Chapter 17. Development of the Heart, with Particular Reference to the Cardiac Conduction Tissues

It is, perhaps, unfortunate that the myocytes that generate and dissipate the cardiac impulse throughout the myocardial body are known as the cardiac conduction tissues, because all myocytes within the heart possess the capacity to conduct. The alternative title, the "specialized tissues," is equally unfortunate, because from a developmental stance, as we will show, the so-called conducting tissues, in particular the nodal components, share important features with the myocytes of the primary heart tube. It is their so-called "working" partners that show more evidence of developmental specialization. In this chapter, therefore, we will describe the processes whereby the initial heart tube grows by addition of myocardium at both its venous and arterial pole, how it loops, and how eventually it becomes converted into the four-chambered organ. In this definitive structure, it is the sinus node that generates the cardiac impulse, this being propagated within the atrial musculature, delayed in the atrioventricular node, and then conducted rapidly to the ventricular myocardium through the atrioventricular bundle, its branches, and the ventricular ramifications known as Purkinje fibers. We will also discuss the fate of the more widespread areas of primary myocardium found in the developing heart because it is almost certainly the remnants of these areas in the postnatal heart that are the substrates for many cardiac arrhythmias. In contrast, with regard to perhaps the most important arrhythmia in the ageing population, namely atrial fibrillation, we will show how the pulmonary myocardial sleeves, known now to be the substrate for many forms of atrial fibrillation, are derived from so-called working rather than specialized conducting tissues.

The overall structure of the developing heart and the basic arrangement of the definitive heart are preserved across the vertebrates, from fish to man. When first seen during development, the heart is laid down as a linear myocardial tube, with venous and arterial poles (Fig. 17–1A). Conduction through this linear tube begins at the venous pole and proceeds toward the arterial pole. With ongoing development, cavities balloon from the tube, to which additional material is rapidly added at both poles. In the most basic plan, the myocardium that will form the atrial chambers balloons from the primary tube into the dorsal direction, whereas the myocardium of the developing ventricles balloons in ventral fashion (Fig. 17–1B). Concomitant with the ballooning of these pouches from the linear, or primary, tube, it becomes possible to record an electrocardiogram.1 At these early stages, it is not possible to recognize any "conduction tissues" when assessing the cardiac structure histologically. It is possible, however, to distinguish the characteristics of the myocytes in the different parts of the developing heart according to their molecular signatures. Thus, the myocytes of the primary heart tube do not express the connexin proteins responsible for the formation of fast-conducting gap junctions.2 As a consequence, at this early stage, the blood is propelled through the developing heart in a sluggish, peristaltic fashion. In contrast, the myocytes forming the walls of the developing atrial and ventricular chambers express the molecules that ensure rapid conduction and synchronous contraction, such as the gap-junctional proteins connexin 40 and 43.3 This myocardium making up the walls of the developing chambers is called working, or secondary, myocardium. It is possible to distinguish subtypes within the working myocardium.4 The myocytes making up the atrial working myocardium conduct faster and express the rapidly contracting atrial myosin isoform, whereas the myocytes within the developing ventricles express the slow myosin isoform,5 which is also expressed in slow-twitch skeletal muscle. Whereas myocytes within both the atrial and ventricular working myocardial walls express atrial natriuretic factor, the atrial myocytes that are added via the so-called dorsal mesocardium, the area that will eventually form the myocardial sleeves of the pulmonary veins, do not express this gene (Fig. 17–2).

