The modified ablation index: a novel determinant of acute pulmonary vein reconnections after pulmonary vein isolation

Abstract

Background: Although pulmonary vein isolation (PVI) guided by the ablation index (AI) has been well-developed, acute PV reconnections (PVRs) still occur. This study aimed to compare the prognostic performance of the modified AI and its optimal cut-off value for the prediction of acute PVRs to ensure durable PVI. Methods: Three-dimensional left atrium (LA) voltage maps created before an extensive encircling PVI in 64 patients with atrial fibrillation (AF) (45 men, 62 ± 10 years) were examined for an association between electrogram voltage amplitude recorded from the PV–LA junction and acute post-PVI PVRs (spontaneous PVRs and/or ATP-provoked dormant PV conduction). Results: Acute PVRs were observed in 22 patients (34%) and 33 (3%) of the 1012 PV segments. Acute PVRs were significantly associated with segments with higher bipolar voltage zones (3.23 ± 1.17 vs. 1.97 ± 1.20 mV, P < 0.0001), lower mean AI values (449 [428–450] vs. 460 [437–486], P = 0.05), and radiofrequency lesion gaps ≥ 6 mm (48 vs. 32%, P = 0.04), but not with contact force, force–time integral, or power. We created the modified AI calculated as AI/LA bipolar voltage, and found it to be significantly lower in areas with acute PVRs than in those without (152 [109–185] vs. 256 [176–413] AU/mV, P < 0.0001). Univariate analysis showed the prognostic performance of the modified AI, with an area under the curve of 0.801 (0.775–0.825), to be the highest of all the significant parameters. Conclusions: Low values of the novel modified AI on the PV-encircling ablation line were strongly associated with acute PVRs.