Dynamic voltage mapping of the ventricular tachycardia substrate

Abstract

Abstract Background Substrate mapping techniques that predict ventricular tachycardia (VT) isthmus components in sinus or paced rhythm are necessary as mapping during VT is often not possible. Voltage maps lack accuracy in differentiating scar, traditionally displayed as tissue <0.50mV, from bordering arrhythmogenic substrate. We developed a practical technique aiming to improve the utility of a voltage map during VT ablation, termed ‘Dynamic Voltage Mapping’ (DVM). Methods Consecutive patients with post-infarct VT at our centre underwent left ventricle (LV) voltage mapping (mean 4762 points) with Carto3 v.7® and a PentaRay®, or Ensite X® and a HD Grid mapping catheter. Both mapping systems offer algorithms that allow substrate activation maps to be superimposed and analysed on the surface of a voltage display (i.e. Carto Ripple Maps, or Ensite X activation vectors). DVM is a process that sequentially adjusts the voltage display to highlight scar tissue in areas devoid of Ripple bars, or in areas where activation vectors point in multiple unfathomable directions (‘vector disarray’). This approach is summarised in Figures 1 & 2 and was retrospectively performed in all cases. This alternative scar voltage limit was termed the ‘DVM scar threshold’ (mV). Tissue below this threshold was termed ‘DVM-scar’, and tissue immediately surrounding (up to 0.5mV) was termed "DVM-Borderzone (BZ)". Results Twenty-five patients (mean age 64.4±11.0 years, 88% male, LV ejection fraction 31.8±6.9%, 16.4±8.5 years since infarct) were studied (n=15 Carto; n=10 EnSite). Across all cases, the mean shell area with voltage <0.50 mV was 36.8±20.3 cm2. The mean DVM scar threshold was 0.24±0.07 mV (Carto: 0.22±0.07 mV [range 0.12–0.35 mV]; EnSite: 0.26±0.08 [range 0.18–0.50 mV]). The mean DVM-scar area was 18.4±14.3 cm2, significantly smaller than the traditional 0.50mV value (p=0.0002). In the 8 patients where stable VT was mapped, the VT isthmus sites co-located within the DVM-BZ in all cases, exemplified in the figures. Pace-mapping at the DVM-BZ matched the remaining unmappable VT morphologies. Ablation lesions were retrospectively superimposed onto the DVM map and generally co-located within the DVM-BZ. No ablation was delivered in tissues >0.5mV. At median follow-up of 8.8 months, VT burden significantly reduced (median 3 episodes in 3-months pre vs. 0 post; p<0.001), with 76% of patients VT-free. Conclusion DVM appears to improve the delineation of the post-infarct VT substrate, and is feasible using two commercially-available mapping systems. The mean voltage threshold that differentiated DVM-Scar from BZ tissue was significantly lower than the traditional 0.50 mV scar definition, at 0.24mV, though ranged widely, suggesting that scar thresholds require individualised substrate definition. Targeted ablation of DVM-BZ may be effective at reducing future VT, though requires prospective validation.Figure 1Figure 2