Automated lesion annotation during pulmonary vein isolation: influence on acute isolation rates and lesion characteristics.

PURPOSE: Recovery of pulmonary vein (PV) conduction is a common mechanism of atrial fibrillation recurrence after PV isolation (PVI), underscoring the need for durable lesion formation. We aimed to evaluate the utility of an automated lesion annotation algorithm (ALAA) on acute isolation rates and resulting lesion characteristics.
METHODS: Fifty patients underwent PVI using a contact force (CF) sensing catheter and ALAA. Single antral circles around ipsilateral PVs were performed with ALAA-1 settings including catheter stability (range of motion ≤2 mm, duration >10 s). Target CF was 10-20 g but not part of ALAA-1 settings. If PV conduction persisted after circle completion, force over time was added to automated settings (ALAA-2). Emerging gaps were subsequently ablated, followed by re-assessment for PVI.
RESULTS: ALAA-1 isolated 70 % of the left and 78 % of the right PVs using 756.3 ± 212.3 s (left) and 737.1 ± 145.9 s (right) of energy delivery. ALAA-2 settings identified 29 gaps in previously unisolated PVs, closure significantly increased isolation rates to 88 % of the left and 96 % of the right PVs with additional 325.4 ± 354.1 s (left) and 266.8 ± 279.5 s (right) of energy delivery (p = 0.001). Lesion characteristics significantly differed between ALAA-1 (n = 3521 lesions) and ALAA-2 (n = 3037 lesions) settings, and between isolated and non-isolated PV segments, particularly with respect to CF. Interlesion distances with ALAA-2 were significantly longer in the left superior, left superior-anterior, and right superior-posterior segments when compared to ALAA-1.
CONCLUSIONS: Settings of an ALAA affect lesion characteristics reveal areas of insufficient lesion formation and influence acute effectiveness of PVI. Combination of CF and stability shows superior performance over stability alone.