Introduction: Patients who present with non-ischemic dilated cardiomyopathy (NIDCM) and midwall fibro- sis, visible as areas of late gadolinium enhancement (LGE) on magnetic resonance imaging, are at high risk of arrhythmic sudden cardiac death. Reentry is the most commonly reported arrhythmic driver in NIDCM, but the role of the fibrosis underlying LGE is poorly understood. We examined the effects of three important fibrosis characteristics (texture, density and conductivity) on the incidence of reentry formation and mechanism in 600 computational models of cardiac electrophysiology in NIDCM.
Methods: 2D-finite element models were generated from a single short axis LGE MRI image of a patient with NIDCM and a septal enhancement pattern. Random fibrosis micro-structures were included in the models, corresponding to areas of LGE, and were guided by image intensity values. Ten levels of overall fibrosis density (0.1 – 1.0), were considered, along with three fibrosis textures, (replacement (R), and interstitial with high (IHA ) and low (ILA) anisotropy) and two levels of conductivity (healthy and reduced). Ten random micro-structures were generated for each density, texture, and conductivity combination. Simulated programmed electrical stimulation was applied to test for reentry inducibility.
Results: Reentries were inducible above the density thresholds (0.8, 0.8, 0.6 for IHA, ILA, R) with healthy conductivity. Reduced conductivity lowered the density thresholds to (0.3, 0.3, 0.4 for IHA, ILA, R) and led to greater reentry incidence (211 vs 71 for normal conductivity). Rotor, micro-reentry, macro-reentry and figure-eight mechanisms were observed. Rotors were more common with lower scar density and interstitial fibrosis, whereas micro and macro reentries occurred mostly with replacement fibrosis and higher scar density.