Could we still improve early and interim outcome after prosthetic systemic-pulmonary shunt? A risk factors analysis

European Journal of Cardio-thoracic Surgery 34 (2008) 545 549 www.elsevier.com/locate/ejcts Could we still improve early and interim outcome after prosthetic systemic-pulmonary shunt? A risk factors analysis Siamak Mohammadi *, Osama Benhameid, Andrew Campbell, Jim Potts, Jacqueline Joza, Hamad Al-Habib, Suvro Sett, Jacques Le Blanc Department of Pediatric Cardiac Surgery, British Columbia Children s Hospital, Vancouver, BC, Canada Received 28 December 2007; received in revised form 27 May 2008; accepted 2 June 2008; Available online 16 July 2008 Abstract Objective: To identify factors associated with in-hospital and interim mortality in children with a systemic-to-pulmonary shunt (SPS). Methods: Between January 1988 and April 2005, 226 children with a median age of 17 days, and weight of 3.4 kg, underwent an isolated SPS for pulmonary atresia (PA)-VSD/ tetralogy (n = 124, 54.9%), functional single ventricle PA (n = 35, 5.5%), PA-intact septum (IS, n = 31, 13.7%), transposition of the great arteries VSD-PA (n = 30, 13.3%), and double outlet right ventricle-pa (n = 6, 2.6%). Surgery was performed through sternotomy (group S, n = 46) or thoracotomy (group T, n = 180). The origin of the SPS was either the innominate artery (n = 38) or ascending aorta (n = 8) in group S, and the subclavian artery (n = 180) in group T. Results: In-hospital mortality was 5.7%. Univariate and logistic regression analysis revealed younger age ( p = 0.01), lower body weight ( p < 0.04), a diagnosis of PA-IS with severe right ventricle hypoplasia ( p = 0.005), preoperative intubation ( p = 0.03), increased length of intubation ( p < 0.0001), longer ICU stay ( p < 0.0001), and group S ( p = 0.03) as risk factors for in-hospital death. Group S had a longer median ventilation time (112 vs 30 h, p < 0.0001) despite the similar median age, weight, mean indexed shunt size (1.19 vs 1.15 mm/kg, p = 0.2), and the number of patients with antegrade pulmonary flow. Interim mortality was 7% (n = 15), and younger age ( p = 0.03), and group T ( p = 0.03) were independent risk factors for death prior to second-stage surgery. Absence of antiplatelet agents or anticoagulants was not a risk factor for interim mortality. Conclusions: In-hospital mortality and longer ventilation time after SPS by sternotomy may be related to pulmonary over circulation due to shunt insertion origin and/or size, and pathologic features. Early and interim outcomes can be improved by using a smaller shunt or changing the SPS insertion origin when using a sternotomy approach. # 2007 European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved. Keywords: Systemic-pulmonary shunt; Pulmonary atresia; Pulmonary blood flow 1. Introduction Despite the progressively higher number of patients who undergo a complete primary repair for complex congenital heart disease, many children with variable anatomy are not amenable to total corrective surgery and will benefit from a systemic-to-pulmonary shunt (SPS) as either a sole or an additional source of pulmonary blood flow [1 3]. Palliation with a SPS provides time for the pulmonary vascular resistance to drop and for pulmonary arteries (PA) growth before a more definitive cardiac repair. Although usually effective, there are still potential risks of complications [4,5] and death during the postoperative and interim period [6,7]. The purpose of the present study was to evaluate the early outcomes and interim mortality by using multiple risk factor Presented at the 21st Annual Meeting of the European Association for Cardio-thoracic Surgery, Geneva, Switzerland, September 16 19, 2007. * Corresponding author. Address: McGill University Health Center, Cardiac Surgery Department, 687 Pine Avenue West, Montreal, QC, Canada H3A 1A1. Tel.: +1 514 843 1463; fax: +1 514 843 1602. E-mail address: siamakmohammadi@yahoo.com (S. Mohammadi). analysis to assess whether we could still improve the outcomes of patients undergoing SPS shunt in cyanotic congenital cardiac malformations. 