Introduction

The atrioventricular node is the atrial component of that part of the heart known to electrophysiologists as the specialised atrioventricular junctional area. For a structure that has been recognised for just less than one hundred years, the node has been the source of a remarkable number of controversies, as indeed has the junctional area itself. The first suggestion of a special bundle of muscle being responsible for atrioventricular conduction was made by Wilhelm His the younger in 18931. In the same year, Stanley Kent² argued, erroneously, that there were multiple muscular pathways crossing the insulating atrioventricular junctions of the normal heart. At this stage, however, there had been no suggestion of "specialisation" of the atrial end of the bundle. Indeed, subsequent to these initial arguments, distinguished anatomists such as Arthur Keith³ had been skeptical of the existence of an atrioventricular bundle. The situation was clarified by Sunao Tawara, a young Japanes pathologist working in the laboratory of Ludwig Aschoff. His monumental monograph was written in the German language⁴. The work is now available in an English translation⁵. With illustrations juxtaposed to the appropriate text in this translation, we can now appreciate fully the accuracy of Tawara's initial descriptions. We can also appreciate that, if we abide by his definitions, then the controversies that have emerged over the last nine decades can readily be resolved. It is attention to the descriptions of Tawara⁴, therefore, coupled with application of the criterions proposed by Aschoff⁶ and Monckeberg⁷ for recognition of histologically "specialised" myocardium, that provide the basis for understanding the structure of the atrioventricular node. This knowledge, nonetheless, needs nowadays to be placed in the context of clinical investigation. This means that the structures should be described as seen with the heart occupying its normal position within the body, in other words, in attitudinally correct orientation⁸. Before describing the location and structure of the node, therefore, we commence by emphasising the attitudinally correct description of the parts of the heart that contain the atrioventricular node and the conduction pathways.

Attitudinally Correct Description

All structures within the body should be described on the basis of the anatomical position. This means that the subject is viewed as though standing upright, and facing the observer. Structures are then described in terms of their three orthogonal planes. These are the coronal planes, running from side to side, the sagittal planes, running from front to back, and the transverse planes, running from top to bottom. When the heart is considered in this context, and evaluated in the left anterior oblique position, it becomes clear that, currently, major problems exist with the conventional descriptions of cardiac structure. This is because the diaphragmatic surface of the ventricular mass is directly below that part of the heart supporting the arterial trunks (Figure 1). Yet, at present, this part of the ventricular mass is usually described as being "posterior". In reality, it is inferior. The posterior part of the ventricular mass is, self-evidently, that closest to the spine. These problems also relate to the atrial chambers. The right atrial appendage is an anterior structure when the heart is normally positioned within the body, and the septal component of the right atrium is posterior. When the heart is opened through the right atrium, and viewed in attitudinally correct position, then the part of the right atrium containing the atrioventricular node, the triangle of Koch, has its apex pointing upwards (Figure 2). In current descriptions, however, the triangle of Koch is usually illustrated with its apex pointing to the right. If we are to put our understanding of nodal structure into the appropriate clinical context, then we must all adopt the attitudinally correct adjectives (Figure 1).

Click on images for legends and to enlarge.
Figure 1.	Figure 2.

