Abstract
Background: Renal cell carcinoma (RCC) is a common malignant tumor worldwide, and immune checkpoint inhibitors are a new therapeutic option for metastatic RCC. Infiltrating immune cells in the tumor microenvironment (TME) play a critical part in RCC biology, which is important for tumor therapy and prediction. Hypoxia is a common condition that occurs in the TME and may lead to RCC immunosuppression and immune escape. This study was conducted to analyze the extent of the hypoxia immune microenvironment in the TME of RCC and develop a hypoxia-related risk model for predicting the prognosis of patients with RCC. Methods: The gene expression profiles of 526 patients with RCC were downloaded from The Cancer Genome Atlas database. Combined with the hallmark-hypoxia gene dataset downloaded from Gene Set Enrichment Analysis, prognosis-related hypoxia genes were selected by survival analysis. A protein-protein interaction network and functional enrichment analysis were performed. A hypoxia-related risk model predicting the prognosis of patients with RCC was established using the least absolute shrinkage and selection operator. Data of 91 cases downloaded from the International Cancer Genome Consortium (ICGC) database were used for validation. CIBERSORT was applied to analyze the fractions of 22 immune cell types in the TME of RCC between low- and high-risk groups. The expression profiles of immunomodulators and immunosuppressive cytokines were also analyzed. Results: Ninety-three genes were significantly associated with poor overall survival of patients with RCC and were mainly involved in 10 pathways. Using the established hypoxia-related risk model, the receiver operating characteristic curves showed an accuracy of 76.1% (95% CI: 0.719-0.804), and Cox proportional hazards regression analysis revealed that the model was an independent predictor of the prognosis of patients with RCC [hazard ratio (HR) = 2.884; 95% CI: 2.090-3.979] (p < 0.001). Using the ICGC database, we verified that the low-risk score group had a better overall survival outcome than the high-risk group. Additionally, dividing the hypoxia risk score into high-risk and low-risk groups could predict the immune microenvironment of RCC. Conclusions: We demonstrated that a hypoxia-related risk model can be used to predict the outcomes of patients with RCC and reflect the immune microenvironment of RCC, which may help improve the overall clinical response to immune checkpoint inhibitors.