European Journal of Cardio-Thoracic Surgery 49 (2016) 176 182 doi:10.1093/ejcts/ezv056 Advance Access publication 19 February 2015 ORIGINAL ARTICLE Cite this article as: Mair R, Sames-Dolzer E, Innerhuber M, Tulzer A, Grohmann E, Tulzer G. Anatomic repair of complex transposition with en bloc rotation of the truncus arteriosus: 10-year experience. Eur J Cardiothorac Surg 2016;49:176 82. Anatomic repair of complex transposition with en bloc rotation of the truncus arteriosus: 10-year experience Rudolf Mair*, Eva Sames-Dolzer, Michaela Innerhuber, Andreas Tulzer, Eva Grohmann and Gerald Tulzer Children s Heart Center Linz, Linz, Austria * Corresponding author. Children s Heart Center Linz, Krankenhausstr. 9, 4020 Linz, Austria. Tel: +43-732-780673265; fax: +43-732-78062154; e-mail: rudolf.mair@akh.linz.at (R. Mair). Received 20 September 2014; received in revised form 13 January 2015; accepted 19 January 2015 Abstract OBJECTIVES: Transposition of the great arteries, ventricular septal defect and left ventricular outflow tract obstruction are commonly called complex transposition. The traditional method of repair is the Rastelli procedure. Aortic translocation (Nikaidoh 1984) provides a more anatomic repair of this malformation. En bloc rotation of the truncus arteriosus (double root translocation, half turn truncal switch procedure) was introduced in 2003 (Yamagishi), and offers a complete anatomic repair with growth potential in all tubular structures. The aim of this study was to analyse our general experience with this method and the mid-term results concerning growth of the tubular structures as well as the function of the semilunar valves, if preserved. METHODS: Nineteen patients with transposition of the great arteries, ventricular septal defect and left ventricular outflow tract obstruction or similar cases of double outlet right ventricle (DORV) have been treated by an en bloc rotation of the truncus arteriosus in our centre since 2003. Patient age ranged between 4 days and 6.46 years. The median age was 0.39 [0.1; 2.25] years. Weight ranged between 3.1 and 18.8 kg. Median weight was 5.6 [3.6; 9] kg. Five patients had received between 1 and 4 palliative procedures prior to the definitive repair. The pulmonary valve could be preserved in 15 cases, whereas a transannular patch was necessary in 4 cases. RESULTS: One patient died of chronic left ventricular failure during the hospital stay. One patient acquired a severe cerebral haemorrhage 3 weeks after the operation. She was discharged and died 6 months later. One patient is not in a follow-up programme. Sixteen patients are now followed over a period of 153 days to 9.96 years. Aortic and pulmonary valves showed proportional growth during the follow-up period. The preserved pulmonary valves were small for age, but kept their competence satisfactorily. Three patients required a reoperation: one aortic valve repair, 1 permanent pacemaker, VSD closure. CONCLUSIONS: Up to now, reoperations had been caused by technical issues. The reoperation rate can be kept low, by understanding some important features of this procedure and avoiding these problems. Complete anatomic repair with growth potential and satisfactory preserved pulmonary valve function is possible. Keywords: Complex transposition Left ventricular outflow tract obstruction En bloc rotation Rastelli procedure INTRODUCTION Transposition of the great arteries (TGA), ventricular septal defect (VSD) and left ventricular outflow tract obstruction (LVOTO) and similar cases of DORV are commonly treated with the Rastelli procedure [1]. In this procedure, the systemic blood flow is directed from the systemic ventricle through the VSD into the aorta by an intracardiac baffle. This tunnel might be long, and could lead to subaortic obstruction later in life depending on the anatomic situation (length of the conus septum, size of the pulmonary root, distance of the VSD to the aortic root). Furthermore, a right ventricle to pulmonary artery (RV-PA) conduit is necessary [2 5]. Presented at the 28th Annual Meeting of the European Association for Cardio- Thoracic Surgery, Milan, Italy, 11 15 October 2014. The Réparation à l étage ventriculaire (REV procedure) avoids the RV-PA conduit by using a Lecompte manoeuvre, and a direct anastomosis of the PA with the RV. Resection