AUTHOR=Zhang Zhen , Yi Lirong , Xiang Bo , Zhou Jie 

TITLE=Analysis of 90-day risk factors for poor prognosis in patients with postoperative CT with hyperdense signs after thrombectomy for acute ischemic stroke

JOURNAL=Frontiers in Neurology

VOLUME=Volume 16 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2025.1609722

DOI=10.3389/fneur.2025.1609722

ISSN=1664-2295

ABSTRACT=ObjectiveTo investigate the risk factors associated with poor 90-day prognosis in patients exhibiting hyperdense signs on the initial CT scan after mechanical thrombectomy for acute ischemic stroke.MethodsWe conducted a retrospective analysis of 96 patients with acute ischemic stroke who underwent mechanical thrombectomy at The Affiliated Yongchuan Hospital of Chongqing Medical University. The patients were divided into a training set (n = 55, from August 2020 to March 2022) and a validation set (n = 41, from November 2022 to December 2023). Based on the 90d mRS scores, patients were categorized into a good prognosis group (n = 46) and a poor prognosis group (n = 50). Clinical and imaging data were compared between the two groups. A prediction model was constructed using the logistic regression algorithm and validated using the temporal validation method. Receiver operating characteristic (ROC) curves were generated, and the area under the curve (AUC) was calculated to evaluate the discriminative ability of the model. The Hosmer–Lemeshow goodness-of-fit test was used to assess calibration, and decision curve analysis was performed to evaluate the clinical net benefit of each model across various threshold probabilities.ResultsLogistic regression analysis identified the following risk factors for poor 90-day prognosis: D-dimer level, admission diastolic blood pressure, preoperative NIHSS score, vascular occlusion site, as well as the volume of hyperdense areas and the total volume of both hyperdense and hypodense areas on the first postoperative CT scan. Two models were developed: one based solely on an imaging indicator (total volume) and another incorporating combined indicators (total volume, D-dimer, admission diastolic blood pressure, preoperative NIHSS score, and vascular occlusion site). The combined-indicator model demonstrated superior performance. In the training set, it achieved an AUC of 0.886 (95% CI: 0.801–0.971, p < 0.001), accuracy of 0.818, sensitivity of 0.818, and specificity of 0.818. In the test set, the AUC was 0.848 (95% CI: 0.730–0.966, p < 0.001), with an accuracy of 0.707, sensitivity of 0.542, and specificity of 0.941. Decision curve analysis confirmed that two models maintained a positive clinical net benefit within a wide range of threshold probability values (10%–90%).ConclusionPatients with hyperdense signs on the first postoperative CT scan exhibit distinct risk factors for poor 90-day prognosis. Combining imaging features with clinical indicators significantly improves the predictive value for 90-day outcome after mechanical thrombectomy.