Abstract
Osteosarcoma (OS), the most prevalent primary malignant bone tumor with a dismal prognosis, exhibits significant heterogeneity in programmed cell death (PCD) pathways, but its subtype-specific functional mechanisms remain poorly characterized. This study integrated PCD-related gene signatures to delineate molecular subtypes in OS via consensus clustering, successfully defining four distinct subtypes with divergent prognostic outcomes and immune microenvironments. Differential expression, functional enrichment, and protein–protein interaction (PPI) network analyses revealed subtype-specific PCD pathway associations (e.g., lysosome-dependent cell death, apoptosis, pyroptosis and anoikis), while comparative immune profiling and clinical characterization further refined subgroup identities. A robust prognostic risk model incorporating five pivotal genes (SERPINE2, CBS, SQLE, UBE2D4, and S100A13) and metastasis status demonstrated superior predictive performance in both training and external validation cohorts. These findings not only elucidate the functional architecture of PCD across OS molecular subtypes but also establish a clinically actionable model for precision risk stratification and tailored therapeutic strategies.