免疫指数：用于肝细胞癌免疫治疗反应预测的基因和细胞预后特征

The heterogeneity of tumor immune microenvironment (TIME) plays important roles in the development and immunotherapy response of hepatocellular carcinoma (HCC). Using machine learning algorithms, we introduced the immune index (IMI), a prognostic model based on the HCC immune landscape.

We found that IMI low HCCs were enriched in stem cell and proliferating signatures, and yielded more TP53 mutation and 17p loss compared with IMI high HCCs. More importantly, patients with high IMI exhibited better immune-checkpoint blockade (ICB) response.

To facilitate clinical application, we employed machine learning algorithms to develop a gene model of the IMI (IMI G ), which contained 10 genes. According to our HCC cohort examination and single-cell level analysis, we found that IMI G high HCCs exhibited favorable survival outcomes and high levels of NK and CD8 + T cells infiltration.

Finally, after coculture with autologous tumor infiltrating lymphocytes, IMI G high tumor cells exhibited a better response to nivolumab treatment. Collectively, the IMI and IMI G may serve as powerful tools for the prognosis, classification and ICB treatment response prediction of HCC.