Quantification of brain tissue injury and prediction of prognosis using serum GFAP and UCH-L1: a multicenter prospective cohort study

Background: It remains unclear whether the serum levels of the brain injury biomarkers (glial fibrillary acidic protein [GFAP] and ubiquitin C-terminal hydrolase-L1 [UCH-L1]), can be used to quantitatively evaluate brain tissue injury and predict prognosis in patients with intravenous thrombolysis (IVT). Aim: This study investigates the association between serum GFAP and UCH-L1 levels with functional outcomes in patients receiving IVT. Methods: Patients were prospectively enrolled from 16 hospitals. We measured serum GFAP and UCH-L1 levels 24 hours after IVT. Infarct volume, hemorrhagic transformation (HT), and short- and long-term prognostic indicators were evaluated. GFAP and UCH-L1 cutoff levels for predicting 3-month unfavorable outcomes were derived, and a biomarker-based model was established and subjected to internal and external validation. Results: This study included 1028 patients. Higher GFAP and UCH-L1 levels were independently associated with larger infarct volume, HT, higher 24-hour and 7-day National Institutes of Health Stroke Scale scores, and 3-month modified Rankin Scale scores. The cutoff levels for GFAP and UCH-L1 (116 and 292 pg/mL, respectively) predicted patients with 3-month unfavorable outcomes with a specificity and positive predictive value (PPV) of 97.56% (95% confidence interval [CI], 94.51–99.00) and 88.68% (95% CI, 76.28–95.31), respectively, in the training cohort. In the testing and validation cohorts, specificity was 97.83% (95% CI, 91.62–99.62) and 96.90% (95% CI, 91.77–99.00), respectively, and PPV was 90.00% (95% CI, 66.87–98.25) and 75.00% (95% CI, 47.41–91.67), respectively. Furthermore, the biomarker-based nomogram model showed good predictability of 3-month prognosis in the different cohorts. Conclusions: Serum GFAP and UCH-L1 levels can be used to quantitatively evaluate brain tissue injury and predict the prognosis of patients with IVT.