- Email: [email protected]
Improving outcomes of the Fontan operation in children with atrial isomerism and heterotaxy syndromes
Improving Outcomes of the Fontan Operation in Children With Atrial Isomerism and Heterotaxy Syndromes Anthony Azakie, MD, Sandra L. Merklinger, MN, William G. Williams, MD, Glen S. Van Arsdell, MD, John G. Coles, MD, and Ian Adatia, MD Departments of Surgery, Pediatrics, and Critical Care Medicine, Divisions of Cardiovascular Surgery and Cardiology, The Hospital for Sick Children, University of Toronto School of Medicine, Toronto, Ontario, Canada
Background. The historic outcome of the Fontan procedure in children with single ventricle and heterotaxy syndrome has been poor, and in the current era it has been incompletely described. Methods. From January 1993 through April 2000, 30 patients (age range, 5.3 ⴞ 3.6 years) had total cavopulmonary connection for single ventricle and isomerism and heterotaxy syndrome. Right atrial isomerism and heterotaxy (n ⴝ 19) or left atrial isomerism and heterotaxy (n ⴝ 11) were associated with a morphologic right ventricle (n ⴝ 26), a common atrioventricular valve (n ⴝ 25), pulmonary atresia (n ⴝ 9) or stenosis (n ⴝ 17), anomalous pulmonary venous drainage (partial, n ⴝ 7; total, n ⴝ 11; obstructed, n ⴝ 4), and interrupted inferior vena cava (n ⴝ 13). A previous systemic to pulmonary artery shunt had been performed in 21 patients, and 3 patients had previous pulmonary artery banding. The Fontan procedure was staged in all but 3 patients at a median age of 12 months (range, 3 to 114 months). Before the Fontan procedure, atrial arrhythmia was present in 7 patients (23%). Results. An extracardiac conduit was constructed in 22 patients (median diameter, 20 mm; range, 16 to 27 mm), and a lateral tunnel was constructed in 4 patients. Asso-
ciated procedures included pulmonary artery augmentation (n ⴝ 17), repair of anomalous pulmonary venous drainage (n ⴝ 7), and repair of a regurgitant atrioventricular valve (n ⴝ 2). The mean cardiopulmonary bypass time was 134 ⴞ 57 minutes. Cardioplegic cardiac arrest was used in 12 patients for a mean duration of 62 ⴞ 28 minutes. There were four hospital deaths (13%), three that occurred in children having concomitant repair of anomalous pulmonary venous drainage (previously undetected in one). Thirteen children (41%) developed early postoperative atrial arrhythmias, 11 of the 13 children required temporary pacing. There was one late death. Follow-up (3.4 ⴞ 2.7 years) was available on 24 of 25 survivors. All but 2 patients were in sinus rhythm. Conclusions. Early and midterm outcomes after the Fontan procedure in this patient group improved and may be further ameliorated by detecting and repairing associated obstruction to pulmonary venous flow and performing atrioventricular valvuloplasty before the Fontan procedure.
C
Fontan procedure in children with single ventricle and visceroatrial heterotaxia are limited. During the staged palliation of children with single ventricle, the presence of heterotaxy syndrome has been shown to increase the risk of adverse outcomes after the Fontan procedure [8, 9]. Early mortality after the Fontan procedure in this patient group was reported to be as high as 43% to 80% [10, 11], but also decreased dramatically over time [12]. We have reported our institutional experience with all children presenting with either left or right atrial ismoerism and heterotaxy syndromes to include management and outcomes for those with normal hearts or those requiring either biventricular repair or staged univentricular palliation [1, 2]. The purpose of the present study is to determine the early and midterm outcomes for total cavopulmonary connections (including hepatic veins) in children with single ventricle and isomerism and heterotaxy syndromes in the current era.
hildren with functional single ventricle and atrial isomerism and visceral heterotaxy syndromes are a challenging group to manage. Thus far the majority of the literature dedicated to this clinical entity has focused on the outcomes of initial palliation. The early clinical course for children who are surgically palliated for left atrial isomerism and single ventricle is characterized by a high mortality and unfavorable midterm survival [1]. Children with single ventricle physiology and right atrial isomerism have an even worse outlook, especially in the presence of total anomalous pulmonary venous drainage with obstruction [2–7]. Detailed reports on the midterm outcome of the Accepted for publication June 20, 2001. Address reprint requests to Dr Adatia, Department of Cardiology and Critical Care Medicine, The Hospital for Sick Children, 555 University Ave, Toronto, Ontario M5G-1X8, Canada.
© 2001 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 2001;72:1636 – 40) © 2001 by The Society of Thoracic Surgeons
0003-4975/01/$20.00 PII S0003-4975(01)03039-9
Ann Thorac Surg 2001;72:1636-40
Patients and Methods Definitions The diagnosis of left or right atrial isomerism with visceral heterotaxia was made according to previously described criteria [1, 2].
