|
 |
Invited article
(0.4M)
|
McCarthy KP*, Ho
SY**, Anderson RH***. Categorisation of ventricular septal defects: review
of the perimembranous morphology. Images Paediatr Cardiol
2000;3:24-40 |
|
*
|
Senior Research Technician, Institute of Child
Health, University College London, Cardiac Unit, 30 Guilford Street, London
WC1N 1EH |
|
**
|
Reader in Cardiac Morphology, National Heart
& Lung Institute, Imperial College, Paediatric Cardiac Morphology,
Dovehouse Street, London SW3 6LY |
|
***
|
Professor Robert H Anderson, Joseph Levy Professor
of Paediatric Cardiac Morphology, Institute of Child Health, University
College London, Cardiac Unit, 30 Guilford Street, London WC1 1EH |
| MeSH |
|
|
| Heart septal defects,
ventricular |
Heart defects, congenital |
Heart ventricle/pathology |
| Heart ventricles/abnormalities |
|
|
Abstract
The Association for European Paediatric Cardiology, the Society of
Thoracic Surgeons, and the European Association for Cardiothoracic Surgery,
have recently published detailed hierarchical listings for the description
of ventricular septal defects. This review details the anatomic basis for
the European codes, illustrating the phenotypic features of the various
holes that can be described as perimembranous ventricular septal defects.
Article
Introduction
There is currently a significant new impetus in
the drive to create a generally acceptable nomenclature to underscore the
creation of databases used for the collection and validation of results
in the treatment of congenital cardiac malformations. Thus, the Association
for European Paediatric Cardiology has published its short and long lists,
along with accompanying commentaries.1-3 The Society of Thoracic
Surgeons, together with the European Association for Cardiothoracic Surgery,
have now published an extensive supplement, containing a short surgical
code and companion of detailed hierarchical listings.4 To move
forward, and to achieve the international consensus desired by all, it
is essential that the potential discrepancies between these listings be
resolved. Most of the differences are semantic, and can be overcome by
the simple process of cross-mapping. But some differences are more profound,
and potentially reflect varying interpretations of the morphology of malformations.
It is possible that interventricular communications fall into this latter
category. If we are to resolve such potential scientific disagreements,
then it is essential that those offering definitions do so on the basis
of clear anatomic descriptions for the phenotypes of the different holes
considered to require distinction. In this review, and a subsequent article,
we will illustrate the features which are used to distinguish the entities
categorised within the European classification,(2-3) emphasising
those differences which might exist in comparison to the surgical hierarchies.4
In this initial discussion, we will concentrate on the defect we define
as being perimembranous. But first of all, we must explain what we perceive
to be the defect.
What is a ventricular
septal defect?
|
The problems which
exist cannot be resolved without first deciding on the boundaries of the
space to be defined as the “septal defect”. When there is a simple hole
within the muscular septum separating the right and left ventricular cavities
(Fig.1), then there is no problem.
|
Difficulties arise when the orifice of an arterial valve overrides
the crest of the muscular ventricular septum (Fig.2).
|
Figure 1: Four-chamber section showing communication between
the two ventricles through a muscular ventricular septal defect (VSD)
|
Figure 2: Overriding of the muscular ventricular septum by the
orifice of the aortic valve
|
|
In this setting, some
might define the upward extension of the plane of the long axis of the
ventricular septum as representing the septal defect (Fig.3).
|
Figure 3: The plane of the long axis of the muscular ventricular
septum extends up into the arterial outflow tract, and reaches to the leaflets
of the aortic valve
|
Such a plane certainly
represents the location of the interventricular communication but, in the
heart illustrated, the leaflets of the aortic valve form the roof of the
defined defect. No surgeon would place a patch on this location to restore
ventricular septal integrity. In fact, in the heart illustrated, the surgeon
would place the patch so as to reconnect the aorta with the left ventricle.
