Search

Various pathologic conditions producing atrioventricular block are reviewed. Calcific atrioventricular block (aortic stenosis, mitral valve annular calcium), active infective endocarditis ("ring" abscess), acute and healed myocarditis, various collagen-vascular diseases, amyloidosis, and tumors are reviewed as causes of atrioventricular block. Various diseases producing bundle-branch block are also discussed.

Twenty-three chicken hearts were used to study the cardiac conduction system by light and electron microscopy. In addition to a sinus node, atrioventricular node (AVN), His bundle, left and right bundle branches (LBB, RBB), the chicken also has an AV Purkinje ring and a special middle bundle branch (MBB). The sinus node lies near the base of the lower portion of the right sinoatrial valve. The AV node is just above the tricuspid valve and anterior to the coronary sinus. The His bundle descends from the anterior and inferior margin of the AV node into the interventricular septum, then dividing into right, left and middle branches some distance below the septal crest. The middle bundle branch turns posteriorly toward the root of the aorta. The AV Purkinje ring originates from the proximal AV node and then encircles the right AV orifice, joining the MBB to form a figure-of-eight loop. The chicken conduction system contains four types of myocytes: 1) The P cell is small and rounded, with a relatively large nucleus and sparse myofibrils. 2) The transitional cell is slender and full of myofibrils. 3) The Purkinje-like cell resembles the typical Purkinje cell, but is smaller and darker. 4) The Purkinje cell is found in the His bundle, its branches, and the periarterial and subendocardial Purkinje network.

The cardiac conduction system may be affected by various pathologic changes. Some of these changes are general and affect the nonconduction system tissues of the heart (atrophy, fibrosis, necrosis, segment deposition, mineral) and others are more localized to the conduction system (Lev and Lenegre disease). Pathologic conditions associated with atrial, junctional, and ventricular arrhythmias are reviewed.

Normal anatomic and histologic features of the atrioventricular junction (transitional cell zone, atrioventricular node, penetrating portion of bundle) and the bifurcation of the penetrating portion into bundle branches are reviewed. Terminal ventricular Purkinje fibers are also discussed.

Normal anatomic and histologic features of the sinus node, atrial myocardium, and interatrial conduction of the cardiac impulse are reviewed. The controversy surrounding atrial conduction via specialized atrial cells versus specific internodal tracts (between sinus and atrioventricular nodes) is discussed.

On 12th randomized specimens of the human heart, of either sex, from 19-84 years of age, I investigated the vascularization of the conducting system of the heart in relation to the type of coronary ramification, by the dissection and injection-corrosive method. The symmetrical type of arterial vascularization was established in 47.5%, the right type in 45% and the left type in 7.5% of the cases investigated. Most often the right anterior sinus nodal artery was in the symmetrical type and the left anterior sinus nodal artery in the right type of arterial supply. The right atrioventricular nodal artery was in 90% of the hearts in the right and symmetrical type of the arterial vascularization.

Serial hand mapping techniques in man have identified 3 to 5 sites of epicardial breaktrough (EBT). However, transmural epicardial excitation from the widely distributed His/Purkinje system suggests a more complicated pattern may exist. Multielectrode arrays used with large mapping systems during surgery often present complicated and sometimes inconsistent activation patterns. The purpose of this work is to reconcile epicardial activation in the normal human heart with anatomical and endocardial/intramural physiological recordings using multichannel computer mapping requiring only a single beat, and rigorously defined and applied activation time detection algorithms. Eighteen subjects undergoing surgery for Wolff-Parkinson-White syndrome were recorded with a 119 site sock array during nonpreexcited sinus rhythm. None had evidence of coronary artery disease and all exhibited a normal 12-lead ECG except during periods of preexcitation or tachycardia. Each was recorded bipolarly and four also were recorded monopolarly. Recordings revealed 8.0 +/- 1.6 EBTs (range 5 to 12). Closely spaced, multiple EBTs often were observed and usually confirmed using different activation time detection algorithms. The earliest EBT always occurred over the anterior right ventricle at 14.3 +/- 6.5 msec (range -1 to 29 msec) after QRS onset. Subsequent EBTs could occur at any ventricular site with variable latencies. In contrast to previous reports describing epicardial spread of activation from a few foci, a mosaic of epicardial activation emerges. These data are consistent with endocardially initiated transmural activation of the epicardium suggested by the anatomy of the His/Purkinje system and intramural recordings.

