CARDIOVASCULAR
Is Complete Heart Block After Surgical Closure of
Ventricular Septum Defects Still an Issue?
Henrik Ø. Andersen, MD, PhD, Marc R. de Leval, MD, Victor T. Tsang, MD, MS,
Martin J. Elliott, MD, Robert H. Anderson, MD, and Andrew C. Cook, PhD
Department of Cardiothoracic Surgery, Rigshospitalet, Denmark; Cardiothoracic Unit, Great Ormond Street Hospital for Children,
National Health Service Trust, London, and Cardiac Unit, Institute of Child Health, London, United Kingdom
Background. A serious complication after surgical closure of ventricular septal defect (VSD) is complete heart
block. In this retrospective study, we reviewed the incidence of complete heart block after surgical closure of a
VSD at Great Ormond Street Hospital from 1976 to 2001
to identify any particular anatomic features that still
predisposed patients to surgically-induced complete
heart block and to provide anatomic guidelines to avoid
this in future.
Methods. Data were extracted from our local database
for patients having (1) isolated VSD or VSD in the
setting of (2) tetralogy of Fallot with pulmonary stenosis
or (3) tetralogy of Fallot with pulmonary atresia; (4)
absent pulmonary valve syndrome; (5 and 6) coarctation
or interruption of the aortic arch; and (7) subaortic fibrous
shelf. We carefully reviewed the operative notes from all
patients with postoperative complete heart block to dis-
cover any predisposing anatomical reasons to explain the
complication.
Results. Two thousand seventy-nine patients had a
VSD closure. Permanent complete heart block developed
in 7 of 996 patients (0.7%) with an isolated defect and in
1 of 847 patients (0.1%) with tetralogy of Fallot. Four
more patients had postoperative complete heart block.
Conclusions. Instances of iatrogenic complete heart
block continue to occur after surgical VSD closure, either
because of unexpected biological variations or because of
unawareness of the disposition of the atrioventricular
conduction axis in particular circumstances. This report
emphasizes the latter aspect in details and suggests a risk
of iatrogenic complete heart block of less than 1%.
V
remains the production of complete heart block, and this
is unequivocally linked to the conduct of the surgical
procedure itself. Closure is usually achieved by insertion
of a patch, anchoring the patch using either a continuous
suture, or interrupted sutures. During this maneuver, be
it performed through atrial, ventricular, or arterial access,
traction and tension are needed to obtain good surgical
exposure. The conduction system, specifically the bundle
of His and its branches, is almost always closely related to
some part of the border of the defect, and is, therefore, at
risk during the insertion of the individual stitches.
Should heart block occur during or after the procedure, it
is most often an indication for insertion of a permanent
pacemaker, which may demand either a limited, partial
sternotomy, performed by opening the lower part of
epigastric portion of the wound, or in some occasions a
full resternotomy, with insertion of epicardial pacemaker
electrodes. There is a recognized increased risk of late
death in such patients with postsurgical complete heart
block [7].
In the account of the follow-up of Lillehei’s pioneering
open heart repairs [2], it is reported that complete heart
block occurred in 4 of the 27 patients in whom closure
was attempted. In reports collected from the last 30 years,
describing series of from 23 to 265 patients, although
some had no incidence of complete heart block [5, 8 –14],
this complication occurred in as many as 4% of the others
[3, 4, 7, 15–28], with an incidence of as high as 8%
entricular septal defects (VSD) are the most frequent
congenital cardiac malformations [1]. Surgical closure of such defects was first performed in 1954, using
cross-circulation [2]. Since then, closure by the surgeon
has become routine. Over the intervening period, however, there have been significant changes in the surgical
strategy for closure. Thus, priority has shifted from a
two-stage approach, with initial banding of the pulmonary trunk to limit the flow of blood to the lungs, with
subsequent surgical closure of the defect, to a singlestage approach, with radical surgery performed at an
early age [3, 4]. The strategy for perfusion has also
changed, from the use of total circulatory arrest with
cooling to 18°C, to an approach without any, or only mild,
cooling and standard cardiopulmonary bypass. Finally,
the surgical procedure itself has changed, from a transventricular procedure [5] to a transatrial procedure,
thereby limiting the potential damage to the right ventricle [4, 6].
Complications are now rare, but still include serious
problems such as cerebral damage, or even death. In
certain circumstances, such problems may still be unavoidable. Another serious complication, nonetheless,
Accepted for publication April 7, 2006.
Address correspondence to Dr Andersen, Department of Cardiothoracic
Surgery, The Heart Centre, Rigshospitalet, Blegdamsvej 9, Copenhagen
2100, Denmark; e-mail: [email protected].