Patients Permission to perform the Health Record review was obtained from the Research Ethics Board, the Hospital for Sick Children, Toronto, Ontario, Canada. The cardiovascular surgery and cardiology databases at the Toronto Hospital for Sick Children were reviewed for all patients having a Fontan operation or patients with a diagnosis of visceroatrial heterotaxia, or both, in the current era of low mortality for the Fontan operation at this hospital as previously defined [13]. From January 1993 through April 2000, among the 212 patients having a modified Fontan operation, 30 patients (14%) had total cavopulmonary connection (including hepatic veins) for single ventricle and isomerism and heterotaxy syndrome. Their records were reviewed for demographic, operative and perioperative features, as well as midterm follow-up echocardiography, Holter electrocardiogram, catheterization, and clinical data (from October 2000 through January 2001). There were 11 males and 19 females with a median age of 4.1 years (range, 23 months to 14.2 years) and a median weight of 14 kg (range, 8 to 37 kg) at the time of the Fontan procedure. Right atrial isomerism and heterotaxy (n ⫽ 19) or left atrial isomerism and heterotaxy (n ⫽ 11) was associated with double outlet right ventricle in 19 patients, common atrioventricular valve in 25 patients, pulmonary atresia in 9 patients, or stenosis in 17 patients, anomalous pulmonary venous drainage in 18 patients (partial, n ⫽ 7, or total, n ⫽ 11; obstructed, n ⫽ 4), interrupted inferior vena cava in 13 patients (azygous continuation, n ⫽ 10; hemiazygous, n ⫽ 3), and absent coronary sinus in 5 patients. A dominant morphologic right ventricle was present in 26 children. The mean preoperative pulmonary artery pressure was 11 ⫾ 3 mm Hg with an average room air oxygen saturation of 81% ⫾ 5%. Preoperatively, 3 patients had qualitatively reduced ventricular function. The mean McGoon index was 2.01 ⫾ 0.54 and Nakata index was 312 ⫾ 169. Before the Fontan procedure, moderate to severe atrioventricular valve regurgitation was present in 3 patients (10%), and atrial arrhythmia was present in 7 patients (23%; junctional rhythm ⫽ 3, supraventricular tachycardia ⫽ 2, pacemaker for atrioventricular block ⫽ 1, wandering atrial pacemaker ⫽ 1).
Previous Procedures A previous systemic to pulmonary artery shunt was performed in 21 patients, and 3 patients had previous pulmonary artery banding. The Fontan procedure was staged in all but 3 children with either a bidirectional cavopulmonary anastomosis (n ⫽ 9), or Kawashima anastomosis (n ⫽ 12), or bilateral bidirectional cavopulmonary anastomosis (n ⫽ 7), or a hemi-Fontan (n ⫽ 1) at a median age of 12 months (range, 3 to 114 months). Other
AZAKIE ET AL FONTAN OUTCOMES FOR HETEROTAXY SYNDROMES
1637
previous procedures included patch augmentation of the pulmonary arteries (n ⫽ 15), resection of subaortic obstruction, repair of a regurgitant atrioventricular valve (n ⫽ 2), and division of a main pulmonary artery (n ⫽ 13).
Operative Approach Standard cardiopulmonary bypass techniques were used to perform the Fontan operation with ascending aortic and bicaval venous cannulation. In 22 patients (73%), an extracardiac conduit (20 ⫾ 2 mm; range, 16 to 27 mm) was used to construct the Fontan circuit. Either Gore-Tex (W. L. Gore and Associates, Flagstaff, AZ) (n ⫽ 12) or aortic homograft tissue (n ⫽ 10) was used for the reconstruction. In 12 patients (43%) cardioplegic cardiac arrest was used for a mean duration of 64 ⫾ 31 minutes. A fenestration between the Fontan circuit and neopulmonary venous atrium with a median size of 4 mm (range, 3 to 5 mm) was created in 20 patients (67%). Associated procedures included pulmonary artery augmentation (n ⫽ 17), repair of anomalous pulmonary venous drainage (n ⫽ 7), repair of a regurgitant atrioventricular valve (n ⫽ 2), bidirectional cavopulmonary anastomosis (n ⫽ 3), and resection of subaortic stenosis (n ⫽ 5).
Statistical Methods Data are described as frequencies, medians with ranges, and means with standard deviations. For missing data, the number of non-missed values is given. Simple (univariable) logistic regression analysis was performed to identify predictors of death and the development of early postoperative atrial arrhythmias. Variables assessed as potential predictors of death include date of operation, age, gender, weight, preoperative McGoon and Nakata indexes, preoperative pulmonary artery pressures, age at superior cavopulmonary anastomosis, right or left atrial isomerism, presence of a common atrioventricular valve, pulmonary atresia or stenosis, anomalous pulmonary venous drainage, presence of noncardiac anomalies, and type of Fontan connection. Variables assessed as potential predictors of the development of early postoperative arrhythmias included presence of preoperative arrhythmias, right or left atrial isomerism, and type of Fontan connection. Time-related survival estimates were calculated using the Kaplan-Meier method. All analyses were performed using SAS (SAS Institute, Inc, Cary, NC) version 8. A p value less than 0.05 was set as the level of statistical significance.
Results Postoperative Morbidity and Mortality There were four hospital deaths (13%) all caused by ventricular failure; two of these deaths had takedown of the Fontan connection. Three of the 4 children that died had concomitant repair of anomalous pulmonary venous drainage, including 1 child who also had concomitant repair of a regurgitant atrioventricular valve. In another child, obstructed anomalous venous drainage was iden-
1638
AZAKIE ET AL FONTAN OUTCOMES FOR HETEROTAXY SYNDROMES
Ann Thorac Surg 2001;72:1636-40
in a child with left atrial isomerism. No children had developed protein-losing enteropathy at the time of follow-up. Twenty-four hour ambulatory electrocardiographic monitoring demonstrated predominantly sinus rhythm in all children except 2 (junctional rhythm). Echocardiographic evaluation (n ⫽ 25) showed qualitatively depressed ventricular function in 5 patients (20%). The remaining 20 children had qualitatively good ventricular function. Atrioventricular valve regurgitation was graded as moderate in 3 children, mild in 16, and absent in 7. Fig 1. Kaplan-Meier survival curve from the time of the Fontan operation in children with functional or morphologic single ventricle and visceroatrial heterotaxia. Numbers of patients at risk are given above the horizontal axis. Survival is expressed as a percentage with 95% confidence intervals.
