Abstract 686: Utility of Transesophageal Echocardiography versus Computed Tomography in Detection of Intracardiac Thrombus in Patients Undergoing Catheter Ablation of Atrial Fibrillation
Anatomic rendering of the left atrium and pulmonary veins and exclusion of intra-cardiac thrombus (ICT) prior to catheter ablation for AF (CAAF) is important yet the most appropriate imaging modality remains unclear. Whether, computed tomography (CT) is similar to transesophageal echocardiography (TEE) in sensitivity and specificity and the relative predictive value of ECG-gated (ECGCT) versus ungated (N-ECGCT) CT is unknown. Therefore, we sought to compare sensitivity and specificity of gated and ungated CT with TEE in detection of ICT in patients (pts) undergoing CAAF. All pts who underwent ECGCT or N-ECGCT and TEE prior to CAAF at Mayo Clinic between March 2004 and October 2007 were included. All data were obtained prospectively and reviewed independently. Image data were independently reviewed by two cardiac radiologists blinded to TEE findings with the left atrial appendage (LAA) defined as normal (fully opacified) or abnormal (underfilled). A total of 657 pts (mean age 57 ± 10 yrs; 77% male; EF 57 ± 9%) were included. These included 255 ECGCT and 402 N-ECGCT scans. Three pts were excluded due to a prior LAA ligation. A diagnosis of possible ICT on the basis of abnormal LAA anatomy was made by CT in 73 pts (11.1%) of which 13 patients were verified to have an LAA clot on TEE. All pts with clot on TEE had an abnormal CT. Sensitivity and specificity for CT overall was 100% and 91%, respectively. Sensitivity and specificity for LAA clot was 100% and 92%, respectively, for N-ECGCT and 100% and 88%, respectively, for ECGCT. All pts with confirmed thrombus were therapeutically anti-coagulated with warfarin at time of image acquisition. CHADS2 score was higher (1.1 versus 0.9) in patients with LAA filling defects versus those without LAA filling defect. No cases of LAA were observed in pts aged <52 years with CHADS score <1. In patients referred for CAAF, ICT is uncommon, especially in pts < 50 yrs with few or no risk factors for stroke. Use of warfarin was not necessarily associated with “no risk” of ICT. CT provides excellent accuracy for ruling out thrombus in the LAA. N-ECGCT has superior specificity when compared to ECGCT for detection of thrombus.