Fetal, infant, and childhood growth and adult blood pressure: a longitudinal study from birth to 22 years of age.

BACKGROUND: People who are small at birth tend to have higher blood pressure in later life. However, it is not clear whether it is fetal growth restriction or the accelerated postnatal growth that often follows it that leads to higher blood pressure. METHODS AND RESULTS: We studied blood pressure in 346 British men and women aged 22 years whose size had been measured at birth and for the first 10 years of life. Their childhood growth was characterized using a conditional method that, free from the effect of regression to the mean, estimated catch-up growth. People who had been small at birth but who gained weight rapidly during early childhood (1 to 5 years) had the highest adult blood pressures. Systolic pressure increased by 1.3 mm Hg (95% CI, 0.3 to 2.3) for every standard deviation score decrease in birth weight and, independently, increased by 1.6 mm Hg (95% CI, 0.6 to 2.7) for every standard deviation score increase in early childhood weight gain. Adjustment for adult body mass index attenuated the effect of early childhood weight gain but not of birth weight. Relationships were smaller for diastolic pressure. Weight gain in the first year of life did not influence adult blood pressure. CONCLUSIONS: Part of the risk of adult hypertension is set in fetal life. Accelerated weight gain in early childhood adds to this risk, which is partly mediated through the prediction of adult fatness. The primary prevention of hypertension may depend on strategies that promote fetal growth and reduce childhood obesity.

L-arginine and substance P reverse the pulmonary endothelial dysfunction caused by congenital heart surgery.

BACKGROUND: The increase in pulmonary vascular resistance (PVR) seen in children after cardiopulmonary bypass has been attributed to transient pulmonary endothelial dysfunction (PED). We therefore examined PED in children with congenital heart disease by assessing the L-arginine-nitric oxide (NO) pathway in terms of substrate supplementation (L-arginine [L-Arg]), stimulation of endogenous NO release (substance P [Sub-P]), and end-product provision (inhaled NO) before and after open heart surgery. METHODS AND RESULTS: Ten patients (aged 0.62+/-0.27 years) with pulmonary hypertension undergoing cardiac catheterization who had not had surgery and 10 patients (aged 0.65+/-0.73 years) who had recently undergone cardiopulmonary bypass were examined. All were sedated and paralyzed and received positive-pressure ventilation. Blood samples and pressure measurements were taken from catheters in the pulmonary artery and the pulmonary vein or left atrium. Respiratory mass spectrometry was used to measure oxygen uptake, and cardiac output was determined by the direct Fick method. PVR was calculated during steady state at ventilation with room air, during FIO(2) of 0.65, then during additional intravenous infusion of L-Arg (15 mg. kg(-1). min(-1)) and Sub-P (1 pmol. kg(-1). min(-1)), and finally during inhalation of NO (20 ppm). In preoperative patients, the lack of an additional significant change of PVR with L-Arg, Sub-P, and inhaled NO suggests little preexisting PED. Postoperative PVR was higher, with an additional pulmonary endothelial contribution that was restorable with L-Arg and Sub-P. CONCLUSIONS: Postoperatively, the rise in PVR suggested PED, which was restorable by L-Arg and Sub-P, with no additional effect of inhaled NO. These results may indicate important new treatment strategies for these patients.

Is banding of the pulmonary trunk obsolete for infants with tricuspid atresia and double inlet ventricle with a discordant ventriculoarterial connection? Role of aortic arch obstruction and subaortic stenosis.

Banding the pulmonary trunk may exacerbate or promote the development of subaortic stenosis in patients with double inlet ventricle or tricuspid atresia with a dominant left ventricle and discordant ventriculoarterial connection and, therefore, may be an inappropriate palliative procedure for such patients. To examine this possibility, 102 consecutive infants were studied who presented with this anatomy between 1972 and 1987. Obstruction of the aortic arch was present in 52 patients. In 28 patients (17 with aortic arch obstruction), subaortic stenosis was already apparent at presentation. Of the remaining 74 patients, 19 received no palliative surgery and 55 underwent banding of the pulmonary trunk either with (n = 22) or without (n = 33) aortic arch repair. Outcome was significantly worse in patients with associated aortic arch obstruction. All such patients either died or developed subaortic stenosis by 3 years of age (survival free of subaortic stenosis 0 of 22 versus 22 of 33 for patients with isolated banding of the pulmonary trunk, p less than 0.001). After isolated banding, there was a lower ratio of the ventricular septal defect to ascending aorta diameters at presentation in the patients who developed subaortic stenosis than in the patients who did not (0.60 +/- 0.08 versus 1.03 +/- 0.15, p less than 0.001). Of the latter, 18 (95%) of 19 patients fulfilled criteria for a Fontan procedure at recatheterization. Thus, the presence of aortic arch obstruction is associated with rapid development of subaortic stenosis after banding of the pulmonary trunk. Alternative initial surgery, even though high risk, may be indicated. In the absence of such obstruction, banding the pulmonary trunk can be performed at reasonable risk and, provided that the ventricular septal defect is of adequate size, satisfactorily prepares most patients for a later Fontan procedure.

