Accuracy of CT-derived interarterial and intramural anomalous coronary arteries anatomy compared with 3D-reconstructed IVUS in-vivo reference.

Abstract

In anomalous aortic origin of a coronary artery (AAOCA), coronary CT angiography (CCTA) is the primary tool for anatomical assessment. However, CCTA measurements lack validation against a high-resolution in-vivo standard. To assess CCTA reliability for lumen dimensions and intramural length (IML) in AAOCA, and to validate these against a three-dimensionally reconstructed intravascular ultrasound (3D-IVUS) reference using three additional IML methodologies. A total of 60 AAOCA patients with an interarterial/intramural course underwent both CCTA and IVUS. A 3D-IVUS was constructed and its geometric reproducibility assessed. CCTA reliability and agreement with 3D-IVUS were evaluated for ostial dimensions, position-dependent minimal lumen area (MLA), and IML, measured by interluminal space thickness (ILS), elliptic ratio, and pericoronary fat sign. 3D-IVUS reference standard showed excellent geometric reproducibility (median point-wise distance 0.19 mm, 95%CI 0.07-0.43 mm) and good-to-excellent intrareader reliability (ICC 0.72-0.96). CCTA intra- and inter-reader reliability was moderate-to-excellent for ostial parameters and ILS-derived IML (ICC 0.69-0.92) and poor-to-good for MLA (ICC 0.28-0.87). Against 3D-IVUS, CCTA ostial area demonstrated moderate correlation (r = 0.73, p < 0.001) with minimal bias, and ILS-derived IML showed moderate agreement (r = 0.59; bias + 0.23 mm), while the fat sign substantially overestimated length (+ 7.88 mm). Using 3D-IVUS as an in-vivo reference standard for the precise assessment of coronary anatomy in AAOCA, CCTA provided reproducible ostial measurements, and the ILS-based method showed the best performance among the evaluated CCTA approaches for estimating intramural length. However, wide limits of agreement preclude precise individual sizing, and position-dependent MLA measurements remain unreliable.