Search

In summary, many of these patients have achieved a normal working capacity, but as a group there is a statistically significant reduction when compared to healthy children (p < 0.01). There was no significant difference between the group of patients who had their Mustard operation in infancy and later. As yet, there is no suggestion that the older patients, or those with the longest time interval between the operation and the exercise test, have a progressive reduction in working capacity. The decreased working capacity and moderate increase in ventilation suggests restricted cardiac output on exercise. This could be related to the abnormal heart rate response rather than indicating poor ventricular function. Equally, a restricted cardiac output on exercise may be due to inefficiency of atrial transport, tricuspid regurgitation or unmasked pulmonary vascular obstructive disease. Further studies may clarify these points. This study demonstrated that the exercise performance of asymptomatic patients 6-13 years after Mustard's operation was somewhat diminished, compared with a group of normal children. Nevertheless, these patients did lead and enjoy a normal life. Only longitudinal studies may bring the final answer to the important question: How many patients will be alive with normal or near normal exercise tolerance 40-50 years after Mustard's operation? Until such an answer is available, we believe that the Mustard operation should be used in the treatment of TGA, while the alternative techniques are carefully explored.

A study was performed to assess the feasibility, additional diagnostic value and potential applications of biplane transesophageal echocardiography in neonates, infants and children. One hundred thirty-two consecutive studies were attempted in 111 anesthetized children with congenital heart disease. Longitudinal and transverse planes were compared using 3 methods: (1) separate 7 mm longitudinal and transverse pediatric transducers used sequentially; (2) an experimental 9 × 8 mm biplane pediatric transducer; and (3) a standard adult biplane transducer (12 × 9 or 13 × 9 mm). In all but 1 patient, a probe could be inserted. The longitudinal plane provided superior visualization of both the right and left ventricular outflow tracts, the interatrial septum, the main pulmonary artery, the ascending aorta and the right coronary artery. In 18 patients (16%), the longitudinal plane provided completely new diagnostic information that was not obtained with combined transthoracic and transverse plane transesophageal echocardiography. However, the transverse plane was mandatory for demonstration of the 4-chamber view, short-axis cross sections through the great arteries, the distal right pulmonary artery and bifurcation of the left coronary artery. The longitudinal plane is complementary to the transverse plane, but cannot substitute for it.

Survival before definitive operations was studied in 191 infants with double-inlet ventricle presenting before 1 year of age (1973 to 1988, median follow-up 8.5 years). The morphologic spectrum was broad, with a great prevalence of associated lesions. The actuarial survival rate before definitive repair was 57% at 1 year, 43% at 5 years, and 42% at 10 years, worse than prior reports because of the younger age at entry into our series. Analysis of univariate risk factors established that right atrial isomerism (18% of the group, relative risk 2.9), common atrioventricular orifice (42%, 2.0), pulmonary atresia (20%, 3.4), obstruction of the systemic outflow tract (18%, 2.5), and extracardiac anomalous pulmonary venous connection (13%, 3.1) were strongly associated with poorer survival. Pulmonary stenosis (40%, 0.35), balanced pulmonary blood flow (9%, 0.40), and presentation at an older age (33%, 0.42 to 0.18) were beneficial (p

OBJECTIVE--To develop, test, and validate an algorithm for diagnosing disease in neonates during an over the telephone referral to a specialist cardiac centre. DESIGN--A draft algorithm requiring only data available to a referring paediatrician was generated. This was modified in the light of a retrospective review of case records. A questionnaire to elicit all the data required by the algorithm was then generated. There followed a prospective three phase evaluation during consecutive over the telephone referrals. This consisted of (a) a conventional phase with unstructured referral consultations, (b) a phase with referrals structured around the questionnaire but independent of the algorithm, and (c) a validation phase with the algorithm (and its previous errors) available during the referral consultation. SETTING--59 paediatric centres in south east England and a central specialist paediatric cardiology unit. PATIENTS--Consecutive neonates (aged less than 31 days) referred with suspected heart disease. The retrospective review was of records of 174 neonates from 1979. In the prospective evaluation (1987-90) the conventional phase comprised 71 neonates (over 5.5 months), the structured phase 203 neonates (over 14 months), and the validation phase 195 neonates (over 12 months). MAIN OUTCOME MEASURES--Diagnostic accuracy (assigning patients to the correct diagnostic category (out of 27)), of the referring paediatrician, the specialist after the referral consultation, and the algorithm as compared with the definitive diagnosis by echocardiography at the specialist centre, and score for the appropriateness of management in transit. RESULTS--Simply structuring the consultation by questionnaire (that is, proceeding from the conventional phase to the structured phase) improved the diagnostic accuracy of both paediatricians (from 34% (24/71 cases) to 48% (97/203) correct) and specialists (from 54% (38/71 cases) to 64% (130/203) correct). The algorithm (structured phase) would have been even more accurate (78% (158/203 cases); p less than 0.01). Management scores in the structured phase were also better than in the conventional phase (80%(162/203 cases)v 58% (41/71) appropriate; p less than 0.01). Management scores would have improved to 91% appropriate (185/203; p less than 0.001) had the algorithmic diagnoses dictated management. The superiority of the algorithm was maintained but not bettered in the validation phase. CONCLUSIONS--Applying the algorithm should reduce the morbidity and mortality of neonates with critical heart disease by aiding clinicians in therapeutic decisions for in transit care.