of the conus septum is also part of this procedure. In this case, the intracardiac tunnel should be somewhat shorter and wider than in a classic Rastelli operation [5 7]. The aortic translocation (Nikaidoh-Bex) brings the aorta directly over the LVOT to its original anatomic position. Pulmonary root and valve are sacrificed in this procedure [8]. In en bloc rotation of the truncus arteriosus (half turn truncal switch procedure Yamagishi, double root translocation etc.) the aorta is brought over the LVOT and the pulmonary root comes over the right ventricular outflow tract (RVOT). The pulmonary valve, not suitable for aortic position, may work properly in pulmonary position and growth potential should be preserved in all The Author 2015. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
R. Mair et al. / European Journal of Cardio-Thoracic Surgery 177 tubular structures. With this procedure, a complete anatomic repair of this malformation should be achieved [9 11]. The aim of this retrospective study is to analyse our experience, as well as the mid-term outcome of this procedure. The study wanted to see whether, and how, the tubular structures grow and also to investigate whether the semilunar valve function can be preserved. PATIENTS AND METHODS Patients Nineteen patients with either TGA/VSD/LVOTO or similar forms of DORV underwent a complete repair by en bloc rotation of the truncus arteriosus from April 2003 to November 2013 (Table 1). Patient age ranged from 4 days to 6.46 years. Median age was 0.39 [0.1; 2.25] years, median weight 5.6 [3.6; 9] kg (3.1 18.8 kg). Anatomy Thirteen patients were classified as TGA, 6 patients as similar forms of DORV. All patients had well-balanced ventricles. The VSD was sub-pulmonary in all cases. In 9 patients, the pulmonary valve was bicuspid: 6 of them were symmetric and hypoplastic; 3 were asymmetric and dysplastic. In 10 patients, the pulmonary valve was tricuspid: in 9 patients, it was symmetric and simply hypoplastic; in 1 patient, it was asymmetric and dysplastic. The LVOT obstruction was short in 8 patients; 11 patients had a tunnel type stenosis of the LVOT. Indication, decision making We regarded en bloc rotation as the procedure of choice in this malformation over the last 10 years. If there was no ductus dependency, the operation was planned electively as a primary repair at an age between 6 weeks and 3 months. The majority of patients however presented later in our centre. All of our patients would have otherwise been candidates for either a Rastelli procedure or other forms of repair of this type of malformation. An arterial switch operation was not possible in any of these cases. A Rastelli procedure was performed, if additional complex cardiac lesions were present. This was the case in 2 patients with additional complete atrioventricular (AV)-canal, one of them with heterotaxy syndrome. Initially, we restricted en bloc rotation of the truncus arteriosus to patients with a normal coronary anatomy for transposition. This means that the left coronary artery (RCA) arises from the left anterior facing sinus, the right coronary artery from the right posterior facing sinus (Yacoub type A). An intraoperative switch to a Rastelli procedure due to unfavourable coronary anatomy was performed in 7 cases during this period (e.g. circumflex artery from the RCA; RCA from the left anterior descending artery (LAD) or single coronary ostium). By gaining more and more experience, we were gradually able to extend the indication regarding coronary anatomy. Based on our present experience, 3 of these 7 patients would still be refused an en bloc rotation of the truncus arteriosus. However, patients with a circumflex artery coming from the right coronary artery and the posterior facing sinus, and patients with a right coronary artery arising from an LAD and crossing the RVOT have undergone an en bloc rotation successfully in the latter part of our series. Univentricular palliation is used only if biventricular repair is precluded either by severely unbalanced ventricles, by unbalanced or CONGENITAL Table 1: Patients and preoperative palliative procedures Patient number Sex Age (years) Weight (kg) Anatomy Palliations number Type of palliations Indication for palliation 1 m 5.62 18.8 DORV, TGA, LVOTO 4 BT shunt + atrioseptectomy, shunt