Congenital aortopulumonary window: diagnosis, surgical technique and

long-term results
T Tkebuchava, LK von Segesser, PR Vogt, U Bauersfeld, R Jenni, A Kunzli, M
Lachat and M Turina
Eur J Cardiothorac Surg 1997;11:293-297

This information is current as of May 31, 2010

The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://ejcts.ctsnetjournals.org

The European Journal of Cardio-thoracic Surgery is the official Journal of the European
Association for Cardio-thoracic Surgery and the European Society of Thoracic Surgeons.
Copyright © 1997 by European Association for Cardio-Thoracic Surgery. Published by
Elsevier. All rights reserved. Print ISSN: 1010-7940.

Downloaded from ejcts.ctsnetjournals.org by on May 31, 2010

 European Journal of Cardio-thoracic Surgery 11 (1997) 293 – 297

Congenital aortopulmonary window: diagnosis, surgical technique and

long-term results

T. Tkebuchavaa,*, L.K. von Segessera, P.R. Vogta, U. Bauersfeldb, R. Jennic, A. Künzlia,

M. Lachata, M. Turinaa
a
Clinic for Cardio6ascular Surgery, Uni6ersity Hospital, Rämistrasse 100, CH-8091 Zurich, Switzerland
b
Department of Pediatrics, Uni6ersity Hospital, Rämistrasse 100, CH-8091 Zurich, Switzerland
c
Department of Cardiology, Uni6ersity Hospital, Rämistrasse 100, CH-8091 Zurich, Switzerland

Received 30 November 1995; revised 18 April 1996; accepted 25 April 1996

Abstract

Objecti6e: Congenital aortopulmonary window is rare, often associated with other cardiac anomalies. Surgical repair as the only
treatment should be performed before pulmonary vascular changes have developed. This study presents the long-term outcome
after surgical correction for this condition. Materials and methods: Between 1971 and 1993, 13 patients with congenital
aortopulmonary window were found. 10 had type I, 2 type II and 1 had type III. Concomitant cardiac anomalies were present
in 10/13. Eleven patients were operated on at a mean age of 31.29 48.3 months (range 6 days – 10 years).
Thoracotomy was used in 3 and sternotomy in 9 patients. In 4/11, the aorto-pulmonary window was simply ligated, 4 had a
transpulmonary approach and 1 combined with a transaortic approach. The aortopulmonary window was closed directly in 2 and
with a Dacron patch in 1. Cardiopulmonary bypass was used in 6 patients. Associated anomalies were corrected in 10/11 patients.
Results: There was one operative death (9%): a six-day old boy with interrupted aortic arch died 6 h postoperatively due to
low-cardiac output. The mean follow-up period is 8.197.3 years (range 2 – 24 years). Clinical examination, transaortic
echocardiography and/or cardiac catheterization were obtained in the follow-up. There was no late death. All are in New York
Heart Association (NYHA) class I. One had to be reoperated on for a recurrent shunt 29 months after ligation and one had
angioplasty after 23 months for residual stenosis of the reimplanted right pulmonary artery. The actuarial survival rate is 90%
after 1, 5 and 10 years. Conclusions: The surgical treatment of aortopulmonary window has a low risk, even if associated with
major cardiac anomalies. Prompt operative treatment achieves excellent long-term results. © 1997 Elsevier Science B.V.