2. Patients and methods Between January 1988 and April 2005, 226 patients underwent an isolated SPS using a polytetrafluoroethylene (Gore-Tex) shunt (W.L. Gore Inc., Elkton, MD, USA) without cardiopulmonary bypass for decreased pulmonary blood flow (PBF) at British Columbia Children s hospital. The median age and weight at SPS insertion was 17 days (range 1 day to 219 months old) and 3.4 kg (range 1.8 30.0 kg), respectively. Indications for surgical palliation were pulmonary stenosis/ atresia (PS/PA) with ventricular septal (VSD)/tetralogy of Fallot in 124 patients (54.9%), functional single ventricle anatomy-ps/pa in 35 patients (5.5%), PA with intact septum in 31 patients (13.7%), transposition of the great arteries (TGA)- VSD-PS/PA in 30 patients (13.3%), and double outlet right ventricle-ps/pa in 6 patients (2.6%). The final diagnosis and 1010-7940/$ see front matter # 2007 European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.ejcts.2008.06.001

546 S. Mohammadi et al. / European Journal of Cardio-thoracic Surgery 34 (2008) 545 549 the presence of coexisting malformations have been confirmed by echocardiography (n = 135 patients, 59.7%), and/ or both echocardiography and cardiac catheterism (n = 91 patients, 40.3%). The associated malformations were: right aortic arch (n = 27), patent ductus arteriosus (n = 24), severe hypoplastic right ventricle (n = 11), left-tga (n = 4), and atrio-ventricular canal (n = 4). The surgical approach was by thoracotomy (group T) in 180 patients (79.6%) and sternotomy (group S, performed since 1997) in 46 patients (20.4%). The systemic side of the shunt insertion was either the innominate artery (n =38)or the ascending aorta (n =8)ingroupS,andtheright(n = 151) and left (n = 29) subclavian artery in group T. The site of the distal anastomosis was: (a) the right pulmonary artery (n = 189) in patients with proximal anastomosis from the innominate and or right subclavian artery, (b) the left pulmonaryartery(n = 29) in patients with proximal anastomosis from the left subclavian artery, and finally (c) the main pulmonary artery (n = 8) in all patients with proximal anastomosis from the ascending aorta. The median shunt size was 4.0 mm (range 3.0 6.0 mm). The mean indexed shunt size (mm/kg) was 1.16 0.37. Type of surgical approach and shunt diameters were the surgeon s preference (Four surgeons in total during the study). Examined variables included cardiac diagnosis, age at operation, weight at operation, the year of operation (before or after 1996), the presence of antegrade PBF, preoperative urgent intubation for any reason, type of surgical approach, indexed shunt size, surgeon (A, B, C, D), duration of intubation, intensive care unit (ICU) stay, peripheral oxygen saturation (SaO 2 ) at discharge, hospital stay, need for reoperation for shunt-related complications (thrombosis or narrowing), ASA use after discharge, in-hospital and interim mortality (death after hospital discharge and before complete repair operation). Shunt thrombosis or narrowing was defined by: (a) need for an urgent repeated SPS as a result of prohibitive cyanosis; (b) absence of murmur in conjunction with increasing cyanosis; (c) absence and or decrease of flow on echo Doppler, angiography, computerized tomography, or magnetic resonance imaging in conjunction with increasing cyanosis; (d) surgical or autopsy finding. All inpatient and outpatient medical records were reviewed. In cases in which an autopsy was performed, the report was reviewed, and correlated with clinical information. 2.1. Statistical analysis Data were collected retrospectively. Summary statistics were expressed using mean 1 standard deviation for continuous variable or as percentage for categorical data and median. According to the distribution of continuous data, Student s t-test or Wilcoxon s rank-sum test was used to compare groups (T and S). The analysis of categorical variables was performed using Fisher s exact test. Multivariate analysis was performed using stepwise logistic regression including only factors identified to be significant during univariate analysis ( p < 0.15) to assess the independent risk factors for in-hospital and interim mortality. The data were analyzed using the statistical package program SAS 9.1 (SAS Institute Inc., Cary, NC). A p-value less than 0.05 was considered significant. 