Four metastasis associated gene signatures for predicting the prognosis of osteosarcoma patients

Abstract

Osteosarcoma (OS) is a rare malignant tumor derived from bone cells. Many studies have reported that metastatic OS is associated with a poorer prognosis. However, there are no current effective methods of predicting survival. We applied gene expression data of OS from Gene Expression Omnibus (GEO) and Therapeutically Applicable Research to Generate Effective Treatments (TARGET) to identify hub genes and construct a risk model with the help of R software. We then verified the identified signature using GEO data. Then we explored the potential relationship between the risk score and immune microenvironment in OS. After a series of analyses, we identified four hub-genes (APBB1IP, UHRF2, PDK1 and CORT). Hub genes were then used to construct a model using a training data set. The model was subsequently tested using a validation data set, and performed well in sensitivity and specificity. In the immune analysis, we found the high-risk score might represent immunosuppression in OS. The identified hub genes may play a critical role in the mechanism of the metastatic OS. Moreover, the model we built was related to immunosuppression in OS. Thus, the gene signature identified could be a potential target for future clinical treatment of OS.