Relation of surface T-wave to vulnerability to ventricular fibrillation in explanted structurally normal hearts

Introduction: Body surface characterization of repolarization gradient substrates for ventricular fibrillation (VF) is still poor. We investigated if VF propensity can be assessed from surface T-wave in an ex-vivo model of purely electrical repolarization dispersion. Methods: To create repolarization gradients, dofetilide and/or pinacidil were separately infused in N=7 isolated pig hearts in a human-shaped torso tank. At each drug state, VF inducibility was quantified by the vulnerability window (VW), i.e. the interval during which VF was triggered by S1 (atrium) S2 (ventricle) stimulation. T-wave duration (the peak-to-end interval, T-PEAK-T-END), and shape, i.e., symmetry (the ratio of the areas under the peak-to-end and onset-to-peak frames, Asy), and flatness (kurtosis, Kurt) were computed from 256 tank potentials. We fitted a linear mixed-effect (LMM) model to link T-wave markers (fixed effects) with VW (response), with random effects on drug states and hearts. Results: VF was induced in 14/23 drug states from at least one ventricle. In vulnerable substrates (higher VW), T-waves were longer (T-PEAK-T-END, p=0.0004), less symmetric (Asy, p