Application of a risk score model based on glycosylation-related genes in the prognosis and treatment of patients with low-grade glioma.

INTRODUCTION

Low-grade gliomas (LGG) represent a heterogeneous and complex group of brain tumors. Despite significant progress in understanding and managing these tumors, there are still many challenges that need to be addressed. Glycosylation, a common post-translational modification of proteins, plays a significant role in tumor transformation. Numerous studies have demonstrated a close relationship between glycosylation modifications and tumor progression. However, the biological function of glycosylation-related genes in LGG remains largely unexplored. Their potential roles within the LGG microenvironment are also not well understood.

METHODS

We collected RNA-seq data and scRNA-seq data from patients with LGG from TCGA and GEO databases. The glycosylation pathway activity scores of each cluster and each patient were calculated by irGSEA and GSVA algorithms, and the differential genes between the high and low glycosylation pathway activity score groups were identified. Prognostic risk profiles of glycosylation-related genes were constructed using univariate Cox and LASSO regression analyses and validated in the CGGA database.

RESULTS

An 8 genes risk score signature including ASPM, CHI3L1, LILRA4, MSN, OCIAD2, PTGER4, SERPING1 and TNFRSF12A was constructed based on the analysis of glycosylation-related genes. Patients with LGG were divided into high risk and low risk groups according to the median risk score. Significant differences in immunological characteristics, TIDE scores, drug sensitivity, and immunotherapy response were observed between these groups. Additionally, survival analysis of clinical medication information in the TCGA cohort indicated that high risk and low risk groups have different sensitivities to drug therapy. The risk score characteristics can thus guide clinical medication decisions for LGG patients.

CONCLUSION

Our study established glycosylation-related gene risk score signatures, providing new perspectives and approaches for prognostic prediction and treatment of LGG.