Long-Term Outcome of Direct Neopulmonary Artery Reconstruction During the Arterial Switch Procedure

Transcription

1 Long-Term Outcome of Direct Neopulmonary Artery Reconstruction During the Arterial Switch Procedure Jacek J. Moll, MD, PhD, Krzysztof W. Michalak, MD, Katarzyna Młudzik, MD, PhD, Tomasz Moszura, MD, PhD, Marek Kopala, MD, PhD, Maciej Moll, MD, PhD, and Jadwiga A. Moll, MD, PhD Departments of Cardiosurgery and Cardiology, Polish Mother s Memorial Hospital, Lodz, Poland Background. Neopulmonary stenosis at anastomosis site is one of the most frequent complications after the arterial switch procedure for transposition of the great arteries. The surgical technique is a crucial factor associated with the frequency of stenotic complications. We present the outcomes of direct neopulmonary anastomosis during the arterial switch procedure in patients with simple transposition. This research was to assess the efficacy of this surgical technique based on the incidence of postprocedural supravalvular neopulmonary stenosis (SVPS). Methods. Among 545 patients operated on in our department between 1992 and 2009, the 346 consecutive survivors who had undergone simple transposition in the first month of life were included in this analysis. Switch procedures were performed with direct neopulmonary artery anastomosis in 318 patients (92%); in the remaining 28 (8%), the risk of coronary artery compression required the use of a pericardial patch for pulmonary reconstruction. Results. Neopulmonary stenosis occurred in 9 patients (2.6%): 5 had undergone direct neopulmonary reconstruction, and 4 had been treated with a patch. Balloon angioplasty of SVPS was performed twice in 1 patient. No patients required reoperation to treat neopulmonary stenosis. In multivariate analysis (logistic regression), patch reconstruction (odds ratio, 27.5; p 0.001) and nonfacing commissures (odds ratio, 11.1; p 0.004) were correlated significantly with the incidence of SVPS. Conclusions. Direct neopulmonary artery anastomosis during arterial switch is an interesting alternative to patch reconstructions and ensures a good postoperative result with low rates of complications and SVPS. (Ann Thorac Surg 2012;93:177 84) 2012 by The Society of Thoracic Surgeons In the last 35 years, the arterial switch operation (ASO), developed and introduced by Jatene and colleagues [1], has become the procedure of choice for transposition of the great arteries (TGA). Since the 1980s, it has been performed successfully in neonates [2], which has significantly changed the late outcomes and quality of life for patients with TGA. During the last decade, excellent clinical results have been reported after the switch procedure, with good cardiac function, psychosomatic development, and quality of life [3 8]. Although the switch procedure has eliminated some complications related to atrial baffle repairs (Mustard and Senning procedures), such as late failure of the right ventricle working in the systemic circulation, other postoperative complications of ASO were discovered during follow-up, most frequently aortic regurgitation, neopulmonary artery stenosis, and coronary artery flow disturbances [7, 9 11]. During anatomic correction of TGA, pulmonary artery (PA) reconstruction may be performed directly or with Accepted for publication Sept 14, Address correspondence to Dr Michalak, Department of Cardiology, Polish Mother s Memorial Hospital, Research Institute, ul. Rzgowska 281/289, Lodz, Poland; krzysiekmichalak@interia.pl. the use of a patch [3, 12 14]. The site of anastomosis is especially susceptible to development of stenosis. Many clinical analyses confirm that neopulmonary stenosis is the most frequent cause of reinterventions and reoperations, occurring in 1% to 35% of patients after ASO [7]. The frequency of neopulmonary stenosis depends on the type of neopulmonary reconstruction and on the experience of the cardiosurgical center. In our department, direct PA anastomosis with modifications developed at our institution is the method of choice for the ASO. This research assessed the efficacy of this surgical technique by analyzing the incidence of supravalvular neopulmonary stenosis (SVNS) and its potential risk factors, including type of surgical reconstruction of the neo-pa. Patients and Methods Study Population Between 1992 and 2009, 545 patients with TGA underwent the ASO procedure in the Cardiosurgery Department of Polish Mother s Memorial Hospital, with 6.8% overall mortality and 0.7% mortality in the last 5 years. In 92% of patients, neopulmonary reconstruction was performed directly, without a patch by The Society of