Coronary vessel wall visualization via three-dimensional turbo spin-echo black blood imaging in Kawasaki disease

Purpose: To evaluate the feasibility of coronary vessel wall visualization using three-dimensional turbo spin-echo black blood imaging (3D-TSE) in children with Kawasaki disease. Materials and methods: Nine patients (6 girls and 3 boys; mean age ± standard deviation, 5.6 ± 3.3 years; range, 1.4–10.3 years) were included. Coronary magnetic resonance angiography (MRA) with an axial slice orientation and 3D-TSE with axial and sagittal slice orientations (3D-TSE-axi and 3D-TSE-sag) were acquired for the whole heart. Coronary vessel walls were evaluated separately in aneurysm and normal-proximal regions. The internal diameter and wall thickness of the reformatted cross-sectional images were measured in both the regions. Reproducibility between MRA and 3D-TSE was evaluated via interclass correlation coefficients (ICCs) and Bland-Altman plots. Results: In total, 164 points (aneurysmal regions, 73; normal-proximal regions, 64; normal-distal regions, 27) were evaluated. The ICC for 3D-TSE-axi was higher than that for 3D-TSE-sag (aneurysmal regions, ICC = 0.88 and 0.81; normal-proximal regions, ICC = 0.90 and 0.32, respectively). Bland-Altman plots of the internal diameter via MRA and 3D-TSE-axi showed a wide 95% limit of agreement (−0.13 to 2.89 mm) and significant fixed and proportional biases (P < 0.001 and P = 0.002) in the aneurysmal regions. However, the 95% limit of agreement was narrow (−0.14 to 0.57 mm) in the normal-proximal regions. If 1 mm was set as the cut-off for a thickened wall, wall thickness via 3D-TSE-axi was found to be abnormal across many points (84.0% of aneurysmal regions; 18.4% of normal-proximal regions). Conclusions: 3D-TSE imaging of the normal-proximal regions of the coronary vessel in individuals with Kawasaki disease was found to be feasible. However, in aneurysmal regions, larger aneurysmal diameters led to an increased bias between MRA and 3D-TSE.