Keywords: Aortopulmonary septal defect; Aortopulmonary window

1. Introduction genital cardiac diseases [19]. Most commonly, the aor-

Aortopulmonary window is a rare congenital mal-
formation resulting from abnormal septation of the
truncus arteriosus into the aorta and the pulmonary
artery. It appears in approximately 0.15% of all con-
* Corresponding author. Tel.: +41 1 2551111; fax: + 41 1
2554446.
Downloaded from ejcts.ctsnetjournals.org by on May 31, 2010
1010-7940/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved
PII S 1 0 1 0 - 7 9 4 0 ( 9 6 ) 0 1 0 4 8 - 2
topulmonary window is a single defect of proximal
type; its size varies from a few mm to several cm.
Distal defects are less common [1,14]. The prognosis
of uncorrected aortopulmonary communication is
poor, with 40% of the infants dying during the first
year of life. A substantial proportion of survivors will
die from congestive heart failure in childhood.
This analysis was performed to assess the long-term
outcome of patients operated for congenital aortopul-
monary window.
294 T. Tkebucha6a et al. / European Journal of Cardio-thoracic Surgery 11 (1997) 293–297
2. Materials and methods
2.1. Patients
Thirteen consecutive patients, 4 females and 9 males,
were investigated at our hospital between 1971 and
1993 because of congenital aortopulmonary septal de-
fect. There were two nonoperated children, the first, in
1976, was a two-day old boy with various congenital
cardiac and noncardiac malformations. This child was
considered as nonoperable and died a few days later
from cardiac failure. The second, in 1978, was a girl,
with complicated mixed congenital anomalies. She died
approximately 15 h after birth.
Eleven patients were operated on. Age at operation
ranged from 6 days to 10 years (mean age 31.29 48.3
months). Among the operated patients, at clinical ex-
amination 6 children presented signs of congestive heart
failure, i.e. exertional dyspnea in 5/11 and recurrent
respiratory infections in 3/11 cases. Three patients had
been referred after a patent ductus arteriosus had been
suspected. Electrocardiography revealed left ventricular
hypertrophy in 2 patients and biventricular hypertro-
phy in another 2. Two-dimensional echocardiography
was performed since the late seventies in 5 patients and
allowed correct diagnosis in all of them. Cardiac
catheterization was performed in order to define the
exact anatomy of the aortopulmonary communication,
exclude associated cardiac malformations, as well as for
the registration of the pulmonary artery pressure, calcu-
lation of pulmonary vascular resistance and for the
determination of shunt volume. From this examination,
mean left to right shunt was calculated to be 2.49 0.6
(pulmonary to systemic ratio), mean pulmonary artery
pressure was 699 7 mmHg and pulmonary vascular
resistance was 4.95 91.2 U · m2. Type I defect was
observed in 10/13 patients, type II in 2/13 and type III
in 1/13 patient (Fig. 1). Only 3 patients of the whole
group showed isolated aortopulmonary window, but
had already developed complications because of hemo-
dynamic changes due to the main disease, such as
pulmonary complications and biventricular hypertro-
phy.
The following concomitant congenital lesions in op-
erated patients were diagnosed: ventricular septal de-
Fig. 1. Diagrammatic representation of aortopulmonary window
localisation according to the modified Richardson’s classification.
Downloaded from ejcts.ctsnetjournals.org by on May 31, 2010
fect, 4/8; atrial septal defect, 3/8; patent ductus
arteriosus, 3/8; persistent left superior vena cava, 2/8;
patent foramen ovale, 2/8; dextrocardia, 2/8; inter-
rupted aortic arch, 1/8; coronary anomaly, 1/8; trans-
position of great arteries, 1/8; and pulmonary valve
atresia, 1/8 (Table 1).
2.2. Surgical management
Eleven of the thirteen patients underwent surgical
intervention for total correction of aortopulmonary
defect. In 3 events the surgical approach to the heart
was performed through a left-sided antero-lateral tho-
racotomy in the fourth intercostal space; in the other 9,
median sternotomy was used. Closure of aortopul-
monary window was performed with extracorporeal
circulation using hypothermia and blood cardioplegia
in 7 children, whereas in 5 correction could be achieved
without cardiopulmonary bypass.
The following types of surgical intervention were
used (Table 1): simple ligation of the windowlike com-
munication (n= 3/11), ligation with additional suturing
of the aorta (n=1/11); transpulmonary approach (n=
4/11) for direct closure (n=1/4) using autologous peri-
cardial patch (n=2/4) or combined with transaortic
approach for direct closure (n= 1/4); division (n= 3/11)
either with oversewing of the aorta and the pulmonary
artery (n=2/11) or use of Dacron patch (n=1/11), and
in one case (aortopulmonary window type III), the
right pulmonary artery was reimplanted to the pul-
monary trunk. The following associated procedures
were performed simultaneously (Table 1): ligation of a
patent ductus arteriosus (n= 3/10), banding of pul-
monary artery (n= 2/10), closure of ventricular (n=2/
10) or atrial septal defect (n=1/10), correction of
interrupted aortic arch (n=1/10), and interposition of
the Hancock-Graft between right ventricle and pul-
monary artery (n= 1/10).
3. Results
There was only one perioperative death after correc-
tion of interrupted aortic arch and aortopulmonary
window. This child was operated on urgently because
of persistent metabolic acidosis and developed refrac-
tory biventricular failure soon after operation. No
other significant perioperative complications were ob-
served. However, one patient had to be reoperated 29
months after the first procedure. He was found to have
a recanalisation of the aortopulmonary communication.
At reoperation, a ruptured ligature was detected. Clo-
sure of the window was performed through a combined
transpulmonary/transaortic approach and the postoper-
ative course was uneventful. In another boy a residual
stenosis was found at the side of the anastomosis
 Table 1
Surgical intervention and results with congenital aortopulmonary windows

