Improved results after the primary repair of interrupted aortic arch: impact of a new management protocol with isolated cerebral perfusion,

Transcription

1 European Journal of Cardio-thoracic Surgery 38 (2010) Improved results after the primary repair of interrupted aortic arch: impact of a new management protocol with isolated cerebral perfusion, Tomas Tlaskal *, Pavel Vojtovic, Oleg Reich, Bohumil Hucin, Roman Gebauer, Vladimir Kucera Kardiocentrum and Cardiovascular Research Center, University Hospital Motol, Prague, Czech Republic Received 29 September 2009; received in revised form 8 January 2010; accepted 18 January 2010; Available online 6 March 2010 Abstract Objective: The aim of this retrospective study was to analyse results and risk factors of death after the repair of the interrupted aortic arch, and to compare results obtained with the use of deep hypothermic circulatory arrest versus isolated cerebral perfusion. Methods: The primary repair of the interrupted aortic arch and associated heart lesions was performed in 50 consecutive patients. The median age was 5 days and the mean weight was kg. The interrupted aortic arch was of type A in 12 (24%) patients, type B in 37 (74%) and type C in one (2%) patient. Ventricular septal defect was present in 48 (96%) patients, subaortic stenosis in 15 (30%), truncus arteriosus in 14 (28%), transposition of the great arteries in two (4%), aortopulmonary window in two (4%) and double-outlet right ventricle in one (2%). The surgery consisted of reconstruction of the aortic arch by direct anastomosis and repair of associated heart lesions. In 25 (50%) patients, aortic arch reconstruction was performed using hypothermic circulatory arrest (group I) and in 25 by isolated cerebral perfusion (group II). The duration of cardiopulmonary bypass, aortic cross-clamping and circulatory arrest or isolated cerebral perfusion was min, min and min, respectively, in group I; and min, min and 31 6 min, respectively, in group II. Results: There were 10 (20%) deaths in this series, eight (32%) in group I and two (8%) in group II. Out of 12 patients operated before 1995, seven (58%) patients died; and out of 38 patients operated between 1995 and 2009, three (8%) patients died ( p = 0.008). By Cox multifactorial analysis, the earlier date of operation represented the only risk factor of death ( p = 0.037). Twelve (71%) survivors in group I and five (22%) survivors in group II required re-intervention, most often for subaortic stenosis, aortic arch obstruction or conduit obstruction. All patients remain in the New York Heart Association (NYHA) class I or II at median 12.6 years in group I, and 1.7 years in group II, respectively, after surgery. Conclusions: Interrupted aortic arch can be repaired in neonates with a mortality of 5 10%. The results depend on experience. Isolated cerebral perfusion was joined with decreased mortality but it did not influence the occurrence of neurological complications. # 2010 European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved. Keywords: Congenital heart defects; Interrupted aortic arch; Primary repair in neonates; Isolated cerebral perfusion; Long-term results 1. Introduction Interrupted aortic arch (IAA) represents a critical ductusdependent congenital heart disease characterised by complete luminal interruption of the aortic arch [1].This lesion is often associated with ventricular septal defect (VSD), subaortic stenosis (SAS) or persistent truncus arteriosus (PTA). The haemodynamics depends on patency of ductus arteriosus (PDA) as the only source of blood flow to the lower part of the body, and on associated heart lesions. The primary repair of IAA and VSD was first reported by Trusler and Izukawa in 1970 [2]. The mortality and long-term results after repair of IAA depend on morphology, clinical status, surgical methods Presented at the 23rd Annual Meeting of the European Association for Cardio-thoracic Surgery, Vienna, Austria, October 18 21, Funding: Supported by a research project of the University Hospital Motol: MZOFNM2005. * Corresponding author. Address: Kardiocentrum, University Hospital Motol, V Uvalu 84, Prague 5, Czech Republic. Tel.: ; fax: address: tomas.tlaskal@lfmotol.cuni.cz (T. Tlaskal). and experience [3 7]. We reported our initial experience with the primary and two-stage repair of IAA and associated heart lesions in 1998 [8]. Until that time, the mortality rate and the rate of complications were rather high in our centre. We therefore concentrate on improving our management protocol and also introduce several new methods including an isolated cerebral perfusion (ICP), which could theoretically improve protection of the brain and other organs by maintaining perfusion through the collateral circulation during an aortic arch reconstruction [9 12]. We started to use ICP in 1997 and reported our preliminary experience with this method in 2005 [13]. In this retrospective study, we analyse results after the primary repair of IAA and associated heart lesions in all consecutive patients operated on in one centre. 2. Patients and methods 2.1. Patients Between May 1990 and April 2009, the primary biventricular repair of IAA and associated heart lesions was /$ see front matter # 2010 European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved. doi: /j.ejcts

