Correlation between left atrial spontaneous echocardiographic contrast and 5-year stroke/death in patients with non-valvular atrial fibrillation.

BACKGROUND: Transthoracic echocardiography (TTE) and transoesophageal echocardiography (TOE) can be used to detect the presence of left atrial thrombus and left atrial spontaneous echocardiographic contrast (LASEC). AIM: To evaluate the prognostic value of TTE and TOE in predicting stroke and all-cause death at 5-year follow-up in patients with non-valvular atrial fibrillation (NVAF). METHODS: This study included patients hospitalised with electrocardiography-diagnosed NVAF in Saint-Antoine University Hospital, Paris, between July 1998 and December 2011, who underwent TTE and TOE evaluation within 24hours of admission. Cox proportional-hazards models were used to identify predictors of the composite outcome (stroke or all-cause death). RESULTS: During 5 years of follow-up, stroke/death occurred in 185/903 patients (20.5%). By multivariable analysis, independent predictors of stroke/death were CHA2DS2-VASc score (hazard ratio [HR] 1.35, 95% confidence interval [CI] 1.25-1.47; P<0.001), left atrial area>20 cm2 (HR 1.59, 95% CI 1.08-2.35; P=0.018), moderate LASEC (HR 1.72, 95% CI 1.13-2.62; P=0.012) and severe LASEC (HR 2.04, 95% CI 1.16-3.58; P=0.013). Independent protective predictors were dyslipidaemia (HR 0.60, 95% CI 0.43-0.83; P=0.002) and discharge prescription of anti-arrhythmics (HR 0.59, 95% CI 0.40-0.87; P=0.008). Adding LASEC to the CHA2DS2-VASc score modestly improved predictive accuracy and risk classification, with a C index of 0.71 vs. 0.69 (P=0.004). CONCLUSIONS: In this retrospective monocentric study, the presence of moderate/severe LASEC was an independent predictor of stroke/death at 5-year follow-up in patients with NVAF. The inclusion of LASEC in stroke risk scores could modestly improve risk stratification.