The use of double inlet as the defining feature for single ventricle, nonetheless, had become the accepted paradigm by the mid-1970s.4 Increasingly, however, the potential logical flaws in proposed systems of nomenclature became exposed5 as groups, such as the one from Europe in which one author of this review was a senior member,6 sought to rationalise the situation by creating rules so as formally to deny ventricular state to the smaller entity in those hearts with one big and one small chamber. Eventually, it was the advent of surgical correction which revealed the unity amongst the various contentious malformations. This was because creation of the Fontan circulation7, 8 became the chosen option whenever it was obvious that one chamber within the ventricular mass was incapable independently of supporting either the pulmonary or the systemic circulation. By that time, it had also been recognised that, in the most contentious lesions, those still described by many as “single ventricle with outlet chamber”, it was the atrioventricular connection rather than the ventricular mass which was truly univentricular.9, 10 Once having appreciated the most logical use of “univentricular”, it then became equally clear that problems of description were made logical not by denying ventricular state to one or other chamber, but rather by appreciating that the entire ventricular mass was functionally univentricular whenever one or other ventricle was incapable, for whatever reason, of supporting either the systemic or the pulmonary circulation. In this review, therefore, we describe the anatomical arrangements that produce these functionally univentricular situations, showing how current nomenclature is not only entirely logical, based on sound anatomical principles, but also is in keeping with recent advances made in understanding cardiac development. Not the least of the requirements of a system of nomenclature is that it be equally applicable and accurate to the clinician applying the various tools of cardiac imaging, to the surgeon, who with the naked eye can appreciate the surface characteristics of cardiac structures as they present themselves, and some additional features as he or she invades the inner reaches of the heart, and to the morphologist, who unpressured by the clock, may methodically reveal the intricate details of myocardial trabeculations and the disposition of the specialised conduction tissues. Like the cardiac structures they symbolise, the terms we use to communicate must be constructed from elements or building blocks such that the nomenclature is ever precise and accurate, though sometimes incomplete. The substance of this essay, therefore, is the description of those elements of anatomy, and thus nomenclature, as applied to functionally univentricular hearts.
Description of the ventricular mass
If we are to provide logical terms for description of ventricles, we must first be precise with what we understand by the ventricular mass. As far as we are concerned, this mass, made up of the entirety of the ventricular myocardium, extends from the distinct anatomical boundaries of the atrioventricular junctions to the equally discrete boundaries of the ventriculo-arterial junctions (Fig. 1). It so happens that the hinge points of the leaflets of the arterial valves cross the anatomic ventriculo-arterial junctions in semilunar fashion, but this does not disturb the anatomic integrity of the junctions. Within the normal ventricular mass as thus defined there are two chambers, best distinguished on the nature of their apical trabecular components. In the past, it was frequent practice to describe these normal ventricles as possessing two components, namely the “sinus” and the “conus”. There are, however, no boundaries, either anatomic or functional, that justify such division, and use of this system creates major problems when attempting logically to describe the patterns seen in the contentious hearts with a functionally univentricular arrangement. It is a simple matter, in contrast, to recognise the functional units of the normal ventricles as the inlet, the apical trabecular component, and the outlet (Fig. 1). Although again there are no clear-cut anatomical boundaries between these parts, as we will show, use of the tripartite approach11 greatly simplifies the problematic functionally univentricular hearts. Not all have accepted this concept, arguing that the “conus” is, in reality, no more than the infundibulum, and that the infundibulum possesses its own apical component which extends to the apex in parallel to the apical part of the sinus. If this concept were valid, then it follows that the wall of myocardium separating this part of the right ventricle from its counterpart in the left ventricle would be the infundibular septum. In reality, this is far from the case, since there is virtually no myocardium to be found in the normal heart that separates the sub-pulmonary infundibulum from the left ventricle, the infundibulum being the free-standing tube that is removed by the surgeon whenever he or she performs the Ross procedure.12 The muscle that separates the ventricular cavities in the normal heart can be described in no greater detail than as representing the muscular ventricular septum, but as we will discuss, this is the structure built from the primary ventricular septum as seen in the developing heart.13 And, because of the deeply wedged location of the normal left ventricular outlet, the inlet component of this muscular septum as seen from the right-side interposes between the right ventricular inlet and the sub-aortic outlet. Due to this, there is no justification, in the normal heart, for dividing the muscular septum into inlet, apical, and outlet components. These individual parts are only seen in congenitally malformed hearts, such as those producing the functionally univentricular arrangements.