3. Results The patient characteristics and examined variables compared in patients with group S and T approach are depicted in Table 1. Median intubation time was significantly longer in group S patients (112 h; range 13 1776 h), compared to group T (30 h; range 0 extubation in operating room to 454 h) respectively, ( p < 0.0001). Median duration of ICU stay was also significantly longer in group S patients (148 h; range 20 1874 h) compared with 54 h (range 16 521 h) for group T ( p < 0.0001). Median intubation time and ICU stay was also significantly higher in patients who were intubated preoperatively ( p = 0.02), and patients with higher indexed shunt size (>1.20 mm/kg, p = 0.03). The two groups were comparable in terms of the number of patients with different morphologies, age, the number of neonates, body weight, the number of patients with antegrade pulmonary flow, the mean indexed shunt size, and the O2 saturation at discharge (Table 1). Retrospective analysis of the surgical approach demonstrated factors leading mainly (60% of cases) to the choice of a sternotomy approach were functional single ventricle anatomy-ps/pa, and surgeon C. The main factors (60% of cases) leading to choice of a thoracotomy approach were PS/ PA with VSD/tetralogy of Fallot, and surgeon A, B. Retrospective analysis also showed that factors leading mainly (60% of cases) to the choice of small shunt size (3.5 mm) were neonates, low body weight (3.5 kg), PA-IS pathology, functional single ventricle anatomy-ps/pa, and operations undertaken more recently (between 1998 and 2005). The factors leading to choice of large shunt size (>3.5 mm) levels were higher body weight (>3.5 kg), PS/PA with VSD/tetralogy of Fallot, surgeon B, and operations undertaken during earlier years (between 1988 and 1997). 3.1. In-hospital mortality There were 13 in-hospital deaths (5.7%), which was significantly higher ( p = 0.03) in group S patients (n =7, 15.2%) compared to group T patients (n = 6, 3.3%). The reasons of death are depicted in Table 2. Univariate analysis revealed younger age ( p = 0.01), lower body weight ( p =0.04), a diagnosis of PA with intact septum ( p = 0.005), preoperative intubation ( p=0.03), length of intubation ( p < 0.0001), longer ICU stay ( p < 0.0001), and group S ( p = 0.03) as risk factors for in-hospital death. In the multivariate analysis, Table 1 Patient characteristics and examined variables compared in sternotomy (group S) and thoracotomy (group T) patients Group S (n = 46) Group T (n = 180) p value Age (median) 13 days 22 days 0.3 Neonates (n, %) 13, 28.3% 44, 24.4% 0.2 Weight (median) 3.5 4.1 0.3 Preoperative intubation (n, %) 4, 8.7% 17, 9.4% 0.4 Antegrade pulmonary flow (n) 30 121 0.4 Indexed shunt size (mm/kg) 1.19 1.15 0.2 Intubation time (median; h) 112 30 <0.0001 ICU stay (median; h) 148 54 <0.0001 Shunt complication 6.1% 5.2% 0.2 O 2 saturation at discharge (% SD) 85.2 4.8 84.0 5.6 0.3

S. Mohammadi et al. / European Journal of Cardio-thoracic Surgery 34 (2008) 545 549 547 Table 2 In-hospital death in both groups Patient no. Main diagnosis Reason of death Stenotomy group (group S, n = 46) 1 PA-IS ARDS 2 Single ventricle- PS/PA Low cardiac output 3 PA-IS Sepsis 4 TGA-VSD-PS/PA Necrotizing pneumonitis 5 PA-IS Myocardial infarction 6 DORV- PS/PA Severe pneumonia 7 Single ventricle-ps/pa Low cardiac output Thoracotomy group (group T, n = 180) 1 TGA-VSD-PS/PA Sepsis 2 PA-IS Sudden death 3 ToF/PA Necrotizing enterocolitis 4 Single ventricle-ps/pa Low cardiac output 5 PA-IS Myocardial infarction 6 DORV-PS/PA Sepsis PA-IS; pulmonary atresia with intact septum, ARDS; acute respiratory distress syndrome, TGA-VSD-PS/PA; transposition of the great arteries-ventricular septal defect-pulmonary stenosis/ atresia, DORV; double outlet right ventricle, ToF; tetralogy of Fallot. younger age ( p = 0.04), PA with intact septum ( p = 0.01), and length of intubation ( p =0.01) were independent risk factors of in-hospital mortality. 