Abstract

Purpose Neuroblastoma (NB) is the most common extracranial solid tumor in children with poor overall survival. Increasing evidence indicates that telomeres contribute to tumorigenesis and influence cancer prognosis. However, the biological and clinical implications of telomere-related genes (TRGs) in NB remain poorly defined. Materials and Methods We integrated data from multiple independent cohorts to elucidate the roles of TRGs in NB. Differential expression and weighted gene co-expression network analyses (WGCNA) were performed to identify telomere-related differentially expressed genes (TRDEGs) linked to patient survival. Consensus clustering based on TRDEG expression patterns was conducted to stratify molecular subtypes, followed by functional enrichment analysis. A prognostic signature was then built using machine-learning algorithms to predict clinical outcomes and potential therapeutic responses. Single-cell RNA sequencing (scRNA-seq) data were used for signature gene expression validation and to guide functional candidate selection. Quantitative RT-PCR was performed to verify the TRDEG signature, and functional assays were performed to explore the role of PSAT1 in NB progression. Results First, we identified 103 telomere-related differentially expressed genes (TRDEGs) significantly linked to NB patient survival. Consensus clustering of TRDEGs revealed two NB molecular subtypes with distinct biological processes and clinical outcomes. We established an eight-gene prognostic signature ( ARHGAP23, CHD5, E2F3, ELOVL6, FEN1, GMPS, LRR1, and PSAT1 ) that demonstrated high predictive accuracy, with 1-, 3-, and 5-year survival AUCs of 0.885, 0.903, and 0.911, respectively. The model showed consistent robustness across validation cohorts. Multivariate Cox regression confirmed the risk score as an independent prognostic factor. Integrating the risk score with clinical parameters within a nomogram yielded superior prognostic performance compared with tradition