Figure 1. This photograph of the morphologically right ventricle shows its extent from atrioventricular (red line) to the ventriculo-arterial (blue line) junctions, and that it is an easy matter to analyse the chambers in terms of its inlet, apical trabecular, and outlet components, albeit that there are no clear-cut boundaries between the components. The dotted lines are inserted for emphasis.
Ventricular description in congenitally malformed hearts
It follows from our description as given above that the normal heart is made up of two ventricles, each possessing inlet, apical trabecular, and outlet components. Analysis of the ventricles in malformed hearts is then simplified when it is appreciated that the inlets and outlets can be shared between the apical trabecular components in any pattern that is anatomically feasible, recognising also that, on occasion, these components can be completely absent. Thus, normal ventricles are also complete ventricles, since they possess all their component parts. In those hearts in which the inlets or outlets are abnormally shared between the apical components, then of necessity one of the ventricles will be incomplete, while in many instances the reciprocal ventricle will possess more than its normal compliment of components, and hence be a “super” ventricle. Such abnormal sharing does not necessarily mean that the abnormal heart will be incapable of supporting the circulations in biventricular fashion. For example, the left ventricle is incomplete in the setting of double outlet right ventricle, since it lacks its outlet component, yet many patients with double outlet can be repaired in biventricular fashion. This is not always the case. And this is the key to logical analysis of the functionally univentricular arrangements. Even hearts with two normally constituted ventricles can, on occasion, produce a functionally univentricular situation. Thus, when we encounter hearts with an imperforate outlet valve, such as the pulmonary valve, then depending on the size of the right ventricle, patients with such an arrangement might be suitable for repair in biventricular fashion, in functionally univentricular fashion, or in the fashion known as the “one and a half” ventricle approach. So as fully to describe ventricles, therefore, it is necessary first to account for their component make-up, and then to take note of their size. As we will show, in almost all instances those lesions in which one ventricle lacks its inlet component will be suitable only for functionally univentricular surgical treatment. Other lesions, in contrast, may be suitable for either biventricular or functionally univentricular therapeutic options depending on the particular anatomic conditions.
Ventricular description in the setting of the univentricular atrioventricular connection
The paradigm of the heart with the univentricular atrioventricular connection is double inlet left ventricle, and as we have already emphasised, this has probably been the most controversial of all hearts previously described, at some time and in some place, as being a “single ventricle”. Indeed, as we have also emphasised, there are still some authorities that maintain that the second chamber seen in the setting of double inlet left ventricle is no more than the infundibulum of the right ventricle. If that were the case, then in hearts in which both inlet components together with both outlet components were supported by the left ventricle, there would be nothing left to form a second chamber. In these hearts with double inlet and double outlet left ventricle (Fig. 2), however, the apical trabecular component of the morphologically right ventricle persists as an incomplete second ventricle (Fig. 3). The possibility that this pouch is no more than an apical extension of the infundibulum can then be discounted since the septum separating it from the left ventricle is known to carry the ventricular conduction tissues on its crest, and is nourished by the septal perforating arteries, these features showing unequivocally that it is the remnant of the muscular ventricular septum.14 Furthermore, we now know that, when first seen in the developing heart,13 this apical part of the forming right ventricle does not yet receive any direct inlet component, as is the case with double inlet left ventricle or tricuspid atresia, and already it carries the primordium of the ventricular conduction axis on its crest. Thus, there can now be little doubt that the so-called “outlet chamber” seen in double inlet left ventricle is no more and no less than the incomplete right ventricle, lacking in most cases its inlet component, and in some cases both its inlet and outlet components (Fig. 3).