Previous descriptions of the four anatomic regions of dissection in Wolff-Parkinson-White syndrome have been largely qualitative. In this study quantitative data describing this anatomy are presented, together with statistical analysis of selected anatomic relationships. Fourteen human hearts were dissected. The borders of the posteroseptal dissection along the mitral anulus, tricuspid anulus, and epicardium were measured. A positive correlation between the mitral and tricuspid annular dimensions was found (r = 0.55; p = 0.04); however, the length of epicardial dissection was more variable. The dimensions of the anteroseptal space and the position of the right coronary artery within this space were measured. These measurements emphasize the proximity of the aortic sinuses of Valsalva to the right atrial endocardium near the posteromedial extent of the dissection. The dimensions of the right and left free walls and the position of the coronary arteries within these spaces were measured. A positive correlation was found between the width of the atrioventricular groove and the length of the dissection necessary to separate the atrioventricular groove fat from the ventricular surface (r = 0.89; p = 0.0001, right free wall; r = 0.87; p = 0.0001, left free wall). The quantitative data presented in this article are intended to enhance the surgeon's appreciation of the anatomy relevant to the treatment of Wolff-Parkinson-White syndrome. The knowledge gained from this quantitative analysis may improve accuracy in the electrophysiologic localization and surgical disruption of accessory atrioventricular connections.

The structure of cholinergic and adrenergic neuromuscular junctions in the sinus venosus of the toad, Bufo marinus, was determined by electron microscopy. From random sections of sinus venosus tissue it appeared that there were variable separations between cholinergic or adrenergic varicosities and the nearest sinus venosus muscle cell. However, when the structure of complete cholinergic and adrenergic varicosities was determined by examining serial electron micrographs, virtually all varicosities that lost their covering of Schwann cell were found to form an area of close apposition with an adjacent muscle cell. At the region of close apposition, the neuromuscular cleft was filled with a single layer of basal lamina to give a neuromuscular separation of about 70 nm. Synaptic vesicles within a varicosity were usually found to be concentrated towards the region of close apposition. These observations are discussed in relationship to the idea that when transmission occurs at these neuromuscular junctions the transmitters act on discrete pools of specialized subsynaptic receptors.

A realistic computer model of propagation of ventricular activation was used to study the effects of varying the position of specific conduction system terminations in the left ventricle and the septum, representing the sites of initial activation, on the resulting simulated spatial heart vectors. Three differently localized foci of initial activation, each of them represented by one model element, were considered: in the central part of the left septal surface, posteriorly at about one third of the distance from the apex to the base, and in the upper part of the anterior free wall. During the model experiments, the positions of the initial activation were shifted +/- 5 model units (ca 5 mm) in the vertical and lateral direction either separately or in different mutual combinations. Small variations of the initial activation site in the basal parts of the left ventricle led to significantly smaller changes of the vectorcardiographic loop than variations of the same extent with the initial activation site located more apically.

The role of G proteins in mediating the responses of the heart to circulating catecholamines and to the influences of the autonomic nervous system is of special interest to cardiologists. It is evident that G proteins are essential links in the cascade of biochemical events that ensure when neurotransmitters and hormones interact with receptors on myocardial cells. It is likely [corrected] that dysfunction of G proteins plays a role in cardiovascular pathophysiology. With current methodologies, especially molecular biological and recombinant DNA techniques, and with transgenic animal models that can relate physiological function and specific gene dosage, some cardiovascular diseases may be traced to G protein-related defects.