The Location of the Atrioventricular Node

As shown by Tawara⁴, the atrioventricular conduction axis is a continuous system of histologically discrete cells which have their origin in the atrial myocardium and their insertion in the ventricular Purkinje cells. It is the atrial components of this axis that constitute the atrioventricular node, or the "knoten" of Tawara⁴. The node, along with its surrounding zones of transitional cells, is positioned at the base of the atrial septum, occupying the upper part of the triangular area first illustrated by Koch (Figure 2). The apex of the triangle is the atrioventricular component of the membranous septum (Figure 3). This fibrous structure is continuous on its left side with the thickened rightward end of the region of fibrous continuity between the leaflets of the aortic and mitral valves (Figure 4). This fibrous thickening, known as the right fibrous trigone, together with the membranous septum, forms the so-called central fibrous body. When seen from the right side, the structures forming the sides of Koch's triangle insert directly to the membranous septum (Figure 3). The more posterior border is the fibrous strand continuation of the valve of the inferior caval vein, the tendon of Todaro¹⁰. The anterior border of the triangle is the line of attachment of the septal leaflet of the tricuspid valve. It is this structure that crosses the right ventricular aspect of the membranous septum in most hearts, dividing it into atrioventricular and interventricular components (Figure 3). The muscular triangle delimited by these two borders separates the hinge of the tricuspid valve in the right ventricle from that of the mitral valve in the left ventricle, and seems to be a septum. In the strictest sense, however, the area is not truly septal, since a fibroadipose continuation of the inferior atrioventricular groove separates the atrial wall forming the floor of the triangle from the underlying ventricular musculature (Figure 5). The artery supplying the atrioventricular node traverses this tissue plane, extending superiorly to the point where the node penetrates the plane of atrioventricular insulation to become the bundle of His. The area within the triangle of Koch, therefore, is better described as an atrioventricular muscular sandwich, with the insulating fibroadipose tissue forming the "meat" between the two muscular "slices of bread"⁹. The base of the triangle is positioned inferiorly, and is occupied by the coronary sinus. Adjacent to the mouth of the sinus, which is guarded by the Thebesian valve, there is an important isthmus bounded by the coronary sinus itself posteriorly, and the hinge of the tricuspid valve anteriorly. This is the septal isthmus. The musculature of the septal isthmus then continues inferiorly as the vestibule of the tricuspid valve, while the musculature around the mouth of the coronary sinus itself continues inferiorly into another isthmus. This second isthmus is confined between the mouth of the inferior caval vein and the leaflets of the tricuspid valve. This is the inferior isthmus, or cavo-tricuspid, isthmus. It has trabecular, membranous, and vestibular components¹¹ (Figure 6). The membranous part is often expanded to form a prominent diverticulum. Although usually known as the sub-Eustachian sinus, with the heart in attitudinally correct position, the sinus is seen to be sub-Thebesian, lying directly beneath the mouth of the coronary sinus (Figure 2). Two other muscular areas are important in the context of the relations of the triangle of Koch and the approaches to the atrioventricular node. The first is the muscular ridge between the mouth of the coronary sinus and the oval fossa, also known as the Eustachian ridge. The second area is the atrial wall between the mouths of the coronary sinus and the inferior caval vein. Frequently described as the "sinus septum", this second area is no more than the fold between the two venous walls at their insertion into the musculature of the right atrium.

Click on images for legends and to enlarge.
Figure 3.	Figure 4.

Figure 5.	Figure 6.

Figure 7.	Figure 8.

The Atrioventricular Node relative to Koch's Triangle

The histologically specialised tissues making up the atrioventricular node occupy a relative small area within the triangle of Koch. The atrioventricular node itself is a half-oval of cells forming the base of the atrial wall, and set against the fibrofatty tissue of the atrioventricular junction (Figure 7). When traced inferiorly, the half-oval breaks up into two extensions, which run towards the attachments of the mitral and tricuspid valves (Figure 8). This dense knot of cells is the compact node. The artery supplying the node usually enters this compact nodal tissue between the two inferior extensions. The strand of compact nodal cells running towards the tricuspid valve is the more prominent of these extensions. Its extent inferiorly, however, varies from heart to heart¹². The compact node itself is then usually surrounded by an additional half-oval of cells intermediate in their morphology between the nodal cells and the ordinary atrial musculature. These are the transitional cells (Figures 7,8). Of necessity, the transitional cells are not insulated by fibrous tissue from the remainder of the atrial myocardium. As we will see, because of this they do not qualify for consideration as conducting tracts, but since they can be traced from section to section, they can be recognised as being histologically specialised (see below). Additional transitional cells are found interposing between the atrial myocardium and the inferior nodal extensions, while still further areas of transitional cells interpose between the left edge of the compact node and the left atrial aspect of the septum. Thus, on all sides the compact node has transitional cells interposed between it and the ordinary atrial myocardium. The ordinary atrial myocardium then forms a further discrete overlay at the apex of Koch's triangle. This layer extends anteriorly to insert into the vestibule of the tricuspid valve (Figure 7). These overlay cells originate from the atrial wall in front of the oval fossa, which corresponds with the area targetted as the so-called fast pathway into the atrioventricular node (Figure 6). It is composed of ordinary atrial myocardium.

