An immune-related eleven-RNA signature-drived risk score model for prognosis of osteosarcoma metastasis

Abstract

Abstract This study aimed to determine an immune-related RNA signature as a prognostic marker, in this study, we developed a risk score model for predicting the prognosis of osteosarcoma metastasis. We first downloaded the clinical information and expression data of osteosarcoma samples from the UCSC Xena and GEO databases, of which the former was the training set and the latter was the validation set. Immune infiltration was assessed using the ssGSEA and ESTIMATE algorithms, and the osteosarcoma samples were divided into the Immunity_L and Immunity_H groups. Then, eleven RNAs were identified as the optimal prognostic RNA signatures using LASSO Cox regression analysis for establishing a risk score (RS) model. Kaplan–Meier approach indicated the high-risk group exhibited a shorter survival. Furthermore, we analyzed the tumor metastasis, age, and RS model status were determined to be independent clinical prognostic factors using Cox regression analysis. Decision curve analysis (DCA) indicated that the prognostic factor + RS model had the best net benefit. Finally, nine tumor-infiltrating immune cells (TIICs) showed significant differences in abundance between high- and low-risk groups via CIBERSORT deconvolution algorithm. In conclusion, the immune-related eleven-RNA signature be could served as a potential prognostic biomarker for osteosarcoma metastasis.

Keywords

• Osteosarcoma
• Metastasis
• Risk score model
• Prognosis