摘要： Predictive value of left atrial emptying fraction and late gadolinium enhancement for the clinical outcomes in atrial fibrillation patients undergoing ablation. Background: A dimension, volume and function of left atrium (LA) and late gadolinium enhancement (LGE) of the LA are known as a prognostic factor for the clinical outcome of atrial fibrillation (AF) ablation. However, which of those parameters provide a best predictive power remains unclear. Purpose: We aimed to compare the efficiency of prediction among those parameters using echocardiography and cardiac magnetic resonance (CMR). Methods: Eighty-eight consecutive patients (67 males; 56±10 years old; 45 persistent AF) underwent transthoracic echocardiography (TTE) and CMR was performed one day before AF ablation. LA anterior-posterior dimension (LAD) was measured at the phase of maximal LA size on M-mode of TTE. Maximal and minimal volume indexed to body surface area (LVAi) from CMR and Area Length Method ALM. Phasic volumes were used to calculate LA emptying fraction (LAEF) ([maximum-minimum LAV]/maximum LAVx100). LGE was defined as areas with thresholds of 6-SD above mean signal in unenhanced LV wall. The LGE stage was defined based on delayed enhancement (DE) volume area divided by LA volume (Stage 1: ≤5%, Stage 2: ≤5%, Stage 3: ≤25%, Stage 4: ≤35%). Results: During mean follow-up duration: 10±3 months, AT/AF recurred in 19/88 patients (21.5%; 3/43 in PAF; 16/45 in PeAF). Compared to patients without recurrence, those with recurrence had larger maximal/minimal LAVi (maximal: 54±21 VS. 45±16, p=0.07; minimal: 40±22 VS. 26±14, p=0.02) and LAD (45.5±6.5 VS. 40.1±5.8, p=0.001) and lower LAEF (24±10 VS. 45±16, P=0.001). ROC analysis for recurrence showed that LAEF had the largest area under curve (AUC =0.86, p=0.0001) compared to maximal LAVi, minimal LAVi and LAD. LGE stage was insignificantly higher in patients with recurrence (3.6±0.7 VS. 2.9±0.9, p=0.46). Cut-off value of LAEF for freedom from recurrence was 31.0% (sensitivity = 0.82, specificity = 0.78, p=0.001). In case of PeAF, only lower LAEF (23±11 VS. 40±7, p=0.001) and larger LAD (45.5±7.1 VS. 40.5±6.3, p=0.02) were significantly related with recurrence. Adjusting clinical risk factor, Cox-regression analysis showed that LAEF was only independent predictor for freedom from recurrence (HR = 0.88; 95% CI; 0.82–0.95; p=0.001). Conclusion: LAEF measured by CMR was superior to maximal/minimal LAVi, LAD and LGE stage in predicting the AT/AF recurrence after AF ablation. Moreover, patients with PeAF and LAEF>31% had excellent clinical outcome.