The slow pathway into the atrioventricular node is located within the septal isthmus (Figure 6), albeit that its anatomic nature remains to be clarified. In two cases we studied in which this pathway had been ablated to cure atrioventricular nodal re-entry tachycardia¹³, in both the lesion placed to ablate the pathway was located in ordinary atrial myocardium, and was distant from the compact atrioventricular node, its inferior extensions, and the areas of histologically discrete transitional cells. Others argue, nonetheless, that the inferior extension from the compact node might itself form the slow pathway¹². In this respect, even in those cases where the inferior extension is not identified histologically, it is certain that cells with an initially nodal phenotype are present in this area. This is because the atrioventricular node and bundle are derived from a ring of cells which initially surrounds the embryonic interventricular foramen¹⁴. With subsequent development, part of this ring becomes incorporated into the vestibule of the tricuspid valve, but loses its histologically discrete phenotype. Remnants of this ring almost certainly form the nodal structures which Kent² mistakenly believed provided multiple muscular atrioventricular connections in the normal heart. It is also the case that the area ablated as the slow pathway to cure nodal reentry tachycardia was initially traversed by this ring, even if, in the postnatal heart, the cells are no longer recognised as being histologically specialised.

There are, however, still other morphologic possibilities that might account for the existence of the slow pathway. In particular, there is marked variation in the alignment of the myofibrils in the septal isthmus from heart to heart (Figure 6). It is now well established that this alignment of the myofibrils is responsible for non-uniform anisotropy, which could well produce a preferential route of conduction into the node through histologically ordinary atrial myocardium. The coupling of the myocardial cells by connexins may also differ in this area. The anatomic substrate for the slow pathway has, therefore, still to be established. At the apex of the triangle of Koch, the compact atrioventricular node is set against the central fibrous body. Study of serial sections shows that, as the axis of specialised musculature extends superiorly, it becomes engulfed by fibrous tissue, thus insulating it from the adjacent atrial myocardium. According to Tawara, it is at the point at which the conduction axis becomes surrounded by fibrous tisuue that the compact atrioventricular node becomes the bundle of His (Figure 9). As noted by Tawara, there is very little histological difference in the human between the cells of the axis still in contact with atrial myocardium (Figure 7) as opposed to that part which is insulated within the central fibrous body (Figure 10). The anatomic difference between these two parts, nonetheless, is unequivocal and clear-cut.

When traced superiorly, this penetrating part of the bundle turns slightly leftward and emerges from the fibrous tissue on the crest of the muscular ventricular septum. In the human, this penetrating part of the axis is remarkably short. Having reached the crest of the muscular septum, and still contained within a fibrous sheath, the bundle runs a variable non-branching course before beginning to give off the fascicles of the left bundle branch (Figure 9). Having run forward along the septal crest, and having given rise to the left bundle branch, the axis gives rise to the cord-like right bundle branch. This turns back through the substance of the muscular ventricular septum, emerging in the right ventricle in relation to the medial papillary muscle. The bundle branches themselves extend in insulated fashion down the two sides of the ventricular septum, with the left bundle branch having a trifascicular form (Figure 11). Only at the apex do the conducting cells lose their fibrous sheath, becoming the ventricular Purkinje cells.

Click on images for legends and to enlarge.
Figure 9.	Figure 10.

Figure 11.	Figure 12.

