Classic options for treating hypoplastic left heart syndrome

Transcription

1 Birth Weight and Complexity Are Significant Factors for the Management of Hypoplastic Left Heart Syndrome John H. Artrip, MD, David N. Campbell, MD, D. Dunbar Ivy, MD, Melvin C. Almodovar, MD, Kak-Chen Chan, MD, Max B. Mitchell, MD, David R. Clarke, MD, and François Lacour-Gayet, MD The Children s Hospital Heart Institute, The Children s Hospital, Denver, Colorado Background. Classic options for treatment of hypoplastic left heart syndrome include the Norwood procedure (NW) and heart transplantation (HT). Recently off-pump palliative procedures were introduced in the management of these patients. Risk factors influencing the decision between the NW with staged reconstruction or off-pump palliation and HT were assessed. Methods. Between January 2002 and January 2006, 69 patients with hypoplastic left heart syndrome were referred for either a NW (n 33) or HT (n 36). Patients referred for HT underwent off-pump palliation (catheterbased, n 20, or surgical hybrid procedures, n 7) until a donor organ became available: including patent ductus arteriosus stenting (80.6%), atrial septostomy (41.7%), and branch pulmonary artery banding (55.6%). Results. Heart transplantation patients were more complex, based on a higher Aristotle Comprehensive Complexity score calculated at birth (HT versus NW ; p < 0.05). Presurgical death was significantly greater for HT referrals (HT 27.8% versus NW 3.0%; p < 0.01); however, there was no difference between the two groups in operative (HT 11.5% versus NW 21.9%; not significant) or overall mortality (HT 36.1% versus NW 24.2%; not significant). No independent risk factors for death were identified with HT; however, a lower birth weight (<2.5 kg) and a higher Aristotle score (>20) correlated with surgical death with NW (p < 0.01). Noticeably, surgical survival was 85.7% for infants with birth weight of 2.5 kg or greater undergoing NW. Conclusions. Overall survival is similar for hypoplastic left heart syndrome patients referred for the NW or HT. Lower birth weight and higher complexity are risk factors for patients undergoing NW. Off-pump palliation followed by HT or further staged palliation should be considered for these high-risk hypoplastic left heart syndrome patients. (Ann Thorac Surg 2006;82:1252 9) 2006 by The Society of Thoracic Surgeons Accepted for publication April 24, Presented at the Forty-second Annual Meeting of The Society of Thoracic Surgeons, Chicago, IL, Jan 30 Feb 1, Address correspondence to Dr Lacour-Gayet, The Children s Hospital Heart Institute, 1056 E 19th Ave, B200, Denver, CO ; lacour-gayet.francois@tchden.org. Classic options for treating hypoplastic left heart syndrome (HLHS) are reconstructive surgery (Norwood procedure; NW) [1] or orthotopic heart transplantation (HT) [2]. Referral for either procedure is largely based on the particular expertise of the treating center and the family s preference. Regardless which pathway is pursued, the survival is similar and typically approaches 64% at 1 year [3]. At centers committed to reconstructive surgery, the NW carries a 23% to 46% surgical fatality rate [4 6]. At centers committed to HT, the surgical procedure carries a fatality rate of 13%; however, 26% to 32% of infants listed for HT die before a donor organ becomes available [7, 8]. Hypoplastic left heart syndrome represents a wide spectrum of cardiac lesions [9] with a high incidence of noncardiac malformations and comorbidities [10]. Factors identified as risk for death with the NW or HT referral are not necessarily the same. Risk factors previously identified for surgical fatality after the NW include aortic atresia, depressed right ventricular function, atrioventricular valve regurgitation, pulmonary venous abnormalities, low birth weight, and prematurity. Limited data exist on the risk of these factors associated with HT referral. The unstable physiology of HLHS in its natural state can result in death before a donor organ becomes available, and for this reason, time on the wait list and blood type O have been important risk factors for mortality. Recent strategies to improve survival to HT include conservative atrial septostomy, patent ductus arteriosus stenting, and branch pulmonary artery (PA) banding [11, 12]. Such off-pump palliative procedures can be performed using interventional techniques (catheter-based) or as part of a surgical hybrid procedure. The introduction of these techniques could impact patient selection for reconstructive surgery or referral for HT. The objective of this study was to identify risk factors for each treatment pathway that may influence the decision between the NW with staged reconstruction and off-pump palliation and HT referral by The Society of Thoracic Surgeons /06/$32.