Abstract
BACKGROUND: Cardiac dysfunction and altered serum biomarker profiles may be associated with left atrioventricular uncoupling (LAU) in patients with acute ischemic stroke (AIS). This study sought to: (I) determine the key independent predictors from these cardiovascular and hematological parameters; and (II) systematically compare the predictive ability and cost-effectiveness in LAU risk stratification. METHODS: A total of 207 patients with AIS who underwent real-time four-dimensional (4D) echocardiography and blood biomarker assay at admission were included in this study. LAU was diagnosed via 4D echocardiography when the left atrioventricular coupling index (LACI) was greater than 0.25. Multivariate logistic regression models were employed to identify the factors independently associated with LAU. A decision tree model was used to compare the cost-effectiveness of these factors. RESULTS: LAU was found in 146 patients with AIS. These patients had higher levels of brain natriuretic peptide (BNP) and greater left atrial (LA) volume, stiffness, and function impairment than did those without LAU (P<0.05). According to multivariate logistic regression analysis, the two independent risk factors for LAU in patients with AIS were the LA stiffness index (LASI) [odds ratio (OR) =822.286; 95% confidence interval (CI): 25.487-182,900.741; P<0.0001] and BNP (OR =1.115; 95% CI: 0.998-1.246; P=0.031). In the prediction of LAU, the performance of LASI [area under the receiver operating characteristic curve (AUC) =0.772; sensitivity, 74.15%; specificity, 73.02%; cutoff value, 0.34] was superior to that of BNP (AUC =0.673; sensitivity, 70.55%; specificity, 60.66%; cutoff value, 39 pg/mL) (P<0.05). Moreover, the LASI echocardiographic measurement was a cost-effective tool for predicting LAU in patients with AIS. CONCLUSIONS: Both LASI and BNP demonstrated independent value for predicting LAU in the AIS cohort. LASI echocardiographic measurement can serve as a cost-effective approach for detecting LAU.