Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Mar 16:13:832944.
doi: 10.3389/fneur.2022.832944. eCollection 2022.

Construction and Validation of an Immune-Related Risk Score Model for Survival Prediction in Glioblastoma

Wei Ren  1 , Weifeng Jin  2 , Zehua Liang  3
Affiliations

Construction and Validation of an Immune-Related Risk Score Model for Survival Prediction in Glioblastoma

Wei Ren et al. Front Neurol. .

Abstract

Background: As one of the most important brain tumors, glioblastoma (GBM) has a poor prognosis, especially in adults. Immune-related genes (IRGs) and immune cell infiltration are responsible for the pathogenesis of GBM. This study aimed to identify new tumor markers to predict the prognosis of patients with GBM.

Methods: The Cancer Genome Atlas (TCGA) database and ImmPort database were used for model construction. The Wilcoxon rank-sum test was applied to identify the differentially expressed IRGs (DEIRGs) between the GBM and normal samples. Univariate Cox regression analysis and Kaplan-Meier analysis was performed to investigate the relationship between each DEIRG and overall survival. Next, multivariate Cox regression analysis was exploited to further explore the prognostic potential of DEIRGs. A risk-score model was constructed based on the above results. The area under the curve (AUC) values were calculated to assess the effect of the model prediction. Furthermore, the Chinese Glioma Genome Atlas (CGGA) dataset was used for model validation. STRING database and functional enrichment analysis were used for exploring the gene interactions and the underlying functions and pathways. The CIBERSORT algorithm was used for correlation analysis of the marker genes and the tumor-infiltrating immune cells.

Results: There were 198 DEIRGs in GBM, including 153 upregulated genes and 45 downregulated genes. Seven marker genes (LYNX1, PRELID1P4, MMP9, TCF12, RGS14, RUNX1, and CCR2) were filtered out by sequential screening for DEIRGs. The regression coefficients (0.0410, 1.335, 0.005, -0.021, 0.123, 0.142, and -0.329) and expression data of the marker genes were used to construct the model. The AUC values for 1, 2, and 3 years were 0.744, 0.737, and 0.749 in the TCGA-GBM cohort and 0.612, 0.602, and 0.594 in the CGGA-GBM cohort, respectively, which indicated a high predictive power. The results of enrichment analysis revealed that these genes were enriched in the activation of T cell and cytokine receptor interaction pathways. The interaction network map demonstrated a close relationship between the marker genes MMP9 and CCR2. Infiltration analysis of the immune cells showed that dendritic cells (DCs) could identify GBM, while LYNX1, RUNX1, and CCR2 were significantly positively correlated with DCs expression.

Conclusion: This study analyzed the expression of IRGs in GBM and identified seven marker genes for the construction of an immune-related risk score model. These marker genes were found to be associated with DCs and were enriched in similar immune response pathways. These findings are likely to provide new insights for the immunotherapy of patients with GBM.

Keywords: glioblastoma; immune-related genes (IRGs); risk score; survival prognosis; tumor-infiltrating immune cells.

PubMed Disclaimer

Conflict of interest statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Figures

Figure 1
Figure 1
The flow chart of the study protocol.
Figure 2
Figure 2
Heatmap and volcano plot of 198 DEIRGs in TCGA-GBM patients. (A) Heatmap of 198 DEIRGs in TCGA-GBM patients. Blue and red indicate the genes of a lower and higher expression. (B) Volcano plot of 198 DEIRGs in TCGA-GBM patients. Blue and red, respectively, represent the significantly downregulated and upregulated genes. The names of genes in the dense regions are hidden.
Figure 3
Figure 3
Forest plots depicting the results of univariate and multivariate Cox analyses. (A) The results of 13 DEIRGs were screened out by the univariate Cox analysis (We removed two of the genes with a large confidence interval). (B) The results of 7 prognostic DEIRGs were screened out by following multivariate Cox analyses.
Figure 4
Figure 4
The time-dependent ROC curves in the TCGA training cohort. (A–C) ROC curves to evaluate the predictive accuracy of the model for OS at 1-, 2-, and 3-years in the TCGA training cohort.
Figure 5
Figure 5
The time-dependent ROC curves in the CGGA validation cohort. (A–C) ROC curves to validate the predictive accuracy of the model for OS at 1-, 2-, and 3-years in the CGGA validation cohort.
Figure 6
Figure 6
Survival analyses of the prognostic model in the TCGA-GBM cohort. (A) Risk score scatter map. (B) Heat map of differential expression of 7 marker genes. (C) The KM curves of OS in the high- and low-risk groups. (D) Survival status scatter map.
Figure 7
Figure 7
The results of survival analysis of the prognostic model in the CGGA-GBM cohort. (A) Risk score scatter map. (B) Heat map of the differential expression level of 7 marker genes. (C) KM curves of OS in the 2 risk subgroups. (D) Survival status scatter map.
Figure 8
Figure 8
The results of functional enrichment analysis of 198 DEIRGs. (A) The results of GO analysis demonstrated the most-enriched biological functions. (B) The results of KEGG analysis show the most enriched biological processes. The X-axis represents the enrichment fraction, while the Y-axis represents the GO term and the KEGG pathway. The color indicates the P-value, while the circle size of the bubble chart represents the gene counts.
Figure 9
Figure 9
The interaction network map of 5 marker genes. The network nodes represent the protein types, the edges represent protein–protein associations, and the line thicknesses represent the strength of the data supporting the interaction between the genes.
Figure 10
Figure 10
The violin plot shows the infiltration levels of 22 types of immune cells (blue and red represent normal and GBM samples, respectively).
Figure 11
Figure 11
The correlation between 6 marker genes and DC. (A) LYNX1. (B) MMP9. (C) TCF12. (D) RGS14. (E) RUNX1. (F) CCR2. (G) Kaplan–Meier analysis of the DC infiltrating levels from the TCGA database.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Ostrom QT, Patil N, Cioffi G, Waite K, Kruchko C, Barnholtz-Sloan JS. CBTRUS statistical report: primary brain and other central nervous system tumors diagnosed in the United States in 2013-2017. Neuro Oncol. (2020) 22 (12 Suppl. 2):iv1–iv96. 10.1093/neuonc/noaa200 - DOI - PMC - PubMed
    1. Lee E, Yong RL, Paddison P, Zhu J. Comparison of glioblastoma (GBM) molecular classification methods. Semin Cancer Biol. (2018) 53:201–11. 10.1016/j.semcancer.2018.07.006 - DOI - PubMed
    1. Yan H, Parsons DW, Jin G, McLendon R, Rasheed BA, Yuan W, et al. . IDH1 and IDH2 mutations in gliomas. N Engl J Med. (2009) 360:765–73. 10.1056/NEJMoa0808710 - DOI - PMC - PubMed
    1. Brat DJ, Verhaak RG, Aldape KD, Yung WK, Salama SR, Cooper LA, et al. . Comprehensive, integrative genomic analysis of diffuse lower-grade gliomas. N Engl J Med. (2015) 372:2481–98. 10.1056/NEJMoa1402121 - DOI - PMC - PubMed
    1. Network TC. Corrigendum: comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature. (2013) 494:506. 10.1038/nature11903 - DOI - PubMed