Only those congenital defects carrying a very low risk of complication (either before or after surgical correction) were considered. Atrial Septal Defects--(a) Ostium primum defects should be treated with caution either before or after surgical correction because of the risk of progressive conduction disorders and mitral regurgitation. (b) Ostium secundum defects could be considered for licensing (if the defect is small) or with surgical repair if the right ventricular systolic pressure is normal. (c) Sinus venosus defects--if too small to require surgical repair, licensing may be considered provided ambulatory electrocardiographic monitoring shows no evidence of arrhythmias. Surgery increases the risk of sino-atrial disease, thus licensing should be permitted only where there is no evidence of arrhythmia and adequate cardiological follow-up is possible. Ventricular Septal Defects--Subjects with very small defects not requiring surgical closure may be considered for licensing. Subjects who have had surgical closure have a risk of arrhythmias and should be carefully evaluated. Pulmonary Stenosis--If mild (either before or after surgery) may be licensed, but regular assessment perhaps including right heart catheterization is needed to demonstrate stability of the lesion. Persistent Ductus Arteriosus--Surgical closure should be recommended on diagnosis and need not affect licensing. Isolated Bicuspid Aortic Valve--Need not debar from licensing, but careful annual examination (with electrocardiogram 2-D echocardiography and fluroscopy ) is required to detect calcification, stenosis or regurgitation. Coarctation of aorta--Subjects who have had a repair before the age of 12 years may be considered for licensing after examination of other risk factors (blood pressure at rest and on exercise in particular). Those repaired over the age of 12 may be considered for restricted licensing if normotensive. These recommendations will need review in the light of further long-term studies currently under way.

Twelve patients with univentricular heart and common atrioventricular valve were identified by two-dimensional echocardiography. Seven had an ostium primum atrial septal defect and five a common atrium. The common atrioventricular valve had the appearance of a free floating anterior leaflet flanked by two lateral leaflets. The position of the rudimentary chamber in relation to the atrioventricular valve was posterior in six and anterior in two; no chamber was identified in four. Two-dimensional echocardiography provides a reliable method for the identification of common atrioventricular valve in univentricular heart.

A new performed, semi-rigid, polyethylene catheter, with built-in torque control, has been devised for entry to the pulmonary artery in complete transposition of the great arteries. It has been used 19 times in 17 patients: 18 times the pulmonary artery was entered from the right atrium (via the left atrium and ventricle) in a time between 40 s and 15 min (median 5 min); the patients' ages were 2 days to 6 years (median 8 months) and their weights were 3.1 to 13.3 kg (median 6.9 (kg: in the remaining 1 day-old patient, the procedures was terminated because of atrial flutter. The catheter was introduced into the axillary vein in 3 patients and thesaphenous or femoral vein in the remainder. It is suitable for angiocardiography, and the other heart chambers and vessels were easily entered. Thus the catheter has certain advantages over previously described methods for entry to the pulmonary artery, particularly when there is inferior vena caval thrombosis, or when angiocardiography is necessary. Its use does depend on the presence of an interatrial communication, so a method for entry to the pulmonary artery by retrograde catheterization from the axillary artery using a different catheter is also presented; this was successful in 2 patients with ventriculal sepatal defect.

In order to study the usefulness of incorporating a subpulmonary ventricular chamber (SPVC) in the Fontan circulation, we compared the hemodynamics of 18 patients who underwent an atriopulmonary connection with those of 17 patients with an atrioventricular connection. Early postoperatively, regardless of whether an SPVC was incorporated or not, the mean pulmonary artery pressure was not higher than the mean right atrial pressure in any of our patients. The stroke work of the right heart was at best zero, and pressure tracings showed an a wave in the pulmonary artery (right atrium-dependent circulation). If an additional gradient was interposed between the right atrium and pulmonary artery, the effective stroke work of the right heart might be negative and the work curve have a negative slope. This can occur irrespective of the use of an SPVC. In the long term, if it grows sufficiently, an SPVC can sometimes provide a right ventricle-dependent pulmonary circulation with a more normal hemodynamic pattern (mean right atrial pressure less than mean pulmonary artery pressure). We concluded that an SPVC should be used in Fontan’s circulation, provided that an obstruction-free atrioventricular connection can be established.

The atrial morphology and venous connections were assessed "blind" in 51 necropsy specimens from patients with visceral heterotaxy. This was compared with bronchial morphology as established by dissection. Six specimens were found to have both atria and bronchi in situs solitus or inversus, and were rejected. In the remainder, atrial isomerism was diagnosed, though this required minor revision of the atrial assessment in two patients. Thirty-four patients had isomeric right atria and bronchi, while 11 had isomeric left atria and bronchi. In seven cases, splenic status was unknown, but in seven of the remaining 38 (18.4%) atrial isomerism was not associated with either asplenia or polysplenia. Nevertheless, right isomerism was strongly associated with total anomalous pulmonary venous drainage (as is asplenia) and left isomerism was likewise associated with interruption of the inferior vena cava (as is polysplenia). Bilateral superior venae cavae and hepatic veins, and absence of the coronary sinus, were frequent in both forms of isomerism (as they are in asplenia and polysplenia). These findings suggest that atrial situs can be defined as solitus inversus, right isomerism, and left isomerism. This determination of atrial situs is quite independent of any other abnormalities of visceral situs. The high incidence of anomalies of both venous return and common atrium resulted in presumed complete mixing of blood at atrial level in all but one patient (97.8%), making the haemodynamic connection between atria and ventricles almost always ambiguous. To describe this anatomical connection as ambiguous when there are two ventricles present is therefore no more than recognition of anatomical and haemodynamic reality.

