Real-world experience in routine usage of second-generation artificial intelligence-guided catheter ablation in long-standing persistent atrial fibrillation in de novo and repeat cases

Introduction: Despite advanced ablation strategies and major technological improvements, treatment of persistent atrial fibrillation (AF) remains challenging and the underlying pathophysiology is not fully understood. Recently, the Tailored-AF trial (NCT04702451) using artificial intelligence (AI)-guided catheter ablation demonstrated superiority to a pulmonary vein isolation (PVI)-only procedure in persistent AF in attaining AF-freedom. 

Purpose: To investigate outcome and safety of routine usage of AI-guided catheter ablation in long-standing persistent AF patients in de novo and repeat cases using the new generation Volta AF-Xplorer.

Methods: In this prospective, single-centre, non-randomized study, 100 consecutives long-standingpersistent AF patients underwent catheter ablation. High density mapping was performed with EnSite NavX^TM, Advisor^TM HD Grid (both Abbott) and the AI-based (Volta AF-Xplorer^TM, Volta Medical) software. This expertise-based AI tool was trained to detect spatio-temporal dispersion (DISPERS) in multipolar intracardiac EGMs. After LA mapping, pulmonary veins were isolated (de novo patients) or re-isolated (repeat procedure) and ablation of automatically-tagged DISPERS-areas was performed to homogenize, dissect, isolate, or connect DISPERS areas to non-conducting anatomical structures. 

Results: Out of 100 patients (mean age 70.9±9.5 years, n=45 female), 56 patients were referred for a de novo procedure while 44 had a repeat procedure. Re-isolation of reconnected PV was performed in 34/44 patients who received an AF ablation before first-time DISPERS-guided ablation. In 85/100 patients, at least one linear lesion was created (anterior line, n=39; roof line, n=38; floor line, n=13; cavotricuspid isthmus, n=4 and posterior box lesion, n=31). Average procedure time was 140.6±45.3 min with a fluoroscopy time of 10.9±7.6 min and a radiofrequency duration of 26.9±10.4 min. Following PVI, ablation of DISPERS areas led to AF cycle length prolongation (163.7±22.4 ms before ablation vs 189.8±26.5 ms after) and AF termination to atrial tachycardia (AT) or sinus rhythm (SR) in 22 patients. No stroke or pericardial tamponade occurred; Groin complications were detected in 5 patients. Early recurrence (≤42 days) was seen in 32 patients (18 organized AT, 10 AF, 4 AF/AT). After a 6-week-blanking period, recurrence was documented in 8 patients. Re-ablation was performed in 6 AT patients. After a mean of 1.5 months (50.1±114.8 days), 92/100 patients remained in stable SR, 32/92 temporarily on antiarrhythmic drugs.

Conclusion: This study reports for the first time the routine use of DISPERS-ablation guided by AI-based software enabling a quick, automatic, operator- and experience-independent annotation of potential AF trigger areas in the LA high-density 3D EAM allowing a signal-based, patient-specific ablation approach in persistent AF patients. While long-term outcome data is needed, short term outcome is promising with low complication rates.   

Legend: Postero-anterior and antero-posterior views showing ablated LA dispersion regions (orange tags). LA mapping and automated tagging with a second-generation artificial intelligence algorithm software