Partial vagal denervation increases vulnerability to vagally induced atrial fibrillation

INTRODUCTION: Cervical vagal stimulation shortens the atrial effective refractory period (ERP) primarily in the high right atrium (HRA) and facilitates induction of atrial fibrillation (AF) by single premature HRA extrastimuli. We hypothesized that vagal denervation of the HRA prevents both ERP shortening in the HRA and AF induction during vagal stimulation. METHODS AND RESULTS: Vagal denervation of the HRA was achieved using radiofrequency catheter ablation (RFA) of the fat pad at the right pulmonary vein-atrial junction (RPV fat pad). Programmed stimulation was performed at each of four atrial sites to measure ERP and inducibility of AF during vagal stimulation. RPV fat pad RFA increased only the HRA ERP during vagal stimulation (70 +/- 8.7 vs 117 +/-14.8, P < 0.05). RPV fat pad RFA increased measures of dispersion of refractoriness, the standard deviation of ERP (24 +/- 2.1 vs 33 +/- 2.0, P < 0.01), and the standard deviation of AF cycle length (11 +/- 0.8 vs 22 +/- 1.7, P < 0.001) during vagal stimulation. RPV fat pad RFA increased the incidence of AF (15/28 vs 24/28, P < 0.05) and the vulnerability (22 +/- 4.7 vs 39 +/- 5.6, P < 0.01) to AF induction during vagal stimulation, particularly from left atrial premature beats. After RPV fat pad RFA, premature beats induced AF by causing conduction block primarily in the HRA and macroreentrant activation around the block. CONCLUSION: Partial right atrial vagal denervation facilitated rather than prevented initiation of vagally mediated AF.