2 T. Tlaskal et al. / European Journal of Cardio-thoracic Surgery 38 (2010) Table 1 Basic characteristics of the patient population. Lesion Group I Group II Period Number of patients (n) Age (days; median, range) 5 (1 85) 6 (1 36) Weight (kg) ( ) ( ) Male/female ratio (n, %) 14/11 (56%/44%) 14/11 (56%/44%) Severe preoperative condition (n, %) 7 (28%) 7 (28%) DiGeorge syndrome (n, %) 9 (36%) 10 (40%) Non-cardiac lesions (n, %) 2 (8%) 5 (20%) IAA type A (n, %) 4 (16%) 8 (32%) IAA type B (n, %) 21 (84%) 16 (64%) IAA type C (n, %) 0 1 (2%) PTA (n, %) 6 (24%) 8 (32%) SAS (n, %) 7 (28%) 8 (32%) TGA (n,%) 2 (8%) 0 Cardiopulmonary bypass (min; mean, range) ( ) ( ) Aortic cross-clamping (min; mean, range) (58 162) (39 125) Circulatory arrest/isolated perfusion (min; mean, range) (18 68) 31 6 (22 45) Temperature (8C; mean, range) 17 2 (14 20) 22 3 (17 27) Open chest (n, %) 23 (92%) 18 (72%) Abbreviations: IAA interrupted aortic arch; PTA persistent truncus arteriosus; SAS subaortic stenosis; TGA transposition of the great arteries. performed in 50 consecutive patients at Kardiocentrum, University Hospital Motol in Prague (Table 1). The median age at operation was 5 days (range 1 85 days) and the mean weight was kg (range kg). IAA was of type A in 12 (24%) patients, type B in 37 (74%) and C in one (2%) patient. Forty-eight (96%) patients had VSD, 15 (30%) SAS, 14 (28%) PTA, two (4%) transposition of the great arteries (TGA), two (4%) aortopulmonary window (APW) and one (2%) doubleoutlet right ventricle (DORV). In 19 (38%) patients, the DiGeorge syndrome was identified. In seven (14%) patients, non-cardiac congenital anomalies were found: anorectal atresia, intestinal malrotation with a Meckel diverticle, hydronephrosis, laryngeal malformation, chondrodystrophy with multiple skeletal malformations, coxa vara and pes calcaneovalgus in one patient each. In all patients, the diagnosis was set up by echocardiography, but in nine (18%) patients, heart catheterisation with angiocardiography was also required. Forty-eight (96%) patients received prostaglandins before surgery. Before the diagnosis of IAA was set up, the clinical condition of 16 (32%) neonates deteriorated and required intubation and mechanical ventilation (13 patients) and/or treatment of renal failure (five patients), treatment of infection (two patients) or cardiopulmonary resuscitation (two patients). Early in our experience, one patient with IAA and PTA, who was admitted in a moribund condition, and in whom it was not possible to stabilise the haemodynamics, was indicated for surgery on the day of admission. In another patient, who also had congenital intestinal malformation, an urgent abdominal surgery was required for signs of an acute abdomen before repair of IAA with VSD Surgical technique Our strategies and techniques have been already described [8,13]. In this article, we would like to summarise the basic principles of our management protocol. An infusion of prostaglandins is introduced when the diagnosis of IAA is made. Before surgery, all complications are treated. After stabilisation, the primary repair of IAA and associated heart lesions is performed through the midline sternotomy approach. Neonatal probes for continuous near-infrared spectroscopy (NIRS) monitoring of cerebral oxygen saturation are placed on the forehead. An arterial line is inserted into the most appropriate artery, usually the right radial artery. The thymus is subtotally excised. Two fine (2.0 or 2.6 mm) aortic cannulas are introduced. The first one is introduced directly into the right innominate or carotid artery. The second aortic cannula is introduced into the ductus arteriosus or into the aortic root, if continuous myocardial perfusion is also intended. The two arterial cannulas are connected using a Y connector to the arterial line. Two 12-Fr venous cannulas are inserted transatrially. The principles of the insertion of cannulas for ICP are shown in Fig. 1. Earlier in our experience, deep hypothermic circulatory arrest (DHCA) at C was used for arch reconstruction. Today, mid-hypothermic continuous ICP in combination with continuous perfusion of the myocardium at C is preferred. On bypass, it is necessary to snare both pulmonary arteries and insert a vent into the left atrium. During the cooling period, the ascending aorta, the aortic arch and its branches, the ductus arteriosus and the descending aorta must be extensively mobilised. After cooling, DHCA is instituted and ice bags applied on the head. Alternatively, ICP is maintained via the cannula in the innominate artery at 1/5 1/3 of the calculated full flow (30 50 ml m 2 min 1 ) so as to maintain the mean arterial pressure at mmhg. For aortic arch reconstruction, it is important to resect all the ductal tissue, which often surrounds subclavian arteries. We do therefore transect the left subclavian artery unless good arterial tissue is present and the anastomosis can be performed safely and without tension. We believe that it is very important to also transect the anomalous right subclavian artery, which could preclude adequate mobilisation and construction of a non-obstructive anastomosis. Direct end-to-side anastomosis between descending and ascending aorta is almost always possible. If the ascending aorta is narrow, or the aortic wall is too thin or fragile or if it is not possible to excise all the ductal tissue and

