Background and Purpose: Contrast-induced encephalopathy (CIE) is a rare and underrecognized complication after endovascular thrombectomy (EVT) for acute ischemic stroke. This study investigated the incidence and risk factors of CIE in patients who underwent EVT. Methods: Consecutive patients with acute ischemic stroke who received EVT between September 2014 and December 2019 at 2 medical centers were included. CIE was diagnosed on clinical criteria of neurological deterioration or delayed improvement within 24 hours after the procedure that was unexplained by the infarct or hemorrhagic transformation and radiological criterion of edematous change extending beyond the infarct core accompanied by contrast staining. Results: Of 421 patients with acute ischemic stroke who received EVT, 7 (1.7%) developed CIE. The manifestations included worsening of focal neurological signs, coma, and seizure. Patients with CIE were more likely to experience contrast-induced acute kidney injury than were those without CIE, but the volume of contrast medium was comparable between the two groups. The independent risk factors for CIE included renal dysfunction (defined as an estimated glomerular filtration rate <45 mL/min per 1.73 m ² ; odds ratio, 5.77 [95% CI, 1.37–24.3]; P =0.02) and history of stroke (odds ratio, 4.96 [95% CI, 1.15–21.3]; P =0.03). Patients with CIE were less likely to achieve favorable functional outcomes (odds ratio, 0.09 [95% CI, 0.01–0.87]; P =0.04). Conclusions: CIE should be suspected in patients with clinical worsening after EVT accompanied by imaging evidence of contrast staining and edematous changes, especially in patients with renal dysfunction or history of stroke.

Background and Purpose: We aim to investigate whether histopathologic examination of thrombi retrieved from acute ischemic stroke patients undergoing endovascular treatment could distinguish cancer-related stroke from other etiologies. Methods: Thrombi from patients undergoing endovascular treatment were analyzed. The etiology of stroke was divided into cardioembolism, large artery atherosclerosis, and active cancer groups. All selected thrombi were subjected to hematoxylin and eosin staining. The percentages of fibrin/platelets, red blood cells, and white blood cells within a thrombus were quantified. Results: One-hundred fifty-two patients (active cancer, 19; cardioembolism, 107; large artery atherosclerosis, 26) were included. Thrombi from the active cancer group exhibited a higher fibrin/platelet composition than did those from the cardioembolism and large artery atherosclerosis groups (median, 85.7% versus 43.9% and 42.5%; P <0.001). Fibrin/platelet composition was the only independent factor (odds ratio, 1.05 [95% CI, 1.02–1.08]) in differentiating cancer-related stroke from stroke caused by cardioembolism and large artery atherosclerosis. A fibrin/platelet proportion of ≥65% accurately predicted cancer-related stroke (area under the curve, 0.84; P <0.001). Conclusions: In thrombi retrieved from patients undergoing endovascular treatment, a high fibrin/platelet composition was a probable indicator of cancer-related stroke.

Timely intravenous (IV) thrombolysis for acute ischemic stroke is associated with better clinical outcomes. Acute stroke care implemented with “Stroke Code” (SC) may increase IV tissue plasminogen activator (tPA) administration. The present study aimed to investigate the impact of SC on thrombolysis. The study period was divided into the “pre-SC era” (January 2006 to July 2010) and “SC era” (August 2010 to July 2013). Demographics, critical times (stroke symptom onset, presentation to the emergency department, neuroimaging, thrombolysis), stroke severity, and clinical outcomes were recorded and compared between the two eras. During the study period, 5957 patients with acute ischemic stroke were admitted; of these, 1301 (21.8%) arrived at the emergency department within 3 h of stroke onset and 307 (5.2%) received IV-tPA. The number and frequency of IV-tPA treatments for patients with an onset-to-door time of <3 h increased from the pre-SC era (n = 91, 13.9%) to the SC era (n = 216, 33.3%) (P<0.001). SC also improved the efficiency of IV-tPA administration; the median door-to-needle time decreased (88 to 51 min, P<0.001) and the percentage of door-to-needle times ≤60 min increased (14.3% to 71.3%, P<0.001). The SC era group tended to have more patients with good outcome (modified Rankin Scale ≤2) at discharge (49.5 vs. 39.6%, P = 0.11), with no difference in symptomatic hemorrhage events or in-hospital mortality. The SC protocol increases the percentage of acute ischemic stroke patients receiving IV-tPA and decreases door-to-needle time.

