Feed

This study aimed to identify a novel prognostic signature derived from an EGFR Tyrosine kinase inhibitors (TKI-resistant) macrophage subpopulation and to evaluate its clinical and therapeutic relevance in HCC. We utilized single-cell RNA sequencing data from HCC patients. An EGFR-TKI resistance score was calculated across all cell types. Macrophages, which exhibited the highest resistance score, were sub-clustered to identify the most resistant subpopulation. Marker genes from this sub-cluster were intersected with differentially expressed genes (DEGs) from the TCGA-LIHC cohort. A robust prognostic model was constructed. The model's performance was rigorously validated, and the signature was further characterized through multi-omics analysis and its correlation with immune checkpoint blockade (ICB) response and drug sensitivity. scRNA-seq analysis unequivocally identified macrophages as possessing the highest EGFR-TKI resistance score. We identified seven key prognostic genes: SLC41A3, DCAF13, PPM1G, NDC80, FAM83D, FUCA2, and UQCRH. A risk model built on these seven genes effectively stratified patients into high- and low-risk groups with significantly different overall survival (OS) in the TCGA cohort, a finding successfully validated in the independent GSE76427 cohort. A clinical nomogram integrating the risk score demonstrated excellent predictive accuracy, with AUC values for 1-, 3-, and 5-year OS of 0.816, 0.781, and 0.799, respectively. The low-risk group was associated with a favorable immune-infiltrated phenotype and was predicted to be more sensitive to immunotherapy. Conversely, the high-risk group exhibited distinct genomic features and was predicted to be more sensitive to specific targeted agents, including Navitoclax and Sorafenib. We identified and validated a novel 7-gene prognostic signature derived from a subpopulation of EGFR-TKI-resistant macrophages. This signature accurately predicts patient survival, offers insights into the molecular mechanisms of therapy resistance in HCC, and provides a promising tool for improved patient stratification and the development of personalized treatment strategies.