Mapping the Progression of Atrial Fibrillation Substrate: Insights from Left Atrial Wall Thickness and High-Density...

Background/Objective: Advances in cardiac computed tomography (CT) now allow high-resolution, three-dimensional reconstruction of the left atrium, enabling precise quantification of left atrial wall thickness (LAWT). However, the relationship between LAWT and electrophysiological properties, particularly as assessed by high-density bipolar voltage mapping, remains poorly defined. The objective was to evaluate left atrial structural remodeling in atrial fibrillation (AF) by integrating three-dimensional CT-derived LAWT maps with high-density bipolar voltage mapping, and to assess the relationship between structural remodeling and electrophysiological changes across AF phenotypes. Methods: In this single-center observational study, 60 patients (30 with paroxysmal AF and 30 with persistent AF) undergoing their first AF ablation were enrolled. Pre-procedural cardiac CT scans were analysed using a validated artificial intelligence (AI)-based method to generate three-dimensional LAWT maps. The left atrium was segmented into five regions (lateral, anterior, septal, posterior, and roof) excluding the pulmonary veins and left atrial appendage. High-density bipolar voltage maps were acquired using the CARTO system and a multipolar catheter. Global and regional LAWT and bipolar voltage values were compared between AF phenotypes. The association between global LAWT and bipolar voltage was assessed at the patient level using Pearson's correlation coefficient, whereas regional comparisons and associations between LAWT and bipolar voltage were evaluated using linear mixed-effects models accounting for repeated measurements within patients. Results: Patients with persistent AF showed larger left atrial volume (119 ± 26 vs. 102 ± 31 mL; p = 0.024) and a trend toward greater CT-derived LAWT, with significant differences observed in selected atrial regions. Overall mean bipolar voltage and LAWT were 1.86 ± 0.59 mV and 1.45 ± 0.32 mm, respectively. No meaningful linear association was observed between these parameters (global Pearson r = -0.018), a finding that persisted across regions and AF phenotypes. No clinical factors were significantly associated with global LAWT, although heart failure and persistent AF showed a trend toward greater global wall thickness. Conclusions: Compared with paroxysmal AF, persistent AF was associated with a tendency toward increased left atrial wall thickness, consistent with more advanced structural remodeling. The lack of a meaningful correlation between wall thickness and bipolar voltage suggests that LAWT provides complementary structural information beyond electroanatomical mapping.