|
|
||||||||
Ann Thorac Surg 2002;74:805-810
© 2002 The Society of Thoracic Surgeons
a Division of Cardiothoracic Surgery, Medical University of South Carolina, Charleston, South Carolina, USA
b Division of Pediatric Cardiology, Medical University of South Carolina, Charleston, South Carolina, USA
c Division of Anesthesia, Medical University of South Carolina, Charleston, South Carolina, USA
* Address reprint requests to Dr Bradley, Division of Cardiothoracic Surgery, Medical University of South Carolina, 96 Jonathan Lucas St, Charleston, SC 29425 USA
e-mail: bradlesm{at}musc.edu
Presented at the Forty-eighth Annual Meeting of the Southern Thoracic Surgical Association, San Antonio, TX, Nov 810, 2001.
| Abstract |
|---|
|
|
|---|
Methods. From January 1996 to June 2001, 22 infants were treated by this strategy. Patients with hypoplastic left heart syndrome were excluded. Median age was 18 days (range 2 days to 6 months). In addition to pulmonary artery division and shunt, 3 of 22 patients underwent a Damus-Kaye-Stansel procedure, and 13 of 22 patients underwent a modified Norwood procedure.
Results. There were no operative deaths, and one late death. Actuarial survival beyond 30 months was 90%. At follow-up catheterization in 22 patients, median transpulmonary gradient was 7 mmHg (range 4 to 18), and median pulmonary vascular resistance 1.9 Wood units (range 0.9 to 3.3). Twenty-one patients have undergone a subsequent bidirectional superior cavopulmonary connection, and 6 a Fontan procedure, with no deaths. No patient developed subaortic stenosis, or aortic arch obstruction. Neoaortic insufficiency was none or trivial in 12 patients, mild in 3, and moderate in 1.
Conclusions. In patients with a functional single ventricle and excessive pulmonary flow, a strategy of pulmonary artery division and shunt, along with prophylactic bypass of systemic outflow obstruction, carries low operative and midterm mortality. It provides consistent protection of the pulmonary vascular bed, avoids subaortic stenosis and aortic arch obstruction, minimizes neoaortic insufficiency, and ensures suitability for progression along a Fontan pathway. These results provide a comparison for alternate strategies, including pulmonary artery banding.
| Introduction |
|---|
|
|
|---|
Patients whose pulmonary blood flow is provided by a systemic-to-pulmonary artery shunt generally demonstrate low pulmonary vascular resistance, and good candidacy for progression along a Fontan pathway [11]. Therefore, in patients with a single ventricle and excessive pulmonary flow, we have pursued a strategy which includes division of the pulmonary artery, and placement of a systemic-to-pulmonary artery shunt. Any actual or potential sites of systemic outflow tract obstruction are simultaneously bypassed, by either a Damus-Kaye-Stansel, or modified Norwood procedure. The aims of this report are to assess the results of this strategy on operative and midterm survival, progression along a Fontan pathway, pulmonary vascular protection, systemic outflow tract obstruction, and neoaortic insufficiency.
| Material and methods |
|---|
|
|
|---|
|
Systemic-to-pulmonary artery shunts of polytetrafluoroethylene (Gore-Tex; W.L. Gore and Assoc, Flagstaff, AZ) were placed from the innominate artery (21 patients) or ascending aorta (1 patient) to the central pulmonary artery confluence or origin of the right pulmonary artery. Shunt diameters were 3.5 mm (12 patients), 4 mm (6 patients), or 5 mm (4 patients). Generally, 4-mm shunts were used in patients over 3.5 kg, and 5-mm shunts over 4.5 kg. In an effort to avoid shunt thrombosis, all patients were administered intravenous heparin once operative bleeding had stopped (target partial thromboplastic time 60 to 80 seconds). When enteric intake resumed, heparin was converted to aspirin, which was continued indefinitely. Associated procedures performed were atrial septectomy in 12 patients, repair of total anomalous pulmonary venous connection in 2, and resection of accessory atrioventricular valve tissue in 1 (Table 1). Cardiopulmonary bypass, cardioplegic myocardial arrest, and periods of circulatory arrest were used in 21 of the 22 operations. The average cardiopulmonary bypass time (mean ± SD) was 164 ± 52 minutes, cross-clamp time 46 ± 17 minutes, and circulatory arrest time 32 ± 21 minutes.