Thus, for the surgeon, it is the right ventricular margin of the cone of
space subtended by the leaflets of the overriding valve that is the locus
of interest (Fig. 4).
Figure 4: Diagrammatic representation of the cone of space found
between the crest of the muscular ventricular septum and the leaflets of
the overriding aortic valve
|
Because of the surgical
focus on this plane, this is the space that we define as THE septal defect.
At the same time, we recognise that the left ventricular margin of the
cone of space is also important as the outflow tract from the left ventricle
(Fig. 5).
|
These differences in definition of the hole chosen to represent
the defect have important consequences for categorisation. Consider again
the situation in which the extension of the long axis of the ventricular
septum is taken as the defect. The postero-inferior rim of this particular
plane is made up of fibrous continuity between the leaflets of the aortic
and mitral valves (Fig. 6). As we will see, it is fibrous continuity between
the aortic valve and an atrioventricular valve that we use as our defining
feature for a certain set of septal defects.
|
Figure 5: The leftward margin of the interventricular
cone of space is the outflow tract of the left ventricle
|
Figure 6: The outflow tract of the left ventricle is roofed
by fibrous continuity between the leaflets of the mitral and aortic valves
(yellow arrow). The long axis of the ventricular septum extends up into
the aortic outflow tract (red arrow)
|
|
But, for us, the significant
atrioventricular valve for the purposes of categorisation is the tricuspid
valve. The reason we choose the tricuspid valve is shown when we study
again our illustrated heart with overriding of the aortic valve. In the
chosen specimen (Fig. 7), a muscular bar interposes between the aortic
and tricuspid valves. The presence of this muscular ledge between the edge
of the defect and the fibrous membranous septum protects the specialised
muscular axis responsible for atrioventricular conduction (See below).
Consequently, we make our categorisation of ventricular septal defects
by viewing the margins of the plane chosen to represent the defect from
the aspect of the surgeon approaching from the right side.
|
Figure 7: The rightward margin of the interventricular
cone of space has a muscular ledge (yellow line) that separates the aortic
(Ao) from the tricuspid valve (TV)
|
What is the significance of atrioventricular junction
morphology?
| Perhaps the major scientific difference between the proposed classifications
for holes between the ventricles is the continuing problem of the “atrioventricular
canal ventricular septal defect”. This particular problem is resolved by
paying attention to the morphology of the atrioventricular junctions. Thus,
there is no question but that hearts can exist with a common atrioventricular
junction, and with shunting confined at ventricular level. These hearts
are fundamentally different nevertheless, from extensive holes between
the ventricles which abut on the membranous part of the septum, and which
open primarily into the inlet component of the right ventricle, being shielded
by the septal leaflet of the tricuspid valve. Within the European code,
it is only the hearts with a common atrioventricular junction (Fig. 8)
which are considered to represent “atrioventricular canal” defects, irrespective
of the level of shunting across the defect. |
Figure 8: Viewed from the apex, the common valve connects to
both ventricles and guards a common atrioventricular junction
|
This is because, in these hearts, the left atrioventricular valve is
half of an essentially common atrioventricular valve, even if the valvar
orifice is divided into separate parts for the right and left ventricles,
as seen in the “ostium primum” defect. This left valve guarding half of
the common junction possesses three leaflets, and closes in trifoliate
fashion. Hearts with junctional morphology of this type do exist with shunting
possible only at ventricular level because the bridging leaflets of the
common atrioventricular valve are firmly attached to the leading edge of
the atrial septum (Fig. 9).
Figure 9: In some hearts with a common atrioventricular valve,
as seen here from the left ventricle, the leaflets of the common valve
are attached to the leading edge of the atrial septum. In this case it
results in shunting only at ventricular level
| When such hearts are examined from the right side, the bridging leaflets
are seen as they cross through the septal defect (Fig. 10). |
Figure 10: From the right ventricular aspect, the bridging leaflets
cross through the septal defect
|
This arrangement is fundamentally different from a ventricular septal
defect that opens primarily to the inlet of the right ventricle in hearts
with separate right and left atrioventricular junctions. In these latter
malformations, the left atrioventricular valve is morphologically a mitral
valve and, like the normal mitral valve, has but one zone of apposition
between its paired leaflets (Fig. 11).