© 2006 by The Society of Thoracic Surgeons
Published by Elsevier Inc
(Ann Thorac Surg 2006;82:948 –57)
© 2006 by The Society of Thoracic Surgeons
0003-4975/06/$32.00
doi:10.1016/j.athoracsur.2006.04.030
ANDERSEN ET AL
COMPLETE HEART BLOCK AFTER VSD CLOSURE
949
Table 1. Surgical Closure of Simple Ventricular Septal Defects Reported in the Literature Since 1971, With Details of the
Incidence of Complete Heart Block and Mortality
Author
Year
Published
Surgical
Period
Number of
Patients
Complete
Heart
Block
Ibach [15]
Ziady [16]
Blackstone [17]
Sigman [18]
Fisher [8]
McNicholas [4]
Hobbins [19]
Rizzoli [20]
Borsu [21]
Arciniegas [5]
Blake [22]
Henze [23]
Yeager [24]
Doty [25]
Houyel [26]
Moller [7]
McGrath [9]
Serraf [27]
Hardin [3]
Backer [10]
Kuribayashi [11]
Mullen [12]
Leao [13]
Nygren [29]
Gaynor [14]
Bol-Raap [28]
Andersen
1971
1972
1976
1977
1978
1979
1979
1980
1980
1980
1982
1984
1984
1985
1990
1991
1991
1992
1992
1993
1994
1996
1996
2000
2001
2003
2005
1960–70
1965–71
1967–76
1957–75
1971–76
1971–76
1973–76
1967–79
1974–79
1973–79
1958–75
1976–82
1973–81
1972–83
1960–87
1954–60
1985–89
1980–90
1986–91
1980–91
1971–82
1993–95
1974–92
1976–96
1996–99
1992–01
1976–01
102
68
88
106
71
65
58
265
54
52
187
46
128
57
100
258
115
130
48
141
49
23
124
256
172
188
996
2%
4%
1.5%
1%
0%
3%
2%
1%
3.7%
0%
3%
2%
2%
2%
4%
3.5%
0%
3%
(2)a
0%
0%
0%
0%
8%
0%
1%
0.7%
a
Mortality
9%
5%, late death
9.8%
22%
4.2%
6%
15%
9.6%
4%
7.6%, transventricular access
Pts selected by survival
6%
7.8%
7%
Pts from pool of 262 pts, atrial/ventricular access
13%
1%
7.7%, isolated multiple VSDs
2% early, 4% late
0%
8%
0%
1.8% transaortic approach
10%
0.6%, tricuspid valve detachment
2%
1.5%
One patient had intermittent atrioventricular block, but died day 4 postoperatively.
Pts ⫽ patients;
VSD ⫽ ventricular septal defect.
reported in one series spanning a period of 21 years [29]
(Table 1).
The aim of our retrospective study, therefore, was to
review the incidence of complete heart block in patients
who had undergone surgical closure of a VSD at Great
Ormond Street Hospital over a period of 26 years. We
hoped to identify any particular anatomic features that
still predisposed to surgically-induced heart block, if
indeed it still occurred, and to provide anatomic guidelines to avoid this complication in future. We analysed
only patients with concordant atrioventricular and ventriculoarterial connections, and with normally positioned
chambers and arterial trunks.
Material and Methods
The data were collected from our local database, which
covered the period from January 1, 1976, through December 31, 2001, a period of 26 years. The database contains
information concerning date of birth, diagnosis, surgical
procedures, date of operation, and so on. The study was
approved by the local Institutional Review Board. Indi-
vidual patient consent was waived in this quality review
study with no patient identifiers used. We extracted data
relevant to patients with otherwise isolated VSDs, along
with those having interventricular communications in
the setting of tetralogy of Fallot with pulmonary stenosis
or atresia, absent pulmonary valve syndrome, coarctation
or interruption of the aortic arch, and those having
resection of a subaortic fibrous shelf. We included those
with multiple as well as single defects. We excluded
patients with discordant ventriculoarterial connections
(“transposition”), double-outlet right ventricle, all forms
of functionally single ventricle, including hypoplasia of
the left heart, and those with single outlet from the heart
except in the setting of tetralogy. We also excluded all
patients having a common atrioventricular junction, irrespective of the level of shunting through the associated
atrioventricular septal defect.
So as to identify the patients who had suffered complete heart block, we matched the procedure “VSD
closure” with the procedure “insertion of pacemaker,”
defining complete heart block for the purposes of this
review as the indication for insertion of a permanent
CARDIOVASCULAR
Ann Thorac Surg
2006;82:948 –57
CARDIOVASCULAR
950
ANDERSEN ET AL
COMPLETE HEART BLOCK AFTER VSD CLOSURE
Ann Thorac Surg
2006;82:948 –57
Table 2. Numbers in the Various Groups, Along With the Incidence of Complete Heart Block and the Rates of Deaths, for
Patients Undergoing Surgical Closure of a Ventricular Septal Defect at Great Ormond Street, London, During the Period
of 26 Years From 1976 to 2001
Diagnostic Groups
Isolated
Subaortic shelf
Previous repair of coarctation
Interrupted aortic arch
Tetralogy of Fallot
Absent pulmonary valve
syndrome
Tetralogy of Fallot with
pulmonary atresia
Number of
Patients
Complete
Heart
Block
Number of Deaths
1976–2001
Number of Deaths
1997–2001
996
32
58
40
847
14
7 (0.7%)
1 (6.4%)
2 (3,4%)
1 (2.5%)
1 (0.1%)
0
15 (1.5%)
1 (3.1%)
8 (13.8%)
12 (30%)
27 (3.2%)
2 (14.3%)
2/263 (0.7%)
0/10 (0%)
0/19 (0%)
1/10 (10%)
1/42 (2.5%)
92
0
11 (12.0%)
1/13 (8%)
pacemaker. We then reviewed in detail the operative
notes from all patients found to have postoperative
complete heart block to discover any predisposing anatomical reasons to explain the complication.