Three patients with ventriculoarterial concordance and an abnormal atrioventricular connexion were investigated. One patient had total anomalous pulmonary venous connexion to a left vertical vein. In 1972 this was repaired and a ventricular septal defect was closed, but the presence of atrioventricular discordance was missed and the patient died. In the other two patients, Mustard's operation resulted in survival. One patient had situs solitus and atrioventricular discordance with an associated ventricular septal defect, which was closed. The other had left isomerism (quasi solitus) with an ambiguous atrioventricular connexion (quasi discordant). A previously noted ventricular septal defect had spontaneously closed.

The left pulmonary artery and ductus arteriosus were ligated in 14 pigs at birth. Animals were sacrificed at intervals from 2 to 24 weeks of age. In the right lung the pulmonary artery and in the left, either the distal pulmonary artery, bronchial arteries or both were injected. The fixed lung specimens were studied by arteriography, dissection and microscopic examination of serial and random sections of lung tissue. The bronchial arterial circulation to, and within the right lung appeared normal and was similar to that described in the human lung. In the left lung, the bronchial arterial circulation hypertrophied rapidly during the first 2 weeks, and large anastomoses between pulmonary and systemic circulations were found at the same sites as in the normal pig lung. The position and structural characteristics of the anastomosing arteries is described in the different types of broncho-pulmonary connection. In most animals aged 16 weeks or more, peripheral bronchial arteries immediately proximal to the anastomotic sites, developed intimal and medial proliferation. The left lung continued to grow although in all animals it was small. The axial pulmonary artery and its branches became smaller with age. These findings help explain how the lung is perfused and grows in children with congenital heart disease and an acquired collateral pulmonary arterial circulation.

Between April, 1975, and December, 1979, 99 modified Blalock-Taussig shunts (MBTSs) were carried out at The Hospital for Sick Children, Great Ormond Street. The operation consists of interposing between the subclavian artery and the pulmonary artery a prosthesis of greater diameter than that of the subclavian artery. The first 13 operations were performed with a prosthesis of woven Dacron. Conduits of expanded polytetrafluorethylene (PTFE) were used for 86 MBTSs, and these form the basis of this report. Forty-four (51.1%) shunts were performed in infancy, 25 in patients under the age of 1 month (29.0%). Prostheses of 4 mm were used in 34 cases, 5 mm in 14, and 6 mm in 38. There were five shunt failures (5.8%) documented at postmortem or angiocardiographic investigation. Including two late deaths for which postmortem examinations were not obtained, the overall failure rate was 8.1% (seven patients). All these patients were operated upon in infancy, four of them in the neonatal period. Although a longer follow-up is necessary to assess the validity of these shunts, the early results are encouraging. We believe we can now recommend MBTS as an alternative when the classical Blalock-Taussig shunt is considered unsuitable.

Interrupted aortic arch is a rare condition, usually lethal in early infancy without treatment. The only characteristic feature on conventional non-invasive investigation is peripheral pulse inequality, which indicates ductal construction, and therefore may be absent or transient and preterminal. We report the cross-sectional echocardiographic findings in seven patients with aortic arch interruption between the left carotid and subclavian arteries. Their ages were 1 day to 7 months (median 7.5 days). The arterial connection was concordant in four, double outlet right ventricle in two, and truncus arteriosus in one. In each case the ascending aorta was small in comparison to the pulmonary trunk. From the suprasternal approach the ascending aorta could be seen to terminate in the left carotid artery, and the ductus to continue smoothly into the descending aorta, with no vestige of an aortic arch linking its ascending and descending portions. The left subclavian artery was seen to arise distal to the ductus in all but one patient. All four patients with ventriculoarterial concordance had pronounced subaortic stenosis caused by posterior displacement of the infundibular septum. Cross-sectional echocardiography therefore provides the only accurate method of non-invasive diagnosis of this condition. It permits early treatment with prostaglandins to prevent ductal closure, a planned approach to cardiac catheterisation, and a further means of investigating the nature of subaortic stenosis in this condition.

Seventy-six hearts were studied having no direct communication between the right atrium and the ventricular mass. The different cardiac chamber combinations producing so-called "tricuspid atresia" were considered. The nature of the atrioventricular valve atresia, the morphology of the ventricular mass, the size and position of the interventricular communication, the ventriculoarterial connection, and the presence of subarterial outflow tract obstruction were all analyzed. The majority of cases were of the "classical" type, i.e., absent right atrioventricular connection with the left atrium connected to the left ventricle. In another group there was absence of the right atrioventricular connection but the left atrium drained into the morphologically right ventricle, which was left-sided. In a third group both atrial chambers connected with the ventricular mass but some structure, an imperforate valve or muscular partition, completely blocked the flow pathway through the right side of the heart All the hearts described are candidates for the Fontan procedure with either an atrioventricular or an atriopulmonary conduit. The surgical options would not be affected by the observed variability at the atrioventricular junction but would be dictated by the feasibility of incorporating the subpulmonary ventricle within the pulmonary circulation.

Precise non-invasive localisation of the site of a small ventricular septal defect was attempted using a new technique that simultaneously combines conventional cross sectional echocardiography with a Doppler system by superimposing the colour coded direction and velocity of blood flow directly on to real time ultrasound images. Twenty three patients with unoperated ventricular septal defects and a further eight after surgical closure were studied; 12 children with normal hearts served as controls. A colour coded blood flow jet entering the right ventricle during systole was identified in all 23 unoperated patients, in 11 of whom the defect was too small to be visualised by conventional cross sectional echocardiography. The colour Doppler technique precisely located 19 perimembranous and five trabecular defects (one patient had two defects). Five of the postoperative patients were without clinical evidence of a significant shunt but had pansystolic murmurs. In each of these five, trans-septal shunt blood flow as demonstrated by colour Doppler images whereas in only three of these patients was the residual defect large enough to be visualised by conventional cross sectional echocardiography. Three postoperative patients had no murmurs and showed no residual shunt on colour Doppler images. This was confirmed at cardiac catheterisation. There were no false positive results among the controls. This technique is useful for the more accurate diagnosis and location of ventricular septal defects and may help in assessing their natural or surgical closure.

