Results and Complications After Single-Stage Repair of Aortopulmonary Window and Interrupted Aortic Arch in a 32-Week Preterm and a Full-Term Neonate

Background

An aortopulmonary window (APW) is a congenital defect of septation between the ascending aorta and the pulmonary trunk. APW accounts for 0.2% of congenital heart disease and occurs in isolation or in association with other cardiac anomalies in 52% of cases, including atrial septal defect, ventricular septal defect, interrupted aortic arch, coarctation of the aorta, and tetralogy of Fallot [1]. The combination of APW and interrupted aortic arch (IAA) is extremely rare, accounting for 0.046% of patients with congenital heart disease and 4% to 6% of patients with IAA [1,2]. The Richardson classification system was used to define the position of aortopulmonary septal defects, which comprises type I (proximal), type II (distal), and type III (proximal and distal) [3,4]. For the classification of IAA in this study, we adopted the system of Celoria and Patten, which categorizes IAA into type A (distal to the left subclavian artery), type B (between the left carotid and left subclavian arteries), and type C (between the innominate artery and the left carotid artery) based on the missing segment of the aortic arch [5,6].

In patients with APW and IAA, an excessive left-to-right shunt across the APW leads to pulmonary edema, elevated left atrial pressure, and left heart volume overload. Additionally, due to the presence of IAA, cerebral perfusion heavily relies on the output from the left heart, which can be significantly diminished by the substantial shunting through the APW. Lower visceral organ perfusion is sustained through the ductus arteriosus, connected to the descending aorta. The gradual decrease in pulmonary vascular resistance further diminishes blood flow through the ductus arteriosus into the descending aorta over time. Symptoms typically develop rapidly within 1 to 2 weeks, and are characterized by respiratory distress, poor feeding, reduced urine output, and cold extremities. Echocardiography is frequently diagnostic. In current practice, an electrocardiography-gated computed tomography (CT) scan serves as a valuable modality for detailed interpretation and classification. Diagnostic cardiac catheterization is often unnecessary when either echocardiography or CT can clearly establish the diagnosis [3–6].

Preoperatively, prostaglandin E1 infusion is mandatory, while inotropic support, mechanical ventilation, and parenteral nutrition supplements are used according to clinical presentation. Currently, a single-stage reconstruction performed within a few weeks of life is the preferred approach, according to published case series. Mortality rates for this rare disease have shown improvement over time, ranging from 9% to 38% in various series [1,2,7,8].

This report is of 2 cases of congenital APW associated with IAA in a 32-week premature neonate with type I APW and type A IAA and a term neonate with type III APW and type A IAA, both treated with a single-stage surgical repair. We introduce the detailed presentation, images, surgeries, complications, and outcomes of these 2 cases.

Case Reports

CASE 1:

A male newborn with weight of 1789 g and at gestation age of 32 weeks presented with a heart murmur and progressive respiratory distress 3 days after birth. Echocardiography revealed a direct connection between the ascending aorta and pulmonary trunk in the high parasternal short axis view, indicating a type I APW (Figure 1A). With the suprasternal arch view, the site of the arch disruption after left subclavian artery was defined and was compatible with the diagnosis of IAA type A (Figure 1B). CT scanning revealed the same results. Prostaglandin E1 infusion, mechanical ventilator, and parenteral nutrition supplementation were started.

Single-stage surgical repair was performed at 3 weeks of age (weight 1948 g) for worsening heart failure (Figure 1C). By the transwindow approach, the APW was divided and defects of the ascending aorta and main pulmonary artery were repaired with 0.6% glutaraldehyde-treated autologous pericardial patches. The aortic arch and descending aorta were then fully mobilized, followed by direct end-to-side anastomosis. The patient developed oliguria, elevated lactate (57 mg/dL), and weak femoral artery pulsation on postoperative day (POD) 1. Echocardiography revealed recurrent arch obstruction with a pressure gradient of 54 mmHg at the anastomosis site of the IAA repair. The treated autologous pericardial patch was applied to augment the aortic arch. The patient was extubated on POD 37 and discharged from the hospital on POD 77.

At a 6-year follow-up, the patient was asymptomatic, with normal growth and development. Echocardiography revealed mild obstruction of the aortic arch, with a gradient of 23 mmHg. No obstruction was detected at the main pulmonary artery or ascending aorta.