tified after the completion of the Fontan connection by an intraoperative transesophageal echocardiogram. No perioperative risk factors for death by univariable logistic regression analysis were identified, probably because of the small sample size and low mortality. Thirteen children (41%) developed early postoperative atrial arrhythmias; 11 of these children required temporary pacing. In 9 children, sinus node dysfunction was present, characterized by relative bradycardia (or the presence of a junctional rhythm; atrioventricular conduction block developed in 2 patients and a wandering atrial pacemaker was present in 2 patients), or a combination thereof. Patients with left atrial isomerism were more likely to develop early postoperative arrhythmias (odds ratio, 3.5) than those with right atrial isomerism, although this did not reach statistical significance (p ⫽ 0.11). The presence of preoperative arrhythmias was also associated with an increased likelihood of early postoperative atrial rhythm disturbances (odds ratio, 6.0; p ⫽ 0.05). Type of Fontan connection was not associated with the development of atrial rhythm disturbances in the early postoperative period. The mean Fontan pressure in the early postoperative period was 13 ⫾ 3 mm Hg and the mean pulmonary venous atrial pressure was 6 ⫾ 2 mm Hg. The median duration of mechanical ventilatory support was 1 day (range, 6 hours to10 days), the median length of stay in the intensive care unit was 2 days (range, 1 to 10 days), and the median duration of hospital admission was 10 days (range, 6 to 19 days). The median duration of chest tube drainage was 9 days (range, 5 to 19 days).
Midterm Follow-up Follow-up was available on 25 survivors for up to 7.1 years. The mean follow-up time was 3.4 ⫾ 2.7 years. Kaplan-Meier survival (Fig 1) at 1 month is 86% (95% confidence interval: 75% to 97%), and at 1 and 5 years is 83% (95% confidence interval: 70% to 96%). There was one death after hospital discharge caused by anastomotic disruption of a direct connection between the hepatic veins and azygous continuation of the inferior vena cava
Comment Despite the generally reported poor outcomes for patients with left or right atrial isomerism with cardiac anatomy that is unsuitable for biventricular repair, there remains a group of patients who survive initial palliation of the neonatal period and require further palliation for limiting and progressive cyanosis. Cardiac tansplantation is one option, yet there are important disadvantages to this approach. In our recent experience with 30 children with visceroatrial heterotaxy and cardiac anatomy, which precludes biventricular repair, the hospital mortality for the modified Fontan procedure is 13%. Early postoperative rhythm disturbances are common. Children with left atrial isomerism are more likely to have congenital atrioventricular block, junctional rhythm, and hypoplasia of the sinoatrial node, whereas those with right atrial isomerism may have bilateral sinoatrial and atrioventricular nodes. In our experience, the development of postoperative atrial dysrhythmias is more likely in children with preoperative rhythm disturbances or a diagnosis of left atrial isomerism, or both. At midterm follow-up, survival is 83%, ventricular function and atrioventricular valve competency are good, and early rhythm disturbances resolve to base line. We have recently reported cohort analyses of all newborns with left [1] or right [2] atrial isomerism and visceral heterotaxy syndromes. In infants with left atrial isomerism, single ventricle primary palliation has been associated with an early mortality rate of 25% and a 5-year survival of 61%. Risk factors for death included congenital atrioventricular block, coarctation of the aorta, single ventricle and the presence of biliary or other gastrointestinal malformations. More than 50% of children within the entire cohort had anomalous pulmonary venous drainage, which by univariate analysis entered as a risk factor for mortality. Our experience with children with single ventricle and right atrial isomerism has also been reported as a cohort analysis with a focus on primary palliation, presurgical outcomes, and biventricular repair. The overall mortality for staged palliation of children with right isomerism and single ventricle [2] was 62% (40 of 64). Independent risk factors for time-related mortality include absence of pulmonary outflow tract obstruction, presence of major atrioventricular valve anomaly, and obstructed pulmonary veins. In fact, the surgical outcomes for primary neonatal repair of anom-
Ann Thorac Surg 2001;72:1636-40
alous pulmonary venous drainage in the setting of single ventricle was 95% mortality. Others have reported comparable outcomes of infants with single ventricle and total anomalous pulmonary venous drainage, also with limited information on midterm outcomes after the modified Fontan procedure in children with isomeric hearts. Sadiq and colleagues [3] reported high mortality rates for early palliation in children with right atrial isomerism and single ventricle. Seven of their 20 patients progressed to a Fontan operation with five survivors. There were no late survivors among children with a history of obstructed pulmonary veins. Gaynor and colleagues [4] reported on the longterm outcomes of 73 infants with single ventricle and total anomalous pulmonary venous connection. Survival was 37% at 1 year and 23% at 5 years. By univariate analysis, time-related mortality was predicted by younger age and total anomalous pulmonary venous connection. Primary palliation was associated with a mortality rate that approximated 50%. Twenty-one of their patients underwent superior cavopulmonary connections with an early mortality rate of 38%. Early mortality after the Fontan operation was 42% (8 of 21 patients). Of the 73 patient cohorts only 9 had successful Fontan operations at the most recent follow-up. Histologic examination of postmortem specimens revealed marked dilatation of pulmonary veins with increased wall thickness for age. As pointed out by the authors, the venous abnormalities were present even in the absence of clinical evidence of pulmonary venous obstruction, suggesting that current diagnostic methods underestimate the prevalence and severity of obstruction. Detailed information regarding outcomes for the Fontan procedure in children with single ventricle and visceral heterotaxy syndromes (right and left atrial isomerism, or asplenia and polysplenia) is