Dynamic Behavior of Left Atrial Low-Voltage Areas in Persistent Atrial Fibrillation: Implications for...

Low-voltage areas (LVAs) identified by electroanatomical mapping are commonly used to characterize the atrial substrate in persistent atrial fibrillation (PeAF). However, voltage measurements may vary according to rhythm and physiological state, and LVAs should not be considered synonymous with histological scar or fibrosis.

To evaluate rhythm- and physiological state-dependent changes in left atrial LVAs during atrial fibrillation (AF), early sinus rhythm after cardioversion (SR1), and sinus rhythm under isoproterenol infusion (SR2).

This retrospective observational study included 49 patients with persistent AF undergoing first-time catheter ablation. High-density left atrial voltage maps were acquired during AF, early sinus rhythm after electrical cardioversion (SR1), and sinus rhythm during isoproterenol infusion (SR2). Bipolar LVAs were defined as voltage amplitudes < 0.3 mV during AF and < 0.5 mV during sinus rhythm. Unipolar LVAs were defined using corresponding rhythm-specific thresholds of < 0.71 mV during AF and < 0.83 mV during sinus rhythm, whereas dense bipolar LVAs were defined as bipolar voltage amplitudes < 0.2 mV. Global and regional LVA burden was quantified, and complete paired global analyses included 48 patients.

Global LVA burden decreased stepwise across mapping conditions. Bipolar LVA ratio decreased from 17.25% (IQR, 11.40-31.76) during AF to 8.70% (IQR, 6.98-16.17) during SR1 and 5.20% (IQR, 2.60-10.93) during SR2 (p < 0.001). Similar reductions were observed for unipolar and dense bipolar LVA ratios (both p < 0.001). Regional analysis showed significant heterogeneity in bipolar LVA reduction, whereas unipolar and dense bipolar changes were more homogeneous. Larger left atrial volume was associated with less pronounced bipolar and unipolar LVA reduction. AF duration category was not significantly associated with LVA reduction.

Voltage-defined LVAs in PeAF vary substantially according to rhythm and physiological state. These findings suggest that single-condition voltage mapping may incompletely characterize the atrial substrate and should be interpreted as dynamic electroanatomical changes rather than evidence of fibrosis or scar regression.