At present there is no reliable method of recognising atrial isomerism by two dimensional echocardiography. We therefore used two dimensional echocardiography to examine 158 patients including 25 with atrial isomerism and four with situs inversus. Particular attention was paid to the short and long axis subcostal scans of the abdomen. Using the position of the inferior vena cava and the aorta with respect to the spine it was possible to separate those with situs solitus from the others. Two false positives for abnormal situs had exomphalos. In situs solitus the aorta lay to the left of the spine and the inferior vena cava lay to the right. One patient with situs solitus and azygos continuation of the inferior vena cava also had inferior vena cava to right atrial connection. In the four patients with situs inversus the mirror image of the normal pattern was present. In nine patients with right isomerism the inferior vena cava and aorta ran together on one or other side of the spine. The inferior vena cava, anterior to the aorta at the level of the diaphragm, received at least the right hepatic veins (normal or partial anomalous hepatic venous connection). Of the 16 patients with left isomerism, 14 had azygos continuation of the inferior vena cava which was visualised posterior to the aorta in all but two. All patients with left isomerism had total anomalous hepatic venous connection to one or both atria via one or two separate veins. Two dimensional echocardiography therefore provides the means of detecting abnormal atrial situs and of diagnosing right or left isomerism in the great majority of patients, if not all.

Traumatic heart block remains a major concern after ventricular septal defect (VSD) closure in patients with atrioventricular discordance. A technique of closing the VSD, placing the suture line on the morphologically right side of the septum without opening the systemic ventricle, is described. This was used in 13 consecutive patients. The VSD was closed through the right atrium in eight patients, the left ventricle in three patients, the right atrium and the left ventricle in one patient, and through the right ventricle in one patient with atrioventricular discordance and ventriculo-arterial concordance. All patients were in sinus rhythm preoperatively, two exhibited atrioventricular dissociation before intracardiac manipulation began, and 11 patients were in sinus rhythm postoperatively. No major arrhythmia could be attributed to the closure of the VSD.

Formalin infiltration of the ductus arteriosus was performed in 13 neonates with pulmonary atresia (three with ventricular septal defect, two with tricuspid atresia, and eight with intact ventricular septum, one of whom had Ebstein's anomaly) in an attempt to maintain duct patency. Nine patients had an additional retrograde pulmonary valvotomy and one a Blalock-Taussig shunt. The mean preoperative systemic PO2 was 31 mmHg and rose significantly to 46 mmHg after operation. However, the PO2 rose in only two of the four patients who had formalin infiltration alone. There were five early deaths and three had clinical evidence of duct closure. Four patients needed further surgery (systemic-pulmonary artery shunt) within eight days. Seven of eight surviving patients were shown to have a closed duct at repeat cardiac catheterisation (two to 13 months after operation). The eighth died before restudy. There were four late deaths, one following late reoperation and the other three within five months of early reoperation. These results suggest that formalin infiltration cannot ensure long-term duct patency. As short-term patency can be maintained more reliably by prostaglandin administration, we have abandoned the operation.

To gain a better understanding of the anatomy of doubly committed subarterial defects and its relation to findings at cross sectional echocardiography and angiocardiography, eight necropsy specimens from patients with this condition were examined, and preoperative echocardiograms and angiocardiograms from 313 patients with surgically or necropsy confirmed outlet defect were reviewed. Of these, 48 had doubly committed subarterial defects. Morphological review showed that doubly committed defects are roofed by the arterial valves in fibrous continuity because of lack of both the outlet septum and the "septal" aspect of the subpulmonary infundibulum. Angiocardiography had a lower sensitivity (50%) than echocardiography (95%) for diagnosis of doubly committed defects, but each was highly specific. In five (14%) of 35 available echocardiograms the arterial valves were normally offset, but in the remainder they were at the same level. The ventriculoarterial connection was concordant in 37/48 (77%), discordant in five (10%) of 48, and double outlet right ventricle in six (13%) of 48. Displacement of the fibrous raphe between the arterial valves in relation to the ventricular septum below was associated with outflow obstruction in 14 patients (pulmonary in nine and aortic in five). These features were readily identified by echocardiography. Thus echocardiography is not only a more accurate method than angiocardiography of recognising these defects, but also shows that the arterial pole of the heart is architecturally abnormal.

If cross sectional echocardiography in isolation is used to diagnose critical left ventricular outflow obstruction in neonates, false positive and false negative diagnoses may result. Continuous wave Doppler velocimetry was used to measure blood flow velocity in the ascending and descending thoracic aorta in six neonates (aged less than 6 weeks) presenting with reduced or absent peripheral pulses in order to determine the important sites of obstruction. This technique demonstrated abnormal high velocity blood flow jets (three times higher than normal) in the ascending aorta in three patients with normal descending aortic flow velocity, suggesting aortic stenosis. In the other three patients velocity in the ascending aorta was normal but high in the descending aorta, suggesting coarctation. The Doppler diagnosis was confirmed in the five patients who required surgery. Two patients had residual high velocity jets after aortic valvotomy. Both had significant pressure gradients across the aortic valve at cardiac catheterisation with good agreement between actual gradients and those predicted by the Doppler technique. Thus a combined anatomical and physiological approach using cross sectional echocardiography and continuous wave Doppler velocimetry enables accurate non-invasive definition of the site of left ventricular outflow obstruction and may obviate the need for invasive investigation in these sick neonates.

