No abstract

Background: Cardioembolic (CE) stroke carries significant morbidity and mortality. Current risk stratification tools such as CHADS2 score do not include any imaging parameters and are based on clinical features, which have limitations. Left atrial (LA) enlargement and remodeling may be associated with CE risk due to predisposition for atrial arrhythmias and thrombus formation. Left atrial cross sectional area (LACSA), a novel echo measure which reflects both LA size and shape, may improve CE stroke risk assessment. Aim: This study examined the value of LACSA in predicting CE stroke risk and the improvement in risk prediction when added to CHADS2 score. Methods: Clinical and echo parameters were examined in a prospective cohort of 1275 consecutive patients with ischemic stroke. Strokes were classified using the Causative Classification of Strokes and 259 (20%) were classified as CE stroke. LACSA was calculated using the formula: π/4*largest measured LA diameter*smallest measured LA diameter where mid LA diameter was measured in the parasternal long axis, 4 chamber and 2 chamber views. Results: Patients with CE stroke had greater LACSA (8.6 ± 2.3 vs 6.4 ± 1.8 cm2/m2; p<0.001) and mean CHADS2 score (2.25 ± 1.28 vs 1.87 ± 1.40; p<0.0001) compared to non-CE stroke patients. LACSA was independently associated with CE strokes (OR 1.21; 95% CI 1.08-1.34; p=0.001) in a multivariable model adjusted for CHADS2, gender, score, BMI, atrial fibrillation, anti-platelet and anti-coagulant use, E/E’ and LVEF. The addition of LACSA to CHADS2 score improved the prediction of CE stroke (c-statistic for predicting CE stroke using CHADS2 alone was 0.59 (95% CI 0.55-0.63) vs CHADS2 and LACSA 0.78 (95% CI 0.72-0.80) (p<0.001). Conclusion: LACSA is a novel measure of LA remodeling and associated with CE stroke. LACSA, an imaging parameter, enhances the risk prediction of the CHADS2 score, a clinical measure of risk, improving risk stratification for CE stroke and impacting therapeutic strategies.