CASE 2:

A male infant at gestational age 38 weeks (term) presented with tachypnea and subcostal retraction at 5 days old. Chest X-ray showed pulmonary edema and cardiomegaly. Echocardiography and CT revealed the deficiency of septation between the right pulmonary artery (RPA) and ascending aorta, forming the W-shaped communication between the great vessels, which indicated APW type III (Figure 2A, 2B). IAA type A was also noticed. Single-stage surgical repair was performed at 16 days of age (weight 3548 g; Figure 2C). The APW type III was divided using the transwindow approach. The defects of the ascending aorta and RPA and partly involved main pulmonary artery were repaired with a 0.6% glutaraldehyde-treated autologous pericardial patch. The descending aorta was mobilized extensively. The end-to-side anastomosis was administrated with a treated autologous pericardial patch augmentation.

For left lung atelectasis, we arranged bronchoscopy on POD 7, which revealed C-shaped tracheal rings with a wider intrusion (<50%) of posterior membrane on expiration and non-pulsa-tile obstruction of left main bronchus (Figure 3A, 3B). There was mild tracheobronchomalacia and compression from tissue swelling. Left vocal cord paralysis was detected simultaneously, which might have been related to recurrent laryngeal nerve injury. Instead of immediate surgery, we extubated on POD 7 and applied prolonged high-flow nasal cannula supports afterward. Bronchoscopy on POD 66 showed resolution of the left bronchial obstruction (Figure 3C), while the left vocal cord remained paralyzed. Notably, during echocardiography follow-ups, the residual pressure gradient at the arch anastomosis increased from 19 mmHg to 27 mmHg to 34 mmHg on PODs 8, 17, and 73, respectively. This patient was discharged from the hospital on POD 75, without oxygen therapy. By echocardiography, there was a pressure gradient of 18 mmHg across the arch anastomosis at the 1-year follow-up, indicating gradual resolution of stenosis at the arch anastomosis.

Discussion

Treatment for neonates with APW and IAA could be challenging. Currently, a single-stage operation within a few weeks of life is the preferred approach, according to published case series. Mortality rates have shown improvement over time, ranging from 9% to 38% in various series [1,2,7,8]. Even in low-birth-weight premature neonates, single-staged repair is feasible [9,10]. However, the single-stage reconstruction is associated with a high rate of re-intervention for complications such as aortic arch, bronchial, and pulmonary artery obstructions [1,2,7,8]. Staged repair can be considered in critical conditions, with stenting of ductus arteriosus and bilateral pulmonary artery banding. It can pose challenges in subsequent operations, including difficulties in stent removal during arch reconstruction and the potential need for pulmonary arterioplasty after de-banding [11,12]. Herein, we would like to compare our 2 cases with previously published cases regarding surgical techniques, complications, and outcomes.

In contemporary published cases about the surgical repair of APW and IAA, surgeons divided the APW through various approaches, such as transpulmonary, transwindow, or transaortic. Closure techniques include direct suture, single-patch repair, and double-patch repair (of both the aorta and pulmonary artery). Intra-aortic baffle was another novel technique for APW repair, proposed by Nguyen et al [7]. Arch reconstruction procedures include anastomosis with or without patch augmentation, or using an interposition graft. Additionally, the ductus arteriosus is ligated or divided [1,2,7–10].

The largest retrospective study was published by Konstantinov et al. Their case series included 20 patients with APW and IAA reported from 1987 to 1997, of which 15 patients (79%) underwent a single-stage repair. The infants weighed 2.1 to 4.3 kg, and arch obstruction or bronchial compression that required re-intervention occurred in 10 patients (51%), with surgical repair for residual arch obstruction in 2 patients, transcatheter balloon dilatation in 6 patients, mobilization of arch and bronchus for bronchial compression in 1 patient, and placement of an aortic graft for relief of bronchial compression in 1 patient. The rate of freedom from aortic re-intervention was 56% at 1 year, 48% at 5 years, and 46% at 10 years. Arch re-intervention was most prevalent within the first year after the initial repair, accounting for 10 events. Notably, the aortic arch was repaired directly without patch augmentation in 15 patients, with pericardial patch augmentation in 2 patients. The mean weight at repair was lower in patients with patch augmentation than without (2.3 kg vs 3.1 kg). The authors stated that avoidance of patch augmentation is associated with an increased risk of subsequent arch obstruction and likely contributes to bronchial compression [1].