The goal of noninvasively imaging the cardiac electrical generator is considerably complicated by the conductive anistropy of the cardiac muscle which, as featured in the oblique dipole layer model of ventricular depolarization, introduces muscle fiber geometry and a conductivity tensor as additional unknowns. The ventricular surface activation map has been previously presented as a description of the cardiac generator in image form, but the integral equation defining its relationship to the body surface potentials is valid only under the less accurate uniform dipole layer hypothesis. Using an argument from differential geometry, which allows the integral equation approach to be bypassed, we show that the critical points of this map can still be localized on the heart surface from the body surface potentials in a manner fully consistent with the oblique dipole layer model. Thus, in principle, a realistic and useful "image-like" output is possible in a limited way even without explicit information regarding conductivity anisotropy. The realization of this output will require improvements in the temporal resolution presently available from existing body surface potential mapping systems.

This is a qualitative and quantitative study of dog bundle branch and Purkinje cell development from day 0 to week 12 of life; we correlate the morphologic data with changes observed in the functional properties of developing dog Purkinje tissue. The bundle branch itself has a roughly cylindrical shape and is surrounded by a collagen sheath covered with endocardium. Within the bundle, Purkinje cells are packed closely together in fascicles distributed evenly around a central artery. Cross-sectional area doubles in the right bundle and increases 5-fold in the left bundle system between day 0 and week 12 of life. About one third of the bundle by volume is Purkinje tissue; the rest is extracellular space containing an increasing amount of collagen as the animal ages. Purkinje cell cross-sectional area is constant during the first week of life, but its length doubles and the cell changes from a relatively round to a more cylindrical shape. Between day 7 and week 12, cell diameter doubles; Purkinje cell surface area increases 5-fold and its volume almost 10-fold. As a consequence, the surface to volume ratio halves and approaches the value reported for adult dogs by week 12 of life. The percent of the intercalated disc occupied by nexal junctions virtually doubles by week 12, the same period over which Purkinje fiber conduction velocity increases. The disc itself becomes less dominant as the cell enlarges; the total percent of sarcolemma involved in its formation decreases by a fourth and has achieved the adult value by week 12 of life. As this happens, the percent of cell membrane facing on clefts increases almost 6-fold, so that the total percent of sarcolemma facing on small spaces (approximately 340A wide) is constant over the age period studied. The paucity of clefts in newborn tissue compared with the value reported for the adult dog may help explain the relative lack of responsiveness to extracellular potassium concentration of the resting membrane potential described for fetal Purkinje tissue. Within the Purkinje cell itself, the percent by volume occupied by mitochondria remains relatively constant over the age span studied, while sarcomeric mass increases 3-fold over the same period of time; these data are consonant with the relative resistance of this tissue to hypoxia.

Accurate anatomic localization of accessory pathways during preoperative electrophysiologic study and during operative mapping depends on a knowledge of the dimensions of the posterior septal space and the left free wall. These dimensions were therefore studied in 48 human cadaver hearts. Mean distance from the coronary sinus orifice to the left margin of the posterior septal space was 2.3 +/- 0.4 cm and mean length of the left free wall was 5.0 +/- 1.0 cm. The posterior septal space at the level of the valve anuli extended a mean of 3.4 +/- 0.5 cm around the epicardium. The width of the posterior septum measured in the coronary sinus was related to heart weight and a combination of body weight and patient age (p less than 0.05). The probability of an accessory pathway being located in the left free wall or the posterior septum during catheter mapping was calculated for various distances from the coronary sinus orifice for adults of different ages and body weights. In adults, accessory pathways located in the proximal 1.5 cm of the coronary sinus are almost always in the posterior septum. Those located between 1.5 and 3 cm from the coronary sinus orifice may be in either the left free wall or the posterior septum, and those located greater than 3 cm from the coronary sinus orifice are almost invariably in the left free wall.

