Effect of linear ablation on spectral components of atrial fibrillation.

BACKGROUND

Spectral components of atrial fibrillation (AF) other than the dominant frequency (DF) may represent macroreentrant circuits that coexist with higher-frequency sources during AF.

OBJECTIVE

The purpose of this study was to determine whether spectral components of AF can be eliminated by targeted linear ablation.

METHODS

Antral pulmonary vein isolation (APVI) and linear ablation were performed in 26 patients (age 60 ± 11 years) to eliminate long-standing persistent AF (duration 3 ± 2 years). Spectral analysis of atrial activation at multiple atrial sites was performed during AF, at baseline, after APVI, and immediately before and after liner ablation along the roof of the left atrium, mitral isthmus, and cavotricuspid isthmus. The prevalence and spatial distribution of spectral components of AF were examined before and after each step of ablation.

RESULTS

Twelve (46%) of 26 patients had conversion of AF to atrial tachycardia (AT) during ablation. Mean cycle length of AT was 237 ± 25 ms. A spectral component of AF (3.7 ± 1.2 Hz) other than the DF (6.0 ± 0.9 Hz) was present in 74 (43%) of 173 baseline AF periodograms at multiple atrial sites. Following APVI, no difference in the prevalence of spectral components was seen (38% vs 43%, P = .38). However, linear ablation resulted in a significant decrease in the prevalence of spectral components (24% vs 43%, P < .01), but only when complete conduction block was achieved.

CONCLUSION

Elimination of spectral components of AF by targeted linear ablation suggests that spectral components may indicate site-specific ATs that coexist with AF despite a lower frequency than the DF of AF.