Figure 2. The illustration shows double inlet to, and double outlet from, a dominant left ventricle.
Figure 3. This illustration shows the second ventricular chamber present in the heart shown in Figure 2 with double inlet and double outlet left ventricle. This ventricle is the persisting apical trabecular component of the incomplete right ventricle.
Having established that the smaller second chamber seen in double inlet left ventricle is no more than the incomplete right ventricle, it follows that this chamber will be directly comparable with the smaller chamber seen in classical tricuspid atresia, since the essence of this lesion, rather than being an imperforate tricuspid valve, is complete absence of the inlet component of the normal right ventricle, in other words, absence of the right atrioventricular connection (Fig. 4). Any differences in the nature of the small incomplete right ventricles seen in the settings of double inlet as opposed to tricuspid atresia have long been known to reflect difference in the ventriculo-arterial rather than the atrioventricular connections.3
Figure 4. The “four-chamber” section shows complete absence of the right atrio-ventricular connection, this being the commonest anatomical substrate for tricuspid atresia, with the heart itself having a dominant left ventricle.
It also follows that, in double inlet right ventricle, it is the left ventricle that is incomplete, lacking its inlet component, and almost always small. Double inlet right ventricle, therefore, also produces a functionally univentricular heart. In rare circumstances, it is not possible, during life, to discover a second chamber within the ventricular mass. These hearts potentially represent the true solitary ventricle. It is possible, nonetheless, that such lesions represent double inlet right ventricle, but in which the incomplete left ventricle, lacking in these instances its inlet and outlet components, is so small as to escape recognition. This is certainly possible, and we are aware of instances in which patients were labelled during life as having solitary ventricles, but in which sectioning of the heart at autopsy revealed the presence of a second small ventricle buried within the diaphragmatic surface of the ventricular mass. We are also aware of other instances in which serial histological sectioning has failed to reveal any evidence of a second ventricle, and the bizarre disposition of the ventricular conduction tissues strongly supports the notion that the chamber in these instances is a true solitary ventricle of indeterminate morphology.15 Our own opinion, therefore, is that the ventricular mass in the setting of either double inlet ventricle, or absence of either the right or left atrioventricular connection, can take the form of a dominant left ventricle with an incomplete right ventricle, a dominant right ventricle with an incomplete left ventricle, or very rarely, a solitary and indeterminate ventricle.
Describing the valves and ventricular positions
In the normal heart, the tricuspid valve is always found in the morphologically right ventricle, whilst the mitral valve is found in the morphologically left ventricle. Indeed, identifying the nature of the atrioventricular valves is often the most reliable means of distinguishing the morphology of the ventricles when each ventricle is joined to its own atrium. In the setting of hearts with univentricular atrioventricular arrangement, the situation is far less clear cut. This is because, when both valves are committed in their entirety to a dominant left ventricle, then both the valves resemble morphologically the mitral valve, since neither is attached directly to the ventricular septum as is the case for the morphologically tricuspid valve. Analysis of a reasonably large series of autopsied specimens with double inlet left ventricle16 showed that, in some instances, one of the valves had some affinities to the tricuspid valve, but this was not sufficiently consistent to provide a reliable means for clinical description. When seeking to describe the atrioventricular valves in hearts with double inlet ventricle, therefore, it is far more accurate, and logical, to describe them as being the right and left valves, depending on whether they drain the right or the left atrium. Similarly, in many instance it is possible to deduce the ventricular topology, or the direction of the loop, in the setting of the univentricular atrioventricular connection. When this is obvious, it is entirely reasonable to describe the position of the incomplete ventricle on the basis of the presumed loop, or topology. Again, nonetheless, it is often not possible accurately to predict the direction of ventricular looping. In such circumstances, it is better simply to state that the incomplete ventricle is right-sided, left-sided, or directly in front of or behind relative to the dominant ventricle. In this respect, in our experience, incomplete right ventricles have always been found antero-superior relative to the dominant left ventricle, while incomplete left ventricles are always positioned postero-inferiorly, albeit that either form of incomplete ventricle can be right-sided or left-sided. Describing the position of the incomplete ventricle as being right-sided or left-sided provides a much more consistent system of nomenclature when one atrioventricular connection is absent. For example, consider the situation in which the left atrioventricular connection is absent, the right atrium is joined to the dominant left ventricle, and the incomplete right ventricle is located antero-superiorly and to the left. In this setting, it is reasonable to draw the conclusion that, had the left atrioventricular valve been formed, it would have been of tricuspid morphology, and the heart would have been an example of congenitally corrected transposition. Due to this, some find it easier to describe the variant with absence of the left atrioventricular connection and dominant left ventricle as “tricuspid atresia”. It is not intuitive, however, that this terminology describes a patient in which it is the pulmonary venous return that is unable directly to reach the ventricular mass. It is arguable, therefore, that describing absence of the left atrioventricular connection is more logical, and more readily understandable.