Although the pig has been used as an experimental model for ischemic heart disease and sudden death, relatively little is known about the anatomy of the conduction system (CS) of this animal. We attempted to correlate electrophysiologic and anatomic differences between the pig and human CS. Invasive electrophysiologic studies were performed in five healthy anesthetized pigs. In contrast to the adult human, the pig has sinus tachycardia, shortened PR and H-V intervals, and a relatively short sinoatrial conduction time. Compared with the human CS, serial sections of the CS of pig hearts showed the following differences: (1) the atrioventricular node is located more to the right of the summit of the ventricular septum; (2) the penetrating bundle is very short, and the bifurcation of the bundle into bundle branches occurs more proximally; (3) there is more connective tissue and less elastic tissue; and (4) there is a copious amount of nerve fibers (about 50 percent throughout the CS). The presence of the abundant neural tissue implies that there is an important neurogenic component to conduction in the pig. Because of the above differences from the human, the pig should be used with caution as an experimental model in ischemic heart disease and sudden death where arrhythmias are studied.

A three-dimensional display is described that allows activation sequences from the epicardium and endocardium to be shown simultaneously on the same image. Three electrode arrays (epicardial sock, left ventricular balloon, right ventricular balloon) are represented in a three-dimensional perspective by an array of dots that are intensified when activated. This arrangement requires fewer calculations and is easier to interpret than sliced-isochronal maps but cannot represent a complete heart cycle in one image. The three-dimensional display eliminates the distortion caused by two-dimensional diagrams and facilitates activation correlation between electrode arrays. A standard, low cost microcomputer has been used to implement the activation display.

Since the last revision of the American Heart Association's guidelines in 1985, several new developments of clinical importance have occurred in the field of cardiopulmonary resuscitation. These include enhanced access to and earlier use of defibrillation, the use of high-dose epinephrine when standard doses fail, the assessment of resuscitative efforts with end-tidal CO2 monitoring and the addition of two new drugs, amiodarone (for refractory ventricular fibrillation) and adenosine (for paroxysmal supraventricular tachycardia). Time will determine the ultimate role of these advancements in the management of cardiac arrest.

Specific features of surgical anatomy of the atrioventricular conduction system (AVCS) in true and false aneurysms of the membranous part of the ventricular septum were studied on 126 heart specimens. The authors succeeded in localisation and dissection of the CS with the aid of an original method of anatomical microdissection which they prefer to the histologic method in studying the surgical anatomy of CS. The advantage of the authors' method consists in an objective anatomical identification of CS while preserving the integrity of adjacent structures.

The ventricular conduction system is a complicated network of specialized muscle cells responsible for the transmission of electrical activity between the atria and the ventricles of the human heart. It has been the focus of numerous electrical and anatomical studies at both the microscopic and macroscopic levels. An understanding of its behavior at both levels is considered important, because it is primarily responsible for the spread of excitation in the ventricles. Previous computer models have been very simple ones that have been primarily adjuncts to models of the ventricles. This paper describes a strategy for the construction of conduction system models which is based on real microscopic and macroscopic features, although the model still is much simpler than reality. The model contains almost 35,000 individual cylindrical elements, each of whose physical dimensions approximate unit bundles of Purkinje and atrioventricular nodal cells. The model, whose physical appearance closely resembles that of the conduction system, was generated from limited anatomical data in less than 2 min CPU time on an IBM 3090 at the Cornell National Supercomputer Facility.

Ten patients with chronic atrial flutter were studied prospectively using electrophysiologic mapping and pacing techniques to assess the mechanism of atrial flutter and the presence of an area of slow conduction in the atria. Electrograms recorded from greater than or equal to 30 right atrial sites for each patient during atrial flutter demonstrated that right atrial free wall activation was craniocaudal and that the interatrial septum activation was caudocranial, consistent with a reentrant circuit involving the right atrium. In six patients, slow conduction occurred during atrial flutter in the inferior right atrium and was spatially associated with fractionated electrographic recordings. In the other four patients, a "missing" interval of electrical activity occurred in the inferior right atrium for an average of 40% of the atrial flutter cycle. Transient entrainment criteria were demonstrated in each patient during rapid high right atrial pacing. The mean activation time from the high right atrial pacing site to the coronary sinus (inferior left atrial) recording site was long (228 ms) and consistent with activation through an area of slow conduction. During rapid pacing of atrial flutter from the coronary sinus site, no transient entrainment criteria could be demonstrated. The mean activation time from the coronary sinus pacing site to the high right atrial recording site was relatively short (134 ms) and consistent with orthodromic activation of the high right atrium not through an area of slow conduction. High right atrial pacing during sinus rhythm at rates similar to atrial flutter demonstrated a short activation time to the coronary sinus and low right atrial sites (mean 169 and 88 ms, respectively), indicating activation that did not traverse an area of slow conduction. Coronary sinus pacing during sinus rhythm demonstrated the same phenomena. Low right atrial electrograms recorded during sinus rhythm and during rapid pacing of sinus rhythm were not fractionated, although they were during atrial flutter. Thus, atrial mapping and pacing data were complementary, indicating that human atrial flutter in the patients studied was generated by a reentrant circuit in the right atrium, with an area of slow conduction in the low right atrium present only during atrial flutter.

