Initial experience of wall thickness guided AF ablation using semiautomated photon counting CT segmentation

Background

Creating a transmural lesion is one of the most important measures when determining AF ablation success. Over the past decades ablation technologies have evolved using different energy sources all with the goal of achieving durable transmural ablation without causing injury extracardiac structures. However, ablation success remains suboptimal and variations of left atrial wall thickness (LAWT) may be a potentially limiting factor. Photon counting CT (PCCT) is a novel imaging modality which allows highly accurate visualization of left atrial wall thickness levels. In this study we describe a novel workflow of combining 3D segmented wall thickness imaging with high resolution electroanatomical mapping for AF ablation.

 Methods and Results

This study represents an initial experience and 10 patients were enrolled scheduled for first time AF ablation. All patients received PCCT imaging (slice thickness 0.2 mm) which was analysed using “inHEART medical” 3D segmentation to quantify LAWT as well as atrial epicardial adipose tissue. Following electro anatomical mapping (HD Grid, Abbott or Pentaray, Biosense Webster) we merged the map with the previously segmented wall thickness CT to be able to display local LAWT during ablation. High power short duration ablation was used in all patients. Energy delivery was titrated according to local wall thickness (< 1mm = 7 sec; 1-2 mm = 10 sec; > 2 mm = 15 seconds). PVI was tailored to avoid the oesophageal sleeve posterior and areas with LAWT > 2.5 mm.

The mean age was 67 ± 11 years with 8 patients being male; 80% paroxysmal AF with a median CHA2DS2-VASc-Score of 3. Mean BMI was 27.3 ± 2. All patients received first time PVI; additional lines were performed in 3 patients (1 anterior line, 1 roof line and 1 CTI Line). Merging was successful in all patients. Mean procedure duration (skin to skin including merging) was 104 ± 16.1 minutes with a mean merging duration of 5.9 minutes. In 7 patients first pass PVI was achieved. No major procedure related complications occurred in this cohort.

Semiautomated PCCT segmentation is a highly accurate and reproducible method to quantify LAWT levels in AF patients. Its integration into a conventional ablation workflow is feasible and does not cause prolonged procedure durations. Future large studies will investigate whether this method will lead to a superior outcome in AF patients with a possibility for reduced complication rates.

Figure 1: LAWT Map using 3D segmentation of PCCT images with wall thickness levels (epi distance) in mm

Figure 2: Merging of LAWT Map with a 3D electroanatomical Map