limited, including our two recent cohort analyses [1, 2] and reports from earlier eras [10, 14 –16]. Humes and colleagues [10] reported results for the modified Fontan operation in 49 patients with asplenia and polysplenia between 1975 and 1986. Hospital mortality was 43%, but improved to approximately 15% in a more recent era. Fontan mortality was associated with right atrial isomerism (asplenia) and need for concomitant atrioventricular valve repair. In the report from Marcelletti and colleagues [11], from a total of 12 children with single ventricle and right and left atrial isomerism, 5 children progressed to a Fontan procedure with only one survivor. Culbertson and colleagues [16] reported a 20% mortality rate for patients with heterotaxy syndrome undergoing the Fontan procedure. Preoperative findings of more than mild atrioventricular valve regurgitation significantly increased the risk of overall mortality. The purpose of the present study was to determine the early and midterm outcomes of total cavopulmonary connections in all children with heterotaxy syndrome and single ventricle at a single institution in the current era. The mortality for the Fontan procedure in patients with heterotaxy syndrome in our recent experience is 13%, which compares favorably with published mortality
AZAKIE ET AL FONTAN OUTCOMES FOR HETEROTAXY SYNDROMES
1639
rates of 20% to 80% in earlier eras [10, 11, 16] but remains higher than our institutional mortality of less than 5% for the Fontan procedure in all children with functional single ventricles. Although the current study only addresses the numerator by reviewing the outcomes for children with heterotaxy syndrome undergoing the Fontan procedure, several important observations can be made. The presence of clinically important atrioventricular valve regurgitation [10, 16] is associated with an adverse outcome after the Fontan procedure for children with functional single ventricles and isomeric hearts. In our experience, one of the 2 children undergoing concomitant atrioventricular valve repair died following the Fontan procedure. The inference from our experience and published reports is that early atrioventricular valve repair may protect the pulmonary vasculature and allow for preservation of ventricular function, and may allow for improved outcomes if performed before the Fontan procedure. Three of the 4 patients that died after the Fontan procedure also had concomitant repair of anomalously draining pulmonary veins. In 1 child, pulmonary venous obstruction was only detected after the Fontan procedure. As pointed out by Gaynor and colleagues [4], persistent or recurrent pulmonary venous obstruction is not uncommon in children with single ventricle and anomalous draining pulmonary veins, and pulmonary venous obstruction may not always be clinically evident as demonstrated by their histopathologic studies. In our experience and that reported by McElhinney and Reddy [17], pulmonary venous obstruction may be unmasked by the increased pulmonary blood flow that results after the Fontan operation. Thus, efforts to improve surgical outcomes in such patients require both close monitoring and a high index of suspicion for persistent, recurrent, or previously undetected pulmonary venous obstruction. Furthermore, repair before the Fontan procedure may protect the pulmonary vasculature and improve the outcomes that follow total cavopulmonary connection.
References 1. Gilljam T, McCrindle BW, Smallhorn JF, et al. Outcomes of left atrial isomerism over a 28-year period at a single institution. J Am Coll Cardiol 2000;36:908–16. 2. Hashmi A, Abu-Sulaiman R, Mcrindle BW, et al. Management and outcomes of right atrial isomerism: a 26-year experience. J Am Coll Cardiol 1998;31:1120– 6. 3. Sadiq M, Stumper O, De Giovanni JV, et al. Management and outcome of infants and children with right atrial isomerism. Heart 1996;75:314–9. 4. Gaynor JW, Collins MH, Rychik J, et al. Long-term outcome of infants with single ventricle, and total anomalous pulmonary venous connection. J Thorac Cardiovasc Surg 1999;117: 506–14. 5. Deleon SY, Gidding SS, Ilbawi MN, et al. Surgical management of infants with complex cardiac anomalies associated with reduced pulmonary blood flow and total anomalous pulmonary venous drainage. Ann Thorac Surg 1987;43: 207–11. 6. Di Donato R, di Carlo D, Squitieri C, et al. Palliation of cardiac malformations associated with right isomerism
1640
7. 8.
9.
10. 11. 12.
AZAKIE ET AL FONTAN OUTCOMES FOR HETEROTAXY SYNDROMES
(asplenia syndrome) in infancy. Ann Thorac Surg 1987;44: 35–9. Heinemann MK, Hanley FL, Van Praagh S, et al. Total anomalous pulmonary venous drainage in newborns with visceral heterotaxy. Ann Thorac Surg 1994;57:88–91. Gentles TL, Mayer JE Jr, Gavreau K, et al. Fontan operation in five hundred consecutive patients: factors influencing early and late outcome. J Thorac Cardiovasc Surg 1997;114: 376–91. Knott-Craig CJ, Danielson GK, Schaff HV, et al. The modified Fontan operation: an analysis of risk factors for early postoperative death or takedown in 702 consecutive patients from one institution. J Thorac Cardiovasc Surg 1995;109: 1237– 43. Humes RA, Feldt RH, Porter CJ, et al. The modified Fontan operation for asplenia and polysplenia syndromes. J Thorac Cardiovasc Surg 1988;96:212– 8. Marcelletti C, Di Donato R, Nijveld A, et al. Right and left isomerism: the cardiac surgeon’s view. Ann Thorac Surg 1983;35:400–5. Cetta F, Feldt RH, O’Leary PW, et al. Improved early
Ann Thorac Surg 2001;72:1636-40
13.
14.
15.
16.
17.
morbidity and mortality after Fontan operation: the Mayo Clinic experience, 1987 to 1992. J Am Coll Cardiol 1996;28: 480– 6. Van Arsdell G, McCrindle BW, Einarson KD, et al. Interventions associated with minimal Fontan mortality. Ann Thor Surg 2000;70:568–74. Julsrud PR, Danielson GK. A modification of the Fontan procedure incorporating anomalies of systemic and pulmonary venous return. J Thorac Cardiovasc Surg 1990;100: 233–9. Michielon G, Gharagozloo F, Julsrud PR, et al. Modified Fontan operation in the presence of anomalies of systemic and pulmonary venous connection. Circulation 1993;88: 141– 8. Culbertson CB, George BL, Day RW, et al. Factors influencing survival of patients with heterotaxy syndrome undergoing the Fontan procedure. J Am Coll Cardiol 1992;20:678– 84. McElhinney DB, Reddy VM. Anomalous pulmonary venous return in the staged palliation of functional univentricular heart defects. Ann Thorac Surg 1998;66:683–7.