Thoracic Surgeons /$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 178 MOLL ET AL Ann Thorac Surg DIRECT PULMONARY RECONSTRUCTION 2012;93: Table 1. Characteristic of the Study Group With Division Into Patients With and Without Neopulmonary Artery Stenosis Without Variable a (N 346) (n 9) (n 337) Total SVPS SVPS p Value Neo-PA reconstruction Direct 318 (92) 5 (55) 313 (93) b Patch 28 (8) 4 (45) 24 (7) b Sex 0.47 Male 244 (70.5) 7 (78) 237 (70.3) Female 102 (29.5) 2 (22) 100 (29.7) Rashkind procedure 215 (62) 6 (67) 209 (62) 0.53 Nonfacing commissures 51 (15) 4 (44) 47 (14) 0.03 b PAV/AoV discrepancy 122 (35) 3 (33) 119 (35) 0.6 Coronary anomalies 118 (34) 5 (56) 113 (34) 0.15 PAV insufficiency 225 (65) 7 (77) 218 (65) 0.43 Age at operation, days Weight at operation, grams 3, , , Aorta cross-clamping time, min Intensive care unit stay, days a Continuous data are presented as number (%), and continuous data as mean standard deviation. b Statistically significant. AoV aortic valve; PA pulmonary artery; PAV pulmonary artery valve; SVPS supravalvular neopulmonary stenosis. To minimize additional factors associated with the surgical technique and for the standardization of the study group, only the 346 survivors with simple TGA operated on in the first month of life were included in this analysis (Table 1). Patients with ventricular septal defect, aortic arch anomalies associated with TGA, or a two-stage correction were excluded. We reviewed retrospective clinical data from operations, routine clinical visits, echocardiographic examinations, and catheterizations to estimate potential risk factors for neopulmonary stenosis. In the case of missing data or more than 1 year since the last follow-up visit, patients were contacted for a hospital follow-up examination with full echocardiographic examination. Surgical Technique All ASO operations were performed by one surgeon (J.J.M.) under moderate hypothermia (24 to 26 C) during cardiopulmonary bypass and with blood cardioplegia infused into the aortic root and afterward reinfused selectively into the coronary artery ostia at regular intervals (approximately 20 minutes). In most cases, the pulmonary trunk was anastomosed to the native aortic root without a patch. The ascending aorta was surgically prepared up to the supraaortic branches, as well as the proximal part of the PA, to release the proximal parts of the great vessels and increase their mobility. To minimize the risk of stenosis at neopulmonary anastomosis site, two modifications to the standard ASO were introduced. A double incision in the pulmonary anterior and posterior walls increased the length of the suture line and allowed filling gaps in the posterior wall of the original aorta created by the harvest of the coronary arteries (Fig 1). To facilitate a direct connection, the pulmonary trunk was transected as low as possible, usually just 2 to 3 mm above the pulmonary valve commissures and the aorta high, usually 5 to 7 mm above the commissures (Fig 2). These modifications of the transection lines allowed for wide and tensionless neopulmonary anastomosis and a shortened aortic arch, which minimized the risk of aortic pressure on the coronary arteries after the Lecompte maneuver (performed in each case). Most of the pulmonary reconstructions were performed with a beating heart. It was necessary to maintain cardioplegic arrest for this procedure due to the right coronary artery being near to the suture line in 2 patients with the coronary anomaly of the right coronary artery arising from the left coronary artery. The coronary arteries were excised with the surrounding circular part of the aortic wall and relocated to a properly adapted part of the pulmonary wall by using a modified trapdoor method. The incisions in the primary pulmonary sinus are very important for proper coronary artery transplantation because the arteries should be transplanted at the same level at which they arise from the aortic sinuses. The right coronary artery should be transplanted to a neoaortic incision, which is usually made high above the valve and parallel to the transection line of the vessel (Fig 3A and B). The left coronary artery is typically transplanted deep into the sinus because of its primary pattern. The incision on the anterior neoaortic wall has, in this case, the shape of a hockey stick (Fig 3A and B). The depth of the incision depends on the primary position of the coronary artery ostium in the aorta. The shape and depth of incisions on the neoaortic wall are important for proper coronary transfer and to avoid potential complications such as tension or torsion of the coronary arteries.

