A novel immune cell signature for predicting osteosarcoma prognosis and guiding therapy

Dysregulation of immune cell infiltration in the tumor microenvironment contributes to the progression of osteosarcoma (OS). In the present study, we explored genes related to immune cell infiltration and constructed a risk model to predict the prognosis of and guide therapeutic strategies for OS. The gene expression profile of OS was obtained from TARGET and Gene Expression Omnibus, which were set as the discovery and verification cohorts. CIBERSORT and Kaplan survival analyses were used to analyze the effects of immune cells on the overall survival rates of OS in the discovery cohort. Differentially expressed gene (DEG) analysis and protein-protein interaction (PPI) networks were used to analyze genes associated with immune cell infiltration. Cox regression analysis was used to select key genes to construct a risk model that classified OS tissues into high- and low-risk groups. The prognostic value of the risk model for survival and metastasis was analyzed by Kaplan-Meier survival analyses, receiver operating characteristic curves, and immunohistochemical experiments. Immunological characteristics and response effects of immune checkpoint blockade (ICB) therapy in OS tissues were analyzed using the ESTIMATE and Tumor Immune Dysfunction and Exclusion algorithms, while sensitivity for both targeted and chemotherapy drugs was analyzed using the OncoPredict algorithm. It was demonstrated that the high infiltration of resting dendritic cells in OS tissues was associated with poor prognosis. A total of 225 DEGs were found between the high- and low-infiltration groups of OS tissues, while 94 genes interacted with others. Through COX analyses, among these 94 genes, four genes (including AOC3, CDK6, COL22A1, and RNASE6) were used to construct a risk model. This risk model showed a remarkable prognostic value for survival rates and metastasis in both the discovery and verification cohorts. Even though a high microsatellite instability score was observed in the high-risk group, the ICB response in the high-risk group was poor. Furthermore, using OncoPredict, we found that the high-risk group OS tissues were resistant to seven drugs and sensitive to 25 drugs. Therefore, our study indicates that the resting dendritic cell signature constructed by AOC3, CDK6, COL22A1, and RNASE6 may contribute to predicting osteosarcoma prognosis and thus therapy guidance.

Automated summary

This study constructed a risk model to predict the prognosis and guide therapeutic strategies for osteosarcoma (OS) by analyzing gene expression profiles and immune cell infiltration.