Figure 11: Viewed from the apex in a normal heart, the aortic
and mural leaflets of the mitral valve close along a solitary zone of apposition
(red dots)
The tricuspid valve in this setting is then located exclusively within
the separate right atrioventricular junction, with no straddling of its
valvar leaflets. The criterion concerning the arrangement of the atrioventricular
junctions, however, serves also to distinguish those hearts with inlet
ventricular septal defects in which there is straddling and overriding
of the tricuspid valve. In these hearts, the right atrioventricular junction
is shared to varying extent between the ventricles. The ventricular septal
defect seen with such straddling tricuspid valves is also frequently categorised
as representing an “atrioventricular canal defect” 5-6. But,
although the right atrioventricular junction overrides the crest of the
muscular ventricular septum, it is not part of a common junction. Indeed,
the separate left atrioventricular junction (Fig. 12) is guarded by a morphologically
mitral valve, distinguishing the hearts from those with a common atrioventricular
junction. It is the common junction, therefore, which should be the hallmark
for distinction of the “atrioventricular canal malformation”. Only those
hearts with exclusively ventricular shunting in the setting of such a common
junction (Fig. 9-10) should be named as “atrioventricular canal VSD’s”.
Figure 12: Four-chamber section highlighting the malalignment
of the atrial and ventricular septums. In the setting of straddling and
overriding of the tricuspid valve. The mitral valve guards a separate left
atrioventricular connections
How
then, do we Categorise Holes between the Ventricles?
On the basis of the features considered above, our observations suggest
that, when the surgeon views the boundaries of holes between the ventricles
from the right side, these holes can be placed into one of three categories.
The largest category seen in the autopsy room, and at surgery, is that
where part of the border of the defined defect is composed of fibrous continuity
between the leaflets of the aortic and tricuspid valves (Fig. 13).
Figure 13: The heart illustrated has tricuspid-aortic valvar
continuity, this being the diagnostic feature of a perimembranous ventricular
septal defect
We call these holes perimembranous defects. This is because they abut
directly on the atrioventricular component of the membranous septum. In
addition, frequently they possess a remnant of the interventricular membranous
septum reinforcing their postero-inferior rim. The remainder of this initial
review will be devoted to these perimembranous defects. The second category
includes all holes which, when viewed from their right side, have exclusively
muscular borders. As already discussed, there is general agreement that
these are muscular ventricular septal defects (Fig. 1). The third category
is the one in which the defining feature is fibrous continuity between
the leaflets of the aortic and pulmonary valves, or else the presence of
a common truncal valve. We call these defects doubly committed and juxtaarterial.
Our second review will detail the phenotypic features of the muscular and
doubly committed defects. We use this categorisation into three basic patterns
in all hearts, irrespective of their segmental interconnections. We also
recognise the defect in which there is a common atrioventricular junction
and shunting exclusively at ventricular level (Fig. 9-10). But, as explained
above, because of the significance of the structure of the trifoliate left
atrioventricular valve, and because the hearts lack atrioventricular septal
structures, we prefer to categorise them as atrioventricular septal defects,
albeit with only ventricular shunting across the defect.
The
Philosophy of Categorisation
When categorising holes
between the ventricles, therefore, we use a system of description which
accounts not only for “isolated” ventricular septal defects, but also for
those seen in the presence of other associated malformations, such as tetralogy
of Fallot, double outlet ventricle, tricuspid atresia, double inlet ventricle,
and so on.If all these holes are accurately categorised, then the description
chosen confers important information regarding their relationship to structures
such as the conduction tissue axis and the arterial valves. If, in addition,
their size and potential for closure are defined, these features having
important consequences upon the haemodynamics of ventricular shunting,
then the strategy for subsequent treatment can properly be determined.
