基于单细胞 RNA 测序的肝细胞癌焦亡相关基因预后模型构建及免疫微环境分析

Hepatocellular carcinoma (HCC) prognosis continues to be challenging due to tumor heterogeneity and dynamic immunosuppressive microenvironments. Although pyroptosis plays a critical role in tumor-immune interactions, its prognostic significance in HCC at single-cell resolution has not been systematically investigated.

We analyzed a publicly available single-cell RNA sequencing (scRNA-seq) data from 10 HCC tumors and paired adjacent tissue samples (60,496 cells) to elucidate pyroptosis-related gene (PRG) profiles. Differential expression and functional pathway analyses revealed PRG expression dynamics across cell subtypes. A LASSO-Cox prognostic model was developed using data from the liver hepatocellular carcinoma (LIHC) cohort of The Cancer Genome Atlas (TCGA) (n=365); the model was externally validated with International Cancer Genome Consortium (ICGC) datasets (n=231). Biological validation comprised reverse transcription quantitative polymerase chain reaction (RT-PCR) in HCC cell lines and immunohistochemical analysis of clinical specimens.

The scRNA-seq atlas identified 10 cellular clusters with enriched expression of 29 PRGs, primarily in natural killer cells, T lymphocytes, monocytes, and macrophages. The prognostic model developed in this study stratified patients into high-risk and low-risk categories based on eight significant genes, achieving area under the curve (AUC) values of 0.73, 0.65, and 0.69 for overall survival at one-year, two-year, and three-year intervals, respectively. Furthermore, external validation using data from the ICGC confirmed the prognostic model's discriminative ability. Notably, high-risk patients demonstrated enhanced sensitivity to immunotherapy, as indicated by decreased tumor immune dysfunction and exclusion (TIDE) scores and increased expression of the immune checkpoints PD-1 and CTLA4.

This study established a scRNA-seq-derived prognostic model based on PRGs, which offers insights into HCC immune landscape remodeling. The risk score and nomogram integrate tumor stages and pyroptosis-associated signatures, providing a clinical tool for personalized prognosis and therapeutic targeting.