Abstract T P178: Echocardiographic Left Atrial Strain Analysis Plus NT-ProBNP to Determine Cardioembolic Strokes

Abstract

Background: Undetected paroxistic atrial fibrillation (PAF) is the main etiology of undetermined ischemic strokes. The left atria functionality assessed by echocardiography Strain analysis (LA-S) is lower in patients with PAF. N-terminal pro-brain natriuretic peptide (NT-ProBNP) has been described as a marker of cardioembolic etiology. We describe the relationship between these surrogates in order to improve the detection of strokes related to PAF. Methods: Consecutive undetermined acute stroke patients underwent specific echocardiography study to determine the LA-S by speckle tracking analysis. All patients were monitored by implantable Reveal TX Holter during three years. In those who were not detected PAF (NO PAF GROUP, n=20) were compared with PAF detected patients. PAF GROUP were defined as those PAF detected patients (n=2) plus acute stroke patients with previously known PAF (n=41). A LA-S cut off point of -3.62% was defined to predict the sensitivity and specificity of PAF detection. NT-proBNP was measured in a group of 25 non AF patients (pro-BNP group) with suspected cardioembolic stroke and these values were compared between patients with predefined high (>3.62%) and low (<3.62%) LA-S. Results: There were no differences in the baseline variables (PAF group vs NO PAF group). PAF group showed a percentage of LA -S significantly lower in the anterior- posterior walls (-2% +/- 1.76 vs -6.9+/-7.45) (p=0.035). LA -S cut off value of -3.62% predicted PAF with a sensitivity of 77% and specificity of 80%. According pro-BNP group those patients with low SL-LA had a trend to a higher NT-ProBNP value compared with high SL-LA group (685 vs 293 pg/mL; p=0.07). There were no differences in the size of the atria between the groups. Conclusions: Measurement of LA-S and NT-ProBNP in patients with undetermined stroke could be a useful tool to select those patients that could benefit from long-term cardiac monitoring in order to diagnose PAF.