Complete Heart Block and Fatal Right Ventricular Failure in an Infant
Drs David W. Sapire and Suzanne M. Lopez
The mother of this child was a 22-year-old primigravida in apparent good health. There was no clinical or biochemical evidence indicating use of ethanol, addictive drugs, or other possibly teratogenic substances. Fetal heart rate on ultrasonograms done early in her pregnancy was about 150 beats per minute. Near the end of her fifth month of pregnancy, she was referred to the John Sealy Hospital at the University of Texas Medical Branch (UTMB) because the fetal heart rate had decreased to 86 beats per minute. Although the atrial rate on subsequent examinations remained about 150 beats per minute, the ventricular rate ranged from 40 to 50 beats per minute. Two weeks before delivery, atrial and ventricular rates were 144 and 47 beats per minute, respectively. When the ventricular rate began to slow even further, it was decided that risk of intrauterine death was imminent, and a Caesarean section delivery was performed at approximately the 28th week of gestation.
There was no significant family history relevant to the case. The mother’s serological studies for lupus erythematosus and antiphospholipid syndrome were negative.
Infant’s Hospital Course
The baby was in no major distress just after delivery, but her heart rate increased very little in response to increasing doses of isoproterenol. Within hours after birth, it was decided that an electronic pacemaker was advisable. During epicardial placement of electrodes near the left ventricular apex, the surgeon noted that the “infiltrated” right ventricular myocardium was thin and that it did not respond to electronic pacing. Over the following weeks, there were no conducted atrial beats, although there were regular P waves. Occasional single and multiple ventricular premature beats were recorded, but the heart was driven primarily by the electronic pacemaker. The main clinical problem was increasing right ventricular failure that did not respond to intensive care in the high-risk neonatal unit. Because of deteriorating performance by the original pacemaker, it was successfully replaced at about the fourth week of life, but the right ventricular failure continued to worsen, and the baby died at the age of 7 weeks.
Drs Thomas N. James, Myron M. Nichols, and Pier Luigi DiPatre
General examination. Body weight was 1750 g, and body length was 45.5 cm. The apparent gestational age was 34 weeks. Pacing electrodes were securely fixed over the left ventricular apex. The entire right ventricle was very thin, with a translucent aneurysmal bulge in the outflow region. All the major coronary arteries were patent, and the four cardiac valves were normal. The interatrial and interventricular septa were intact.
Anasarca was severe and generalized. Some acute tubular necrosis was present in the kidneys. The liver exhibited cardiac cirrhosis. The ductus arteriosus was closed. The heart and lungs together weighed 55 g. On microscopic examination, the thymus exhibited severe lymphocyte depletion. The heart was preserved in neutral formalin.
Special examination of the heart. The entire heart was routinely processed for embedding in paraffin and subsequent serial sectioning at 8 μm. Sectioning was begun from the anterior surface of the heart roughly parallel to the free wall of both the right and left ventricles. Every 10th section was saved, and every 30th section was routinely stained with the Goldner trichrome method. After initial screening, selected additional sections were stained with the Verhoeff–van Gieson’s elastic or periodic acid–Schiff method.
For immunohistochemical identification of apoptotic cells, we used the TUNEL (TdT-mediated dUTP-biotin nick end labeling) method,1 which identifies early DNA fragmentation in the nucleus on the basis of the specific binding of terminal deoxynucleotidyl transferase (TdT) to 3′-OH ends of DNA. Commercially available staining kits (Apoptag Plus, ONCOR) were used for this purpose. Interpretation was facilitated by the presence of counterstained nuclei (blue) of nonapoptotic neighboring cells and by the absence of inflammatory cellular infiltration in the vicinity of the apoptotic cells (brown nuclei).
Massive apoptotic destruction of the right ventricle. The selective nature of this process is illustrated in Figs 1 through 3⇓⇓⇓. Destruction of the right ventricular myocardium included both the septal and parietal bands of the crista supraventricularis.2 There were no histological abnormalities of the myocardium in the left ventricle, in either atrium, or in the interatrial septum.
Apoptotic destruction of the His bundle but not the AV node. The only other anatomic structure undergoing apoptotic destruction, except the right ventricle and its crista supraventricularis, was the His bundle. A contrasting comparison between cytology and histology of the His bundle and AV node is demonstrated in Figs 4⇓ and 5⇓, with normal myocytes of the AV node and only apoptotic debris and a slight increase in fibrosis of the His bundle; the human His bundle normally has more collagen than the AV node. An abnormal fibrous band anatomically separates the AV node and His bundle (Fig 4⇓).
