Low-energy (360 J) radiofrequency catheter ablation using moderate power − short duration: proof of concept based on in silico modeling

Abstract

Background

Pilot clinical studies suggest that very high power–very short duration (vHPvSD, 90 W/4 s, 360 J energy) is a feasible and safe technique for ablation of atrial fibrillation (AF), compared with standard applications using moderate power–moderate duration (30 W/30 s, 900 J energy). However, it is unclear whether alternate power and duration settings for the delivery of the same total energy would result in similar lesion formation. This study compares temperature dynamics and lesion size at different power-duration settings for the delivery of equivalent total energy (360 J).

Methods

An in silico model of radiofrequency (RF) ablation was created using the Arrhenius function to estimate lesion size under different power-duration settings with energy balanced at 360 J: 30 W/12 s (MPSD), 50 W/7.2 s (HPSD), and 90 W/4 s (vHPvSD). Three catheter orientations were considered: parallel, 45°, and perpendicular.

Results

In homogenous tissue, vHPvSD and HPSD produced similar size lesions independent of catheter orientation, both of which were slightly larger than MPSD (lesion size 0.1 mm deeper, ~ 0.7 mm wider, and ~ 25 mm³ larger volume). When considering heterogeneous tissue, these differences were smaller. Tissue reached higher absolute temperature with vHPvSD and HPSD (5–8 °C higher), which might increase risk of collateral tissue injury or steam pops.

Conclusion

Ablation for AF using MPSD or HPSD may be a feasible alternative to vHPvSD ablation given similar size lesions with similar total energy delivery (360 J). Lower absolute tissue temperature and slower heating may reduce risk of collateral tissue injury and steam pops associated with vHPvSD and longer applications using moderate power.

Appendix 1

1.1 Material properties

Table 3 Physical characteristics of tissues and materials of each element used in the models

1.2 Boundary conditions

Each model (half of the entire domain, see Fig. 1) was made of ~ 214,000 tetrahedral elements and ~ 300,000 nodes. The outer dimensions (4 cm around the ablation electrode), mesh size (minimum of 118 μm around the electrode and maximum of 12 mm in the periphery), and the time step (ranging from 20 to 100 ms) were verified by means of convergence tests. The model solved a coupled electric-thermal problem numerically using the finite element method with ANSYS software (ANSYS, Canonsburg, PA, USA). Laplace’s equation and bioheat equation were used to solve the electrical and thermal problem, respectively. Details of these equations and boundary conditions are described in detail elsewhere [7]. The dispersive electrode was simulated using conditions of 0 V on all the outer surfaces except the symmetry plane. Electrode irrigation was modeled by fixing a value of 25 °C in the cylindrical zone of the electrode tip, and leaving the semispherical tip free, which mimics a multi-hole electrode since we are assuming that irrigation occupies almost the entire surface of the electrode [15]. The thermal problem was not solving in the blood, and thermal transfer coefficients for low blood flow condition (0.1 m/s) were used on the interfaces electrode-blood and myocardium-blood instead (3310 W/m²·K and 694 W/m²·K, respectively). A temperature of 37 °C was set at the outer contours of the tissue.

1.3 Thermal injury model

Lesion size (depth, maximum width, and volume) was computed using the thermal injury index Ω based on the Arrhenius model, which relates the number of undamaged cells C(0) present before the heating to the remaining number of undamaged cells at time t indicated by C(t) as follows:

$$\Omega \left(t\right)=\mathrm{ln}\left(\frac{C\left(o\right)}{C\left(t\right)}\right)$$

(1)

It can be computed from the “thermal history” to which the tissue is subjected, specifically from the temperature T (in Kelvin) reached at each instant t (s) of the heating period:

$$\Omega \left(t,T\right)={\int }_{0}^{t}A{e}^{\left[{-E}_{a}/RT\left(\tau \right)\right]}d\tau$$

(2)

where A is the frequency factor, E_a is the activation energy, and R the universal gas constant (8.3143 J/mol·K). The values of A and E_a are particular for each tissue type and analyzed process. In this study, the Ω = 1 isoline was considered to represent the thermal lesion contour. We considered A = 7.39·10³⁹ s⁻¹ and E_a = 2.577·10⁵ J/mol, since these values were checked in a previous study[10] in order to provide a Ω = 1 isoline more or less coincident with the 72 °C isotherm for 5 s heating and more or less coincident with the 55 °C isotherm after 60 s of heating [21, 22]. To sum up, lesion size was volumetrically calculated by considering all the elements of the mesh in which Ω ≥ 1.