Artificial Intelligence for Individualized Treatment of Persistent Atrial Fibrillation: a Randomized Controlled Trial

Deisenhofer, I.

doi:10.1038/s41591-025-03517-w

TAILORED-AF

Artificial Intelligence for Individualized Treatment of Persistent Atrial Fibrillation: a Randomized Controlled Trial

Patient / Population Intervention / Exposure Comparison Outcome

In 370 patients with drug-refractory persistent atrial fibrillation, AI-guided ablation of electrogram dispersion added to pulmonary vein isolation improved freedom from atrial fibrillation at 12 months compared with pulmonary vein isolation alone.

N

370patients

Design

Multicentre double-blind superiority RCT

Endpoint

Freedom from documented AF, with or without antiarrhythmic drugs, at 12 months after a single ablation procedure

Relevance

2Important — one of several pillars.

ResultFreedom from AF 88% vs 70% (log-rank P<0.0001 for superiority); no significant difference in freedom from any atrial arrhythmia after one procedure; safety composite did not differ; procedure and ablation times twice as long in the tailored arm.

Deisenhofer I, et al. Nat Med. 2025;31(4):1286-1293. 10.1038/s41591-025-03517-w

Discussion & critique

The first randomised trial in which an AI-identified ablation target beat pulmonary vein isolation alone in persistent AF. The gain was specific to AF: freedom from any atrial arrhythmia was not different, because tailored ablation can leave organised atrial tachycardias that need a further procedure, and the procedures took twice as long. Follow-up was 12 months after a single procedure.