Apoptotic destruction and anatomic configuration of the septal band of the crista supraventricularis. Since it is an integral normal component of the right ventricular myocardium, it is unsurprising that the septal band of the crista supraventricularis was destroyed along with all other myocardium of the right ventricle, as seen in Figs 6 through 8⇓⇓⇓. However, this clearly delimited configuration, surrounded by the other (left ventricular) normal myocardium within the interventricular septum, provides a unique opportunity to visualize how the septal band of the crista supraventricularis normally ‘fits’ into the interventricular septum and to appreciate the hemodynamic advantage of this anatomic configuration. In normal human hearts, it is very difficult to determine which portion of myocardium at the interventricular septal crest is provided by left ventricle or right ventricle.
Cytological and immunohistochemical nature of the apoptotic debris. All the areas of apoptotic destruction exhibited essentially the same cellular changes (Fig 9⇓). This includes the many examples of TUNEL-positive responses by apoptotic cells or apoptotic bodies and other extracellular and intracellular apoptotic debris, as shown in Figs 10⇓ and 11⇓.
Other significant findings. Despite a careful search in serial sections, we were unable to identify either a right or left bundle branch, nor was there any replacement tissue, such as subendocardial fibrosis, that might have represented the residuum of previous bundle branches that had earlier been destroyed. There were also no identifiable Purkinje cells in either the right or left endocardial regions. All other components of the cardiac conduction system except the His bundle were normal, although the sinus node was slightly smaller than usual.
Dr Thomas N. James
The cardiac anomaly named for Henry Stephen Magraw Uhl was described by him as a case report written in his first year out of medical school while he was working in the Department of Pathology at Johns Hopkins University.3 Although Uhl’s anomaly is relatively rare, it has fascinated cardiologists worldwide because of its great instructive value relative to other cardiac conditions. A special example is arrhythmogenic right ventricular dysplasia,4 5 6 7 8 9 10 11 12 13 14 15 a more frequently encountered entity. Despite growing clinical interest in them, however, the cause and pathogenesis of both conditions selectively involving the right ventricle remain frustratingly elusive.
Findings in the present case shed useful light on the pathogenesis of both Uhl’s anomaly and arrhythmogenic right ventricular dysplasia as well as congenital heart block and the functional importance of the crista supraventricularis. These and related questions will be discussed in the following sections.
Postnatal Morphogenesis of the Right Ventricle
A dramatic change in the work of the newborn’s heart begins shortly after birth, due to the abrupt but normal reduction in pressure against which the right ventricle must pump. A prevailing interpretation of the nature of adaptive postnatal changes in the two ventricles holds that the left ventricle steadily thickens to compensate for its new hemodynamic burden but that the “right ventricular mass remains stable,”16 thus accounting for the eventual normal disparity in right and left ventricular thickness in later life. This very logical and simple concept may be incorrect, since there are numerous examples in human development, especially in the early years of life, when mechanisms come into play to remove cells and tissue no longer useful in the body economy. Postnatal morphogenesis of the brain is a striking example with its normal loss (selective deletion) of enormous numbers of “surplus” neurons.
Something of this nature also happens to the normally but disadvantageously thick postnatal right ventricle. As but one example of a metabolic handicap, coronary flow would need to be maintained at a larger volume than the right ventricle of optimal thickness would require for its sharply diminished hemodynamic burden. In postnatal morphogenesis of the brain, it is now generally held that the surplus neurons are eliminated by apoptosis (programmed cell death). Similar events could lead to an active normal postnatal involution of the right ventricle on the reasonable assumption that nature would not needlessly postpone that sort of change. Findings in our case support this concept, except that the mechanisms normally limiting the extent of right ventricular apoptosis failed in this infant. This could be caused by either incessant triggering of apoptotic signals or loss of antiapoptotic mechanisms, or a combination of both. Although massive apoptosis has been reported among hemopoietic cells of the liver,17 in bone marrow,18 and in the central nervous system,17 to the best of our knowledge this extent of selective destruction of myocytes within the human heart has not been described previously.
When Did the Apoptotic Right Ventricular Destruction Begin?
If this occurred early in fetal life, it is improbable that the fetus would have survived. However, the grossly abnormal and electrically inexcitable right ventricle was apparent at the time of the first pacemaker implantation just after birth. From then on, the infant’s brief survival was secured by electronic pacing of the left ventricle and by major support therapy in the neonatal intensive care unit.
