Identification of prognostic biomarkers in neuroblastoma using WGCNA and multi-omics analysis.

Background: Neuroblastoma (NB) is one of the most frequent parenchymal tumors among children, with a high degree of heterogeneity and wide variation in clinical presentation. Despite significant therapeutic advances in recent years, long-term survival in high-risk patients remains low, emphasizing the urgent need to find new biomarkers and construct reliable prognostic models. Methods: In this study, data from neuroblastoma samples in the ArrayExpress database were utilized to identify key gene modules and pivotal genes associated with NB prognosis by weighted gene co-expression network analysis (WGCNA). The Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) functional enrichment analysis was performed using the DAVID database. Based on these hub genes, survival prognosis models were constructed and validated on an independent validation set in the Gene Expression Omnibus (GEO) database. Differences in biological functions and immune microenvironments and the sensitivity to pharmacological and immunotherapeutic treatments of patients in the high- and low-risk groups were examined by gene set enrichment analysis (GSEA) and immune infiltration analysis. Results: WGCNA identified 14 gene modules and screened the module with the highest relevance to the International Neuroblastoma Staging System (INSS), containing 60 pivotal genes. GO and KEGG analyses demonstrated that these pivotal genes were mainly implicated in biological processes and signaling pathways including DNA replication, cell division, mitotic cell cycle, and cell cycle. Based on Lasso regression and COX regression analysis, a prognostic model containing DHFR, GMPS and E2F3 was constructed, and the RiskScore was significantly correlated with the 1-, 3- and 5-year survival of the patients. GSEA and immune infiltration analyses revealed significant differences in the levels of cell cycle-related pathways and immune cell infiltration between the high and low RiskScore groups. In particular, patients in the high-risk group are less likely to benefit from immunotherapy and may be better suited for treatment with drugs such as Oxaliplatin and Alpelisib. Conclusion: This research systematically identified biomarkers related to NB prognosis and developed a reliable prognostic model applying WGCNA and multiple bioinformatics methods. The model has important application value in predicting patients' prognosis, evaluating drug sensitivity and immunotherapy effect, and provides new ideas and directions for precise treatment of neuroblastoma. [ABSTRACT FROM AUTHOR]