Background Symptomatic intracranial hemorrhage (sICH) is a common and severe complication in patients with acute ischemic stroke (AIS) after treatment with thrombectomy. Objective To explore the ability of admission plasma D-dimer levels to predict sICH after thrombectomy. Methods Between February 2018 and August 2021, consecutive patients with AIS who underwent thrombectomy at our single comprehensive stroke center were retrospectively enrolled. sICH was defined according to the criterion of the Heidelberg Bleeding Classification. Logistic regression analysis was performed to determine the risk factors of sICH. The overall discriminative ability of D-dimer levels in predicting sICH was evaluated by adopting a receiver operating characteristic (ROC) curve. Results Of the 395 enrolled patients, 48 (12.2%) had sICH. Patients with sICH were older (72.9 vs 69.3 years, P=0.037), more often female (62.5% vs 45.5%, P=0.027), had higher D-dimer levels (2.70 vs 0.74 mg/L, P<0.001), higher National Institutes of Health Stroke Scale score (20 vs 15, P<0.001), lower Alberta Stroke Program Early CT Score (8 vs 9, P<0.001), a higher proportion of internal carotid artery occlusions (56.2% vs 30.3%, P<0.001), and less commonly had large-artery atherosclerosis stroke etiology (12.5% vs 32.3%, P=0.010) than patients without sICH. After adjustment for potential confounders, D-dimer levels (adjusted OR=2.45, 95% CI 1.75 to 3.43, P<0.001) remained significantly associated with sICH. Based on the ROC, the D-dimer as a predictor for predicting sICH, presented with a specificity of 86.2%, a negative predictive value of 94.6%, and an area under the curve of 0.774. Conclusion Elevated admission D-dimer levels are an independent predictor of sICH in patients with AIS after thrombectomy.
Background: Flat-panel computed tomography (CT) is an available imaging modality immediately after endovascular thrombectomy without transferring patients to the CT room. Purpose: To determine the accuracy of flat-panel CT scans in differentiating hemorrhagic transformation (HT) from contrast exudation after thrombectomy in patients with acute ischemic stroke (AIS). Material and Methods: From January 2019 to December 2021, consecutive patients with AIS who received an immediate flat-panel CT scan and follow-up neuroimaging after thrombectomy were enrolled in our study. The receiver operating characteristic curve was adopted to assess the discriminating accuracy of characteristics of flat-panel CT for HT. Results: A total of 108 patients were enrolled in the study; 58 (53.7%) patients presented with hyperdense lesions on flat-panel CT. Patients with hyperdense lesions experienced a higher proportion of HT than patients without (58.7% vs. 10.0%; P < 0.001). Among all patients with hyperdensity on flat-panel CT, patients who experienced HT had higher average Hounsfield units (HUavg) (125 vs. 93; P = 0.001) and a higher proportion of mass effect (67.6 vs. 12.5; P < 0.001). The flat-panel CT differentiating HT from contrast exudation yielded a sensitivity of 87.2% and a negative predictive value of 90.0%. The area under the curve of HUavg, mass effect, and combination for differentiation of HT were 0.74, 0.78, and 0.83, respectively. Conclusion: The hyperdensity on immediately post-thrombectomy flat-panel CT could differentiate HT from contrast exudation with an excellent negative predictive value. The ability of flat-panel CT in differentiating HT from contrast exudation was improved when combined with HUavg and mass effect.
Background: Intracranial atherosclerosis-related large vessel occlusion (ICAS+LVO) poses an important technical challenge for endovascular thrombectomy (EVT). Purpose: To evaluate the value of D-dimer in predicting ICAS+LVO alone and in combination with other clinical and imaging predictors. Material and Methods: Consecutive patients who underwent EVT at our center between January 2018 and June 2021 were retrospectively reviewed. Patients were classified to the ICAS+LVO or ICAS–LVO group according to angiographic findings. Collateral gradings were evaluated based on computed tomography angiography and categorized as follows: score 0–1 unfavorable collaterals and score 2–3 favorable collaterals. Receiver operating characteristic curve was analyzed to evaluate the predictive value of D-dimer and the combination of other predictors for ICAS+LVO. Results: A total of 374 patients were enrolled, among them, 107 (28.6%) had an ICAS+LVO, while ICAS–LVO was determined in 267 (71.4%) patients. Median D-dimer levels were lower (0.36 vs. 1.18 mg/L; P < 0.001) while the proportion of favorable collaterals was higher (85.0% vs. 22.5%; P < 0.001) in the ICAS+LVO group than the ICAS–LVO group. After multivariable analysis, D-dimer (adjusted odds ratio [OR]=0.32, 95% confidence interval [CI]=0.21–0.50; P < 0.001) and collaterals (adjusted OR=16.25, 95% CI=7.58–34.84; P < 0.001) remained independent predictors of ICAS+LVO. The area under the curve of D-dimer, collaterals, and combination for identification of ICAS+LVO was 0.82, 0.85, and 0.92, respectively. Conclusion: Low early plasma D-dimer levels are a significant and independent predictor of ICAS+LVO, and predictive value strengthens when in a combined model using D-dimer and collateral grading.