Background. The historic outcome of the Fontan procedure in children with single ventricle and heterotaxy syndrome has been poor, and in the current era it has been incompletely described. Methods. From January 1993 through April 2000, 30 patients (age range, 5.3 ⴞ 3.6 years) had total cavopulmonary connection for single ventricle and isomerism and heterotaxy syndrome. Right atrial isomerism and heterotaxy (n ⴝ 19) or left atrial isomerism and heterotaxy (n ⴝ 11) were associated with a morphologic right ventricle (n ⴝ 26), a common atrioventricular valve (n ⴝ 25), pulmonary atresia (n ⴝ 9) or stenosis (n ⴝ 17), anomalous pulmonary venous drainage (partial, n ⴝ 7; total, n ⴝ 11; obstructed, n ⴝ 4), and interrupted inferior vena cava (n ⴝ 13). A previous systemic to pulmonary artery shunt had been performed in 21 patients, and 3 patients had previous pulmonary artery banding. The Fontan procedure was staged in all but 3 patients at a median age of 12 months (range, 3 to 114 months). Before the Fontan procedure, atrial arrhythmia was present in 7 patients (23%). Results. An extracardiac conduit was constructed in 22 patients (median diameter, 20 mm; range, 16 to 27 mm), and a lateral tunnel was constructed in 4 patients. Asso-
ciated procedures included pulmonary artery augmentation (n ⴝ 17), repair of anomalous pulmonary venous drainage (n ⴝ 7), and repair of a regurgitant atrioventricular valve (n ⴝ 2). The mean cardiopulmonary bypass time was 134 ⴞ 57 minutes. Cardioplegic cardiac arrest was used in 12 patients for a mean duration of 62 ⴞ 28 minutes. There were four hospital deaths (13%), three that occurred in children having concomitant repair of anomalous pulmonary venous drainage (previously undetected in one). Thirteen children (41%) developed early postoperative atrial arrhythmias, 11 of the 13 children required temporary pacing. There was one late death. Follow-up (3.4 ⴞ 2.7 years) was available on 24 of 25 survivors. All but 2 patients were in sinus rhythm. Conclusions. Early and midterm outcomes after the Fontan procedure in this patient group improved and may be further ameliorated by detecting and repairing associated obstruction to pulmonary venous flow and performing atrioventricular valvuloplasty before the Fontan procedure.
C
Fontan procedure in children with single ventricle and visceroatrial heterotaxia are limited. During the staged palliation of children with single ventricle, the presence of heterotaxy syndrome has been shown to increase the risk of adverse outcomes after the Fontan procedure [8, 9]. Early mortality after the Fontan procedure in this patient group was reported to be as high as 43% to 80% [10, 11], but also decreased dramatically over time [12]. We have reported our institutional experience with all children presenting with either left or right atrial ismoerism and heterotaxy syndromes to include management and outcomes for those with normal hearts or those requiring either biventricular repair or staged univentricular palliation [1, 2]. The purpose of the present study is to determine the early and midterm outcomes for total cavopulmonary connections (including hepatic veins) in children with single ventricle and isomerism and heterotaxy syndromes in the current era.
hildren with functional single ventricle and atrial isomerism and visceral heterotaxy syndromes are a challenging group to manage. Thus far the majority of the literature dedicated to this clinical entity has focused on the outcomes of initial palliation. The early clinical course for children who are surgically palliated for left atrial isomerism and single ventricle is characterized by a high mortality and unfavorable midterm survival [1]. Children with single ventricle physiology and right atrial isomerism have an even worse outlook, especially in the presence of total anomalous pulmonary venous drainage with obstruction [2–7]. Detailed reports on the midterm outcome of the Accepted for publication June 20, 2001. Address reprint requests to Dr Adatia, Department of Cardiology and Critical Care Medicine, The Hospital for Sick Children, 555 University Ave, Toronto, Ontario M5G-1X8, Canada.
© 2001 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 2001;72:1636 – 40) © 2001 by The Society of Thoracic Surgeons
0003-4975/01/$20.00 PII S0003-4975(01)03039-9
Ann Thorac Surg 2001;72:1636-40
Patients and Methods Definitions The diagnosis of left or right atrial isomerism with visceral heterotaxia was made according to previously described criteria [1, 2].
Patients Permission to perform the Health Record review was obtained from the Research Ethics Board, the Hospital for Sick Children, Toronto, Ontario, Canada. The cardiovascular surgery and cardiology databases at the Toronto Hospital for Sick Children were reviewed for all patients having a Fontan operation or patients with a diagnosis of visceroatrial heterotaxia, or both, in the current era of low mortality for the Fontan operation at this hospital as previously defined [13]. From January 1993 through April 2000, among the 212 patients having a modified Fontan operation, 30 patients (14%) had total cavopulmonary connection (including hepatic veins) for single ventricle and isomerism and heterotaxy syndrome. Their records were reviewed for demographic, operative and perioperative features, as well as midterm follow-up echocardiography, Holter electrocardiogram, catheterization, and clinical data (from October 2000 through January 2001). There were 11 males and 19 females with a median age of 4.1 years (range, 23 months to 14.2 years) and a median weight of 14 kg (range, 8 to 37 kg) at the time of the Fontan procedure. Right atrial isomerism and heterotaxy (n ⫽ 19) or left atrial isomerism and heterotaxy (n ⫽ 11) was associated with double outlet right ventricle in 19 patients, common atrioventricular valve in 25 patients, pulmonary atresia in 9 patients, or stenosis in 17 patients, anomalous pulmonary venous drainage in 18 patients (partial, n ⫽ 7, or total, n ⫽ 11; obstructed, n ⫽ 4), interrupted inferior vena cava in 13 patients (azygous continuation, n ⫽ 10; hemiazygous, n ⫽ 3), and absent coronary sinus in 5 patients. A dominant morphologic right ventricle was present in 26 children. The mean preoperative pulmonary artery pressure was 11 ⫾ 3 mm Hg with an average room air oxygen saturation of 81% ⫾ 5%. Preoperatively, 3 patients had qualitatively reduced ventricular function. The mean McGoon index was 2.01 ⫾ 0.54 and Nakata index was 312 ⫾ 169. Before the Fontan procedure, moderate to severe atrioventricular valve regurgitation was present in 3 patients (10%), and atrial arrhythmia was present in 7 patients (23%; junctional rhythm ⫽ 3, supraventricular tachycardia ⫽ 2, pacemaker for atrioventricular block ⫽ 1, wandering atrial pacemaker ⫽ 1).