The Membranous Septum and Perimembranous Septal Defects
| The normal membranous septum is usually a small structure (Fig. 14). |
Figure 14: Four-chamber section showing the normal location
of the small membranous septum
|
It is an integral
part of the aortic root, where it is confluent with the right fibrous trigone,
forming the base of the fibrous triangle between the right and the non-coronary
leaflets of the aortic valve (Fig. 15).
Figure 15: This figure shows the close relationship of the membranous
septum to the right fibrous trigone, which is adjacent to the area of fibrous
continuity between the aortic and mitral valve leaflets (red dots)
The atrioventricular conduction axis penetrates through the atrioventricular
part of this septum, this being demarcated by the hinge of the septal leaflet
of the tricuspid valve attached to the right side of the septum. The bundle
itself is then sandwiched between the interventricular component of the
septum and the muscular septum (Fig. 16).
Figure 16: Histology section showing the conduction tissue on
the crest of the ventricular septum, along with its relationship to the
aortic and tricuspid valves
When the septum “comes apart” in the setting of perimembranous septal defects,
it does so between its fibrous and muscular components. The conduction
bundle stays on the crest of the muscular ventricular septum. Such perimembranous
defects can then be sub-classified according to the direction in which
they open to the right ventricle, but the location of the conduction axis
stays constant. To account for the variations, the right ventricle itself
can be divided into three regions, inlet, outlet, and apical. These parts
correspond to their relationship with the tricuspid valve, the sub-pulmonary
infundibulum, and the apical trabeculated portion of the morphologically
right ventricle, respectively (Fig. 17).
Figure 17: Diagrammatic representation showing how the right
ventricle can be described in terms of three components
Should the defect open into more than one of these regions, then it
is described as a confluent defect. Previously, we have considered the
muscular septum itself as having inlet, apical, and outlet components.7
We have revised this description. Because of the wedged location of the
normal subaortic outflow tract, the inlet of the right ventricle is mostly
separated by the muscular septum from the outlet of the left ventricle
(Fig. 11). Thus, the septum we considered to be “inlet” is, in reality,
an inlet/outlet septal structure. Furthermore, the larger part of the right
ventricular outlet is made up of the free-standing sleeve of subpulmonary
infundibulum. The “outlet septum”, therefore, is not nearly as extensive
as was initially believed. In fact, it is not possible with any certainty
to distinguish the precise location of the small outlet septal component
of the normal heart.8 These facts call into question the wisdom
of dividing the ventricular septum in tripartite fashion. It remains entirely
accurate and simple, nonetheless, to describe the inlet, apical, and outlet
components of the right ventricle (Fig. 17).
Perimembranous defects
opening into the inlet of the right ventricle are seen beneath the septal
leaflet of the tricuspid valve (Fig. 18). Here, there is fibrous continuity
in the roof of the defect between the leaflets of the tricuspid and the
mitral valves, and this fibrous tissue forms the postero-inferior border
of the defect. The hole between the ventricles can partially be blocked
by the septal leaflet of the tricuspid valve, by aneurysms of the membranous
septum, or by accessory tissue tags derived from the tricuspid valve. The
superior border is made up of the remnant of the membranous septum and
the fibrous tissue of the aortic root.
Figure 18: From the right ventricle, the perimembranous defect
opening to the inlet is seen under the septal leaflet of the tricuspid
valve
Pure outlet defects,
without malalignment of the septal structures (Fig.19), are rare. This
is because, as emphasised, the outlet part of the muscular septum is very
small in the normal heart. Much more commonly, outlet perimembranous defects
are associated with malalignment of the muscular outlet septum and overriding
of the aortic valve. When seen, they are usually, but not always, associated
with obstructive lesions of the ventricular outflow tracts (see below).
