Characterization of AF Recurrence and Substrate Health by Electrographic Flow (EGF) Consistency in EGF Mapping

Background

Electrographic flow (EGF) mapping is an FDA 510(k)‐cleared method for visualizing atrial activation wavefronts in atrial fibrillation (AF). Its clinical efficacy in detecting AF sources was demonstrated in the FLOW‐AF randomized controlled trial, and the underlying machine learning strategy used to develop and refine EGF source detection has been recently detailed. However, EGF mapping metrics can also characterize other properties of atrial wavefront propagation—most notably Electrographic Flow Consistency (EGFC) as a measure of the health of the atrial substrate. Patients with low EGFC have been found to have recurrent AF more frequently than those with high EGFC. EGFC may therefore be additive to source presence to phenotype AF patients based on mechanism of disease and recurrence likelihood.

Objectives

Present the EGFC‐based Kaplan–Meier curves for each study arm of FLOW‐AF and explore the clinical relevance of EGFC using prospective studies.

Methods

Unipolar electrograms were recorded using 64‐electrode basket catheters in AF patients across three clinical trials. The EGF algorithm then processed these recordings to reconstruct wavefront propagation patterns. EGFC was quantified from the vector length at each point in space and averaged across spatial coordinates and mapping positions to determine the mean EGFC for each patient. As applicable, EGFC and related EGF map characteristics were compared with 12‐month freedom from AF, patient demographic factors, and bipolar voltage mapping data.

Results

Primarily, mean biatrial EGFC < 0.62 was found to be the best metric at predicting recurrent AF in No Source and Treatment patients in FLOW‐AF: it had 86% ± 11% accuracy after fivefold cross‐validation. Among these patients, those with EGFC ≥ 0.62 had 6% recurrence vs. 51% recurrence in those with EGFC < 0.62; p  < 0.001. Control patients with sources post‐procedure had 77% vs. 100% recurrence in respective EGFC groups; p  = 0.032. Follow‐up analyses showed that EGFC dropped from 0.58 ± 0.15 to 0.51 ± 0.14 (paired p  = 0.041) in a subgroup of patients who underwent repeat mapping 103 ± 22 days later, but it was not affected by pulmonary vein isolation (PVI) or source ablation (paired p  > 0.30). It was also higher among de novo and paroxysmal AF patients than redo ( p  = 0.043) and non‐paroxysmal AF patients ( p  = 0.022), respectively. Among the group of patients with low EGFC and no sources post‐procedure, recurrent patients had increased EGFC spatial variability ( p  = 0.026) and more near‐threshold sources with source activity ≥ 25% but < 26.5% ( p  = 0.038). Lastly, EGFC was globally and locally correlated to bipolar voltage across multiple rhythms, atria, mapping software and analysis techniques ( p  = 0.009).

Conclusions

EGFC may provide insights on the substrate health of the patient and is predictive of recurrence likelihood. An understanding of EGFC may therefore increase mechanistic understanding of the progression and therapeutic strategies for AF ablation that can be tailored to the individual patient.

Clinical Trial Registration: FLOW‐AF: NCT04473963; FLOW EVAL‐AF: NCT06260670; AF‐FLOW Global Registry: NCT05481359.