P3774Prediction of left atrial appendage thrombosis in nonvalvular atrial fibrillation by means of predictors not related to CHA2DS2-VASc score

Background Nonvalvular atrial fibrillation (AF) is the most common sustained cardiac dysrhythmia and the most frequent cause of cardio-embolic stroke. It is well known that left atrial appendage thrombi (LAAT) are the source of most embolic strokes in patients with AF, and can be easily identified by transesophageal echocardiography (TEE). Purpose To determine additional LAAT predictors not included in the CHA2DS2-VASc score in patients with nonvalvular AF. Methods Retrospective study enrolled 636 patients with nonvalvular AF (400 males) aged 24–84 years (mean age 57.8±9.1), admitted to our Research Center in 2014–2017 for catheter ablation or electrical cardioversion. All patients had scheduled transthoracic echocardiography (TTE), as well as TEE performed to exclude LAAT. Four forms of cardiac geometry were revealed with the help of TTE according to Recommendations of ASE and EACI: normal geometry, concentric remodeling, eccentric hypertrophy and concentric hypertrophy. Results According to TEE results, LAAT (6.6%) was detected in 42 patients from 636. Patients with LAAT more often had persistent and permanent AF (assigned as “AF stability”), had bigger sizes and volumes of left and right cardiac chambers, lower left ventricular ejection fraction, more expressed LV hypertrophy and lower blood flow velocity in the LA appendage compared to patients without LAAT. Remodeling types also varied: less LAAT patients had normal cardiac geometry and more LAAT patients had LV eccentric hypertrophy. To analyze factors associated with AF, we used a multivariable logistic regression model, involving the potential independent, clinically relevant variables and echocardiographic parameters. Logistic regression analysis identified the latter three as independent predictors for LAAT (Table 1). According to ROC-analysis the quality of the received model was assessed as good: AUC=0.763 (p<0.01), sensitivity – 75.8%, specificity – 72.2%. Table 1 Predictors B Wald statistics χ2 P OR 95% CI AF stability 0.913 4.143 0.042 2.491 1.034–6.000 LA diameter, mm 0.149 8.684 0.003 1.160 1.051–1.281 Eccentric hypertrophy 1.440 7.411 0.006 4.222 1.497–11.908 Constant −10.613 24.088 0.000 Conclusion From risk factors, not included in the CHA2DS2-VASc score, in addition to such predictable values as AF stability and left atrial diameter, such predictor as eccentric left ventricular hypertrophy was revealed, which in our study was associated with more than a 4-fold increase in the risk of LAAT.