Subsequent bidirectional superior cavopulmonary connection (BSCC) consisted of a bidirectional Glenn shunt in 16 patients, hemi-Fontan procedure in 4, and Kawashima procedure in 1. Additional procedures at the time of BSCC were connection of a left superior vena cava to the left pulmonary artery (left bidirectional Glenn shunt) in 5 patients, atrioventricular valve repair in 2, reimplantation of an anomalously draining right upper pulmonary vein in 1, and patch closure of an insufficient left atrioventricular valve in 1. Completion Fontan procedures were fenestrated intra-atrial lateral tunnel connections in 4 patients, and extracardiac conduits in 2.
Data collection/follow-up
All records were retrospectively reviewed. Cardiac catheterization was performed before each BSCC, and again before each Fontan procedure. Subaortic stenosis or aortic arch obstruction during follow-up were defined as a gradient of 20 mmHg or greater by echocardiogram or catheterization. Neoaortic insufficiency and atrioventricular valve regurgitation were qualitatively graded by echocardiogram on a scale from 0 (none) to 4 (severe). Ventricular function was qualitatively graded by echocardiogram and angiography. Actuarial survival curve was generated using NCSS 2000 for Windows (Number Cruncher Statistics Systems, Kaysville, UT). All patients were followed within the Pediatric Heart Center of South Carolina, a pediatric cardiology network providing coordinated congenital cardiac care to the state.
| Results |
|---|
|
|
|---|
|
Pulmonary vascular protection
All patients have undergone cardiac catheterization in preparation for BSCC (Table 2).
A single patient had elevated pulmonary pressure (32 mmHg) and transpulmonary gradient (18 mmHg). This patient, with unbalanced atrioventricular septal defect and trisomy 21, presented late. He underwent initial palliation at 3.4 months of age, with pulmonary artery ligation and placement of a 5-mm shunt. At subsequent catheterization, his pulmonary-to-systemic flow ratio (Qp:Qs) was 2.6, so that his calculated pulmonary vascular resistance was low (2.2 Wood units), in spite of elevated pulmonary pressure. He has not yet undergone BSCC.
|
Systemic outflow tract obstruction; valvular insufficiency
No patient developed either subaortic stenosis or aortic arch obstruction, as assessed by either echocardiography or catheterization. At most recent follow-up, atrioventricular valve insufficiency was grade 0 to 1 (none to trivial) in 9 patients, grade 2 (mild) in 12 patients, and grade 3 (moderate) in 1 patient. Ventricular function was normal in 19, and mildly depressed in 3 patients. In the 16 patients who underwent either a Damus-Kaye-Stansel or modified Norwood procedure, function of the pulmonary (neoaortic) valve was assessed during follow-up by echocardiography. The 3 patients who had a Damus-Kaye-Stansel procedure had no neoaortic insufficiency. The 13 Norwood patients had grade 0 to 1 (none to trivial) in 9 patients, grade 2 (mild) in 3 patients, and grade 3 (moderate) in 1 patient.
| Comment |
|---|
|
|
|---|
The use of a pulmonary band in patients with a functional single ventricle and excessive pulmonary flow can be complicated by the occurrence of subaortic stenosis. Freedom and coworkers observed the development of subaortic stenosis in 72% of patients with a univentricular heart treated with a pulmonary band [2]. Subaortic stenosis can result in ventricular hypertrophy, which is a risk factor for a Fontan operation [9, 10]. The factors leading to subaortic stenosis have been extensively and expertly discussed [3, 8]. Morphologic factors include origin of the aorta from an outlet chamber which is supplied by a bulboventricular foramen or ventricular septal defect. Matitiau and associates further refined this observation by determining that a bulboventricular foramen size of less than 2 cm2/m2 is associated with the later development of subaortic stenosis [14]. The presence of aortic arch obstruction is also associated with both a small bulboventricular foramen and the subsequent development of subaortic stenosis [14]. Several groups have reported that subaortic stenosis developed in 100% of survivors of palliation by a pulmonary band and aortic arch repair [4, 7, 13].
More recent series have reported improved results of pulmonary artery banding in patients with a functional single ventricle [1518]. The approach taken in these series has been to leave the band in place for a short time, and carefully follow the patients for the development of subaortic stenosis. The operative mortality of banding, with or without concomitant arch repair, was low, ranging from 0% to 8% [1618]. However, subaortic stenosis occurred frequently, being seen in 64% to 100% of patients at a median interval of 3 to 8 months [15, 17], and within 1 month in several cases [1618]. Thus, frequent early reoperation for subaortic stenosis remains a drawback to pulmonary banding.
