Related Experiment Video
Updated: Sep 9, 2026
Evaluation of Biomarkers in Glioma by Immunohistochemistry on Paraffin-Embedded 3D Glioma Neurosphere Cultures
Published on: January 9, 2019
An integrative clinical-molecular model as an auxiliary predictive tool for glioma malignancy grade
Wei Wen1, Xiaoli Zhang1, Lin Yue1
1Department of Neurosurgery, The people's Hospital of Leshan, Leshan City, Sichuan, China.
Objective:
An integrative auxiliary predictive model incorporating clinical parameters, serum biomarkers, and molecular pathological markers was developed to assess glioma malignancy grade.
Methods:
This single-center retrospective observational study consecutively enrolled 400 glioma patients. A total of 26 variables, including demographic characteristics, clinical parameters, laboratory indicators, and serum biomarkers, were analyzed. Predictors were selected using univariate analysis, followed by Least Absolute Shrinkage and Selection Operator (LASSO) regression and machine learning models to construct the model. Model interpretability was assessed via SHapley Additive exPlanations (SHAP) analysis. Performance was evaluated using the area under the receiver operating characteristic curve (AUC), calibration curves, and decision curve analysis (DCA).
Results:
Independent predictors for high-grade glioma (WHO grades III-IV) included age (OR = 1.085, 95%CI: 1.035-1.138), KPS score (OR = 0.928, 95%CI: 0.886-0.972), maximum tumor diameter (OR = 1.649, 95%CI: 1.233-2.205), neutrophil-to-lymphocyte ratio (OR = 2.310, 95%CI: 1.487-3.589), albumin-to-globulin ratio (OR = 0.163, 95%CI: 0.031-0.847), IDH mutation status (wild-type vs. mutant, OR = 4.502, 95%CI: 1.694-11.966), and Ki-67 proliferation index (OR = 1.176, 95%CI: 1.106-1.250). The Random Forest achieved an AUC of 0.864 (95%CI: 0.815-0.914) in the training set and 0.820 (95%CI: 0.732-0.909) in the validation set. Calibration was excellent (Hosmer-Lemeshow test, p > 0.05), and DCA demonstrated clinical net benefit within a 0.2-0.8 threshold probability range.
Conclusion:
A novel hybrid clinical-molecular predictive model integrating serum biomarkers and clinical parameters was developed via internal validation for glioma grading. This exploratory auxiliary tool can assist preoperative assessment and guide individualized treatment strategies; however, it cannot replace pathological and molecular diagnosis, as it incorporates tissue-derived markers such as IDH mutation status and Ki-67 index.