Results
During the period of 26 years spanning from 1976 to 2001,
closure of a VSD was attempted in 2,079 patients with
concordant atrioventricular and ventriculoarterial connections (Table 2). Of these, the defect in 996 patients was
an isolated perimembranous, muscular, or doubly committed defect, whereas in 847 patients the defect was
closed as part of the repair of tetralogy of Fallot. The
numbers of patients with other complicating lesions,
such as fibrous subaortic shelves, absent pulmonary
valve syndrome, and so on, are shown in Table 2.
In terms of the numbers of patients in the two major
categories undergoing closure each year, the surgical
activity has been relatively stable since 1988, with an
average of 50 and 40 operations yearly for the two
respective groups (Fig 1). The distribution of the ages of
the patients in these groups is shown in Figure 2. Most
patients underwent closure of isolated defects between
the ages of 3 and 12 months, whereas the most frequent
age at closure for those with tetralogy of Fallot was a little
later, from 6 months to 2 years. The overall mortality for
the patients with isolated defects was 1.5%, and that for
the patients with tetralogy of Fallot was 3.2%. The major
cause of postoperative death was low cardiac output. The
rates of mortality for those with other complicating
lesions are shown in Table 2.
In the 996 patients with an isolated defect, permanent
complete heart block developed postoperatively in 7,
whereas only 1 of the 847 patients with tetralogy of Fallot
suffered from this complication (Table 2). Another 4
patients also had heart block. Thus, among the 58 patients with previous repair of coarctation, 2 had heart
block, whereas 1 patient among the 32 patients undergoing resection of a subaortic fibrous shelf had heart block.
Heart block also developed in 1 of 40 patients who
needed additional surgery for repair of an interrupted
aortic arch. It is noteworthy, however, that in closing the
VSD in 92 patients needing procedures on the pulmonary
arteries in the setting of tetralogy with pulmonary atresia,
and 14 with absent pulmonary valve syndrome, there
were no incidences of surgically-induced complete heart
block (Table 2).
The findings from the detailed review of the operative
and case notes are shown in Table 3. In 1 patient (no. 12),
heart block occurred subsequent to resection of the
subaortic fibrous shelf that was producing obstruction
within the left ventricular outflow tract. Such shelves are
known to overlie the left bundle branch, albeit that the
obstructing fibrous tissue can usually be enucleated
without damaging the underlying conduction tissues, as
occurred in the other 31 patients with such fibrous
shelves.
In 2 further patients (nos. 4 and 5), there was associated
overriding and straddling of the tricuspid valve (Fig 3),
including the only patient in our series in whom heart
block occurred in the setting of tetralogy of Fallot. The
Fig 1. The graph shows the numbers of patients undergoing surgery
in each year at Great Ormond Street Hospital in the period from
1976 to 2001 for patients having isolated ventricular septal defect
(open columns) and patients with tetralogy of Fallot (shaded
columns).
ANDERSEN ET AL
COMPLETE HEART BLOCK AFTER VSD CLOSURE
951
Fig 2. The graph shows the distribution of ages at the time of surgical closure of the patients with
isolated ventricular septal defect
(open columns) and patients with
tetralogy of Fallot (shaded columns) operated on in the period
from 1976 to 2001. (mth ⫽
months; yrs ⫽ years.)
operative notes give no indication that the technique
used to close the interventricular communication was
modified once the surgeon had noted the presence of
straddling of the tricuspid valve, yet the conduction axis
is known to be abnormally located when there is straddling and overriding of the tricuspid valve (see discussion). In another 2 cases (nos. 1 and 8), heart block
occurred after closure of multiple muscular defects. This
is another known risk factor for heart block, yet in one
instance (patient no. 8), the muscular bar separating the
defects was purposely removed. This maneuver is almost
certain to induce surgical heart block (see discussion).
Heart block also occurred in 1 patient (no. 3) when a
muscular defect opening to the inlet was closed with
sutures placed close to the muscular margins. This is
another known danger situation (see discussion). When
closing a perimembranous defect in 1 further patient (no.
9), the surgeon commented in his operative note that he
was suspicious about a suture placed in the fibrous tissue
forming the central fibrous body, yet he did not remove
the suture. This patient also had postoperative complete
heart block. In the other 3 patients (nos. 2, 7, and 10) with
perimembranous defects, and 2 (nos. 6 and 11) with
muscular defects, there were no obvious anatomical
reasons why the patient should have had complete heart
block.