The alternative of prophylactically bypassing the subaortic area in patients at risk for subaortic stenosis has been advocated for some time [1, 1921]. By avoiding subaortic stenosis, rather than treating it once it has occurred, this approach has the potential of avoiding ventricular hypertrophy, thereby improving suitability for a Fontan procedure [9, 10]. The main drawback to this approach is a complex operation utilizing cardiopulmonary bypass, and potentially circulatory arrest, in a neonate. Several groups pioneered this approach in the late 1980s and early 1990s [1, 14, 19, 20]. More recently, operative mortalities below 20% have been reported for Norwood procedures carried out for defects other than hypoplastic left heart syndrome [2225]. Mosca and associates reported the use of a modified Norwood procedure in 38 patients with a single left ventricle and ventriculoarterial discordance [12]. In this fairly uniform group, there were only three early deaths (8%), and five late deaths (overall mortality 21%). Our results further support the use of pulmonary artery division, shunt, and prophylactic bypass of systemic outflow tract obstruction as a management strategy.
In patients who are not at risk for subaortic stenosis, the distinction between the current strategy and pulmonary artery banding is less clear. Placement of a pulmonary band is a relatively "simple" operation, which can be performed without cardiopulmonary bypass. The use of a pulmonary band in this situation has been suggested by groups which advocate a Damus-Kaye-Stansel or Norwood approach in patients at risk for subaortic stenosis [25, 26]. However, there is little data in the literature on the results of pulmonary banding in patients without systemic outflow obstruction [1, 7]. There are also several arguments favoring pulmonary artery division and shunt over a pulmonary band. In the operating room, it can be difficult to predict the eventual effectiveness of a pulmonary band. Following banding, the pulmonary artery can remodel, resulting in a lower gradient than intended. A band is adjusted under nonphysiologic conditions, including general anesthesia, positive pressure ventilation, and an open chest. Under physiologic conditions several days later, the band gradient may well be different [27]. Furthermore, when a band is placed in a neonate, pulmonary vascular resistance can be expected to fall over the following days or weeks, also affecting the band gradient. In contrast, a shunt has a uniform diameter, and is not generally adjusted in the operating room. A great deal of experience with the Norwood operation is now available to guide the selection of shunt size in neonates. Follow-up catheterization in a large number of patients with shunts placed in the neonatal period has confirmed consistent pulmonary vascular protection [11]. Of interest, in the Boston Childrens Hospital experience, 17% of patients who had a pulmonary band required subsequent division of the pulmonary artery and substitution of a shunt [1]. Pulmonary banding may well be considered in selected patients. However, we have been pleased with the results of an initial palliation which results in a shunt as the sole source of pulmonary blood flow.
Neoaortic insufficiency is a potential concern following any operation which places the pulmonary valve in the aortic position [28]. Significant insufficiency has been unusual following a primary Damus-Kaye-Stansel or modified Norwood procedure [12, 22, 25, 28]. Our results are in agreement, with 1 of 16 patients having more than mild neoaortic insufficiency during follow-up. Several groups have examined neoaortic valve function in patients who have had a Damus-Kaye-Stansel procedure after a previous pulmonary band [18, 28, 29]. Jenkins and colleagues observed moderate insufficiency in 2 of 13 patients; Amin and coworkers found moderate insufficiency in 1 of 15 patients; Daenen and associates noted more than mild insufficiency in 0 of 12 patients [18, 28, 29]. Significant pulmonary valve damage by a band is thus unusual. However, there have been anecdotal reports of the pulmonary valve being so damaged by a band, that a planned Damus-Kaye-Stansel procedure had to be aborted [26].
There are several other potential disadvantages to a strategy of pulmonary artery division and shunt placement. Some of these relate to the shunt, and include shunt thrombosis, pulmonary overcirculation with systemic hypoperfusion, low diastolic blood pressure with resulting coronary ischemia, and pulmonary artery distortion. The patients in the current report had no episodes of shunt thrombosis, which may have been due to the postoperative use of intravenous heparin, followed by conversion to aspirin. Pulmonary overcirculation and low diastolic pressure are unusual with proper shunt sizing, and generally respond to appropriate ventilator management and inotropic support. Pulmonary artery distortion in this experience was minor, and well dealt with at subsequent BSCC. Finally, most of the modified Norwood operations in this report were performed using circulatory arrest, which may carry a risk of neurologic damage. Late in this experience, we adopted regional low flow perfusion via the innominate artery to avoid circulatory arrest. Whether this will result in improved neurologic outcomes is currently unknown.