Figure 19: Perimembranous ventricular septal defect opening
to the outlet of the right ventricle
When the outlet septum
is maligned in antero-cephalad fashion, but there is no stenosis of the
outflow tract, the hearts are classified as “Eisenmenger” ventricular septal
defects.9 In this setting, it is an easy matter for the blood
from the right ventricle to be directed through the ventricular septal
defect to the aorta when the pulmonary resistance exceeds the systemic,
and they are typically seen in the setting of pulmonary hypertension. More
frequently, the defect is found with antero-cephalad deviation of the muscular
outlet septum in the setting of tetralogy of Fallot. The hallmark of this
defect is obstruction of the subpulmonary outflow tract produced by the
“squeeze” between the deviated outlet septum and the hypertropied septoparietal
trabeculations.10 This is the ventricular septal defect termed
by some a “conoventricular defect”. Significantly, although the defect
is most usually perimembranous (Fig. 20), it can also be found with a muscular
postero-inferior rim (Fig. 21).
Figure 20: In tetralogy of Fallot, the ventricular septal defect
(VSD) usually has a perimembranous postero-inferior margin
|
Figure 21: In this case of tetralogy of Fallot, a muscular bar
separates the aortic valvar leaflets from the defect, producing a muscular
postero-inferior rim
|
It is the precise morphology of this postero-inferior margin which
dictates the vulnerability of the atrioventricular conduction axis. The
European classification provides system of categorisation that differentiates
between these features.2-3 Defects opening exclusively to the
apical component of the ventricle are relatively rare compared to those
opening to either inlet or outlet, or those which are confluent. The defects
opening towards the apex and remaining perimembranous are small, frequently
being slit-like, with the medial papillary muscle marking the apical extent
of the defect. The conduction bundle is typically closely adjacent to the
postero-inferior margins of the defect in these cases, but a fan-like right
bundle branch originating at the same level from the main axis as the left
bundle branch has also been described.11 Oftentimes the septal
leaflet of the tricuspid valve is cleft. Much more common are the larger
defects opening towards the apex which are confluent with either inlet
or outlet perimembranous extensions.
The Conduction Axis
in Perimembranous Septal Defects
Defining the margins of the defect as we have suggested provides
an accurate guide to the route taken by the conduction axis.12
In perimembranous septal defects, the conduction tissue is always found
postero-inferior to the defect, the only variation being seen in association
with straddling and overriding of the tricuspid valve. In the typical situation,
the atrioventricular node is found in its expected location at the apex
of the triangle of Koch. The conduction bundle then penetrates through
the myocardium from the apex of the triangle of Koch towards the left ventricular
outflow tract (Fig. 22).
Figure 22: The conduction axis passes from the atrioventricular
node in the right atrium (yellow dot), towards the crest of the ventricular
septum (blue dots), and is positioned postero-inferiorly relative to the
perimembranous ventricular septal defect
The conduction axis then branches on the crest of the septum, with the
right bundle branch emerging beneath the medial papillary muscle (Fig.
23).
Figure 23: Diagrammatic representation of the direction of the
conduction axis as seen from the right atrial aspect in a perimembranous
ventricular septal defect, seen here in the setting of tetralogy of Fallot
The area for the surgeon to avoid is along the right ventricular border,
and is demarcated by a line drawn from the apex of the triangle of Koch
to the medial papillary muscle. The phenotypic feature in cases with straddling
and overriding of the tricuspid valve is the malalignment between the atrial
and ventricular septums (Fig. 12). Because of this, the atrioventricular
node is no longer found at the apex of the triangle of Koch, but is abnormally
positioned, being formed at the point where the muscular ventricular septum
meets the atrioventricular junction (Fig. 24).
