Percutaneous pulmonary debanding for an infant complicated by spontaneously closing muscular ventricular septal defect: A case report and in vitro study

There are several types of congenital heart disease which may be complicated by pulmonary hypertension (PH) following excessive pulmonary blood flow, such as large ventricular septal defect (VSD), double outlet right ventricle, and single ventricle. Primary repair is generally accepted as the best option for such diseases, although pulmonary artery banding (PAB) is occasionally indicated as a palliative procedure for situations in which primary repair has a considerable risk. We report on the case of a patient with coarctation of the aorta (CoA) complicated by multiple muscular VSDs (Swiss cheese type) for whom we performed arch repair and PAB as the first-stage surgery. His VSDs spontaneously became smaller following PAB, consequently, we chose to perform percutaneous pulmonary debanding rather than surgical debanding. We also created a model of the PAB and studied the anatomic development.

The boy had been referred to our hospital on day 3, because of cyanosis and a heart murmur after uneventful vaginal delivery with birth weight of 3080 g and Apgar score of 9 points. Transthoracic echocardiography demonstrated severe CoA and multiple muscular VSDs. Physical findings suggested acute peripheral circulatory failure complicated by peripheral cyanosis, coldness, and hypotension following complete closure of the ductus arteriosus. After stabilization with medical treatment, we repaired his aortic arch and banded his pulmonary artery (PAB) on day 6, using an expanded polytetrafluoroethylene (ePTFE) patch of 3 mm width and 0.4 mm thickness to decrease the pulmonary artery circumference to 21 mm employing polyester thread (4-0 Ticron^® (Medtronic, Dublin, Ireland)) and hexafluoropropene thread (6-0 Pronova^® (Johnson & Johnson, New Brunswick, New Jersey, United States)). The band was fixed to the pulmonary artery wall. Initially we planned to close the muscular VSDs between 6 and 12 months of age. However, outpatient department follow-up echocardiography showed that all VSDs diminished in size, so left to right shunt was reduced. Consequently, at 5 months, we scheduled cardiac catheterization to evaluate hemodynamics and VSD morphology, intending transcatheter debanding if the left to right shunt became small.

On admission, his body weight was 5.8 kg, while height was 64.0 cm. SpO₂ was 99%. No pressure difference was found between the upper and lower extremities, while there was a grade 3/6 systolic murmur over the left upper sternal border. There was no specific abnormality in his complete blood count and biochemistry, while B-type natriuretic peptide was 39.6 pg/ml. Cardiothoracic ratio was 54% with normal pulmonary vasculature. Electrocardiogram showed right ventricular hypertrophy. On echocardiography, left ventricular diastolic dimension was 22.6 mm, which was 100% of the estimated normal value, while the ejection fraction was 64%. There was no sign of re-CoA, while peak velocity in the main pulmonary artery distal to the PAB was 3.2 m/s. The pulmonary vessel diameters of main PA, PA banding, left PA, and right PA were 10 mm, 3.2 mm, 8.2 mm, and 7.3 mm, respectively. Although we could not determine how many VSDs remained open, the diameter of the largest remaining muscular VSD was only 1.1 mm with a few tiny VSDs on two-dimensional image, while peak velocity through it was 3.6 m/s.

We obtained informed consent from his parents to perform catheter intervention. Right ventricular systolic pressure, main pulmonary artery pressure, and left ventricular systolic pressures were 58, 25, and 95 mmHg, respectively. Pulmonary to systemic flow ratio measured by Fick’s method was 1.0, which gave a pulmonary vascular resistance of 1.7 wood units (Table 1). A left ventriculogram revealed trivial left-to-right shunt through multiple muscular VSDs. Based on these findings, we decided to do percutaneous debanding. As the lateral projection of the pulmonary angiogram showed a pulmonary valve diameter of 12 mm, and a PAB diameter of 4 mm (Fig. 1a), we chose the balloon diameter which was more than 50% of PAB circumference, and did not exceed a pulmonary valve annulus diameter. Consequently, we decided to dilate the PAB using a Mustang^® (Boston Scientific, Marlborough, Massachusetts, United State) balloon (12 mm diameter and 2 cm length). With balloon dilatation at 14 atmospheres, the diameter at the band site was dilated to 8.2 mm. However, a waist remained in the middle portion of the balloon (Fig. 1b). Further dilation using the same size of an extra-high pressure Conquest^® (Medicon, Osaka, Japan) balloon completely eliminated the waist (Fig. 1c). Final pulmonary angiogram documented complete debanding with 0 mmHg pressure gradient across the previous PAB (Fig. 1d). After debanding, there was no significant pulmonary regurgitation, while pulmonary to systemic flow ratio remained 1.0. At the subsequent outpatient clinic, muscular VSDs had become further smaller, while some tiny VSDs had closed.