3 54 T. Tlaskal et al. / European Journal of Cardio-thoracic Surgery 38 (2010) Table 2 Surgical procedures (n = 50). Total n % Procedures for reconstruction of the aortic arch Anastomosis of descending and ascending aorta Anastomosis with patch augmentation of the aorta 6 12 Reconstruction using a vascular graft 1 2 Procedures for correction of associated heart lesions Patch closure of ventricular septal defect Repair of persistent truncus arteriosus Myectomy or myotomy of the connal septum Plasty of the truncal or aortic valve 3 6 Plasty of the aortic root in persistent truncus arteriosus 3 6 Patch plasty of pulmonary arteries 3 6 Yasui operation for severe subaortic stenosis 2 4 Excision of a subaortic membrane 2 4 Arterial switch operation 2 4 Repair of aortopulmonary window 2 4 Repair of double-outlet right ventricle 1 2 Reconstruction of the right pulmonary artery using a graft 1 2 using Cox multivariate logistic regression analysis. An actuarial survival was set up using Kaplan Meier estimation. Fig. 1. Schema of cannulation for primary repair of interrupted aortic arch with the use of isolated cerebral perfusion for aortic arch reconstruction. Legend: Aortic cannulas were inserted into the brachiocephalic trunk and ductus arteriosus. After cooling to 25 8C, continuous isolated cerebral perfusion was instituted. Continuous perfusion of the myocardium is permitted by applying a partially beating clamp for aortic arch reconstruction. when the anastomosis is under tension, the anastomosis, including adjacent part of the ascending and descending aorta, is enlarged using an autologous pericardial patch pretreated in a glutaraldehyde solution. The anastomosis is performed with 7/0 polypropylene monofilament suture. The suture line is treated using a biological glue. When the ascending aorta is not hypoplastic, it is sometimes possible to perform aortic arch reconstruction on a beating heart, using a side-beating clamp. This permits to maintain continuous flow to the aortic root and coronary arteries from the cannula in the innominate artery. Under these settings, the second aortic cannula may not be required. When the aortic arch reconstruction is completed, a standard cross-clamp is applied and crystalloid cardioplegia instituted. Associated lesions are corrected. Surgical procedures are summarised in Table 2. The chest is usually left open for 1 3 days Data collection and statistics For this retrospective study, all hospital and outpatient department records were examined. Data for approximately 50% of the patients of this series had been reported earlier [6,11]. The data collection was complete because all patients have been followed up in the outpatient department of our hospital. Variables are expressed as median or mean stanstandard deviation, as appropriate, and a range of values. Statistical univariate analysis of risk factors for mortality was performed using Fisher s exact test or x 2 as appropriate. Multivariate analysis of risk factors of death was performed 3. Results 3.1. Mortality A total of 10 (20%) patients died at median 4 days (range 0 39) after surgery. Out of 12 patients operated before 1995, seven (58%) patients died, but out of 38 patients operated between 1995 and 2009, only three (5%) patients died ( p = 0.008). Out of 25 patients operated with the use of DHCA, eight (32%) patients died, and out of 25 patients operated with the use of ICP two (8%) patients died. During an overlapping period between 1997 and 2004, when both methods were used, 15 patients were operated with one (7%) mortality: in six patients DHCA was used with one (17%) death, and in nine patients ICP was used with no (0%) death. By univariate and Cox multivariate analysis, a total of 16 risk factors were analysed (Table 3). The surgery before 1995 was disclosed as the only independent risk factor of mortality ( p = 0.037). The hazard of the surgery before 1995 was times more than after that period (95% confidence limit of ). The causes of death were heart failure in five, sepsis in three, multi-organ failure in one and sudden death from respiratory failure in one patient. Two patients with PTA died from heart failure; it was not possible to wean the patients from cardiopulmonary bypass. As mentioned above, one of these patients was in a moribund clinical condition after several cardiopulmonary resuscitations at the time of admission. In one patient with a very unfavourable anatomy, hypoplastic ascending aorta and aortic annulus with valvar and subvalvar stenosis, the valve was partially destroyed by surgical excision of severe subvalvar stenosis. Because of signs of congestive heart failure from a combination of severe aortic insufficiency and stenosis, the Ross Konno operation was performed on the third postoperative day; however, the