banding, Cyanosis bidirectional Glenn, right class. Glenn 2 f 1.01 6.7 DORV, TGA, LVOTO 3 m 0.02 3.6 TGA/VSD/LVOTO 4 f 0.01 3.2 TGA, VSD, LVOTO 5 f 0.17 5.2 DORV, TGA, LVOTO 6 m 0.01 3.1 TGA/VSD/LVOTO 7 f 0.20 3.38 TGA, VSD, LVOTO 8 m 1.00 5.8 DORV, TGA, VSD, LVOTO 9 f 0.16 3.8 TGA/VSD/LVOTO 10 f 0.39 4.7 TGA/VSD/LVOTO 1 BT shunt + atrioseptectomy Lost stent 11 f 0.28 5.6 TGA/VSDLVOTO 12 f 6.47 14 TGA/VSD/LVOTO 13 m 1.96 9 DORV, TGA, LVOTO 14 f 3.10 14.8 DORV, TGA, VSD, LVOTO, 2 BT shunt, bidirectional Glenn Cyanosis 15 f 2.53 8.4 TGA/VSD/LVOTO 16 m 0.94 8.3 TGA/VSD/LVOTO 17 f 3.65 13.4 TGA, VSD, LVOTO 2 BT shunt, central shunt Cyanosis 18 m 0.04 3.66 TGA/VSD/LVOTO 1 BT shunt + atrioseptectomy Desperate condition after transport 19 m 0.02 3.6 TGA/VSD/LVOTO F: female; m: male.
178 R. Mair et al. / European Journal of Cardio-Thoracic Surgery straddling AV-valves or by other situations making biventricular reconstruction impossible. Nevertheless, 5 patients had between 1 and 4 palliative procedures prior to en bloc rotation of the truncus arteriosus (Table 1). Three palliated patients came from other centres, where complete repair was not offered or possible (Table 1,patients 1,14 and 17). Two newborn patients received a primary palliation in our centre due to emergencies (Table 1, patients 10 and 18). Operative technique The operation is done on cardiopulmonary bypass (CPB) in moderate hypothermia with double venous cannulation. Cold crystalloid cardioplegia is used, and usually repeated at hourly intervals. Both PAs are extensively mobilized. After cross-clamping the aorta and instillation of cardioplegia, the aorta and PA are cut transversely at the same level. Marking stitches are placed at the most anterior aspect of the aortic root and at the most posterior aspect of the pulmonary root. The coronaries are excised and mobilized in the same fashion as an arterial switch procedure. The complete truncus arteriosus (aortic and pulmonary root together) is excised (Fig. 1). The conus septum is either totally resected together with the aortic and pulmonary root, or partially resected and divided. The incision of the conus septum is made leftwards to the papillary muscle of the conus (Fig. 2). The truncus arteriosus is rotated by 180 and reimplanted, so that the aorta comes over the LVOT and the pulmonary root comes over the RVOT. After implanting the posterior part of the aortic root, the LVOT is completed by closing the VSD with a patch. The LVOT is thereby enlarged to its appropriate size. It is important to carry out the suture line to both ends of the divided or resected conus septum at this level (Fig. 2). By doing so, the residual circumference of the aortic root can determine the exact width of the VSD patch. The next step is to reconnect the coronaries to their contralateral orifices, which are located anteriorly after rotation. A Lecompte manoeuvre is performed and the anastomosis between the aortic root and ascending aorta is made. Pulmonary root and RV are then anastomosed. Finally, the pulmonary root and PA are reconnected [10]. The pulmonary valve could be preserved in 15 patients. Commissurotomy, subannular and supravalvular patch enlargement was used in most of these cases. A transannular patch was necessary on 4 patients. All patches were of autologous pericardium, fixated in glutaraldehyde. Operative data The median CPB time was 289 min [263; 357] ranging from 213 to 419 min. Median aortic cross-clamp time was 157 min [142; 171] ranging from 128 to 190 min. Delayed chest closure was performed on 5 patients. Follow-up One patient, who was referred by an aid programme, does not have a fixed address, and we were therefore unable to get data from her. Four patients were referred by aid organizations from foreign countries. They are being followed by hospitals in their home countries. We have reliable echocardiographic reports and clinical follow-up examinations on all 4 patients. Seven patients are being followed in our centre and 5 patients are in the follow-up programmes coordinated with our centre, providing complete echocardiographic and clinical data. Data analysis Routine echocardiographic data of the LVOT and RVOT were analysed. To assess growth, the diameter of the aortic valve and the diameter of the pulmonary valve in the first