Conduction velocity-based functional substrate mapping during atrial fibrillation (AF) enhances identification of AF drivers compared to mapping during sinus rhythm

Abstract

Abstract Background Conduction velocity (CV) heterogeneity is known to play a role in the initiation and maintenance of atrial fibrillation (AF) [1,2]. CV has been shown to decrease with cycle lengths (CLs) <600ms in low voltage areas (LVAs) [3,4], and with CLs <300ms in non-LVAs [4]. Yet, the impact of rhythm and CL on voltage-independent, CV-based atrial substrate mapping remains uncertain. Purpose This study evaluates CV differences based on mapping protocol and explores the relationship between CV and AF pathologic conduction patterns (PCPs) identified through global, non-contact, charge density mapping (CDM) [5]. Methods CDM was performed in twelve patients with persistent AF, presenting for de novo ablation. All mapping was performed prior to any ablation. Sinus rhythm (SR) CV was assessed during atrial pacing at two CLs (380ms and 800ms). To mitigate the impact of wavefront direction, pacing was performed at three sites: mid coronary sinus, left atrial appendage, and right atrial appendage. AF CV was determined from a 20-second continuous recording of global propagation. PCPs were identified as: focal firing (FF), rotational propagation (Localised Rotational Activation [LRA]), and pivoting propagation (Localised Irregular Activity [LIA]) [5]. Patients’ left atrial geometries consisted of approximately 3000 vertices, facilitating paired comparisons of spatial data. At each vertex, values for median CV dependent on mapping protocol and frequency of PCPs were utilised for analysis. Variables are presented as mean±standard deviation, median (interquartile range), and count (percentage) as appropriate. All analyses were performed using R (version 4.1.2). CV distributions were compared using the Analysis of Variance test. Results Mean age was 64±12 years, eight (67%) were male, median time since AF diagnosis was 3.1 (2-3.5) years, and median CHA2DS2-VASc score was 2 (2-4). Median CV across our cohort was significantly lower during AF than SR with either long or short CL pacing (0.37mm/ms vs 0.66mm/ms vs 0.61mm/ms, p<0.001). Histograms of CV vs left atrial surface area demonstrated similar distributions for both CLs. However, there was poor overlap with AF, which was prominently skewed towards slower CVs (Figure 1). Slower CVs during AF were associated with an increased frequency of PCPs (Figure 2), particularly FF with CVs <0.2mm/ms, and LRA and LIA with CVs <0.4mm/ms. This association was absent during SR. Conclusions Rhythm has a significant impact on CV-based functional substrate mapping. Slower CVs observed during AF likely result from the presence of CLs shorter than 300ms, which impact both LVA and non-LVA areas. Crucially, areas of slow CV during AF were associated with the location of PCPs. Mapping performed during AF may improve identification of areas where functional remodelling has preceded structural. Such areas, characterised by CV heterogeneity, may serve as promising targets for catheter ablation.Figure 1.Figure 2.