Data Sheet 1_Intracranial hemorrhage prediction in acute ischemic stroke patients with anterior circulation tandem lesions following endovascular thrombectomy.docx

BackgroundAcute ischemic stroke (AIS) patients with anterior circulation tandem lesions (TL) face a heightened risk of hemorrhage following endovascular thrombectomy (EVT). Predictive models specifically for this complication in the TL population are currently lacking.
MethodsThis retrospective cohort study analyzed 200 AIS patients with anterior circulation TL who underwent EVT. Least Absolute Shrinkage and Selection Operator regression was used for feature selection. Multivariable logistic regression (LR) models predicting intracranial hemorrhage (ICH) and symptomatic intracranial hemorrhage (sICH) risk were developed and visualized as nomograms. Model discrimination was evaluated using the area under the receiver operating characteristic curve (AUC).
ResultsAfter EVT, ICH occurred in 92 patients (46%) and sICH in 24 patients (12%). The LR model for ICH identified diabetes [odd ratio (OR) 2.454, 95% CI 1.137–5.297], drinking history (OR 2.230, 95% CI 1.160–4.288), and lower modified Thrombolysis in Cerebral Infarction (mTICI) score (OR 0.547, 95% CI 0.311–0.961) as significant independent predictors (AUC = 0.712). The LR model for sICH identified lower Glasgow Coma Scale (GCS) score (OR 0.820, 95% CI 0.695–0.968), lower mTICI score (OR 0.398, 95% CI 0.182–0.868), and lower Alberta Stroke Program Early CT Score (ASPECTS) (OR 0.795, 95% CI 0.641–0.984) as significant independent predictors (AUC = 0.830). Nomograms effectively quantified the contribution of predictors to outcome probabilities.
ConclusionIn AIS patients with anterior circulation TL undergoing EVT, diabetes, drinking history, and lower mTICI score independently predict ICH risk, while lower GCS score, lower mTICI score, and lower ASPECTS independently predict sICH risk. The nomograms provide practical tools for individualized risk assessment, aiding clinical decision-making and perioperative management in this high-risk cohort.