Po-05-089 comparison of virtual heart arrhythmia ablation targeting predictions with areas of isochronal crowding in scar-dependent ventricular tachycardia

Isochronal late activation mapping (ILAM) is a functional substrate mapping technique which detects deceleration zones (DZ) in those with scar-dependent VT. DZ have been shown to correlate with the critical isthmus during re-entrant VT. Personalised virtual heart arrhythmia ablation targeting (VAAT) is a technique which aims to non-invasively identify the optimum ablation targets through computational modelling. The association of the VAAT predictions with DZ is unknown, nor is the effect of extrastimulus pacing on identification of DZ. To assess to relationship between VAAT predicted targets with areas of slow conduction on ILAM in those with scar-dependent VT. Gadolinium-enhanced MRI scans were used to reconstruct virtual heart models. Simulated ventricular pacing was performed to demonstrate VT inducibility and predict optimum lesion sets to terminate all likely VTs in the models. ILAMs were generated by annotating the latest activation of ventricular electrograms during right ventricular and extrastimulus pacing at 20ms above VERP. DZ were defined as > 3 isochrones in a 1cm radius. VAAT predictions were superimposed on the invasive ILAM. The spatial relationship was assessed. DZ identified during RV and extrastimulus pacing were compared. 11 patients (100% male) with scar-dependent VT underwent invasive ILAM mapping and VAAT protocol. 3195 ± 935 EAM points per patient were analysed. Mean VAAT-predicted area was 11.2 ± 4.3 cm2 representing 22.3% of the total MRI infarct area. RV pacing identified 13 DZ (mean 1.2 DZ per patient, range 0-3). 10/13 (76.9%) were located within 5mm of a VAAT predicted site, whereas 3/13 (23.1%) were remote (>5mm from a predicted site). Extrastimulus pacing identified 25 DZ (mean 2.3 DZ per patient, range 1-4). 21/25 (84%) were within 5mm of a VAAT predicted site, 4/25 (16%) were remote. More DZ were identified during extrastimulus compared to RV pacing (25 vs 13, p = 0.04). The mean distance of RV pacing-identified DZ and extrastimulus-identified DZ from VAAT predictions was not significantly different (3.3 ± 6.7 mm vs 2.9 ± 5.4 mm respectively, p = 0.84). 11/13 (84.6%) of RV pacing-identified DZ were also identified by extrastimulus pacing. 14/25 (56%) of extrastimulus-identified DZ were novel, having not been detected by RV pacing. Deceleration zones visualised on ILAM are closely spatially associated with VAAT-predicted optimum targets for VT ablation. Extrastimulus pacing identifies significantly more DZ compared to RV pacing.