3 Ann Thorac Surg MOLL ET AL 2012;93: DIRECT PULMONARY RECONSTRUCTION 179 Echocardiographic Examination All echocardiographic records were reviewed to estimate the transpulmonary Doppler gradient and to confirm pulmonary stenosis and its potential development. To evaluate SVPS, the preoperative echocardiograms were compared with all postoperative echocardiographic observations. After their operations, patients were evaluated in ambulatory care each year, with periodic examinations in the cardiology department, typically each year during the first 5 years and every 3 to 4 years thereafter. When a significant neopulmonary stenosis gradient or another complication, such as neoaortic regurgitation, was detected in the ambulatory screening echocardiographic assessment, the patient was admitted to the cardiology department for a full clinical and echocardiographic examination. All instances of significant neopulmonary stenosis were evaluated and confirmed by a senior pediatric cardiologist (J.A.M.). Neopulmonary stenosis was estimated using peak flow velocity in the PA in 2-dimensional echocardiography with Doppler imaging. Because of the heart anatomy after the ASO procedure with the Lecompte maneuver, the subcostal projection for pulmonary flow estimation was preferred. The Bernoulli formula was used to estimate pressure gradient. Stenosis with a peak gradient (in mm HG) of less than 25 (PA peak flow velocity 2.5 m/s) was considered insignificant, 25 to 50 was mild, 50 to 75 was moderate, and exceeding 75 was severe [4, 15]. Cardiac catheterization was performed when the peak flow gradient exceeded 40 mm Hg [15]. Fig 1. Direct anastomosis of the pulmonary artery: double incisions on the anterior and posterior wall (A) facilitate direct connection of the transected vessel (B). In the remaining cases, when local adaptation of the pulmonary trunk was ineffective and the risk of compression of the coronary arteries was significant, a large pantaloon-shaped patch of fresh autologous pericardium was used to reconstruct the sinus portion of the neo-pa. The factors associated with choice of patch reconstruction are complex and include evaluation of the coronary artery pattern after transplantation and possible compression of the coronary arteries by the posterior wall of the anastomosed neo-pa. Generally, after transection of the PA and mobilization of its proximal part, it is essential to evaluate the possibility of a tensionless suture of its posterior wall, which is not always possible in the case of specific local anatomic conditions (ie, high position of the PAs or large disproportion between the pulmonary and aortic valves). In cases of a possible threat of coronary artery compression by a tight neopulmonary posterior wall, it is better to use a pericardial patch for a tensionless anastomosis. Fig 2. Transection lines of the great vessels during the arterial switch operation. The white dots show the usual location of pulmonary artery transection line, and the yellow dots show our modification. The white lines show the usual location of the aorta transection line, and the yellow lines show our modification.

4 180 MOLL ET AL Ann Thorac Surg DIRECT PULMONARY RECONSTRUCTION 2012;93: Fig 3. (A, B) Incisions on the anterior wall of the pulmonary sinus (neoaortic) for coronary artery transplantation. Incision for the right coronary artery is placed high above the valve and parallel to the transection line. A second incision for the left coronary artery goes deep to the sinus of the neoaorta and has the shape of a hockey stick. Statistical Analysis Statistical analysis was performed using Statistica 9.0 software (StatSoft Inc, Tulsa, OK). Frequencies are presented as percentages, and continuous data are presented as means standard deviations. The occurrence of pulmonary stenosis was analyzed and presented as a survival probability (Kaplan-Meier), with log-rank test analysis for the difference between patients with direct and patch reconstruction of the neo-pa. PA peak gradient was analyzed separately for each year of follow-up using the analysis of variance (ANOVA) test presented as error bar plots and correlated with time of observation using Spearman rank-order correlation coefficient. Potential risk factors for SVPS were analyzed using univariate analysis: two-tailed Fisher exact test or 2 test for qualitative data and the Student t test or Mann-Whitney U test for quantitative data. Multivariate analysis was performed using logistic regression and modified Cox multivariate regression for all analyzed risk factors. A value of p 0.05 was considered statistically significant. Results The mean follow-up