We now know that, in the mammalian heart, the initial component of the linear heart tube forms little more than the embryonic left ventricle and atrioventricular canal. Recent evidence from the mouse heart shows that, from the stance of lineage, the ventricular septum is derived from this embryonic left ventricle, whereas the entire mural component of the left ventricle is derived from the atrioventricular canal.6 Additional material remains to be added to the developing heart at both the venous and arterial poles from the so-called second heart field, although it is debated whether this is derived from a separate field or rather from ongoing temporal migrations of cells from a single source.7,8 Regardless, the end result is that the heart tube elongates and bends concomitant with the addition of the new material at its venous and arterial poles.9 As it bends, the cavities of the atrial and ventricular chambers balloon from the linear tube,10 with the atrial appendages in the mammalian heart ballooning in parallel from the atrial component of the primary tube, but with the apical components of the ventricles ballooning in series from the ventricular part of the developing heart (Fig. 17–3A). The molecular signatures of the developing myocardial components are well seen in the mouse heart on the 12th day of development, albeit the apical trabecular component of the right ventricle is only just beginning to balloon at this stage (Fig. 17–3B,C). The evidence in the mouse heart, now known to be replicated during development of the human heart, shows that the areas of primary myocardium are extensive. It is from these areas that, with time, there will be formation of the components of the so-called conduction system as recognized in the definitive heart.11

As we will discuss, our studies on the developing heart, as yet unpublished, showed the location of the remnants of the initial primary myocardium, now providing insights into the location of arrhythmic substrates in the postnatal heart. The parts of the primary myocardium, including those of the sinus venosus persisting as the components of the so-called cardiac conduction tissues in the definitive postnatal heart, can be recognized morphologically based on the criteria established by Aschoff12 and Mönckeberg13 at the beginning of the twentieth century.12,13 The conduction tissues, according to these criteria, should be histologically distinguishable from surrounding myocardium, should be traceable from one section of the specimen to another, and, if forming tracts, should be insulated from the other myocardium by connective tissue. Thus, all different components of the cardiac conduction tissues in the definitive postnatal heart (Fig. 17–4) can be recognized morphologically using histologic stainings and the fulfillment of the criteria, established by Mönckeberg13 and Aschoff.12 The exemplar of such a conduction tract, as shown by Tawara,14 is the atrioventricular conduction axis (Fig. 17–5A). The sinus and atrioventricular nodes, in contrast, are not insulated from the atrial musculature, thus fulfilling only two of the three criteria (Fig. 17–5B,C). The morphologic features that permit recognition of the regular cardiac nodes also permit additional areas of atrial myocardium to be recognized as being histologically specialized. However, use of these same criteria shows that the pulmonary venous myocardial sleeves and the internodal atrial myocardium, at least in terms of gross histology, are made up of working myocytes.15

The sinus and atrioventricular nodes and the pulmonary myocardial sleeves are all components of the definitive atrial chambers.16 Knowledge of the development of these components, therefore, requires an appreciation of the morphologic stages involved in formation and separation of the morphologically right and left atrial chambers. It requires, in particular, an understanding of the formation of the atrial septum. Therefore, prior to considering the steps involved in formation of the cardiac nodes themselves, we will review our own understanding of the developmental heritage of the left and right atrial chamber myocardium.

The entire linear heart tube, including the venous pole and the part from which the atrial appendages will balloon, is made up of primary myocardium. The appendages, with walls made of chamber myocardium, balloon in parallel fashion from the primary myocardium. Subsequent to looping of the heart tube, they are positioned rightward and leftward relative to the developing outlet component of the heart (see Fig. 17–3B,C). At the stage at which the systemic venous tributaries enter the atrial component of the tube, which they do in relatively symmetrical fashion, there has been no formation of the lungs. However, it is important to appreciate that the atrial component of the tube remains connected dorsally to the mediastinal mesenchyme through the so-called dorsal mesocardium, where the pulmonary vein will develop (Fig. 17–6). With ongoing development, there is a shift in the relationship of the systemic venous tributaries so that, eventually, they open to the right side of the developing atrial component. In the mouse heart, anatomic boundaries of a discrete intrapericardial systemic venous sinus possessing myocardial walls do not become recognizable until after the tributaries have come to open to the prospective right atrium.4 In the human heart, in contrast, the venous sinus is recognizable morphologically, as a compartment within the pericardial cavity, albeit its walls initially are not myocardial. In both mouse and man, nonetheless, the left-sided venous tributary, which will become the coronary sinus in man, retains its own myocardial walls, passing inferior to the dorsal mesocardial connection as it extends to open in the right atrium (Fig. 17–7). Concomitant with this rightward shift of the atrial connections of the systemic venous tributaries, the dorsal mesocardium itself also undergoes significant remodeling.