The mean time from closure of the defect to insertion
of the pacemaker was 27 days, with a range from 6 to 96
days. In 2 patients, however, the course was unusual. Our
first patient (no. 1) initially had sinus rhythm, which
converted to complete atrioventricular dissociation 9
days after closure of the defect. The other patient (no. 2)
suffered a serious postoperative infection, and the pacemaker was not inserted until 96 days after surgery.
Excluding these 2 patients, the mean period for insertion
of the pacemaker was 17 days after surgery, with a range
from 6 to 42 days.
Comment
It is salutary to note that, when Lillehei first performed
surgical closure of VSDs, conventional wisdom [30, 31]
suggested that there was no single axis responsible for
atrioventricular conduction, despite the earlier exemplary description of the atrioventricular bundle provided
by Tawara [32]. It was Lev [33], in fact, who clarified the
course of the atrioventricular bundle in hearts with the
typical VSD, with Copenhaver and Truex [34] shortly
thereafter showing the distinction between the course of
the conduction axis in the setting of what we now call
perimembranous as opposed to muscular defects opening to the inlet of the right ventricle. Had the rules
established by Copenhaver and Truex [34] been respected, these subsequently being confirmed by Latham
and Anderson [35] in 1972, before the commencement of
our current series, then heart block could at least have
been avoided in our patient with a muscular defect
opening close to the annulus of the tricuspid valve
(patient no. 1). Indeed, on the basis of these early studies,
the disposition of the atrioventricular conduction axis has
been well established for all the various types of VSD
[36], and surgeons are well aware of the significance of
these anatomical findings [37]. Despite this knowledge,
as our analysis shows, occasional instances of iatrogenic
complete heart block continue to occur after surgical
closure of VSD, either because of unexpected biological
variations, or because of unawareness of the known
disposition of the atrioventricular conduction axis in
CARDIOVASCULAR
Ann Thorac Surg
2006;82:948 –57
CARDIOVASCULAR
952
ANDERSEN ET AL
COMPLETE HEART BLOCK AFTER VSD CLOSURE
Ann Thorac Surg
2006;82:948 –57
Table 3. Details of the 12 Patients From 2,079 Who Had Complete Heart Block After Surgical Closure of an Interventricular
Communication
Patient No. and
Age at Surgery
Date of
Surgery
Diagnosis
Description of Defect and
Significant Lesions
Description of
Surgery
One defect just underneath
the aortic valve in close
proximity to the tricuspid
valve, with additional
apical defects. All 3
apical defects were
covered with one patch
placed through left
ventriculotomy.
Small perimembranous
defect opening toward
apex and large ASD.
Access through right
atrium and left
ventriculotomy.
Upper defect
closed with
interrupted
sutures inferiorly
and with running
suture superiorly.
Access through right
atrium. Standard
closure.
Access through right
atrium. Sutures
placed close to the
margins of
muscular defect.
Access through right
atrium. In the
vicinity of the
straddling
papillary muscle
to the septal
leaflet of the
tricuspid valve, a
small cut was
made in the patch.
The stitches were
then brought out
through the
annulus of the
tricuspid valve.
Access through right
atrium. Running
suture, patch split
and draped
around the
papillary muscle.
Right ventricular
muscle bundles
resected.
Access through right
atrium. Standard
closure. CHB
noted at end of
bypass.
(1) 7 years
9/6/77
Multiple muscular
VSDs, earlier
PAB
(2) 6 months
5/13/85
Perimembranous
VSD
(3) 1 1/2 years
3/22/90
Muscular VSD,
earlier PAB
Muscular defect opening to
RV inlet.
(4) 2 1/2 years
4/11/94
Perimembranous
VSD
Overriding of tricuspid
valve and straddling
tricuspid valve papillary
muscle.
(5) 1 year
1/30/95
Tetralogy of Fallot
Huge confluent
perimembranous defect.
Large papillary muscle to
tricuspid valve was
straddling the septum
and arising from the left
ventricular surface.
(6) 6 months
3/16/95
Muscular VSD
Large muscular defect.
Abnormal AV
conduction with left
bundle branch block.
particular circumstances. It is the latter circumstance that
are the focus of our current review, since in more than
half of our own cases, it is possible that damage to the
conduction axis could have been avoided had full advantage been taken of the extant anatomical knowledge.
In this respect, our own experience, coupled with that
culled from review of the literature, shows that the incidence of complete heart block is independent of the route
Time of
Pacemaker
Insertion
PM 42 days after
surgery
CHB
postoperatively;
temporary PM
wires due to
infection in
postoperative
course; final
PM 96 days
after surgery
CHB
postoperatively;
PM 12 days
after surgery
Postoperative
CHB; PM 7
days after
surgery
PM 23 days after
surgery
PM 9 days after
surgery
of surgical access, and whether or not the surgeon chooses
to detach the septal leaflet of the tricuspid valve [14]. Some
of our own patients, nonetheless, would not have had heart
block if knowledge about the course of the atrioventricular
bundle had been respected [36, 37]. Thus, in 1 patient (no. 8)
with both perimembranous and muscular defects, the surgeon chose to divide the muscle bar between the defects. It
is well established that the bundle of His almost certainly
ANDERSEN ET AL
COMPLETE HEART BLOCK AFTER VSD CLOSURE
953
Table 3. (Continued)
Patient No. and
Age at Surgery
Date of
Surgery
Diagnosis
Description of Defect and
Significant Lesions
(7) 6 months
3/27/95
Perimembranous
VSD
Earlier PAB ⫹ CoA
Large perimembranous
inlet defect.