Age (months) Sex Associated anomaly L/R Rp Rp/Rs ECC Operations and types of APW Outcome Follow-up
U · m2 (years)

3 M ASD − 1971 (Type I) midline sternotomy Alive and well 24

APW-ligation, additional suture
of Ao.
8 M VSD, PDA, PFO large 1.1:1 5 0.35 − 1972 (Type I) lateral thoraco- Recanalisation of APW 20
tomy. APW-ligation, PDA-liga-
tion, PA-banding.
+ 1975 reoperation: midline ster- Alive and well
notomy. APW-transaortic and
transpulmonary direct closure,
PA-debanding, VSD-Dacron
patch closure.
119 M 70% 5.6 1/3 + 1975 (Type II) midline ster- Alive and well 2
notomy. Division, Dacron patch
closure.
0.8 M Dextrocardia, VSD, PLSVC, − 1975 (Type I) lateral thoraco- Alive and well 2
PDA, CA tomy. APW-ligation, PA-bending.
0.2 F IAA severe high − 1978 (Type I) lateral thoraco- Death after 6 h
tomy. Division, oversuturing of
Ao, PA. Anastomosis of aortic
arch with Ao descendens, PDA-
ligation.
129 M Dextrocardia, TGA, Pulm. atre- + 1979 (Type I) midline sternotomy Alive and well 5
sia, ASD, VSD APW-ligation, Hancock-Graft
RV-PA. Dacronvelour-Patch of
VSD, ASD direct closure.
3 M ASD 83% 6:1 0.8 + 1985 (Type I) midline sternotomy Alive and well 10
Division, oversuturing of Ao, PA.
48 M 65% 2.85:1 4.5 0.27 + 1990 (Type I) midline sternotomy Alive and well 6

Downloaded from ejcts.ctsnetjournals.org by on May 31, 2010

Transpulmonary direct closure.
0.2 M PDA, PFO − 1990 (Type III) midline ster- Stenosis of the right PA anasto- 5
notomy. Reimplantation of the mosis. Alive and well
right PA to the pulmonary trunk,
PDA-ligation.1992 Angioplasty of
the anastomosis
T. Tkebucha6a et al. / European Journal of Cardio-thoracic Surgery 11 (1997) 293–297

30 M 95% + 1991 (Type I) midline ster- Alive and well 4

notomy. Transpulmonary pericar-
dial patch.
2 F VSD, RAA, PLSVC 77% 5 0.19 + 1992 (Type I) midline sternotomy Alive and well 3
Transpulmonary pericardial
patch.