The Criterions of Aschoff and Monckeberg

Controversies relative to the recognition and definition of the atrioventricular node largely devolve on the criterions to be used for "histological specialisation". In this respect, we should remember that Aschoff and Monckeberg, in 1910, provided excellent interpretations of the criterions which needed to be established before specialised muscular structures could be described within the heart. These criterions^6,7 were threefold. First, it was necessary that the cells be histologically distinct from their neighbours. Second, they proposed that it should always be possible to trace the cells from section to section in serially prepared histological material. Third, the cells within any proposed tract should be insulated by fibrous tissue. The right bundle branch is an excellent example of a bundle fulfilling these criterions (Figure 12). Judged relative to these criterions, it now becomes clear that the atrioventricular node and the transitional cells, whilst satisfying two of the criterions, are not insulated by fibrous tissue from the remainder of the atrial myocardium. Thus, the node and the transitional cells do not constitute anatomically discrete "conducting tracts", but rather are the recognisable atrial component of the axis of atrioventricular conduction.

[back to top]

References

1. His W, Jr 1893 Die Thätigkeit des embryonalen Herzens und deren Bedeutung für die Lehre von der Herzbewegung beim Erwachsenen. Arb aus d med Klinik zu Leipzig 1:14-49.

2. Kent AFS 1893 Researches on the structure and function of the mammalian heart. J Physiol 14:233-254.

3. Keith A 1950 An Autobiography. Philosophical Library, New York. pp 254-259.

4. Tawara S 1906 Das Reizleitungssystem des Säugetierherzens. Eine Anatomisch-Histologische Studie Über das Atrioventrikularbündel und die Purkinjeschen Fäden. Gustav Fischer, Jena.

5. Tawara S 2000 The Conduction System of the Mammalian Heart. An Anatomico-histological Study of the Atrioventricular Bundle and the Purkinje Fibers. Translated by K Suma and M Shimada. Imperial College Press, London.

6. Aschoff L 1910 Referat uber die Herzstorungen in ihren Beziehungen zu den Spezifischen Muskelsystem des Herzens. Verh Dtsch Pathol Ges 14:3-35.

7. Monckeberg JG 1910 Beitrage zur normalen und pathologischen Anatomie des Herzens. Verh Dtsch Pathol Ges 14:64-71.

8. Cosio FC, Anderson RH, Kuck K, Becker A, Borggrefe M, Campbell RWF, Gaita F, Haïssaguerre M, RufilanchasJJ, Thiene G, Wellens HJJ, Langberg J, Benditt DG, Bharati S, Klein G, Marchlinski F, Saksena S. Living anatomy of the atrioventricular junctions. A guide to electrophysiological mapping. A consensus statement from the Cardiac Nomenclature Study Group, Working Group of Arrhythmias, European Society of Cardiology, and the Task Force on Cardiac Nomenclature from NASPE Circulation 1999;100:e31-e37.

9. Anderson RH, Ho SY, Becker AE. Anatomy of the human atrioventricular junctions revisited. Anat Rec 2000;260:81-91.

10. Ho SY, Anderson RH 2000 How constant anatomically is the tendon of Todaro as a marker for the triangle of Koch? J Cardiovasc Electrophysiol 11:83-89.

11. Cabrera JA, Sanchez-Quintana D, Ho SY, Medina A, Anderson RH. The architecture of the atrial musculature between the orifice of the inferior caval vein and the tricuspid valve: the anatomy of the isthmus. J Cardiovasc Electrophysiol 1998;9:1186-1195.

12. Inoue S, Becker AE. Posterior extensions of the human compact atrioventricular node. A neglected anatomic feature of potential clinical significance. Circulation 1998;97:188-193

13. Sanchez-Quintana D, Davies DW, Ho SY, Oslizlok P, Anderson RH. Architecture of the atrial musculature in and around the Triangle of Koch: its potential relevance to atrioventricular nodal reentry. J Cardiovasc Electrophysiol 1997;8:1396-1407.

14. Lamers WH, Wessels A, Verbeek FJ, et al 1992 New findings concerning ventricular septation in the human heart. Implications for maldevelopment. Circulation 86:1194-1205.