Arnav Amit
In dilated cardiomyopathy (DCM), sudden cardiac death from ventricular arrhythmia occurs unpredictably among patients with similar ejection fraction (EF), and EF performs poorly as an individual risk predictor. EF is a cavity-volume ratio and is, by construction, blind to the spatial distribution of wall thinning. I hypothesized that the pattern of remodeling, specifically asymmetric septal versus lateral wall thinning, produces spatially heterogeneous diastolic wall stress and stretch, which through stretch-activated channel (SAC) activation creates an arrhythmogenic substrate that EF cannot encode. I developed a three-component in silico model coupling a parametric prolate-ellipsoid left ventricular geometry, regional biaxial (Laplace) wall stress, and the O'Hara-Rudy 2011 human ventricular action potential model with an ohmic SAC current engaged during diastole. Across five severity by five asymmetry levels, ejection fraction was computed and depended on severity alone, not on asymmetry. At matched EF, asymmetric remodeling produced a regional gradient of end-diastolic stretch that uniform dilation did not: at severe asymmetric DCM the thinned septum was depolarized by 2.25 mV relative to baseline, with 1.32 mV of regional resting-potential dispersion and 2.19 ms of action potential duration (APD) dispersion, against essentially none for uniform dilation at the same EF. The substrate rose approximately linearly with the septal/lateral wall-stress gradient, which equals the lateral-to-septal wall-thickness ratio measurable on a standard echocardiogram (slope 1.81 mV per unit gradient), and saturated at the sarcomere stretch ceiling. This in silico study quantifies a known mechanism and converts it into a specific, falsifiable prediction: an echo-derived wall-thickness ratio indexes an arrhythmic substrate orthogonal to EF, motivating retrospective and prospective clinical validation.