Incomplete Atrioventricular Canal Complicated by Cardiac Tamponade and Bidirectional Shunting in an Adult Dog

Introduction

The primitive heart consists of a simple endothelial tube. The caudal part of this tube later forms the atria and sinus venosus, while the cranial part forms the ventricles and the truncus arteriosus.^1,2 The atria are initially divided from each other by the septum primum, which grows ventrad from the dorsocranial wall of the developing atrium. The ventral portion of this structure eventually fuses with the endocardial cushions (after sealing off the early ostium primum) to partially eliminate the previous communication between the two developing atria.^1,2 Amore dorsal opening between the two atria, the ostium secundum, is formed by coalescing perforations in the superior aspect of the septum primum.¹ As this dorsal foramen appears, a second membranous fold grows ventrally into the atrium, to the right of the septum primum. This septum secundum overlaps the ostium secundum, but it leaves a communication between the two atria in the form of the oval foramen. Persistence of this foramen into adulthood is known as “patent foramen ovale.”¹ Other more severe forms of interatrial communications include atrial septal defects of the primum and secundum areas, which are distinguished from each other mostly by location.² The most severe congenital anomaly in this category is a complete absence of the interatrial septum, resulting in a condition known as common atrium, or cor triloculare biventriculare.^3–7 When accompanied by a cleft in the septal leaflet of the mitral valve, the common atrium has been referred to as an incomplete or partial atrioventricular (AV) canal.²

The purpose of this report is to describe the clinical, echocardiographic, and pathological findings in an adult dog that had an incomplete AV canal and some unusual complications.

Case Report

A 6-year-old, 33-kg, spayed female Weimaraner was presented to the emergency service at the University Hospital of the Koret School of Veterinary Medicine; the dog had a 3-day history of anorexia, lethargy, labored breathing, and progressive abdominal distention. Physical examination findings included pale, gray-to-cyanotic mucous membranes; a heart rate of 132 beats per minute (bpm); and muffled heart sounds. The femoral arterial pulses were weak but constant in intensity throughout the respiratory cycle. Dyspnea and tachypnea (respiratory rate of 48 breaths per minute, marked abdominal effort) were noted. The abdomen was distended, and a noticeable fluid wave occurred upon abdominal ballottement. No abnormal lung sounds, heart murmurs, or split heart sounds were ausculted.

Arterial blood gas analysis showed a partial oxygen pressure (PaO₂) of 48.8 mm Hg (reference range ~100 mm Hg in room air), with an oxygen saturation of 78.8% (reference range >96%). Arterial blood pH was 7.46 (reference range 7.36 to 7.44); bicarbonate (HCO₃⁻) was 21.1 mmol/L (reference range 18 to 26 mmol/L); and partial carbon dioxide pressure (PaCO₂) was 30.2 mm Hg (reference range 36 to 44 mm Hg). These values were suggestive of hypoxia and mild respiratory alkalosis, which were probably secondary to tachypnea. A complete blood count, serum creatinine, and electrolyte concentrations were normal. A packed cell volume (PCV) was 44% (reference range 37% to 55%), which suggested that hypoxia was acute, as there was no evidence of compensatory polycythemia.

Electrocardiography [Figure 1] revealed a sinus rhythm of 142 bpm, with evidence of first-degree AV block and a right bundle branch block. Abdominocentesis yielded 3 liters of a clear, pale yellow, modified transudate, with negligible red blood cells and total solids of 5 g/dL. The creatinine level of the abdominal effusate was 0.95 mg/dL. Thoracocentesis yielded 400 mL of a serosanguinotic fluid, which was not analyzed.

The dog was supplemented with oxygen via a nasal catheter and sedated with intramuscular morphine^a (0.3 mg/kg intramuscularly [IM]), but cyanosis only mildly improved. Pulmonary thromboembolism was considered unlikely, as there was no history of possible risk factors (e.g., protein-losing nephropathy, protein-losing enteropathy, hyperadrenocorticism, immune-mediated diseases, diabetes mellitus). A right-to-left shunting cardiac anomaly was suspected, with congenital heart disease considered likely despite the animal’s age.