3.2. Interim mortality Among the survivors discharged from hospital (n = 213), 15 patients (7%) died before complete repair in the interim period. All deaths happened in group T. Among them, 10 patients had small (3.5 mm) shunt size (4 patients with 3 mm and 6 patients with 3.5 mm shunt size), and 5 patients had large (>3.5 mm) shunt size (2 patients with 4 mm, 2 patients 5 mm, and 1 patients with 6 mm shunt size). The nature of death was sudden in the majority of our patients (n = 12, 80%), 9 out of which (75%) had small shunt size (3.5 mm). Two patients had admission for severe pulmonary infection; both were in hospital at the time of death. One premature neonate died from sepsis. Among them univariate analysis revealed younger age ( p=0.03), a diagnosis of PA with intact septum ( p=0.04), group T ( p=0.03), and small (3.5 mm) shunt size ( p=0.04) as risk factors for death prior to second-stage or complete repair surgery. In the multivariate analysis, younger age ( p = 0.03), and group T ( p=0.03) were independent risk factors of interim mortality. ASA had been started in 128 survivors from hospital (60.1%), which was not significantly different in the 2 groups ( p = 0.4). Absence of antiplatelet therapy was not a risk factor for interim mortality. Autopsy has been performed in nine patients in whom the diagnosis of shunt thrombosis was confirmed. 3.3. Early and interim shunt-related complications There were 19 re-operations in the early (n = 12; 8 in group T, and 4 in group S) and interim (n = 7; all in group T) period. There were 12 episodes (5.3%) of early shunt-related complications, with 8 (4.4%) episodes in group T and 4 (8.7%) episodes in group S ( p = 0.3). Four out of 12 patients (33.3%) with early shunt-related complications (thrombosis) died during hospitalization, 2 patients in each group ( p = 0.2). All four patients had undergone surgical revision of the shunt. All other early shunt-related complications except one (six episodes in group T and one episode in group S) were thrombosis or stenosis that needed shunt revision and/or replacement. There was one patient of group S who underwent reduction of the shunt diameter with the use of a metallic clip on day 3 postoperation for excessive PBF and cardiac failure. Univariate analysis revealed younger age ( p = 0.03) and lower body weight ( p = 0.04) as risk factors for early shunt-related complications. Surgical approach was not a risk factor for early shunt thrombosis or stenosis ( p=0.3). In this interim period among the survivors discharged from the hospital, there were seven shunt re-operations. None of them were in group S. Five of those patients had another SPS on the opposite side in order to delay the complete repair between 1988 and 1993, and two patients had SPS for early shunt stenosis. The interval between the first SPS and complete repair were 19.9 9.6 months and 22.3 17.3 months for group S and T, respectively ( p=0.04). The pulmonary artery required repair at the site of the SPS shunt in 27 patients (11.9%). There was a trend towards having more PA distortion needing repair in patients with thoracotomy (n = 23, 12.7%), compared to patients with sternotomy (n = 4, 8.7%) SPS ( p = 0.12) during complete repair. 4. Discussion The systemic to pulmonary artery shunt is an effective surgical therapy for initial palliation of cyanotic congenital heart disease. With improvements in surgical, anesthetic, and intensive care management, outcomes after palliative surgery have improved. As the knowledge of patient outcomes with a SPS has increased, efforts to improve overall results have focused on early and interim mortality. 4.1. Early outcomes In the present study, univariate analysis showed that general patient characteristics (younger age, and lower body weight), original pathologic feature (in particular patients with a PA-intact septum), surgical approach (sternotomy with SPS originating from the innominate artery and/or ascending aorta), and finally complicated postoperative course with longer intubation periods and ICU stay have an impact on inhospital death. However only the PA-intact septum, longer intubation period, and ICU stay were independent risk factors for in-hospital mortality. Some investigators [3,8] have studied the negative impact of younger age, and lower body weight in the early outcome of palliation with SPS. The higher mortality rate in these younger patients with lower body weight could be related to the small size of the pulmonary arteries. Weight at the time of the procedure has been identified by all of the larger studies as an independent risk factor [9,10]. However, a number of investigators have demonstrated early primary repair of cyanotic congenital defects with good results [11 13]. The staged approach remains a surgical option in young patients in whom definitive repair may not be possible.

