P ulmonary vein electric isolation (PVI) is an effective therapy for atrial fibrillation. Bidirectional conduction block between left atrium (LA) and the pulmonary veins (PV) is an accepted end point for PVI. We report a case of rate-dependent unidirectional exit conduction from PV to LA after PVI by large area circumferential ablation, despite entrance and exit conduction block.A 63-year-old man with symptomatic paroxysmal atrial fibrillation refractory to flecainide underwent PVI. An 8-Fr irrigated ablation catheter (Thermocool RMT) connected to a CARTO3 (Biosense Webster, Diamond Bar, CA) electroanatomic mapping system and Stereotaxis remote navigation system were used for PVI. A circular mapping catheter was used to confirm electrogram-guided PVI by large area circumferential ablation. Isolation of the ipsilateral left PVs was obtained, and bidirectional conduction block between the left PVs and LA was confirmed. Subsequently, isolation of the right superior PV (RSPV) was noted by entrance conduction block. Of note, the left PVs and the RSPV exhibited dissociated spontaneous electric activity suggestive of exit conduction block (Figure 1). Pacing from the distal bipoles of the ablation catheter at the anterior aspect of the RSPV was performed at a cycle length of 600 ms (10 mV/2 ms). Pacing captured the PV by suppressing its dissociated activity and unexpectedly revealed exit conduction from the RSPV to LA. The same response was obtained by pacing the posterior aspect of the RSPV without changing the atrial activation sequence, which was earlier in the distal coronary sinus than the right atrium. This rate-dependent exit conduction was reproducible, and it was maintained irrespective of lowering the pacing output down to 1 mV, ruling out far-field capture of the superior vena cava. To further assess this rate-dependent unidirectional exit conduction, the cycle length was changed from 600 to 1500 ms and to 1000 ms. Despite PV capture at 1500 and 1000 ms, exit block persisted. However, by pacing at 600 ms, the rate-dependent exit conduction reappeared. Additional ablation along the RSPV large area circumferential ablation lesion set eliminated the rate-dependent exit conduction at baseline and during both adenosine and isoproterenol challenge (Figure 2).Weerasooriya et al demonstrated that dissociated spontaneous PV activity after PVI represented exit conduction block. Duytschaever et al showed spontaneous isolated PV activity in 171 PVs among 135 patients (35% of 378 patients). However, only 1 of the 171 (0.6%) spontaneous potentials exited to the LA.2 In our patient, 2 behaviors of dissociated spontaneous PV activity were present; the left superior PV exhibited the typical exit block, despite its capture by pacing, and the RSPV showed this unique rate-dependent exit PV-LA conduction. This raterelated conduction abnormality may share some mechanisms found to explain bradycardia-dependent conduction block in Purkinje tissue. Even though phase 4 depolarization was initially implicated, El-Sherif and Jalife have...