A 69-year-old woman with a 10-year history of paroxysmal atrial fibrillation was initially treated with propafenone and amiodarone. Four years ago her arrhythmia burden worsened and she developed recurrent episodes of symptomatic persistent atrial fibrillation. After several cardioversions, she was referred to our institution for radiofrequency catheter ablation. Her amiodarone was discontinued, and 1 year ago she underwent radiofrequency catheter ablation using a pulmonary vein antral isolation approach. Unfortunately, she developed recurrent atrial fibrillation and remained extremely symptomatic despite adequate control of her heart rate. Six months ago, she underwent a second radiofrequency catheter ablation procedure. Electrophysiology study revealed recurrent conduction between all four pulmonary veins and the left atrium, and the veins were reisolated and linear ablations were performed on the roof of the left atrium and the cavotricuspid isthmus. After the ablation, no atrial tachycardia or atrial fibrillation could be induced with aggressive pacing protocols. Unfortunately, she developed recurrent persistent atrial fibrillation within 1 week of the procedure. Four months later she underwent cardioversion with initiation of dofetilide. Two days after starting dofetilide she developed atypical atrial flutter and was referred for a third electrophysiology study. Surface ECG revealed upright P waves in leads V1 and V2 but were poorly seen in the other leads. At electrophysiology study, a circular mapping catheter placed in the four pulmonary veins demonstrated no electrical activity during atrial tachycardia, confirming adequate pulmonary vein isolation. At baseline the atrial tachycardia had stable alternating cycle lengths of 365 ms and 410 ms. Initial pacing maneuvers from the anterolateral right atrium and the proximal and distal coronary sinus yielded post-pacing intervals with long return cycle lengths. In addition, constant fusion and progressive fusion during pacing from the right atrium and the coronary sinus were observed. In order to delineate the tachycardia circuit a 64-electrode basket catheter was placed in the posterior left atrium (Figure 1). At baseline, electrograms could be recorded for 318 ms of the 363-ms and 411-ms cycle lengths (88% and 77% respectively) over a large area in the posterior left atrium. The deflection of the P wave corresponded to activation of the superior and anterior portions of the left atrium. We mapped the tachycardia circuit by identifying the initial site of cycle length change that localized the critical isthmus to the posterior roof of left atrium near the right superior pulmonary vein (Figure 2). When the mapping catheter was moved septally or laterally, split potentials suggestive of a line of conduction block were observed and a 20-electrode catheter revealed a fractionated electrogram that was the first site of cycle length change. Unfortunately, adequate capture could not be obtained despite pacing at maximal outputs from either electrode pair 17 and 18 of the 20-electrode reference catheter or an ablation catheter placed at the same fluoroscopic site. In addition, despite maximal gains, low amplitude electrograms could not be recorded with the The authors report no conflicts of interest for the published content. Manuscript received January 23, 2012, final version accepted February 13, 2012. Address correspondence to: Fred M. Kusumoto, MD, Electrophysiology and Pacing Service, Division of Cardiovascular Disease, Mayo Clinic, 4500 San Pablo Ave, Jacksonville, FL 32224. E-mail: Kusumoto.fred@mayo.edu The Journal of Innovations in Cardiac Rhythm Management, 3 (2012), 727–730