* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Pulmonary artery banding and ventricular septal defect enlargement
Survey
Document related concepts
History of invasive and interventional cardiology wikipedia , lookup
Cardiac contractility modulation wikipedia , lookup
Coronary artery disease wikipedia , lookup
Management of acute coronary syndrome wikipedia , lookup
Mitral insufficiency wikipedia , lookup
Lutembacher's syndrome wikipedia , lookup
Hypertrophic cardiomyopathy wikipedia , lookup
Arrhythmogenic right ventricular dysplasia wikipedia , lookup
Cardiac surgery wikipedia , lookup
Aortic stenosis wikipedia , lookup
Atrial septal defect wikipedia , lookup
Quantium Medical Cardiac Output wikipedia , lookup
Dextro-Transposition of the great arteries wikipedia , lookup
Transcript
European Journal of Cardio-thoracic Surgery 22 (2002) 192–199 www.elsevier.com/locate/ejcts Pulmonary artery banding and ventricular septal defect enlargement in patients with univentricular atrioventricular connection and the aorta originating from an incomplete ventricle Alfredo Giuseppe Cerillo a,*, Bruno Murzi a, Sandra Giusti b, Adrian Crucean a, Sofia Redaelli b, Vittorio Vanini a a Department of Paediatric Cardiac Surgery, Ospedale Pediatrico Apuano ‘G. Pasquinucci’, Via Aurelia Sud, 54100 Massa, Italy b Department of Paediatric Cardiology, Ospedale Pediatrico Apuano ‘G. Pasquinucci’, Via Aurelia Sud, 54100 Massa, Italy Received 20 August 2001; received in revised form 9 April 2002; accepted 22 April 2002 Abstract Background: In patients with univentricular atrioventricular connection and the aorta originating from an incomplete ventricle, subaortic stenosis is generally due to a restrictive ventricular septal defect (RVSD), that may be present at birth or develop after palliative procedures. In particular, a primary role in the genesis of the RVSD has been ascribed to pulmonary artery banding (PAB). The aim of this paper is to analyse the possible risk factors for the development of an RVSD, including PAB, and the results of one of the proposed procedures for treatment of this condition (RVSD enlargement). Methods: We retrospectively reviewed clinical records and outpatient records of 24 consecutive patients with univentricular atrioventricular connection and the aorta originating from the incomplete ventricle that received their first treatment at our institution from January 1991 to April 2000. The variables age, sex, weight, diagnosis, surgical procedures, associated anomalies, associated surgical procedures, were considered. Results: Four patients (16.7%) had absent left atrioventricular connection, seven (29.7%) had absent right atrioventricular connection and discordant ventriculo-arterial connection, and 13 (54.7%) had double inlet left ventricle and discordant ventriculo-arterial connection. Five patients (20.8%) had associated coarctation or hypoplasia of the aorta, and eight (33.3%) had pulmonary stenosis or atresia. Median age at the first operation was 7.5 days (range: 1–376). Median weight was 3.5 kg (range: 1.9–6.3). Seventeen patients underwent pulmonary artery banding, one underwent a Damus–Kaye–Stansel connection, one received a Glenn shunt and five a modified Blalock–Taussig shunt. Early mortality was 12.5%. The only variable associated with operative mortality was the presence of coarctation or hypoplasia of the aorta (P ¼ 0:004). Ten patients (41.6%) developed subaortic stenosis. None of the tested variables, including pulmonary artery banding, was associated with the development of subaortic stenosis. Subaortic stenosis was due to a restrictive VSD in eight patients, six of whom underwent direct VSD enlargement by muscular resection and are well at last followup (four complete repairs). None of the procedures was complicated by complete heart block. In two cases subaortic stenosis was treated by a Damus–Kaye–Stansel connection. A single patient died during follow-up, and 11 patients have achieved a complete one-ventricle repair. Conclusion: In our experience, pulmonary artery banding was not associated with an increased risk of developing an RVSD. VSD enlargement proved to be safe and effective for treatment of subaortic stenosis due to an RVSD. q 2002 Elsevier Science B.V. All rights reserved. Keywords: Single ventricle; Pulmonary artery banding; VSD enlargement 1. Introduction In hearts whit univentricular atrioventricular connection both atria connect to a dominant ventricle that may be morphological right, left or indeterminate. A second ventricle, of the opposite morphological type, is usually present. This second ventricle is, according to a widely accepted nomenclature [1], incomplete, since it lacks an inlet portion. * Corresponding author. Tel.: 139-585-493522; fax: 139-585-493616. E-mail address: [email protected] (A.G. Cerillo). In this setting, when the aorta originates from the incomplete ventricle, subaortic outflow tract obstruction is generally due to a restriction to flow at the level of the ventricular septal defect (restrictive VSD) [1–3]. Patients with a morphological left dominant ventricle and transposed great arteries, as in double inlet left ventricle and discordant ventriculo-arterial connection or tricuspid atresia and discordant ventriculo-arterial connection, have been previously reported to be at greater risk of developing this complication [1,3,4]. Moreover, subaortic stenosis is parti- 1010-7940/02/$ - see front matter q 2002 Elsevier Science B.V. All rights reserved. PII: S10 10- 7940(02)0026 1-0 A.G. Cerillo et al. / European Journal of Cardio-thoracic Surgery 22 (2002) 192–199 cularly frequent in the presence of associated coarctation or hypoplasia of the aorta [3,5]. Subaortic stenosis may be present at birth or develop after palliative procedures [4,6]. In particular, pulmonary artery banding has been indicated as to predispose to the development of a restrictive VSD. By increasing afterload, pulmonary artery banding causes hypertrophy of the dominant ventricle [4]. It also acutely reduces the QP/QS ratio, thus altering ventricular geometry [6], reducing ventricular volume, and increasing ventricular wall thickness. Once established, subaortic stenosis brings about myocardial hypertrophy, increases myocardial wall thickness and reduces ventricular compliance, and these alterations constitute known risk factors for the outcome of a one-ventricle repair [7]. Since the first reports of the potentially dangerous effect of pulmonary artery banding, many alternative surgical procedures have been proposed [7–13]. Palliative arterial switch operation, Damus–Kaye–Stansel connection (DKS), and Norwood-like procedures, all enable to protect pulmonary vascular bed and to bypass the possible obstruction site represented by the ventricular septal defect. The purpose of this study was to analyse the impact of preoperative and procedural characteristics, including pulmonary artery banding, that could increase the risk of developing a restrictive ventricular septal defect. We also report our operative and long-term results with a staged approach based on initial pulmonary artery banding and, eventually, treatment of the restrictive VSD by direct muscular resection. 193 posed great arteries (DILV-TGA). Seven patients (29.7%) had absent right atrioventricular connection and transposed great arteries (tricuspid atresia and transposed great arteries, TA-TGA), and four (16.7%) had absent left atrioventricular connection and concordant ventriculo-arterial connection (mitral atresia without aortic atresia, MA). Associated anomalies were present in 15 patients (Table 1). Five patients (20.83%) had coarctation of the aorta (DILVTGA: 1) or aortic hypoplasia (MA: 2; DILV-TGA: 1; TATGA: 1). Three patients (DILV-TGA: 2; TA-TGA: 1) had mild pulmonary stenosis, and five (TA-TGA: 3; DILVTGA: 2) had severe pulmonary stenosis or pulmonary atresia. 2.2. Subaortic stenosis The VSD was considered restrictive when it gave rise to a measurable Doppler gradient from the dominant ventricle to the aorta [11], the presence of which was always confirmed at cardiac catheterization. A VSD that appeared at echo less than half in size of the aortic annulus was also considered indication to cardiac catheterization. When indicated, pharmacological challenge with isoproterenol was performed during the procedure. Subaortic stenosis was present at diagnosis in three patients (12.5%): one with TA-TGA and aortic hypoplasia, in which the VSD was restrictive at birth, one with DILVTGA, infracardiac total anomalous pulmonary venous connection (TAPVC) and a restrictive VSD, and a third with MA and a subaortic fibromuscular ring (Table 1). 2.3. Operative management 2. Materials and methods 2.1. Patients Patients with univentricular atrioventricular connection and the aorta originating from the incomplete ventricle, that received their first treatment at our institution between January 1991 and March 2000 were considered eligible. To avoid possible bias deriving from a different initial therapeutic approach, patients who had previous interventions in other institutions were not included. Neither patients with hypoplastic left heart syndrome nor patients with biventricular atrioventricular connection initially treated by pulmonary artery banding (PAB) and addressed to a one-ventricle repair were considered, since our analysis was focused on hearts in which the systemic output passes by necessity through a VSD. Two expert paediatric cardiologists (S.G. and S.R.) independently reviewed the admission echocardiogram tape of each patient, in order to confirm the diagnosis. Twenty-four patients were finally included in the analysis (Table 1). There were 12 females (50%) and 12 males. Median age at operation was 7.5 days (range: 1–376). Median weight was 3.5 kg (range: 1.9–6.3). Thirteen (54.7%) patients had double inlet left ventricle and trans- Sixteen patients (66.66%; DILV-TGA: 10; MA: 4; TATGA: 2) with pulmonary overflow were initially treated by pulmonary artery banding (Fig. 1). PAB was performed through median sternotomy in all but one case (see below). An umbilical tape was passed around the pulmonary trunk and tightened to achieve a distal pulmonary artery pressure 30–50% of the systemic blood pressure in the presence of an arterial oxygen saturation of 80–90%. Six patients undergoing PAB required additional surgical procedures for the presence of associated anomalies (Table 1). One with MA and a subaortic fibromuscular ring also underwent subaortic fibromuscular ring resection through the aorta, during cardiopulmonary bypass and cardioplegic arrest. One with DILV-TGA and coarctation of the aorta underwent coarctation repair by end to end anastomosis. This is the only patient in our series in which PAB was performed through a left thoracotomy. Three patients (DILV-TGA: 1; MA: 2) with associated aortic hypoplasia also underwent aortic enlargement with autologous pericardium during total circulatory arrest and deep hypothermia. A patient with DILV-TGA, restrictive VSD and TAPVC underwent correction of TAPVC and VSD enlargement through the atriotomy, during cardiopulmonary bypass and cardioplegic arrest. On rewarming, the pulmonary artery 194 Table 1 Patients’ characteristics a Associated anomalies PBF First procedure Associated procedure SAS SAS treatment Last procedure Current status 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 M F M F M F F F M M M F M M M F 3 6 21 7 15 119 31 90 43 1 376 28 1 32 8 5 3.7 3.5 3.4 2.6 3.7 4.2 3.8 3.3 3.8 2.0 6.3 3.1 3.5 3.4 3.6 2.7 DILV-TGA MA DILV-TGA MA DILV-TGA DILV-TGA DILV-TGA MA TA-TGA DILV-TGA DILV-TGA TA-TGA MA DILV-TGA DILV-TGA DILV-TGA Overflow Overflow Overflow Overflow Overflow Overflow Overflow Overflow Overflow Overflow Restricted Overflow Overflow Overflow Overflow Overflow PAB PAB PAB PAB PAB PAB PAB PAB PAB PAB BDGS PAB PAB PAB PAB PAB – Transaortic resection – – – VSDE – – VSDE – VSDE VSDE Transaortic resection – – VSDE Complete repair Fontan take-down Complete repair Operative death Complete repair Complete repair Complete repair Complete repair Complete repair Operative death Complete repair BDGS BDGS BDGS BDGS BDGS Asymptomatic Mild cyanosis Asymptomatic Dead Asymptomatic Asymptomatic Asymptomatic Asymptomatic Asymptomatic Dead Asymptomatic Asymptomatic Mild CHF Asymptomatic Asymptomatic Asymptomatic F M 2 2 2.5 3.7 DILV-TGA TA-TGA Overflow Overflow PAB DKS – SAS resection – Aortic enlargement – – – – – Aortic enlargement – – Aortic enlargement – Coartectomy RVSDE, TAPVC correction – Aortic enlargement NO Fibrous ring NO NO NO RVSD NO NO RVSD NO RVSD RVSD Fibromuscular NO NO RVSD 17 18 NO RVSD – DKS PAB Operative death Asymptomatic Dead 19 20 21 22 23 24 F F M M F F 73 1 7 2 3 309 4.8 3.5 1.9 3.1 3.6 4.8 TA-TGA DILV-TGA DILV-TGA TA-TGA TA-TGA TA-TGA NO SAS (fibrous) NO Hypoplastic aorta NO Mild PS NO NO NO Hypoplastic aorta PS NO Hypoplastic aorta NO CoA SAS (RVSD), TAPVC NO SAS (RVSD), hypoplastic aorta PS PS PS PS PS PS Reduced Reduced Reduced Reduced Reduced Restricted MBT MBT MBT MBT MBT BDGS – – – – – – NO NO NO NO RVSD RVSD – – – – VSDE VSDE Complete repair BDGS Complete repair BDGS BDGS Complete repair Asymptomatic Late death Asymptomatic Asymptomatic Asymptomatic Asymptomatic a See text for explanation. PBF, pulmonary blood flow; SAS, subaortic stenosis; DILV-TGA, double inlet left ventricle and transposed great arteries; MA, mitral atresia without aortic atresia; TA-TGA, tricuspid atresia and transposed great arteries; PS, pulmonary stenosis; RVSD, restrictive ventricular septal defect; CoA, coartation of the aorta; TAPVC, total anomalous pulmonary venous connection; PAB, pulmonary artery banding; DKS, Damus–Kaye–Stansel connection; MBT, Modified Blalock–Taussig shunt; BDGS, bi-directional Glenn shunt; RVSDE, restrictive VSD enlargement; CHF, congestive heart failure. A.G. Cerillo et al. / European Journal of Cardio-thoracic Surgery 22 (2002) 192–199 No. Sex Age Weight Diagnosis (days) (kg) A.G. Cerillo et al. / European Journal of Cardio-thoracic Surgery 22 (2002) 192–199 195 mosed to the superior aspect of the right pulmonary artery with a continuous 6/0 polypropylene suture in the posterior and lateral aspects and with simple interrupted 6/0 polypropylene stitches in the anterior aspect. 