1037Self-navigated free-breathing isotropic 3D whole heart MRI for the characterization of complex congenital heart malformations: first experience with 52 patients

Le résumé fourni par la source

Introduction: Cardiac anatomy of patients with congenital heart disease (CHD) is often complex and accurate evaluation mandates MR image acquisition in several non-standard planes requiring multiple breath-holds. The utility of self-navigated isotropic 3D-free-breathing whole-heart MRI (SN-3D) was assessed for the visualization of the heart, coronary arteries (CA) and great vessels (GV) in CHD. Methods: Data acquisition was performed during free breathing on a 1.5T-MRI scanner with a 3D-radial trajectory modified for respiratory self-navigation after gadobutrol injection; it was ECG-triggered, with a T2-prep pulse and fat-saturated SSFP readout. The isotropic image data (resolution 1.1x1.1x1.1mm3) were reformatted offline; image quality was graded and GV diameters were measured at several levels in all patients. 3D MR-angiography (MRA) was performed in a subset of patients for comparison. Results: 52 patients (64% male, age 24 ± 11y) were included. Heart rate was 71 ± 13 bpm, scan duration 9.7 ± 4.2min. Image quality was good in 27, moderate in 15, mediocre in 7, and poor in 3. 3D datasets allowed the assessment of the arrangement of heart chambers and great vessels, the venous return anatomy after atrial switch for transposition of the great arteries (TGA), the morphology of cavo-pulmonary connections after Fontan operation, the morphology of the pulmonary arteries in tetralogy of Fallot or morphology of the great vessels after arterial switch for TGA (Fig1). GV diameters with SN-3D were accurate using conventional MRA as the gold standard (n = 37, r2 = 0.94, bias 0.1 ± 1.7mm), with a low intra- and inter-observer variability (3.7% and 5.4%) (Fig 2). Moreover, the proximal 3 cm of the CA could be well visualized and its 3D course followed in most of the patients with moderate or good image quality (n = 42: LAD 98%, LCx 74%, RCA 91%), making SN-3D useful to detect anomalous CA.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.