Abstract
OBJECTIVE: Using model algorithms, we constructed an immune-related long non-coding RNAs (lncRNAs) risk coefficient model to predict outcomes for patients with clear cell renal cell carcinoma (ccRCC) to understand the infiltration of tumor immune cells and the sensitivity to immune-targeted drugs. METHODS: Open genes data were downloaded from The Cancer Genome Atlas and The Immunology Database and Analysis Portal, and immune-related lncRNAs were obtained through Pearson correlation analysis. R language software was used to obtain differentially expressed immune-related lncRNAs and immune-related lncRNA pairs. The model was constructed using least absolute shrinkage and selector operation regression analysis, and receiver operator characteristic curves were drawn. The Akaike information criterion was used to distinguish the high-risk from the low-risk group. We also conducted correlation analysis for the high- and low-risk subgroups. RESULTS: We identified 27 immune-related lncRNAs pairs, 16 of which were included in the model construction. After merging clinical data, the areas under the curve of 1 -year, 3-year, and 5-year survival times of ccRCC patients were 0.867, 0.832, and 0.838, respectively. Subgroup analyses were conducted according to the cut-off value. We found that the high-risk group was associated with poor outcomes. The risk score and tumor stage were independent predictors of the outcome of ccRCC. The risk model predicted specific immune cell infiltration, immune checkpoint gene expression levels, and high-risk groups more sensitive to sunitinib targeted therapy. CONCLUSION: We obtained prognostic-related novel ccRCC markers and risk model that predicts the outcome of patients with ccRCC and helps identify those who can benefit from sunitinib.