基于 NK 细胞标记基因的模型在肝细胞癌的预后和免疫治疗反应中显示出良好的预测能力

Hepatocellular carcinoma (HCC) is the fourth leading cause of malignancy worldwide, and its progression is influenced by the immune microenvironment. Natural killer (NK) cells are essential in the anti-tumor response and have been linked to immunotherapies for cancers.

Therefore, it is important to unify and validate the role of NK cell-related gene signatures in HCC. In this study, we used RNA-seq analysis on HCC samples from public databases.

We applied the ConsensusClusterPlus tool to construct the consensus matrix and cluster the samples based on their NK cell-related expression profile data.

We employed the least absolute shrinkage and selection operator regression analysis to identify the hub genes.

Additionally, we utilized the CIBERSORT and ESTIMATE web-based methods to perform immune-related evaluations.

Our results showed that the NK cell-related gene-based classification divided HCC patients into three clusters. The C3 cluster was activated in immune activation signaling pathways and showed better prognosis and good clinical features. In contrast, the C1 cluster was remarkably enriched in cell cycle pathways. The stromal score, immune score, and ESTIMATE score in C3 were much higher than those in C2 and C1.

Furthermore, we identified six hub genes: CDC20, HMOX1, S100A9, CFHR3, PCN1, and GZMA. The NK cell-related genes-based risk score subgroups demonstrated that a higher risk score subgroup showed poorer prognosis. In summary, our findings suggest that NK cell-related genes play an essential role in HCC prognosis prediction and have therapeutic potential in promoting NK cell antitumor immunity. The six identified hub genes may serve as useful biomarkers for novel therapeutic targets.