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg ARTRIP ET AL 2006;82: TREATMENT OF HLHS: NORWOOD OR TRANSPLANTATION 1253 Patients and Methods Patients The protocol for this study was reviewed by the Colorado Multi-Institutional Review Board (COMIRB Protocol No ) and classified as exempt because patient identifying information was not used. Additionally, the requirement for individual patient consent was waived. Between January 1, 2002, and January 1, 2006, 69 consecutive patients with HLHS were referred for treatment. The mean birth weight was kg (range, 1.4 to 4.4 kg). Antenatal diagnosis was suspected by fetal echocardiography in just over half of the patients (n 35) and confirmed by standard two-dimensional echocardiography at the time of birth. Hypoplastic left heart syndrome was characterized by ductal-dependent circulation and a left ventricle insufficient to support systemic flow (calculated left ventricular volumes 15 ml/m 2 ). Patients with dominant single left ventricle were excluded (three tricuspid atresia with transposition and one severe Ebstein s anomaly). Sixty-one patients had typical aortic and mitral valve atresia. Three patients had severe aortic and mitral valve stenosis, 2 patients had double-outlet right ventricle with a hypoplastic left ventricle and multiple ventricular septal defects, 2 patients had an unbalanced atrioventricular septal defect with a hypoplastic aortic arch, and 1 patient had severe Shone s complex. At presentation, all were on conventional prostaglandin E 1 therapy ( to 0.05 mg kg 1 min 1 ). Reconstructive NW or HT was offered as the principal therapeutic option to the patient s family. Patients referred for HT followed the clinical protocol as previously published [13, 14]. Briefly, echocardiography was performed every 1 to 2 weeks or as clinically indicated. Prostaglandin E 1 therapy was continued until ductal stenting was performed or until a donor heart became available. Atrial sepal defects were left moderately restrictive ( 2.5 m/s Doppler velocity). Hypoxic ventilation (18% inspired oxygen and 82% inspired nitrogen) was used as needed to maintain elevated pulmonary vascular resistance and preserve systemic perfusion (goal of pulmonary blood flow to systemic blood flow of approximately 1:1). Patients who were intubated at presentation were extubated, and inotropic agents and diuretics were withdrawn as soon as possible. Patients with stable hemodynamics were discharged from the hospital and followed up with weekly outpatient visits. Sustained saturations less than 60% or echocardiographic findings of restrictive atrial blood flow or excessive pulmonary blood flow were indications for balloon or blade atrial septostomy or PA banding, respectively. Surgical and Interventional Techniques NORWOOD PROCEDURE. Neoaorta reconstruction was performed with cardiopulmonary bypass ( minutes) and moderate hypothermia to deep hypothermia as needed (20 to 25 C). Selective antegrade cerebral perfusion was used in all patients and, when necessary, brief periods of circulatory arrest (n 10, 18 9 minutes). Pulmonary blood flow was established by right ventricleto-pa conduit (5 mm Gore-Tex; W. L. Gore and Associates, Inc, Flagstaff, AZ) in all but 8 patients who received standard Blalock-Taussig shunts (3.5 mm Gore-Tex). HEART TRANSPLANTATION. Biatrial HT with arch reconstruction was performed with cardiopulmonary bypass ( minutes) and deep hypothermia (approximately 20 C). Selective antegrade cerebral perfusion was used and, when necessary, brief periods of circulatory arrest (n 15, minutes). Standard organ procurement was performed with preservation using Roe s solution (approximately 30 ml/kg). Donor ischemia time was minutes. OFF-PUMP PALLIATIVE PROCEDURES. Biplane cineangiography was performed in patients referred for off-pump palliation. Ductal stenting with or without balloon atrial septostomy was performed in 29 patients at days of life. Self-expanding nitinol stents (Cordis Precise RX; Cordis Corp, Miami Lakes, FL, or EV3 protégé) with a length of 20 mm were used. Stents were sized to be approximately 2 mm greater than the diameter of the patent ductus arteriosus. Final diameter ranged from 6 to 12 mm. Bilateral PA banding was performed in 20 patients at days of life. Catheter-based internal banding was performed in 13 patients under angiographic guidance. Pulmonary blood flow restrictor devices (7 mm diameter) with two 2-mm holes (Amplatzer; AGA Medical Corp, Golden Valley, MN) were placed in each branch PA. Bands were sized to be approximately 120% of the branch PAs and ranged from 6 to 12 mm in diameter. Seven patients underwent surgical hybrid procedures with sternotomy and placement of external PA bands (W. L. Gore and Associates, Inc, Flagstaff, AZ) [14]. The proximal right and left PAs were banded while the distal pulmonary blood pressure was monitored by a long 20F catheter placed through the PA trunk. The band was tightened to bring mean branch PA pressure to approximately half mean systemic pressure. Calculation of Basic and Comprehensive Aristotle Scores The Aristotle score is a complexity score based on the surgical procedure and its potential for mortality and morbidity, and the anticipated technical difficulty [15, 16]. The basic and comprehensive Aristotle score values were previously published (15, 23). For the stage 1 NW, the values are summarized in Table 1. The basic score value is The Aristotle comprehensive score further adjusts the complexity according to specific patient and procedural characteristics including anatomic factors, associated procedures, age, and other clinical factors. The comprehensive score can add 10 additional points to the 14.5 points of the basic score. A maximum of 5 points can be added respectively for the procedure-dependent and procedure-independent factors. The most frequent comorbidities met in the series are shown in Table 1. Because the score is procedural-based, for comparison of complexity between the two groups, all scores were calculated based on birth characteristics and assuming a NW would be performed within the first 2 weeks of life.

3 1254 ARTRIP ET AL Ann Thorac Surg TREATMENT OF HLHS: NORWOOD OR TRANSPLANTATION 2006;82: Table 1. Aristotle Scores for Norwood Procedure Basic score (max 15) 14.5 Comprehensive score (max 25) Procedure-dependent factors (max 5) Obstructed pulmonary venous return (no ASD) 4 TAPVD repair 4 AV valve regurgitation, grades 3 and Aortic atresia 3 Age 1 mo 3 IAA repair 2.5 Most frequent procedure-independent factors (max 5) Extreme prematurity 32 weeks 4 Prematurity weeks 2 Weight 2.5 kg 2 Cardiopulmonary resuscitation 2 Mechanical ventilation to treat cardiorespiratory failure 2 Shock, resolved at time of surgery 1 Necrotizing enterocolitis treated medically 1 ASD atrial septal defect; AV atrioventricular; IAA interrupted aortic arch; max maximum; TAPVD total anomalous pulmonary venous drainage. Statistical Analysis Statistical analysis was performed with SPSS 13.0 for Windows (SPSS Inc, Chicago, IL). Differences between means were compared with a standard Student s t test. Differences between proportions were compared using Fisher s exact test. Multivariate risk factor analysis was performed using logistic regression. Kaplan Meier actuarial survival curves were constructed and difference assessed by log-rank test. Differences were considered statistically significant at a probability value less than Results Table 2 lists the characteristics of the patients with HLHS referred for either stage 1 NW palliation or primary HT. No statistical differences were observed between the two groups with respect to the cardiac lesions, birth weight, or associated genetic syndromes. However, the comprehensive Aristotle score was significantly higher for the patients referred for HT, suggesting that collectively the HT referral group was more complex and at increased risk for death had a NW been performed. At the time of the NW, the mean age and weight were days and kg, respectively. One child died the morning of the scheduled NW because of uncorrectable acidosis and myocardial failure. For the patients referred for HT, the mean age and weight were days and kg, respectively, at the time of the transplantation. These values are statistically significant (p 0.01) compared with age and weight of the patients at time of the NW. Catheterization procedures performed before organ transplantation included patent ductus arteriosus stenting (80.6%), atrial septostomy (44.4%), and PA banding (55.6%). Ten patients died waiting for a donor heart. The average age at death for these patients was days (range, 13 to 176 days). Survival to hospital discharge for HLHS patients referred for treatment was 69.9%, and was similar between the reconstructive and transplant groups (75.8% for the NW and 63.9% for HT). Table 3 lists the deaths of the two groups relating to the primary surgical procedure. Presurgical death correlated with referral for HT. There was a trend for surgical fatality to correlate with the NW, although statistical significance was not achieved. Multivariate analysis included 25 factors listed in Table 4. The analysis was performed for all 69 HLHS patients Table 2. Sixty-Nine Patients With Hypoplastic Left Heart Syndrome Referred for Norwood Procedure or Heart Transplantation Variable NW (n 33) HT (n 36) p Value Birth weight (kg) (range, ) (range, ) Sex Male 19 (57.6%) 19 (52.8%) Female 14 (42.4%) 17 (47.2%) Cardiac lesions Aortic atresia 27 (81.8%) 34 (94.4%) Anomalous pulmonary venous drainage b 1 (3.0%) 7 (19.4%) Anomalous systemic venous drainage c 1 (3.0%) 2 (5.6%) Moderate-to-severe AV valve regurgitation 1 (3.0%) 5 (13.9%) Noncardiac risk factors Genetic syndromes d 6 (18.2%) 5 (13.9%) Low birth weight ( 2.5 kg) 4 (12.1%) 9 (25.0%) Prematurity ( 36 weeks gestation) 2 (6.1%) 6 (16.7%) Complexity-adjusted score Aristotle score a a Statistically significant for p b Includes total anomalous pulmonary venous return (supracardiac [n 1], infracardiac [n 2], drainage to the coronary sinus [n 1], and mixed drainage [n 3]) and partial anomalous drainage of right upper lobe to SVC (n 1). c Includes bilateral SVC (n 2) and interrupted IVC with azygous continuation (n 1). d Includes Down syndrome (n 1), Turner syndrome (n 3), DiGeorge syndrome (n 1), heterotaxy syndrome (n 2), and other chromosomal deletions (n 4). AV atrioventricular; HT heart transplantation; IVC inferior vena cava; NW Norwood procedure; SVC superior vena cava.

