P1389Periodicity and Spatial Stability of Complex Propagation Patterns in Atrial Fibrillation

Abstract Introduction Non-contact charge density mapping identifies complex wavefront propagation including localised rotational activation (LRA), localised irregular activation (LIA) and focal firing (FF). However, the duration of mapping required to reveal underlying patterns and their temporal stability is unknown. Purpose We sought to evaluate the variability in propagation patterns over increasing durations of recordings up to 30 seconds and examine the stability of these patterns between 2 separate maps with the aim of identifying the minimum duration required to reveal underlying patterns and how they represent the stable arrhythmia substrate. Methods Patients undergoing first time AcQMap guided catheter ablation were studied. 30s recordings of left atrial propagation were analysed. LIA, LRA, and FF were quantified for frequency, percentage time present and percentage surface area affected (for FF only frequency was assessed) at increasing durations up to 30s in 1s increments. At each incremental recording duration the percentage change in each variable was calculated. For occurrence frequency the results for every possible combination of maps of increasing duration within the 30s recording were compared whilst for occurrence time and surface area a 5s moving average at 1s increments was calculated. The point at which variability was seen to plateau represents the minimum optimal mapping duration. Spatial stability was assessed by correlating the frequency of patterns at each vertex of the anatomy over 2 separate 30s recordings. Stability of regions with the most repetitive patterns were compared using Cohen’s kappa statistic. Results 15 patients were analysed (age 63 ± 9, 10 male, BMI 30 ± 5, CHA2DS2Vasc 1 ± 1.3, ejection fraction 54 ± 12%, left atrial diameter 46 ± 7mm, paroxysmal 1, persistent 14) with 11 included in the spatial stability analysis due to availability of recordings of sufficient duration. LRA demonstrated most variability followed by LIA and FF. Variability in LIA, LRA and FF decrease at increasing durations. LIA and FF variability plateau by 13 and 17s respectively. LRA plateaus at 23s. Variability of <10% is reached in all parameters at 18s. LIA demonstrated the greatest stability with average R2 of 0.76 ± 0.14 (figure). Average R2 for LRA and FF were 0.45 ± 0.16 and 0.47 ± 0.12.  Low frequency focal firings were widely distributed across the atrial surface. For FF occurring at a frequency ≥10 over the 30s, average R2 value was 0.65 ± 0.14. Cohen kappa statistic was 0.70 for LIA and 0.45 for LRA. Conclusion Mapping durations of ≥23s are required to identify all temporally variable propagation patterns although shorter durations will identify less variable LIA and FF.  LIA demonstrates high spatiotemporal stability and may best reflect disrupted conduction caused by the underlying atrial substrate and tissue architecture. Regions of high frequency FF are temporally stable and may represent important targets for ablation. Abstract Figure 1