Fibrosis Entropy Is Associated With Life‐Threatening Arrhythmia in Nonischemic Cardiomyopathy

Greater precision is required for arrhythmic risk stratification of patients with nonischemic cardiomyopathy (NICM). We sought to evaluate whether fibrosis entropy, a measure of scar texture heterogeneity derived from late gadolinium enhancement cardiovascular magnetic resonance, has incremental utility to fibrosis presence for arrhythmic risk prediction in NICM. In this prospective observational cohort study, fibrosis entropy was calculated for patients with NICM and fibrosis (late gadolinium enhancement positive, LGE+), including regions of core fibrosis, gray zone fibrosis and combined core and gray zone fibrosis. Patients with NICM and no fibrosis (LGE-) were included as a comparator group. Adjudicated follow-up for life-threatening arrhythmia included sudden cardiac death, aborted sudden cardiac death, or sustained ventricular tachycardia. Of 291 patients with LGE+ NICM, 38 (13.1%) experienced life-threatening arrhythmia over a median follow-up of 6.3 years. Core fibrosis entropy (per-SD hazard ratio [HR], 1.77 [95% CI, 1.25-2.52]; P=0.001), gray zone fibrosis entropy (HR, 1.97 [95% CI, 1.20-2.54]; P=0.004), and combined fibrosis entropy (HR, 1.98 [95% CI, 1.30-3.02]; P=0.004) were each associated with life-threatening arrhythmia after adjustment for variables used to determine implantable cardioverter-defibrillator candidacy in clinical practice (left ventricular ejection fraction ≤35% and New York Heart Association class >1) and remained associated after accounting for core and gray zone fibrosis mass. Left ventricular ejection fraction ≤35% was not associated with life-threatening arrhythmia (HR, 1.45 [95% CI, 0.77-2.74]; P=0.250). Integration of fibrosis presence with fibrosis entropy classified patients into low-, intermediate-, and high-arrhythmic-risk groups. Deeper phenotypic characterization of scar using fibrosis entropy offers incremental utility to left ventricular ejection fraction and fibrosis presence for arrhythmic risk stratification in NICM.