Although thoracic radiography was clinically indicated based on the dyspnea, it was declined. Following additional sedation with midazolam^b (0.1 mg/kg IM), echocardiography was performed^c and revealed no evidence of an interatrial septum on any of the standard views (i.e., right parasternal short axis, right parasternal long axis, left apical four-chamber, left apical five-chamber, left cranial parasternal short axis) [Figure 2; Video, http://www.jaaha.org/cgi/content/full/43/4/223] These findings were compatible with a diagnosis of a common atrium rather than an incomplete AV canal, since there was no echocardiographic evidence of a cleft mitral valve or of mitral valve insufficiency. A moderate pericardial effusion was seen, along with evidence of diastolic collapse of the right dorsal (basilar) aspect of the suspected common atrium; these findings were consistent with cardiac tamponade. Right ventricular enlargement and diastolic interventricular septal flattening were demonstrated in the right parasternal short-axis view, which were consistent with chronic, right-sided volume overload. The latter finding was supportive of right-sided congestive heart failure (CHF) as a cause, rather than a result, of the pericardial effusion.

Aortic and pulmonary arterial peak flow velocities were normal. When agitated saline (used as contrast echocardiographic material) was injected through the right cephalic vein, it flowed from the right atrium to the right ventricle. When injected into the left saphenous vein, it flowed from the right atrium into the left atrium and then into the left ventricle.

Pericardiocentesis yielded 240 mL of serosanguineous fluid, which was later interpreted as pleural effusion because it was identical in content to the pleural effusion and different from the clear pericardial effusion found at necropsy. The dog’s breathing pattern improved, cyanosis was markedly decreased, and the dog showed more interest in food. Heart rate decreased to 110 bpm, and femoral arterial pulses were subjectively stronger. Enalapril^d (0.5 mg/kg per os [PO] q 12 hours), furosemide^e (1.5 mg/kg PO q 8 hours), and digoxin^f (0.01 mg/kg PO q 12 hours) therapy was initiated to minimize signs of right-sided CHF. Symptomatic relief occurred for 8 hours; but progressive and severe cyanosis, dyspnea, and tachypnea (120 respirations per minute) ensued and were attributed to recurrent cardiac tamponade. The dog was euthanized and a necropsy was performed.

Necropsy revealed no extrathoracic abnormalities. A serosanguinous pleural effusion (1.5 L) was found, and 50 mL of clear, amber-colored pericardial fluid was present in the pericardial sac. The heart weighed 380 grams, which was 1.15%of the dog’s total body weight (reference range in adult dogs is 0.75% to 0.83%).⁸ The right atrial body and appendage, as well as the right ventricle, were severely enlarged [Figure 3A]. Fibrin tags were adhered to the epicardium and were suggestive of pericardial effusion of at least several days’ duration. When the atria were opened, there was no visible division between them [Figures 3B, 3C, 3D]. A 4-mm wide, endocardial fold extended ventrad to the superior aspect of the “common atrium” [Figures 3B, 3C]; this was thought to be the dorsal aspect of the primitive septum primum, the under-developed septum secundum, or a fusion between the two. The interventricular septumappeared fully intact, as were both the semilunar valves. A cleft was identified in the septal leaflet of the mitral valve [Figure 3D], which resulted in the final diagnosis of an incomplete AV canal rather than a common atrium, despite no clinical or echocardiographic evidence of mitral valve regurgitation.

Discussion

The late onset of symptoms in this dog was not very surprising, in light of the fact that many humans diagnosed with an atrial septal defect are reportedly asymptomatic early in life, and a common atrium has been previously diagnosed in late adulthood in humans.^9–11 It is possible that some small animals born with this condition remain asymptomatic throughout their adult lives, making the defect underdiagnosed or unreported.

The electrocardiogram initially recorded in this dog demonstrated a severe right-axis deviation, which was attributed to a right bundle branch block, as previously reported in canine atrial septal defect.¹² It was possible, however, that right ventricular enlargement may also have contributed to the axis shift. The right bundle branch block may also have resulted from right ventricular volume overload and ventricular wall stretching, rather than solely from a primary delay in the right-sided ventricular conduction system.¹³ The presence of a first-degree AV block with the bundle branch block made it tempting to speculate that both aberrancies were triggered by a common etiology, and these may have reflected a potential impact by the incomplete AV canal on the conduction system. An AV canal, however, is not necessarily an underlying cause of severe, aberrant conduction, because the embryonic conduction system does not develop before the AV canal and is not necessarily interrupted by the development of the canal defect.¹³

Interestingly, a small strand of tissue at the dorsocaudal aspect of the “common atrium” (similar to the dorsocaudal fold observed in this dog) was also detected in the only other reported case in a dog.¹⁴ Such a fold may have persisted from a point in time when early development of the embryonic septum had ceased. The absence of an interatrial septum in the presence of a dorsocaudal fold may reflect a large-bore, primum atrial septal defect. The coexistence of such a large defect and a cleft mitral valve is compatible with the diagnosis of incomplete AV canal (i.e., incomplete AV septal defect).^15–17 The lack of clinical or echocardiographic evidence of a cleft mitral valve or mitral valve insufficiency prior to necropsy may be attributed to a high intra-atrial pressure during the terminal phase of the disease course. Under such circumstances, a low systolic pressure gradient between the left ventricle and the left atrium may have made the premortem detection of mitral valve regurgitation difficult.

