Predictors of unfavorable 3-month functional outcome following intravenous thrombolysis with alteplase in anterior circulation acute ischemic stroke: a prospective cohort study

BackgroundIntravenous thrombolysis with alteplase remains the standard reperfusion strategy for acute ischemic stroke (AIS), yet many patients still experience unfavorable outcomes despite timely treatment, underscoring the need for reliable multimodal prognostic markers.ObjectiveTo identify independent clinical, laboratory, and neuroimaging predictors of unfavorable 3-month functional outcome in anterior circulation AIS treated with alteplase, and to develop, internally validate, and benchmark an integrated multivariable model in accordance with TRIPOD.MethodsThis prospective single-center cohort study enrolled 268 consecutive patients with anterior circulation AIS receiving intravenous alteplase within 4.5 h of symptom onset (March 2022, February 2025). Three-month outcome was assessed by the modified Rankin Scale (mRS 0, 2 favorable; 3, 6 unfavorable) using validated structured instruments administered by blinded raters. Multivariable logistic regression was performed, and the final model was internally validated by 1,000-replicate bootstrap with optimism correction and shrinkage, evaluated by decision curve analysis (DCA), rendered into a nomogram, and benchmarked head-to-head against three previously published reference models using the DeLong test.ResultsOf 268 patients, 99 (36.9%) experienced unfavorable outcomes. Seven independent predictors were identified: early neurological deterioration (adjusted OR 3.45, 95% CI 2.01, 5.92), large infarction exceeding one-third of the middle cerebral artery territory (OR 2.78, 1.58, 4.89), baseline NIHSS (OR 1.14 per point, 1.07, 1.22), poor Tan collateral score (OR 2.31, 1.34, 3.98), low clot burden score (OR 2.15, 1.28, 3.61), elevated D-dimer (OR 2.19, 1.29, 3.72), and elevated CRP (OR 1.87, 1.11, 3.15). The model achieved an apparent AUC of 0.847 (95% CI 0.801, 0.893) and an optimism-corrected AUC of 0.831 (0.785, 0.877) on bootstrap validation, with satisfactory calibration (Hosmer, Lemeshow P = 0.394). DCA showed positive net benefit across threshold probabilities of 0.15, 0.75, and the model exceeded the recalibrated Hu, Ping, and Lv models.ConclusionA multimodal panel integrating clinical, inflammatory, coagulation, and neuroimaging parameters independently predicts unfavorable 3-month outcome following intravenous thrombolysis in anterior circulation AIS. The findings are hypothesis-generating pending external validation in independent multicenter cohorts.