Development of a prognostic gene signature based on an immunogenomic infiltration analysis of osteosarcoma

Abstract

Osteosarcoma is the most common primary malignant bone tumour predominantly occurring in children and adolescents with a high tendency of local invasion and early metastases. Currently, tumour immune microenvironment (TME) is becoming the focus of studying of malignant tumours. However, no sound evidence shows a specific immune molecular target in osteosarcoma. We downloaded the gene expression profile and clinical data of osteosarcoma from the TARGET portal, and extracted and normalized via R software. Then, the immune cell infiltration assessed by CIBERSORT and ESTIMATE algorithms. Three survival-related immune cells and immune score were obtained via Kaplan-Meier survival analysis, and 232 immune-related genes were obtained as candidate genes. Enrichment and protein-protein interaction co-expression analyses were performed to identify 13 hub genes. Lastly, a seven gene prognostic signature was identified by univariate and multivariate Cox regression analyses. More importantly, our validations and TIMER algorithm suggested this immune-related prognostic signature a good predictive tool. Our findings have provided novel insights that could demonstrate new targets of immunotherapy in osteosarcoma.