Identification of the mapping-based voltage-derived parameter for accurate arrhythmia outcome prediction in patients undergoing first PVI
T Huang, C R Rodriguez, A G Gamer, S H Hess, F C Cuculi, M E Eichenlaub, H L Lehrmann, A L Loewe, D T Trenk, D W Westermann, T A Arentz, A J JadidiAbstract
Background
Voltage reduction in left atrium, as reflected by LA global mean voltage (GMV) and low voltage area (LVA), have been previously reported as key electrophysiological characteristics differentiating paroxysmal from persistent atrial fibrillation (AF) type.
Purpose
The current study aims to identify the LA voltage marker with the highest predictive value with regard to arrhythmia recurrence in AF patients undergoing de-novo circumferential pulmonary vein isolation (CPVI).
Methods
278 patients (36% female, 55% persistent AF) undergoing de-novo CPVI were enrolled in the study. High-density electro-anatomical mapping was performed strictly under sinus rhythm (SR) to determine the following parameters GMV, LVA, left atrial volume (LAV). All three parameters were extracted and quantified using an AI-algorithm developed by an Institute of Technology. Parameters were determined after exclusion of PV from the map and LVA was quantified using bipolar threshold of 1.0mV based on previous publications. Arrhythmia recurrence included occurrence of AF, atrial flutter and atrial tachycardia at 6-, 12- and 24-month FU respectively.
Results
Arrhythmia recurrence occurred in 48 (17.3%) patients within 12-month FU and in 83 patients (29.9%) within 24-month FU. Multivariate COX regression identified persistent AF and LVA at 1.0mV threshold (continuous variable) as the only two significant predictors for arrhythmia recurrence at both 12-month FU (Persistent AF: HR=2.11, p=0.027; LVA: HR=1.02, p=0.034) and 24-month FU (Persistent AF: HR=1.84, p=0.013; LVA: HR=1.02, p=0.003). C-statistics identified the cut-off thresholds of LVA of >1.5cm2 and >3.0cm2 for paroxysmal and persistent AF cohorts, respectively. As a result, LVA >1.5 cm in paroxysmal cohort was associated with significant difference in arrhythmia freedom rate at 12-month (97.8% vs. 87.6%, p=0.015) and marginally significant difference at 24-month (82.7% vs. 67.6%, p=0.051). LVA >3.0cm2 in persistent cohort was associated with significant difference in arrhythmia freedom rate at both 12-month (89.2% vs. 68.5%, p=0.008) and 24-month (72.5% vs. 43.2%, p=0.002).
Conclusion
Persistent AF and LVA at 1.0mV bipolar threshold in sinus map are highly predictive of arrhythmia recurrence after CPVI, however, more robust predictive performance is observed in persistent cohorts.comparison of selected varibleKM curves of recurrence predictor