After the catheterization, we investigated the mechanism of debanding in vitro. A cardiovascular surgeon created a model of PAB resembling that in the patient using a similar ePTFE patch and threads (4-0 Ticron^®, and 6-0 Pronova^®; Fig. 2). We dilated these bands using similar balloons to those used in the reported patient. When we dilated with a Mustang^® balloon (12 mm diameter, 2 cm length) at 14 atmospheres, the band became loose as 6-0 Pronova^® knot partially tore the ePTFE (Fig. 3a, b). After further dilation with a Conquest^® (12 mm, 2 cm length), ePTFE was torn completely at 10 atmospheres (Fig. 3c, d). Primary dilatation of a similar model with the Conquest^® balloon reproducibly debanded the ePTFE in a similar fashion.

In this patient, the PAB became unnecessary following spontaneous near closure of his muscular VSDs. Percutaneous balloon artery debanding successfully eliminated the narrowing of the main pulmonary artery.

Percutaneous balloon debanding was first reported by Bjørnstad in 1990 []. They debanded a Dacron^® band of 4 mm width attached with polypropylene thread in patients aged 6 months and 4 years. Morgan et al. reported experimental debanding of 40 bands, those circumferences ranged from 25 to 30 mm, and those bands were made of Dacron^® fixed in position with 5-0 Prolene^® (Johnson & Johnson, New Brunswick, New Jersey, United States) []. A balloon of 15 mm in diameter could deband all models at 4–7 atmospheres, while in 34 models, the mechanism of debanding was snapping of the Prolene^® knot, in 5 models, the Dacron^® band was torn by the thread, and in one model the thread knot became loose following balloon dilatation. In our actual practice, we use an ePTFE patch tied with 4-0 Ticron^® and 6-0 Pronova^® for PAB. Balloon dilatation using 12 mm Mustang^® at 14 atmospheres could not completely abolish the band narrowing, while a 12 mm Conquest^® could completely open the band at 10 atmospheres. Specific in vitro experiments have documented that tearing of the ePTFE patch by the suture thread explains the mechanism of debanding. As ePTFE is softer than Dacron^®, we supposed that ePTFE but not the thread was ruptured by balloon dilatation. The extremely non-compliant property of the Conquest^® balloon may ensure more robust stretch than the Mustang^® balloon, which has more compliant properties, once it has been dilated by nominal pressure.

Life-threatening rupture of the pulmonary artery following percutaneous debanding has been reported [] from Western counties, where the band was commonly made from a Dacron^® patch. Meanwhile there have been a few reports of staged or complete balloon debanding of bilateral PAB in hybrid approach for hypoplastic left heart syndrome from Japan where the band is always made with ePTFE. Hoshino et al. recommended the balloon diameter of 30–40% of the circumference of the band for staged dilatation of the band [], while in our experimental model, ePTFE band could be completely debanded when we used a balloon diameter larger than 50% of the circumference of the band associated with similar model as described in this study []. Dacron^® may be too solid to allow balloon debanding with a large size balloon, and attempts may be complicated by tearing of the pulmonary arterial wall unless the suture threads do not break. Vazquez-Garcia et al. reported debanding for ePTFE band using a balloon 1.2–2.0 times to the narrowest segment of pulmonary artery []. In their report, a small but a significant pressure gradient remained after balloon debanding. Judging from their description, their balloon diameter might not exceed 50% to the circumference of the band. In contrast, in our patient and in the in vitro experiments, the band made from ePTFE broke by the suture threads following balloon dilatation with the diameter of 57% to the band circumference and almost the same as the reference vessel diameter.

In conclusion, transcatheter pulmonary debanding was successful in a patient whose muscular VSD had become negligible. This may be one of the options to avoid surgery in patients whose PAB is no longer necessary. Although balloon debanding for ePTFE band is potentially safer than for Dacron^® band, further study will be mandatory to clarify optimal combination of the band and the thread for safe and effective percutaneous balloon debanding.

No potential conflicts of interest are disclosed.

We deeply appreciate Dr Peter M. Olley, Professor Emeritus of Pediatrics, University of Alberta, and Dr Setsuko Olley for kind language consultation.