Determination of atrial situs is of cardinal importance in the analysis of complex congenital heart lesions, and is best predicted from bronchial situs. Previous methods for assessing bronchial morphology, however, are unsuited to the very young patient. To assess bronchial morphology, 100 consecutive patients under 18 months of age (medium 57.5 days) with suspected congenital heart disease were studied by high kilovoltage filtered beam radiographs, before cardiac catheterisation. This low radiation dose technique clearly defined bronchial anatomy in 95 patients. The lengths of the left and right main bronchi were compared and 10 cases (10%) had a ratio less than 1.5 suggesting bronchial isomerism. Discriminant function analysis based on tracheal width and bronchial length enabled clear distinction of right from left bronchi. Four patients had bilateral right and six had bilateral left bronchi. Four of these 10 cases died and had necropsy confirmation of the radiological diagnosis. Practical prediction about cardiac anatomy, particularly the systemic and pulmonary venous return, may be made when bronchial morphology is known.

In light of the recent developments in surgical treatment of double-outlet right ventricle, the anatomic observations on this lesion were reevaluated. For this review, double-outlet right ventricle was diagnosed when more than half of both arterial valves were connected to the same ventricle, although appreciating the reasons for using, in a clinical context, a "90% rule" rather than the "50% rule" used in this review. Although this ventriculo-arterial connection can exist with any segmental combination, most often it is found in the setting of usually arranged atrial chambers (solitus) and atrioventricular concordance. Categorization of this subset is then done on the basis of the relationships of the arterial trunks. Three main groups stand out: intertwining arterial trunks and "normally related" arterial valves, and parallel trunks, but with the aortic valve to the right side or left side. These groupings give information concerning the site of the ventricular septal defect, which in any group may be perimembranous, muscular or be doubly committed and subarterial. Infundibular morphology is also variable, and the proximity of the arterial valves to the roof of the defect is largely determined by the extent of the ventriculo-infundibular fold.

Azygos continuation of the inferior vena cava has importance for both the invasive diagnosis of congenital heart disease by catheterization and for surgical treatment. Cross-sectional echocardiography was used to examine 1,000 patients (1 day to 16 years, mean 3.3 years) who also had angiographic or surgical confirmation. Twenty-eight patients (3%) had azygos continuation (left 13, right 14, bilateral 1) and, in 26 patients, the hepatic portion of the inferior vena cava was absent. Azygos continuation was prospectively detected in all and was directly visualized in subcostal scans as a venous structure posterior to the aorta coursing behind the heart and not entering the inferior aspect of either atrium in 26/28 (93%). Azygos connection to the ipsilateral superior vena cava or atrium was correctly predicted in all. The inferior vena cava was visualized in all patients without azygos continuation, except one neonate with omphalocele. We conclude that cross-sectional echocardiography can accurately detect azygos continuation of the inferior vena cava and predict its side and connection.

The roles of posterior bulging of the interventricular septum (septal bulge) and of systolic septal mitral apposition in patients with simple transposition of the great arteries are not known. Cross sectional echocardiograms of 40 such patients were reviewed (after exclusion of those with fixed left ventricular outflow tract obstruction) and haemodynamic findings were compared with long and short axis measurements within the left ventricle. There was no significant correlation between the degree of septal bulge and systolic gradient across the left ventricular outflow tract, but septal bulge correlated weakly with systolic right ventricular pressure and inversely with pulmonary arteriolar resistance index. Systolic left ventricular outflow gradient was inversely related to the minimum systolic distance between the anterior mitral leaflet and interventricular septum. No patients without complete systolic apposition of the anterior mitral leaflet and interventricular septum had a left ventricular outflow gradient greater than 20 mm Hg. Conversely, even when cross sectional echocardiography showed apparently total obstruction of the left ventricular outflow tract at some time in systole there was often no significant gradient detected during haemodynamic study. In the short axis cuts closeness of the papillary muscles to the interventricular septum or to each other was unrelated to systolic gradient. This study shows that (a) cross sectional echocardiography can identify fixed obstruction of the left ventricular outflow tract in simple transposition of the great arteries; (b) the degree of septal bulge, unless complicated by fibrous thickening of the anterior mitral leaflet and interventricular septum, is unrelated to the gradient across the left ventricular outflow tract; (c) the absence of systolic septal/mitral apposition excludes a significant gradient at that site across the left ventricular outflow tract; and (d) papillary muscle geometry is unrelated to dynamic gradients across the left ventricular outflow in this condition.

Ninety-six patients with an atrioventricular septal defect were assessed by two dimensional echocardiography. Forty-eight patients were judged as having two discrete valve orifices. In this group, 37 had an intact ventricular septum and 11 a ventricular septal defect. Two patterns of attachment of the valve leaflets to the interventricular septum were noted. Forty-eight patients had a common valve orifice. In 29 patients there was absence of tethering of the anterior bridging leaflet to the interventricular septum. A further 19 had varying degrees of tethering of the anterior bridging leaflet. A ventricular septal defect was identified under the anterior bridging leaflet in all cases. In 58 cases where the posterior bridging leaflet was identified, a ventricular septal defect was visualised in eight and missed in two. Seventy-eight patients had an ostium primum defect, 12 a common atrium, and a further six an intact interatrial septum. Two dimensional echocardiography provides in depth morphological information about the type of atrioventricular defect and can correctly identify the presence of a ventricular septal defect beneath the anterior or posterior bridging leaflet, unless it exists between short crowded chordae.

