Ablation of the interatrial connections in atrial fibrillation: insights from a patient-specific computational study

Background Interatrial connections (IACs) have been suggested as pathways contributing to the maintenance of atrial fibrillation (AF). Whether IACs should be targeted for ablation remains uncertain, owing to the difficulty of identifying reentrant circuits involving IACs and the potential risk of interatrial conduction block. Patient-specific atrial computer models can provide mechanistic insights into evaluating IAC contribution, and help determine whether they should be ablated. Purpose To study the characteristics of IAC-dependent reentries and evaluate the role of IACs in reentry maintenance by performing in silico IAC ablations. Methods Six patient-specific biatrial bilayer models were reconstructed from CT and MRI-derived geometries, each incorporating four IACs between the right (RA) and left atrium (LA): Bachmann’s bundle (BB), fossa ovalis (FO), upper posterior (UP), and coronary sinus (CS) connections. The Courtemanche ionic model was used to simulate mild (M) and severe (S) electrical remodeling due to AF, with fibrosis burden ranging from 9.0% (M) to 27.6% (S) in both atria. Conduction velocity was set to 1.0 (M) and 0.8 m/s (S). Electrical propagation was simulated using the monodomain equation and solved with openCARP. Across the 12 models, 110 sustained reentries were induced from different pacing locations via an extrastimulus protocol, followed by circumferential pulmonary vein isolation (PVI). To account for the influence of reentry stability on ablation outcome, 16 IAC ablation strategies were applied at three different timings, yielding a total of 110×16×3=5280 simulations. The number of interatrial rotations across the six possible interatrial loops was quantified by computing topological indices derived from phase maps. The tachycardia cycle length (TCL) and number of phase singularities (PS) in both atria were quantified before and after IAC ablation. Results Following IAC ablation, 11.8% of reentries were IAC-dependent, 40% were sustained only in the LA, 20% only in the RA, and 28.2% had independent reentries in both atria. The number of interatrial rotations was higher in the IAC-dependent group (7.6 ± 8.3 vs 1.8 ± 0.9; Δ=5.8). The TCL of IAC-dependent reentries was slightly longer compared with IAC-independent reentries (198.2 ms vs 192.7 ms; ΔTCL=5.5 ms). The numbers of PS in the LA and RA were comparable between groups. Ablation timing had no significant effect on termination rates. Conclusion IAC-dependent reentries exhibited a markedly higher number of interatrial rotations. IAC-independent reentries also showed occasional interatrial rotations which were secondary to fibrotic conduction barriers and anatomical boundaries. Interatrial rotations were mostly observed in cases with M remodeling. Understanding the distinctive characteristics of IAC-dependent reentries may aid in identifying patients who could benefit from targeted IAC ablation.