Gene expression-based immune infiltration analyses of liver cancer and their associations with survival outcomes

Liver cancer is one of the most lethal malignancies, presenting inferior survival outcomes once diagnosed. Current therapeutic approaches mainly target the tumor cells or vasculature, but rarely take the immune factors into consideration.

In our study, the compositions of tumor-infiltrating immune cells (TIICs) in liver cancer and paracancer samples were analyzed based on the gene expression profiles by CIBERSORT. After calculating the proportions of 22 TIICs subtypes in 51 paired cancer and paracancer samples, we found their proportions varied between intragroup and intergroup. Compared with the paracancer tissues, the relative proportions of macrophages M0 and resting mast cells in liver cancer samples were significantly elevated, while that of M2 macrophages were reduced.

Univariate Cox regression analysis with the 22 TIICs subtypes as continuous variables showed increased B cells memory and resting NK cells were significantly associated with poor survival outcome. Besides, hierarchical clustering analysis based on the proportions of 22 TIICs subtypes identified 3 clusters, which exhibited distinct prognosis. Among them, cluster 1 had superior survival outcomes, while cluster 3 had inferior survival outcomes.

Collectively, our research suggested certain TIICs subpopulations proportions, as well as cluster patterns were associated with the prognosis of liver cancer, which provided potential therapeutic targets for liver cancer.