P441Targeted ablation of residual luminal pulmonary vein potentials through high density mapping: preliminary results from the CHARISMA registry

Abstract Background A high incidence of pulmonary vein (PV) reconnection has been reported in patients (pts) with clinical recurrences of AF. Low-voltage activity beyond PVs (e.g. antral activity) may contribute to ablation failures in the long term. Detailed characterization of PV antra through high density mapping (HDM) and automated algorithm is still lacking. Purpose to characterize PV gaps and the low-voltage activity in tissue such as the PV antra during and after ablation of PVs in AF pts. Methods Consecutive pts undergoing AF ablation from the CHARISMA registry with complete characterization of residual PV antral activity were included. A complete map of the left atrium and PVs was performed prior and after ablation through the Rhythmia HDM system. A novel map analysis tool (Lumipoint - LM -) that automatically identifies split potentials and continuous activation was used sequentially on each PV component, in order to assess the presence of gaps (PVG) and residual potential within the antral scar (RAP, defined as any low voltage high frequency fractionated signal propagating within the antral scar without conduction into the vein) and characterize electrical propagation. After ablation we reassessed with repeat voltage and propagation maps that electrical quiescence was achieved. Ablation endpoint was PV isolation. Results Thirty-six cases of AF ablation were analyzed (11 de novo, 25 redo). A total of 36 PVG in 13 (36%) patients were detected after remap (1 case of de novo) or initial map of redo patients (12 cases). A total of 34 RAP in 20 cases (56%) were found: 4 (36%) cases of de novo (all after ablation and remap) and 16 (64%) cases of redo (all after initial map). In 7 (19%) cases we found at least one RAP in pts with complete absence of PV conduction. 100% of PVG (n = 36) and 89% of RAP (n = 29) were fully detected though a first pass automated annotation. In 5 RAPs (11%) an additional temporal consistency of low-voltage signal relative to neighboring activation was needed due to the very low voltage EGM (≤0.1 mV). PVGs were more common at right PV sites (n = 26, 72%) and anterior PV sites (n = 20, 55.6%) whereas RAPs were detected more frequently at left PV sites (n = 20, 59%) and anterior PV sites (n = 21, 62%). RAP showed a lower median voltage compared with PVG (0.22[0.2-0.3]mV for RAP vs 0.97[0.6-1.3]mV for PVG, p < 0.0001) whereas the median number of EGM peaks were higher (6.5[5-8] for RAP vs 3[2-4] for PVG, p < 0.0001). No complications during the procedures were reported. The acute procedural success was 100%, with all PVs successfully isolated and RAPs completely abolished in all study pts. Conclusion In our preliminary experience, local vulnerabilities in antral lesion sets were commonly discernible using HDM system both in de novo or redo patients when no PV conduction was present. The applied workflow seemed to be useful to quickly pinpoint and accelerate the search of local PV activity or concealed low-voltage activity.