Integrating Machine Learning and Single‐Cell Analysis to Reveal the Diagnostic and Therapeutic Value of Regulated Cell Death Mechanisms in Hepatocellular Carcinoma

ABSTRACT

Hepatocellular carcinoma (HCC) treatment faces significant challenges, particularly in tumor growth, metastasis, and drug resistance. While several predictive models exist, effective models that accurately predict patient prognosis and guide targeted therapy decisions remain insufficient. Regulated cell death (RCD) pathways play a pivotal role in the development and progression of various cancers, offering potential prognostic indicators and biomarkers of drug sensitivity for HCC patients. We analyzed multi‐cohort transcriptomic data (TCGA, GSE14520, ICGC) and single‐cell RNA sequencing data (GSE149614) to identify differentially expressed RCD‐related genes (DEGs). A prognostic model, the Regulated Cell Death Index (RCDI), was constructed using machine learning algorithms to stratify HCC patients into high‐ and low‐RCDI groups. Single‐cell analysis was employed to examine tumor microenvironment heterogeneity between these groups, and drug sensitivity analysis assessed differences in immune therapy, targeted therapy, and chemotherapy responses based on RCDI subgroups. RCDI was significantly associated with poor clinical features and shorter overall survival, with results validated across all cohorts. Enrichment analysis revealed that high RCDI is correlated with key cancer‐related pathways, including the PI3K‐Akt pathway and cell cycle regulation. High RCDI was also associated with immune cell infiltration and the expression of immune checkpoint molecules, as validated through single‐cell RNA sequencing. Patients with high RCDI exhibited higher sensitivity to several targeted therapies, including Vorinostat and Trametinib. Further prioritization analyses identified EEF1E1, ITGB3BP, and SPP1 as promising candidate biomarkers with potential diagnostic and prognostic relevance. The RCDI model effectively stratifies HCC patients based on RCD‐related molecular features, providing a valuable tool for predicting survival and therapeutic responses. The identification of key genes offers new insights into the molecular mechanisms of HCC and potential therapeutic targets.