2.4. Definitive repair Fig. 1. Management of patients with pulmonary overflow, See text for explanation. PAB, pulmonary artery banding; DKS, Damus–Kaye–Stansel connection; BDGS, bi-directional Glenn shunt; TCPC, total cavopulmonary connection. was banded as described above. This is the only patient in our series in which direct VSD enlargement was performed during neonatal life. A second patient with TA-TGA, restrictive VSD and aortic hypoplasia (Fig. 1 and Table 1) underwent Damus–Kaye–Stansel connection and aortic enlargement with autologous pericardium (see below). Five patients (20.83%; TA-TGA: 3; DILV-TGA: 2) initially required a modified Blalock–Taussig shunt for the presence of pulmonary stenosis or atresia (Table 1 and Fig. 2). This was performed through a median sternotomy, with a 4- or 5-mm polytetrafluoroethylene graft anastomosed to the right subclavian artery proximally and to the superior aspect of the right pulmonary artery distally with two continuous 6/0 polypropylene sutures. Two patients (DILV-TGA: 1; TA-TGA: 1) with moderate pulmonary stenosis were allowed to grow until the age of 1 year, when they underwent bi-directional Glenn shunt (Table 1 and Fig. 2). Bi-directional Glenn shunt was performed during cardiopulmonary bypass with a beating heart. Selective venous cannulation of the inferior vena cava and of the innominate vein achieved venous drainage. The superior vena cava was transected at his junction with the right atrium and anasto- The presence of an incomplete ventricle is considered indication to a one-ventricle repair. This was obtained by a staged approach involving a bi-directional Glenn shunt, performed at around 6 months of age and subsequent completion of the operation, at 2–4 years of age. Total cavopulmonary connection was performed during cardiopulmonary bypass with a beating heart. The inferior vena cava was transected and anastomosed to a 16–22-mm GoreTex conduit with a continuous 5/0 Gore-Tex suture. This was then anastomosed to the inferior aspect of the right pulmonary artery with a continuous 5/0 Gore-Tex suture. 2.5. Management of subaortic stenosis (Table 1) In two patients with MA (Table 1), subaortic stenosis was due to the presence of fibromuscular tissue along the systemic outflow tract. They were treated by fibromuscular tissue resection trough the aorta, the first in the neonatal period and the second at Glenn. Subaortic stenosis due to a restrictive VSD was preferably managed by direct VSD enlargement, performed during cardiopulmonary bypass with cardioplegic arrest, either through the atrium (five cases) or through the aorta (two cases). In order to avoid the conduction tissue, the VSD was enlarged in a superior and anterior direction, as described in the literature [3,7]. One patient with DILVTGA, restrictive VSD and TAPVC (see above) underwent VSD enlargement through the atrium during the neonatal era. Despite the small size of the heart, the VSD was easily visualized through the right-sided AV valve when the right atrium was opened for TAPVC correction. A Damus–Kaye–Stansel connection was employed when VSD enlargement was considered not feasible through the atrium or the aorta. In our series, a DKS was performed in two cases. The first one was a neonate with TA-TGA and aortic hypoplasia that also underwent aortic enlargement. The second was an infant with TA-TGA and banded pulmonary artery, that underwent bi-directional Glenn plus DKS. In both cases the VSD was not easily accessible from the aorta or the atrium. DKS was performed during cardiopulmonary bypass with cardioplegic arrest. After excision of a portion of the proximal aortic wall, the pulmonary artery was transected and end-to-side anastomosed to the aorta with a continuous 5/0 polypropylene suture. 2.6. Data Fig. 2. Management of patients with pulmonary stenosis. See text for explanation. PS, pulmonary stenosis; PA, pulmonary atresia; MBT, modified Blalock–Taussig shunt; BDGS, bi-directional Glenn shunt; TCPC, total cavopulmonary connection. Considered variables were obtained by a retrospective review of clinical records, outpatient records, and operative registry. Continuous variables were analysed by the Mann– 196 A.G. Cerillo et al. / European Journal of Cardio-thoracic Surgery 22 (2002) 192–199 3.2. Subsequent operations (Figs. 1, 2 and Table 1) Table 2 Operative mortality (univariate analysis) a Variable P Morphological right vs. left dominant ventricle First operation (PAB vs. others) Native subaortic stenosis Coarctation or hypoplasia of the aorta Pulmonary stenosis or atresia Weight Age 0.43 (FET) . 0.99 (FET) 0.34 (FET) 0.004 (FET) 0.52 (FET) 0.20 (M-W) 0.09 (M-W) a See text for explanation. PAB, pulmonary artery banding; FET, Fisher’s exact test; M-W, Mann–Whitney U-test. Whitney U-test. Dichotomous variables were analysed by Fisher’s exact test. Overall survival was estimated by the Kaplan–Meier method and comparison between unadjusted overall group survival relative to baseline characteristics was assessed by a log-rank test. Long-term freedom from subaortic obstruction was evaluated by the Kaplan–Meier method. Significance of considered variables was assessed by a log-rank test. P # 0:05 was considered significant. 