(8) 3 1/2 years
6/6/96
Perimembranous
and
muscular
VSDs
Previous PAB ⫹
CoA
Enormous defect. Septum
almost absent, but 2
defects.
(9) 6 months
10/17/
96
Perimembranous
VSD
Large perimembranous
outlet defect with an
aortic valve prolapsing
through the superior
margin of the defect. The
suture in the fibrous rim
was suspected by
surgeon to be suspicious,
but he did not remove in
spite of 2:1 block.
(10) 2 1/2 months
21.01.99
Perimembranous
VSD
Inlet VSD extending to the
subaortic area, diameter
larger than the aorta.
(11) 1 year
02.03.00
IAA (type A) and
muscular VSD
Large muscular outlet
defect with posterior
deviation of the
infundibular septum,
producing a significant
narrowing of the
subaortic area.
(12) 6 months
08.01.01
Perimembranous
VSD and
subaortic shelf
Large perimembranous
defect and subaortic
shelf.
ASD ⫽ atrial septal defect;
band;
PM ⫽ pacemaker;
Access through right
atrium. Standard
closure. CHB
peroperative.
Access through right
atrium. Defects
joined by cutting
the muscle
between them,
and the holes
were closed as
one.
Access through right
atrium. There was
a zone of
deficiency of
valvar tissue, and
sutures were
placed in the
fibrous rim of this
defect in the belief
that this was far
away from the
conduction tissue.
Access through right
atrium. Defect
larger than aorta.
Standard closure.
Access through right
atrium. Sutures
were placed
deeply into the
left side of the
outlet septum in
an attempt to pull
it to the right side.
Access through right
atrium. Subaortic
shelf resected
from aortotomy.
CHB ⫽ complete heart block;
CoA ⫽ Coarctation;
IAA ⫽ interrupted aortic arch;
SR ⫽ sinus rhythm;
VSD ⫽ ventricular septal defect.
will traverse such a muscle bar (Fig 4) [36]. In another
patient (no. 9), a zone of deficiency was noted between the
septal and antero-superior leaflets of the tricuspid valve. A
suture was placed in the fibrous rim at this point, which
almost certainly was the site at which the atrioventricular
bundle is known to penetrate from the apex of the triangle
of Koch to the crest of the muscular ventricular septum
(Figs 5 and 6) [6, 36].
In 2 further cases, the surgeon recognized the presence
of straddling and overriding of the tricuspid valve (patient nos. 4 and 5). Overriding of the right atrioventricular
junction, with consequent malalignment between the
atrial and ventricular septal components, is known to be
the one situation in which, with a perimembranous
defect, the conduction axis does not arise from the
regular atrioventricular node, but instead arises from an
Time of
Pacemaker
Insertion
Description of
Surgery
PM 12 days after
surgery
PM 6 days after
surgery
PM 36 days after
surgery; Had a
short-termed
occurrence of
SR then CHB
CHB noted.
February 1,
1999; PM 60
days after
surgery
CHB during
operation; PM
12 days after
surgery
CHB
postoperatively;
PM 8 days
after surgery
PAB ⫽ pulmonary artery
anomalous node. The node is formed at the site where
the malaligned ventricular septum meets the right atrioventricular junction (Fig 3) [38]. In this respect, it is surely
significant that 1 of the patients with straddling and
overriding of the tricuspid valve was the only 1 among
847 patients with pulmonary stenosis, and a further 92
with pulmonary atresia, who had complete heart block in
the setting of tetralogy of Fallot. This difference between
the incidence of heart block between otherwise isolated
defects, and that occurring in the setting of tetralogy, was
highly significant (p ⫽ 0.04, Fishers test). It almost certainly reflects the fact that the bundle of His is better
protected in the setting of tetralogy of Fallot, the known
danger situations being combined perimembranous and
muscular defects [36], and straddling and overriding of
the tricuspid valve [38].
CARDIOVASCULAR
Ann Thorac Surg
2006;82:948 –57
CARDIOVASCULAR
954
ANDERSEN ET AL
COMPLETE HEART BLOCK AFTER VSD CLOSURE
Fig 3. The drawing shows the known disposition of the conduction
axis, as seen by the surgeon operating through the right atrium,
when there is straddling and overriding of the tricuspid valve [38].
Note that the bundle does not originate from the regular atrioventricular node located at the apex of the triangle of Koch.
The choice and timing of insertion, of a permanent
pacemaker after closure of a VSD can be difficult.
