Numerical modeling of structural scar homogenization with pulsed field ablation in infarcted swine ventricles

Background

Pulsed Field Ablation (PFA) is a new ablation method rapidly being adopted for the treatment of cardiac arrhythmias. Scar homogenization is a commonly used ablation strategy for scar-related ventricular tachycardia (VT). This strategy involves multiple PFA applications in border zone regions to eliminate re-entry circuits. However, quantifying the tissue-specific extent of these lesions in vivo is challenging, and assessment is largely limited to histology.

Purpose

Predict and evaluate the ablated volume after PFA scar homogenization in the infarcted swine left ventricle (LV) using numerical modeling.

Methods

MRI-guided structural scar homogenization with PFA was performed in five infarcted swine ventricles (4–6 weeks post-infarct) using an 8 Fr, 5 mm tip focal catheter in the LV and a return catheter in the inferior vena cava (IVC). Between 21 and 27 applications per animal were delivered, each consisting of eight R-wave–gated biphasic pulse trains at 1500 V.

Pre-ablation 3D late gadolinium enhancement magnetic resonance images (LGE MRI) were processed to obtain animal-specific ventricular geometry and fibrotic tissue distribution (scar map). In the scar map, normalized pixel intensity thresholds of 40% and 60% differentiated healthy tissue (<40%), border zone (40%-60%), and dense scar (>60%). The numerical model incorporated the animal-specific geometry (LV, RV, and IVC) and scar map, which was used to assign distinct, tissue-specific electrical properties (Figure 1, Step 1).

The electric field distribution was calculated for each PFA application with the ablation catheter positioned according to the ablation tags from the cardiac mapping system (Figure 1, Step 2). The ablated volume was determined by combining regions of tissue in the infarcted LV where the electric field exceeded the lethal threshold of 456 V/cm (as determined in the previous study [1]) for each PFA application, accounting for overlaps (Figure 1, Step 3).

Results

The modeling framework provided detailed, animal-specific predictions of the ablated volume and its spatial distribution relative to the fibrotic tissue distribution (Figure 2, panels A and B). Across the five animals, the models predicted a mean total ablated volume of 2.8±0.3 cm³, consisting of 1±0.4 cm³ healthy tissue, 1.2±0.2 cm³ border zone, and 0.6±0.4 cm³ dense scar (Figure 2, panel C). These data suggest a new mean dense scar volume of 2.2±0.2 cm³, assuming conversion of the initially ablated healthy tissue and border zone.

Conclusion

The presented in silico framework provides a method to quantify and visualize the predicted tissue-specific impact of PFA scar homogenization, demonstrating potential for valuable insights into treatment outcomes.Numerical modeling workflowNumerical modeling results