3. Results 3.1. Initial palliation (Figs. 1, 2 and Tables 1, 2) Three patients died after the first procedure (overall hospital mortality 12.5%). The first one was a neonate with MA and hypoplastic aorta that underwent aortic enlargement with autologous pericardium and PAB. The second was a patient with DILV-TGA and hypoplastic aorta that underwent aortic enlargement with autologous pericardium and PAB. A third patient with TA-TGA, hypoplastic aorta and restrictive VSD underwent DKS and aortic enlargement with autologous pericardium and died. The only factor significantly associated with a poor outcome was aortic hypoplasia (P ¼ 0:004 by Fisher’s exact test, Table 2). Age, weight, morphology of the dominant ventricle, associated anomalies other than aortic hypoplasia, including subaortic stenosis, were not associated with early mortality. Fig. 3. Cumulative survival. See text for explanation. Of the 21 survived patients, 20 have received a bi-directional Glenn anastomosis, either as a first procedure (2) or as the second one (18). One of them, with DILV-TGA and severe pulmonary stenosis requiring a shunt in the neonatal age, died early postoperatively (cumulative mortality: 16.67%). Nineteen patients survived after bi-directional Glenn shunt. Twelve of these subsequently underwent total cavopulmonary connection, with no operative deaths. One patient with DILV-TGA and no associated anomalies already treated by pulmonary artery banding and bi-directional Glenn, required Fontan take-down for the appearance, early postoperatively, of a severe low output syndrome. 3.3. Follow-up Overall, four patients (16.67%) died during the study period. All 20 surviving patients are periodically seen at our outpatients’ service, and actually are asymptomatic or paucisymptomatic (Table 1). Eleven patients (45.83%) have achieved a complete repair. One additional patient required Fontan take-down, and up to now has been doing well. Eight patients are awaiting a definitive repair. Five-year cumulative survival (Kaplan–Meier) was 85.7% (95% confidence limit (CL): 70.7–100%) (Fig. 3). The only variable associated with long-term outcome was the presence of coarctation or hypoplasia of the aorta (Table 3). 3.4. Subaortic stenosis (Fig. 4 and Tables 1, 4) Overall, ten patients (41.66%; TA-TGA: 5; DILV-TGA: 3; MA: 2) presented subaortic stenosis during the study period. Subaortic stenosis was present at birth in three patients (12.5%, two with restrictive VSD). In three it appeared after a first palliative operation (two restrictive VSD), and in the remaining four it was first observed at total cavopulmonary connection (restrictive VSD in four). Of the seven patients that developed subaortic stenosis after one or more palliative procedures, six had a restrictive VSD. Four of them had previously received pulmonary artery banding. The mean time between the diagnosis and the development of subaortic stenosis was 1.8 ^ 1.9 years. Among the eight patients with a restrictive VSD, two received a Damus–Kaye–Stansel connection. The first one was a neonate with TA-TGA and aortic hypoplasia that also underwent aortic arch enlargement. The second was an infant with TA-TGA and banded pulmonary artery, that underwent bi-directional Glenn plus DKS. As already stated, in this patient the VSD was partially covered by accessory tissue from the left-sided AV valve, and VSD enlargement was judged not feasible. Six patients underwent VSD enlargement, in the neonatal period (1), at Glenn (1), or in concomitance with total cavopulmonary connection (4). One of the patients treated at total cavopulmonary connection presented at postoperative A.G. Cerillo et al. / European Journal of Cardio-thoracic Surgery 22 (2002) 192–199 197 Table 3 Factors affecting long term survival (univariate analysis) a Variable Diagnosis Morphology of the dominant ventricle First operation Native subaortic stenosis Coarctation or hypoplasia of the aorta Subaortic stenosis Pulmonary stenosis or atresia DILV-TGA TA-TGA MA Right Left PAB Other Yes No Yes No Yes No Yes No 2-Year survival (%) P (Mantel–Cox) 84.6 85.7 75.0 75.0 85.0 87.5 75.0 87.7 66.6 40.0 94.7 90.0 78.5 87.5 81.2 0.87 0.59 0.44 0.31 0.0009 0.51 0.51 a See text for explanation. DILV-TGA, double inlet left ventricle and transposed great arteries; TA-TGA, tricuspid atresia and transposed great arteries; MA, mitral atresia; PAB, pulmonary artery banding. 