When first seen, the internal atrial location of the mesocardial connection is marked by the so-called pulmonary ridges, which flank an imperforate pit in the dorsal wall of the undivided atrium.17 With further development, it is possible to trace from this pit an endocardial strand through the mediastinal mesenchyme to the ventral surface of the foregut, where lung buds will eventually develop. With ongoing development, this strand lumenizes to form the pulmonary vein (see Fig. 17–7), which enters the atrium through the dorsal mesocardial connection (Fig. 17–8). At later stages of development, the walls of the newly formed pulmonary vein become myocardial, and neither in mouse nor man do they have any direct connection with the myocardial walls clothing the tributaries of the systemic venous sinus.4,17

Concomitant with the rightward shift of the tributaries of the systemic venous sinus, septation of the atrial component of the developing heart commences. This is achieved initially by growth toward the atrioventricular canal of a crescentic muscular shelf, the primary atrial septum, or "septum primum." As it grows toward the atrioventricular canal, which undergoes division by formation of two atrioventricular endocardial cushions, there is a space between the leading edge of the developing septum, which carries a mesenchymal cap, and the atrial surfaces of the cushions. This space is the primary atrial foramen, or "foramen primum" (Fig. 17–9). The foramen is closed by union of the mesenchymal cap on the leading edge of the primary septum and the extracardiac dorsal mesenchyme, or vestibular spine, with the endocardial cushions.18 By the time of fusion, the upper edge of the septum has already broken down to form the secondary atrial foramen, or "foramen secundum" (see Fig. 17–9). As the primary septum fuses with the cushions to close the primary foramen, it completes the confinement of the pulmonary venous return to the left atrium. During this period, the right pulmonary ridge becomes very pronounced and now is recognizable as the vestibular spine, or "spina vestibuli,"19 also known as the dorsal mesenchymal protrusion20 (Fig. 17–10). The area occupied initially by the spine and the mesenchymal cap most probably becomes the extensive muscular anteroinferior rim of the oval foramen. This process of formation of the anteroinferior rim of the oval fossa, as we will see, is particularly important in forming the atrial septal connections of the atrioventricular node, albeit the mechanisms of formation have yet to be established.

During and after division of the atrial chambers and formation of the definitive anteroinferior border of the oval foramen, marked changes have also occurred in the location of the pulmonary veins. The initially solitary venous opening is first seen adjacent to the developing left atrioventricular junction (see Fig. 17–7). Eventually, four veins open to the atrial roof (Fig. 17–11).21 Only after the venous openings have become translocated to the atrial roof does it become possible to recognize the infolded superior atrial fold that eventually separates the opening of the superior caval vein to the right atrium from the right-sided pulmonary venoatrial junctions. This area, although usually known as the "septum secundum," is no more than a fold in the superior atrial roof, a feature made more obvious in the postnatal heart by the fat that accumulates within the groove.22

Concomitant with the rightward shift of the systemic venous tributaries and their incorporation into the right atrium, its floor, composed of primary myocardium, becomes an integral part of the morphologically right atrium. The boundary between the floor of the venous sinus and the appendage becomes particularly prominent and is recognizable in the formed heart as the terminal crest.16 During fetal life, the entirety of the boundary between the systemic venous sinus and the remainder of the right atrium is marked by extensive muscular folds known as the venous valves. These structures usually regress to greater or lesser extent in the postnatal heart, with parts of the right valve persisting as the Eustachian and Thebesian valves, guarding the orifices of the inferior caval vein and coronary sinus, respectively. The left venous valve usually disappears in its entirety in the human heart. By the time of birth, the musculature forming the floor of the systemic venous sinus has become indistinguishable in terms of gross histology from the remaining atrial musculature, but its initial location within the sinus is indicative of its primary myocardial heritage. Further myocardium of primary origin, but converted into working myocardium at birth, is also found within the vestibules of both atria just above the atrioventricular valves. This is the atrioventricular canal musculature, which is sequestrated within the atrial component concomitant with formation of the fibroadipose plane of atrioventricular insulation.23 Apart from this vestibular myocardium, the remainder of the left atrial musculature is derived either from chamber or dorsal mediastinal myocardium and has never possessed a primary myocardial phenotype.4,24 As we will discuss, these differences in developmental heritage are crucial in determining the potential anatomic substrates for atrial arrhythmias.