postoperative echo were compared with the diameter at the last follow-up. To see whether growth in the two vessels was proportional, we compared the z-scores of the aortic valve and pulmonary valve diameter in the first postoperative echo with those in the last follow-up. These data, diameter and z-scores were only available in patients followed by our own centre and related clinics (n = 12). Figure 1: Aorta and pulmonary artery are cut transversely, coronaries are mobilized, the line of excision of the truncus arteriosus is depicted by the dashed lined. Figure 2: Truncus arteriosus is excised, rotated by 180, reimplantation started posteriorly, conus septum transected leftward to the papillary muscle of the conus.
R. Mair et al. / European Journal of Cardio-Thoracic Surgery 179 Statistical analysis Basic data are described as median and quartiles. For the comparison of z-scores, a paired t-test was applied. A Kaplan Meier plot was drawn for freedom from reoperation and reintervention. RESULTS Perioperative results Mortality. One patient, a 3-year-old child referred from a foreign centre, died due to chronic left ventricular failure during her hospital stay. She had a single ventricle palliation at the stage of a bidirectional Glenn. She was severely cyanotic due to multiple veno-venous fistulas to the coronary sinus. For this reason, we decided not to do a Fontan procedure, but to go for biventricular repair. After the operation, the patient suffered from left ventricular failure, and we had to take her on extracorporeal membrane oxygenation (ECMO). She could be weaned, but remained in chronic pulmonary congestion. She needed another ECMO support 6 weeks later, suffering from symptoms similar to an acute respiratory distress syndrome. She could be weaned again, but she died 8 weeks after the operation in our intensive care unit. There were no signs of ischaemic myocardial lesions. Although the left ventricle was used to systemic pressure, it was obviously too small to support systemic circulation without the help of the right ventricle. Morbidity. One patient acquired cerebral bleeding 3 weeks after the operation when she was already completely mobilized. This was presumably caused by cerebral arterio venous malformations. She was our oldest patient and she was extremely cyanotic prior to the operation. She could be discharged with a severe neurological deficit. Residual lesions One patient had a residual VSD. One patient had an intermittent AV block II. Perioperative results: left ventricular outflow tract None of the patients showed a gradient in the LVOT. Six patients showed some minor degree of aortic valve regurgitation of <II. One patient demonstrated mild-to-moderate aortic valve insufficiency (II +). Almost all patients had an increased aortic valve diameter in the first postoperative echo: median z-score of the aortic valve diameter in the first postoperative echocardiography was 3.97 [2.74; 4.32]. Perioperative results: right ventricular outflow tract Four patients had extremely hypoplastic pulmonary valves, such that a transannular patch was necessary. The pulmonary valve could be preserved in 15 patients. Subcommissural patchplasty and supra-annular patch enlargement of the pulmonary root had to be performed on some of these however. The pulmonary valves or roots of these patients were still small in the first postoperative echo. Median z-score was 0.66 [ 0.9; 0.29]. The pulmonary valve remained competent in 14 patients, and showed trivial regurgitation in 1 patient. One patient demonstrated a significant gradient (4 m/s) in the RVOT. This was due to a residual VSD (see: residual lesions, reoperations and reinterventions). Four other patients showed flow velocities up to 3 m/s. Flow velocities were <2.7 m/s for 14 patients, which is equivalent to a gradient of <30 mmhg. At discharge, all patients showed good ventricular function in both ventricles in the postoperative echocardiographic examinations. No serious AV-valve incompetence (>I ) could be detected. According to our protocol for patients after repair of complex cardiac malformations (prolonged CPB time, prolonged cross-clamp time), all patients were on a combination of angiotensin converting enzyme inhibitors and diuretics for 3 6 months postoperatively. Aside from the patient with the severe