The intrinsic cardiac ganglia of the neonatal rat heart in situ were studied using electrophysiological and histochemical techniques. The vagal branches innervating the atrial myocardium and cardiac ganglia were identified and individual ganglion cells visualized using Hoffman modulation contrast optics. Histochemical studies revealed the presence of acetylcholinesterase activity associated with neuronal cell bodies and fibers, catecholamine-containing, small intensely fluorescent cells, and cell bodies and nerve fibers immunoreactive for vasoactive intestinal polypeptide. Intracellular recordings from the "principal" cells of the rat cardiac ganglion in situ revealed a fast excitatory postsynaptic potential (EPSP) evoked after electrical stimulation of the vagus nerve, which was inhibited by the nicotinic receptor antagonist, mecamylamine. No spontaneously firing neurons were found, although infrequent (less than 2 min-1) spontaneous miniature EPSPs were observed in most neurons. The quantal content of neurally evoked responses was between 10 and 30 quanta, and the presence of multiple EPSPs in some cells suggested polyneuronal innervation. The neurally evoked EPSP amplitude was dependent on the rate of nerve stimulation, decreasing with increasing frequency of stimulation. Neurons exhibited a sustained depolarization during high frequency stimulation (greater than 1 Hz), and in approximately 15% of the cells a slow depolarization lasting 1-3 min was observed after a train of stimuli. The presence of catecholamine- and neuropeptide-containing neuronal cell body fibers in neonatal rat cardiac ganglia in situ, along with neurally evoked postsynaptic responses resistant to cholinergic ganglionic blockers, suggests a role for noncholinergic transmission in the regulation of the mammalian heart beat.

The natural variability in physiological form and function is herein related to the geometric concept of a fractal. The average dimensions of the branches in the tracheobronchial tree, long thought to be exponential, are shown to be an inverse power law of the generation number modulated by a harmonic variation. A similar functional form is found for the power spectrum of the QRS-complex of the healthy human heart. These results follow from the assumption that the bronchial tree and the cardiac conduction system are fractal forms. The fractal concept provides a mechanism for the morphogenesis of complex structures which are more stable than those generated by classical scaling (i.e., they are more error tolerant).

The atrioventricular node artery is a major contributor to the arterial supply of the atrioventricular conducting pathway and is an important vessel in the pathogenesis of heart block. The terminal ramifications of this artery were studied in detail by serial sectioning techniques in 50 human infant hearts. The artery provided branches to the posterior interventricular septum in all hearts (100%) and to the interatrial septum in 22 hearts (44%). The vessel supplied the atrioventricular node in 45 hearts (90%) but it terminated before reaching the node in five (10%). It supplied the penetrating bundle in 32 hearts (64%). Alternative sources of arterial supply to the atrioventricular conducting pathway include: the first septal branch of the left anterior descending coronary artery; the descending septal artery; and anterior atrial branches. Although the arterial supply to this region of the heart is variable, it is possible to make hypothetical predictions of conducting tissue involvement in myocardial infarction of various types.