As the complete heart block persisted and the rhythm of the ventricles (probably of AV junctional origin) began to fail late in pregnancy, apoptotic destruction of both the His bundle and right ventricle may have commenced, possibly triggered by an abnormally early delivery of a signal for apoptosis to begin. Why that apoptosis did not “turn off” normally, certainly before complete destruction of the right ventricle, is unknown.
On the Congenital Heart Block in This Patient
After we had discovered the unusual histology and immunohistochemical features of the right ventricle and a similar process in the His bundle, our surprise at this coincidence was diminished by learning that heart block is not that rare in conjunction with Uhl’s anomaly11 or in association with arrhythmogenic right ventricular dysplasia.8 10
In a well-studied case of Uhl’s anomaly and complete heart block occurring in a 29-year-old woman, Bharati et al11 found extensive destruction of both bundle branches and in the dividing portion of the His bundle. In addition to an electrically inexcitable right ventricle during life and extensive loss of myocytes in the right ventricle demonstrated postmortem, they specifically noted similar destruction of myocytes on the right side of the crest of the interventricular septum, closely resembling our own findings (Figs 6 through 8⇑⇑⇑). In that heart, there were foci of inflammatory infiltration at many sites, but the investigators commented that this was more particularly true in older individuals and that inflammation was often absent in young subjects who died of Uhl’s anomaly. Thus, there are three similarities between their case and ours: (1) the presence of complete heart block and eventual death with intractable heart failure; (2) extensive destruction of the right ventricle, including the crista supraventricularis; and (3) major destruction of the His bundle with only minor abnormalities in other portions of the conduction system.
At least two components of normal cardiac morphogenesis may play a role in the pathogenesis of the conduction system changes we found. If normal development of the two bundle branches originates from the His bundle, as has been proposed,19 then their absence in our case may be attributed to the destroyed His bundle from which they could not arise normally, although this would probably require earlier destruction of the His bundle than we believe happened. A corollary consideration could be that the His bundle was selectively deleted in part because its connection with the AV node was severed, but there is also a problem with this interpretation. In previous examinations of hearts with unconnected AV node and His bundle, we have not seen such destruction of the His bundle20 ; however, that may be because severing of the connection in those cases occurred in adult life. The presence of a normal AV node in our case is a point against heart block here being related to unrecognized maternal lupus erythematosus, since the anatomic abnormality in such cases is absence of the AV node.21
In estimating when the destruction of the His bundle happened, which according to the baby’s clinical course almost certainly began in utero, we also need to explain where the ventricular rhythm necessary for survival of the fetus originated after heart block was established. For that purpose, the ratio between ventricular rate (47 beats per minute) and atrial rate (144 beats per minute) recorded 2 weeks before delivery gives us a clue. The ratio of 47/144 is 0.326, making the ventricular rate 33% of atrial rate. In animal experiments, we determined that the rhythm controlling the ventricles originating spontaneously during complete heart block, selectively produced either pharmacologically22 or by surgical incision in the AV junction,23 is exactly 33% of atrial rate. In dogs, that rhythm originates from the proximal end of the His bundle or the junction of AV node and His bundle.23
This leads us to suspect that the His bundle in our case retained some functional integrity until shortly before delivery, at which time its impulse-forming capacity began to fail. The complete heart block probably began considerably earlier as a consequence of completion of the fibrous separation of AV node and His bundle (Fig 4⇑) and would not require that the subsequent apoptotic destruction of the His bundle also commenced at that earlier time. But how such an impulse of AV junctional origin propagated to the ventricles in the absence of bundle branches remains a mystery.
It is possible that the structure that we are interpreting as the His bundle undergoing apoptotic destruction is actually just another component of the septal band of the crista supraventricularis, given that all those cells seem to be intermingled. But we believe that this is not true, for two reasons. First, its location atop the ventricular septal crest and just beneath the normal AV node (to which it is not connected), as well as its cross-sectional configuration, all favor an interpretation as His bundle. Second, if all this tissue is crista supraventricularis, then there is no His bundle, making it difficult if not impossible to explain what the fetal escape rhythm could have been to permit survival in utero when complete heart block developed.
Returning to the question of when the complete heart block developed during fetal life, it is important to note that the human cardiac conduction system is in essence fully formed by week 6 or 8 of gestation.19 24 25 In this child, when the fetal heart rate was 150 beats per minute early in pregnancy, it is likely that normal sinus rhythm was being regularly conducted to the ventricles. The mother was referred to UTMB near the end of her fifth month of pregnancy specifically because the fetal heart rate had dropped (on the ultrasonogram) to 86 beats per minute. Our subsequent observations never demonstrated conducted sinus rhythm during the remainder of the baby’s life. This makes it probable that heart block, most likely due to completion of the fibrous separation of AV node and His bundle, occurred at or near the end of the fifth month of gestation.