4 T. Tlaskal et al. / European Journal of Cardio-thoracic Surgery 38 (2010) Table 3 Univariate analysis of risk factors of death after repair of interrupted aortic arch. Risk factor Operated Died Operated Died p n n n n Surgery before 1995/ DHCA/ICP NS Weight <2.5/2.5 (kg) NS Complex CHD/VSD NS IAA type B, C/IAA type A NS PTA +/PTA * AS +/AS NS TGA +/TGA NS Yasui operation +/Yasui operation NS Severe condition +/severe condition NS DiGeorge +/DiGeorge NS Non-cardiac lesion +/non-cardiac lesion * AOX >90/AOX 90 (min) NS CPB >200/CPB 200 (min) NS Open chest +/open chest NS Complications +/complications NS Abbreviations: AOX aortic cross-clamping; AS aortic stenosis; CHD congenital heart disease; CPB cardiopulmonary bypass; DHCA deep hypothermic circulatory arrest; IAA interrupted aortic arch; ICP isolated cerebral perfusion; PTA persistent truncus arteriosus; TGA transposition of the great arteries; VSD ventricular septal defect. * The difference is marginally significant. patient died 2 days later from multi-organ failure. In three patients with signs of low cardiac output (LCO), infection occurred and terminated with sepsis. The only death occurring more than 30 days after surgery was in a patient with IAA, VSD and SAS who was managed with the use of ICP. She had also congenital laryngeal malformation and DiGeorge syndrome. After surgery, she suffered from chronic respiratory problems. She could be extubated and transferred to another hospital for further treatment of a respiratory tract infection. This patient died suddenly from an acute onset of respiratory distress 39 days after surgery. For statistical analysis of risk factors, this patient was considered in the group of early deaths Re-operations During the follow-up, 23 re-operations and nine catheter interventions were required in 17 patients (Table 4). The most common indications for re-intervention were aortic arch obstruction (AAO) (nine patients), conduit obstruction (seven patients), SAS (seven patients) and aortic insufficiency (AI) (five patients). The first re-operation was performed at median 0.9 years (range 1 day 7.2 years) after surgery. The first re-intervention for SAS was performed at median 1.7 years (range 1 day 5.5 years) and the first re-intervention for AAO was performed at median 0.6 years (range years) after repair. In three patients, two re-operations, and in two patients three re-operations were required. Nine catheter interventions were performed in seven patients: eight balloon dilatations for recurrent AAO were performed in six patients, and one balloon angioplasty of right and left pulmonary arteries was done in one patient. Five reoperations were necessary during the early postoperative period (Ross Konno operation, myectomy, aortopexy for bronchus obstruction and two plications of the diaphragm) and 14 re-interventions (nine surgeries and five catheter interventions) in 10 patients were required during the first year after surgery. Eight re-interventions were performed in five patients after repair of PTA. In three of them, simultaneous conduit replacement and aortic valve replacement or plasty were performed. After previous intervention for SAS, a total of eight re-interventions were necessary in six patients; in three of them re-intervention for SAS as well as for recurrent AAO was required. In two patients, a second reoperation for SAS was necessary. At the same time as intervention for SAS, resection of AAO was done in two patients, and in one patient, patch plasty of the aortic arch was performed. Six balloon dilatations of recurrent AAO were Table 4 Re-operations and catheter interventions. Procedures n Aortic stenosis and insufficiency Aortic valve replacement 6 Myectomy 4 Excision of subaortic membrane 4 Interposition graft at the ascending aorta 2 Valvotomy of the aortic valve 2 Valvuloplasty of the aortic valve 1 Ross Konno operation 1 Patch plasty of the aortic root 1 Recurrent aortic arch obstruction Balloon dilatation of aortic arch obstruction 7 Resection of aortic arch obstruction 2 Patch plasty of the aortic arch 1 Pulmonary artery and right ventricular outflow tract Conduit replacement 11 Plasty of the right ventricular outflow tract 5 Patch plasty of the right pulmonary artery 3 Resection of an aneurysm of the right ventricle 2 Transannular patch plasty 1 Balloon angioplasty of right and left pulmonary artery 1 Other intracardiac procedures Closure of residual ventricular septal defect 5 Extracardiac procedures Plication of the diaphragm 2 Aortopexy 1

5 56 T. Tlaskal et al. / European Journal of Cardio-thoracic Surgery 38 (2010) Fig. 2. Long-term actuarial survival after repair of interrupted aortic arch. Comparison of results with deep hypothermic circulatory arrest (DHCA) versus isolated cerebral perfusion (ICP) for aortic arch reconstruction. Legend: Actuarial survival after repair of interrupted aortic arch expressed using a Kapplan Meier estimation. Comparison of results with the use of deep hypothermic circulatory arrest versus isolated cerebral perfusion for aortic arch reconstruction. Abbreviations: DHCA deep hypothermic circulatory arrest, ICP isolated cerebral perfusion. indicated in four patients who later required surgical intervention Follow-up In this series there were no late deaths. Forty (80%) early survivors were followed up for median 5.7 years (range years). Actuarial 1 year, 10 years and up to 16 years survival was 73% in the group I, and 1 year, 10 years and up to 11 years survival was 92% in the group II, respectively (Fig. 2). Actuarialised re-operation-free survival at 1 year, 5 years and 10 years was 63%, 46% and 34%, respectively, in group I, and 95%, 89% and 44%, respectively, in the group II. The clinical condition of patients is better in group II but the difference is not statistically significant. During follow-up, all 40 survivors after the primary repair of IAA remain in good or very good clinical condition; Regarding cardiovascular health, 32 (80%) patients are in the New York Heart Association (NYHA) class I and eight (20%) in class II. Nine (23%) patients, four in group I and five in group II, are followed up by a neurologist for signs of mild psychomotor retardation in five patients, history of epileptic seizures in two, moderate neurologic deficit in one and hydrocephalus in one patient. There was no difference in occurrence of neurological findings between the two groups. 