Blood pressure in four and five-year-old children: the effects of environment and other factors in it's measurement--the Brompton study.

Systolic blood pressure was measured on a total of 1855 occasions in 1307 children aged four and five years, and compared with values obtained since birth in the same children. There was a rapid rise in blood pressure in the first month of life. The mean blood pressure then only rose from 93 mmHg at six months to 98 mmHg at five years. The 95th percentile was 113 to 114 mmHg over this period. In children aged four and five years, over the ranges studied, blood pressure was not importantly affected by place of measurement, time of day, time since previous meal, or ambient temperature. However, blood pressure was approximately 1.6 mmHg higher in winter than in summer (P less than 0.01). Nevertheless, it is unlikely that these factors are of significance when making clinical measurements. Blood pressure was correlated with weight at all ages. Between the ages of four and five years, the index, weight/height 1.70 was the best function of adiposity tested that was independent of age between four and five years. It is suggested that this or the Quetelet Index (weight/height2) are suitable indices for adjusting blood pressure for body build in children aged four and five years.

Systolic blood pressure in a population of infants in the first year of life: the Brompton study.

Systolic blood pressure (BP) was measured by the Doppler technique and random zero sphygmomanometer in a sample of infants between the ages of 4 days (n = 1,740) and 1 year (n = 1,338). Mean systolic BP rose from 76 mm Hg at age 4 days to 96 mm Hg at age 6 weeks in babies awake, and showed little further variation at 6 months and 1 year. The BP was approximately 6 mm Hg higher in babies awake than asleep. Blood pressure was nearly normally distributed at all ages and the 95th percentile for BP of babies awake was 95 mm Hg at 4 days and 113 mm Hg between 6 weeks and 1 year. A comparison of intra-arterial and Doppler BP confirmed that both inflation bag length and cuff width are important for accurate measurement.

Frequency and outcome of disorders of cardiac rhythm and conduction in a population of newborn infants.

Of 3,383 apparently healthy newborn infants studied in one region of England, 33 showed cardiac arrhythmias or preexcitation on a standard ECG. Twenty-six infants had multiple atrial or ventricular premature beats. One infant had atrial flutter, two had multifocal atrial tachycardia, and two had supraventricular tachycardia. Two infants showed the ECG pattern of the Wolff-Parkinson-White syndrome, but did not demonstrate cardiac arrhythmias. Performance of 24-hour ECG recordings of 15 of the 26 subjects with premature beats showed additional supraventricular or ventricular tachycardia in five cases. In one of two infants with multifocal atrial tachycardia, 24-hour recordings also demonstrated long episodes of rapid supraventricular tachycardia. Follow-up recordings (standard and 24-hour ECG) showed that in the majority of infants arrhythmias or preexcitation could not be demonstrated after 12 weeks of age. One infant with atrial premature beats, one infant with ventricular premature beats and ventricular tachycardia, and one infant with multifocal atrial tachycardia showed persisting arrhythmias. Four infants received antiarrhythmic therapy which probably influenced the natural history of their arrhythmias. Disorders of cardiac rhythm and conduction occur in apparently healthy infants. This study provides a basis for further prospective research into their natural history.

Sequential 22-hour profiles of breathing patterns and heart rate in 110 full-term infants during their first 6 months of life.

Sequential recordings (total number 365, mean duration 22 hours) of ECG and abdominal wall movement were obtained from 110 full-term infants up to 6 months of age. The longest pause in breathing movement per recording (maximum 21.6 seconds) decreased in duration over the first 2 weeks of life (P less than .005). Pauses greater than 18.0 seconds were not detected after seven days. The spread of values for pauses greater than or equal to 3.6 seconds duration was widest during the first 2 weeks, and their number decreased with age (P less than .001). Periodic breathing, detected in 69% to 80% of infants in all age groups, showed decreasing trends with age in total duration and maximum length of episode (P less than .005 for both). The spread of values was widest during the first 2 weeks (range for total duration 0 to 4.7 hours) and decreased with age. The mean respiratory rate during regular breathing decreased after 4 weeks (P less than .001). The spread of values was widest during the first 2 weeks and decreased with age. Birth weight was positively correlated with mean respiratory rate during the first three days of life (r = +.64, P less than .001). The mean heart rate during regular breathing increased during the first 15 days (P less than .001) and then decreased after 4 weeks (P less than .001). Higher mean heart rates were found in male infants (P less than .01).

