Answers are grounded in this paper and cite their sources.
Background Osteosarcoma (OS) is an aggressive primary bone malignancy with poor outcomes in patients with recurrent or metastatic disease. Although molecular profiling has revealed substantial heterogeneity in OS, the relationships between stemness-associated gene expression, patient prognosis, and the tumor immune microenvironment remain incompletely understood. Methods In this study, non-negative matrix factorization of stemness-associated gene expression was used to identify molecular subtypes in TARGET-OS. Functional enrichment, weighted gene co-expression network analysis, somatic mutation profiling, and drug response prediction characterized subtype differences. After joint batch correction of TARGET-OS and GSE21257, an overall survival model based on stemness-associated gene expression was developed using LASSO and Cox regression in TARGET-OS and evaluated in GSE21257. Immune infiltration and single-cell transcriptomic analyses characterized the associated immune features, cellular distribution of the signature, and potential cell–cell interactions. Results Our analysis identified two exploratory stemness-associated OS subtypes with distinct prognoses in the TARGET-OS cohort. C1 showed poorer survival and enrichment of metabolic and DNA repair programs, whereas C2 was enriched for extracellular matrix and immune-related processes. Drug response predictions indicated greater sensitivity to fixed-ratio combinations containing selumetinib in C1 and to cabozantinib in C2. We developed a nine-gene signature that stratified overall survival in both cohorts, with 1-, 3-, and 5-year AUCs of 0.653, 0.751, and 0.711, respectively, in GSE21257. High-risk tumors exhibited lower immune scores and reduced estimated infiltration of CD8 + T Cells and B Cells. Single-cell analysis revealed relatively high signature scores in OS cells and cancer-associated fibroblasts. The high-scoring OS1 subpopulation was enriched for ossification and extracellular matrix-related processes. Communication analysis further predicted COL1A1–CD44 interactions from OS cells and fibroblasts to monocytes. Conclusion Our study characterizes stemness-associated transcriptional heterogeneity in OS and identifies a nine-gene prognostic signature. These findings associate survival differences with metabolic and immune programs and reveal the cellular expression patterns of the signature in tumor and stromal cells.
No takes yet. Share an insight, caveat, or question.
Cui et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: