Abstract 11387: Four-Dimensional Spatiotemporal Characterization of Cardiomyopathy in Duchenne Muscular Dystrophy Using Cardiac Magnetic Resonance Imaging

Abstract

Introduction: Cardiomyopathy (CM) is the most common cause of mortality in Duchenne muscular dystrophy (DMD). Because CM progression is variable, there is a critical need for biomarkers to detect early onset or rapid progression. Our objective was to evaluate localized kinematic parameters that correlate with function utilizing spatiotemporal analysis of 4D (3D plus time) cardiac magnetic resonance (CMR). We hypothesized that novel regional strain metrics would correlate with left ventricular ejection fraction (LVEF), providing insight for future DMD CM characterization studies . Methods: Sequential short axis cine CMR images of DMD patients (n=10) with a range of CM severity were compiled into 4D sequences. Epi- and endocardial borders were then segmented across one cardiac cycle for kinematic analysis. We evaluated Green-Lagrange circumferential strain (ε cc ) at 60 equally spaced short axis slices and across 60 time points to generate a spatiotemporal map. We also evaluated a novel “hybrid strain index” combining basal ε cc with posterior free-wall longitudinal strain. Spearman’s rho was used to determine statistical correlation. Results: DMD subjects had a median age of 11 years (8-24); 3 had LVEF<55% and 7 had late gadolinium enhancement. Regional peak averages of ε cc decreased with LVEF for basal ( rho =-0.94, p <.001) and mid-LV ( rho =-0.84, p =.004) but not at the apex ( rho =-0.55, p =.106). The hybrid strain index also decreased with LVEF ( rho =0.89, p <.001). Qualitative spatiotemporal ε cc maps of mild, moderate, and severe DMD cases demonstrate stark differences in regional wall motion (Figure 1). Conclusion: 4D spatiotemporal analysis of CMR images in DMD patients allows for robust comparison of regional kinematic parameters that correlate with LVEF. Figure 1: 4D CMR spatiotemporal analysis of DMD cardiomyopathy. A,B) Schematic of circumferential strain (ε cc ). C,D,E) Example mild, moderate, and severe patient spatiotemporal ε cc maps.