4. Discussion Results after repair of IAA and associated heart lesions may be affected by many risk factors, which have been analysed by numerous investigators [3 8]. In an extensive multicentric study, McCrindle et al. determined by a multivariate logistic regression analysis lower birth weight, younger age, type B IAA and major associated heart lesions to be incremental risk factors for death after repair of IAA [5]. Fulton et al. detected prolonged DHCA to be a predictor of death [6]. In our previous study, we found severe preoperative complications, earlier year of surgery, severe metabolic acidosis and bad clinical condition to be the most important risk factors of death [8]. In 1999, we summarised our experience with the primary repair of IAA in 19 patients operated with the use of DHCA and compared our early and midterm results of this method with the previous experience with treatment of IAA by a two-stage method in 21 patients [8]. The mortality rate in our primary repair group reported in 1999 was 37% but it was still lower than in the two-stage group. Continuous improvement of the protocol of management of neonates with IAA and associated heart lesions led to further decrease of the early mortality rate [7]. In our experience, the diagnosis of IAA can be done by foetal echocardiographic examination. If the diagnosis is not disclosed prenatally, it should be set up as soon as possible after birth. An infusion of prostaglandins must be always introduced so as to prevent circulatory collapse and subsequent renal failure associated with closure of ductus arteriosus. Before surgery, it is essential to stabilise the patient and treat all serious complications such as infection, respiratory tract disorders or renal failure. In case of instability, it is preferable to put the patient on mechanical ventilation, monitor pressure and blood gases and maintain on prostaglandin and dopamine infusion. We believe that the learning curve and improvement of our management protocol with the introduction of ICP, NIRS monitoring, improvement of extracorporeal circulation, heart protection and intensive postoperative care constituted the basis for improvement of mid- and long-term outcomes of patients with IAA and associated heart lesions. In this retrospective study, we analysed risk factors of mortality after the primary repair of IAA and associated heart lesions in all 50 consecutive patients operated on in one institution. Surgery before 1995 represented the most important predictor of death in our series ( p = 0.037). By multivariate analysis the difference between the use of DHCA in comparison with ICP was not statistically significant. In the group of patients operated with the use of ICP the 30-day mortality rate was, however, lower (4% vs 28%, p = 0.032). In this group, we found also lower need for re-operation, which was influenced especially by growing experience. The difference between the results could be explained first of all by a learning effect. We believe, however, that there were also other factors which influenced favourably the outcome in the group II. These are: (1) better protection of all the organs by maintaining perfusion through the collateral circulation, (2) improvement of heart protection by shortening of the cross-clamping time in patients in whom continuous myocardial perfusion was used, (3) not using deep hypothermia which has some unfavourable consequences and by (4) better conditions for the surgeon, who is not working under such a time-related stress. No clear long-term advantages in the prevention of neurological complications by continuous ICP in comparison with DHCA were proved. Although we have not found any

6 T. Tlaskal et al. / European Journal of Cardio-thoracic Surgery 38 (2010) severe neurological complications that could be attributed to the surgical method, perfusion or method of brain protection, in four (24%) patients in group I and in five (22%) patients in group II, signs of a mild psychomotor retardation or a neurological deficit occurred. These findings could be attributed to the fact that seven (78%) out of the nine patients had DiGeorge syndrome, three (33%) patients survived severe circulatory collapse after closure of the ductus arteriosus and five (56%) had complicated postoperative recovery. No relationship between these events and ICP or DHCA was proved. These findings are in accordance with observations of the Boston group, who were not able to determine any significant difference between the neurological outcomes in patients operated in DHCA and continuous low-flow deep hypothermic perfusion [14,15]. Based on this experience, we would recommend rather deeper hypothermia of at least 22 8C even for aortic arch reconstruction in ICP. The mid- to long-term follow-up of patients after repair of IAA shows an increased risk of recurrent AAO, progression of SAS and complications related to individual-associated heart diseases, such as obstruction of a