Re-intervention and Late Events
No late deaths occurred during the follow-up. Recoarctation occurred only in one patient who underwent percutaneous aortic balloon angioplasty 13 months after the initial surgery. No false or true aortic arch aneurysms were detected in patients undergoing patch reconstruction (Fig. 1). Three neoaortic valves showed moderate regurgitation and were under close follow-up. Mild pulmonary stenosis was observed in one patient. None of the patients had an RVOT pressure gradient >30 mmHg. At the last follow-up, no reoperation was required for neoaortic valve dysfunction or right-sided obstructions. One tricuspid valve showed more than moderate regurgitation. This was noted in patient 7 who had DORV with a non-committed VSD and type B IAA. The patient has no significant symptoms and remained under close monitoring. For the two patients who had 5 mm VSD fenestrations, a 2 mm and 4 mm residual shunt could still be detected at the latest follow-up echocardiograms.
All surviving patients were in New York Heart Association (NYHA) class I, except for patient 10, who was 4-year-old and had reserved VSD fenestrations at the initial surgery. The patient’s condition remained NYHA class II, and the patient was still receiving bosentan orally for postoperative PHT. No cardiac medication was required for the other patients.
Discussion:
Only a few case reports and series, including single-digit patient numbers, exclusively documented the surgical procedures in the transposition complex associated with IAA.3–5 Pigott et al. first reported the single-stage strategy for TGA with aortic arch obstruction (AAO) in 1987.6 Moreover, in 1993, Planché et al. demonstrated superior outcomes of the single-stage repair over the two-stage repair for TGA, VSD, and AAO.2 The advantages of the single-stage repair may be due to the early reestablishment of the physiological circulation and aversion of detrimental effects of the pulmonary artery banding. Moreover, the hypoplastic aortic arch, coarctation, and interruption can be treated efficiently through a median sternotomy incision with mild tension on the anastomosis at the time of the ASO.7 Since 2000, single-stage repair for ventriculoarterial discordance associated with IAA has become the treatment of choice at our center.
Compared to other series, the age of the patients at the initial repair was significantly older in this series, which is closely related to the social and economic developmental factors in our country. Because of poor medical and health conditions at the grass-root level, most of the patients in this series were beyond the neonatal period when they were first referred to our center; moreover, there were two patients, a 4- and 6-year-old. To some extent, these patients were somewhat “naturally selected.” This may explain why the incidence of coronary artery anomaly was obviously low, and no right ventricle hypoplasia was detected preoperatively in our series, both of which remain important risk factors for this complex subgroup of patients, as stated in the report by Pocar et al.8 To complete accurate coronary artery transfer during ASO, coronary reimplantation after neoaortic reconstruction was recommended by some previous studies.9 In our series, the trap-door technique was used, but coronary malperfusion did not seem to be the direct cause of death. Of the three early deaths, only one had abnormal coronary artery anatomy, which was a single coronary artery originating from sinus 1. The premature neonate died of severe capillary leakage subsequently. In the other two patients, the direct causes of death were sepsis and treatment withdrawal because of multiple organ dysfunction.
Although coronary artery anomalies were uncommon, there were other risk factors in our series. Compared to patients with noncomplex TGA or TGA with arch coarctation, PHT is more common and persistent in this population, with complex forms of TGA associated with IAA. Postoperative PHT has been believed to be correlated with a prolonged stay in the ICU among patients with coexisting VSD,8 and implantation of a fenestrated atrial or ventricular septal patch or a one-way valved atrial patch has been advocated in patients with hypoplastic right ventricle (RV) or PHT.10,11 In our series, postoperative inhaled nitric oxide and oral sildenafil and bosentan were used to reduce pulmonary artery pressure in three patients who were >1 year old, among whom VSD fenestration was performed in two older children. In two of these three children, medications were stopped for a satisfactory decrease in pulmonary artery pressure and good exercise tolerance. This shows that PHT is reversible, even in older children with complex anomalies. However, a 4-year-old child at the time of operation required continued oral bosentan for persisting PHT. This demonstrates the complexity and diversity of PHT.
Size discrepancy between the great arteries has been reported to significantly complicate surgery and influence early and late outcomes, especially neoaortic valve regurgitation. The fate of the neoaortic root and valve is of particular concern after the ASO.12 A higher postoperative neoaortic root/ascending aorta ratio may be a risk factor for neoaortic regurgitation evolution at follow-up.13 Anastomosis of a dilated neoaortic root with a diminutive ascending aorta may cause supravalvular aortic stenosis, which has been reported as a cause of early death after an ASO.14 In previous studies, several procedures have been performed to overcome the mismatch in size, such as V-shaped resection of the posterior sinus,13 extensive patch enlargement of the entire arch and ascending aorta,7and relocation of the right coronary above the aortic anastomosis.15 Herein, sinus plasty after transfer of the coronary arteries was performed in 7 of the 11 patients. Without using additional material in the anastomosis of the neoaorta, the probability of major distortion and bleeding of the reconstructed sinotubular junction was minimized through a simple procedure. The integrity of the neoaortic root and apparent diminutive size of the ascending aorta could pose a concern. However, no supravalvular aortic stenosis or neoaortic root dilation was found on postoperative or follow-up echocardiography. Although three patients developed moderate neoaortic valve regurgitation, no reoperation was necessary.
Transpulmonary VSD closure has also been reported to be a significant risk factor for neoaortic valve regurgitation.13Moreover, in a previous study, it was believed that it was possible to close the VSDs through the tricuspid valve in nearly all cases.16 However, in our series, we closed the VSD through the right atrium in two patients with TGA, through the right atrium and the original pulmonary valve in seven patients with TBA, and through an infundibulotomy incision in one patient with TBA and one patient with DORV and non-committed VSD. The high probability of transpulmonary VSD closure in our series may be the reason for the relatively high rate of aortic valve regurgitation (3 of 8 surviving patients). However, we believe that a decent approach route depends on anatomic features, and irrespective of the route used, it is essential to carefully trim the VSD patch to avoid interference with the neoaortic valve.
In a report from Australia, the end-to-side arch repair was performed in four of five patients with transposition complex and IAA.5 The authors suggested that pericardial patch augmentation of the arch should be avoided, because it may result in dilation of the neoaorta and subsequent neoaortic valve insufficiency.5 However, in a series from France, extended end-to-end anastomosis was reported to be a risk factor for recoarctation; in another report from England, augmentation of the arch with a homograft patch was advocated for IAA repair.7Repair procedures performed for IAA at our center varied over time. In the first three cases before 2012, extended end-to-end anastomosis was performed. Since then, glutaraldehyde-treated autologous pericardial patch enlargement has been performed in the following eight patients, except for the 6-year-old child in whom a 16-mm interposition graft was implanted; therefore, this approach is our preferred choice for IAA repair in patients with associated anomalies. We believe that this approach may significantly reduce the tension of the anastomosis and the distortion of the neoaortic root.
In our series, we often completed IAA repair prior to ASO; hence, it is extremely important to tailor the patch to provide less of an acute angle of the arch, especially after performing the Lecompte maneuver and moving the ascending aorta posteriorly. Sometimes, two separate patches are required, with one moving up the ascending aorta and one moving down the descending aorta to create a smoother curve. Re-intervention for recoarctation was necessary for only one patient in whom percutaneous balloon angioplasty was performed uneventfully 13 months after the initial surgery, in contrast to the findings of several previous series with a high rate of re-intervention.7,13 Differences in surgical technique and thresholds of balloon dilation for recoarctation may explain such differences.