for the study group was years (range, 1 to 18 years). Pulmonary stenosis occurred in 9 patients (2.6%; Table 2). In 8 of 9 patients, significant pulmonary stenosis occurred in the first year after the operation and did not increase with the time of observation. Table 2. Patients With Significant Pulmonary Artery Stenosis Pt Operative Data Age (days) Weight (g) Operation Date Neo-PA Anastomosis SVPS Pressure Gradient (mm Hg) Reinterventions Clinical Data 1 6 3,300 7/14/1993 Pericardial patch ,900 7/19/1992 Pericardial patch 36 None Cx inverted, nonfacing commissures, PI: I0 60 BVP, 1999r; BVP, 2001r Intramural pattern of LCA 3 4 3,800 1/3/1992 Direct 30 None Nonfacing commissures, PI: II ,000 7/2/1997 Direct 28 None PI trivial 5 7 3,800 7/71997 Pericardial patch 31 None PAV AoV discrepancy, coronary anomalies (LCA, RCA from right sinus of aorta), PI: I ,100 7/24/1997 Direct 32 None Nonfacing commissures, PI: trivial 7 5 3,700 2/12/1997 Direct 30 None Cx inverted, PI: trivial 8 4 2,900 2/23/1997 Direct 29 None PAV AoV discrepancy, nonfacing commissures 9 4 3,500 12/25/1995 Pericardial patch 28 None PAV AoV discrepancy, Cx inverted, PI: trivial AoV aortic valve; BVP balloon valvuloplasty; Cx circumflex coronary artery; LCA left coronary artery; PA pulmonary artery; PAV pulmonary artery valve; PI pulmonary artery valve insufficiency; RCA right coronary artery; SVPS supravalvular neopulmonary stenosis.

5 Ann Thorac Surg MOLL ET AL 2012;93: DIRECT PULMONARY RECONSTRUCTION 181 Fig 4. Kaplan-Meier analysis shows freedom from significant pulmonary stenosis (PS) in patients with direct pulmonary artery (PA) reconstruction and reconstruction with patch. SVPS occurred in 4 patients (14.2%) with patch reconstruction of the neo-pa artery and in 5 patients (1.6%) with direct PA reconstruction. This difference was statistically significant in univariate analysis (Fisher exact test p 0.003) and is shown as a survival probability graph with statistically significant differences (log-rank test p 0.001; Fig 4). As summarized in Table 2, in patients with direct PA reconstruction, the 5 patients with significant neopulmonary stenosis had a mild stenotic gradient, whereas in patients with patch reconstruction, 3 patients had mild and 1 patient had medium pulmonary stenosis. The patient with medium pulmonary stenosis underwent balloon angioplasty twice. The first angioplasty procedure was successful, with the initial pulmonary stenosis gradient of 60 mm Hg decreased to 33 mm Hg, but restenosis was detected in an echocardiographic examination 1 year later. After 1 year of observation with echocardiographic evaluation every 6 months, the pulmonary stenosis gradient reached 57 mm Hg. A second angioplasty was performed 2 years after the first angioplasty, with a gradient reduction of approximately 35 mm Hg. The patient currently has stable pulmonary stenosis, with a peak flow velocity of 2.8 m/s (pressure gradient, 32 mm Hg), and in the 9 years after the second angioplasty, he has not required any interventions. None of the patients in the study group required reoperation to treat pulmonary stenosis. Among the 545 patients with TGA, there were only 2 reoperations (0.3%) due to right ventricle outflow tract obstruction in 2 patients excluded from the present study (TGA associated with aortic arch anomalies). Time-related analysis of PA flow peak gradient did not reveal significant differences between particular years of observation (ANOVA test; F 0.2; p 0.99; Fig 5) when we analyzed all data collected in yearly intervals. However, there was a significant positive correlation between the average peak pulmonary gradient and the time of observation (R 0.53, p 0.018). In the 9 patients with significant postoperative pulmonary stenosis, we did not observe a significant increase of stenosis gradient after the first year of observation. The exception was the patient with moderate stenosis who was found to have restenosis after the first balloon angioplasty. Among the risk factors analyzed (patch reconstruction, neopulmonary valve insufficiency, PA valve aortic valve discrepancy, nonfacing commissures, age at operation, weight at operation, aortic cross-clamp time, coronary anomalies, and intensive care unit length of stay) only patch reconstruction (p 0.003) and non-facing commissures (p 0.03) were significantly correlated with neopulmonary stenosis occurrence in univariate analysis (Table 1). Multivariate logistic regression confirmed patch reconstruction (odds ratio, 27.5; 95% confidence interval [CI], 3.5 to 215; p 0.002) and nonfacing commissures (odds ratio, 11.1; 95% CI, 2.1 to 58; p 0.007) were both independent risk