4 Ann Thorac Surg ARTRIP ET AL 2006;82: TREATMENT OF HLHS: NORWOOD OR TRANSPLANTATION 1255 Table 3. Mortality and Fatality Data Variable NW (n 33) HT (n 36) p Value Presurgical death b 1 (3.0%) 10 (27.8%) a Surgical death c 7/32 (21.9%) 3/26 (11.5%) Hospital mortality 8 (24.2%) 13 (36.1%) a Statistically significant for p b Presurgical death is defined as death before NW or HT. c Surgical death is defined as death within 30 days after NW or HT combined with discharge mortality. Patients who died before the primary surgery were excluded. HT heart transplantation; NW Norwood procedure. and then stratified by the pathway of treatment (NW or HT). Outcome variables were presurgical death (death before NW or HT), surgical death ( 30 days of NW or HT or discharge mortality), and all death (presurgical and surgical death). Analysis of all HLHS patients regardless of treatment pathway failed to identify aortic atresia, complex anomalous systemic or pulmonary venous drainage, moderate-to-severe atrioventricular valve regurgitation, birth weight, or the presence of genetic syndromes as predictors of presurgical, surgical, or all death. The risk factor of prematurity was statistically significant for all death (p 0.04). Stratifying patients by treatment pathway identified lower birth weights, prematurity, higher Aristotle scores, and longer cardiopulmonary bypass times correlating with surgical fatality with the NW (p 0.01 for all). Similar analysis performed on patients referred for HT failed to identify any risk factors correlating with presurgical death, and only longer donor ischemic times correlated with surgical death (p 0.04). An Aristotle score greater than 20 and a birth weight less than 2.5 kg (low birth weight [LBW]) were both strong individual predictors of surgical death with a NW (p 0.01 for both). Importantly, Aristotle score greater than 20 and LBW were not identified as risk factors for wait list (presurgical), surgical, or all death for patients referred for HT. For patients presenting with Aristotle scores greater than 20, all 3 patients treated with a NW died with surgery compared with 3 of 8 patients referred for HT who died waiting for an available organ. Similarly for LBW patients, all 4 patients treated by a NW died within days after surgery. Of the 9 LBW patients referred for HT, 2 died waiting for an available organ and none died after HT. For the 7 patients surviving until a donor heart became available, at the time of HT the mean age was days and the mean weight was kg. This represents an increase of weight of kg during the waiting period. Midterm results for both pathways are illustrated in the Kaplan Meier survival curves (Fig 1). There is no significant survival difference between the NW and HT groups. Of the 25 patients who survived their NW, there was one interim death. This child was found at home unresponsive at 4 months of age. One patient survived stage 2 cavopulmonary anastomosis but subsequently underwent HT owing to sustained pulmonary hypertension and cyanosis. Twenty patients have completed and 4 Table 4. Twenty-Five Factors Identified for Multivariate Analysis Patient-specific factors OR age OR weight Sex Birth weight Prematurity ( 36 weeks gestation) Genetic syndromes Blood type Anatomic factors Aortic atresia Pulmonary venous anomalies Systemic venous anomalies AV regurgitation Restrictive ASD Complexity-adjusted score Aristotle score Surgery-specific factors Date of surgery Surgeon CPB time X-clamp time Circulatory arrest time HT-specific factors PDA stent PA banding Atrial septostomy Donor ischemic time Donor blood type NW-specific factors Type of shunt Size of shunt ASD atrial septal defect; AV atrioventricular; CPB cardiopulmonary bypass; HT heart transplantation; NW Norwood procedure; OR operating room; PA pulmonary artery; PDA patent ductus arteriosus. Fig 1. Kaplan Meier survival curves. There was no statistical difference between midterm survival for patients referred for a Norwood procedure or heart transplantation.

