Digital twin reconstruction of ventricular repolarisation identifies regional causes of T-wave abnormalities in hypertrophic cardiomyopathy

Background: Abnormal ventricular repolarisation in hypertrophic cardiomyopathy (HCM) may predispose patients to lethal arrhythmias, but repolarisation in HCM remains poorly spatially characterised. Integrating spatial cardiac magnetic resonance with ECG data has the potential to map electrical function throughout the ventricles. This study applied novel digital twin inverse ECG methods to map repolarisation patterns underlying abnormal T-waves in HCM. Methods: Patient-specific full ventricular electrophysiological models were iteratively refined to match the patient 12-lead ECG. Data from 32 healthy volunteers and 69 HCM patients were analysed, with inferred substrates incorporating activation times, repolarisation times, and rate-corrected action potential durations (APDcs). Patients were stratified by T-wave phenotype to identify distinct spatial repolarisation signatures associated with different ECG presentations. Results: Clinical 12-lead ECGs were accurately reproduced by the inferred ventricular models in 95 of 101 cases. Healthy volunteers (N=30) and HCM patients with normal T-waves (N=33) were characterised by apex-to-base APDc gradients of 60 ms (40?80) and 60 ms (30?80), respectively. HCM patients with V1-V3 T-wave abnormalities (N=6) had attenuated apex-to-base APDc gradients of 30 ms (-20?40) driven by apical-to-mid anterior APDc prolongation, greatest at the apical segment (?APDc vs. healthy: 54 ms; 95% CI: 23?84 ms). HCM patients with V4-V6 T wave abnormalities (N=21) had reversed apex-to-base APDc gradients of -20 ms (-40?0) driven by apical-to-mid APDc prolongation, most severe at the apical segment (?APDc vs. healthy: 94 ms; 95% CI: 74?120 ms). Despite significant APDc prolongation, only 4 of 69 HCM patients had QTc > 480 ms, due to masking by the intrinsically healthy longer APDcs at the ventricular base. Conclusions: Distinct spatial distributions of APDc prolongation, not necessarily mirroring the distribution of hypertrophy, underlie different ECG repolarisation phenotypes in HCM and may be missed by the QTc interval.