Non-invasive driver stability mapping predicts ablation sites of atrial fibrillation termination

Success rates of atrial fibrillation (AF) are still not satisfactory, particularly in persistent AF. Local drivers of variable spatial and temporal stability are key in AF initiation and maintenance. However, current invasive mapping systems lack the capability to provide a comprehensive assessment of individual spatio-temporal AF patterns. Non-invasive image-less ECGi allows for global mapping of spatial and temporal AF dynamics in real-time (1). This enables mapping of AF drivers and their spatio-temporal stability as potential ablation targets.

This study tested whether stable drivers identified by noninvasive mapping coincide with ablation sites resulting in AF termination.

In this substudy of the European multicenter trial SAVE-COR consecutive AF-ablation patients that presented in AF underwent peri- and intraprocedural non-invasive mapping. ECGi-maps based on local AF cycle lengths were repetitively recorded throughout the procedure over periods of 20-40 seconds each.

Stable AF drivers were defined as locations that harboured the lowest AF cycle lenghts (below the 10th percentile of all local cycle lengths at a given timepoint) over more than 50% of the recording time. Potential drivers were visualized according to their spatio-temporal stability ("driver stability maps") and their mean cycle lengths (Fig. 1).

In all patients pulmonary vein (PV) isolation was performed. Additional ablations like left atrial (LA) posterior wall or LA appendage isolation were at the investigators’ discretion. In a regional analysis, the presence or absence of stable drivers was determined for 15 predefined left and right atrial regions. In patients in whom a regional ablation (isolation of the left PV antrum, the right PV antrum, the LA posterior wall or the LA appendage) resulted in AF termination, it was determined whether in this "termination region" a stable driver had been present.

A total of 90 patients with non-invasive mapping during AF ablation were included included. In 18 of those (20%), isolation of either the left PV antrum, the right PV antrum, the posterior wall or the LAA resulted in AF termination – either conversion to sinus rhythm (10 pts) or transition to an organized flutter (8 pts). In 16 patients (88.9%), a stable driver had been present in the "termination region" according to non-invasive driver stability mapping immediately prior to ablation and could thus have been defined as ablation target a priori. On per-patient average, stable drivers were present in 2.8 of the 15 predefined left and right atrial regions. Drivers detected in "termination regions" displayed a mean cycle length of 149 ms and spatio-temporal stability over 60.8 % of the recording period.

Non-invasive driver stability mapping identifies relevant AF drivers and allows for a priori prediction of AF termination sites. It may therefore be a powerful tool to improve AF therapy through personalized ablation approaches.Figure 1