548 S. Mohammadi et al. / European Journal of Cardio-thoracic Surgery 34 (2008) 545 549 It is generally agreed that in patients with a PA and intact septum, with a small right ventricle and/or ventricularcoronary artery fistulae, palliation with SPS may be the first surgical approach. In patients with this rare pathology, there are a variety of anatomic subtypes and considerable morphologic heterogeneity [14,15]. Patients with the severe form of PA and intact septum are at increased risk of death despite the type of treatment. Fenton et al. [16] showed that the interim mortality after palliation with SPS in infants with PA and intact septum and severe right ventricle hypoplasia is 24%. In the present study, PA and intact septum was an independent risk factor for in-hospital mortality in which severe right ventricle hypoplasia was present in 61.3% of patients. However, it was not a risk factor for interim mortality. The latter may be related to the small number of survivors in this subgroup of patients. The surgical approach and technique might be one of the most important factors to modify for improving the early outcome. In our study of univariate analysis, the sternotomy approach was a risk factor for in-hospital mortality despite the similar median age, the number of neonates, body weight, the mean indexed shunt size, and the number of patients with antegrade pulmonary flow, compared with the thoracotomy group. However, this was not confirmed in the multivariate analysis. Group S patients experienced significantly longer median intubation time and ICU stay than group T patients. Group S patients had a more proximal insertion shunt site than group T patients. The characteristics of the shunt on pulmonary blood flow and the influence of shunt site on negative early outcome risk are complex. The length and diameters of the graft are also determinants of blood flow through the shunt. The longer median intubation time and ICU stay in patients with sternotomy may be related to the pulmonary over circulation. The use of the innominate artery and/or ascending aorta as the site of shunt insertion renders the length of the graft shorter, and puts the distal anastomosis closer to the PA bifurcation. However, the latter allows even distribution of blood flow to both lungs and potentially better PA growth, but the combination of both may allow for increased pulmonary blood flow. These shunt-related physiologic factors may contribute to in-hospital mortality. The use of smaller shunt size and more distal insertion site in group T patients may result in less pulmonary blood flow. We recommended use of a 3.5 (in cases of small pulmonary arteries) or a 4.0 mm graft for full term infants (up to 4 kg) and a 3.0 mm graft for premature babies weighing less than 3.0 kg, even if the shunt comes from the innominate artery. Because most of these children go on to a complete repair at 6 12 months of age, the shunt flow is adequate and not excessive. The more favorable outcome in patients with smaller shunt size is consistent with previous studies in neonates [3], patients with tricuspid atresia [17], and canine models of univentricular hearts [18]. 4.2. Interim outcomes Interim mortality, defined as death between the first stage of palliation and complete surgical repair, has an important impact on overall outcome of patients with cyanotic congenital disease undergoing SPS. In the present study, univariate analysis showed that general characteristics of the patients (younger age, and lower body weight), smaller shunt size (3.5 mm), and surgical approach (thoracotomy approach) were risk factors for interim mortality, however only the younger age and thoracotomy approach were independent risk factors for in-hospital mortality. These data demonstrated that infants with a SPS would have an increased risk of interim mortality (7%), which is comparable with interim death in patients with hypoplastic left heart syndrome [19,20]. Thoracotomy approach was an independent risk factor for interim death. Determination of the mechanism by which thoracotomy approach portends a poor interim outcome, is beyond the scope of this observational study. However, a possible contributing factor deserves mention. The longer length of the graft and the use of a smaller insertion site of both systemic and pulmonary sites may make SPS more susceptible to thrombosis and death. The overall mortality (in-hospital plus interim mortality) are comparable in both sternotomy and thoracotomy groups. Using a sternotomy approach with a smaller shunt size to avoid pulmonary over circulation may improve the interim mortality by preventing a thoracotomy approach. More than 65% of the interim mortality happened in patients with a thoracotomy approach and a small shunt size (3.5 mm), so thus the use of a bigger shunt size in patients with the thoracotomy approach may also avoid shunt thrombosis and improve interim mortality. Pulmonary over circulation may also be avoided by a smaller diameter of the subclavian artery to some extent. In nine autopsies performed in patients that died during the interim period, shunt thrombosis was identified in all of them. The nature of death was sudden in the majority of our patients (80%). In another study [7], interim deaths were unexpected in 81% of patients who were doing well. Our study fails to demonstrate any beneficial effect of aspirin to prevent shunt thrombosis, probably due to the insufficient number of patients. Aspirin has been used in approximately 65% of our patients during the interim period. The use of anticoagulation in patients with SPS is a subject of debate. Li et al. [21] in a multicenter study showed infants palliated with SPS to have a poor outcome. In this study, the use of aspirin appeared to be associated with a reduction in shunt thrombosis and death, which was not confirmed in other studies. Investigations into additional mechanisms of antiplatelet and antithrombotic therapies in the interim period are needed. Considerations also must be given to planning the complete repair surgery as early as possible. 4.3. Study limitations Limitations of our study include its retrospective nature, the wide variety of anatomic subtypes, and the absence of autopsies in some patients. The data do not allow us to extract more specific, concrete, or more beneficial measures or strategies that might favorably impact survival in this challenging group of patients. 4.4. Conclusion Our data demonstrate that patients with SPS have a significant mortality rate before undergoing complete repair.