Insertion of a permanent pacemaker may introduce
increased morbidity, and even mortality [22, 39]. Furthermore, on the one hand, recovery of atrioventricular
conduction after temporary postoperative block has
been seen up to 6 weeks after surgery, and on the other
hand, occurrence of block has been seen as late as 25
years after closure of the septal defect [7]. One study,
investigating patients after surgery for congenital heart
disease, reported that two thirds of patients with
temporary postoperative block regained atrioventricular conduction, and that in 97%, this occurred within 9
days [40]. They also reported that of the patients who
needed a permanent pacemaker, 11 of 31 patients had
recovery of atrioventricular conduction after insertion
of the pacemaker [40]. This sequence occurred in 1 of
our own patients, who had a pacemaker inserted 6 days
Ann Thorac Surg
2006;82:948 –57
Fig 5. The drawing shows the expected site of the conduction axis
when a ventricular septal defect (VSD) is perimembranous [36], as it
would be seen by the surgeon lifting up the septal leaflet of the tricuspid valve having approached the defect through the right atrium.
after surgery, and then regained atrioventricular conduction 2 days later.
Temporary pacemakers, placed epicardially during surgery, normally should function for at least 3 weeks after
surgery. Thus, in this respect, the therapeutic window for
placement of a permanent epicardial system extends between 9 and 21 days after surgery. That is also a period
during which postsurgical adherences are not fully developed. Most would prefer to insert an epicardial system with
steroid-eluting electrodes, which should have a lifespan
similar to permanent endocardial systems [39]. Our mean
time for insertion of the permanent pacemaker was 17 days.
It is also the case, of course, that patients suffering transient
postoperative block should always be followed carefully in
order to detect any possible later development of complete
heart block [7, 22].
Limitations of the Study
In this study, we chose to focus only on the development
of complete heart block after surgical closure of so-called
Fig 4. The drawings show the known disposition of the conduction system when the atrioventricular conduction axis descends through a muscle bar separating perimembranous and muscular inlet defects [36]. In (a), the leaflets of the tricuspid valve are shown in-situ, whereas they
are retracted in (b). (VSD ⫽ ventricular septal defect.)
ANDERSEN ET AL
COMPLETE HEART BLOCK AFTER VSD CLOSURE
955
ate to expect rates of mortality at around 0%, and the risk
of iatrogenic heart block at less than 1%. It is against
these results that cardiologists should now evaluate the
anticipated results from interventional closure.
Research at the Great Ormond Street Hospital for Children,
National Health Service Trust, and the Institute of Child Health
and benefits from research and development funding were
received from the NHS Executive. Andrew C. Cook, PhD, and
Robert H. Anderson, MD, are supported by grants from the
British Heart Foundation, together with the Joseph Levy Foundation in the case of Dr Anderson.
References
Fig 6. The drawing shows how the conduction axis penetrates
through the central fibrous body at the apex of the triangle of Koch
when a defect is perimembranous [36], and how this is more obvious
to the surgeon when there is a deficiency of leaflet tissue at the zone
of apposition between the septal and anterosuperior leaflets of the
tricuspid valve.
“simple” VSDs. Other diagnostic groups with VSDs
might also have been included in the survey. Such
groups would be those with interventricular communications in the setting of regular or congenitally corrected
transposition, double-outlet right ventricle, atrioventricular septal defect with common atrioventricular junction,
and common arterial trunk. In our opinion, however,
inclusion of these groups would have unduly confounded the data. Furthermore, there are differences in
the course of the atrioventricular bundle in congenitally
corrected transposition and atrioventricular septal defects when compared with isolated defects. Patients with
double-outlet right ventricle constitute a very heterogeneous group, with a spectrum from double-outlet right
ventricle of the Fallot type to the Taussig-Bing malformation, and in a retrospective review, it would be difficult
to distinguish between these different anatomic types. To
limit our population, therefore, we included only patients
with concordant atrioventricular and ventriculoarterial
connections.
In the process of selection , we matched the procedures
of closure of VSD with insertion of a pacemaker. In some
patients, of course, complete heart block could have
developed at a time when they were no longer under the
surveillance of our unit, and a pacemaker could have
been inserted elsewhere. Such patients, nonetheless,
most likely would have been readmitted to our unit. We
also chose to omit any considerations of transient heart
block, information that might also have been of interest,
but which was not uniformly available.
In conclusion, we see no reason to suppose that our
experience at Great Ormond Street does not reflect
surgical experience worldwide for closure of VSDs. Thus,
our review of literature suggests that it is now appropri-
1. Hoffman JIE. Incidence, mortality and natural history. In:
Anderson RH, Baker EJ, Macartney FJ, Rigby ML, Shinebourne EA, eds. Paediatric cardiology. 2nd ed. Edinburgh:
Churchill Livingstone, 2005:111– 40.
2. Lillehei CW, Varco RL, Cohen M, Warden HE, Patton C,
Moller JH. The first open-heart repairs of ventricular septal
defect, atrioventricular communis, and tetralogy of Fallot
using extracorporeal circulation by cross-circulation: a 30year follow-up. Ann Thorac Surg 1986;41:4 –21.