cerebral haemorrhage, no further patient presented any new neurological abnormalities. Two patients received seizure medication for a seizure disease that existed before admission. Follow-up, mid-term and late results Mortality. One patient, who was discharged with a severe neurological deficit, died after 6 months due to aspiration. Sixteen patients were followed either by our own unit or by their referral centres. The follow-up period ranged from 153 days to 9.96 years. The median follow-up was 3.95 years [1.49; 6.45]. Reoperations and reinterventions Three patients needed reoperations. One patient needed a pacemaker for AV block II 3 months after en bloc rotation (Patient 6). One patient had a residual VSD and was reoperated on after 8.7 months (Patient 11). One patient had aortic valve regurgitation, which was mild to moderate immediately after the operation and deteriorated to moderate to severe over the next few years. She required reoperation after 4 years. The valve could be repaired by a subcommissural annuloplasty (Patient 5). One patient had a catheter intervention for supravalvular PA stenosis and kinking of the left PA 6 months postoperatively (Patient 6). Pictures of this catheterization are shown in Figs 3 and 4. Freedom from reoperation and reintervention is expressed in a Kaplan Meier curve (Fig. 5). All patients are presently in New York Heart Association class I. Left ventricular outflow tract None of the patients developed a gradient in the LVOT. Six patients kept their minor degree of aortic valve regurgitation of <II. The moderate aortic valve insufficiency (II +) in the initial postoperative echo of 1 patient deteriorated throughout the follow-up and needed reoperation (see reoperations and reinterventions). Growth of the aortic valve could be demonstrated in absolute metric values throughout the follow-up (Fig. 6). Median z-score at the last follow-up was 3.62 [0.76; 4.58]. Although the z-score of the aortic valve decreased in some patients, this did not reach significance (P = 0.22 paired t-test). This means, that in most children the aortic valve is still too big for their age and size. Right ventricular outflow tract During the follow-up period, growth of the pulmonary valve or root could be shown in all patients (Fig. 7). However, the z-score CONGENITAL
180 R. Mair et al. / European Journal of Cardio-Thoracic Surgery Figure 5: Kaplan Meier curve for freedom from reoperation and reintervention: two reinterventions occurred in 1 patient (Patient 6). The second reintervention (118 days after the first one) cannot be displayed in a Kaplan Meier curve. Figure 3: Ventriculography of the left ventricle 6 months after en bloc rotation of the truncus arteriosus. Figure 6: Aortic valve growth after en bloc rotation of the truncus arteriosus. Figure 4: Ventriculography of the right ventricle 6 months after en bloc rotation of the truncus arteriosus and subcommisural patch plasty in the right ventricular outflow tract. in the whole group did not change significantly: median z-score at the last follow-up examination was 0.26 [ 1.82; 0.45] (P = 0.49). Significant flow accelerations of >3 m/s in the pulmonary root could be demonstrated by Doppler echocardiography in 2 patients. In both patients, the flow accelerations appeared supravalvar, at the site of the anatomosis to the pulmonary bifurcation. One patient had a flow velocity of 3.7 m/s (calculated gradient: 57 mmhg), and had a catheter intervention after 6 months (see reoperations and reinterventions). Figure 7: Pulmonary valve growth after en bloc rotation of the truncus arteriosus. The other patient has a peak velocity of 4 m/s now after 9.16 years (calculated gradient of 60 mmhg). A catheterization is planned. All other patients showed either normal flow patterns or minor flow velocity accelerations with calculated gradients between 10 and 31 mmhg, remaining constant up to now. A free pulmonary regurgitation has been seen in 3 patients with a transannular patch, the fourth patient died after 6 months due to aspiration. The pulmonary valve was competent in 5 patients. Trivial incompetence (I ) was found in 4 patients. A mild insufficiency (II ) was seen in 4 patients.