factors for SVPS. By Cox proportional hazards regression, patch reconstruction ( 2.77; 95% CI, 1.03 to 4.51; p 0.002) and nonfacing commissures ( 1.96; 95% CI, 0.54 to 3.37; p 0.006) were also significantly correlated with SVPS. Pulmonary insufficiency was a common finding in our study group, occurring in 225 patients (65%). At the end of follow-up, we observed trivial insufficiency in 128 patients (37%), and mild regurgitation in 97 (28%). Although our surgical technique slightly reduces the length of the aortic arch, we did not observe any patients with of newly developed stenosis or a flow velocity increase in the aortic arch during the entire follow-up period. Comment Pulmonary artery stenosis is a common complication of the ASO and still remains the most frequent reason for

6 182 MOLL ET AL Ann Thorac Surg DIRECT PULMONARY RECONSTRUCTION 2012;93: Fig 5. Average pulmonary artery flow peak gradient during follow-up after the arterial switch procedure. reinterventions and reoperations. The incidence of PA stenosis varies from 1% to 35% and is responsible for about 75% of all reoperations and reinterventions [3, 4, 6, 7, 11 19]. The reoperation rate due to PA stenosis ranges from 3% to 25% [6,7, 11 20]. This wide range is related to varying definitions and PA stenosis criteria, as well as the length of follow-up. The high incidence of PA stenosis is the reason that surgical methods of neopulmonary reconstruction still remain a matter of concern [4, 14]. In the classic neo-pa reconstruction, a large bifurcated patch [4, 21] or 2 button patches [1] are used to fill the gaps left by the excision of the coronary ostia. Despite good postoperative results and tensionless anastomosis, using a patch may be the reason behind the development of postoperative neopulmonary stenosis [13, 16, 19]. However, numerous articles support the thesis that reconstruction of the neo-pa with a patch helps avoid late pulmonary stenosis [3, 4, 18]. The number and size of patches used in neo-pa reconstruction may also affect the frequency of neo-pa stenosis. For example, very good results with a low incidence of SVPS have been described for large, pantaloon-shaped patches [3, 14], as well as for double-patch reconstruction of the neo-pa [6, 22]. Ullmann and colleagues [4] reported a modified surgical technique with 2 patches used in the ASO, one small circular patch and the other a large triangular-shaped patch used for extended reconstruction of the pulmonary root and the posterior wall of the pulmonary trunk [4]. In this study, there were no reinterventions because of pulmonary stenosis during almost 9 years of follow-up; however, more than 25% patients had signs of pulmonary stenosis, with the peak flow gradient exceeding 25 mm Hg. Direct anastomosis, described for the first time by Pacifico and colleagues [12], is an interesting alternative to PA reconstruction with a pericardial or prosthetic patch. It ensures a good postoperative outcome, with low rates of stenotic complications and need for reintervention [13, 23]. In contrast, Serraf and colleagues [3] and Prifti and colleagues [14] indicate that direct reconstruction of the neo-pa increases the risk of pulmonary stenosis development. It seems that the best surgical technique for neopulmonary reconstruction is the one that is best known by the surgeon. All available techniques have advantages, and their excellent results are confirmed by a large number of studies. Our results show that direct neopulmonary reconstruction is a good option with an acceptable long-term incidence of SVPS and a low rate of complications and reinterventions. During this procedure, no graft is used to reconstruct the neo-pa, so the lower probability of complications may make this method preferable to patch reconstruction. We believe that even fresh autologous pericardium used as a graft has no or limited growth potential compared with the surrounding tissues. Pericardial patches fill approximately two-thirds of the pulmonary circumference; thus, with less growth potential compared with the surrounding tissues, even with some patch dilatation, it may not be sufficient for proper neopulmonary development. With time, the pericardial patch can also shrink, stiffen, and undergo calcification [24 26], a process that may be responsible for patch retraction and distortion as well as stretching of the PAs. Thus, the patch can be a reason for late local complications, such as a high SVPS incidence and an increasing occurrence of pulmonary stenosis, as follow-up continues. In our series, most of the pulmonary stenosis occurred in the first year after the switch procedure, with no progression of stenosis during follow-up; that is, significant stenosis did not develop in the long-term follow-up in patients with no or trivial gradient in the first years of observation. There was, however, a significant correlation of each year s average peak pulmonary gradient with

