Hearts with biventricular connections and deficient atrioventricular septation are readily recognizable from their characteristic morphology.1 As yet, however, there is still no consensus on precisely what to call them,2 how to categorize them, and, perhaps most importantly, how to approach their surgical repair.

A historical review highlights an interest in the etiology of birth defects over the course of several centuries and recalls specific hypotheses on the origin of cardiovascular defects postulated 60 years ago. The “State of the Art” of the epidemiology of cardiovascular malformations is characterized by an expanding base of knowledge which holds promise for eventual preventive interventions. Epidemiologic investigations conducted in the past decade have led to an increased understanding of the benefits and problems of studies of populations and of methodologic issues in surveillance of disease and study of risk factors. Comparative evaluations of reported studies to an ongoing population-based case-control investigation (the Baltimore-Washington Infant Study) demonstrate variations in prevalence by time and place, evidence for a strong genetic predisposition to cardiac maldevelopment and new hypotheses regarding environmental risks in the home and place of occupation.

Prevalence of atrioventricular septal defect in liveborn infants at birth was calculated in Bohemia (population of 6.3 million). Of 664,208 liveborn infants, 4,409 had a congenital malformation of the heart and in 126 of them (0.19/1,000 live births, 2.9% of all heart defects) this was found to be an atrioventricular septal defect. The ratio between those with a common valvar orifice (complete form) versus those with separate right and left atrioventricular valvar orifices (incomplete form) was 1.14:1. The ratio of girls to boys was 1.17:1. Children born in March, July and October were more likely to have the defect (0.26/1,000 live births) than those born in May (0.10/ 1,000 live births). The defect was the eighth most frequent critical congenital cardiac malformation. Eight in each 100,000 liveborn children, and 3.56% of all patients with congenital cardiac disease, are at high risk of dying due to deficient atrioventricular septation, which in 82.5% resulted in the presence of a common atrioventricular valve (complete form). Unconditional death rate was studied in 1,008 liveborn children who died with a cardiac malformation before the age of 15 before the advent of cardiac surgery. Of these, 64 (6.35%) died with an atrioventricular septal defect which was complete in 5.95% and incomplete in 0.40%. All but one of 60 children with the complete form (common atrioventricular valve) died in the first year of life. The last child died between the age of 2 and 5 years. The estimate for natural survival was calculated using our own data on the prevalence and age distribution of mortality. Of all the liveborn children with atrioventricular septal defect, 91% are expected to survive the first month, 81% (78−85) the first trimester, 63% (56−69) six months, and 49% (35−54) the first year. The survival curve remains stable thereafter.

In this review, we discuss and describe those features which distinguish hearts with abnormal atrioventricular septation from the normal heart. The hearts, best described as atrioventricular septal defects, are unified by having a common atrioventricular junction guarded by a valve having five leaflets. The left component has three leaflets and cannot be interpreted in terms of a cleft in a normal mitral valve. The papillary muscles supporting this valve are also markedly dissimilar from the arrangement seen in normal hearts. The subaortic outflow tract is displaced superiorly compared to the normal heart, and is no longer wedged between the left valve and the septum. There is marked discrepancy in the inlet and outlet lengths of the ventricular mass, these dimensions being equal in hearts with normal atrioventricular septation. Although having the above features in common, atrioventricular septal defects show anatomic variations related to the arrangement of the bridging leaflets and their relationship to the septal structures. There may be a common valvar orifice or separate right and left orifices. The anatomic potential for shunting may be at atrial or ventricular levels, or both. Rarely, the septal structures may be intact. Other important features include ventricular dominance, the left ventricular outflow tract, and the disposition of the atrioventricular conduction tissues.

Defects of the atrioventricular septum impose changes in the septal architecture, atrioventricular valves, and conduction system that must be well understood for optimal repair. The atrioventricular node and bundles are shifted to more posterior and inferior positions that can be anticipated by the surgeon so as to maintain the integrity of atrioventricular conduction. The atrioventricular valve is a five leaflet structure that possesses a common orifice or separate orifices depending on characteristics of the superior and inferior bridging leaflets. These features are demonstrated by intraoperative photographs viewed from the surgeon's position.

Cross-sectional echocardiography has made a major contribution to the preoperative evaluation of the patient with an atrioventricular septal defect,1−5 such that in many instances this modality has been the sole imaging technique used to formulate a medical or surgical plan6. With the advent of Doppler echocardiography additional hemodynamic information is now available, which enables the cardiologist and surgeon to evaluate the site and magnitude of associated atrioventricular valvar regurgitation, and obtain an accurate measurement of pulmonary arterial pressure along with an assessment of outflow tract obstruction, if present.

In a series of over 7,000 pregnancies referred for fetal echocardiography, atrioventricular septal defects were detected in 103 fetuses, (17% of 594 anomalies). A large proportion (42%) also had isomerism of the atrial appendages, abnormalities in arrangement and morphology of the thoracoabdominal organs, and a high incidence of associated cardiac malformations. When this was associated with abnormalities of rhythm, the fetus often showed signs of intrauterine congestive heart failure; hydrops, pleural effusion and ascites. Those with trisomic karyotypes (predominantly trisomy 21) were associated with usual atrial arrangement and concordant connections of the cardiac segments. Of the 103 fetuses, 53 were examined at autopsy. Extracardiac abnormalities were documented, and detailed morphologic examination and measurements of the heart were made. The ratios of inlet to outlet septal disproportion were not significantly different from those found in postnatal hearts. Again, as in postnatal life, those hearts with common atrioventricular orifice had a greater degree of ‘scooping’ of the ventricular septum compared to those with separate right and left atrioventricular valves. There were no significant differences in these findings in relation to the karyotype of the fetus. There was an increased incidence of ventricular dominance, and also of coarctation or hypoplasia of the aortic arch, in fetal hearts when compared to autopsied hearts obtained in postnatal life. In summary, the distinguishing morphological characteristics of atrioventricular septal defect are present from 16 weeks of gestation.