R. Mair et al. / European Journal of Cardio-Thoracic Surgery 181 DISCUSSION The classic repair for TGA/VSD/LVOTO is the Rastelli procedure. Its long-term problems are well documented. A long intracardiac tunnel has an inherent tendency to obstruct. This might be one reason for the substantial late mortality [2, 5]. An RV-PA conduit leads to multiple reoperations or at least reinterventions. Kreutzer et al. [2] published a series of 101 Rastelli procedures over a period of 25 years. They reported 17 late deaths, 11 reoperations for LVOT obstruction, 44 conduit reoperations and 28 conduit reinterventions. The REV procedure does not need an RV-PA conduit, but it has no valve in the RVOT. By resection of the conus septum, the intracardiac tunnel is shortened. However, it is still not an anatomic repair regarding the LVOT. This procedure showed better results in the multicentre study by Hazekamp et al. [5] than the Rastelli procedure. In en bloc rotation of the truncus arteriosus, the aortic root comes over the left ventricle and the pulmonary root over the right ventricle. Due to the shorter follow-up period and the small number of patients, a comparison of our study to the above-mentioned studies is not justified. However, the incidence of reoperations and reinterventions for LVOT or RVOT obstruction is low. Most of our reoperations were due to either technical reasons or are not procedure specific. To avoid any possible LVOT obstruction, we used oversized VSD patches on our first patients. This led to central aortic valve regurgitation in some of them. We were able to solve this problem upon reoperation using plicating sutures in the subcommissural area in 1 patient, thereby narrowing the proximal valve annulus to its appropriate size. Since we use the implantation technique described above, carrying out the proximal suture line of the aortic root up to the remnants of the conus septum, thereby determining the width of the patch exactly, we did not have those aortic valve problems. We regard this as a technical problem, which can be avoided by following this principle of implantation and avoiding oversized patches in an oversized aortic root. The aortic annulus usually remains unchanged by this kind of operation. The pulmonary annulus however might be enlarged if subcommissural and supravalvular or transannular patches are used. We therefore used the first postoperative echocardiographic measurements for comparison with the follow-up examinations. Both the aortic and pulmonary valve showed growth. Almost all patients with this malformation started with an oversized aorta. This is explained by the preoperative flow patterns, which are similar to those in tetralogy of Fallot. Although blood flow normalized after the operation, as there were no shunts anymore, the aortic valve remained big and the z-score did not decrease significantly during the follow-up. On the RV-side, these patients started with a hypoplastic pulmonary root, which was still small after a valve sparing operation despite enlargement plasties. Growth was clearly demonstrated for the pulmonary valve and annulus; however, they remained small for the size and age of the patient. Z-scores did not show significant change. This discrepancy in size of the great arteries after the long follow-up periods might be explained by the fact that despite flow being normalized and equal in both vessels, wall tension is higher in bigger vessels according to the law of Laplace. This may increase growth. Although z-scores are usually not normalized during the followup, an occurrence of a significant stenosis, which required treatment, was rare in this series. It was predominantly supravalvular at the site of the anastomosis when it occurred, which is similar to the findings recorded after an arterial switch procedure. Therefore, this might be a result of scar formation, and not an effect of growth deficit. As we demonstrated in our series, the pulmonary valve, though dysplastic and not usable as an aortic valve, can function well in the pulmonary position in many cases. Most of the preserved pulmonary valves acceptably kept their competence and up to now no patient has required reoperation for a pulmonary valve problem. The virtual axis of rotation of the truncus arteriosus is either on the right side of the infundibular septum or slightly within posterior part of the aortic root. Therefore the coronary ostia, which are located in