Right Ventricular Apoptosis, Uhl’s Anomaly, and Arrhythmogenic Right Ventricular Dysplasia
Although the probable relevance of our observations to the pathogenesis of Uhl’s anomaly seems straightforward, how it might relate to arrhythmogenic right ventricular dysplasia is less obvious. The frequent occurrence of paroxysmal arrhythmias and occasional sudden unexpected death in arrhythmogenic right ventricular dysplasia,4 6 8 11 combined with the likelihood that postnatal changes in the right ventricle are not simply a passive phenomenon but rather active involution, has previously led us to suggest that arrhythmogenic right ventricular dysplasia is also mediated by apoptosis.26
A recent study of Uhl’s anomaly and arrhythmogenic dysplasia9 concluded that these are separate and distinct morphological entities. However, the same authors wrote that although they may be distinct morphological entities, it is possible that they share a common pathogenesis. We submit that Uhl’s anomaly and arrhythmogenic right ventricular dysplasia have in common their pathogenesis as apoptotic dysplasia but that they differ in that Uhl’s anomaly is an incessant process ending in complete destruction of the right ventricle, whereas arrhythmogenic right ventricular dysplasia represents focal episodic apoptosis within the right ventricle beginning at any time, including later in life. If it continues to recur, it would be expected in many cases to end in sudden death or in intractable right ventricular failure.
Findings of focally recurring apoptosis have recently been described as responsible for gradual development of complete heart block associated with periodic arrhythmias ending in sudden death,27 except that in those cases it was the AV node, sinus node, and internodal pathways that were destroyed, whereas the His bundle and right ventricle were spared. There are undoubtedly more than one trigger mechanism for apoptosis in the human heart and several different patterns of recognition by which any elements of the conduction system are selected for destruction.
Why the apoptosis in Uhl’s anomaly becomes incessant and why apoptosis in arrhythmogenic right ventricular dysplasia begins and ends just when it does are equally important questions, for which there is at present no answer. It is similarly unclear why some reported cases of right ventricular dysplasia have either hemodynamic or histological evidence of concomitant left ventricular involvement.4 However, we suggest that examples of putative left ventricular involvement that specifically cite the interventricular septum11 probably represent unrecognized tissue of the septal band of the crista supraventricularis (Figs 6 through 8⇑⇑⇑), which is normally of right and not left ventricular origin.
Brief periods of apoptosis focally distributed in the right ventricle may occur in some individuals at any age and may recur over long periods of time. During such apoptotic bouts, there are at least two consequences that may be arrhythmogenic: First, the dying cells may themselves exhibit transiently increased excitability or automaticity; second, the normal route of right ventricular activation would necessarily be altered when the dying cells cease functioning, thus providing an anatomic substrate for a microreentrant or macroreentrant circuit.
On the Nature of Apoptosis in the Human Heart
Apoptosis has not, to the best of our knowledge, been previously described in relation to either Uhl’s anomaly or arrhythmogenic right ventricular dysplasia. One reason may be that apoptosis is characteristically a rather brief (minutes to an hour or so) but remarkably tidy process in which the dying cell is rapidly engulfed by macrophages28 or even adjacent cells, such as cardiac myocytes.26 This quick scavenging means that in apoptosis there is no release of noxious intracellular debris to provoke inflammatory cellular response, one important (and easily recognized) difference from the histopathological changes attending necrosis. These histological characteristics of apoptosis also closely resemble most published descriptions of what is found in arrhythmogenic right ventricular dysplasia, in which inflammatory changes or active necrosis is sometimes seen but is seldom prominent.
A recent review of fine structural features of myocytes in right ventricular biopsies obtained from patients with arrhythmogenic right ventricular dysplasia29 described mitochondrial abnormalities in particular but no features indicative of apoptosis. However, unlike our patient, in whom the extent of apoptosis was massive, it is often difficult to find individual or small groups of apoptotic cells because of their typically rapid removal.28
During normal cardiac morphogenesis in utero, programmed selective deletion of certain cells and tissue almost certainly involves apoptosis often and perhaps predominantly. But when apoptosis begins and fails to stop, as in the right ventricle in our patient, it can be fatally destructive. There seems to be little doubt that this baby would have died in utero if the emergency delivery and subsequent electronic pacing had not been used. The fact that the baby died anyway with progressively increasing and ultimately intractable right ventricular failure provides some useful lessons.