conduit and aortic regurgitation after repair of PTA [13,16]. Our recent experience is in accordance with the findings of McCrindle et al. that the risk of AAO is higher than all types of aortic arch reconstruction other than a direct anastomosis with pericardial patch augmentation [5]. However, we do prefer to perform a direct end-to-side anastomosis after an extensive mobilisation of descending aorta, ascending aorta and all branches. We do transect the left subclavian artery, as well as the aberrant right subclavian artery, if present, because they are almost always surrounded by a fragile ductal tissue. We believe that this method substantially improves the possibility of construction of a large aortic anastomosis without tension and with a lower risk of restenosis. If recoarctation develops, balloon dilatation represents the method of choice. It was performed eight times in six of our patients. In some patients, however, repeated catheter interventions are required and if balloon dilatations are not effective, surgical resection or patch plasty must be considered [17]. SAS represents one of the most frequent and haemodynamically important lesions associated with IAA, which occurs in a wide spectrum of changes as far as the type, location and severity is concerned [3,4,6,17 20]. Most often, it is caused by hypertrophy and posterior deviation of the connal septum. Sometimes, the hypertrophy of the connal septum is extreme and the diameter of the left ventricular outflow tract (LVOT) may be less than 3 mm. Myotomy or myectomy is usually indicated. It is, however, difficult to set up exact criteria and indications for intervention under borderline conditions. Different methods of echocardiographic assessment of SAS were proposed; however, none of them is ideal [18 20]. Fromourpractice, we think that the subaortic obstruction should be addressed ifthediameteroflvotislessthanabout3 4mm.Wedo agree with Jonas et al. that it is better to tolerate a mild-tomoderate subaortic narrowing after the primary repair if there is an increased risk of aortic valve injury [4].Anexact and effective myectomy in a tiny subaortic region without injury to the aortic valve is sometimes difficult. At reoperation, a subaortic membrane is frequently detected. Usually it can be completely excised, but sometimes it is located so close to the aortic valve that the risk of injury to the valve is extremely high. Valvar stenosis and hypoplastic aortic annulus occur less often but, generally, all types of SASsaremorecommonintypeBIAA,inthepresenceof aberrant subclavian artery or bicuspid aortic valve, and in restrictive VSD [6,7,18]. Luciani et al. recommended placing the superior part of the VSD patch to the left and to suture it from the left ventricular aspect to the connal septum so as to make the path from the left ventricle to the aorta smoother and without turbulence [19]. Flowturbulence in LVOT is believed to be an important factor stimulating progression of SAS and membrane in particular [4,19]. InpatientswithVSDandLVOTdiameterlessthan 3 mm or hypoplastic aortic annulus Yasui operation which uses the pulmonary valve and the main pulmonary artery as an outflow from the left ventricle should be considered [21]. The method requires the use of a valved conduit for reconstruction of the right ventricular to pulmonary artery continuity. Recurrence rate of SAS ranges between 4% and 67% [3,4,5,18,19]. Sell reported 42% incidence of this complication occurring within 3 years after the repair [3]. However, there are patients who require repeated reoperations for recurrent subaortic stenoses. If repeated myectomies or membrane excisions are not effective, Ross or Ross Konno operation represents sometimes the only suitable alternative. As far as long-term survival is concerned, we have reached 73% actuarial survival at 16 years in the DHCA group, and 92% actuarial survival at 12 years in the ICP group. In contrast to our data, McCrindle et al. found an actuarial survival of 64%, 61% and 59%, respectively, at 6 months, 5 years and 15 years [5]. Oosterhof et al. reported even lower actuarial survival with only 34% patients alive at 5 years after repair [22]. Survival and long-term results after repair of IAA and associated heart lesions are continuously improving in many centres. IAA continues to be a challenging congenital cardiovascular disease requiring close lifelong follow-up with an increased risk of surgical and cardiological reinterventions [5,7,17,23 25]. Prevention of neurological complications remains one of the most important tasks in the management of IAA. ICP represents a promising method in this respect, although our study did not confirm any significant difference in neurological outcomes. Further studies to clarify the impact of ICP on early and long-term results after the repair of IAA will be required. 5. Conclusions Congenital heart diseases joined with IAA can be repaired with an early mortality of 5 10%. The results in terms of mortality, occurrence of postoperative complications and need for re-operation depend on experience and management protocol. The use of ICP, as a part of this protocol, probably has a favourable influence on mortality and postoperative morbidity. However, we were not able to provide any data supporting the theory that ICP has a beneficial effect in the prevention of neurological complications.

