Staged correction of pulmonary atresia, ventricular septal defect, and collateral arteries

Abstract Objectives Pulmonary atresia (PA) with ventricular septal defect (VSD) and systemic‐pulmonary collateral arteries (SPCAs) presents with variable anatomy with regard to the pulmonary vasculature, requiring personalized surgical treatment. A protocol consisting of staged unifocalization and correction was employed. Methods Since 1989, 39 consecutive patients were included (median age at first operation 13 months). In selected cases, a central aorto‐pulmonary shunt was performed as the first procedure. Unifocalization procedures were performed through a lateral thoracotomy. Correction consisted of shunt takedown, VSD closure, and interposition of an allograft between the right ventricle and the reconstructed pulmonary artery. Echocardiographic data were obtained postoperatively and at interval follow‐up. Results In 39 patients 66 unifocalization procedures were performed. Early mortality was 5%. Seven patients were considered not suitable for correction, of which four have since died. One patient is awaiting further correction. A correction was performed successfully in 28 patients. Operative mortality was 3% and late mortality was 11%. Median follow‐up after the correction was 19 years. Eleven patients required homograft replacement. Freedom from conduit replacement was 88%, 73%, and 60% at 5, 10, and 15 years respectively. Right ventricular function was reasonable or good in 75% of patients. All but one patient were in NYHA Class I or II. Conclusions After complete unifocalization 30/37 patients (81%) were considered correctable. The staged approach of PA, VSD, and SPCAs results in adequate correction and good functional capacity. RV function after correction remains reasonable or good in the majority of patients.

graphy before and after each procedure.
To evaluate the growth of the pulmonary arterial system, we retrospectively measured the pre-and postunifocalization  The change in lung perfusion on angiogram pre-and postunifocalization was studied retrospectively to evaluate the outcome of the procedure. We studied the total lung perfusion including perfusion by the SPCAs versus the lung perfusion by flow through the confluent pulmonary artery alone.
Patients were selected for total correction based on the angiography findings and measurement data from catheterization. Pulmonary hypertension or unfavorable anatomical result of unifocalization at angiogram were contraindications for total correction. The total correction was performed through a median sternotomy with the use of extracorporeal circulation and moderate hypothermia. The modified Blalock-Taussig and central shunts were divided. The VSD was closed with a Gore-Tex ® patch and a cryopreserved pulmonary homograft was interposed between the RVOT and the proximal pulmonary arterial system. Postoperative recovery and hospital or 30-day mortality are reported.
Long-term follow-up was obtained from the records. In 24 survivors with a complete repair, echocardiographic data were available with the exception of one patient who was lost to follow-up.
Magnetic Resonance Imaging (MRI) data for 17 patients after the successful correction was available to assess right ventricular function.
This study was approved by the Ethical Committee with no need for informed consent.
Statistical analysis was performed using Statistical Package for the Social Sciences (SPSS) software, version 24.0 (SPSS Inc.). Frequencies were given as absolute numbers and percentages. The data were expressed as median with range. The paired t-test was performed for statistical analysis. We applied the χ 2 test to compare frequencies in the two groups. The Kaplan-Meier method was applied to estimate freedom from reintervention and for survival. A p value less than .05 was considered to indicate statistical significance.

| Unifocalization
After initiation of the unifocalization program, older patients were retroactively introduced into the cohort, reflected by a higher age at first surgery and the age at correction. During the first 5 years, the mean age at first unifocalization was 5.9 years (range 3 months-14 years) and after this period the mean age was 1.8 years (range 1 month-16 years with only one older patient of 16 who came to our hospital from abroad). Thirty-nine consecutive patients with PA, VSD, and SPCAs were entered in the protocol. Of these 39 patients, 13 patients received a central shunt as the first procedure followed by 66 unifocalization procedures in which 129 collateral arteries were treated either with unifocalization or ligation. In seven patients a collateral artery was closed percutaneously.
In 50 unifocalization procedures, a modified Blalock-Taussig shunt was placed employing a 5-10 mm diameter tube, depending on the size of the patient. Later procedures most commonly featured a 5 mm tube prosthesis.
Based on several reports and our own clinical experience of the behavior of SPCAs, we tend to introduce more intrapulmonary VAN  anastomoses to avoid long segments of remaining SPCA tissue which could lead to stenosis or dilatation. 10,11 In one patient, the modified Blalock-Taussig shunt became obstructed 2 weeks after unifocalization, and the patient subsequently died after a reoperation in which the modified Blalock-Taussig shunt was replaced. One other patient died in hospital 3 months after unifocalization from an unknown cause. After unifocalization 20 additional procedures were required.
Apart from the coil closure of SPCAs, several other interventions were performed. In four patients a perigraft seroma around the modified Blalock-Taussig shunt was removed. 12 In one patient the modified Blalock-Taussig shunt was incorrectly placed on the pulmonary vein, and was revised 2 weeks later. Two patients received a modified Blalock-Taussig shunt and a further two received a central shunt, owing to the presence of cyanosis after the second unifocalization. In one patient a transannular patch was placed between the RVOT and the pulmonary trunk, which was later followed by balloon dilatation and stent implantation in the left pulmonary artery.
In two patients balloon dilatation of the modified Blalock-Taussig shunt was performed, with one patient also undergoing stent implantation. Two patients underwent revision of a central shunt with augmentation of the pulmonary artery, one with stent implantation and the other with balloon dilatation of the left pulmonary artery.
One patient underwent revision of the modified Blalock-Taussig shunt because of shunt occlusion. One patient underwent a sliding plasty of the distal trachea for stenosis with complete tracheal rings.