Describing the interventricular communication
In the past, the foramen between the large and small chambers in so-called single ventricles has been described in various ways, using circumlocutions as complex as the “bulboventricular foramen”, or the “trabecular foramen”. Once it is accepted that both chambers in functionally univentricular hearts made up of one big and one small chamber are ventricles, albeit that one is incomplete, it becomes rather straight forward to described the communication between them as the ventricular septal defect! In hearts with a dominant left ventricle, when viewed from the incomplete right ventricle, this defect will be seen to have exclusively muscular borders, except for the rare circumstance in which the muscular outlet septum is replaced by a fibrous raphe and the defect is juxtaarterial and doubly committed. Such defects can safely be enlarged by resecting the margin closest to the obtuse margin of the ventricular mass (Fig. 5). In hearts with a dominant right ventricle, in contrast, the defect is typically perimembranous.
Figure 5. The illustration shows the incomplete right ventricle supporting the aorta in a patient with double inlet left ventricle and discordant ventriculo-arterial connections as it would be seen by the surgeon at operation. The course of the conduction tissue is marked relative to the ventricular septal defect, with the star showing the site of the atrioventricular node, and the red lines demonstrating the course of the atrioventricular bundle. The crosshatched area shows the safe area should a surgeon wish to enlarge the defect. Note that the defect has exclusively muscular borders. VSD: ventricular septal defect.
Conclusions
In our review, we have highlighted the problems that, in the past, created confusion and disagreements when describing the so-called “univentricular heart”. All agree that the problems with the “univentricular” situation are clarified simply by adding the adverb “functionally” to the term. The morphological disagreements are resolved, first, by noting that the ventricular mass extends from the atrioventricular to the ventriculo-arterial junctions, and then by dividing the normal ventricles within this mass in tripartite rather than bipartite fashion. Thereafter, it is a simple matter to note that, according to the way that the inlets and outlets are shared between the apical trabecular components, ventricles can be considered complete or incomplete. The real key to analysis is then provided by the appreciation that hearts can be functionally univentricular even when the ventricles possess all their component parts should one of the normally constituted ventricles be of insufficient size to support independently either the pulmonary or the systemic circulation.
Extant and popular systems of nomenclature, as currently used in the databases for paediatric cardiology and paediatric cardiac surgery, are either derived from commonly used but often imprecise terms, thus rendering them familiar but at best incomplete, and at worst inaccurate, or else take the “atomic”, or building block, approach but lack clarity with respect to the definitions of terms necessary for an all-inclusive approach to functionally univentricular hearts.
Description of all ventricles according to the principles outlined herein allows for nomenclature to be ever accurate within the confines of the ability of the clinician to discern features of the anatomy, and to be as complete as the extent of investigation and observation permits, even unto histological sectioning and microscopic study. Application of these principles then resolves any ongoing controversies in the nomenclature used for patients having functionally univentricular hearts.
Acknowledgment
Professor Anderson is supported by grants from the British Heart Foundation together with the Joseph Levy Foundation.
Research at the Institute of Child Health and Great Ormond Street Hospital for Children NHS Trust benefits from R&D funding received from the NHS Executive.