The dismal outlook for patients with pulmonary atresia with intact ventricular septum may be related to associated right ventricular hypoplasia. Study of 32 autopsy specimens and 46 angiocardiograms of neonates with this lesion suggested that the cavitary hypoplasia was related to massive hypertrophy of the right ventricular wall. This hypertrophy was sufficient to obliterate the trabecular and/or infundibular portion of the ventricular cavity entirely in one-third of the cases; this observation forms the basis for a revised classification of these hearts. Three autopsies and 14 angiograms of neonates with critical pulmonary stenosis were examined. Hearts with obliterated infundibular and trabecular cavities had thicker walls and smaller tricuspid valves, as estimated angiographically or at autopsy, than those in which the normal three portions of the ventricular cavity were represented.

All 212 neonates undergoing cardiac surgery at this hospital during the 5-year period from 1976 to 1980 inclusive were reviewed. Forty required open heart surgery with 23 (57%) deaths. One hundred and seventy-four neonates underwent non-bypass procedures and could be divided into three groups: group 1 (82 patients) had inadequate pulmonary blood flow, group 2 (33 patients) had increased pulmonary blood flow or inadequate mixing, and group 3 (59 patients) had coarctation of the aorta, alone or with associated lesions. Forty-four (25%) of the neonates undergoing non-bypass procedures died. Two required bypass surgery later in the first month of life. Metabolic acidosis and the need for preoperative respiratory support were appreciably greater in non-surviving patients. The spectrum of diagnoses encountered and types of operative procedures performed are analysed.

A nomenclature is described for congenital heart disease employing sequential chamber localization. It is an eclectic system based in part upon the previous classifications of Van Praagh and Kirklin. It links together the atrial, ventricular, and arterial segments of the heart and then permits tabulation of associated anomalies. The atrial segment of the heart can exist as situs solitus, situs inversus, or situs ambiguus. Atrioventricular connexions can be concordant or discordant. In certain circumstances the terms concordant and discordant cannot be used. These are in the presence of primitive ventricle and in the presence of situs ambiguus. Alternative terms are described for these contingencies. Ventriculo-arterial connexions can be (a) normal; (b) transposition; (c) double outlet ventricle; or (d) single arterial trunk. These are defined as connexions; relations are relegated to secondary position. Associated anomalies are categorized in terms of venous return, atria, atrioventricular junction, ventricles, and great arteries. Controversial topics are discussed with regard to previous definitions.

Between 1971 and 1982, 249 patients received right heart extracardiac conduits. The 173 patients who survived 30 days were reviewed retrospectively; 72 had antibiotic-sterilized aortic homografts, 97 had xenografts of various types, and four had valveless tubes. Reoperation for conduit obstruction was required as early as 13 months after the original operation, but actuarial analysis showed no differences in overall performance of homograft and heterograft conduit groups with respect to late death or survival free of conduit obstruction. More important risk factors for late death or the need for reoperation were the severity of the underlying cardiac lesions and the influence of early postoperative complications. When analyzed statistically, the performance of conduits bearing homografts was disappointing, influenced sometimes by complications in the Dacron extension tubes. Repairs dispensing with conduits entirely should be evaluated.

Results of primary closure of ventricular septal defects are compared with those of two-stage repair, with banding of the pulmonary artery followed by debanding and closure. Apart from the high incidence of unsatisfactory results after banding and a significant morbidity with the two-stage approach, the mortality for primary repair (2.4%) is considerably lower than that achieved with the staged repair (19.3%). Primary repair of ventricular septal defect is advocated for infants resistant to maximal medical treatment. A more flexible policy is adopted for patients with multiple ventricular septal defects and those with associated anomalies.

From May, 1972, to May, 1978, right ventricular outflow tract reconstruction was performed in 20 patients with congenital heart disease. A monocusp patch constructed entirely of glutaraldehyde-stabilized calf pericardium, was employed in 19 patients, and a composite conduit consisting of a three-cusp pericardial xenograft valve in a Dacron tube was used in 1 patient. There were 2 hospital deaths (10%) and no late deaths. The follow-up ranged from 5 to 75 months (mean, 55.7 ± 4.6 months; total, 1,002 patient months). Eleven patients had completed 5 years of follow-up at the time of writing. Hemodynamic studies were performed in 10 patients at an average of 40 months after operation and the right ventricle-pulmonary artery systolic gradient was 9.3 ± 1.7 mm Hg. This was not significantly different in 6 patients who had a second postoperative catheterization at 51.8 ± 2.0 months following operation. Angiography showed fully mobile, thin valve cusps. These results compare well with those reported with other types of conduits, particularly with reference to relief of obstruction.

It is evident that the term “tricuspid atresia” is imprecise and can be used to describe hearts with widely varying morphology. It might be preferable if the term were reserved for one or other of these variants, probably that which is the most commonly encountered. In this anomaly there is total absence of the connection between the right atrium and the ventricular mass, and the ventricular morphology is that of a univentricular heart. Only if a detailed segmental analysis of the heart is performed can the distinction be made between this and other malformations in which the tricuspid valve is imperforate and in which any ventricular morphology can be encountered. Tricuspid atresia is usually, but not always, univentricular.