Malfunctioning of the left atrioventricular valve has always been, and remains, the major incremental risk factor in the repair of atrioventricular septal defect. Now that the cardiac surgeon has ample time to assess the anatomy and function of the left valve, results have improved, but are still less than ideal. On the presumption that the anterior leaflet of the mitral valve is “cleft” in this anomaly, it used to be common practice to close the “cleft”. Currently, a substantial number of surgeons employ this technique, often irrespective of the individual anatomy, and in the majority of cases with success.

Despite improvement in the early outcome for patients undergoing complete repair of atrioventricular septal defects with a common atrioventricular orifice, the incidence of reoperation for failure of the repair of the left atrioventricular valve remains high. The surgical techniques employed in our most recent experience, including routine suture of the septal commissure and liberal use of valvar annuloplasty, were successful in achieving a very satisfactory competence of left atrioventricular valve with a iow incidence of reoperation for failure of the repair.

During a 10 year interval, between January 1, 1981 and January 1, 1991, primary repair of atrioventricular septal defect with a common atrioventricular orifice was performed in 95 patients younger than 1 year. Patients with atrioventricular septal defect having two atrioventricular orifices and a small ventricular septal defect, and those with severe hypoplasia of the left ventricle, were not included in the present study. According to the presence of associated anomalies, patients were divided into those with simple forms (72%) and those with complex forms (28%). These included parachute deformity of the left atrioventricular valve (10 cases), multiple muscular ventricular septal defects (5 cases), low-lying infundibular stenosis of the right ventricle (4 cases), obstruction of the left ventricular outflow tract (3 cases), tetralogy ofFallot (2 cases) and hypoplasia of the left ventricle (2 cases).

Between 1968 and December 1990, 29 patients underwent 33 reoperations for dysfunction of the left atrioventricular valve after correction ofatrioventricular septal defect. Repair of the valve was possible in all 10 patients who, initially, had a common atrioventricular orifice, using straightforward procedures such as closure of the septal commissure and annuloplasties according to Wooler. Additional valvar abnormalities (double orifice; fenestration of leaflets; malpositioned or malformed papillary muscles; and additional clefts) were rare in this group of patients. In contrast, these anomalies were frequent in the 19 patients who, initially, had separate right and left atrioventricular valves.

We advocate primary repair of atrioventricular septal defect with a common atrioventricular orifice in symptomatic children presenting with intractable congestive heart failure. The single patch technique has been utilized as our procedure of choice. A low operative mortality and a low incidence of residual ventricular septal defect and regurgitation of the left atrioventricular valve requiring reoperation has been achieved. This report emphasizes the technical details of the repair. Careful delineation of anatomic detail is considered to be of the utmost importance. Precise suture placement, tailoring the patch, and reconstruction of the left atrioventricular valve are essential to achieve complete closure of the atrial and ventricular septal defects and functional integrity of the left atrioventricular valve.

A retrospective analysis of 55 patients who underwent replacement of the left atrioventricular valve following surgery for atrioventricular septal defect with separate right and left atrioventricular valves or with a common valvar orifice has revealed a marked decrease in the need for valvar replacement over time. Between 1983 and 1987, the incidence of replacement following repair has been 3% in infants. Overall early mortality for the entire series, extending from 1970 to 1987, was 22%. Mortality in the first year of life was 58%, which was significantly greater than mortality beyond one year of age (14%, p<0.01). Obstruction of the left ventricular outflow tract was a frequent incremental risk factor for death during infancy. The incidence of complete heart block (25%) is disappointingly high, and may reflect distortion of the left atrioventricular junction in an attempt to place as large a prosthesis as possible. The incidence of complete heart block has not decreased with time. These results confirm the desirability of secondary valvoplasty procedures to enable a child to be larger before ultimate replacement of the valve is required.

The outcome of primary repair in young infants having atrioventricular septal defect with a common atrioventricular orifice and regurgitation across the left atrioventricular valve is not yet satisfactory. We studied the significance of the characteristics of left ventricular volume and mass and the predicted wall stress for the outcome of repair in 13 infants with this lesion. Three patients died of left heart failure after operation, although neither residual shunting at ventricular level nor regurgitation across the left atrioventricular valve was present. End-diastolic volume and ejection fraction of the left ventricle were 228 ±66% and 0.65 ±0.06 of normal, respectively, with no difference between the survivors and non-survivors. End-diastolic thickness of the posterior ventricular wall, determined by echocardiogram, was within normal range for body size in all patients.

We present a case of atrioventricular septal defect associated with tetralogy of Fallot which was diagnosed by cross-sectional echocardiography and angiography. The diagnosis was confirmed at necropsy. This case is unusual for the absence of Down's syndrome. We discuss the role of clinical, echocardiographic, hemodynamic and angiographic studies.