Prolonged apnea and cardiac arrhythmias in infants discharged from neonatal intensive care units: failure to predict an increased risk for sudden infant death syndrome.

Twenty-four hour recordings of respiratory wave form and ECG were made on low-birth-weight and/or premature infants within one week of discharge from eight neonatal intensive care units. Eight infants (0.7%) had episodes of apnea greater than 30 seconds in duration, all of which were accompanied by bradycardia less than 100 beats per minute; 25 infants (2.3%) had a total of 36 apneic episodes between 20 and 30 seconds in duration, 29 of which were accompanied by bradycardia less than or equal to 100 beats per minute; and 19 infants (1.7%) had episodes of bradycardia less than or equal to 50 beats per minute without prolonged apnea (as shown by a lack of breathing movement). Five infants had ventricular premature beats (including one with ventricular tachycardia). Eleven infants had supraventricular premature beats (including two with supraventricular tachycardia and one with preexcitation). Four infants had both supraventricular and ventricular premature beats. Two infants had preexcitation. Eleven infants who underwent 24-hour recordings died. Five infants were victims of sudden infant death syndrome. One infant death was sudden and unexpected and was attributed to bronchopneumonia. Two deaths were associated with congenital heart disease and three were associated with major cerebral disorders. None of the six babies who died suddenly and unexpectedly had apnea greater than or equal to 20 seconds, bradycardia less than or equal to 50 beats per minute, or cardiac arrhythmias on their 24-hour recordings.

Use of flecainide acetate for refractory junctional tachycardias in children with the Wolff-Parkinson-White syndrome.

The response of sustained supraventricular tachycardia to intravenous and oral flecainide acetate was investigated in 5 children, aged 5.5 to 11.5 years, who had tachycardias associated with Wolff-Parkinson-White syndrome. All children had failed to respond to at least 2 conventional agents. The effect of flecainide was studied using intracardiac techniques. Intravenous flecainide terminated tachycardia in all 5 patients. After drug infusion, slow, sustained tachycardia could be initiated in 1 patient. With oral treatment, slow, sustained tachycardia was started in 2 children and nonsustained in 2. One child had no inducible tachycardias. In 4 of 5 patients, long-term treatment has reduced the frequency of episodes and the drug is well tolerated. Thus, flecainide may be used to terminate and suppress junctional tachycardias in children who have failed to respond to conventional therapy.

Univentricular heart of right ventricular type: clinical, angiographic and electocardiographic features.

The sequential chamber analysis and angiographic, hemodynamic and clinical features, including those seen on electrocardiography, are presented in 18 patients with a univentricular heart of right ventricular type. In all cases a main chamber with a right ventricular trabecular pattern received both (or the sole) atrioventricular (A-V) connections, and a rudimentary chamber with a left ventricular trabecular pattern receiving no A-V connection was also present. The 18 patients were seen between 1971 and 1979; 6 died and the diagnosis was confirmed at autopsy. Twelve patients had double inlet, four absent right and two absent left A-V connections. Electrocardiography revealed right ventricular hypertrophy in all, and 11 had a superior mean frontal QRS axis. Chest roentgenography disclosed dextrocardia in 7 and levocardia in 11. Of 12 infants presenting with a marked decrease in pulmonary flow, 10 presented as ill neonates with hypoxia and acidosis. Four infants with increased pulmonary flow presented before 8 weeks with heart failure and mild cyanosis. Two with moderate pulmonary stenosis had mild cyanosis, but were asymptomatic. All had associated cardiac anomalies, pulmonary stenosis and persistent ductus arteriosus being the most common. For accurate diagnosis, angiographic demonstration of a posterior rudimentary chamber in addition to the main chamber of right ventricular trabecular pattern is mandatory.

Exercise echocardiography in children with Kawasaki disease: ventricular long axis is selectively abnormal.

In asymptomatic children with Kawasaki disease, left ventricular traditional markers of systolic and diastolic function are maintained. However, long-axis function, which represents the subendocardium, is abnormal during stress, particularly in patients with versus without coronary aneurysm.