4. Discussion artery banding provokes concentric myocardial hypertrophy, increased myocardial mass, reduced ventricular compliance and finally ventricular diastolic dysfunction [4,11]. Moreover, it acutely alters QP/QS ratio, thus reducing volume overload of the dominant ventricle, altering ventricular geometry and increasing myocardial wall thickness [6,11]. These two factors have been indicated as to predispose to the development of a restrictive VSD. In our series, neonatal pulmonary artery banding was not associated with an increased risk of developing a restrictive VSD, nor with an increased risk of long term failure of the dominant ventricle. In our opinion, there are three possible explanations for this. The first one resides in the nature of patients that actually did not receive pulmonary artery banding as a first operation. Apart from the case of a patient with TA-TGA, hypoplastic aorta and restrictive VSD that underwent neonatal Damus–Kaye–Stansel, all the others had pulmonary stenosis severe enough to protect the pulmonary vascular bed, an associated lesion that we encountered in a surprising number of cases [1]. As pulmonary stenosis could itself have acted as a pulmonary band, this may have biased Subaortic stenosis frequently complicates the clinical history of patients with univentricular atrioventricular connection and aorta originating from the incomplete ventricle, and it has been long recognized that in this setting it is generally due to the presence of a restrictive VSD. By abruptly increasing systemic resistance, subaortic stenosis is a frequent cause of acute cardiac failure. Moreover, it represents a stimulus for myocardial hypertrophy, an adaptive mechanism that in the long term can seriously jeopardize candidacy to a one-ventricle repair. The relationships between pulmonary artery banding, myocardial hypertrophy and the appearance of a restrictive VSD were clearly explained by the group of Toronto [2,4]. By increasing the dominant ventricle afterload, pulmonary Fig. 4. Freedom from subaortic stenosis. See text for explanation. SAS, subaortic stenosis. cardiac catheterization with an isoproterenol provoked gradient of about 10 mmHg, in the absence of a basal gradient, and received no further treatment. A second patient had a basal gradient of about 20 mmHg and underwent VSD reenlargement through an atriotomy, with no complications. None of the seven procedures performed was complicated by the occurrence of complete heart block. Cumulative freedom from subaortic stenosis (Kaplan– Meier) at 6 months, 2 and 5 years was 83.3% (95% CL: 68.4–98.2%), 74.8% (95% CL: 57.3–92.3%), and 54.0% (95% CL: 32.5–75.5%), respectively (Fig. 4). None of the tested variables, including previous pulmonary artery banding, was significantly associated with the development of subaortic stenosis (Table 4). Five patients with a previous history of subaortic stenosis have achieved a definitive repair, including one that subsequently underwent Fontan takedown. Four are awaiting the completion of total cavopulmonary connection. 198 A.G. Cerillo et al. / European Journal of Cardio-thoracic Surgery 22 (2002) 192–199 Table 4 Factors influencing the development of subaortic stenosis (univariate analysis) a Variable Diagnosis Morphology of the dominant ventricle First operation Coarctation or hypoplasia of the aorta Pulmonary stenosis or atresia DILV-TGA TA-TGA MA Right Left PAB Other Yes No Yes No 2 Years freedom from SAS (%) P (Mantel–Cox) 92.3 50.0 57.1 50.0 79.7 74.5 75.0 60.0 78.6 87.5 68.2 0.11 0.69 0.71 0.83 0.87 a See text for explanation. SAS, subaortic stenosis; DILV-TGA, double inlet left ventricle and transposed great arteries; TA-TGA, tricuspid atresia and transposed great arteries; MA, mitral atresia; PAB, pulmonary artery banding. our statistics. The second reason probably relates to the time interval that occurred between pulmonary artery banding and bi-directional Glenn shunt. We use to perform this operation around the age of six months, leaving a time interval possibly too short to allow the appearance of myocardial hypertrophy severe enough to restrict the VSD. Careful monitoring of patient conditions and repeated echocardiograms should suffice to prevent the appearance of this complication. Finally, it is reasonable to suppose that factors other than myocardial hypertrophy also play a role in the genesis of the restrictive VSD. It is noteworthy that in 50% of our patients that developed a restrictive VSD, it appeared after a bi-directional Glenn shunt. Rychik and co-workers [6] clearly demonstrated the role of ventricular volume and geometry alterations carried out by volume reducing surgery in determining the appearance of subaortic obstruction. Our data seem to support the idea that volume reduction may actually play a major role in the genesis of the restrictive VSD. At present, the best therapeutic solution for patient with univentricular atrioventricular connection is represented by