Insights from Temperature Feedback and Current in Ablation Index-Guided Pulmonary Vein Isolation

Objectives: To gain further knowledge of in-vivo lesion assessment and the complex regional interplay of lesion parameters during pulmonary vein isolation (PVI). Background: Although ablation index (AI) results in favourable outcomes of CLOSE protocol-guided PVI but its' relationship with other lesion parameters is not fully understood. Methods: 62 patients undergoing CLOSE-PVI (35-50W, {less than or equal to}6mm, AI 400-550) using a QDOT catheter were prospectively included. Lesion-specific data (impedance change [∆-IMP], current [mA], delivered energy [J], power [W], contact force [CF], thermocouple temperature [Temp] and micro-electrogram amplitude [ME-EGM]) were collected. Results: Although AI targets were reached for all 3166 lesions, 479 (15%) had very low ∆-IMP (<5%) and 851 (27%) had low mean current (<550 mA). The strongest predictors of a very low ∆-IMP were slower rise (OR 0.82) and max Temp (OR 1.44). A complex regional interaction was seen between parameters [right posterior: most sub-optimal values and right anterior: most optimal values]. ME-EGM disappeared rapidly after RF initiation, despite very low ∆-IMP [anterior: 3.4 Ω (IQR:2.1) vs. posterior: 2.9 Ω (IQR:2.0), p{less than or equal to}0.001]. AI correlated well with LSI (r=0.7), but LSI was strongly CF-dependent (r=0.8) and AI was strongly W-dependent (r=0.8). Mean current correlated weakly with max Temp (r= 0.3). Higher max Temp was seen in shorter applications and in higher mean current (both r=0.3). Conclusion: AI-guided PVI is associated with 15% of very low ∆-IMP, best predicted by insufficient thermocouple heating. The potential additional roles for lesion evaluation of time to AI, thermocouple heating and tip-current density should be investigated.