5 1256 ARTRIP ET AL Ann Thorac Surg TREATMENT OF HLHS: NORWOOD OR TRANSPLANTATION 2006;82: are awaiting their stage 2 procedures. Of the 23 patients who survived their HT, there have been 2 late deaths and no retransplantations. The deaths occurred at 9 and 15 months after HT both after rejection episodes. Comment This was a nonrandomized study with patients referred for NW or HT based on individual preferences of the family and referring cardiologist. The Aristotle score was significantly higher for the patients referred for HT suggesting a relative bias for more complex patients being referred for HT. However, statistically significant differences in individual risk factors (cardiac lesions, birth weight, or associated genetic syndromes) were not observed between the NW or HT groups. The study lacked necessary power to demonstrate statistical differences in such individual risk factors. A collective risk score such as Aristotle should increase the sensitivity in detecting differences in group characteristics, and will likely prove useful with the small sample sizes encountered in congenital heart surgery. Despite the difference in Aristotle score between groups, early and midterm survival was similar between the NW and HT groups. Preoperative death correlated with referral for HT and there was a trend for surgical death to correlate with the NW. These findings were anticipated and similar to previously reported series of HLHS patients [3]. For patients referred for HT, the risk of death is shifted from the surgical procedure to the preoperative waiting period. Previous studies on infant heart transplantation identified recipient blood type O and time on the wait list as an important risk factor for wait list mortality [17]. This was not true in our study. There was no correlation between recipient blood type and death, and for the 10 patients who died waiting for a donor heart, the average age at death was days compared with days for the 26 patients surviving until HT. The introduction of prostaglandin E 1 and subsequent off-pump palliation may stabilize this group of patients until a donor organ is available, thus allowing for infant growth and potentially lessening the risk of the surgical procedure. Our study failed to identify any differences in the frequency of patent ductus arteriosus stenting, PA banding, or atrial septostomy for infants dying on the wait list versus those surviving until HT. Right ventricular failure with uncorrectable acidosis was the most frequent cause of wait list death; however, 4 patients had catheter-related events with subsequent death. As with all new technologies, there is a relative learning curve, and with growing experience, the morbidity and mortality related to the procedure should improve. Other causes of death included gastrointestinal sepsis and apnea with subsequent cardiac arrest while on prostaglandin E 1 therapy. Although our wait list mortality approximates the wait list mortality presented in previous studies, the actual wait list times have increased. An earlier study from our institution had an average wait list time of days [18], approximately 40 days fewer than the current waiting period with off-pump palliation. In other words, off-pump palliation has not affected the wait list mortality but rather extended the safe period of waiting. Our discharge mortality for a NW was 24.2%. Other reported series have cited rates as low as 6% [19]; however, recent analysis of The Society of Thoracic Surgeons Congenital Heart Surgery Database lists 31.4% discharge mortality for a NW [20] results consistent with our study. Of the previously identified risk factors for surgical death with the NW, lower birth weight, prematurity, and higher Aristotle score were statistically significant in our study. These three factors are interrelated with the obvious correlation between prematurity and birth weight. Additionally both of these factors are tallied in the calculation of the Aristotle score, which alone may account for its correlation with surgical death. An ad hoc analysis performed after removing the LBW patients lacked statistical power to demonstrate significant correlation between Aristotle score and surgical death. Lower birth weight has consistently proved to be a strong predictor of surgical death with the NW [21]. In our series a birth weight less than 2.5kg was uniformly fatal with the NW; however, surgical survival for patients weighing at least 2.5 kg at birth was 85.7%. Because the NW is performed within the first 2 weeks of life, birth weight approximates the weight at operation. There was no difference between birth weight and operative weight for our series of patients undergoing the NW. For patients referred for HT, this is not true, and the weight at operation significantly differs from the birth weight. The average waiting period for a donor heart was days. During this period there was an increase of weight of kg. Of the LBW infants that were referred for HT, 7 of 9 survived to receive a donor organ and gained an average of 1.4 kg during the listing period. The 2 LBW infants who died on the wait list had birth weights of 1.4 and 1.6 kg the lowest of the group. Although LBW was not identified as a risk factor for death with patients palliated to HT, the power of our study was insufficient to achieve statistical significance. It is quite possible that LBW presents a risk for off-pump palliation and HT referral, but at a reduced level compared with the NW pathway. In summary, LBW and an Aristotle score greater than 20 pose increased risk for a NW as opposed to off-pump palliation and HT. For this high-risk subset of HLHS infants, alternative treatment strategies should be considered, including off-pump palliation and HT. Figure 2 illustrates the treatment protocol for HLHS patients presenting at Denver Children s Hospital. On presentation, patients are categorized as low or high risk for a subsequent NW based on birth characteristics, anatomy, and Aristotle comprehensive complexity score. Low-risk patients are referred for NW. High-risk patients are referred for off-pump palliation including patent ductus arteriosus stenting and, if needed, atrial septostomy. Timing of branch PA banding presents a critical issue. Early banding ( 4 weeks age) may better protect the pulmonary vasculature from high pressures and flow of the unrestricted circulation; however, with increasing infant growth the bands may become restrictive with

