Purpose or Case Report: DMD is X-linked cardiac and skeletal myopathy caused by dystrophin protein deficiency with early death by cardiomyopathy. Cardiac magnetic resonance imaging (CMR) with myocardial fibrosis by late gadolinium enhancement (LGE) precedes decline in ejection fraction (LVEF) and may alter management in DMD. Muscle weakness limits ability to perform consistent breath-hold (BH) in traditional LGE sequences. Free-breathing (FB) LGE with motion correction (MOCO) allows improved visualization. We hypothesize that FB LGE with MOCO provides equivalent diagnostic quality with similar scan time to traditional LGE sequence.
Methods & Materials: IRB approved retrospective analysis of 28 patients with DMD undergoing clinical CMR from April to October 2019. All subjects scanned on 3T Seimens Skyra with 1) traditional BH, phase sensitive inversion recovery, spoiled gradient echo (BH-SGRE) and 2) FB-MOCO, PSIR, single shot, balanced SSFP sequences in short-axis plane (FB-MOCO), 10-15 minutes after 0.1 mL/kg Gadavist injection for LGE. BH-SGRE slices were acquired over 10-12 heart beats (~50 ms/beat), while FB-MOCO slices were acquired in one heartbeat (~260 ms). Both sequences had spatial resolution of 1.1x1.1x8 mm3. 139 slices were compared for image quality in blinded fashion by two independent readers using Likert scale scoring system of 1-5 (1=non-diagnostic, 5=excellent). Presence of artifacts and DME was similarly assessed.
Results: Mean±SD age of sample was 13.2±2.8 years, BSA 1.4±0.4 m2, heart rate of 90±15 bpm, and LVEF of 55±5 %. Mean±SD scan duration was 1.4±0.4 min for BH-SGRE and 1.6±0.6 min for FB-MOCO. Image quality score was comparable with median [IQR] BH-SGRE: 3.5 [3.4 - 3.6] vs. FB-MOCO: 3.6 [3.5 -3.8]. Both techniques identified fibrosis equally in 13/28 patients (BH-SGRE:88 segments vs FB- MOCO:91 segments of LGE). Blurring of myocardium seen in 37% of BH-SGRE and 11% of FB- MOCO. Pericardial artifacts related to magnetic inhomogeneity seen in 4% of BH-SGRE and 36% of FB- MOCO. Ghosting artifacts seen in 7% of BH-SGRE and 2% of FB- MOCO.
Conclusions: The FB-MOCO sequence provides similar diagnostic quality and acquisition duration compared to BH-SGRE LGE in patients with DMD. There were no significant diagnostic discrepancies between techniques. There is decreased motion blur/ghosting artifact, but increased magnetic inhomogeneity artifact in the FB-MOCO technique. This technique may be suitable in patients with difficulty breath holding and requiring LGE assessment.