Previous Procedures A previous systemic to pulmonary artery shunt was performed in 21 patients, and 3 patients had previous pulmonary artery banding. The Fontan procedure was staged in all but 3 children with either a bidirectional cavopulmonary anastomosis (n ⫽ 9), or Kawashima anastomosis (n ⫽ 12), or bilateral bidirectional cavopulmonary anastomosis (n ⫽ 7), or a hemi-Fontan (n ⫽ 1) at a median age of 12 months (range, 3 to 114 months). Other
AZAKIE ET AL FONTAN OUTCOMES FOR HETEROTAXY SYNDROMES
1637
previous procedures included patch augmentation of the pulmonary arteries (n ⫽ 15), resection of subaortic obstruction, repair of a regurgitant atrioventricular valve (n ⫽ 2), and division of a main pulmonary artery (n ⫽ 13).
Operative Approach Standard cardiopulmonary bypass techniques were used to perform the Fontan operation with ascending aortic and bicaval venous cannulation. In 22 patients (73%), an extracardiac conduit (20 ⫾ 2 mm; range, 16 to 27 mm) was used to construct the Fontan circuit. Either Gore-Tex (W. L. Gore and Associates, Flagstaff, AZ) (n ⫽ 12) or aortic homograft tissue (n ⫽ 10) was used for the reconstruction. In 12 patients (43%) cardioplegic cardiac arrest was used for a mean duration of 64 ⫾ 31 minutes. A fenestration between the Fontan circuit and neopulmonary venous atrium with a median size of 4 mm (range, 3 to 5 mm) was created in 20 patients (67%). Associated procedures included pulmonary artery augmentation (n ⫽ 17), repair of anomalous pulmonary venous drainage (n ⫽ 7), repair of a regurgitant atrioventricular valve (n ⫽ 2), bidirectional cavopulmonary anastomosis (n ⫽ 3), and resection of subaortic stenosis (n ⫽ 5).
Statistical Methods Data are described as frequencies, medians with ranges, and means with standard deviations. For missing data, the number of non-missed values is given. Simple (univariable) logistic regression analysis was performed to identify predictors of death and the development of early postoperative atrial arrhythmias. Variables assessed as potential predictors of death include date of operation, age, gender, weight, preoperative McGoon and Nakata indexes, preoperative pulmonary artery pressures, age at superior cavopulmonary anastomosis, right or left atrial isomerism, presence of a common atrioventricular valve, pulmonary atresia or stenosis, anomalous pulmonary venous drainage, presence of noncardiac anomalies, and type of Fontan connection. Variables assessed as potential predictors of the development of early postoperative arrhythmias included presence of preoperative arrhythmias, right or left atrial isomerism, and type of Fontan connection. Time-related survival estimates were calculated using the Kaplan-Meier method. All analyses were performed using SAS (SAS Institute, Inc, Cary, NC) version 8. A p value less than 0.05 was set as the level of statistical significance.
Results Postoperative Morbidity and Mortality There were four hospital deaths (13%) all caused by ventricular failure; two of these deaths had takedown of the Fontan connection. Three of the 4 children that died had concomitant repair of anomalous pulmonary venous drainage, including 1 child who also had concomitant repair of a regurgitant atrioventricular valve. In another child, obstructed anomalous venous drainage was iden-
1638
AZAKIE ET AL FONTAN OUTCOMES FOR HETEROTAXY SYNDROMES
Ann Thorac Surg 2001;72:1636-40
in a child with left atrial isomerism. No children had developed protein-losing enteropathy at the time of follow-up. Twenty-four hour ambulatory electrocardiographic monitoring demonstrated predominantly sinus rhythm in all children except 2 (junctional rhythm). Echocardiographic evaluation (n ⫽ 25) showed qualitatively depressed ventricular function in 5 patients (20%). The remaining 20 children had qualitatively good ventricular function. Atrioventricular valve regurgitation was graded as moderate in 3 children, mild in 16, and absent in 7. Fig 1. Kaplan-Meier survival curve from the time of the Fontan operation in children with functional or morphologic single ventricle and visceroatrial heterotaxia. Numbers of patients at risk are given above the horizontal axis. Survival is expressed as a percentage with 95% confidence intervals.