Hemodynamic Importance of the Crista Supraventricularis
It is tempting but probably mistaken to attribute the right ventricular failure directly to the destruction of the free right ventricular wall. Several studies have demonstrated that experimental destruction (eg, by cautery) of the free right ventricular wall is insufficient in itself to cause death and, in fact, is often survived with relatively little hemodynamic change.30 Other investigations have demonstrated that experimental right coronary occlusion must be very proximal to produce right ventricular failure, whereas more distal occlusions fail to do so, although they do cause some infarction of the right ventricular free wall.31 Only very proximal occlusion of the right coronary artery will interrupt all the primary blood supply to the crista supraventricularis.2 Spanish investigators have specifically described the destruction of the crista supraventricularis in a case of Uhl’s anomaly with terminal right ventricular failure.32
On the basis of anatomic study of human and other mammalian hearts,2 it has been suggested that it is crucial to the understanding of right ventricular failure to consider the normal functions of the crista supraventricularis. As long as the crista supraventricularis remains intact, left ventricular systole can tether even the damaged right ventricular free wall sufficiently to empty the right ventricle into the pulmonary artery. An intact crista supraventricularis is also necessary for normal function of the tricuspid valve, both by its purse-string action on the tricuspid orifice and by providing essential support for the anterior leaflet of the tricuspid valve. Thus, in our case it was the destruction of the crista supraventricularis, along with the free wall of the right ventricle, that made the right ventricular failure ultimately intractable. That a normal right atrium could hypothetically propel blood from the right ventricle into the main pulmonary artery is improbable because in our patient the right atrium was normal.
For its optimal mechanical effectiveness in coordinating and facilitating hemodynamic efficiency of the right ventricle with that of the left ventricle (including most of the interventricular septum), the anatomic configuration of the apoptotic tissue (septal band of the crista supraventricularis) in the upper right side of the crest of the interventricular septum (Figs 6 through 8⇑⇑⇑) explains some probable contractile advantages. Although the connection of the septal band with the rest of the crista supraventricularis is thin when seen in cross section (Fig 7A⇑), it is actually a broad sheet extending from an area posterior to the membranous interventricular septum to a region well anterior to the normal location of the His bundle. Furthermore, the bottom of the septal band tissue seen in cross section has a pronglike configuration (Fig 8A⇑), suggesting a firm anchor deep within the mass of interventricular septal myocardium.
Other Considerations Regarding Pathogenesis of the Destruction of the Right Ventricle
It is theoretically possible that the right ventricle in our patient underwent selective destruction as a surplus tissue because the complete heart block prevented its functional usefulness, leading to “disuse atrophy.” But this is improbable because the complete heart block had to affect the normal activation of the left ventricle as well (until electrically paced), and the left ventricle was histologically and functionally normal.
Similar considerations about hypoxia, acidosis, ischemia, or other serious complications in the baby’s hospital course, as contributing to the right ventricular destruction, also become implausible because the histologically normal right and left atria and entire left ventricle were inevitably subjected to these same noxious events. And there was no anatomic interruption of the blood supply to the right ventricle or His bundle.
On the Absence of Both Bundle Branches and the Distal Purkinje System
Although its histological appearance explains why the right ventricle in this case was inexcitable by pacing, the successful pacing of the left ventricle for a long period of time (several weeks) is not so easily understood. Absence of both a left bundle branch and any recognizable subendocardial Purkinje network means that the epicardial pacing impulse would not have the usual route of penetrating a Purkinje net and traveling rapidly to the rest of the left ventricle. Physiotherapists electrically stimulate large masses of skeletal muscle to contract vigorously without any available special electrical distribution net in that striated muscle, and we assume that something analogous may have happened in the left ventricular stimulation in our case, but that is not a fully satisfying explanation.
Summation for Apoptosis as the Explanation
Whatever name one may wish to give to the process of cell death in the heart of this infant, it was unequivocally selective within her heart but extensive in its distribution in the selected areas. It was associated with little or no significant inflammatory response, occurred in the presence of normally patent coronary arteries and impressively normal myocardium in the uninvolved areas, was associated with no recognizable systemic illness, and exhibited no earmarks of infiltrative lesions such as seen in amyloidosis, sarcoidosis, or scleroderma. This constellation of features and the extensive presence of TUNEL-positive cells in the areas of selective destruction leave us with no fully logical diagnosis except apoptosis.
