Atrial Fibrillation Drivers: Redefining the Electrophysiological Substrate

BackgroundWe performed high-density mapping of persistent atrial fibrillation (AF) in animals and patients (1) to test that AF is due to [&ge;]1 reentries, and (2) to characterize activation delay and reentries pre/ post pulmonary vein isolation (PVI). We determined electrophysiological characteristics that may predispose to the induction, maintenance, and reduction of AF. Methods and ResultsThis study includes 48 dogs and nine patients. 43 AF- and five sinus/ paced rhythm dogs (3-14 weeks rapid atrial pacing) were studied at open chest surgery with 117 epicardial electrograms (EGMs) (2.5mm dist.) in 6 bi-atrial regions. Rotational activity automatically detected with a new algorithm tracking the earliest and latest activation in all regions (5{+/-}2 per region) were stable over 424{+/-}505ms [120- 4940ms]. Reentry stability was highest in the right atrial appendage (RAA) (405{+/-}219ms) and the posterior left atrium (PLA) (267{+/-}115ms) and anchored between >=3 zones of activation delay (15{+/-}5ms, median 13ms) defined as >10ms per 2.5mm. Cycle length (CL) and degree of focal fibrosis were highest in the PLA and left atrial free wall (LAFW) with 94{+/-}7ms, 96{+/-}5ms, and 49{+/-}14%, 47{+/-}19%. Fiber crossing density correlated with the stability of rotational activity (R=0.6, P<0.05). Activation delay was 2x higher in AF compared to sinus rhythm/paced rhythm (interval 200-500ms). Activation delay zones > 10ms were at the same locations, but increased 4x during AF vs. SR and were located at fiber crossings, fibrosis/ fat zones. Stability of rotational activity correlated with Organization Index (OI), Fraction Index (FI), Shannons Entropy (ShEn), and CL (R>0.5, p< 0.0001). PVI in five hearts increased CL [2-14%] and reduced stability of rotational activity in nearly all regions remote to the pulmonary veins (PVs). Also in the clinical evaluation in nine patients using the HD-catheter (16 electrodes, 3mm dist.) activation delay at the reentrant trajectory was 2x higher at edges with maximal delay (20.5{+/-}8.1ms, median 19.6ms) vs (9.3{+/-}8.8ms, median 9.2ms) and 1.4 x higher during AF (13.0{+/-}18.7ms, median 7.2ms) compared to SR/ CS-pacing (18.0{+/-}11.6ms, median 17.7ms). ConclusionRotational activities in all bi-atrial regions anchored between small frequency-dependent activation delay zones in AF. PVI led to beneficial remodeling in bi-atrial regions remote to the PVs. These data may identify a new paradigm for persistent AF. Subject TermsArrhythmias, Atrial Fibrillation, Cardiac Electrophysiology, High-Density Mapping, Catheter Ablation, Pulmonary Vein Isolation, Fibrosis Clinical PerspectiveO_ST_ABSWhat Is New?C_ST_ABSO_LIRotational activity trajectories based on high-resolution mapping follow propagation line patterns. C_LIO_LIRotational activities anchor frequently between small frequency-dependent slow conduction zones in all bi-atrial regions. C_LIO_LISlow conduction zones are fiber crossings zones and develop into fibrosis and fat regions over time. C_LIO_LIPVI reduces slow conduction zones and AF drivers in regions remote to the PVs in both atria. C_LI What Are the Clinical Implications?O_LIThe new method for the robust detection of rotational activity based on the earliest and latest activation may be useful for an improved AF treatment. C_LIO_LIStability of rotational activity may be predicted with the correlated substrate characteristic fiber crossings density, with slow conduction zones, and with established electrogram measures in the different atrial regions. C_LIO_LIPVI leads to beneficial remodeling in all regions remote to the PVs in the left atrium and right atrium. C_LI