the posterior aspect of the aortic root, are in the anterior aspect afterwards, but almost at the same level. The shift of the coronary arteries is therefore minimal [9, 10]. The procedure itself can generally be described as invasive. Cross-clamp periods and CPB periods are long and myocardial protection has to be done thoroughly. It is technically demanding and the process of en bloc excision of the truncus arteriosus requires excellent anatomic orientation. For these reasons, we did not extend the indication for en bloc rotation to patients with additional malformations, as was mentioned in the patients and methods section. Concerning coronary anatomy, the indication was extended during the final years. Despite this there are still problems, which we have been unable to solve. For example, we had to switch to a Rastelli procedure in 2013 for a patient with a single coronary artery arising close to the commissure between the two facing sinuses and an anterior intramural course, despite this patient being an otherwise an excellent candidate for en bloc rotation. This is only one example and there will be more, due to the wide variation in coronary anatomy of this malformation. Due to its complexity and its dependency on the coronary anatomy, en bloc rotation of the truncus arteriosus is not currently as universally applicable to these malformations as the Rastelli procedure or the REV procedure. SUMMARY En bloc rotation of the truncus arteriosus is definitely an option of complete anatomic repair of transposition with VSD and left ventricular outflow tract obstruction or similar forms of DORV. It is important to understand and to follow a number of technical principles in reimplantation of the truncus arteriosus and sizing of the VSD patch to avoid problems with the aortic valve. It is an invasive procedure and it is technically demanding. Its late effects are similar to those found after repair of tetralogy of Fallot and the arterial switch procedure. Limitations The small number of patients limits the study. ACKNOWLEDGEMENTS We thank Wolfgang Schimetta for his statistical support. Conflict of interest: none declared. CONGENITAL
182 R. Mair et al. / European Journal of Cardio-Thoracic Surgery REFERENCES [1] Rastelli GC, Wallace RB, Ongley PA. Complete repair of transposition of the great arteries with pulmonary stenosis. Circulation 1969;39:83 95. [2] Kreutzer C, De Vive J, Oppido G, Kreutzer J, Gauvreau K, Freed M et al. Twenty-five-year experience with rastelli repair for transposition of the great arteries. J Thorac Cardiovasc Surg 2000;120:211 23. [3] Emani SM, Beroukhim R, Zurakowski D, Pigula FA, Mayer JE, del Nido PJ et al. Outcomes after anatomic repair for d-transposition of the great arteries with left ventricular outflow tract obstruction. Circulation 2009;120:S53 8. [4] Hörer J, Schreiber C, Dworak E, Cleuziou J, Prodan Z, Vogt M et al. Long-term results after the Rastelli repair for transposition of the great arteries. Ann Thorac Surg 2007;84:894 9. [5] Hazekamp MG, Gomez AA, Koolbergen DR, Hraska V, Metras DR, Mattila IP et al. European Congenital Heart Surgeons Association. Surgery for transposition of the great arteries, ventricular septal defect and left ventricular outflow tract obstruction: European Congenital Heart Surgeons Association multicenter study. Eur J Cardiothorac Surg 2010;38:699 706. [6] Lecompte Y, Neveux JY, Leca F, Zannini L, Tu TV, Dubois Y et al. Reconstruction of the pulmonary outflow tract without prosthetic conduit. J Thorac Cardiovasc Surg 1982;84:727 33. [7] Vouhe PR, Tamisier D, Leca F, Ouaknine R, Vernant F, Neveux JY. Transposition of the great arteries, ventricular septal defect and pulmonary outflow tract obstruction. Rastelli or Lecompte procedure? J Thorac Cardiovasc Surg 1992;103:428 36. [8] Nikaidoh H. Aortic translocation and biventricular outflow tract reconstruction. J Thorac Cardiovasc Surg 1984;88:365 72. [9] Yamagishi M, Shuntoh K, Matsushita T, Fujiwara K, Shinkawa T, Miyazaki T et al. Half- turned truncal switch operation for complete transposition of the great arteries with ventricular septal defect and pulmonary stenosis. J Thorac Cardiovasc Surg 2003;125:966 8. [10] Mair R, Sames-Dolzer E, Vondrys D, Lechner E, Tulzer G. En bloc rotation of the truncus arteriosus-an option for anatomic repair of