We investigated by means of coronarography and dissection the arterial vascularization of the human heart in relation to the type of coronary ramification in 200 human hearts taken at random from males and females between 20 and 85 years of age. Examination of the selective coronarography findings was subsequently performed by means of careful dissection of the injected arteries and their branches in order to compare the results obtained with coronarographic findings in living human beings.

The distribution of sympathetic nerve fibers in the regions of the bundle branches of bovine and rat hearts was examined by the glyoxylic acid-induced method for histofluorescence demonstration of catecholamines and by acetylcholinesterase (AChE) histochemistry. The studies on the bovine heart were concentrated on the nerve fascicles, the arteries and the ganglionic cells, and showed that (1) the AChE-positive nerve fascicles that occur just outside the bundle branches show a positive catecholamine-fluorescence reaction to varying degrees, the paraarterial nerve fascicles showing a uniform reaction; (2) the AChE-positive nerve fascicles within the bundle branches contain a few, mainly varicose, sympathetic nerve fibers; (3) extensive plexuses of sympathetic nerve fibers supply the arterial branches; and (4) sympathetic nerve fibers occur close to some of the ganglionic cells. The pattern of distribution of sympathetic nerve fibers in the region of the SA node was found to be essentially the same. The studies on the rat heart showed that (1) the septal arteries that occur in the proximity of the bundle branches are accompanied by the sympathetic component of innervation; and (2) there is a substantial number of varicose sympathetic nerve fibers within the bundle branches. These observations show that there is a pronounced sympathetic innervation in bundle branch regions and suggest that the paraarterial route is the most important for sympathetic nerve fibers to reach these regions.

Although mice infected with Trypanosoma cruzi develop a wide variety of electrocardiographic (ECG) alterations, the typical isolated right bundle branch block or its association with the left anterior hemiblock patterns are not found in this model. This has been explained as related to topographic differences in the anatomy of the murine conducting system. However, there is no conclusive evidence that the murine conducting system differs from the human system. In this study, the anatomy of the murine conducting system is described, as well as its involvement in the chronic stages of experimental infection. 24 three-month-old C3H mice were infected with 50 bloodstream forms of T. cruzi, Tulahuén strain. Animals were killed after 3, 8 and 12 months. Whole frontal sections of the heart, including the conducting system, were serially studied. The sinoatrial node was located in the right atrial appendage, or in the junction between the superior vena cava and the right atrium, or "riding" on the interatrial septum. The atrioventricular (A-V) node and the His bundle showed a similar anatomic course to that in man. Therefore, there was no important anatomical difference that might have explained the lack of the ECG patterns observed in human chagasic myocardiopathy. The inflammatory involvement and the lesions of the conducting system were diverse and rarely severe. No significant difference was observed in animals killed at different times. The lesions in the working myocardium were similar to those observed in humans (chronic inflammatory infiltrates). Nevertheless, the topography of lesions was different: there was a selective involvement in the neighbourhood of the A-V groove.(ABSTRACT TRUNCATED AT 250 WORDS)

The morphology of the human atrioventricular node, atrioventricular bundle and bundle branches is described. A block of tissue bounded by the ostium of the coronary sinus, the pars membranacea, the septal leaflet of the tricuspid valve and the atrial and ventricular septa is removed. The block is then sectioned serially from the right endocardial surface in the frontal plane of the heart. Sectioning in this way produces fewer sections than from techniques previously described. Outlines of the atrioventricular node, atrioventricular bundle and proximal bundle branches are digitally registered and stored in a computer. Three dimensional reconstructions of the structures are then generated by computer and displayed on an oscilloscope so that the entire three dimensional image can be rotated in any plane. Stereoscopic image pairs are produced to assist perception of the shape of the atrioventricular node, bundle and branching patterns of the bundles. This technique is unique in that it describes a method from which a relatively small number of histologic sections are generated permitting not only a complete histologic examination, but also a study of the morphology of the area.