L/R, left to right shunt; Rp, pulmonary vascular resistance; Rp/Rs, the pulmonary to systemic resistance ratio; ECC, extracorporeal circulation; APW, aortopulmonary window; ASD, atrial septal
defect; PDA, patent ductus arteriosus; PA, pulmonary artery; VSD, ventricular septal defect; PFO, patent foramen ovale; PLSVC, persistent left superior vena cava; CA, coronary anomaly; IAA,
interrupted aortic arch; Ao, aorta; RV, right ventricle; TGA, transposition of the great arteries; RAA, right aortic arch.
295
296 T. Tkebucha6a et al. / European Journal of Cardio-thoracic Surgery 11 (1997) 293–297
between the reimplanted right pulmonary artery and
the pulmonary trunk. Twenty-three months later an
angioplasty of the stenosis was performed successfully.
Since then no other residual defects were documented
at the follow-up.
Actuarial survival analysis provides a 90% survival
rate at 1, 5 and 10 years of follow-up. Postoperative
regular examinations consisted in treadmill-exercise,
echocardiography and/or cardiac catheterization. The
long-term outcome up to 24 years is satisfactory in all
patients. All patients are active and without any car-
diac-related symptoms. After one year, 7/8 patients
were in New York Heart Association (NYHA) func-
tional class I, 6/6 after 3, 3/3 after 10 and 2/2 after 22
years.
4. Discussion
Surgical closure of a congenital aortopulmonary win-
dow should be performed immediately after the diagno-
sis has been established and regardless of the patient’s
age.
In some patients the operation can be done through
the anterolateral thoracotomy without the use of car-
diopulmonary bypass even if the repair has to be
combined with ligation of a persistent ductus arteriosus,
pulmonary artery banding or correction of an inter-
rupted aortic arch.
Currently, early closure of congenital aortopul-
monary window has a low operative mortality and the
prognosis mainly depends on the presence of associated
cardiac malformations. In older patients, the outcome
is determined by the pulmonary vascular resistance at
the time of the repair [23].
An aortopulmonary window is the result of malfor-
mation in the division of the aortopulmonary trunk
during embryogenesis. This defect can show various
sizes and localisations. Among different classifications
of this malformation we used Richardson’s anatomic
classification [18] which presents as follows (Fig. 1):
type I, the defect is situated between the ascending
aorta and the main pulmonary artery just above the
sinuses of Valsalva; genetically it is due to incomplete
septation of the aortopulmonary trunk. Type II, the
defect is more distal, between the ascending aorta and
the origin of the right pulmonary artery from the main
pulmonary artery (due to abnormal migration of the
sixth aortic arch), whereas type III consists of anoma-
lous origin of the right pulmonary artery from the
aorta. This latter type is the result of unequal septation
of the aortopulmonary trunc or truncus arteriosus. The
most frequent form, as in our study, is type I (n= 9/13).
Elliotson [6] was the first to describe congenital aor-
topulmonary window in 1830. Since then approxi-
mately 250 cases were presented, most as case reports.
Downloaded from ejcts.ctsnetjournals.org by on May 31, 2010
Table 2
Historical review of surgical management in congenital aortopul-
monary window
1952 Simple ligation (Gross) [8];
1953 Division with following suturing (Scott) [21];
1957 Repair of the defect under the visual control using bypass
(Cooley) [3];
1961 Total correction in type III (Armer) [1];
1966 Patching through pulmonary artery (Putnam) [15];
1968 Transaortic direct closure (Wright) [24];
1969 Transaortic closure with Dacron patch (Deverall) [4];
1978 ‘Sandwich’ patch closure (Johansson) [9];
1989 U-shaped felt strip technique (Schmid) [20];
1991 Rerouting of the right pulmonary artery (Kitagawa) [10];
1994 Closure with pulmonary artery flap (Messmer) [13];
