atrioventricular channels challenging. Use of autologous or allograft pericardium facilitates the placement of a Mustard-like patch with the atrium, and obstruction to venous return has been rare in our patients. In patients with low pulmonary blood flow, the atrial volume is frequently small, further compromising repair.
Although this series represents the largest reported group of patients undergoing biventricular repair of the combined lesions, the rarity of the lesion still results in a relatively small number of patients, and extensive statistical analysis is not possible. We submit, nonetheless, that our results support the approach of biventricular repair, even in the presence of complex venous connections and a remote interventricular communication. Although the rate of reoperation will be high, particularly for replacement of the conduit, it is hoped that this approach will offer long-term benefits when compared with functionally univentricular palliation. More extensive follow-up of this cohort of patients will be important to justify this strategy.
| Discussion |
|---|
|
|
|---|
One, regarding the VSD enlargement, how do you really decide on the need for this, as it wasn't done in all patients, and in our limited experience with an inlet VSD with extension to the outlet septum, we have not had to use enlargement techniques.
Two, could you expand on the rationale for a two-patch or composite LVOT/VSD patch as opposed to the more commonly described comma-shaped patch that is used in this entity?
In summary, I enjoyed the paper. It was clear and concise. Very good results with a rare and difficult subset of patients ... and thanks for sending me the manuscript in advance.
DR DEVANEY: Jim, thank you very much for your kind comments and your questions. The first question regarded how do we decide which patients need to have their VSDs enlarged, and I think really you highlighted the issue. It really depends on how much outlet extension there is. In our patient series, certainly the majority of the patients who needed VSD enlargement, we felt needed VSD enlargement, was because the defect was so remote from the aorta that we felt that we couldn't easily create the baffle without precipitating LV outflow tract obstruction. So I can't say in any more detail how we do that. It is just something that we have gotten more comfortable with as we have done more of these cases. We are confident that if we do the muscle resection that we can create an unobstructed outlet. And so that is why it has become a more frequent technique that we have used, especially more recently. As I mentioned, I think 8 of the last 8 we have used this technique.
Now, your second question involved why do we do two separate VSD patches rather than just one. For the patients with tetralogy of Fallot with complete atrioventricular septal defect, I think it is easy enough to see the entire VSD transatrially and to cut the patch accurately. For this lesion, I can't see it well enough. I think it is difficult to see the aortic valve transatrially, and that is why we use a two-patch approach. So this is just a technical feature that makes it easier to achieve consistent good results.
DR CHARLES D. FRASER (Houston, TX): That is a wonderful series and a testimony to excellent technical surgery. However, 2 of your 16 patients died, and I would predict that over that same time period in your institution, the survival for Fontan was virtually 100%. So is this really better than a Fontan? And in an individual patient, how would you justify a more high risk approach like this as opposed to a Fontan operation?
DR DEVANEY: Well, I think you have to look at the denominator and decide. We are not just comparing it to Fontans. We are looking at Fontans in heterotaxy patients, especially those with total anomalous venous connections. I think the risks are higher for that patient group. I think that turns out to be a risk factor for almost any analysis we do, whether it is a single ventricle or a biventricular approach to any surgical lesion. The presence of heterotaxy syndrome and total veins tends to cloud the issue. The deaths and complications are occurring in those patients. So I think we will have tough results no matter which approach we go with. But I think your question is fair, and I really think we need to follow these patients longer term to see whether it is really a justifiable strategy.
DR HARALD L. LINDBERG (Oslo, Norway): Very nice paper, nice results. I just wondered, were there any patients with Down syndrome in this series? Probably not.
DR DEVANEY: There were not.
DR LINDBERG: What was your technique on the zone of opposition or the cleft? Was that closed in all patients?
DR DEVANEY: The zone of opposition was closed, correct.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. A. Jonas Surgical Management of the Neonate With Heterotaxy and Long-Term Outcomes of Heterotaxy World Journal for Pediatric and Congenital Heart Surgery, April 1, 2011; 2(2): 264 - 274. [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 |