Abstract 18514: A Simulation Study on the Inducibility of Atrial Fibrillation Drivers From Intra-Atrial Electrode Catheters Using Personalized Computational Models Based on LGE-MRI

Abstract

Introduction: Although non-inducibility of atrial fibrillation (AF) drivers just after pulmonary vein isolation (PVI) with or without substrate ablation for persistent AF (PsAF) has been proposed as an acute outcome suggesting a successful long-term outcome, recent studies have found no association with long-term freedom from atrial tachyarrhythmias. It is unclear whether the latter finding is due to complete elimination of all AF drivers or the inadequacy of an AF induction test. Recently, patient-specific computer simulation based on late gadolinium enhancement (LGE)-MRI has been proposed to be useful for detecting rotors as PsAF drivers. Hypothesis: By investigating the accuracy of an induction test following PsAF ablation from the sites commonly used in clinical practice, particularly coronary sinus (CS), we might establish the promising method to confirm non-inducibility during ablation procedure. Methods: Thirty-five consecutive PsAF patients underwent LGE-MRI tests for fibrosis-based computational models. Rotors were induced by rapid pacing from 40 bi-atrial sites near fibrosis after PVI then ablated, repeated until obtaining non-inducibility of rotors. Each induced rotor was investigated for whether it was induced by pacing at lateral right atrium (RA), high RA or CS where electrode catheters were typically placed and used for induction test (Figure). Results: Except for 9 patients having poor LGE-MRI images or low fibrosis burden, 21 patients had a total of 136 locations capable of sustaining extra-PVI rotors. Of the total amount of rotors induced, 24% were induced by pacing at lateral RA, 11% at high RA and 45% at CS, respectively. The capture of all arrhythmogenic locations capable of sustaining rotors improved to 59% by performing induction test from all 3 pacing sites instead of a single one (p<0.01). Conclusions: Although CS is the reasonable pacing site for induction test, that from RA should be also conducted for the optimal rotor detection.