Aortopulmonary window is frequently found as an
isolated lesion, but can be associated with various other
congenital cardiac anomalies [7,16,22], as our experi-
ence showed. Commonly, the defect is large and an
important left-to-right shunt is present, resulting in
congestive heart failure, pulmonary hypertension and
early development of pulmonary vascular obstructive
disease.
Recently, two-dimensional echocardiography has
considerably facilitated the diagnosis of this defect. In
transthoracic studies, the communication between the
aorta and pulmonary artery can be best visualised in
the high parasternal short-axis view. This non-invasive
investigation allows the exact localisation of the defect
and demonstrates its size. Further, it is useful for
evaluation of associated anomalies, but sometimes it
can be difficult to differentiate it from truncus arterio-
sus. Cardiac catheterization is very important in these
patients, because it provides information on pulmonary
artery pressure and pulmonary vascular resistance.
Surgical closure is indicated in all patients with aor-
topulmonary window; it should be undertaken as early
as possible after diagnosis because of the risk of pul-
monary vascular disease. Since the first report about
successful repair by Gross in 1952 [8], several authors
have reported their results based on a variety of tech-
niques, from simple ligation to closure of the defect
with a pulmonary artery flap (Table 2). Some authors
have presented their experience even without the use of
cardiopulmonary bypass [20]. The defect may be best
visualised through a longitudinal aortotomy which al-
lows exact localisation of the coronary ostia. Most
defects require closure with a patch in order to avoid
narrowing of the pulmonary artery or even of the aorta.
The best method probably includes separate closure of
the aorta and the pulmonary artery with two different
patches through a transpulmonary and transaortic ap-
proach. Therefore, currently we prefer closure under
direct vision using cardiopulmonary bypass. Simple
ligation and simultaneous pulmonary artery banding
 T. Tkebucha6a et al. / European Journal of Cardio-thoracic Surgery 11 (1997) 293–297
should not be done as this may result in distraction of
the anatomic structures which may impair secondary
corrections.
Unlike other reports [12,17,11], 77% of our patients
had associated cardiac anomalies. Nevertheless, com-
plete surgical correction did not increase perioperative
mortality nor its complication rate, except in one pa-
tient with associated interrupted aortic arch, who died
from refractory metabolic acidosis. Autoptic examina-
tion revealed myocardial infarction and subarachnoidal
bleeding.
Long-term outcome after operation is predominantly
dependent on the level of pulmonary resistance at the
time of closure of the defect. Concerning the periopera-
tive mortality, our results compare favourably with
those reported in the literature [2,5].
All patients were long-term survivors except one and
follow-up examinations demonstrated excellent late
functional results, regardless of the operative technique.
Regular postoperative control is recommended
throughout the growth period of these patients in order
to detect recurrent stenosis which can be corrected
readily, resulting in a good long-term outcome as
demonstrated in this study.
References
[1] Armer RM, Shumacker HB, Klatte EC. Origin of the right
pulmonary artery from the ascending aorta. Circulation
1961;24:662 – 668.
[2] Bertolini A, Dalmonte P, Bava GL, Moretti R, Cervo G,
Marasini M. Aortopulmonary septal defects. A review of the
literature and report of ten cases. J Cardiovasc Surg
1994;35:207 – 213.
[3] Cooley DA, McNamara DG, Latson JR. Aortopulmonary sep-
tal defect: diagnosis and surgical treatment. Surgery
1957;42:101 – 120.
[4] Deverall PB, Lincoln JCR, Aberdeen E, Bonham-Carter RE,
Waterston DJ. Aortopulmonary window. J Thorac Cardiovasc
Surg 1969;57:479–486.
