Cross-platform transcriptomic integration identifies PDPN for precision prognostic stratification of 1p/19q non-codeleted adult diffuse gliomas

Abstract

Molecular stratification is central to precision oncology, enabling targeted diagnosis and risk-adapted management of glioma. However, clinically robust transcriptomic biomarkers that refine molecular risk stratification according to 1p/19q status and support personalized diagnostic decision-making remain limited. Microarray and RNA-seq datasets from GEO and CGGA databases representing glioma patients with or without 1p/19q co-deletion were analyzed. Differentially expressed genes (DEGs) were identified with strict criteria (adjusted p < 0.01, |log2FC| ≥ 3). Weighted gene co-expression network analysis (WGCNA) was performed to detect gene modules and hub genes. Integration of DEGs and WGCNA hub genes through intersection revealed key real hub genes as candidate biomarkers. Functional pathway enrichment and protein–protein interaction network (PPIN) analyses characterized biological relevance. Validation of candidate biomarkers was performed via, Kaplan–Meier survival analyses, ROC curve assessments, multivariable Cox regression, prognostic nomograms, and decision curve analyses. Cross-platform integration of differential expression and WGCNA identified PDPN as the principal hub gene associated with 1p/19q non-codeleted gliomas. PDPN was overexpressed in non-codeleted tumors and accurately discriminated 1p/19q status (AUC > 0.83) also, survival-based prognostic evaluation revealed that elevated PDPN expression exhibited a high statistically significant association with overall survival outcomes ( p < 0.0001). Functional enrichment and protein–protein interaction analyses linked PDPN to extracellular matrix remodeling, collagen organization, and invasion-related pathways. Prognostic nomograms supported individualized survival prediction, whereas decision curve analysis showed PDPN matched 1p/19q codeletion without incremental clinical benefit. PDPN complements prognostic assessment by identifying aggressive 1p/19q non-codeleted gliomas, supporting precision medicine and molecular stratification.