Ductus Arteriosus Morphology in Pulmonary Atresia With Ventricular Septal Defect: A Retrospective Case Series Comparing Echocardiography and Digital Subtraction Angiography.

objectiveThis study aims to describe ductus arteriosus (DA) morphology in patients with pulmonary atresia with ventricular septal defect (PA-VSD) and to assess exploratory paired differences between echocardiography and digital subtraction angiography (DSA). This study was primarily descriptive and was not designed to establish diagnostic agreement, equivalence, or interchangeability between the two imaging modalities. MATERIALS AND

methodsThis retrospective descriptive case series was conducted at Children's Hospital 1, Ho Chi Minh City, Vietnam, from April 2024 to April 2025. Consecutive patients aged from birth to under 18 years with confirmed PA-VSD, a patent DA, and available echocardiography and DSA images adequately showing DA morphology were included. PA-VSD was confirmed by echocardiographic evidence of pulmonary atresia, a ventricular septal defect, and absence of antegrade flow from the right ventricle to the pulmonary arteries, with further verification by DSA and catheterization records. Patients with incomplete records, inadequate images, or cardiovascular anomalies substantially altering ductal anatomy were excluded. DA origin, orientation, number, and diameters at the aortic and pulmonary ends were retrospectively remeasured from archived images and independently assessed for each modality. Paired statistical tests were used to evaluate modality-related differences, and exact concordance was described for categorical findings.

resultsThirty patients were included. The median age at admission was 92 days, and the median body weight was 3.6 kg. There were 20 female patients (66.7%) and 10 male patients (33.3%). PGE1 infusion was administered in 22 patients (73.3%). All patients had a single DA on both modalities. On echocardiography, DA origin was classified as the descending aorta in 16 patients (53.3%), the brachiocephalic artery in 11 patients (36.7%), and the subclavian artery in three patients (10.0%). On DSA, the DA origin was classified as the descending aorta in 15 patients (50.0%), the brachiocephalic artery in 12 patients (40.0%), and the subclavian artery in three patients (10.0%). The distribution of DA origin did not differ significantly between modalities (McNemar-Bowker test, χ² = 1.00, p = 0.608), but this result was not interpreted as evidence of agreement. Vertical orientation was observed in 17 patients (56.7%) on both modalities, with exact concordance for DA orientation and DA number. The mean aortic-end diameter was significantly larger on DSA than on echocardiography (3.86 ± 1.43 mm vs. 3.12 ± 0.97 mm; paired-samples t-test, t = 3.18, p = 0.001). The mean pulmonary-end diameter did not differ significantly between DSA and echocardiography (1.98 ± 0.83 mm vs. 2.18 ± 0.62 mm; t = 1.32, p = 0.198).

conclusionsDA morphology in PA-VSD was variable, with descending aortic origin and vertical orientation being common. Echocardiography and DSA showed similar categorical distributions for DA origin, orientation, and number, but these findings do not establish diagnostic agreement or interchangeability. DSA demonstrated larger aortic-end measurements, although the clinical impact of this difference on stent sizing, procedural success, or complications was not directly evaluated. Future studies should assess agreement, reproducibility, and procedural outcomes using standardized imaging protocols.