Detection of coronary artery aneurysm in children with Kawasaki disease: a prospective study using contrast-enhanced coronary MRA with combined diastolic and systolic phases imaging at 3T.

Abstract

Background: Contrast-enhanced coronary magnetic resonance angiography (MRA) images can be obtained during two comparatively quiescent phases in the cardiac cycle (mid-diastole and end-systole) to evaluate coronary artery aneurysms (CAA) in children with Kawasaki disease (KD), but each phase of coronary MRA has its own advantages and disadvantages. Integrating the diastolic and systolic phases images may improve the clinical utility of coronary MRA.

Methods: 101 participants with KD were prospectively enrolled. Coronary MRA images were acquired twice in the diastole and systole acquired within the late gadolinium enhancement (LGE) delay window (10-15min), eliminating the need for extended scan time. Scan-relevant parameters and image quality were assessed. The diagnostic efficacy of CAA was compared among three groups (systole, diastole and dual-phase). The detection performance of CAA with thrombosis was also calculated. Myocardial injury is detected on LGE images. The association between myocardial injury and baseline or CAA with thrombosis were identified with logistic regression analyses.

Results: 101 participants (aged 7.4 ± 3.4, 35.6% female) were evaluated. The mean MRA acquisition times were 241.5 ± 51.9s and 251 ± 79s in systole and diastole, and the average total scan time of two-phase coronary MRA image was 492.6 ± 112.4s, which are integrated into the LGE delay period. The subjective overall image quality in systole were better than that during the diastole. When comparing with the diastolic-phase MRA on a per-segment basis, dual-phase coronary MRA showed greater sensitivity (85.92% [244/284] vs. 79.58% [226/284], p < 0.001) and accuracy (91.97% [828/909] vs. 90.54% [823/909], p < 0.001) in detecting CAA, also the area under the receiver operating characteristic curve (AUC-ROC) of dual-phase MRA was improved (0.903 vs.0.875, p < 0.001). Consistent improvements were observed on per-vessel basis (sensitivity: 88.55% [201/227] vs. 84.58% [192/227]; accuracy: 87.62% [354/404] vs. 86.14% [348/404]; both p < 0.05). Dual-phase coronary MRA also demonstrated excellent diagnostic performance in detecting CAA with thrombosis, achieving an AUC-ROC of 0.863 in per-patient analysis. Intracoronary aneurysmal thrombus was independently associated with myocardial injury (OR: 4.5, [95% CI: 1.8-11.1], p < 0.001; adjusted OR: 3.7, [95% CI: 1.4-9.7], p < 0.05).

Conclusion: Dual-phase 3T contrast-enhanced coronary MRA has satisfactory diagnostic performance in detecting CAA. It is a feasible option for the clinical application of CMRA in patients with Kawasaki disease.