Right atrial structural remodeling outweighs frequency-defined drivers in persistent atrial fibrillation: the multicenter randomized PAD-AF trial

Background Outcomes of catheter ablation for persistent atrial fibrillation (AF) remain suboptimal despite durable pulmonary vein isolation (PVI). Dominant frequency (DF) mapping has been proposed to identify AF-maintaining drivers, yet its incremental value beyond PVI has not been tested in a multicenter randomized trial. The relative importance of functional high-frequency activity versus structural remodeling in determining recurrence remains unclear. Purpose To evaluate whether DF-guided ablation improves outcomes after PVI and to define the mechanistic hierarchy between functional frequency domains and structural substrate in persistent AF. Methods PAD-AF was a prospective, investigator-initiated, multicenter randomized trial conducted at six centers. Patients undergoing first ablation for persistent or long-standing persistent AF were enrolled. After successful PVI, bi-atrial high-density mapping with fast Fourier transform analysis was performed during ongoing AF. High-DF sites were defined as ≥7 Hz with regularity index ≥0.2. Patients with high-DF sites were randomized 1:1 to adjunctive DF ablation or PVI alone. Patients without high-DF sites formed a predefined exploratory cohort. Low-voltage areas (LVAs; <0.5 mV) were quantified during sinus rhythm. The primary endpoint was freedom from AF recurrence (>30 s) off antiarrhythmic drugs at 12 months after a single procedure. Results A total of 103 patients were enrolled; 64 with high-DF sites were randomized (DF n=32; PVI n=32), and 39 had no high-DF sites. AF-free survival off antiarrhythmic drugs at 12 months was 81.3% in the DF group versus 68.9% in the PVI group (P=0.228). Freedom from any atrial arrhythmia was 78.1% versus 65.6% (P=0.263). Patients without high-DF sites demonstrated the most favorable outcome (83.3%; P=0.065 vs PVI), identifying absence of high-frequency activity as a favorable substrate phenotype. High-DF sites showed minimal spatial concordance with LVAs (LA overlap 9–19%; RA 3–6%), indicating functional–structural dissociation. In multivariable analysis, right atrial LVA burden independently predicted recurrence (HR 1.031 per 1% increase; 95% CI 1.005–1.058; P=0.018), whereas DF ablation did not. No major complications occurred. Conclusion Adjunctive DF-guided ablation did not significantly improve outcomes beyond PVI. Right atrial structural remodeling—not frequency-defined drivers—emerged as the dominant determinant of recurrence, redefining the mechanistic hierarchy of substrate relevance in persistent AF. These findings represent a paradigm shift, suggesting that frequency mapping may serve primarily for substrate stratification rather than empirical ablation and highlighting right atrial remodeling as a key target in the era of pulsed field ablation.