About one-third of acute stroke patients may experience stroke-in-evolution, which is often associated with a worse outcome. Recently, we showed that multiscale entropy (MSE), a non-linear method for analysis of heart rate variability (HRV), is an early outcome predictor in non-atrial fibrillation (non-AF) stroke patients. We aimed to further investigate MSE as a predictor of SIE. We included 90 non-AF ischemic stroke patients admitted to the intensive care unit (ICU). Nineteen (21.1%) patients met the criteria of SIE, which was defined as an increase in the National Institutes of Health Stroke Scale score of ≥2 points within 3 days of admission. The MSE of HRV was analyzed from 1-hour continuous ECG signals during the first 24 hours of admission. The complexity index was defined as the area under the MSE curve. Compared with patients without SIE, those with SIE had a significantly lower complexity index value (21.3 ± 8.5 vs 26.5 ± 7.7, P = 0.012). After adjustment for clinical variables, patients with higher complexity index values were significantly less likely to have SIE (odds ratio = 0.897, 95% confidence interval 0.818–0.983, P = 0.020). In summary, early assessment of HRV by MSE can be a potential predictor of SIE in ICU-admitted non-AF ischemic stroke patients.

Background: D-dimer is a fibrin degradation product and a possible marker of thromboembolic events. The aim of this study was to investigate the relationship between D-dimer levels and outcome in acute ischemic stroke (AIS) patients receiving intravenous thrombolysis. Methods: This retrospective study included AIS patients who received intravenous thrombolysis with recombinant tissue plasminogen activator (rt-PA) and provided plasma D-dimer level within 24 h after stroke onset during 2009 and 2014 at a single medical center. Unfavorable outcome was defined as modified Rankin scale ≥3 at 3 months after stroke. Symptomatic intracerebral hemorrhage (ICH) was defined as a deterioration of at least 4 points on the National Institutes of Health Stroke Scale within 36 h post thrombolysis. Results: Of 347 patients receiving intravenous rt-PA, 159 (mean age 67.6 ± 13.1 year, 59.7% male) fulfilled the inclusion criteria. In univariate analysis, patients with unfavorable outcome (n = 79) had significantly higher levels of D-dimer than those with favorable outcome (median ln D-dimer = 1.4 vs. 0.7 μg/ml, p < 0.001). After adjustment for clinical variables, a higher level of D-dimer remained significantly associated with an unfavorable outcome (OR 1.90, 95% CI 1.27-2.86, p = 0.002) and the occurrence of symptomatic ICH (OR 2.97, 95% CI 1.15-7.70, p = 0.025). Conclusion: The D-dimer level within 24 h after stroke onset can be an early outcome indicator in AIS patients receiving rt-PA therapy.

Background: Conventional prognostic scores usually require predefined clinical variables to predict outcome. The advancement of natural language processing has made it feasible to derive meaning from unstructured data. We aimed to test whether using unstructured text in electronic health records can improve the prediction of functional outcome after acute ischemic stroke. Methods and Results: Patients hospitalized for acute ischemic stroke were identified from 2 hospital stroke registries (3847 and 2668 patients, respectively). Prediction models developed using the first cohort were externally validated using the second cohort, and vice versa. Free text in the history of present illness and computed tomography reports was used to build machine learning models using natural language processing to predict poor functional outcome at 90 days poststroke. Four conventional prognostic models were used as baseline models. The area under the receiver operating characteristic curves of the model using history of present illness in the internal and external validation sets were 0.820 and 0.792, respectively, which were comparable to the National Institutes of Health Stroke Scale score (0.811 and 0.807). The model using computed tomography reports achieved area under the receiver operating characteristic curves of 0.758 and 0.658. Adding information from clinical text significantly improved the predictive performance of each baseline model in terms of area under the receiver operating characteristic curves, net reclassification improvement, and integrated discrimination improvement indices (all P <0.001). Swapping the study cohorts led to similar results. Conclusions: By using natural language processing, unstructured text in electronic health records can provide an alternative tool for stroke prognostication, and even enhance the performance of existing prognostic scores.