| Nakata-index and lung perfusion
The median Nakata-indices (n = first 22 consecutive patients), indicating the relative size of the pulmonary arteries were 143 (range 49-345) before and 190 (range 49-401) mm 2 /BSA after unifocalization (p = .097). The normal value of this PA-index is 330 mm 2 /BSA. 9 From 30 patients we obtained detailed information about their lung perfusion by studying their cardiac catheterization images before and after their unifocalization procedure(s). The median total lung perfusion was 99% (range 80%-100%) before and 97% (range 80%-100%) after unifocalization. Median lung perfusion through the confluent pulmonary artery was 52% (range 0%-100%) before and 89% (range 40%-100%) after unifocalization. Although there seems to be an increase in perfusion through the confluent pulmonary artery, this difference was not statistically significant (p = .12), likely due to small group sizes.

| Corrective surgery
About 30 of the 37 patients (81%) were considered correctable. One patient is awaiting correction after complete unifocalization. Twentynine patients underwent a total correction as described earlier which was successful in 28 patients (97%). Median age at correction was 2.6 years (range 1-17 years) and median interval between unifocalization and correction was 8 months (range 1-48 months). The VSD was closed in all patients. One patient required a takedown during operation because of suprasystemic pressures of the right ventricle making it impossible to wean the patient from the extracorporeal circulation. In retrospect we didn't find clear factors predicting this failure in comparison with other patients. One patient had a resternotomy for persistent bleeding the same day and recovered without complications.
Operative mortality was 3% (n = 1). This patient had a reoperation for leakage of the proximal suture of the homograft at the same day of the correction. However, due to extended cerebral hypoperfusion, the patient suffered a fatal cerebrovascular accident.
Seven patients were not eligible for correction. In three patients this was due to pulmonary hypertension. One patient had malperfusion of the complete left lung, one patient had a hypoplastic right ventricle which was not suitable for repair, one patient had hypoplasia of the lungs and airways in combination a neuro-cognitive disorder, and a further patient was considered to have an unacceptably high intraoperative risk of morbidity and mortality due to developmental disorders. Figure 1 shows a flowchart of our patient population.
In the 10 patients with 22q11 deletion, one is awaiting correction, three were ineligible (two died and one remains alive) for final correction, and six patients had a definitive correction.
One patient died postoperatively due to a cerebrovascular accident.  Four of the seven patients who were not suitable for correction died.

| Follow-up
One patient died 2 years after the last unifocalization due to respiratory failure and infection. One patient died of unknown cause 9 years after the last unifocalization procedure. One patient died 14 years after the last unifocalization of multiorgan failure and sepsis and one patient died of massive intracranial bleeding 9 years after the last unifocalization.
Among the five survivors with 22q11 deletion, one is awaiting correction, one is palliated in a reasonable condition, three had a definitive correction (one in reasonable condition and two in good condition). The overall survival in the 22q11 deletion patient was significantly (p = .041) lower compared to nonsyndromic patients,

| Interventions after correction
After the final correction in 21 patients, other interventions were performed. They are listed in Table 1

| Echocardiography
Echocardiographic data after correction at the last check, showed in 18 patients (75%) a reasonable or good right ventricular function (RVF). Four patients (17%) had a moderate RVF. Only two patients (8%) had a severely impaired RVF.
In seven patients a small residual VSD was present, without hemodynamic significance in terms of flow.   During long-term follow-up allograft degeneration and replacement occurred as expected. An explanation for this could be that these allografts are placed in a more extra-anatomical position, in patients with a younger mean age at correction. Also elevated right ventricular pressures, probably due to an abnormal pulmonary vascular bed, can contribute to this degenerative process. 28 Considering the complexities of this patient group, we are cognizant that these patients' life expectancies fall below normal, but lack long-term data comparable with others with right-sided allograft placement, that is, tetralogy of Fallot. 29 This study is limited by the relatively small group of patients.

| MR imaging
However, we treated all cases consecutively and without selection over the course of 30 years, employing an identical strategy by the same team. Given the limited case numbers, we did not perform subanalysis for example for patients operated on at the beginning of the case series, versus those who underwent this process more recently, after the team had gained considerable experience in the interim.
In conclusion, staged repair of PA, VSD, and SPCAs offers an adequate solution for most patients with a high correction rate and low operative mortality. Surgical or catheter-based interventions are necessary during follow-up for pulmonary artery stenosis and allograft degeneration. Long-term survival appears to be diminished.
We continue to follow these patients to collect more data on long-term outcomes. The function of the right ventricle is key a point of concern, as echocardiography and cardiac MRI show moderate and severe dilatation of the right ventricle (75%) and impairment of the function in 25% of the patients. Additional genetic abnormalities, in particular 22q11 deletion, appears to be risk factor for adverse outcomes. 30 Another limitation is the retrospective collection of data which results in incomplete data on exercise testing and quality of life studies. We are also looking to prospectively compare our long-term results to those of other strategies, to facilitate a better comparison between the different methods for the management of this challenging patient group.