On the basis of our collective experience we have reviewed the disposition of the cardiac conduction tissues as they might be observed by the surgeon in both normal and abnormal hearts. The sinus node lies subepicardially in the terminal sulcus; because of its variable blood supply the entire superior cavoatrial junction is a potential danger area. There are no morphologically discrete tracts extending through the atrial tissues between sinus and atrioventricular nodes. The atrioventricular node, the atrial extent of the atrioventricular conduction axis, is contained exclusively within the triangle of Koch. The axis penetrates through the central fibrous body and branches on the muscular ventricular septum immediately beneath the interventricular component of the membranous septum. The landmarks to these structures are described as they might be seen through the right atrium, left atrium, and aorta. Consideration is then given to the surgical anatomy of the abnormal muscular atrioventricular connections that underscore the ventricular pre-excitation syndromes. Finally, rules are developed whereby the disposition of the conduction tissues can be predicted with accuracy in congenitally malformed hearts, in the settings of both normal and abnormal chamber connections. The most important variables in this respect are alignment between the atrial and ventricular septal structures and the pattern of ventricular architecture present.

The previous observations of differences between different cardiac regions (ventricular myocardium, atrial myocardium, Purkinje fibre system) with respect to the maturation of the M-line region and the establishment of mature metabolic characteristics, have been extended. It was found that M-line maturation proceeds differently also between different regions of the conduction system. The M-line proteins, myomesin and MM-creatine kinase, were detected earlier, by means of immunohistochemistry, in the AV bundle and bundle branch cells than in the AV node cells. Also, a difference was observed in large foetuses. Striations in the AV node were less evident than in the AV bundle and the bundle branches in sections incubated with antibodies against myomesin as well as against MM-creatine kinase. Using enzyme histochemistry it was observed that the differences in metabolic properties between the AV node, the AV bundle and the bundle branches on the one hand, and the ordinary myocardium on the other, of adult hearts, are not established at the early stages. No clear difference in activity of succinate dehydrogenase was seen between the conduction tissues and the ordinary myocardium in the foetal hearts, while the conduction tissues showed a lower activity in the adult hearts. Furthermore, the pattern of activity of mitochondrial glycerol-3-phosphate dehydrogenase between the conduction tissues and the atrial and ventricular myocardium was quite different in early foetal stages compared with the adult stage.

A new method for vital staining of the conduction system during heart operations, the iodine gas method, was used in 12 patients. The right bundle branch stained well in all cases, and could be confirmed by the naked eye. The conduction system was not damaged by the staining procedure. The iodine gas used in this method had no adverse effects on thyroid function (thyroxine, triiodothyronine, thyroxine binding globulin, protein bound iodine) or liver function (serum glutamate oxaloacetate transaminase, serum glutamate pyruvate transaminase, lactic dehydrogenase). This method, therefore, was found to be valuable and without complication for preventing conduction disturbances during cardiac operations.

We studied the normal spread of excitation on the anterior and posterior ventricular surface of open-chest dogs by recording unipolar electrograms from an array of 1124 electrodes spaced 2 mm apart. The array had the shape of the ventricular surface of the heart. The electrograms were processed by a computer and displayed as epicardial equipotential maps at 1-msec intervals. Isochrone maps also were drawn. Several new features of epicardial potential fields were identified: (1) a high number of breakthrough points; (2) the topography, apparent widths, velocities of the wavefronts and the related potential drop; (3) the topography of positive potential peaks in relation to the wavefronts. Fifteen to 24 breakthrough points were located on the anterior, and 10 to 13 on the posterior ventricular surface. Some were in previously described locations and many others in new locations. Specifically, 3 to 5 breakthrough points appeared close to the atrioventricular groove on the anterior right ventricle and 2 to 4 on the posterior heart aspect; these basal breakthrough points appeared when a large portion of ventricular surface was still unexcited. Due to the presence of numerous breakthrough points on the anterior and posterior aspect of the heart which had not previously been described, the spread of excitation on the ventricular surface was "mosaic-like," with activation wavefronts spreading in all directions, rather than radially from the two breakthrough points, as traditionally described. The positive potential peaks which lay ahead of the expanding wavefronts moved along preferential directions which were probably related to the myocardial fiber direction.