As we have emphasized already, it is possible to record an electrocardiogram from the developing embryo long before the final appearance of all components of the conduction system can be recognized histologically in the postnatal heart.1 This is because, again as already emphasized, all myocardial cells within the heart have the ability to conduct the cardiac impulse, albeit to a different extent.25 The ability to record an adult type of electrocardiogram from the embryonic heart without a definitive conduction system coincides with the initiation of the local ballooning and differentiation of the fast-conducting chambers from the slow-conducting primary heart tube.26 The rapid conduction between the myocytes making up the secondary myocardium of the chambers is due to their linkage by large and fast-conducting gap junctional complexes,27 which are absent from the slowly conducting primary myocardium.28 Subsequent to the ballooning of the atrial appendages and the growth of the ventricles, the remaining primary myocardium becomes recognizable as the outflow tract, atrioventricular canal, and the corridor of myocardium, which extends from the atrioventricular canal to incorporate the orifices of the systemic venous tributaries. Our knowledge of the subsequent changes to this primary myocardium, which for the most part becomes converted in terms of its histologic appearance to so-called working myocardium, has accrued from experiments carried out largely in the mouse. Thus, in the developing mouse heart, the primary myocardium that will not develop into chamber myocardium is distinguished by the expression of the gene encoding Tbx3, a T-box transcription factor, which represses the chamber program of gene expression (Fig. 17–12).29 By tracing the location of this transcription factor and related factors such as Tbx1824,30 in the myocardial walls of the systemic venous sinus, it has been possible, in the mouse heart, to chart with accuracy the location of those areas of primary myocardium that persist as the definitive postnatal conduction tissues. Recent expressional analyses in the developing human heart from our own laboratory, as yet unpublished, largely confirm the findings in the mouse, taking account of the morphologic differences in the arrangement of the postnatal conduction tissues between the two species.

The Sinus Node

From the outset of the formation of the cardiac tube, pacemaker activity always originates at the intake of the heart.31 Recent molecular experiments have revealed the molecular mechanism whereby this is achieved.30,32 With ongoing development, the systemic venous sinus and its tributaries, or left and right common cardinal veins, become incorporated into the pericardial cavity and acquire myocardial sleeves. The myocardium of the systemic venous tributaries is characterized by the expression of the T-box transcription factor Tbx18, by the expression of the important pacemaker channel Hcn4, and by the absence of expression of the cardiac transcription factor Nkx2-5 (Fig. 17–13).24,32 Under direction of the transcription factor Pitx2c, the pacemaker phenotype becomes restricted to the right sinuatrial junction and is recognizable as the sinus node (Fig. 17–14).32 The remaining areas become working myocardium when assessed both histologically and molecularly. Within 6 weeks of development in the human embryo, the primordium of the sinus node can be recognized in sections processed in standard histologic fashion, which coincides with positivity for the pacemaker channel Hcn4 (Fig. 17–15). If the findings in the mouse are, indeed, paralleled in the human heart, then the location of the primary myocardium around the mouth of the coronary sinus and along the terminal crest offers some explanation for the origins of abnormal electrical activity in patients with atrial arrhythmias.

