Hypercoagulation on Thromboelastography Predicts Early Neurological Deterioration in Patients with Acute Ischemic Stroke

Background: Thromboelastography (TEG) provides an integrated measurement of blood coagulation function and has been reported to be a useful tool for predicting clinical outcomes in patients with cardiovascular diseases. We aimed to investigate the application of TEG on admission for predicting early neurological deterioration (END) in patients with acute ischemic stroke and its potential correlation with the evolution of ischemic lesions. Methods: Among patients consecutively admitted between January 1, 2016, and September 31, 2017, those presenting with mild and moderate acute ischemic stroke (National Institutes of Health Stroke Scale [NIHSS] score 14) within 24 h of stroke onset were identified and included in this study. TEG was performed on the first day of admission. END was defined as an increase of 1 on subitems of the NIHSS or the emergence of new symptoms within 72 h of admission. Demographics, lab test results, and TEG values were compared according to whether END occurred. A multiple logistic regression model was then developed to investigate the predictive power of TEG for END. Receiver operating characteristic (ROC) curves were then plotted to evaluate the optimal cutoff values. Results: Of the 246 eligible patients (mean age 65.3 +/- 12.9 years, 73.6% male), END was identified in 72 (29.3%) patients. Patients with END corresponded to a higher proportion of females, a more prevalent history of diabetes mellitus (DM), higher baseline NIHSS scores, higher serum high-sensitivity C-reactive protein (hsCRP) levels, and significantly shorter R on TEG (4.0 +/- 1.0 vs. 4.7 +/- 1.2 min, p< 0.001). In further comparisons stratified by R tertiles, significant trends were found between shorter R and being female and older and being more likely to exhibit diffusion weighted imaging progression on follow-up MRI. After adjusting for female sex, baseline NIHSS score, DM, and hsCRP, the lower tertile of R (R 3.8 min) was strongly associated with END (OR 3.556, 95% CI 1.165-10.856, p< 0.001). ROC analysis demonstrated that R 3.45 min had the best predictive value for END with 87.9% sensitivity and 40.3% specificity. Conclusion: Decreased R time on admission TEG is associated with END within 3 days in patients with acute ischemic stroke.