3. Hardin.J.T., Muskett AD, Canter CE, Martin TC, Spray TL.
Primary surgical closure of large ventricular septal defects in
small infants. Ann Thorac Surg 1992;53:397– 401.
4. McNicholas K, de Leval M, Stark J, Taylor JFN, Macartney FJ.
Surgical treatment of ventricular septal defect in infancy.
Primary repair versus banding of pulmonary artery and later
repair. Br Heart J 1979;41:133– 8.
5. Arciniegas E, Farooki ZQ, Hakimi M, Perry BL, Green EW.
Surgical closure of ventricular septal defect during the first
twelve months of life. J Thorac Cardiovasc Surg 1980;80:
921– 8.
6. de Leval M. Transatrial closure of perimembranous ventricular septal defect in infancy [Letter]. J Thorac Cardiovasc
Surg 1984;87:640 –2.
7. Moller JH, Patton C, Varco RL, Lillehei CW. Late results (30
to 35 years) after operative closure of isolated ventricular
septal defect from 1954 to 1960. Am J Cardiol 1991;68:1491–7.
8. Fisher RD, Faulkner SL, Sell CG, Graham TP, Bender HW.
Operative closure of isolated defectsof the ventricular septum: planned delay. Ann Thorac Surg 1978;26:351– 6.
9. McGrath LB. Methods for repair of simple isolated ventricular septal defects. J Card Surg 1991;6:13–23.
10. Backer CL, Winters RC, Zales VR, et al. Restrictive ventricular septal defect: how small is too small to close? Ann
Thorac Surg 1993;56:1014 – 8.
11. Kuribayashi R, Sekine S, Aida H, et al. Long-term results of
primary closure for ventricular septal defects in the first year
of life. Surg Today 1994;24:389 –92.
12. Mullen JC, Lemermeyer G, Schipper SA, Bentley MJ. Perimembranous ventricular septal defect repair: keeping it
simple. Can J Cardiol 1996;12:817–21.
13. Leao LE, Buffolo E, Coto AE, Maluf MA, Andrade JC.
Transaortic approach has a role in the surgical treatment of
ventricular septal defects. Cardiovasc Surg 1996;4:250 – 4.
14. Gaynor JW, O’Brien JE Jr, Rychik J, Sanchez GR, DeCampli
WM, Spray TL. Outcome following tricuspid valve detachment for ventricular septal defects closure. Eur J Cardiothorac Surg 2001;19:279 – 82.
15. Ibach JR Jr, Bartley TD, Daicoff GR, et al. Correction of
ventricular septal defect in childhood. Ann Thorac Surg
1971;11:499 –507.
16. Ziady GM, Hallidie-Smith KA, Goodwin JF. Conduction
disturbances after surgical closure of ventricular septal defect. Br Heart J 1972;34:1199 –204.
17. Blackstone EH, Kirklin JW, Bradley EL, DuShane JW, Appelbaum A. Optimal age and results in repair of large
CARDIOVASCULAR
Ann Thorac Surg
2006;82:948 –57
CARDIOVASCULAR
956
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
ANDERSEN ET AL
COMPLETE HEART BLOCK AFTER VSD CLOSURE
ventricular septal defects. J Thorac Cardiovasc Surg
1976;72:661–79.
Sigmann JM, Perry BL, Gehrendt DM, Stern AM, Kirsh MM,
Sloan HE. Ventricular septal defect: results after repair in
infancy. Am J Cardiol 1977;39:66 –71.
Hobbins SM, Izukawa T, Radford DJ, Williams WG, Trusler
GA. Conduction disturbances after surgical correction of
ventricular septal defect by the atrial approach. Br Heart J
1979;41:289 –93.
Rizzoli G, Blackstone EH, Kirklin JW, Pacifico AD, Bargeron
LM Jr. Incremental risk factors in hospital mortality rate
after repair of ventricular septal defect. J Thorac Cardiovasc
Surg 1980;80:494 –505.
Borst HG, Oelert H, Bernsau U, Kallfelz HC. Closure of
ventricular septal defects during infancy. Thorac Cardiovasc
Surg 1980;28:109 –14.
Blake RS, Chung EE, Wesley H, Hallidie-Smith KA. Conduction defects, ventricular arrhythmias, and late death after
surgical closure of ventricular septal defect. Br Heart J
1982;47:305–15.
Henze A, Koul BL, Wallgren G, Settergren G, Bjork VO.
Repair of ventricular septal defect in the first year of life.
Scand J Thorac Cardiovasc Surg 1984;18:151– 4.
Yeager SB, Freed MD, Keane JF, Norwood WI, Castaneda
AR. Primary surgical closure of ventricular septal defect in
the first year of life: results in 128 infants. J Am Coll Cardiol
1984;3:1269 –76.
Doty DB, Lambert WC. Repair of ventricular septal defects.
World J Surg 1985;9:516 –21.
Houyel L, Vaksmann G, Fournier A, Davignon A. Ventricular arrhythmias after correction of ventricular septal defects:
importance of surgical approach. J Am Coll Cardiol 1990;16:
1224 – 8.