The Atrioventricular Node

As with the sinus node, most of our knowledge concerning the genetic regulation of the development of the atrioventricular node has emerged over the past decade subsequent to multiple investigations of the mouse heart. There are known important differences in the function of the atrioventricular node between mouse and man, most probably related to the size of the heart.33 There are also significant differences in the architectural arrangement of the myocytes making up the compact node and its transitional zones in human34 as compared with the mouse.35,36 There are also differences in the relationship of the atrioventricular muscular sandwich relative to the aortic root. Because of the marked off setting of the leaflets of the mitral and tricuspid valves in man, the compact node is a half-oval set against the central fibrous body, with transitional cells forming a marked overlay not seen in the mouse. The penetrating atrioventricular bundle is also appreciably shorter in man compared with mouse. It is important, therefore, to remember interspecies diversity when considering the developmental aspects.

In the mouse, recent lineage studies have shown that the compact part of the node is derived from the atrioventricular canal musculature, with its primary phenotype.6 This part of the atrioventricular node, therefore, as with the sinus node, is distinguished by its content of Tbx3.29 During further development, it acquires a highly specific transcriptional profile that is distinct from both the working myocardium and the original primary myocardium.37 The atrioventricular canal musculature can also be recognized histologically in the human heart at 6 weeks of development (Fig. 17–16). Initially, after the primary foramen has been closed, the mesenchymal tissues that eventually form the central fibrous body interpose between the atrioventricular canal musculature, which will form the compact part of the atrioventricular node and its inferior extensions, and the musculature of the atrial septum (Fig. 17–17). The connection of the definitive node, via transitional cells, with the musculature of the atrial septum requires subsequent muscularization of the tissues contributed via the vestibular spine and the mesenchymal cap. As discussed earlier, an extensive array of transitional cells interpose between the working atrial myocytes and the cells of the compact atrioventricular node in man.34 It remains to be established with certainty whether it is these transitional cells or the nodal inferior extensions that form the clinically important slow and fast pathways for atrioventricular nodal conduction. In terms of development, nonetheless, the slow pathway represents the continuation of the atrioventricular canal musculature into the vestibule of the tricuspid valve. The larger part of this ring, in histologic terms, becomes transformed into working myocardium, but its origin as primary myocardium probably accounts for its role as the slow nodal pathway.38 The fast pathway, in contrast, extends from the anterosuperior rim of the atrial septum to make contact with the atrioventricular nodal myocytes. This connection cannot be formed developmentally until after the muscularization of the vestibular spine and the mesenchymal cap. The precise origin of this pathway remains to be established.

The Nature of the Interatrial Myocardium

The criteria for histologic recognition of conducting tracts were established to defuse a suggestion that "specialized" pathways exist to preferentially conduct the sinus impulse through the atrial myocardium into the atrioventricular node. These sensible criteria were ignored by investigators who subsequently suggested that such "specialized" tracts did exist within the atrial musculature. In terms of gross histology, there are no insulated tracts extending between the cardiac nodes.39 The preferential conduction that does exist within the muscular pathways is the consequence of the packing of the individual atrial myocytes.40 In developmental terms, there is a corridor of primary myocardium extending from the sites of formation of the sinus and atrioventricular nodes.4 In the formed heart, the terminal crest is the boundary between the floor of the atrial component of the primary heart tube and the right atrial appendage. Remnants of this primary myocardium could serve as substrates for abnormal automaticity in the right atrium. In terms of gross histology in the postnatal heart, however, these areas are indistinguishable from the remainder of the working atrial myocardium.39

