Identification of an immune subtype predicting survival risk and immune activity in hepatocellular carcinoma

Immune checkpoint inhibitors (ICI) prolong the survival for advanced/metastatic patients with lung cancer or melanoma; however, for hepatocellular carcinoma (HCC) patients, a durable response has not been reported. Herein, we used a total of 719 HCC patients with public genomic data to determine potential prognostic and immunogenic subtypes. The non-negative matrix factorization (NMF) method was applied to identify the immune classes and potential subtypes. The proportion of tumor infiltration immune cells was estimated using the CIBERSORT algorithm. Gene set enrichment analysis (GSEA) was utilized to calculate the dysregulated pathways. By using NMF analysis for the gene expression profile of the top immune genes, one HCC subtype with better survival (i.e., low-risk subtype) and another with worse survival (i.e., high-risk subtype) were identified in 3 HCC cohorts (all P < 0.05). Better immune cell infiltration, increased enrichment of immune signatures, higher expression of checkpoints, and elevated tumor mutation load (TML) were significantly enriched in the low-risk subtype (all P < 0.05). Higher mutation rates of immune response genes (e.g., TP53 and MUC16) were also observed in the low-risk subtype (both P < 0.05). Discovery of the HCC low-risk subtype might provide clues for HCC prognosis and immunotherapy prediction.