Surgical risk factors in total anomalous pulmonary venous connection.

Eighty-three patients underwent surgical correction of total anomalous pulmonary venous connection (TAPVC) between 1973 and 1986. There were 46 boys and 37 girls. Median age at operation was 60 days (1 to 240) and median weight 3.9 kg (1 to 22). The anatomic types encountered included infracardiac connection (16 patients), supracardiac connection (32) and pulmonary venous drainage connected directly to the coronary sinus (27). Mixed anomalous drainage or pulmonary venous return connected directly to the right atrium occurred in 8 patients. Diagnosis was established by cardiac catheterization and angiography (56 patients), clinical examination (3) and cross-sectional echocardiography alone in 24 of the last consecutive 28 patients. Pulmonary hypertension was present in 26 (55%) of those who underwent cardiac catheterization. The median pulmonary vascular resistance was 4.2 units/m2 (body surface area) for all the patients, whereas in those with infracardiac pulmonary venous connection the median value was 10 units/m2. The median interval between admission and operation was 72 hours. Surgical correction was performed using profound hypothermia and circulatory arrest in 68; for the remainder, conventional cardiopulmonary bypass with profound to moderate hypothermia was used. Ten patients developed 1 or more pulmonary hypertensive crises during the early postoperative period. These were diagnosed in 8 by direct pulmonary artery pressure measurement and in 2 by clinical examination. Late reoperation was necessary in 6 patients (10%). Analyses of risk factors for 30-day survival for all patients showed that the risk of early death was associated with the type of anomaly (infradiaphragmatic), occurrence of pulmonary hypertensive crises, year of the operation, set of the patient and pressure of preoperative congestive heart failure.(ABSTRACT TRUNCATED AT 250 WORDS)

Subclavian flap angioplasty: does the arch look after itself?

OBJECTIVES: We sought to assess the early and long-term results of subclavian flap angioplasty in neonates and infants, with particular attention to growth of the hypoplastic arch. METHODS: A retrospective analysis of 185 consecutive patients who underwent subclavian flap angioplasty between 1974 and 1998 was carried out. The patients included 125 neonates and 60 infants, with a median age of 18 days. Sixty-six (36%) patients had an additional ventricular septal defect, 41 (22%) patients had aortic arch hypoplasia diagnosed preoperatively, 141 (76%) had an associated patent ductus arteriosus, and 41 (22%) had additional complex heart disease. Follow-up was with transthoracic Doppler echocardiography in all patients. RESULTS: The early mortality was 3%. Recoarctation, defined as a Doppler gradient of 25 mm Hg or more, occurred in 11 (6%) patients at a median follow-up of 6.2 years (6.2 +/- 4.6 years). This included 4 of the 41 patients in whom arch hypoplasia was diagnosed preoperatively. There were no complications with the left arm. By multivariate analysis, risk factors for death were determined to be residual arch hypoplasia and low birth weight. The only risk factor for recoarctation was persistent arch hypoplasia after surgical treatment. However, angiographic imaging of the aorta showed that recoarctation was not due to a hypoplastic transverse arch, and it was probably at the site of ductal tissue. Survival at 5 and 10 years was 98% and 96%, respectively. Freedom from reoperation for recoarctation at 2 years was 95%, and at 5, 10, and 15 years, it was 92%. CONCLUSIONS: Subclavian flap repair remains an effective technique for repair of aortic coarctation with excellent results and low mortality. In the majority of patients, arch hypoplasia regresses after this procedure.

Bedside assessment of phrenic nerve function in infants and children.

Phrenic nerve damage is widely recognized after cardiac operations and is associated with an increased morbidity and mortality. Retrospective studies in children have estimated a prevalence of phrenic damage of between 0.5% and 2.2%, but all these studies have limited their investigations to patients who have clinical problems postoperatively, and they have all used clinical or radiologic methods to diagnose the damage. These methods are indirect and hard to evaluate, making the accurate early diagnosis of phrenic nerve damage very difficult. Measurement of phrenic nerve conduction time (phrenic latency) allows direct evaluation of phrenic function and integrity, making it potentially much more specific than indirect methods. Using a simplified method of direct phrenic nerve stimulation, we have developed a method of measuring phrenic latency at the bedside in infants and children. A 1 Hz electrical stimulus is applied over the phrenic nerve in the neck and a diaphragmatic electromyogram from the seventh and eighth intercostal spaces is displayed on a storage oscilloscope. Phrenic latency can be measured directly from the screen. Using these methods we have now studied 37 children (aged 2 days to 15 years) before and after cardiac operations. Mean phrenic latency was 5.4 +/- 1.0 msec on the right and 5.4 +/- 1.0 msec on the left. Prolongation of phrenic latency by more than 2 msec was found in seven of 66 postoperative measurements (10.6%). These patients had a substantially worse postoperative course. We believe this technique to be an important contribution to the diagnosis of postoperative phrenic nerve damage, and one that can help in clinical practice to elucidate an important cause of postoperative morbidity and mortality.

Surgical correction in mirror-image atrial arrangement or left atrial isomerism with systemic venous return to the left-sided atrium.

We have recently encountered three patients with the hemodynamics of complete transposition accompanied by unusual atrial arrangements. The flow patterns produced all required correction by means of a "mirror-image" Mustard operation. Two of the patients had left atrial isomerism as evidenced by the finding of bilateral atrial appendages of morphologically left type. In both patients the systemic venous return was to the left-sided atrium and the pulmonary venous return to the right-sided atrium. The major effect of the isomerism was to reduce the volume of the left-sided chamber. This initially produced concern, since we now try to avoid enlargement of the pulmonary venous atrium when performing Mustard's operation. However, despite the small volume, it was possible to place the baffle without producing postoperative pulmonary or systemic venous obstruction. The other significant effect of left atrial isomerism is to deviate the position of the sinus node, but thus far we have not encountered any postoperative rhythm problems. The third patient had a mirror-image arrangement of the atrial chambers and the venous connections (situs inversus), and the anatomy was corrected by straightforward mirror-image reversal of our current technique for Mustard's procedure.

Anatomic correction for complete transposition and double-outlet right ventricle.

Between February, 1981, and December, 1984, 30 patients underwent anatomic correction of transposition of the great arteries with intact ventricular septum (n = 8), transposition with ventricular septal defect (n = 15), and double-outlet right ventricle with subpulmonary ventricular septal defect, the Taussig-Bing anomaly (n = 7). At operation, ages ranged from 18 hours to 6 years (mean 11.3 months) and weights ranged from 2.6 to 16.4 kg (mean 6.1 kg). The group with transposition and intact ventricular septum on average was younger (mean 1.2 months) and smaller (mean 3.5 kg) than the other two groups. Associated congenital heart defects were seen in 12 patients, including five with coarctation, three with multiple ventricular septal defects, two with right ventricular hypoplasia, two with juxtaposed atrial appendages, and one each with interrupted aortic arch, Wolff-Parkinson-White syndrome, and left ventricular outflow tract obstruction. All 10 patients who had undergone prior palliative operations had pulmonary artery banding. In addition, four of these patients had coarctation repairs, four had atrial septectomy, and one had systemic/pulmonary shunting. All recognized patterns of coronary anatomy were encountered. The aorta and pulmonary artery were side by side in 14 patients and anteroposterior in 16 patients. The Lecompte maneuver to establish right ventricular-pulmonary arterial continuity was successfully used in 12 of 13 patients with anteroposterior great vessels but in none of those with side-by-side arteries. Seven patients had subvalvular right ventricular outflow tract obstruction, recognized either at operation (five) or postoperatively (two). This was responsible for death in three patients. The 30 day hospital mortalities were as follows: one death (12.5%) in the group with transposition and intact ventricular septum, six deaths (40%) in the group with transposition plus ventricular septal defect, and one death (14.3%) among patients with double-outlet right ventricle and subpulmonary ventricular septal defect; the overall mortality was eight deaths (26.7%). There have been no late deaths (mean follow-up 17.2 months). Ninety-five percent of the survivors are in New York Heart Association Functional Class I. Postoperative catheterization in 13 patients has shown normal left ventricular function, no coronary stenosis, and no aortic incompetence. Sixty-nine percent of these patients had clinically unsuspected gradients across the right ventricular outflow tract, which may be prevented by avoiding the Lecompte maneuver or the use of conduits.

Options for surgical repair in hearts with univentricular atrioventricular connection and subaortic stenosis.