Anatomical studies were made on 114 necropsy specimens of atrioventricular defects with atrioventricular concordance. The malformation is characterised by disproportion between the ventricular inlet and outlet dimensions and a malorientation of the aortic valve relative to the atrioventricular valve or valves. Associated with this there is a characteristic 'scopped-out' appearance of the muscular ventricular septum, gross abnormalities of the membranous components of the septum as compared with the normal heart, and narrowing of the aortic outflow tract. Hearts with these anatomical features can be divided into partial and complete forms depending on the morphology of the atrioventricular annuli. In the partial form the septal leaflets are conjoined to give separate mitral and tricuspid orifices, the conjoined leaflets being displaced into the ventricles and usually attached to the crest of the septum. In the complete form, anterior and posterior components of the 'septal' leaflets are separate, so that a single valve orifice connects the atrial to the ventricular chambers. Further subdivision of the complete form, apart from the morphology of the anterior leaflet, is dependent upon the presence or absence of an ostium primum atrial septal defect.

The value of two dimensional echocardiography in identifying communications between the ascending aorta and pulmonary trunk or individual pulmonary arteries was assessed in 24 children, all of whom had either angiocardiographic and surgical or angiocardiographic confirmation alone. Fourteen cases had truncus arteriosus, four aortopulmonary window, four anomalous origin of the left pulmonary artery from the ascending aorta, and two anomalous origin of the right pulmonary artery from the ascending aorta. It was possible to identify reliably each individual abnormality with a combination of suprasternal, precordial, and subcostal cuts. Problems only arose in differentiating truncus arteriosus from pulmonary atresia and ventricular septal defect when the main pulmonary artery and infundibular region of the right ventricle were extremely hypoplastic.

Most hearts described as "single ventricle" or "univentricular heart" possess 2 ventricular chambers, even though 1 is usually described as an "outlet chamber." This stems from the wide acceptance that the criterion of a single ventricle is the presence of a double-inlet atrioventricular (AV) connection. In recent years, using this criterion, an attempt was made to show how hearts with double-inlet right ventricle or "classic tricuspid atresia" were (in terms of ventricular morphology) just as univentricular as "single ventricle with outlet chamber." This attempt brought still further confusion to an already contentious topic. The root of the problem clearly is the injudicious use of the adjective "single" or "univentricular." Conventionally it is used to describe the ventricular mass. In most hearts with double-inlet connection it is not the ventricles that are univentricular; it is the AV connection. The concept of a univentricular AV connection, then, appropriately groups hearts with double-inlet along with those having absence of 1 AV connection. It distinguishes this entire group from those other hearts with biventricular AV connections (each atrium connected to its own ventricle). The term "univentricular AV connection" is thus a collective one for all those hearts in which the atria connect to only 1 ventricle. Confusion will be completely removed if individual hearts are described for what they are in terms of AV connection and ventricular morphology (for example, double-inlet left ventricle with rudimentary right ventricle rather than single ventricle with outlet chamber).

Eight patients were catheterized between 1.2 and 2.4 years after reconstruction of the right ventricular outflow tract with autologous fascia lata. Whereas the immediate post-bypass pressures had demonstrated a maximum gradient of 15 mm. Hg across the fascial valve at the time of recatheterization, this varied from 55 to 142 mm. Hg (mean 83 mm. Hg). Right ventricular systolic pressure varied between 80 and 160 mm. Hg (mean 106 mm. Hg). All but one patient had clinical evidence of pulmonary incompetence. Selective angiocardiography with injection into the right ventricle and pulmonary artery demonstrated shrunken, thickened, immobile valve cusps with an abrupt stenosis of the fascial tube or a diaphragm across it. Re-operation has been done in five patients. In each, the graft tube functioned satisfactorily, but at the site of the “valve”, no leaflets were visible, and the orifice of the conduit narrowed abruptly to 6 to 10 mm. in diameter. It is concluded that autologous fascia lata is unsuitable for reconstruction of the right ventricular outflow tract.

Twenty-five autopsy specimens of complete transposition of the great arteries with intact ventricular septum (VS) were categorized as “bulging” (11 cases) or “nonbulging” (14 cases) according to the curvature of the VS. A fibrous ridge was observed on the VS, especially at the site of mitral apposition in 82% of the bulging group. No fibrous ridge was seen in the nonbulging group. An objective index of anteroposteriorness was then designed to measure the distance between the midpoint of the nonfacing aortic cusp line and the left anterior descending coronary artery. In the bulging group, the aorta lay more anterior to the pulmonary trunk, whereas in the hearts with a straight VS, the aorta tended to lie side by side and to the right of the pulmonary trunk. The midmitral line is an imaginary line in the middle of the anterior mitral leaflet. The more the pulmonary valve is wedged toward the right atrioventricular junction, the more the midmitral line will pass through the nonfacing pulmonary cusp rather than right pulmonary cusp or the nonfacing/right commissure. This was so in 100% of the bulging group, but in only 36% of the nonbulging group. These differences between the groups in terms of antero-posterior index and extent of wedging were statistically significant. In conclusion, if the aorta lies more anterior and to the left of pulmonary trunk rather than side by side and to the right, the “wedged” subpulmonary area will be more susceptible to obstruction caused by septal bulging. Cross-sectional echocardiography is the best means of diagnosing these features.

