基于内质网应激应答的胶质瘤风险模型构建.

Objective This study aimed to explore the predictive value of endoplasmic reticulum stress response （ERS）-related regulatory gene expression for the pathological grading, prognosis, and malignant progression phenotype of gliomas （excluding glioblastomas）. Methods After excluding duplicate and non-survival samples, clinical sequencing data of glioma samples from the American Cancer Genome Atlas and the Chinese Brain Glioma Genome Atlas were selected as the training and validation sets. A prognosis-related ERS risk regression model was constructed through differential gene enrichment and protein interaction analysis. The predictive value of ERS Cox model for glioma prognosis and malignant progression phenotype was validated using ROC curve analysis, real-time fluorescence quantitative PCR, and immunohistochemistry. Results The study findings reveal that the expression of 7 ERS-related risk factors increases with the rise in glioma grade and accurately predicts unfavorable patient prognosis （with accuracies higher than 0.7 for predictions at 1, 3, and 5 years）, all of which are statistically significant. Further validation demonstrates a positive correlation between ERS risk genes and the glioma malignant phenotype marker CD44, as well as a negative correlation with the clinically favorable prognosis marker GPR158, both of which have statistical differences （P < 0.05）. Finally, gene expression and immunohistochemistry analysis of clinical samples confirm that ERS-related risk factors are highly expressed in higher-grade gliomas, positively correlated with CD44 expression, with statistical significance （P < 0.05）.Conclusion The preliminary results of the study suggest that the risk regression model based on ERS response exhibits the capacity to predict pathological grading and prognosis. Moreover, it demonstrates a positive correlation with the malignant progression phenotype of gliomas. These findings offer insights for precise and targeted diagnosis and treatment of gliomas. [ABSTRACT FROM AUTHOR]