Download Pulmonary artery banding and ventricular septal defect enlargement

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.