S. Mohammadi et al. / European Journal of Cardio-thoracic Surgery 34 (2008) 545 549 549 Patients with younger age, lower body weight and pulmonary atresia and intact septum diagnosis are more susceptible to in-hospital death. Patients with a thoracotomy approach have higher interim mortality. Deploying smaller shunt and/ or changing systemic to pulmonary shunt insertion origin through sternotomy may improve in-hospital and interim outcome. The latter technique may also avoid future pulmonary arteries distortion. References [1] Lawless CE, Smith EEJ, Hallidie-Smith K, Sapsford RN. The modified Blalock-Taussig shunt using microporous expanded polytetrafluoroethylene (PTFE). J Thorac Cardiovasc Surg 1982;23:287 92. [2] Tamisier D, Vouhé PR, Vernant F, Leca F, Massot C, Neveux JY. Modified Blalock-Taussig shunts: results in infants less than 3 months of age. Ann Thorac Surg 1990;49:797 801. [3] Alkhulaifi AM, Lacour-Gayet F, Serraf A, Belli E, Planché C. Systemic pulmonary shunts in neonates: early clinical outcome and choice of surgical approach. Ann Thorac Surg 2000;69:1499 504. [4] Gladman G, McCrindle BW, Williams WG, Freedom RM, Benson LN. The modified Blalock-Taussig shunt: clinical impact and morbidity in Fallot s tetralogy in the current era. J Thorac Cardiovasc Surg 1997;114: 25 30. [5] Okita Y, Miki S, Kusuhara K, Ueda Y, Tahata T, Yamanaka K, Tamura T. Acute pulmonary edema after Blalock-Taussig anastomosis. Ann Thorac Surg 1992;53:684 5. [6] Fermanis GG, Ekangaki AK, Salmon AP, Keeton BR, Shore DF, Lamb RK, Monro JL. Twelve year experience with the modified Blalock-Taussig shunt in neonates. Eur J Cardiothorac Surg 1992;6:586 9. [7] Fenton KN, Siewers RD, Rebovich B, Pigula FA. Interim mortality in infants with systemic-to-pulmonary artery shunts. Ann Thorac Surg 2003;76: 152 6. [8] Ibawi MN, Grieco J, DeLeon SY, Idriss FS, Muster AJ, Berry TE, Klich J. Modified Blalock-Taussig shunt in newborn infants. J Thoracic Cardiovasc Surg 1984;88:770 5. [9] Daubeney PE, Wang D, Delany DJ, Keeton BR, Anderson RH, Slavik Z, Flather M, Webber SA. UK and Ireland Collaborative Study of Pulmonary Atresia with Intact Ventricular Septum. Pulmonary atresia with intact ventricular septum: predictors of early and medium-term outcome in a population-based study. J Thorac Cardiovasc Surg 2005;130:1071 708. [10] Ashburn DA, Blackstone EH, Wells WJ, Jonas RA, Pigula FA, Manning PB, Lofland GK, Williams WG, McCrindle BW, Congenital Heart Surgeons Study members. Determinants of mortality and type of repair in neonates with pulmonary atresia and intact ventricular septum. J Thorac Cardiovasc Surg 2004;127:1000 8. [11] van Dongen EI, Glansdorp AG, Mildner RJ, McCrindle BW, Sakopoulos AG, VanArsdell G, Williams WG, Bohn D. The influence of perioperative factors on outcomes in children aged less than 18 months after repair of tetralogy of Fallot. J Thorac Cardiovasc Surg 2003;126:703 10. [12] Lee JR, Kim JS, Lim HG, Hwang HY, Kim YJ, Rho JR, Ahn C. Complete repair of tetralogy of Fallot in infancy. Interact Cardiovasc Thorac Surg 2004;3:470 4. [13] Hirsch JC, Mosca RS, Bove EL. Complete repair of tetralogy of Fallot in the neonate: results in the modern era. Ann Surg 2000;232:508 14. [14] Daubeney PE, Delany DJ, Anderson RH, Sandor GG, Slavik Z, Keeton BR, Webber SA. United Kingdom and Ireland Collaborative Study of Pulmonary Atresia with Intact Ventricular Septum. Pulmonary atresia with intact ventricular septum range of morphology in a population-based study. J Am Coll Cardiol 2002;39:1670 9. [15] Anderson RH, Anderson C, Zuberbuhler JR. Further morphologic studies on hearts with pulmonary atresia and intact ventricular septum. Cardiol Young 1991;1:105 13. [16] Fenton KN, Pigula FA, Gandhi SK, Russo L, Duncan KF. Interim mortality in pulmonary atresia with intact ventricular septum. Ann Thorac Surg 2004;78:1994 8. [17] Karamlou T, Ashburn DA, Caldarone CA, Blackstone