As a possible explanation for the changes found in this dog, the following sequence of events may have taken place [Figure 4]. Chronic, right-sided volume overload may have occurred from left-to-right intra-atrial flow across the large defect (i.e., across the absent inter-atrial septum), triggering progressive pulmonary overcirculation and eventual secondary pulmonary arteriolar hypertension. This process may have culminated in elevation of right ventricular diastolic (filling) pressure. A gradual decrease in right ventricular compliance may have occurred that affected right atrial drainage into the right ventricle. As right ventricular compliance gradually decreased, blood flow from the right atrium was directed, at least during certain components of the cardiac cycle, toward the path of least resistance (i.e., through the left-sided portion of the “common atrium” into the left ventricle). The resulting hypoxia may have exacerbated progressive pulmonary arteriolar vasoconstriction and increased pulmonary vascular resistance even further, thereby increasing right ventricular afterload and right ventricular filling pressure, which increased the risk of right-sided CHF. Progressive hypoxia, along with progressive right-sided CHF, may have gradually increased the heart rate enough to abbreviate the duration of diastole, further diminishing right ventricular compliance and facilitating right-to-left shunting of blood across the “common atrium.”¹⁸ A positive feedback mechanism may have been triggered in which the contribution of the right-to-left component of an intra-atrial bidirectional shunt continued to increase. The gradual development of pericardial effusion, possibly secondary to right-sided CHF, could have culminated in cardiac tamponade, which increased intra-atrial pressure even further. Symptoms of right-sided CHF and right-to-left shunting of blood may have progressed from cardiac tamponade. The normal PCV measured upon presentation suggested that significant right-to-left shunting of blood was not chronic.

The proposed gradual development of hypoxia and cyanosis in this case was compatible with Eisenmenger physiology, which is considered rare in dogs >6 months of age.^2,17 If the proposed scenario did lead to the dog’s demise, cardiac tamponade ultimately developed as a complication of pericardial effusion secondary to right-sided CHF (a rare sequela).¹⁹ Cardiac tamponade may have been the last component of the positive feedback mechanism and the one directly responsible for the final exacerbation of the ongoing right-to-left shunting, with terminal cyanosis and dyspnea.

An additional contributor to pulmonary arteriolar hypertension (other than pulmonary overcirculation) cannot be excluded.^20–24 While the presence of a large atrial septal defect may underload the left ventricle, the left atrium remains overloaded from recycling of blood between the right-sided cardiac chambers and pulmonary vasculature. Under the circumstances described in this dog, the left atrium may also have been pressure overloaded, as both atria may have hemodynamically operated as a single chamber (both being equally over-pressurized). Such mechanisms may have further compromised left atrial compliance, leading to pulmonary venous congestion and possibly pulmonary edema. Late, left-sided CHF may also have contributed to the terminal cyanosis and dyspnea.

To the authors’ knowledge, this is the first report of cyanotic heart disease from suspected Eisenmenger physiology complicating right-sided CHF secondary to a congenital incomplete AV canal in an adult dog. The lack of invasive pressure measurements, as well as histopathological assessment of tissues from this dog, did not allow a full understanding of the pathophysiology of the disease process in this dog. Invasive diagnostic tests (e.g., measurements of central venous pressure, right ventricular systolic and diastolic pressures, and/or pulmonary capillary wedge pressures) may have been useful to rule out or confirm progressive CHF prior to the development of terminal stages of the disease. Histopathological evaluation of pulmonary tissue may have helped to confirm changes compatible with chronic, progressive pulmonary hypertension.

Conclusion

An incomplete AV canal with bidirectional shunting and cardiac tamponade in a 6-year-old dog was initially diagnosed echocardiographically as a common atrium. The dog failed to respond to medical therapy and was euthanized. Upon necropsy, the defect was confirmed to be an incomplete AV canal. Earlier recognition of, and prompt reaction to, the development of relatively late-onset complications of the defect may prove effective in changing the outcomes of future cases of incomplete AV canal.

Acknowledgment

The authors thank James W. Buchanan, Emeritus Professor, for having carefully reviewed some of the figures in this case report.