A Predictive Nomogram Model Incorporating Contrast Extravasation for Early Prediction of Postthrombectomy Malignant Cerebral Edema.

Abstract

Early prediction of malignant cerebral edema (MCE) following endovascular thrombectomy (EVT) is critical for guiding timely interventions to improve outcomes; however, existing prediction models predominantly require complex algorithms. Contrast extravasation (CE) is a novel imaging biomarker. We aimed to develop a practical nomogram incorporating CE for predicting MCE to enhance postoperative management. We reviewed the records of individuals with anterior circulation acute ischemic stroke who exhibited CE on dual-energy computed tomography after successful recanalization via EVT. The prediction nomogram was developed via multivariable logistic regression and internally validated with the Hosmer-Lemeshow test. The final cohort comprised 89 patients (median age, 72 [interquartile range, 67-79] years; male, 62.92%), with 35 (39.33%) patients developing MCE. After adjusting for confounding variables, the CE score in the Alberta Stroke Program Early CT Score region (CE-ASPECTS) retained independent predictive significance for MCE. It was incorporated into the nomogram with variables according to clinical relevance. The nomogram demonstrated excellent discriminative performance, with an area under the curve of 0.961 (95% confidence interval: 0.927-0.995), and good calibration accuracy according to the Hosmer-Lemeshow test (P = 0.869). CE-ASPECTS is an accessible and independent predictor of MCE. The nomogram, which is composed of age, admission fasting blood glucose level, Trial of ORG 10,172 in Acute Stroke Treatment classification, occlusion site, and CE-ASPECTS, may serve as a practical tool for predicting the probability of MCE in patients with acute ischemic stroke who achieved successful recanalization after EVT.