7 Ann Thorac Surg MOLL ET AL 2012;93: DIRECT PULMONARY RECONSTRUCTION 183 the length of follow-up, but taking into consideration not only averages but all data collected in each year of observation gives opposite results, with no differences between particular years of observation (Fig 5). Analysis of average differences in Figure 5 (red dots) shows that even if we consider this linear correlation as statistically significant, the difference between average peak gradient during 18 years of observation is between 1 and 3 mm Hg, which is clinically insignificant. Furthermore, these values are within the range of acceptable error of echocardiographic Doppler measurements. However, the ANOVA test is more trustworthy because it analyzes all data collected, not only the averages. Such observations in patients with simple TGA after ASO were reported previously by Choi and colleagues [11] and Serraf and colleagues [3]. Direct neopulmonary anastomosis with the Lecompte maneuver is the procedure of choice during the ASO in our Cardiosurgery Department, and thus, patients with patch reconstruction comprised highly selective cases with poor local condition, which necessitated using a patch for PA reconstruction because of the risk of postsurgical coronary compression. Nevertheless, in our study group, patch reconstruction was an independent risk factor for pulmonary stenosis in multivariate analysis, which may reflect a true causative relationship or the effect of other factors not analyzed in our series. In keeping with the reports from centers where patch reconstruction is a primary method, our results demonstrate the value of direct anastomosis. In our study, presence of nonfacing commissures was also a significant risk factor for neopulmonary stenosis. This factor is directly related to coronary artery transplantation and its postoperative pattern. In some cases, it forces coronary arteries transplantation to the same sinus of the neoaortic root and, thus, modification of neo-pa anastomosis. Numerous descriptions of other risk factors for SVPS had been reported, including a two-stage repair with previous PA banding and hypoplastic aortic arch [3]. Our study did not confirm these possible risk factors because we excluded all patients with an aortic arch anomaly, previous PA banding, and ventricle septal defect associated with TGA. However, we had few patients in our study group with pulmonary stenosis, and other potential risk factors may appear to be statistically insignificant because of the small number of patients with significant SVPS. Neopulmonary valve regurgitation was a common finding in our study group. Typically, it manifested as an early diastolic, tinny jet, classified as trivial or mild insufficiency with no hemodynamic significance and no progression with follow-up. None of our patients required reoperation because of neopulmonary insufficiency. The incidence of neopulmonary regurgitation is, however, surprising because primary aortic valves usually have no signs of insufficiency. On the other hand, trivial or mild pulmonary insufficiency is also a common finding in healthy children [27], suggesting that lowpressure circulation may predispose to worse coaptation of the semilunar valves. Data on pulmonary insufficiency incidence in the published literature are limited [11, 28], and its frequency varies from 6% to 50%. In most patients, pulmonary insufficiency is trivial or mild, does not progress with time, has no hemodynamic implications, and does not require any reinterventions [11, 28]. In conclusion, direct PA anastomosis during ASO is an interesting alternative to patch reconstructions that ensures a good postoperative outcome with low rates of complications and SVPS. Two modifications to the switch procedure in this series of patients improved on the known surgical technique in an effort to minimize the risk of stenosis at the site of anastomosis. The main value of direct anastomosis is that no graft is used during the operation, which ensures a good prognosis for the future growth and development of the anastomosed neo-pa. References 1. Jatene AD, Fontes VF, Paulista PP, et al. Anatomic correction of transposition of the great vessels. J Thorac Cardiovasc Surg 1976;72: Castaneda AR, Norwood WI, Jonas RA, Colon SD, Sanders SP, Lang P. Transposition of the great arteries and intact ventricular septum: anatomical