The Atrioventricular Conduction Axis

It was the monumental monograph of Sunao Tawara,14 now also available in an excellent English translation,41 that clarified the arrangement of the atrioventricular conduction axis. He showed that the atrioventricular node was but the commencement of a tract of histologically distinct myocytes that could be traced eventually into the ventricular Purkinje cells. The proximal ventricular part of this axis is directly continuous with the compact atrioventricular node, which, in turn, is part of a larger developmental entity, recognized in the human heart by the expression of the GlN2 antigen42 and that can be traced round the vestibule of the tricuspid valve and the aortic root, eventually returning to the atrioventricular axis itself as a "dead-end tract." The entirety of the axis is initially part of a ring of primary myocardium that surrounds the embryonic interventricular foramen (Fig. 17–18). In the mouse heart, when taking the border between node and penetrating bundle as the site at which the axis becomes insulated by the fibrous tissue of the central fibrous body,14 a tongue of the ventricular component can be traced as the inferior part of the compact atrioventricular node. Such a dual origin of the proximal atrioventricular conduction axis is also supported circumstantially in human hearts by evidence from the arrangements seen in patients with rare forms of congenitally complete atrioventricular dissociation.43 Whether the arrangement seen in the mouse also exists in the developing human heart remains to be established.

The Ventricular Conducting Tissues

In the normal situation, it is the atrioventricular conduction axis that provides the only muscular continuity between the atrial and ventricular muscles masses. There have also been discussions as to whether the myocardial cells making up this axis, including the so-called Purkinje myocytes, are derived by recruitment of working myocardial cells or differentiate directly from primary myocardium.44,45 Lineage studies, as yet unpublished, suggest that the myocytes making up the proximal part of the ventricular component of the axis are derived from the initial primary myocardium of the heart tube. The primary myocardium, however, is slowly conducting. The purpose of the ventricular bundle branches and their peripheral ramifications, in contrast, is to ensure rapid activation of the working ventricular myocardium, with the activation commencing at the ventricular apexes so as to ensure systolic ejection of the blood into the arterial trunks. The bundle branches and the ventricular Purkinje myocytes, therefore, are derived from the initial trabeculated component of the ventricular walls,46 with their myocytes joined together by rapidly conducting connexins.

The Atrioventricular Ring Tissues

Another debate relating to cardiac conduction concerned the presence of recognizable node-like structures in the right atrial vestibular myocardium. These entities were recognized by Kent47 at the turn of the nineteenth century, albeit erroneously interpreted by him as providing multiple muscular connections across the atrioventricular junctions of the normal heart. The structures are the remnants of the ring of primary atrioventricular canal myocardium that forms the vestibule of the tricuspid valve (see Fig. 17–18). The T-box transcription factor Tbx3, which represses the chamber program of gene expression, is expressed in the entirety of the initial ring (see Fig. 17–12). In very rare circumstances, the remnants can persist as histologically recognizable structures that function as substrates for ventricular preexcitation in otherwise normally structured hearts.48 They can also form anomalous atrioventricular nodes in the setting of congenital malformations such as congenitally corrected transposition and double inlet left ventricle.49,50 In the developing heart, the ring of atrioventricular canal musculature crosses over the atrioventricular conduction axis, forming a mass of histologically recognizable tissue in the atrial musculature directly dorsal to the aortic root (see Fig. 17–18B). This so-called retroaortic node is separated from the atrioventricular node itself by the insulating tissues of the central fibrous body. It does not become an integral part of the normal atrioventricular node.

The Atrioventricular Insulating Planes

In the postnatal heart, it is only the atrioventricular conduction axis that provides a muscular pathway between the atrial and ventricular myocardial masses. As we have discussed, it is possible to record an adult-type electrocardiogram long before specialized conducting tracts can be recognized.1 At these initial stages, muscular continuity is found throughout both atrioventricular junctions, with the muscular continuity being provided by the primary atrioventricular canal myocardium (see Fig. 17–9). Recent experiments using lineage data have shown that the atrioventricular canal region contributes significant material to the left ventricular inlet component.6 Part of the initial canal musculature, nonetheless, becomes sequestered in the atrial tissues subsequent to the formation of the fibrous insulating planes.23 The separating planes themselves are often considered part of an extensive fibrous skeleton. In reality, this is far from the case. In the right atrioventricular junction of the postnatal human heart, it is rare to find a well-formed fibrous annulus supporting the hinges of the leaflets of the tricuspid valve. Instead, it is the fibroadipose tissue of the atrioventricular groove that separates the atrial from the ventricular myocytes. On the left side, the fibrous annulus is better formed, supporting the mural leaflet of the mitral valve. In relation to the anterior, or aortic, leaflet of the mitral valve, it is initially possible to see the musculature of the inner heart curve. With ongoing development, this muscle disappears, leaving aortic-to-mitral fibrous continuity. The precise timing of the disruption of the extensive areas of initial myocardial continuity around the atrioventricular junctions has still to be established for the human heart, as does the mechanism permitting retention of those muscular bridges51 known to be the substrates for ventricular preexcitation.52

