Ou Hongru, Wu Huanhua, Wu Shuolong, Cao Qian, Cao Xiaozheng, Zhang Guanye, Mo Jiabin, Hu Youzhu
Department of Radiology, The Affiliated Shunde Hospital of Jinan University, Foshan, Guangdong, P. R. China.
Central Laboratory, The Affiliated Shunde Hospital of Jinan University, Foshan, Guangdong, P. R. China.
Brain Behav. 2025 May;15(5):e70544. doi: 10.1002/brb3.70544.
The outcomes of endovascular reperfusion in acute large vessel occlusion stroke (ALVOS) vary, with some patients recovering fully while others face disability or mortality despite recanalization. Alberta stroke program early CT score-net water uptake (ASPECTS-NWU), a quantitative imaging metric assessing tissue edema and infarct progression, may improve prognostic accuracy.
This study included 96 ALVOS patients between December 2020 and March 2024. Patients were categorized into good (modified Rankin Scale [mRS] 0-2) and poor (mRS 3-6) prognosis groups based on 90-day mRS outcomes. Feature selection using Least Absolute Shrinkage and Selection Operator (LASSO), Boruta, and logistic regression (LR) identified key predictors, including LVO, Alberta Stroke Program Early CT Score (ASPECTS), ASPECTS from the follow-up CT (ASPECTSFCT), and National Institutes of Health Stroke Scale (NIHSS) scores. Predictive performance was validated with cross-validation, and model calibration was assessed via calibration curves, receiver operating characteristic (ROC) curves, area under the curve (AUC), and the Spiegelhalter Z-test (significance set at p < 0.05).
LR highlighted LVO, ASPECTS, ASPECTSFCT, and NIHSS as significant predictors of poor prognosis. The constructed nomogram enables individualized risk assessment, with total points correlating to poor outcome probability. ROC analysis showed good discriminatory ability in the training set (AUC 0.815, 95% confidence interval [CI]: 0.714-0.916) but moderate performance in the test set (AUC 0.688, 95% CI: 0.484-0.891). Calibration was strong in the training set (Spiegelhalter Z < 0.0001) but showed minor issues in the test set (Spiegelhalter Z = 1.222).
This study highlights the prognostic value of ASPECTS-NWU in ALVOS and its integration into a predictive nomogram for individualized risk assessment. By refining ischemic injury stratification, ASPECTS-NWU can guide therapeutic decisions, optimize post-reperfusion management, and improve patient outcomes.
急性大血管闭塞性卒中（ALVOS）血管内再灌注的结果各不相同，一些患者完全康复，而另一些患者尽管血管再通仍面临残疾或死亡。阿尔伯塔卒中项目早期CT评分-净水摄取量（ASPECTS-NWU）是一种评估组织水肿和梗死进展的定量成像指标，可能会提高预后准确性。
本研究纳入了2020年12月至2024年3月期间的96例ALVOS患者。根据90天改良Rankin量表（mRS）结果，将患者分为预后良好（mRS 0-2）和预后不良（mRS 3-6）两组。使用最小绝对收缩和选择算子（LASSO）、Boruta和逻辑回归（LR）进行特征选择，确定了关键预测因素，包括大血管闭塞（LVO）、阿尔伯塔卒中项目早期CT评分（ASPECTS）、随访CT的ASPECTS（ASPECTSFCT）和美国国立卫生研究院卒中量表（NIHSS）评分。通过交叉验证验证预测性能，并通过校准曲线、受试者工作特征（ROC）曲线、曲线下面积（AUC）和Spiegelhalter Z检验（显著性设定为p < 0.05）评估模型校准。
LR强调LVO、ASPECTS、ASPECTSFCT和NIHSS是预后不良的重要预测因素。构建的列线图能够进行个体化风险评估，总分与不良结局概率相关。ROC分析显示，在训练集中具有良好的区分能力（AUC 0.815，95%置信区间[CI]：0.714-0.916），但在测试集中表现中等（AUC 0.688，95%CI：0.484-0.891）。训练集中校准良好（Spiegelhalter Z < 0.0001），但在测试集中存在一些小问题（Spiegelhalter Z = 1.222）。
本研究强调了ASPECTS-NWU在ALVOS中的预后价值及其纳入预测列线图以进行个体化风险评估。通过完善缺血性损伤分层，ASPECTS-NWU可以指导治疗决策，优化再灌注后管理，并改善患者结局。