EH, Jonas RA, Jacobs ML, Williams WG, Ungerleider RM, McCrindle BW, Members of the Congenital Heart Surgeons Society. Matching procedure to morphology improves outcomes in neonates with tricuspid atresia. J Thorac Cardiovasc Surg 2005;130:1503 10. [18] Kitaichi T, Chikugo F, Kawahito T, Hori T, Masuda Y, Kitagawa T. Suitable shunt size for regulation of pulmonary blood flow in a canine model of univentricular parallel circulations. J Thorac Cardiovasc Surg 2003; 125:71 8. [19] Ghanayem NS, Hoffman GM, Mussatto KA, Cava JR, Frommelt PC, Rudd NA, Steltzer MM, Bevandic SM, Frisbee SS, Jaquiss RD, Litwin SB, Tweddell JS. Home surveillance program prevents interstage mortality after the Norwood procedure. J Thorac Cardiovasc Surg 2003;126:1367 77. [20] Ashburn DA, McCrindle BW, Tchervenkov CI, Jacobs ML, Lofland GK, Bove EL, Spray TL, Williams WG, Blackstone EH. Outcomes after the Norwood operation in neonates with critical aortic stenosis or aortic valve atresia. J Thorac Cardiovasc Surg 2003;125:1070 82. [21] Li JS, Yow E, Berezny KY, Rhodes JF, Bokesch PM, Charpie JR, Forbus GA, Mahony L, Boshkov L, Lambert V, Bonnet D, Michel-Behnke I, Graham TP, Takahashi M, Jaggers J, Califf RM, Rakhit A, Fontecave S, Sanders SP. Clinical outcomes of palliative surgery including a systemic-to-pulmonary artery shunt in infants with cyanotic congenital heart disease: does aspirin make a difference? Circulation 2007;116:293 7. Appendix A. Conference discussion Dr M.A. Navabi (Tehran, Iran): I think in the sternotomy group it s very important to clarify the origin of the aortic site, because if you do a sternotomy and do a right Gore-Tex shunt, you may put the proximal end on the ascending aorta or you may put it on the brachiocephalic, and that determines the flow of the shunt. So in your sternotomy group you probably have to clarify the site of aortic anastomosis. Dr Mohammadi: In the majority of patients with sternotomy, the insertion site was the innominate artery, however, in eight patients it was the ascending aorta. In our series, these patients experienced a longer postoperative intubation period and ICU stay, but whether it was really related to the pulmonary overcirculation cannot be explained by the present study. I, personally, believe systemic to pulmonary shunts with a more proximal insertion site have more pulmonary blood flow. However, the flow through the modified BT shunt is a complex issue, and for a given diameter of prosthesis, it is also dependent on the length of graft and distal bed, which is different in the different types of surgery, and pathologies. Dr W. Daenen (Linden, Belgium): Based on your results now, what size would you recommend for a 4 kg patient treated with a systemic shunt from the brachiocephalic artery to the pulmonary artery? Dr Mohammadi: I think the best recommendations are to perform the procedure through a median sternotomy, to use a smaller shunt size that corresponds to the weight of the patient, and to perform the graft as distal as possible in the systemic arterial tree towards the subclavian artery. The shunt size is one of the factors which plays a role in pulmonary blood flow. Dr C. Sebening (Heidelberg, Germany): Did you heparinize or anticoagulate in the mid or interim phase? Did you anticoagulate your shunt? Dr Mohammadi: About 60% of our patients received aspirin, which was equally distributed in both groups, and the absence of using aspirin was not a risk factor for shunt-related complications and death. However, the number of patients is not enough to make a statement in this regard. Dr Sebening: If you say 65%, what were the criteria for the others not to receive it? Dr Mohammadi: There were no strict criteria due to the very long duration of the study. During this period, which started in 1988 and ended in 2005, the practice evolved in terms of the use of anticoagulation and the use of aspirin in the interim phase.