tified after the completion of the Fontan connection by an intraoperative transesophageal echocardiogram. No perioperative risk factors for death by univariable logistic regression analysis were identified, probably because of the small sample size and low mortality. Thirteen children (41%) developed early postoperative atrial arrhythmias; 11 of these children required temporary pacing. In 9 children, sinus node dysfunction was present, characterized by relative bradycardia (or the presence of a junctional rhythm; atrioventricular conduction block developed in 2 patients and a wandering atrial pacemaker was present in 2 patients), or a combination thereof. Patients with left atrial isomerism were more likely to develop early postoperative arrhythmias (odds ratio, 3.5) than those with right atrial isomerism, although this did not reach statistical significance (p ⫽ 0.11). The presence of preoperative arrhythmias was also associated with an increased likelihood of early postoperative atrial rhythm disturbances (odds ratio, 6.0; p ⫽ 0.05). Type of Fontan connection was not associated with the development of atrial rhythm disturbances in the early postoperative period. The mean Fontan pressure in the early postoperative period was 13 ⫾ 3 mm Hg and the mean pulmonary venous atrial pressure was 6 ⫾ 2 mm Hg. The median duration of mechanical ventilatory support was 1 day (range, 6 hours to10 days), the median length of stay in the intensive care unit was 2 days (range, 1 to 10 days), and the median duration of hospital admission was 10 days (range, 6 to 19 days). The median duration of chest tube drainage was 9 days (range, 5 to 19 days).
Midterm Follow-up Follow-up was available on 25 survivors for up to 7.1 years. The mean follow-up time was 3.4 ⫾ 2.7 years. Kaplan-Meier survival (Fig 1) at 1 month is 86% (95% confidence interval: 75% to 97%), and at 1 and 5 years is 83% (95% confidence interval: 70% to 96%). There was one death after hospital discharge caused by anastomotic disruption of a direct connection between the hepatic veins and azygous continuation of the inferior vena cava
Comment Despite the generally reported poor outcomes for patients with left or right atrial isomerism with cardiac anatomy that is unsuitable for biventricular repair, there remains a group of patients who survive initial palliation of the neonatal period and require further palliation for limiting and progressive cyanosis. Cardiac tansplantation is one option, yet there are important disadvantages to this approach. In our recent experience with 30 children with visceroatrial heterotaxy and cardiac anatomy, which precludes biventricular repair, the hospital mortality for the modified Fontan procedure is 13%. Early postoperative rhythm disturbances are common. Children with left atrial isomerism are more likely to have congenital atrioventricular block, junctional rhythm, and hypoplasia of the sinoatrial node, whereas those with right atrial isomerism may have bilateral sinoatrial and atrioventricular nodes. In our experience, the development of postoperative atrial dysrhythmias is more likely in children with preoperative rhythm disturbances or a diagnosis of left atrial isomerism, or both. At midterm follow-up, survival is 83%, ventricular function and atrioventricular valve competency are good, and early rhythm disturbances resolve to base line. We have recently reported cohort analyses of all newborns with left [1] or right [2] atrial isomerism and visceral heterotaxy syndromes. In infants with left atrial isomerism, single ventricle primary palliation has been associated with an early mortality rate of 25% and a 5-year survival of 61%. Risk factors for death included congenital atrioventricular block, coarctation of the aorta, single ventricle and the presence of biliary or other gastrointestinal malformations. More than 50% of children within the entire cohort had anomalous pulmonary venous drainage, which by univariate analysis entered as a risk factor for mortality. Our experience with children with single ventricle and right atrial isomerism has also been reported as a cohort analysis with a focus on primary palliation, presurgical outcomes, and biventricular repair. The overall mortality for staged palliation of children with right isomerism and single ventricle [2] was 62% (40 of 64). Independent risk factors for time-related mortality include absence of pulmonary outflow tract obstruction, presence of major atrioventricular valve anomaly, and obstructed pulmonary veins. In fact, the surgical outcomes for primary neonatal repair of anom-
Ann Thorac Surg 2001;72:1636-40
alous pulmonary venous drainage in the setting of single ventricle was 95% mortality. Others have reported comparable outcomes of infants with single ventricle and total anomalous pulmonary venous drainage, also with limited information on midterm outcomes after the modified Fontan procedure in children with isomeric hearts. Sadiq and colleagues [3] reported high mortality rates for early palliation in children with right atrial isomerism and single ventricle. Seven of their 20 patients progressed to a Fontan operation with five survivors. There were no late survivors among children with a history of obstructed pulmonary veins. Gaynor and colleagues [4] reported on the longterm outcomes of 73 infants with single ventricle and total anomalous pulmonary venous connection. Survival was 37% at 1 year and 23% at 5 years. By univariate analysis, time-related mortality was predicted by younger age and total anomalous pulmonary venous connection. Primary palliation was associated with a mortality rate that approximated 50%. Twenty-one of their patients underwent superior cavopulmonary connections with an early mortality rate of 38%. Early mortality after the Fontan operation was 42% (8 of 21 patients). Of the 73 patient cohorts only 9 had successful Fontan operations at the most recent follow-up. Histologic examination of postmortem specimens revealed marked dilatation of pulmonary veins with increased wall thickness for age. As pointed out by the authors, the venous abnormalities were present even in the absence of clinical evidence of pulmonary venous obstruction, suggesting that current diagnostic methods underestimate the prevalence and severity of obstruction. Detailed information regarding outcomes for the Fontan procedure in children with single ventricle and visceral heterotaxy syndromes (right and left atrial isomerism, or asplenia and