Figure 24: In this figure, which shows deviation of the ventricular
septum, the location of the atrioventricular node is not at the apex of
the triangle of Koch, but is found where the ventricular septum meets the
right atrium (yellow dot). The conduction axis then passes towards the
crest of the ventricular septum (blue dots)
Even when seen in isolation, the ventricular septal defect is a common
congenital cardiac malformation. Consensus is needed to clarify description
of the defect, and this requires agreement on the fundamental features
that underscore definition. The first step in this process is to agree
the plane of space which is to be defined as “the” defect. Further steps
towards consensus can then be taken by providing a coherent definition
of the "atrioventricular canal ventricular septal defect", and agreeing
on what precisely is meant by the "conoventricular" defect. The field of
diagnosis is becoming ever more advanced with the advent of techniques
such as 3-dimensional echocardiography and magnetic resonance imaging.
These sophisticated diagnostic methods permit the defect to be viewed in
the planes necessary to define precisely the postero-inferior margin of
the defect, its size, and its location. This information then permits identification
of the position of the conduction axis. The categorisation as suggested
here concentrates on these features of clinical significance, and at the
same time facilitates assessment of the probability of spontaneous closure
of the hole, and the likelihood of developing aortic regurgitation. Use
of these accurate definitions, therefore, coupled with exact descriptions
of the defect, should help the diagnostic team, the surgeon, and ultimately
the patient.
References
-
Franklin RCG, Anderson RH, Daniëls
O, Elliott M, Gewillig MHML, Ghisla R, Krogman ON, Ulmer HE, Stocker FP.
Report of the Coding Committee of the Association for European Paediatric
Cardiology. Cardiol Young 2000; 10: Suppl 1; 1-7
-
Franklin RCG, Anderson RH, Daniëls
O, Elliott M, Gewillig MHML, Ghisla R, Krogman ON, Ulmer HE, Stocker FP.
The European Paediatric Cardiac Code – The Short List. Cardiol Young 2000;
10: Suppl 1; 8-26
-
Franklin RCG. The European Paediatric
Cardiac Code Long List: structure and function. Cardiol Young 2000; 10:
Suppl 1; 27-146
-
Jacobs JP, Burke RP, Quintessenza JA,
Mavroudis C. Congenital Heart Surgery Nomenclature and Database Project:
ventricular septal defect. Ann Thorac Surg 2000; 69 (4 Suppl): S25-35
-
Neufeld HN, Titus JL, DuShane JW, Burchell
HB, Edwards JE. Isolated ventricular septal defect of the persistant common
atrioventricular type. Circulation 1961; 23: 685-696
-
LaCorte MA, Fellows KE, Williams RG.
Overriding tricuspid valve: echocardiographic and angiographic features.
8 cases of ventricular septal defect of atrioventricular canal type. Am
J Cardiol 1976; 37: 911-919
-
Soto B, Becker AE, Moulaert AJ, Lie
JT, Anderson RH. Classification of ventricular septal defects. Br Heart
J 1980; 43: 332-343
-
Merrick AF. Yacoub MH. Ho SY. Anderson
RH. Anatomy of the muscular subpulmonary infundibulum with regard to the
Ross procedure. Ann Thor Surg 2000: 69: 556-61
-
Fukuda T, Suzuki T, Ito T. Clinical
and morphologic features of perimembranous ventricular septal defect with
overriding of the aorta - the so-called Eisenmenger ventricular septal
defect. A study making comparisons with tetralogy of Fallot and perimembranous
ventricular defect without aortic overriding. Cardiol Young 2000;10: 343-352
-
Gatzoulis MA, Soukias N, Ho SY, Josen
M, Anderson RH. Echocardiographic and morphological correlations in tetralogy
of Fallot. Eur Heart J 1999; 20: 221-231
-
Ueda M, Becker AE. Morphological characteristics
of perimembranous ventricular septal defects and their surgical significance.
Int J Cardiol 1985; 8: 149-157
-
Anderson RH, Wilcox BR. The surgical
anatomy of ventricular septal defect. J Card Surg 1992; 7: 17-35
www.impaedcard.com