Sixty-six hearts were examined in which a single arterial trunk, leaving the base of the heart through a single semilunar valve, supplied the aorta, pulmonary artery and coronary arteries. Careful attention was paid to the infundibular morphologic features in these hearts, and these were compared with findings in 24 hearts with single aortic trunk, pulmonary atresia and ventricular septal defect. It was concluded that the two anomalies represented morphologically discrete conditions. This was particularly true with regard to the disposition of the infundibular septum, the ventriculo-infundibular fold and the relation of the coronary arteries to the semilunar sinuses. Although it is theoretically possible for a heart with true persistent truncus arteriosus to have absence of the pulmonary trunk and right and left pulmonary arteries, it is argued that such hearts (“truncus type IV”) are best classified as pulmonary atresia with ventricular septal defect. It is also argued that hearts with a common arterial trunk supplied through discrete ventricular outflow tracts and two semilunar valves are best considered examples of aorticopulmonary window. It is suggested that persistent truncus arteriosus is best defined as that condition in which a single arterial trunk leaves the heart through a single semilunar valve and supplies the aorta, one or both pulmonary arteries and the coronary arteries.

Critical aortic valvular stenosis presents in infancy with severe congestive heart failure. Clinical assessment and electrocardiography are of value, but cardiac catheterization with angiography has been considered mandatory prior to surgical treatment. With cross-section echocardiography an accurate diagnosis of aortic stenosis and associated lesions is possible. Over the past 2 years, we have established a protocol according to which, if a clinical diagnosis of critical aortic stenosis is confirmed by cross-sectional echocardiography in the absence of major associated cardiac anomalies, infants are submitted for aortic valvotomy under inflow occlusion without invasive studies. This protocol was used in an effort to decrease the mortality rate by avoiding the preoperative stress of cardiac catheterization and angiography, as well as the hazards of cardiopulmonary bypass in the severely ill infant. Eight infants with critical aortic stenosis have been operated upon, five without prior cardiac catheterization. Ages at operation ranged from 2 days to 7 months, with six children less than 2 weeks of age. The noninvasive diagnosis was confirmed at operation in each case. There was one early postoperative death and one late death. No death has been related to the technique of inflow occlusion. A decision tree for the noninvasive assessment of suspected critical aortic stenosis based on the clinical features and echocardiographic findings is presented.

Closing the ventricular septal defect in patients with atrioventricular discordance with sutures placed through the defect onto the morphologically right septal surface should avoid production of complete heart block. To discover whether this procedure otherwise affects conduction, standard electrocardiogram (ECG) were compared preoperatively and postoperatively in 11 such patients. Operation had lengthened the PR interval in one and shortened it in another. The QRS interval remained unchanged in 5 patients thought 2 of these showed minor changes in QRS configuration. Four patients showed QRS prolongation (increase greater than 20 msec) with delayed conduction towards the right ventricle and the initial QRS vector preserved, i.e., morphologically right bundle branch block. Two patients showed QRS prolongation with initial QRS vector alteration and delayed conduction towards the left ventricle, i.e., morphologically left bundle branch block. Repair through the morphologically left ventricle invariably caused bundle branch block. Repair through a right atriotomy caused bundle branch block in only 3 of 8 patients. Therefore closure of a ventricular septal defect by this method, particularly transatrially, need not affect conduction. Morphologically left ventriculotomy rarely if ever causes morphologically left bundle branch block. When present, morphologically right bundle branch block presumably results from interruption of the proximal right bundle.

Actual and potential factors that create an obstruction to the flow within the systemic side of the heart have been analyzed in 114 specimens with atrioventricular septal defects. Hearts with separate right and left atrioventricular orifices and with common atrioventricular orifice were included. Potential inlet obstruction was due to tethering of the "cleft" of the left atrioventricular valve to the septum in 17.5%, incomplete "cleft" in 8%, and deformed valve leaflets in 70% of cases. Potential factors predisposing to left ventricular outflow tract obstruction were seen in 70% of cases, 21% being due to abnormal position of the papillary muscle and the remainder due to intrinsic hypoplasia of the outflow tract. Anatomic factors producing unequivocal obstruction were seen in 29% of cases. Twenty-two percent had obstructions at the inlet level, of which 13% were due to malformations of the left atrioventricular valve or the left component of a common valve. The other 9% had overall inlet obstruction due to severe hypoplasia of the left ventricle itself. Unequivocal outlet obstruction was observed in 7% of cases, 5% due to abnormal disposition of papillary muscles in the outflow tract and 2% to aneurysm of the interventricular septal remnant. Although anatomic obstruction occurs in the minority of cases (29%), this report shows that potential obstruction must be taken into account when considering surgical repairs.

To determine the sensitivity and specificity of cross-sectional echocardiography in diagnosing anomalous systemic venous return we used the technique in 800 consecutive children with congenital heart disease and whom the diagnosis was ultimately confirmed by angiography. Cross-sectional echocardiography was performed without prior knowledge of the diagnosis in all but 11 patients, who were recalled because of a known abnormality of atrial situs. The sensitivity of cross-sectional echocardiographic detection of various structures was as follows: right superior vena cava 792/792 (100%); left superior vena cava 46/48 (96%); bilateral superior vena cava 38/40 (95%); bridging innominate vein with bilateral superior vena cava 13/18 (72%); connection of superior caval segment to heart (coronary sinus or either atrium) (100%); absence of suprarenal inferior vena cava 23/23 (100%); azygos continuation of the inferior vena cava 31/33 (91%); downstream connection of azygos continuation, once seen, 21/21 (100%); partial anomalous hepatic venous connection (one hepatic vein not connected to the inferior vena cava) 1/1 (100%); total anomalous hepatic venous connection (invariably associated with left isomerism) 23/23 (100%). The specificity of each above diagnoses was 100% except in one infant with exomphalos in whom absence of the suprarenal inferior vena cava was incorrectly diagnosed. Thus cross-sectional echocardiography is an extremely specific and highly sensitive method of recognizing anomalous systemic venous return. It is therefore of great value of planning both cardiac catheterisation and cannulation for open heart surgery.