A sub-cluster of cancer cells indicates poor prognosis in advanced hepatocellular carcinoma.

Background: For patients with advanced hepatocellular carcinoma (HCC), it is emergent to focus on elucidating different classifications of intratumoral heterogeneity and understanding the mechanisms of treatment resistance to improve prognosis. This study aims to classify the sub-clusters of cancer cells in patients diagnosed with advanced HCC, and identify the genes, pathways and microenvironment associated with prognosis.

Methods: Single-cell transcriptomic profiling were conducted on advanced HCC samples. Bulk RNA-seq transcriptome data were obtained from The Cancer Genome Atlas (TCGA) database. Prognosis analysis was performed for survival stratification within the identified sub-clusters. Enriched genes and pathways were determined, and the association and underlying mechanism of the identified sub-cluster of cancer cells were elucidate with other cellular components within advanced HCC tumor tissue.

Results: A total of 26,800 cells were obtained from two patients with advanced HCC. Seven sub-clusters were identified, and only the CancerCell 6 cluster were exhibited significant disparities in both overall survival (OS) and progression-free survival (PFS) rates according to the 370 HCC patients from the TCGA database. The CancerCell 6 cluster was associated with adenosine triphosphate (ATP)-related pathways and the high expression of pyruvate kinase M (PKM). It also demonstrated a strong interaction ability with mononuclear phagocytes (MPs). The MPs_MMP9 cluster showed strong interactions between the CancerCell 6 cluster and related signaling pathways of EDN and NOTCH, indicating a tendency toward M2 polarization.

Conclusions: This study identified the CancerCell 6 cluster associated with a poor prognosis and characterized by ATP-related metabolism. This cluster demonstrated interactions with MPs inclined to M2 polarization.