[5] Doty DB, Richardson JV, Falkovsky GE, Gordonova MI,
Burakovsky VI. Aortopulmonary septal defect: hemodynamics,
angiography and operation. Ann Thorac Surg 1981;32:244 – 250.
[6] Elliotson S. Case of malformation of the pulmonary artery and
aorta. Lancet 1930;1:247–251.
Downloaded from ejcts.ctsnetjournals.org by on May 31, 2010
297
[7] Gargiulo G, Zannini L, Albanese S et al. Interrupted aortic arch
and aorto-pulmonary window: one-stage repair in the first week
of life. Ann Thorac Surg 1993;56:554 – 556.
[8] Gross RE. Surgical closure of an aortic septal defect. Circulation
1952;5:858 – 863.
[9] Johansson L, Michaelsson M, Westernholm C-J, Åberg T. Aor-
topulmonary window: a new operative approach. Ann Thorac
Surg 1978;25:564 – 567.
[10] Kitagawa T, Katoh I, Taki H et al. New operative method for
distal aortopulmonary septal defect. Ann Thorac Surg
1991;51:680 – 682.
[11] Kutsche LM, Van Mierop LH. Anatomy and pathogenesis of
aorticopulmonary septal defect. Am J Cardiol 1987;59:443–447.
[12] Luisi SV, Ashraf MH, Gula G, Radley-Smith, Yacoub M.
Anomalous origin of the right coronary artery with aortopul-
monary window: functional and surgical consideration. Thorax
1980;35:446 – 448.
[13] Messmer BJ. Pulmonary artery flap for closure of aortopul-
monary window. Ann Thorac Surg 1994;57:498 – 501.
[14] Mori K, Ando M, Takao A, Ishikawa S, Imai Y. Distal type of
aortopulmonary window: report of 4 cases. Br Heart J
1978;40:681 – 689.
[15] Putnam TC, Gross RE. Surgical management of aortopul-
monary fenestration. Surgery 1966;59:727 – 735.
[16] Rein AJ, Gotsman MS, Simcha A. Echocardiographic diagnosis
of interrupted aortic arch with an aortopulmonary communica-
tion. Int J Cardiol 1989;24:238 – 241.
[17] Redington AN, Rigby ML, Ho SY, Gunthard J, Anderson RH.
Aortic atresia with aortopulmonary window and interruption of
the aortic arch. Pediatr Cardiol 1991;12:49 – 51.
[18] Richardson JV, Doty DB, Rossi NP, Ehrenhaft JL. The spec-
trum of anomalies of aortopulmonary septation. J Thorac Car-
diovasc Surg 1979;78:21 – 27.
[19] Rowe RD. Aortopulmonary septal defect. In: Keith JD, Rowe
RD, Vlad P, editors. Heart Disease in infancy and childhood,
third ed. New York: Mcmillan, 1978.
[20] Schmid FX, Hake U, Iversen S, Schranz D, Oelert H. Surgical
closure of aorto-pulmonary window without cardiopulmonary
bypass. Pediatr cardiol 1989;10:166 – 169.
[21] Scott HW Jr, Sabiston DC Jr. Surgical treatment for congenital
aorticopulmonary fistula: experimental and clinical aspects. J
Thorac Surg 1953;25:26.
[22] Tabak C, Moskowitz W, Wagner H, Weinberg P, Edmunds HL.
Aorto-pulmonary window and aortic isthmic hypoplasia: opera-
tive management in newborn infants. J Thorac Cardiovasc Surg
1983;86:273 – 279.
[23] Van Son JAM, Puga FJ, Danielson GK et al. Aortopulmonary
window: factors associated with early and late success after
surgical treatment. Mayo Clin Proc 1993;68:128 –133.
[24] Wright JS, Freeman R, Johnston JB. Aortopulmonary fenestra-
tion: a technique of surgical management. J Thorac Cardiovasc
Surg 1968;55:280 – 283.
 Congenital aortopulumonary window: diagnosis, surgical technique and
long-term results
T Tkebuchava, LK von Segesser, PR Vogt, U Bauersfeld, R Jenni, A Kunzli, M
Lachat and M Turina
Eur J Cardiothorac Surg 1997;11:293-297
This information is current as of May 31, 2010

Updated Information including high-resolution figures, can be found at:

& Services http://ejcts.ctsnetjournals.org
Citations This article has been cited by 1 HighWire-hosted articles:
http://ejcts.ctsnetjournals.org#otherarticles
Permissions & Licensing Information about reproducing this article in parts (figures,
tables) or in its entirety can be found online at:
http://ejcts.ctsnetjournals.org/misc/Permissions.shtml
Reprints Information about ordering reprints can be found online:
http://ejcts.ctsnetjournals.org/misc/reprints.shtml

Downloaded from ejcts.ctsnetjournals.org by on May 31, 2010