the total cavopulmonary connection. Diastolic dysfunction secondary to myocardial hypertrophy, and pulmonary vascular disease caused by long-standing pulmonary overflow are two among the most powerful risk factors for late failure of a one-ventricle repair. For this reason, preparative palliative procedures should be able to protect the pulmonary vascular bed without damaging the dominant ventricle. The rationale for palliative arterial switch operation [7,9] is to protect the pulmonary circulation by translocating the pulmonary artery over the obstructed incomplete ventricle outflow tract, creating in the mean time an unobstructed path from the dominant ventricle to aorta. This approach has been largely abandoned, mostly because of the scarce reliability of the VSD in regulating pulmonary blood flow, resulting either in pulmonary overflow or in cyanosis. Damus–Kaye–Stansel connection [10,12] or, in the presence of an hypoplastic aorta, Norwood-like procedures [13,14], allow to bypass the obstruction site represented by the VSD. In recent years tremendous progresses have been reported in the management of hypoplastic left heart syndrome, and in selected institutions the Norwood operation is now performed with an acceptable mortality, making this procedure, at last in theory, the one of choice for patients with an hypoplastic aorta originating from a rudimentary ventricle. Unfortunately, in our hands the Norwood stage I operation still carries a high mortality (to give a figure, operative mortality after stage I Norwood procedure has been 50% in 1997 and 40% in 1999). For this reason at our institution the recourse to this procedure has been limited to classic hypoplastic left heart syndrome. The improved perioperative management of Norwood patients will probably alter this figure in the next years [14]. Recently, good results have been reported with a pulmonary artery banding protocol similar to ours. Jensen and coworkers [15] reported 19% mortality in a series of 26 patients with DILV-TGA or TA-TGA initially treated by pulmonary artery banding, 16 of who subsequently developed subaortic obstruction. Webber and co-workers [16] reported similar results. This approach allows treating neonate patients with a simple, fast, and rather sure procedure, without avoiding the possibility to treat subaortic stenosis should it appear. Amin and co-workers [17] and Deanen et al. [18] have in fact recently demonstrated that pulmonary artery banding does not compromise the pulmonary valve, thus allowing surgeons to perform a Damus–Kaye–Stansel procedure should it became necessary. Treatment of restrictive VSD by VSD enlargement has been reported to expose patients to the occurrence of various complications, some of which are related to the ventriculotomic access (poor ventricular function, aneurysm formation), while others are explained by surgical damage of the conduction tissue (complete heart block [19]). For this reason subaortic obstruction is now preferably managed by a Damus–Kaye–Stansel connection at many institutions [11]. In our series, VSD enlargement was always accomplished either through the aorta or through the atrium, and was never complicated by the occurrence of complete heart block. This A.G. Cerillo et al. / European Journal of Cardio-thoracic Surgery 22 (2002) 192–199 is at least in part due to the fact that the only neonate in our series that required enlargement of the VSD was a girl with DILV-TGA and associated TAPVC, in which the VSD was easily visualized from the atrium at operation. As the transaortic and the transatrial approach are not always feasible in the neonatal era, it is probable that in the neonate with a restrictive VSD other strategies (Damus–Kaye–Stansel or Norwood procedure) are preferable. Patients with associated coarctation or hypoplasia of the aorta constitute a particularly high-risk subgroup. In our series, three of five such patients died after the first palliation. The coexistence of an obstruction to flow at the aortic and subaortic level should probably be considered as an indication to the Norwood procedure. Our current strategy for patients with univentricular atrioventricular connection and aorta originating from the incomplete ventricle is based on the following considerations. (1) Protect pulmonary circulation by pulmonary artery banding (in the absence of pulmonary stenosis or atresia). (2) Create a Glenn anastomosis around the sixth month of age, thus avoiding excessive myocardial hypertrophy. During this period, careful echocardiographic monitoring of the VSD is mandatory. (3) Complete the one-ventricle repair by an extracardiac conduit, between the age of 2–4 years. Subaortic stenosis is preferably managed by direct VSD enlargement, associating, when possible, this procedure to one of the surgical