Nguyen et al [7] published intermediate outcomes for single-stage surgical repair in 11 patients with APW and IAA. They performed a novel surgical technique of single-stage repair by using an intra-aortic baffle with a boomerang patch for aortic reconstruction. Specifically, this intra-aortic baffle technique was applied in 3 patients with APW type II and 3 patients with APW type III. We considered that this novel technique could significantly decrease the bypass time, which contributes to less end-organ damage in neonates. Furthermore, the combination of intra-aortic baffle and patch augmentation during arch repair might decrease the rate of recurrent aortic arch stenosis. However, the study revealed that 2 patients with APW type III developed RPA stenosis, necessitating RPA balloon angioplasty shortly after the operation. The authors attributed this to the patch shifting toward the pulmonary artery, causing RPA stenosis and limiting potential RPA growth. Additionally, in a 2.5-kg neonate with APW type II and IAA type A, the intra-aortic baffle technique with direct suture of the arch was used. This patient later required aortic arch re-intervention with balloon angioplasty [7].

In our report, patient 1, weight 1948 g, who received end-to-side anastomosis without patch for IAA developed oliguria and elevated lactate levels from residual aortic arch obstruction within 24 h of surgery. Patch augmentation for anastomosis was performed promptly on POD 1. The aorta of patient 2 was repaired by end-to-side anastomosis with patch augmentation. He presented with progressive arch stenosis at anastomosis shortly after operation but it resolved after 1 year of follow-up. In patients with lower weight (<2.0 kg), patchy augmentation with aortic arch anastomosis may be a better approach for IAA repair. For APW repair, both of our cases used the transwindow approach with double-patch closure. There was no postoperative pulmonary artery obstruction. We considered that the transwindow approach, especially in APW type III, can greatly expose the defect of the RPA during APW repair. The double-patch closure technique could prevent patch shifting or compression from the aorta toward the RPA.

Bronchial compression is one of the major complications observed after arch anastomosis or prosthetic graft interposition. This results from tension between the aortic arch and descending aorta. This complication could be resolved with extensive mobilization of the descending aorta, patch augmentation for anastomosis, or placement of an interposition aortic graft [1,2]. Konstantinov et al encountered bronchial compression in 2 patients approximately 1 month after the repair. The diagnosis was made through bronchoscopy in 1 patient, with prolonged ventilator dependence, and by magnetic resonance imaging in the other patient, with refractory collapse of the left lower lobe. Neither patient underwent patch augmentation during the anastomosis of the IAA. The compression of the left bronchus airway could be related to undue tension at the end-to-side aortic anastomosis [1]. Our patient in case 2, with patch augmentation for arch anastomosis, developed left bronchial narrowing, but it resolved spontaneously 66 days later. Bronchoscopy on POD 7 revealed left main bronchus narrowing and intrusion (<50%) of posterior membrane of the trachea, especially on expiration. It may have been related to tracheobronchomalacia and tissue edematous compression around the surgical site. Gradual resolution with somatic growth could be possible. This presentation implied that bronchial compression may not warrant immediate re-intervention. Immediate re-intervention can be necessary in the following conditions: (1) external, pulsatile compression that results in complete obstruction and (2) significant compression that causes persistent left lung atelectasis or recurrent infection.

Conclusions

In conclusion, a single-stage surgery for APW and IAA is preferred. In the first case, we encountered postoperative arch obstruction and re-intervened promptly. This baby is one of the 3 reported cases with a lower birth weight, below 1800 g [6,7]. According to current experiences, lower body weight and direct end-to-side anastomosis without patch augmentation could be risk factors for re-intervention. The patient in the second reported case developed postoperative left vocal cord paralysis and left bronchial non-pulsatile narrowing, which improved over time after prolonged respiratory supports. He was discharged without oxygen therapy. This presentation implied that bronchial compression may not warrant immediate re-intervention unless there is complete obstruction, recurrent atelectasis, or infection. In both cases, we used the transwindow approach in conjunction with a double patch repair technique for APW repair. No instances of postoperative RPA stenosis were observed in either case. These 2 reports highlighted the feasibility of single-stage surgical repair while addressing challenges, including recurrent arch stenosis, bronchial compression, and RPA stenosis that need intervention. Further studies on the long-term outcomes of different surgical procedures would help us to clarify the proper way to avoid re-intervention.