Identification of left atrial fibrosis with a late-enhancement MR sequence (LE-MR): preliminary results

This study was done to identify left atrial fibrosis in a group of consecutive patients with atrial fibrillation (AF) candidate for percutaneous radiofrequency catheter ablation (RFCA) by using a late-enhancement magnetic resonance (LE-MR) sequence, and to validate the technique by comparison with electroanatomical mapping (EAM).

We enrolled 37 patients (29 males; mean age, 61 years) candidate for percutaneous RFCA of AF, who were studied with LE-MR and EAM. To identify left AF we used a three-dimensional LE sequence with cardiac gating and respiratory navigator. The EAM study involved the acquisition of 200 points in the left atrium (LA). The LA was divided into seven segments (pulmonary vein antra, floor, anterior wall, posterior wall-roof). Two blinded radiologists assessed the presence of fibrosis (area of hyperintense signal), reaching a consensus in discordant cases. Inter-observer variability was also evaluated to estimate the reproducibility of the method. We analysed the anatomical agreement between the results obtained with LE-MR imaging and EAM.

Five patients were excluded because of to poor image quality. As for the other 32 patients, inter-observer agreement was good [Cohen's kappa κ = 0.72 with 95 % confidence interval (CI) of 0.55, 0.89]. In the classification of LA segments affected by fibrosis, LE-MR had a sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy of 66 % (95 % CI 53.7 %, 77.2 %); 87 % (95 % CI 80.9 %, 91.9 %); 69 % (95 % CI 56.5 %, 80.1 %); 85.5 % (95 % CI 79.1 %, 90.6 %) and 81 % (95 % CU 75.1 %, 85.7 %).

Despite the small size of the sample studied, the LE-MR sequence proved more useful for excluding the presence of AF than for confirming its existence. Identification of AF prior to RFCA is paramount to select those patients who are truly amenable to the ablation procedure, which is expensive and not entirely free of risks.