steps mentioned above. It is not possible, from our data, to draw conclusions about the best strategy for treatment of the neonate with a restrictive VSD. 5. Conclusion Subaortic stenosis may be expected to occur in about 25– 75% of patients with univentricular atrioventricular connection and aorta originating from the incomplete ventricle by the age of 5 years, and tends to appear after procedures that reduce the volume of the dominant ventricle. Pulmonary artery banding constitutes a valid alternative for the initial palliative management of patients with pulmonary overflow. VSD enlargement is effective in resolving subaortic gradients, and, in our series, did not lead to relevant complications. Neonates with subaortic stenosis constitute a particularly high-risk subgroup, and are probably best managed by different approaches (Damus–Kaye–Stansel or Norwood operation). References [1] Becker AE, Anderson RH. Pathology of congenital heart disease, London: Butterworths, 1981. pp. 241–278. [2] Penkoske PA, Freedom RM, Williams WG, Trusler GA, Rowe RD. Surgical palliation of subaortic stenosis in the univentricular heart. J Thorac Cardiovasc Surg 1984;87(5):767–781. 199 [3] Kirklin JW, Barratt-Boyes BG. Double inlet left ventricle and atretic atrioventricular valve. In: Kirklin JW, Barratt-Boyes BG, editors. Cardiac surgery, London: Churchill Livingstone, 1993. pp. 1549– 1580. [4] Freedom RM, Benson LN, Smallhorn JF, Williams WG, Trusler GA, Rowe RD. Subaortic stenosis, the univentricular heart, and banding of the pulmonary artery: an analysis of the courses of 43 patients with univentricular heart palliated by pulmonary artery banding. Circulation 1986;73(4):758–764. [5] Freedom RM, Dische MR, Rowe RD. Pathologic anatomy of subaortic stenosis and atresia in the first year of life. Am J Cardiol 1977;39:1035–1044. [6] Rychik J, Jacobs ML, Norwood Jr WI. Acute changes in left ventricular geometry after volume reduction operation. Ann Thorac Surg 1995;60(5):1267–1274. [7] Karl TR, Watterson KG, Sano S, Mee RB. Operations for subaortic stenosis in univentricular hearts. Ann Thorac Surg 1991;52(3):420– 428. [8] Serraf A, Conte S, Lacour-Gayet F, Bruniaux J, Sousa-Uva M, Roussin R, Panche C. Systemic obstruction in univentricular hearts: surgical options for neonates. Ann Thorac Surg 1995;60(4):970–977. [9] Lacour-Gayet F, Serraf A, Fermont L, Bruniaux J, Rey C, Touchot A, Petit J, Planche C. Early palliation of univentricular hearts with subaortic stenosis and ventriculoarterial discordance. The arterial switch option. J Thorac Cardiovasc Surg 1992;104(5):1238–1245. [10] Brawn WJ, Sethia B, Jagtap R, Stumper OF, Wright JG, De Giovanni JV, Silove ED, Jackson M, Sreeram N. Univentricular heart with systemic outflow obstruction: palliation by primary Damus procedure. Ann Thorac Surg 1995;59(6):1441–1447. [11] Odim JN, Laks H, Drinkwater Jr DC, George BL, Yun J, Salem M, Allada V. Staged surgical approach to neonates with aortic obstruction and single-ventricle physiology. Ann Thorac Surg 1999;68(3):962– 968. [12] Gates RN, Laks H, Elami A, Drinkwater Jr DC, Pearl JM, George BL, Jarmakani JM, Williams RG. Damus–Kaye–Stansel procedure: current indications and results. Ann Thorac Surg 1993;56(1):111– 119. [13] Rychik J, Murdison KA, Chin AJ, Norwood WI. Surgical management of severe aortic outflow obstruction in lesions other than hypoplastic left heart syndrome: use of a pulmonary artery to aorta anastomosis. J Am Coll Cardiol 1991;18(3):809–816. [14] Tchervenkov CI, Shum-tim D, Beland MJ, Jutras L, Platt R. Single ventricle with systemic obstruction in early life: comparison of initial pulmonary artery banding versus the Norwood operation. Eur J Cardiothorac Surg 2001;19(5):671–677. [15] Jensen RA, Williams RG, Laks H, Drinkwater D, Kaplan S. Usefulness of banding of the pulmonary trunk with single ventricle physiology at risk for subaortic obstruction. Am J Cardiol 1996;77:1089– 1093. [16] Webber SA, LeBlanc JG, Keeton BR, Salmon AP, Sandor GG, Lamb RK, Monro JL. Pulmonary artery banding is not contraindicated in double inlet left ventricle with transposition and aortic arch reconstruction. Eur J Cardiothorac Surg 1995;9:515–520. [17] Amin Z, Backer CL, Duffy CE, Mavroudis C. Does banding of the pulmonary artery affect pulmonary valve function after the Damus– Kaye–Stansel operation? Ann Thorac Surg 1998;66:836–841. [18] Daenen W, Eyskens B, Meyns B, Gewillig M. Neonatal pulmonary artery banding does not compromise the short term function of a Damus–Kaye–Stansel connection. Eur J Cardiothorac Surg 2000;17: 655–657. [19] O’Leary PW, Driscoll DJ, Conner AR, Puga FJ, Danielson GK. Subaortic obstruction in hearts with a univentricular connection to a dominant left ventricle or an anterior subaortic outlet chamber. J Thorac Cardiovasc Surg 1992;104:1231–1238.