6 Ann Thorac Surg ARTRIP ET AL 2006;82: TREATMENT OF HLHS: NORWOOD OR TRANSPLANTATION 1257 Fig 2. Denver Children s Hospital 2005 protocol for the treatment of hypoplastic left heart syndrome (HLHS) patients. (PA pulmonary artery.) time, resulting in cyanosis. Our average waiting period for a donor heart is just under 4 months, with some periods extending past 6 months. Delaying branch PA banding beyond 4 weeks of age and into the second and third months of life may extend the waiting times and improve survival to HT. The timing of band placement is effectively balancing the risk between irreversible pulmonary vascular injury and progressive cyanosis. In our study, branch PA banding was performed at days well beyond a suggested 4-week window. However, our institutional bias is for performing subsequent HT and not subsequent arch reconstruction with a cavopulmonary anastomosis as is being performed at other centers [22]. Finally, off-pump palliative procedures should be strongly considered for high-risk HLHS patients, enabling further infant growth and development before more definitive surgical procedures, whether HT or further staged palliation [23]. We would like to thank James Todd, MD, Department of Epidemiology, for his assistance with the statistical analysis of our manuscript. References 1. Norwood WJ, Lang P, Hansen DD. Physiologic repair of aortic atresia-hypoplastic left heart syndrome. N Engl J Med 1983;308: Bailey LL, Nehlsen-Cannarella SL, Doroshow RW, et al. Cardiac allotransplantation in newborns as therapy for hypoplastic left heart syndrome. N Engl J Med 1986;315: Jacobs ML, Blackstone EH, Bailey LL. Intermediate survival in neonates with aortic atresia: a multi-institutional study. The Congenital Heart Surgeons Society. J Thorac Cardiovasc Surg 1998;116: Gaynor JW, Mahle WT, Cohen MI, et al. Risk factors for mortality after the Norwood procedure. Eur J Cardiothorac Surg 2002;22: Forbess JM, Cook N, Roth SJ, Serraf A, Mayer JE Jr, Jonas RA. Ten-year institutional experience with palliative surgery for hypoplastic left heart syndrome: risk factors related to stage I mortality. Circulation 1995;92(Suppl):II Ashburn DA, McCrindle BW, Tchervenkov CI, et al. Outcomes after the Norwood procedure in neonates with critical aortic stenosis or aortic atresia. J Thorac Cardiovasc Surg 2003;125: Chrisant MR, Naftel DC, Drummond-Webb J, et al. Fate of infants with hypoplastic left heart syndrome listed for cardiac transplantation: a multicenter study. J Heart Lung Transplant 2005;24: Chiavarelli M, Gundry SR, Razzouk AJ, Bailey LL. Cardiac transplantation for infants with hypoplastic left-heart syndrome. JAMA 1993;270: Noonan JA, Nadas AS. The hypoplastic left heart syndrome: an analysis of 101 cases. Pediatr Clin North Am 1958;5: Nantowicz M, Chatten J, Clancy R, Conard K, Glauser T, Huff D. Genetic disorders and major extracardiac anomalies associated with hypoplastic left heart syndrome. Pediatrics 1988;82: Ruiz CE, Gamra H, Zhang HP, Garcia EJ, Boucek MM. Brief report: stenting of the ductus arteriosus as a bridge to cardiac transplantation in infants with the hypoplastic leftheart syndrome. N Engl J Med 1993;328: Gibbs JL, Wren C, Watterson KG, Hunter S, Hamilton JR. Stenting of the arterial duct combined with banding of the pulmonary arteries and atrial septectomy or septostomy: a new approach to palliation for the hypoplastic left heart syndrome. Br Heart J 1993;69: Bourke KM, Sondheimer HM, Ivy DD, et al. Improved pretransplant management of infants with hypoplastic left heart syndrome enables discharge home while waiting for transplantation. Pediatr Cardiol 2003;24: Mitchell MB, Campbell DN, Boucek MM, et al. Mechanical limitation of pulmonary blood flow facilitates heart transplantation in older infants with hypoplastic left heart syndrome. Eur J Cardiothorac Surg 2003;23: Lacour-Gayet F, Clarke D, Jacobs J, et al. The Aristotle score: a complexity-adjusted method to evaluate surgical results. Eur J Cardiothorac Surg 2004;25: Jacobs JP, Lacour-Gayet FG, Jacobs ML, et al. Initial application in the STS congenital database of complexity adjustment to evaluate surgical case mix and results. Ann Thorac Surg 2005;79: Morrow WR, Naftel D, Chinnock R, et al. Outcome of listing for heart transplantation in infants younger than six months: predictors of death and interval to transplantation. The Pediatric Heart Transplantation Study Group. Heart Lung Transplant 1997;16: Mitchell MB, Campbell DN, Clarke DR, et al. Infant heart transplantation: improved intermediate results. J Thorac Cardiovasc Surg 1998;116: Tweddell JS, Hoffman GM, Fedderly RT, et al. Patients at risk for low systemic oxygen delivery after the Norwood procedure. Ann Thorac Surg 2000;69: Jacobs JP, Jacobs ML, Mavroudis C, Lacour-Gayet FG. Executive Summary: The Society of Thoracic Surgeons Congenital Heart Surgery Database Second Harvest ( ) Beta Site Test. The Society of Thoracic Surgeons (STS) and Duke Clinical Research Institute (DCRI), Duke University Medical Center, Durham, North Carolina, United States, Fall 2002 Harvest. Available at Weinstein S, Gaynor JW, Bridges ND, et al. Early survival of infants weighing 2.5 kilograms or less undergoing first-stage reconstruction for hypoplastic left heart syndrome. Circulation 1999;100(Suppl):II Michel-Behnke I, Akintuerk H, Marquardt I, et al. Stenting of the ductus arteriosus and banding of the pulmonary arteries: basis for various surgical strategies in newborns with multiple left heart obstructive lesions. Heart 2003;89: Aristotle Institute Web site. Available at aristotleinstitute.org (accessed June 26, 2006).