7 58 T. Tlaskal et al. / European Journal of Cardio-thoracic Surgery 38 (2010) References [1] Celoria GC, Patton RB. Congenital absence of the aortic arch. Am Heart J 1959;58: [2] Trusler GA, Izukawa T. Interrupted aortic arch and ventricular septal defect: direct repair through a median sternotomy incision in a 13-dayold infant. J Thorac Cardiovasc Surg 1975;69: [3] Sell JE, Jonas RA, Mayer JE, Blackstone EH, Kirklin JW, Castaneda AR. The results of a surgical programme for interrupted aortic arch. J Thorac Cardiovascular Surg 1988;96: [4] Jonas RA, Quaegebeur JM, Kirklin JW, Blackstone EH, Daicoff G. Outcomes in patients with interrupted aortic arch and ventricular septal defect. A multi-institutional study. Congenital Heart Surgeons Society. J Thorac Cardiovasc Surg 1994;107: [5] McCrindle BW, Tchervenkov CI, Konstantinov IE, Williams WG, Neirotti RA, Jacobs ML, Blackstone EH, Congenital Heart Surgeons Society. Risk factors associated with mortality and interventions in 472 neonates with interrupted aortic arch: a Congenital Heart Surgeons Society study. J Thorac Cardiovasc Surg 2005;129: [6] Fulton JO, Mas C, Brizard CPR, Cochrane AD, Karl TR. Does left ventricular outflow tract obstruction influence outcome of interrupted aortic arch repair? Ann Thorac Surg 1999;67: [7] Mishra PK. Management strategies for interrupted aortic arch with associated anomalies. Eur J Cardiothorac Surg 2009;35: [8] Tlaskal T, Hucin B, Hruda J, Chaloupecky V, Kostelka M, Janousek J, Skovranek J. Results of primary and two-stage repair of interrupted aortic arch. Eur J Cardiothorac Surg 1998;14: [9] Pigula FA, Gandhi SK, Siewers RD, Davis PJ, Webber SA, Nemoto EM. Regional low-flow perfusion provides somatic circulatory support during neonatal aortic arch surgery. Ann Thorac Surg 2001;72: [10] Asou T, Kado H, Imoto Y, Shiokawa Y, Tominga R, Kawachi Y, Yasui H. Selective cerebral perfusion technique during aortic arch repair in neonates. Ann Thorac Surg 1996;61: [11] Pigula FA, Nemoto EM, Griffith BP, Siewers RD. Regional low-flow perfusion provides cerebral circulatory support during neonatal aortic arch repair reconstruction. J Thorac Cardiovasc Surg 2000;119: [12] Ishino K, Kawada M, Irie H, Kino K, Sano S. Single-stage repair of aortic coarctation with ventricular septal defect using isolated cerebral and myocardial perfusion. Eur J Cardiothorac Surg 2000;17: [13] Tlaskal T, Hucin B, Kucera V, Vojtovic P, Gebauer R, Chaloupecky V, Skovranek J. Repair of persistent truncus arteriosus with interrupted aortic arch. Eur J Cardiothorac Surg 2005;28: [14] Newburger JW, Jonas RA, Wernovsky G. A comparison of the perioperative neurologic effects of hypothermic circulatory arrest versus low-flow cardipulmonary bypass in infant heart surgery. N Engl J Med 1993;329: [15] Visconti KJ, Rimmer D, Gauvreau K, del Nido P, Mayer JE, Hagino I, Pigula FA. Regional low-flow perfusion versus circulatory arrest in neonates: one-year neurodevelopmental outcome. Ann Thorac Surg 2006;82: [16] Konstantinov IE, Karamlou T, Blackstone EH, Mosca RS, Lofland GK, Caldarone CA, Williams WG, Mackie AS, McCrindle BW. Truncus arteriosus associated with interrupted aortic arch in 50 neonates: a Congenital Heart Surgeons study. Ann Thorac Surg 2006;81: [17] Serraf A, Lacour-Gayet F, Robotin M, Bruniaux J, Sousa-Uva M, Roussin R, Planche C. Repair of interrupted aortic arch: a ten-year experience. J Thorac Cardiovasc Surg 1996;112: [18] Suzuki T, Ohye RG, Devaney EJ, Ishizaka T, Nathan PN, Goldberg CS, Gomez CA, Bove EL. Selective management of the left ventricular outflow tract for repair of interrupted aortic arch with ventricular septal defect: management of left ventricular outflow tract obstruction. J Thorac Cardiovasc Surg 2006;131: [19] Luciani GB, Ackerman RJ, Chang AC, Wells WJ, Starnes VA. One-stage repair of interrupted aortic arch, ventricular septal defect, and subaortic obstruction in the neonate: a novel approach. J Thorac Cardiovasc Surg 1996;111: [20] GavaT,HornbergerLK,SandersSP,JonasRA,OttDA,ColanSD.Echocardiographic predictors of left ventricular outflow tract obstruction after repair of interrupted aortic arch. J Am Coll Cardiol 1993;22: [21] Yasui H, Kado H, Nakano E, Yonenaga K, Mitani A, Tomita Y, Iwao H, Yoshii K, Mizoguchi Y, Sunagawa H. Primary repair of interrupted aortic arch and severe aortic stenosis in neonates. J Thorac Cardiovasc Surg 1987;93: [22] Oosterhof T, Azakie A, Freedom RM, Williams WG, McCrindle BW. Associated factors and trends in outcomes of interrupted aortic arch. Ann Thorac Surg 2004;78: [23] Vouhe P, Mace L, Vernant F, Jayais P, Pouard P, Mauriat P, Leca F, Neveux JY. Primary definitive repair of interrupted aortic arch with ventricular septal defect. Eur J Cardiothorac Surg 1990;4: [24] Brown JW, Ruzmetov M, Okada Y, Vijay P, Rodefeld MD, Turrentine MW. Outcomes in patients with interrupted aortic arch and associated anomalies: a 20-year experience. Eur J Cardiothorac Surg 2006;29: [25] Kobayashi M, Ando M, Wada N, Takahashi Y. Outcomes following surgical repair of aortic arch obstructions with associated cardiac anomalies. Eur J Cardiothorac Surg 2009;35: Appendix A. Conference discussion Dr S. Sano (Okayama, Japan): Your technique is very similar to ours except for temperature. We repair the interrupted aortic arch and VSD with mild hypothermia at a temperature of 34 or even at normothermia now. It is always difficult to say whether the improvement of outcome is only due to the adoption of cerebral perfusion or not, because you have adopted this technique since 1997 and you analysed 50 consecutive patients from So during the 20-year period, perioperative and postoperative management have changed in many ways. And also, most of group I are until 2004 in your series. I have two questions. One is that in your manuscript, you mentioned you use crystalloid cardioplegia and also blood cardioplegia. Did you change the cardioplegia during the 20-year period? And also, did you use modified ultrafiltration since 1990, or you don t use any ultrafiltration in the last 20 years or has it changed? Dr Tlaskal: I agree absolutely that the improvement can t be attributed to this method alone. There were many changes during this period of time. But as for cardioplegia, it remained more or less the same. We used crystalloid cardioplegia, and I think that this was not the major difference. And the second question? Dr Sano: The modified ultrafiltration. Dr Tlaskal: Modified ultrafiltration? We are using all the time classical ultrafiltration. Dr Sano: And the third question is from careful review of your manuscript, during your follow-up 21 re-operations were required due to residual valvular or subvalvular aortic stenosis or incompetence, and 17 were in group I and only 4 were in group II. The same thing happened with recurrent aortic arch obstruction. Ten patients required either balloon dilation or surgical repair due to residual arch obstruction in group I and zero in group II. So all these results seem to me to affect the surgical results, which means your technique has improved according to your experience. What do you think of this? And the final question is that it is always difficult to say that the improvement of such outcome is due to only one technique, as long as the analysis is retrospective and nonrandomised. However, I congratulate your remarkable improvement of surgical outcome in this complex anomaly. Dr M. Pozzi (Ancona, Italy): Just a short technical comment. I have seen in the antegrade group that you had direct cannulation of the innominate artery. Correct? Dr Tlaskal: Yes. Dr Pozzi: We ve been using cannulation with the use of the interposition of a Gore-Tex conduit, and I think that makes life much easier. I think it gives you more reliable, even distribution of flow. It gives you more room and much more practical mobility of the ascending aorta. Dr Tlaskal: Yes. Probably you are right. We have been using always the direct cannulation of the innominate artery.