transposition of the great arteries, ventricular septal defect and left ventricular outflow tract obstruction. J Thorac Cardiovasc Surg 2006;131:740 1. [11] Hu S, Xie Y, Li S, Wang X, Yan F, Li Y et al. Double root translocation, for double outlet right ventricle with noncommitted ventricular septal defect or double outlet right ventricle with subpulmonary ventricular septal defect and associated subpulmonary stenosis: an optimized solution. Ann Thorac Surg 2010;89:1360 5. APPENDIX. CONFERENCE DISCUSSION Scan to your mobile or go to http://www.oxfordjournals.org/page/6153/1 to search for the presentation on the EACTS library Dr V. Hraska (Sankt Augustin, Germany): The optimal therapeutic approach to complex transposition remains challenging and controversial, particularly in patients with non-committed/restrictive VSD, complex AV valve anatomy, borderline right ventricle, and left ventricular outflow tract obstruction. Under these circumstances, usually intraventricular rerouting using the REV or Rastelli operation is suboptimal and consideration is given either to conversion to the Fontan or to posterior translocation of the aorta or eventually to en bloc rotation of truncus arteriosus, as was demonstrated. That leads to the first question. Is en bloc rotation the method of choice even in patients with committed VSD to the aorta who are the perfect candidates for intraventricular rerouting in your institution? In other words, what are the contraindications for en bloc rotation? Dr Mair: We used this method for all patients with transposition of the great arteries or similar forms of DROV who otherwise would have needed a Rastelli procedure or a REV procedure. We restricted it in our initial series to forms with normal coronary anatomy for transposition. That means if the left coronary artery arises from the left coronary sinus and the right one from the right posterior coronary sinus. But with more and more experience we got more courageous and we are now doing it in almost all coronary anatomies. Dr Hraska: So, in other words, it is a method of choice in your institution? Dr Mair: Yes, the method of choice. Dr Hraska: The major advantage of Nikaidoh or en bloc rotation is creation of straight alignment of both outflows. However, in comparison with posterior aortic translocation, en bloc rotation should provide a functioning neopulmonary valve. That is basically the reason why the rotation is done. The question is, what is the functional performance of the pulmonary valve, which is very frequently stenotic and dysplastic in this group of patients? Dr Mair: The pulmonary valve was, of course, dysplastic, because this is one part of the indication of the operation, but you can do a lot to the pulmonary valve, even to a dysplastic pulmonary valve. If you think of tetralogy, you can also do a lot of things to the pulmonary valve that works in this more or less low pressure system. You can use subvalvular patches; you can use supravalvular patch plasty, and so on. As I said, the gradient which was found in most patients was between 10 and 31 mmhg and completely constant over the years, and in the patients in whom the valve was kept intact, we did not find any pulmonary valve incompetence of more than mild. Dr Hraska: Unfortunately data regarding pulmonary valve function has not been provided. I think this is very important information, because this is the reason why the en bloc rotation, in comparison with Nikaidoh, should be done. Dr Mair: This may be one reason. One other reason is maybe the coronary arteries, because the axis of rotation is about in the plane of the coronary arteries. So you don t have to mobilize them as far as in the Nikaidoh procedure. Dr Hraska: But if the Nikaidoh procedure is done correctly, there is no need for that. Dr P. Vouhe (Paris, France): You said that there is no contraindication related to coronary arteries. Is it true also with an artery crossing in front of the aorta, and in this case how do you deal with it? Dr Mair: When the coronaries cross in front of the aorta, when the right coronary comes in front of the aorta, you have to mobilize the coronary artery very extensively, but it works. The only contraindication we found in the last years was a single intramural coronary artery which came out of the posterior facing sinus.