Abstract 18526: Electrogram Amplitude is Reduced at Rotor Sites Critical to Maintenance of Human Persistent Atrial Fibrillation

Abstract

Introduction: Increasing studies find human AF is driven by rotors, but their detection remains challenging. Theoretically, rotor cores display low amplitude and multiphasic signals, but human AF rotors fail to display such signatures. This may reflect rotor precession (obscuring analysis site chosen), poor basket contact or bipoles (that do not register orthogonal wavefronts). We hypothesized that studying unipolar signals at rotor precession areas may reveal hitherto hidden electrogram signatures. Methods: We studied 22 persistent AF patients (60±11 years, CHADS2=1.9), in whom phase mapping of 64 pole basket signals revealed rotors/sources where ablation terminated AF (fig A). Unipolar electrograms over the rotor precession area were measured by 3 blinded observers for spectral dominant frequency (DF), peak-to-peak amplitude and duration. This was compared to a paired source in each patient where ablation did not terminate AF (‘controls’) (fig B). Results: Patients had 3.3±1.7 AF rotor/focal sources. At the index source, AF terminated to sinus rhythm (n=14) or atrial tachycardia by targeted ablation (mean 4.0 mins). Compared to control areas, AF sustaining rotors/sources had lower amplitude (fig C) (0.60±0.30mV vs. 0.87±0.70mV, p=0.02). Conversely, electrogram spectra (90±45Hz vs. 80±46Hz, p=ns), duration (125±45ms vs. 116±47ms, p=ns) and AF cycle length (174±8ms vs. 174±6ms, p=ns) did not differ. Conclusions: Rotors/sources that maintain human persistent AF have lower unipolar amplitude than non-termination sites. This is the first identified electrogram signature for functionally critical human AF rotors, and motivates definition of structure/function correlations at these sites.