Work

Pre-procedural three-dimensional computed tomography vs. angiographic assessment for patent ductus arteriosus stenting: a comparative analysis

Year: 2026

Type: article

Abstract: Background: Patent Ductus Arteriosus (PDA) stenting has become an established albeit complex intervention in ductal dependent congenital heart disease. Accurate pre-procedural planning is essential for successful stent placement, yet the optimal measurement methodology remains debated. This study compared the accuracy and clinical utility of routine pre-procedural three-dimensional (3D) computed tomography-derived measurements with traditional angiographic assessments for PDA stent sizing. Methods: = 12). PDA length measurements from CTA-derived 3D reconstruction and 2D angiography were compared against the actual stent length used. Statistical analysis included Pearson correlation, Bland-Altman agreement, and concordance correlation coefficients (CCC). Results: = 0.811) measurements correlated with stent length. Performance varied significantly by anatomical complexity. In Group 1 (straight), angiography demonstrated favorable accuracy with lower bias (-0.28 mm vs. 1.33 mm) and higher concordance (CCC: 0.807 vs. 0.681). In Group 2 (moderate tortuosity), angiographic performance showed greater underestimation and lower concordance (Bias -5.67 mm; CCC: 0.168), while 3D remained reliable (Bias -2.33 mm; CCC: 0.526). In Group 3 (severe tortuosity), 3D overestimated length (+3.1 mm) while angiography underestimated it (-2.88 mm), though angiographic consistency was higher (CCC: 0.704 vs. 0.559). Conclusion: Pre-procedural 3D measurements provide consistent reliability across varying degrees of anatomical complexity, offering an advantage in cases with moderate tortuosity where angiography may be vulnerable to foreshortening. Angiography remains highly accurate for straightforward anatomy. A complexity-stratified imaging approach is recommended to optimize stent selection.

Source: Frontiers in Cardiovascular Medicine

DOI: 10.3389/fcvm.2026.1767627