Automated CT-derived assessment of left atrial wall thickness before and after pulsed-field and radiofrequency ablation for atrial fibrillation

Background

Catheter ablation of atrial fibrillation (AF) may induce variable structural remodeling of the left atrium (LA) depending on the ablation energy. Pulsed field ablation (PFA) induces non-thermal electroporation with selective myocardial targeting, whereas radiofrequency (RF) ablation generates thermal lesions that can result in diffuse atrial shrinkage. Automated quantification of LA wall thickness (LAWT), wall volume, LA volume, sphericity and LAA volume from cardiac CT provides a reproducible way to assess this remodeling.

Objectives

To evaluate post-ablation changes in LA wall morphology and chamber geometry after AF ablation using either PFA or RF energy, based on semi-automated 3D CT-derived measurements.

Methods

This single-center retrospective study included 90 consecutive patients (63.6 ± 11.0 years; 64% men; 53% paroxysmal AF) who underwent AF ablation between 2022 and 2024 with both pre- and post-procedure cardiac CT. Pre- and post-ablation CTs were performed at 0.5 [0.3–0.9] months before and 3.1 [3.0–4.1] months after ablation, respectively. Patients were divided into RF (n = 30) and PFA (n = 60) groups. Patients in RF group received PVI. Patients in PFA group received PVI or PVI with other lesions. Quantitative analyses were performed using ADAS3D software with a standardized 8-region segmentation of the LA.

Results

RF ablation led to a significant global reduction in LA geometry total LA wall volume −12.6 % (p < 0.0001), LA chamber volume −16.7 % (p < 0.001), LA body volume −17.2 % (p < 0.001), except for mean LAWT –0,5%, p = 0,8148. These changes were predominantly located in ablated areas: posterior wall (7a–7d), roof (3b–3d), intervein regions (1–2), while non-ablated regions such as the septum (8) remained unchanged. PFA did not produce global changes in mean LAWT (-2,6 %, p = 0,054), or volumes (LA wall volume – 2,6 %; LA body volume 0 %, LA chamber volume – 0,8 %) but induced localized remodeling limited to treated regions. Wall thinning occurred in the ridge (6a) and intervein areas/roof (2, 3a–3b) (−6 % to −16 %, p ≤ 0.003), with a slight increase in LA sphericity (+0.5 %, p = 0.028).

Remote, non-ablated regions (anterior wall, septum) remained stable. Finally, there was no significant difference in LAA volume before and after ablation in each group, even if the effect seemed different between groups (+ 7,5 % in RF group, p = 0,089; - 4,2 % in PFA group, p = 0,33)

Conclusions

RF ablation induced global, homogeneous remodeling of the LA, with significant reductions in wall and cavity geometries probably reflecting diffuse atrial contraction secondary to thermal injury.

In contrast, PFA led to localized and spatially confined changes in ablated regions, sparing remote zones and preserving overall atrial size.

Mean LAWT did not significantly decrease with either energy source, even if some ablated segments decreased with both techniques, suggesting a chronic local thinning caused by ablation.LA 3D segmentation