polysplenia) is limited, including our two recent cohort analyses [1, 2] and reports from earlier eras [10, 14 –16]. Humes and colleagues [10] reported results for the modified Fontan operation in 49 patients with asplenia and polysplenia between 1975 and 1986. Hospital mortality was 43%, but improved to approximately 15% in a more recent era. Fontan mortality was associated with right atrial isomerism (asplenia) and need for concomitant atrioventricular valve repair. In the report from Marcelletti and colleagues [11], from a total of 12 children with single ventricle and right and left atrial isomerism, 5 children progressed to a Fontan procedure with only one survivor. Culbertson and colleagues [16] reported a 20% mortality rate for patients with heterotaxy syndrome undergoing the Fontan procedure. Preoperative findings of more than mild atrioventricular valve regurgitation significantly increased the risk of overall mortality. The purpose of the present study was to determine the early and midterm outcomes of total cavopulmonary connections in all children with heterotaxy syndrome and single ventricle at a single institution in the current era. The mortality for the Fontan procedure in patients with heterotaxy syndrome in our recent experience is 13%, which compares favorably with published mortality
AZAKIE ET AL FONTAN OUTCOMES FOR HETEROTAXY SYNDROMES
1639
rates of 20% to 80% in earlier eras [10, 11, 16] but remains higher than our institutional mortality of less than 5% for the Fontan procedure in all children with functional single ventricles. Although the current study only addresses the numerator by reviewing the outcomes for children with heterotaxy syndrome undergoing the Fontan procedure, several important observations can be made. The presence of clinically important atrioventricular valve regurgitation [10, 16] is associated with an adverse outcome after the Fontan procedure for children with functional single ventricles and isomeric hearts. In our experience, one of the 2 children undergoing concomitant atrioventricular valve repair died following the Fontan procedure. The inference from our experience and published reports is that early atrioventricular valve repair may protect the pulmonary vasculature and allow for preservation of ventricular function, and may allow for improved outcomes if performed before the Fontan procedure. Three of the 4 patients that died after the Fontan procedure also had concomitant repair of anomalously draining pulmonary veins. In 1 child, pulmonary venous obstruction was only detected after the Fontan procedure. As pointed out by Gaynor and colleagues [4], persistent or recurrent pulmonary venous obstruction is not uncommon in children with single ventricle and anomalous draining pulmonary veins, and pulmonary venous obstruction may not always be clinically evident as demonstrated by their histopathologic studies. In our experience and that reported by McElhinney and Reddy [17], pulmonary venous obstruction may be unmasked by the increased pulmonary blood flow that results after the Fontan operation. Thus, efforts to improve surgical outcomes in such patients require both close monitoring and a high index of suspicion for persistent, recurrent, or previously undetected pulmonary venous obstruction. Furthermore, repair before the Fontan procedure may protect the pulmonary vasculature and improve the outcomes that follow total cavopulmonary connection.
References 1. Gilljam T, McCrindle BW, Smallhorn JF, et al. Outcomes of left atrial isomerism over a 28-year period at a single institution. J Am Coll Cardiol 2000;36:908–16. 2. Hashmi A, Abu-Sulaiman R, Mcrindle BW, et al. Management and outcomes of right atrial isomerism: a 26-year experience. J Am Coll Cardiol 1998;31:1120– 6. 3. Sadiq M, Stumper O, De Giovanni JV, et al. Management and outcome of infants and children with right atrial isomerism. Heart 1996;75:314–9. 4. Gaynor JW, Collins MH, Rychik J, et al. Long-term outcome of infants with single ventricle, and total anomalous pulmonary venous connection. J Thorac Cardiovasc Surg 1999;117: 506–14. 5. Deleon SY, Gidding SS, Ilbawi MN, et al. Surgical management of infants with complex cardiac anomalies associated with reduced pulmonary blood flow and total anomalous pulmonary venous drainage. Ann Thorac Surg 1987;43: 207–11. 6. Di Donato R, di Carlo D, Squitieri C, et al. Palliation of cardiac malformations associated with right isomerism
1640
7. 8.
9.
10. 11. 12.
AZAKIE ET AL FONTAN OUTCOMES FOR HETEROTAXY SYNDROMES
(asplenia syndrome) in infancy. Ann Thorac Surg 1987;44: 35–9. Heinemann MK, Hanley FL, Van Praagh S, et al. Total anomalous pulmonary venous drainage in newborns with visceral heterotaxy. Ann Thorac Surg 1994;57:88–91. Gentles TL, Mayer JE Jr, Gavreau K, et al. Fontan operation in five hundred consecutive patients: factors influencing early and late outcome. J Thorac Cardiovasc Surg 1997;114: 376–91. Knott-Craig CJ, Danielson GK, Schaff HV, et al. The modified Fontan operation: an analysis of risk factors for early postoperative death or takedown in 702 consecutive patients from one institution. J Thorac Cardiovasc Surg 1995;109: 1237– 43. Humes RA, Feldt RH, Porter CJ, et al. The modified Fontan operation for asplenia and polysplenia syndromes. J Thorac Cardiovasc Surg 1988;96:212– 8. Marcelletti C, Di Donato R, Nijveld A, et al. Right and left isomerism: the cardiac surgeon’s view. Ann Thorac Surg 1983;35:400–5. Cetta F, Feldt RH, O’Leary PW, et al. Improved early
Ann Thorac Surg 2001;72:1636-40
13.
14.
15.
16.
17.
morbidity and mortality after Fontan operation: the Mayo Clinic experience, 1987 to 1992. J Am Coll Cardiol 1996;28: 480– 6. Van Arsdell G, McCrindle BW, Einarson KD, et al. Interventions associated with minimal Fontan mortality. Ann Thor Surg 2000;70:568–74. Julsrud PR, Danielson GK. A modification of the Fontan procedure incorporating anomalies of systemic and pulmonary venous return. J Thorac Cardiovasc Surg 1990;100: 233–9. Michielon G, Gharagozloo F, Julsrud PR, et al. Modified Fontan operation in the presence of anomalies of systemic and pulmonary venous connection. Circulation 1993;88: 141– 8. Culbertson CB, George BL, Day RW, et al. Factors influencing survival of patients with heterotaxy syndrome undergoing the Fontan procedure. J Am Coll Cardiol 1992;20:678– 84. McElhinney DB, Reddy VM. Anomalous pulmonary venous return in the staged palliation of functional univentricular heart defects. Ann Thorac Surg 1998;66:683–7.