Identification of a Prognostic Model Based on Immune Cell Signatures in Clear Cell Renal Cell Carcinoma

Background. Accumulating evidence substantiated that the immune cells were intricately intertwined with the prognosis and therapy of clear cell renal cell carcinoma (ccRCC). We aimed to construct an immune cell signatures (ICS) score model to predict the prognosis of ccRCC patients and furnish guidance for finding appropriate treatment strategies. Methods. Based on The Cancer Genome Atlas (TCGA) database, the normalized enrichment score (NES) of 184 ICSf was calculated using single-sample gene set enrichment analysis (ssGSEA). An ICS score model was generated in light of univariate Cox regression and Least absolute shrinkage and selection operator (Lasso)-Cox regression, which was independently validated in ArrayExpress database. In addition, we appraised the predictive power of the model via Kaplan-Meier (K-M) curves and time-dependent receiver operating characteristic (ROC) curves. Eventually, immune infiltration, genomic alterations and immunotherapy were analyzed between high and low ICS score groups. Results. Initially, we screened 11 ICS with prognostic impact based on 515 ccRCC patients. K-M curves presented that the high ICS score group experienced a poorer prognosis (P < 0.001). In parallel, ROC curves revealed a satisfactory reliability of model to predict individual survival at 1, 3, and 5 years, with area under the curves (AUCs) of 0.744, 0.713, and 0.742, respectively. In addition, we revealed that the high ICS score group was characterized by increased infiltration of immune cells, strengthened BAP1 mutation frequency, and enhanced expression of immune checkpoint genes. Conclusion. The ICS score model has higher predictive power for patients' prognosis and can instruct ccRCC patients in seeking suitable treatment.