Ablation of Reentry-Vulnerable Zones Determined by Left Ventricular Activation From Multiple Directions: A Novel Approach for Ventricular Tachycardia Ablation

Abstract

Background: The optimal method to identify the arrhythmogenic substrate of scar-related ventricular tachycardia (VT) is unknown. Sites of activation slowing during sinus rhythm (SR) often colocalize with the VT circuit. However, the utility and limitations of such approach for guiding ablation are unknown. Methods: We conducted a multicenter study in patients with infarct-related VT. The left ventricular (LV) was mapped during activation from 3 directions: SR (or atrial pacing), right ventricular, and LV pacing at 600 ms. Ablation was applied selectively to the cumulative area of slow activation, defined as the sum of all regions with activation times of ≥40 ms per 10 mm. Hemodynamically tolerated VTs were mapped with activation or entrainment. The primary outcome was a composite of appropriate implanted cardioverter-defibrillator therapies and cardiovascular death. Results: In 85 patients, the LV was mapped during activation from 2.4±0.6 directions. The direction of LV activation influenced the location and magnitude of activation slowing. The spatial overlap of activation slowing between SR and right ventricular pacing was 84.2±7.1%, between SR and LV pacing was 61.4±8.8%, and between right ventricular and LV pacing was 71.3±9.6% ( P <0.05 between all comparisons). Mapping during SR identified only 66.2±8.2% of the entire area of activation slowing and 58% critical isthmus sites. Activation from other directions by right ventricular and LV stimulation unmasked an additional 33% of slowly conducting zones and 25% critical isthmus sites. The area of maximal activation slowing often corresponded to the site where the wavefront first interacted with the infarct. During a follow-up period of 3.6 years, the primary end point occurred in 14 out of 85 (16.5%) patients. Conclusions: The spatial distribution of activation slowing is dependent on the direction of LV activation with the area of maximal slowing corresponding to the site where the wavefront first interacts with the infarct. This data may have implications for VT substrate mapping strategies.