repair in the neonate. Ann Thorac Surg 1984;38: Serraf A, Roux D, Lacour-Gayet F, et al. Reoperation after the arterial switch operation for transposition of the great arteries. J Thorac Cardiovasc Surg 1995;110: Ullmann MV, Gorenflo M, Bolenz C, et al. Late results after extended pulmonary artery reconstruction in the arterial switch operation. Ann Thorac Surg 2006;81: Williams WG, McCrindle BW, Ashburn DA, Jonas RA, Mavroudis C, Blackstone EH. Outcomes of 829 neonates with complete transposition of the great arteries years after repair. Eur J Cardiothorac Surg 2003;24: Prêtre R, Tamisier D, Bonhoeffer P, et al. Results of the arterial switch operation in neonates with transposed great arteries. Lancet 2001;357: Losay J, Touchot A, Serraf A, et al. Late outcome after arterial switch operation for transposition of the great arteries. Circulation 2001;104: Culbert EL, Ashburn DA, Cullen-Dean G, et al. Quality of life of children after repair of transposition of the great arteries. Circulation 2003;108: Pasquali SK, Hasselblad V, Li JS, Kong DF, Sanders SP. Coronary artery pattern and outcome of arterial switch operation for transposition of the great arteries: a metaanalysis. Circulation 2002;106: Castaneda A. Arterial switch operation for simple and complex TGA indication criteria and limitations relevant to surgery. Thorac Cardiovasc Surg 1991;39: Choi BS, Kwon BS, Kim GB, et al. Long-term outcomes after an arterial switch operation for simple complete transposition of the great arteries. Korean Circ J 2010;40: Pacifico AD, Stewart RW, Bargeron LM. Repair of transposition of the great arteries with ventricular septal defect by an arterial switch operation. Circulation 1983;68: Carrel T, Mattila I, Pfammatter JP, Leijala M. Direct reconstruction of the pulmonary artery during the arterial switch operation: an interesting surgical option with excellent hemodynamic results. Ann Thorac Surg 1998;65: Prifti E, Crucean A, Bonacchi M, et al. Early and long term outcome of the arterial switch operation for transposition of the great arteries: predictors and functional evaluation. Eur J Cardiothorac Surg 2002;22: Chen YS, Tsai SK, Chang CI, et al. Prediction of early pulmonary artery stenosis after arterial switch operation: the

8 184 MOLL ET AL Ann Thorac Surg DIRECT PULMONARY RECONSTRUCTION 2012;93: role of intraoperative transesophageal echocardiography. Cardiology 2008;109: Brown JW, Park HJ, Turrentine MW. Arterial switch operation: factors impacting survival in the current era. Ann Thorac Surg 2001;71: Yamaguchi M, Hosokawa Y, Imai Y, et al. Early and midterm results of the arterial switch operation for transposition of the great arteries in Japan. J Thorac Cardiovasc Surg 1990; 100: Jatene MB, Jatene IB, Oliveira PM, et al. Prevalence and surgical approach of supravalvular pulmonary stenosis after Jatene operation for transposition of great arteries. Arq Bras Cardiol 2008;91: Nakano T, Kado H, Hinokiyama K, et al. Long-term results of arterial switch operation. Kyobu Geka 2008;61: Prandstetter C, Hofer A, Lechner E, Mair R, Sames-Dolzer E, Tulzer G. Early and mid-term outcome of the arterial switch operation in 114 consecutive patients: a single centre experience. Clin Res Cardiol 2007;96: Paillole C, Sidi D, Kachaner J, et al. Fate of pulmonary artery after anatomic correction of simple transposition of great arteries in newborn infants. Circulation 1988;78: Hövels-Gürich HH, Seghaye MC, Ma Q, et al. Long-term results of cardiac and general health status in children after neonatal arterial switch operation. Ann Thorac Surg 2003;75: Kawata H, Kishimoto H, Iwai S, et al. Long term outcome of arterial switch surgery for transposition of the great arteries: evaluation of the reconstruction of the pulmonary artery. Kyobu Geka 2008;61: Eloakley RM, Grotte GJ. Severe calcification of a parietal pericardial patch used in an aortic root enlargement: case report. J Thorac Cardiovasc Surg 1994;107: Kadowaki MH, Levett JM, Manjoney DL, Grina NM, Glagov S. Comparison of prosthetic graft materials as intracardiac right atrial patches. J Surg Res 1986;41: Ross EM, McIntosh CL, Roberts WC. Massive calcification of a right ventricular outflow tract parietal pericardial patch in tetralogy of Fallot. Am J Cardiol 1984;54: Takao S, Miyatake K, Izumi S, et al. Clinical implications of pulmonary regurgitation in healthy individuals: detection by cross sectional pulsed Doppler echocardiography. Br Heart J 1988;59: Hwang HY, Kim WH, Kwak JG, et al. Mid-term follow-up of neoaortic regurgitation after the arterial switch operation for transposition of the great arteries. Eur J Cardiothorac Surg 2006;29:162 7.