Background

Thromboembolic complications in atrial fibrillation ablation due to mobilization of preexisting thrombus formation (TF) in the left atrium (LA) are devastating. Gold standard to exclude LA TF is a transesophageal echocardiography (TEE). Aim of this study was to compare sensitivity and specificity of a dual source cardiac computed tomography (DS-CT) with TEE for TF exclusion prior to ablation (Figure 1). In addition,  CT protocols with and without ECG triggering were compared. 

Methods

We studied 622 patients that underwent AF ablation. In all patients, a DS-CT as well as an TOE was performed <48h prior to ablation to exclude TF. Mean age of the patients was 60 +/-10 years (69% male, 61% paroxysmal AF). During the DS-CT, 280 patients (50%) were in AF. An ECG triggered DS-CT was performed in 332 patients compared to 290 patients that underwent DS-CT without ECG triggering. 

Results 

In all patients without suspected TF on DS-CT (n=552; 88.7%), no thrombus was found on  TEE. In patients with suspected TF on DS-CT (n=70), only three patients showed TF on TEE whereas no TF was detected in 67 patients on TEE. Overall, sensitivity for TF detection in DS-CT was 100%, specificity was 89.2% (positive predictive value: 4.3%, negative predictive value: 100%). The CT protocol (triggered vs. non- triggered) had no significant influence on diagnostic accuracy. Radiation exposure was 3.91 ± 4.02 mSv (triggered) versus 1.52 ± 1.02 mSv (non- triggered) ; p<0.0001. 

Discussion 

DS-CT is a highly sensitive method for LA thrombus detection in patients undergoing LA ablation and delivers additional anatomic details of pulmonary veins and LA anatomy with an acceptable radiation exposure (Figure 2). The moderate specifity could probably be optimized by delayed LAA acquisition. Non ECG triggered DS-CT showed a significantly lower radiation exposure whereas diagnostic accuracy was comparable. DS-CT might therefore be used as primary method to exclude LA TF in patients undergoing LA ablation.