Six-gene T cell signature predicts survival in hepatocellular carcinoma

Researchers have built a prognostic model for hepatocellular carcinoma (HCC) that draws on T cell features identified through single-cell transcriptomics. The study, published in Discover Oncology on 27 October 2026, analyzed single-cell RNA-seq data from six HCC patients to identify T cell-specific genes, then used differential expression and Cox/LASSO regression in the TCGA-LIHC cohort to construct a six-gene risk model that was validated in the GSE76427 dataset.

In the TCGA cohort, the model stratified patients into high-risk and low-risk groups with a log-rank p value below 0.001 and achieved a C-index of 0.709. The area under the ROC curve for 1-year, 3-year and 5-year overall survival was 0.790, 0.740 and 0.750, respectively, and the risk score emerged as an independent prognostic factor (p < 0.001). When combined into a clinical nomogram, predictive performance improved further, reaching a C-index of 0.78. In the external GSE76427 validation set, the model again separated patients by survival (log-rank p = 0.021) and outperformed several previously published models.

The analysis identified 751 T-cell-related differentially expressed genes in HCC, enriched primarily in immune response and T-cell activation pathways. High-risk patients showed enrichment of pro-proliferative pathways, and their immune microenvironment was characterized by elevated TIDE scores, restricted T-cell infiltration, enrichment of myeloid-derived suppressor cells and regulatory T cells, and high expression of PD-L1 and multiple other immune checkpoints. The low-risk group, by contrast, displayed higher IPS scores, suggesting greater potential responsiveness to immunotherapy. Although tumor mutation burden was higher in the high-risk group, the risk score remained an independent prognostic factor regardless of TMB status.

One of the six genes, SLC4A10, was experimentally validated in the laboratory. Quantitative RT-PCR confirmed higher SLC4A10 expression in adjacent non-tumor tissues compared with tumor tissue, and functional assays in vitro showed that low expression promoted HCC cell proliferation, migration and clonogenic growth, collectively supporting a role for SLC4A10 as a tumor suppressor gene. The authors nominate SLC4A10 as a potential therapeutic target, though these findings come from cell-based experiments and require further study before any clinical application.

The study’s conclusion emphasizes that the model enables effective risk stratification in HCC and delineates distinct immune and metabolic landscapes between risk groups, offering new insight into tumor biology. Because the model was developed and validated in retrospective cohorts, and the single-cell analysis was based on only six patients, prospective validation would be needed before clinical use. The authors note that existing prognostic models often lack an immunological perspective, and their approach was designed to produce a biologically interpretable signature for more precise risk stratification in this highly heterogeneous cancer.

Subject of Research: A single-cell transcriptomics-based six-gene T cell prognostic signature for hepatocellular carcinoma

Article Title: Development and validation of a six-gene T cell-derived prognostic signature based on single-cell transcriptomics for hepatocellular carcinoma

Article References: Wu, K., Wu, J., & Wang, D. (2026). Development and validation of a six-gene T cell-derived prognostic signature based on single-cell transcriptomics for hepatocellular carcinoma. Discover Oncology. https://doi.org/10.1007/s12672-026-06161-z

Image Credits: AI Generated

DOI: 10.1007/s12672-026-06161-z

Keywords: hepatocellular carcinoma, single-cell transcriptomics, prognostic signature, T cells, TCGA-LIHC, SLC4A10, tumor microenvironment, immunotherapy, nomogram, risk stratification, bioinformatics, Discover Oncology