Serraf A, Lacour-Gayet F, Bruniaux J, et al. Surgical management of isolated multiple ventricular septal defects. Logical approach in 130 cases. J Thorac Cardiovasc Surg 1992;
103:437– 42.
Bol-Raap G, Weerheim J, Kappetein AP, Witsenburg M,
Bogers AJ. Follow-up after surgical closure of congenital
ventricular septal defect. Eur J Cardiothorac Surg 2003;24:
511–5.
Ann Thorac Surg
2006;82:948 –57
29. Nygren A, Sunnegardh J, Berggren H. Preoperative evaluation and surgery in isolated ventricular septal defects: a 21
year perspective. Heart 2000;83:198 –204.
30. Glomsett DJ, Glomsett ATA. A morphologic study of the
conduction system in ungulates, dog and man: I, the sinoatrial node. Am Heart J 1940;20:389 –98.
31. Glomsett DJ, Glomsett ATA. A morphologic study of the
conduction system in ungulates, dog and man: II, the Purkinje system. Am Heart J 1940;20:677–700.
32. Tawara S. Das Reizleitungssystem des Saugetierherzens:
Eine anatomisch-histologischeStudie über das Atrioventrikularbundel und die Purkinjeschen Faden. Jena, Germany: Gustav Fischer, 1906:9-70.
33. Lev M. The architecture of the conduction system in congenital heart disease: III, ventricular septal defect. Arch
Pathol 1960;70:530 – 49.
34. Copenhaver WM, Truex RC. Histology of the atrial portion
of the cardiac conduction system in man and other mammals. Anat Rec 1952;114:601–25.
35. Latham RA, Anderson RH. Anatomical variations in atrioventricular conduction system with reference to ventricular
septal defects. Br Heart J 1972;34:185–90.
36. Anderson RH, Becker AE. The anatomy of ventricular septal
defects and their conduction tissues. In: Stark J, de Leval M,
eds. Surgery for congenital heart defects. 2nd ed. Philadelphia: WB Saunders, 1983:115–38.
37. de Leval M. Ventricular septal defects. In: Stark J, de Leval
M, eds. Surgery for congenital heart defects. 2nd ed. Philadelphia: WB Saunders, 1983:355–71.
38. Milo S, Ho SY, Macartney FJ, et al. Straddling and overriding
atrioventricular valves: morphology and classification. Am J
Cardiol 1979;44:1122–34.
39. Thomson JD, Blackburn ME, Van Doorn C, Nicholls A,
Watterson KG. Pacing activity, patient and lead survival over
20 years of permanent epicardial pacing in children. Ann
Thorac Surg 2004;77:1366 –70.
40. Weindling SN, Saul JP, Gamble WJ, Mayer JE, Wessel D,
Walsh EP. Duration of complete atrioventricular block after
congenital heart disease surgery. Am J Cardiol 1998;82:525–7.
INVITED COMMENTARY
Andersen and colleagues [1] report the incidence of
complete atrioventricular (AV) conduction block in a
selected group of patients who had surgical closure of
ventricular septal defect (VSD) at Great Ormond Street
during a 26-year period. The results are excellent with an
overall incidence of less than 1% for pacemaker insertions after postsurgical AV conduction block. Based on
important anatomic and morphologic data, the authors
make recommendations to further reduce the risk of
injury to the AV conduction axis, supporting the notion
that the incidence of the postoperative need for a pacemaker may in fact be less than 0.5%.
The anatomy of the conduction tissue in congenital
heart disease is fundamental to surgical practice. Thanks
to the many contributions of Anderson and colleagues
[1], Becker, Kurosawa, de Leval, and others, many of the
morphologic details of the AV conduction axis in congenital heart disease have been defined. Biologic studies of
morphologists are a good example of how basic science
can impact clinical practice and outcomes. The principles
of surgical closure of perimembranous VSD, in general,
should include the use of oversized patches, suturing to
© 2006 by The Society of Thoracic Surgeons
Published by Elsevier Inc
the right septal surface at least 2 mm away from the rim
of the ventricular septum, especially around danger
zones, and placement of superficial sutures when transitioning around the posterior inferior rim or crest. Uncomplicated closure of septation defects also requires an
understanding of the anatomic variations of special situations including atrial ventricular septal defect (AVSD),
double outlet right ventricle (DORV), AV discordance,
straddling or overriding AV valves, as well as malaligned
VSDs and trabecular, inlet or outlet extensions of perimembranous defects. Surgical AV conduction block necessitating pacemaker insertion should only occur due to
undefined anatomic variants or in patients who have a
genetic predisposition to AV conduction delay such as in
Tbx 5 or Nkx 2.5 mutations. Both Tbx-5 and Nkx 2.5 are
important for the development, maturation, and maintenance of the conduction system and haploinssufficiency
or heterozygous mutations result in AV conduction
block.
The study may demonstrate some biases due to (1) the
database search criteria matching VSD closure to pacemaker insertion, (2) the decision not to include septal
0003-4975/06/$32.00
doi:10.1016/j.athoracsur.2006.06.027