Massive apoptosis selectively destroying the entire right ventricle and the His bundle.
This work was supported in part by the Pegasus Fund of the University of Texas Medical Branch, Galveston.
This clinicopathological conference was presented at pediatric grand rounds at the University of Texas Medical Branch, Galveston, on January 27, 1995.
- Copyright © 1996 by American Heart Association
Gavrieli Y, Sherman Y, Ben-Sasson SA. Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J Cell Biol. 1992;119:493-501.
Uhl HSM. A previously undescribed congenital malformation of the heart: almost total absence of the myocardium of the right ventricle. Bull Johns Hopkins Hosp. 1972;91:107-209.
Fontaine G, Fontaliran F, Lascault G, Aouate P, Tonet J, Frank R. Arrhythmogenic right ventricular dysplasia. In: Zipes DP, Jalife J, eds. Cardiac Electrophysiology: From Cell to Bedside. Philadelphia, Pa: WB Saunders Co; 1995:754-769.
Marcus FI, Fontaine GH, Guiraudon G, Frank R, Laurenceau JL, Malergue C, Grosgogeat Y. Right ventricular dysplasia: a report of 24 adult cases. Circulation. 1982;65:384-398.
Gerlis LM, Schmidt-Ott SC, Ho SY, Anderson RH. Dysplastic conditions of the right ventricular myocardium: Uhl’s anomaly vs arrhythmogenic right ventricular dysplasia. Br Heart J. 1993;69:142-150.
Bharati S, Ciraulo DA, Bilitch M, Rosen KM, Lev M. Inexcitable right ventricle and bilateral bundle branch block in Uhl’s disease. Circulation. 1978;57:636-644.
Nugent EW, Plauth WH, Edwards JE, Williams WH. The pathology, pathophysiology, recognition, and treatment of congenital heart disease. In: Schlant RC, Alexander RW, eds. The Heart. New York, NY: McGraw-Hill; 1994:1763.
Motoyama N, Wang F, Roth KA, Sawa H, Nakayama K, Nakayama K, Negishi I, Senju S, Zhang Q, Fujii S, Loh DY. Massive cell death of immature hematopoietic cells and neurons in Bcl-x-deficient mice. Science. 1995;267:1506-1510.
Raza A, Gezer S, Mundle S, Gao X-Z, Alvi S, Borok R, Rifkin S, Iftikhar A, Shetty V, Parcharidou A, Loew J, Marcus B, Khan Z, Chaney C, Showel J, Gregory S, Preisler H. Apoptosis in bone marrow biopsy samples involving stromal and hematopoietic cells in 50 patients with myelodysplastic syndromes. Blood. 1995;86:268-276.
Patten BM. The development of the sinoventricular conduction system. Univ Mich Med Bull. 1956;22:1-21.
James TN, Spencer MS, Kloepfer JC. De subitaneis mortibus, XXI: adult onset syncope, with comments on the nature of congenital heart block and the morphogenesis of the human atrioventricular septal junction. Circulation. 1976;54:1001-1009.
James TN, Isobe JH, Urthaler F. Correlative electrophysiological and anatomical studies concerning the site of origin of escape rhythm during complete atrioventricular block in the dog. Circ Res. 1979;45:108-119.
Duckworth JWA. The development of the sinu-atrial and atrioventricular nodes of the human heart. Edinburgh, Scotland: Department of Anatomy, University of Edinburgh; 1952. MD thesis.
Boyd JD. Development of the heart. In: Hamilton WF, Dow P, eds. Handbook of Physiology, Section 2: Circulation, Volume III. Washington, DC: American Physiological Society; 1965:2511-2543.
James TN. Normal and abnormal consequences of apoptosis in the human heart: from postnatal morphogenesis to paroxysmal arrhythmias. Circulation. 1994;90:556-573.
James TN, St Martin E, Willis PW, Lohr TO. Apoptosis as a possible cause of gradual development of complete heart block and fatal arrhythmias associated with absence of the AV node, the sinus node and the internodal pathways. Circulation. In press.
Donald DE, Essex HE. Pressure studies after inactivation of the major portion of the canine right ventricle. Am J Physiol. 1954;176:155-161.
Pérez Díaz L, Quero Jiménez M, Moreno Granados F, Pérez Martínez V, Merino Batres G. Congenital absence of myocardium of right ventricle: Uhl’s anomaly. Br Heart J. 1973;35:570-572.