The Pulmonary Venous Myocardial Sleeves

With the recognition that atrial fibrillation in many patients can be cured by creating electrical discontinuity between the pulmonary venous myocardial sleeves and the remainder of the left atrial musculature,53 suggestions have been made that this myocardium is itself histologically specialized. Some have argued that the myocytes making up the sleeves are "Purkinje-like" and hence histologically specialized.54 Rigorous examination of the sleeves in the human heart shows this not to be the case.55 The pulmonary sleeves are made up of aggregated working myocytes, and there is no evidence that the sleeves are histologically distinct or specialized. An alternative explanation for the "specialization" is that the myocardium itself is initially derived from the systemic venous sinus.56 There is no evidence to support this notion either, with a wealth of evidence showing that the myocytes within the sleeves have never had a primary myocardial heritage.57

The Myocardium of the Outflow Tracts

In contrast to the pulmonary venous myocardial sleeves, which have no relationship to the embryonic primary myocardium, the myocardium making up the ventricular outflow tracts had the primary myocardial phenotype during the embryonic phase of development. Much interest has been focused on this myocardium as the substrate for ventricular tachycardias, the more so because some such tachycardias can be cured by ablation, placing the lesions distal to the anatomic ventriculoarterial junction.58 There is no question that the entirety of the myocardium of the common ventricular outflow tract is initially of the primary variant or that this myocardium initially extends to the margins of the pericardial cavity (Fig. 17–19A).59 During development, the distal part of the outflow tract rapidly changes to an arterial phenotype,60 albeit the mechanisms of the transformation have yet to be established. Subsequent to the formation of the intrapericardial arterial trunks, a sleeve of primary myocardium encloses the entirety of the developing aortic and pulmonary roots (Fig. 17–19B).60 Eventually this myocardium also regresses in a proximal direction, but even in the postnatal heart, it is possible to recognize crescents of ventricular myocardium at the bases of all the pulmonary valvar sinuses and in two of the aortic valvar sinuses.61 These crescentic areas were initially made up of primary myocardium. In the postnatal heart, they cannot be distinguished histologically from the remainder of the ventricular myocardium. Parts of the muscular sleeve could persist on the epicardial aspect of the arterial valves, even extending into the walls of the arterial trunks. The dead-end tract of the primary interventricular ring also persists on the crest of the ventricular septum in the aortic root.62 It remains to be established which, if any, of these muscular structures are the substrates for ventricular tachycardias.

Much has been learned over the last two decades concerning the development of the heart and the formation of the specialized conducting tissues. The majority of this evidence has come from the study of the murine heart. We have sought to summarize this evidence in our chapter, showing its relevance to the understanding of normal cardiac conduction and the understanding of clinical cardiac arrhythmias. At present, the majority of the evidence concerning development is derived from study of the mouse heart. Furthermore, as yet, it has not been possible to achieve consensus as to how the time-honored criteria of Aschoff and Mönckeberg can best be modified to take note of the huge advances made in the understanding of the molecular signatures of the cardiac specialized tissues.15 Within the next few years, we should be able to extend these investigations to the developing human heart. In particular, comparable advances in imaging should permit studies to be made of the developing human heart, ideally correlating these findings with electrocardiographic changes. Comparisons of these changes in the development of the normally structured heart, for example, with the morphology seen in the setting of common atrioventricular junction63 should permit clarification of the roles played by the atrial septal inputs to the atrioventricular node. We anticipate that these future investigations will confirm the hypotheses advanced in this chapter.