Prospective 3-dimensional computed tomography segmentation of the pericardiac right phrenic nerve in the setting of atrial fibrillation ablation.

A 58-year-old man was referred for catheter ablation of symptomatic persistent atrial fibrillation without structural heart disease. Informed consent and local committee approval were obtained. He underwent preprocedural cardiac computed tomography (CT) angiography after administration of intravenous contrast for evaluation of left atrium and pulmonary vein (PV) anatomy. The pericardiac course of the right phrenic nerve (RPN) was delineated on the CT imaging by an experienced radiologist, either using direct RPN visual-ization or inferring the position of the right pericardiophrenic bundle by visualizing the right pericardiophrenic artery when RPN could not be directly seen. A new CT imaging sequence was created containing the original CT images with the path of the RPN overwritten by a 5-pixel wide 3-dimensional curved line at 900 HU value in attenuation to facilitate seg-mentation of the RPN in the electrophysiology laboratory. Three-dimensional reconstruction of the CT images in the new sequence was performed using the CARTO 3 software (Biosense Webster, Diamond Bar, CA). The cardiac chambers , PVs, and the RPN were segmented using the applied specific attenuation threshold. Detailed contact electroanatomic mapping of the left atrium, PVs, right atrium, and superior vena cava was performed using the fast anatomic mapping feature of CARTO 3, and the fast anatomic mapping was then merged with the CT images using CARTOMERGE. Differential output pacing (10, 30, and 50 mA at 2-ms pulse width) was performed along the course of the RPN in right superior PV, left atrium, superior vena cava, and right atrium. RPN capture was assessed by manual palpation of the patient's right hypochondrium for diaphragmatic contraction during pacing. Noncapture and capture points were tagged on the electroanatomic mapping using different colors depending on RPN's threshold. RPN capture points were marked in the cardiac chambers adjacent to the merged RPN CT image, and capture threshold correlated visually with the actual distance between capture points and the RPN (Figure). In general, capture points at 10 mA were located ≤5 to 10 mm from the RPN, whereas capture points at 50 mA were located ≤10 to 20 mm from the RPN. Noncapture points at 50 mA were distant from the RPN. Ablations surrounding the right PVs were not performed where the RPN could be captured at any output. PV isolation using circumferential antral ablation was subsequently performed without complications. RPN injury is a highly symptomatic but uncommon complication of right superior PV isolation or superior vena cava ablation. …