Ablation and antiarrhythmic drug effects on PITX2 +/- deficient atrial fibrillation: A computational modeling study

Abstract

Introduction: Atrial fibrillation (AF) is a heritable disease, and the paired-like homeodomain transcription factor 2 (PITX2) gene is highly associated with AF. We explored the differences in the circumferential pulmonary vein isolation (CPVI), which is the cornerstone procedure for AF catheter ablation, additional high dominant frequency (DF) site ablation, and antiarrhythmic drug (AAD) effects according to the patient genotype (wild-type and PITX2 ^+/- deficient) using computational modeling.

Methods: We included 25 patients with AF (68% men, 59.8 ± 9.8 years of age, 32% paroxysmal AF) who underwent AF catheter ablation to develop a realistic computational AF model. The ion currents for baseline AF and the amiodarone, dronedarone, and flecainide AADs according to the patient genotype (wild type and PITX2 ^+/- deficient) were defined by relevant publications. We tested the virtual CPVI (V-CPVI) with and without DF ablation (±DFA) and three virtual AADs (V-AADs, amiodarone, dronedarone, and flecainide) and evaluated the AF defragmentation rates (AF termination or changes to regular atrial tachycardia (AT), DF, and maximal slope of the action potential duration restitution curves (Smax), which indicates the vulnerability of wave-breaks.

Results: At the baseline AF, mean DF (p = 0.003), and Smax (p < 0.001) were significantly lower in PITX2 ^+/- deficient patients than wild-type patients. In the overall AF episodes, V-CPVI (±DFA) resulted in a higher AF defragmentation relative to V-AADs (65 vs. 42%, p < 0.001) without changing the DF or Smax. Although a PITX2 ^+/- deficiency did not affect the AF defragmentation rate after the V-CPVI (±DFA), V-AADs had a higher AF defragmentation rate (p = 0.014), lower DF (p < 0.001), and lower Smax (p = 0.001) in PITX2 ^+/- deficient AF than in wild-type patients. In the clinical setting, the PITX2 ^+/- genetic risk score did not affect the AF ablation rhythm outcome (Log-rank p = 0.273).

Conclusion: Consistent with previous clinical studies, the V-CPVI had effective anti-AF effects regardless of the PITX2 genotype, whereas V-AADs exhibited more significant defragmentation or wave-dynamic change in the PITX2 ^+/- deficient patients.

Keywords: PITX2; antiarrhythmic drug; atrial fibrillation; computational modeling; dominant frequency.

Figure 1

Study protocol of the computational atrial modeling, AF simulation, and virtual interventions. (A) Integration of the CT imaging and electroanatomical map. (B) Computational modeling integrating the anatomy, fibrosis, fiber orientation, and LAT map. (C) Protocol of the AF simulation and wave-dynamic analysis. AF was induced in each case using AF pacing from 200 to 120 ms with eight beats per cycle lasting a total of 11,520 ms based on the wild-type PITX2^+/− deficient AF baseline ion current settings. AF maintenance was observed for 20,480 ms after induction (overall 32 s including pacing), and the wave dynamics of the DF and Smax were analyzed from 17,000 to 23,000 ms. (D) Baseline AF induction under wild-type and PITX2^+/− deficient backgrounds. The voltage maps and ECGs indicate a successful AF induction during the wild-type and PITX2^+/− deficient baselines. (E) 3D DF map of the baseline AF under wild-type and PITX2^+/− deficient backgrounds. The black arrows indicate the locations of the high DF sites on the 3D DF map. (F) Virtual CPVI with a high DF site ablation. The green lines indicate the CPVI and yellow arrows indicate the ablated regions of the high DF sites. The pink sites indicate the pacing site. (G) Ion current changes with the high and low doses of the three types of AADs. Ion current changes with the high and low doses of the three types of AADs under the wild-type and PITX2^+/− deficient backgrounds. For PITX2^+/− deficiency, the I_K1 decreased by 25% and the I_Kr increased by 100% as compared to that with the wild-type status, while the other ion currents remain the same as the wild-type. (H) Smax and DF analysis after AADs and the CPVI. The ECGs indicate AF was maintained after AADs, and AF converted to AT after the CPVI. CT, computed tomography; EP, electrophysiology; LAT, local activation time; PITX2, paired-like homeodomain transcription factor 2; CPVI, complete pulmonary vein isolation; DF, dominant frequency; Smax, the Maximal slope of the restitution curves; AF, atrial fibrillation; AT, atrial tachycardia; ECG, electrocardiogram; AAD, antiarrhythmic drug.

Figure 2

Characteristics of wild-type and PITX2^+/− deficient baseline AF. Genotype-dependent comparisons of the APD₉₀, CV, mean Smax, mean DF, and AFCL depend on the baseline AF. Every group includes an identical number of samples (n = 25). APD₉₀, action potential duration 90%; CV, Conduction velocity; Smax, the Maximal slope of the restitution curves; AFCL, AF cycle length; DF, Dominant frequency; PITX2, paired-like homeodomain transcription factor 2.

Figure 3

Wave-dynamic change after a virtual CPVI and AADs. (A) The ECGs were obtained at the black ^* sites in the DF maps and indicate that AF converted to AT after the CPVI during a wild-type condition was still maintained during a PITX2^+/− deficient condition. (B) The ECGs were obtained at the black ^* sites in the 3D DF maps and indicate that AF was still maintained after high dose amiodarone under both wild-type and PITX2^+/− deficient backgrounds. (C) The ECGs were obtained at the black ^* sites in the DF maps and indicate that the AF was still maintained after high dose flecainide under both wild-type and PITX2^+/− deficient backgrounds. DF, dominant frequency; Smax, the maximal slope of the restitution curves; CPVI, complete pulmonary vein isolation; AF, atrial fibrillation; AT, atrial tachycardia; ECG, electrocardiogram; PITX2, paired-like homeodomain transcription factor 2.

Figure 4

Genotype-dependent comparisons of the AF defragmentation (A) and AF termination (B) rates, mean Smax (C), and DF (D) depending on the high and low doses of the three types of AADs. Every group includes identical number of samples (n = 25). Smax indicates the maximal slope of the restitution curves; DF, Dominant frequency; PITX2, paired-like homeodomain transcription factor 2.