Thirteen patients have undergone surgical treatment because of subaortic obstruction in hearts with a univentricular atrioventricular connection. Nine patients underwent surgical enlargement of the ventricular septal defect and four patients had construction of an aortopulmonary anastomosis and closure of the pulmonary trunk (the Damus-Kaye-Stansel procedure). Two patients undergoing enlargement of the septal defect and two having the Damus-Kaye-Stansel procedure also had a modified Fontan procedure. One patient had complete atrioventricular dissociation after direct enlargement of the ventricular septal defect, which necessitated insertion of an epicardial pacemaker. One patient died within 30 days of the operation after enlargement of the defect and two patients after the Damus-Kaye-Stansel procedure. There was one late death, occurring in a patient who underwent enlargement of the ventricular septal defect. Ten patients have subsequently undergone conventional cardiac catheterization and angiography or transcutaneous Doppler flow studies to assess the relief of the subaortic obstruction. The result has been satisfactory in all. Because of this experience, we now recommend direct surgical enlargement of the restrictive ventricular septal defect for direct relief of subaortic stenosis occurring with a univentricular atrioventricular connection to a dominant left ventricle, inasmuch as it appears to be hemodynamically effective with a low operative mortality and morbidity. The Damus-Kaye-Stansel procedure can also have a role in relieving subaortic stenosis when the atria are connected to a dominant right ventricle.

Does the modified Blalock-Taussig shunt cause growth of the contralateral pulmonary artery?

BACKGROUND: Although some pediatric cardiology departments have a policy of adopting primary correction of tetralogy of Fallot in all symptomatic infants, we and others still palliate neonates and infants. Effective palliation should ameliorate symptoms and allow growth of the pulmonary arteries. Although studies on the growth of the ipsilateral and contralateral pulmonary arteries after a classic Blalock-Taussig shunt have been reported, pulmonary artery growth after a modified Blalock-Taussig shunt has not been studied as thoroughly. Therefore, we examined whether there is equal growth of the contralateral pulmonary artery after a modified Blalock-Taussig shunt. METHODS: We retrospectively analyzed the records of 140 patients with symptomatic tetralogy of Fallot who had a modified Blalock-Taussig shunt between October 1985 and October 1995. The median age at the time of the Blalock-Taussig shunt was 1.6 months. All patients had corrective procedures at a median age of 1.7 years. Cineangiography was done before the corrective procedure. From the angiograms the diameter of the right and left pulmonary arteries before their first lobar branches and the diameter of the descending thoracic aorta at the level of the diaphragm were measured. For each patient the ratios of right pulmonary artery to descending thoracic aorta and left pulmonary artery to descending thoracic aorta were determined and compared using Student's t test. RESULTS: Of the 140 patients, 114 had a left-sided Blalock-Taussig shunt, 20 had a right-sided shunt, and 6 patients had bilateral shunts. The mean right pulmonary artery to descending thoracic aorta ratio was 1.10 and the mean left pulmonary artery to descending thoracic aorta ratio was 0.98. This difference was not significant. CONCLUSION: We showed equal growth of the right and left pulmonary arteries with no distortion after a modified Blalock-Taussig shunt. If palliation is considered, the modified Blalock-Taussig shunt remains our choice.

Anatomic correction for post-mustard pulmonary venous obstruction.

Pulmonary artery banding and arterial switch operation as a two-stage approach have offered a surgical solution to the failing right ventricle after intraatrial repair of the transposition of the great arteries. Banding of the pulmonary artery increases the resistance to the left ventricular output and, therefore, "prepares" the left ventricle for supporting the systemic circulation and arterial switch operation. We report a case of anatomic correction for transposition of the great arteries after intraatrial repair (Mustard's operation) in which the left ventricle was retrained serendipitously as a consequence of pulmonary venous obstruction complicating Mustard's procedure.

Thromboembolism after the Fontan procedure and its modifications.

The frequency of thromboembolism after the Fontan procedure has not been specifically addressed. Seventy patients underwent this operation between 1987 and 1992, and complete information was available on 64. Forty-five had a right atrium-pulmonary artery anastomosis, 17 had a total cavopulmonary connection, and 2 had a right atrium-right ventricle anastomosis. No patient received anticoagulants for more than 8 months postoperatively. Seven patients died in the early postoperative period (< 30 days). Ten cases of thromboembolism were identified in the follow-up, which ranged from 6 to 55 months (mean follow-up, 24.5 +/- 17 months). They occurred 7 days to 4.5 years after the Fontan repair. Seven thombi were on the systemic venous or pulmonary aspect of the circulation and three, on the systemic arterial aspect. Freedom from thromboembolism was 75% at 4 years. We conclude that thromboembolism is common after the Fontan procedure; it can occur several days to years postoperatively; it is an important cause of morbidity; and it may partly explain late death after the Fontan operation. Our data suggest that all patients undergoing this procedure require long-term anticoagulant therapy.