Identification of Key Genes Involved in the Recurrence of Glioblastoma Multiforme Using Weighted Gene Co-expression Network Analysis and Differential Expression Analysis

Glioblastoma multiforme (GBM) is the most fatal malignancy, and despite extensive treatment, tumors inevitably recur. This study aimed to identify recurrence-associated molecules in GBM. The gene expression profile GSE139533, containing 70 primary and 47 recurrent GBM tissues and their corresponding clinical traits, was downloaded from the Gene Expression Omnibus (GEO) database and used for weighted gene co-expression network analysis (WGCNA) and differentially expressed gene (DEG) analysis. After identifying the hub genes which differentially expressed in recurrent GBM tissues and in the gene modules correlated with recurrence, data from the Chinese Glioma Genome Atlas (CCGA) and The Cancer Genome Atlas (TCGA) databases were analyzed with GSE43378 to determine the relationship between hub genes and patient prognosis. The diagnostic value of the identified hub genes was verified using 52 GBM tissues. Three gene modules were correlated with recurrence and 2623 genes were clustered in these clinically significant modules. Among these, 13 genes - and - were differentially expressed in recurrent GBM samples; low and expression predicted poor outcomes. and expression showed weak positive relationships with Karnofsky performance scale scores. Additionally, and expression was found to be decreased in our recurrent GBM samples compared with that in primary samples; these genes exhibited high diagnostic value in distinguishing recurrent samples from primary samples. Our findings indicate that and might be involved in GBM recurrence and act as potential biomarkers for this condition.