7 1258 ARTRIP ET AL Ann Thorac Surg TREATMENT OF HLHS: NORWOOD OR TRANSPLANTATION 2006;82: DISCUSSION DR PHILIP C. SMITH (Akron, OH): How are you handling very low birth weight babies, say, less than 1,500 g, 32 weeks, where catheter-based technologies and the associated instrumentation really preclude your being able to do these things? DR ARTRIP: Well, our two lowest birth weights, 1.4 kg and 1.6 kg, were both maintained on prostaglandin therapy until ductus stenting was able to be performed. So they essentially remained in the hospital on prostaglandin therapy; and if they could be extubated, they would be. We tried to titrate down the inotropic agents as needed, but some may require longer periods of inotropic agents. Low birth weight in a catheter-based intervention may be a risk factor for death. Now, it may be these extreme low birth weights that are risk factors as opposed to something like 2.5 kg with a surgical procedure. DR BRADLEY S. ALLEN (Houston, TX): I would like to focus on the time that patients were waiting for a heart transplant. If I remember the numbers correctly, there were 35 patients, and 10 died before transplantation. However, the majority of these patients received a hybrid procedure as a palliative procedure. That seems to indicate these new hybrid procedures don t work very well. If these patients were proceeding down a delayed Norwood tract, this would mean an almost 30% mortality after hybrid palliation. In addition, these patients would still have to undergo a Norwood and Glenn with their risk of mortality. That seems to say these hybrid procedures don t work very well for Norwood palliation, at least at Denver, which is one of the original institutions. I wondered if you could comment. DR ARTRIP: Well, you re absolutely right. The hybrid procedures do carry risks. We had four catheter-related deaths. In particular, in one death we caused an intimal flap and obstructed the isthmus; in one case the patient has a supraventricular arrhythmia and subsequently died of that; and two other catheter-related injuries. So these are not risk-free procedures. The thing that I can tell you from the data is that if you look at patients who are smaller birth weights, perhaps the catheter interventions carry less risk than an actual Norwood procedure. DR MARSHALL L. JACOBS (Philadelphia, PA): I think that s a fascinating analysis and a very, very intelligent proposal regarding diverse therapies and finding specific indications. My question goes to the low birth weight infants. I was very, very impressed that in the patients awaiting transplantation, you were able to grow these patients from 2.4 kg to 3.7 kg. I suspect that with the internal or surgical banding you had altered the risk factors for necrotizing enterocolitis and were able to enterally feed these children. That s really my first question, is what is the strategy by which you re able to get them on a thriving growth curve? But in the same context, Dr Boucek, from your institution, has many times shown a slide, the data basically saying that even when hypothermic circulatory arrest was the primary support mechanism for transplantation at Denver, the most important risk factor for adverse neurologic outcome in the transplant patients was not bypass-related or circulatory arrest related, but waiting time related. So tell us about the positive aspects of waiting time management and the growth and thriving and what you have done to mitigate that seemingly negative impact of waiting time on neurologic outcomes. DR ARTRIP: The waiting times actually have increased. A few years ago Dr Mitchell published a paper where his waiting times were about 40 days less than our current waiting times for donor organ. Now, the actual wait list mortalities remain about the same. So I think the impact of catheter-based interventions and PA 9pulmonary artery) banding have been predominantly to increase the waiting times, not to really reduce the mortality as we initially hoped. As for the first question about thriving, I m not sure we do anything different than any other institution. We do try to feed these patients as early as possible. And we do have problems with necrotizing enterocolitis. As a matter of fact, one of the patients who was on prostaglandin therapy died because of necrotizing enterocolitis. So yes, we do have the same problems. I m not sure we do anything really different. DR CHRISTOPHER A. CALDARONE (Toronto, Ontario, Canada): I just wanted to rise to maybe defend the hybrid palliation approach a little bit. I think with regard to the timing of hybrid palliation in patients awaiting heart transplantation, the timing is very important. And we have wrestled with this quite a bit in Toronto. Clearly, the rockier and more unstable the patients are going into that procedure, the more rocky and unstable they are going out. And it s a difficult clinical question as to pick the time for palliation. As you know, a donor organ might come the next day, so you don t want to do an unnecessary procedure. On the other hand, if you wait until the patient s status has deteriorated, the mortality goes up with the palliative procedure. How do you decide when to palliate them? DR ARTRIP: Well, we try to do the ductal stenting as early as possible. One of the limitations is actually size. As mentioned in the earlier question, extremely small birth weights, you have to wait until a greater birth weight. DR CALDARONE: Well, let s leave the small birth weight babies out. Just your standard 3.5-kg neonate with hypoplastic left heart syndrome awaiting heart transplantation. DR ARTRIP: We ll try to do a ductal stenting within the first few weeks of life, that way we can get them off prostaglandins and extubated. However, as for atrial septectomies and pulmonary artery banding, our decision is based on clinical criteria. Our cardiologists follow these patients quite closely. They re seen on a weekly basis. They have echocardiograms. And if they develop signs of restrictive ASD (atrial septal defect), obviously they are referred for an atrial septectomy, and if overcirculated, then they are referred for pulmonary artery banding. I cannot say that we have a defined day that the patient is going to get a ductal stenting at 2 weeks, and at 4 weeks is going to get a PA banding. It is a clinical algorithm, and based on whether the patients become cyanotic or if the patients are overloaded, and that s how we make our decisions. DR BOULOS ASFOUR (St. Augustin, Germany): Actually, we already did what you suggested in your conclusion and analyzed

8 Ann Thorac Surg ARTRIP ET AL 2006;82: TREATMENT OF HLHS: NORWOOD OR TRANSPLANTATION 1259 the outcome of your Norwood patients in respect to the Aristotle score. We divided our patients by Aristotle score in high-risk and low-risk groups. We found a mortality rate of about 40% in the high-risk group (Aristotle score over 20) and less than 10% in the low-risk group (score less than 20). We don t do any hybrid procedures or transplantations. Because of that we developed the modified Norwood procedure on the beating heart (published October 2005 in The Annals). I would like to ask if you evaluated doing these kinds of procedures for your high-risk group patients. DR ARTRIP: Yes, I ve read that paper actually. It s a very interesting concept, and it might be a future aspect for very high-risk patients. We have not done that in Denver, and we do have at our disposal catheter-based interventions and heart transplantations. But obviously, if your institution does not have that pathway, the beating heart palliation might be a viable alternative for very complex patients. DR ASFOUR: With this procedure, it didn t have the survival, so we have to see. DR CARL L. BACKER (Chicago, IL): I would like to poll the audience to find out how many people are offering heart transplantation to patients with hypoplastic left heart syndrome. Could I have a show of hands, how many people are still offering heart transplantation? (Show of hands.) I see two hands. Is that it? Let s keep going with the poll. Norwood-Sano versus Norwood-modified BT (Blalock-Taussig) shunt, how many people are doing the Norwood with the Sano modification? (Show of hands.) Quite a few, at least 50%. And how about Norwood with the modified BT shunt? (Show of hands.) Probably about % Sano, 40% modified BT. And I saw two hands for heart transplantations. François, do you want to make a comment? You appear to be swimming upstream with these heart transplants. There are only two other centers in this audience that are routinely offering them for hypoplastic left heart syndrome. DR FRANÇOIS G. LACOUR-GAYET (Denver, CO): When I came to Denver, there was 100% transplantation for HLHS (hypoplastic left heart syndrome). As John Artrip presented very well, there is a culture of transplantation in Denver that will take time to reverse. But I believe that for complex Norwood, as published in several papers, including the excellent paper from Boulos Asfour from St. Augustin, we don t control this surgery. When dealing with HLHS with isolated mitral and aortic atresia and a weight of 3.5 kg, the mortality is probably less than 10%. But when dealing with patients with a high comprehensive Aristotle score (meaning that it can be associated to total anomalous pulmonary venous return, interrupted aortic arc, and weight less than 2.5 kg), I have not seen a single publication with a mortality less than 10%. So, I think that for these patients, there is still an indication for heart transplantation. Now, whether in the long term this is the best solution, I don t have the answer. But we are fighting to get this patient surviving and this is a way to do it. Now, it could be that doing a pulmonary branch banding with a hybrid procedure is another solution and may work, but it s not easy to band the pulmonary branch of a baby that is 2.1 kg, at least in our experience.