• Xiaodong Chen, Fen Tian, and Zeyu Wu.
• Neurosurgery Department, The Affiliated Hospital of Qingdao University, Shandong, China.
• World Neurosurg. 2022 Nov 1; 167: e515e526e515-e526.

BackgroundGenomic instability and aberrant tumor mutation burden are widely accepted hallmarks of cancer. Glioblastoma (GBM) is a common brain tumor in adults, and survival of patients with GBM is poor. This study aimed to investigate the prognostic value of genomic instability-derived genes in GBM.MethodsGBM data were downloaded from The Cancer Genome Atlas and Chinese Glioma Genome Atlas databases. Differential expression analysis of all samples with different tumor mutation burden was performed. Univariate Cox and LASSO Cox regression analyses were integrated to determine the optimal genes for constructing a risk score model. Multivariate Cox regression analysis and survival analysis determined independent prognostic indicators. Immune cell infiltration was analyzed by CIBERSORT algorithm.ResultsIn GMB patients with high and low tumor mutation burden, we identified 154 differentially expressed genes, which were significantly enriched in 47 Gene Ontology terms and 6 Kyoto Encyclopedia of Genes and Genomes pathways. To establish a risk score, 9 genes were further screened, including SDC1, CXCL1, CXCL6, RGS4, PCDHGB2, CA9, ZAR1, CHRM3, and SLN. High-risk patients had worse prognosis in two databases. The performance of a nomogram including prognostic factors (risk score and age) was good. Moreover, mast cells resting was significantly differentially infiltrated between high- and low-risk GBM samples.ConclusionsThe risk score constructed by 9 genomic instability-derived genes could reliably predict prognosis of GBM patients. The nomogram based on age and risk score also had a good prognostic predictive value.Copyright © 2022 Elsevier Inc. All rights reserved.

17 keywords...

hide…