Abstract:
Objective To identify the potential prognostic biomarkers of the immune-related genes signature for patients with hepatocellular carcinoma (HCC).
Methods Original HCC data were downloaded from TCGA, and the immune activity of each sample was calculated by ssGSEA. HCC samples were divided into high and low immune cell infiltration groups by "GSVA" package and "hclust" package. The ESTIMATE algorithm scored the tumor microenvironment in each HCC sample. The "limma" package and Venn diagram identified effective immune-related genes. Univariate Cox, Lasso regression and multivariate Cox regression analyses were used to explore key genes. The "rms" package was used to create nomograms and draw calibration curves.
Results Compared with the high immune cell infiltration group, the tumor purity of the samples in the low immune cell infiltration group was higher, the immune score, ESTIMATE score and stromal score were lower. In the high immune cell infiltration group, the immune components were more abundant, and the expression levels of TIGIT, PD-L1, PD-1, LAG3, TIM-3, CTLA4 and HLA family were higher. Multivariate Cox regression analysis showed that four immune-related genes (S100A9, HMOX1, IL18RAP and FCER1G) were used to construct the prognosis model. Compared with other clinical features, the risk score of this prognostic model was recognized as an independent prognostic factor.
Conclusion This study identified the immune-related core genes which may be used in targeted therapy and immunotherapy of HCC.
下载:
