Identification of amino acid metabolism-based molecular subtypes and prognostic signature to predict immune landscape and guide clinical drug treatment in osteosarcoma

Background: Osteosarcoma (OS) is the most common bone malignancy in children and adolescents. The 5-year survival rate is only approximately 20% in patients with metastatic and recurrent OS. There is a critical necessity to ascertain prognostic markers that can predict the survival outcomes of OS patients, as well as to explore new therapeutic targets for the treatment of these patients. Method: In this study, the RNA sequencing profile of osteosarcoma patients and a total of 101 genes relevant to amino acid metabolism were subjected to analyze. Cox analysis and the NMF algorithm were utilized to identified clusters based on genes involved in amino acid metabolism. Then, we performed LASSO regression analysis based on these genes to construct a prognostic signature and a nomogram to predict the prognosis of OS patients. Additionally, tumor microenvironment, tumor mutation burden, tumor immunity responses and drug sensitivity were analyzed by R packages. The expression of these genes in osteosarcoma and normal tissues was determined by immunohistochemistry assays. Results: Four clusters of OS were identified based on genes related to amino acid metabolism. Cluster 2 exhibited the poorest outcomes, may better benefit from immunotherapy, and exhibits specific sensitivity differences to conventional drugs. Therefore, we developed a classifier score calculator based on the amino acid metabolism-related genes. A prognostic signature containing three risk genes and five protective genes and a nomogram were developed to predict the prognosis of OS patients, which had excellent predictive power. Immunohistochemistry assays illustrated that ATF4 and DIO1 were highly expressed in OS tissues while SLC7A7 was suppressed in OS tissues. Conclusion: In conclusion, multiple databases and analytical techniques were utilized to develop a robust eight-gene signature model for predicting OS outcomes. A distinct cluster of OS patients was identified based on genes related to amino acid metabolism, suggesting potential benefits from immunotherapy and targeted drug therapies. This model has the potential to enhance clinical decision-making for the treatment of OS patients.