To summarize, in infants with a functional single ventricle and excessive pulmonary flow, we have utilized a strategy of pulmonary artery division, shunt placement, and prophylactic bypass of systemic outflow tract obstruction. This strategy was carried out with no operative deaths, and 90% survival beyond 30 months. Follow-up cardiac catheterization has demonstrated consistent protection of the pulmonary vascular bed. No patient has developed subaortic stenosis or aortic arch obstruction. Function of the pulmonary valve in the neoaortic position has been good, with moderate neoaortic insufficiency in 1 patient. Twenty-one patients have progressed to BSCC, and 6 to a completion Fontan procedure, without operative mortality or failure. The results of this strategy compare favorably with those of other approaches, such as pulmonary banding.
| Discussion |
|---|
|
|
|---|
Thank you again for a nice series and this information.
DR BRADLEY: Thank you very much for your comments. We basically start by looking at the aortic arch. Any patient who has coarctation of the aorta, with or without hypoplasia of the aortic arch, undergoes a modified Norwood procedure. In the remaining patients, we look at the internal anatomy of the heart. Any patient who has origination of the aorta from an outlet chamber, in other words, a chamber served not by an atrioventricular valve, but instead by a bulboventricular foramen or a ventricular septal defect, undergoes a Damus-Kaye-Stansel procedure. One other patient in our series, with double-outlet right-ventricle and mitral atresia, had subaortic conal tissue that was judged to be potentially restrictive, and also underwent a Damus-Kaye-Stansel procedure. The remaining patients have been shown in a number of series to be at low risk for the development of subaortic stenosis, and they undergo simple division of the main pulmonary artery and shunt placement.
DR ROSS M. UNGERLEIDER (Portland, OR): Scott, thanks for sharing that. I think you have presented three distinctly different groups of patients, and I would have no argument personally with the strategy you have employed for the groups that need either aortic arch reconstruction with a modified Norwood or a Damus-Kaye-Stansel for the prevention of subsequent subaortic obstruction.
I am most interested in your first set of patients, the 6 patients who received oversewing of the pulmonary artery and a shunt for what was simply excessive pulmonary blood flow without the risk of aortic obstruction. Were you doing these procedures on bypass or off bypass? If you were doing them on bypass, have you compared your bypass operation of oversewing of the pulmonary artery and a shunt to patients who simply receive a pulmonary artery band with respect to outcome and other perioperative morbidity factors? My point is, I think, that that subset of patients might do pretty well with simply a band done off bypass, unless you are doing your initial operation off bypass. Thank you.
DR BRADLEY: Thank you for your comments; you have hit on the potentially controversial issue in this talk. Cardiopulmonary bypass was utilized in 5 out of the 6 patients who were treated by pulmonary artery division and shunt placement. Bypass was used in those 5 cases because of associated procedures: atrial septectomy in 3, and repair of total anomalous pulmonary venous return in 2. In the absence of associated procedures, pulmonary artery division and shunt placement can be carried out without cardiopulmonary bypass. Whether there is an important difference between a pulmonary band and the approach that I have described in that particular set of 6 patients is a good question. For a variety of reasons, which are detailed in the paper, it can be difficult to accurately predict the eventual effectiveness of a band placed in a neonate. In contrast, a shunt, once properly placed, provides a uniform and controlled source of pulmonary flow. Our bias has been that the most consistent way to attain protection of the pulmonary vascular bed is to come out of the operating room with a shunt providing the only source of pulmonary blood flow. However, the point of this is not to say that pulmonary bands are never useful. It is always important to maintain flexibility in approach, and we present these results for potential comparisons.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. Masuda, Y. Tanoue, T. Ohno, and R. Tominaga Modified Damus-Kaye-Stansel procedure using aortic flap technique for systemic ventricular outflow tract obstruction in functionally univentricular heart. Eur. J. Cardiothorac. Surg., June 1, 2006; 29(6): 1056 - 1058. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Mavroudis and R. M. Sade The Southern Thoracic Surgical Association 50th anniversary celebration: the impact of STSA pediatric cardiothoracic surgery manuscripts on surgical practice Ann. Thorac. Surg., November 1, 2003; 76(90050): S47 - 67. [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 |