Stage-Specific Transcriptome Landscape of Hepatocellular Carcinoma: Insights from Super and Poor Survivors with Prognostic Signature Identification

Abstract

BackgroundThe prognosis for hepatocellular carcinoma (HCC) patients can vary significantly even within the same clinical stage. This study aims to characterize the stage-specific transcriptomic landscape of super survivors with HCC and identify a prognostic gene signature. MethodsWe analyzed data from The Cancer Genome Atlas (TCGA) for 76 matched super survivors (alive at 5 years) and poor survivors (deceased within 1 year) with HCC. Gene set enrichment analysis stratified by stage was conducted and a key prognostic gene signature was developed. The signature was then applied to the full TCGA cohort and independently validated in International Cancer Genome Consortium (ICGC) data. FindingsStage-specific transcriptomic profiling revealed distinct patterns in super survivors. Stage I and II showed positive enrichment in immune response pathways, while stage III exhibited enhanced catabolic activities and reduced glycolysis. Across all stages, cell cycle processes were less active in super survivors. A 19-gene signature, incorporating immune, metabolic, and cell cycle genes, accurately distinguished super from poor survivors with 91% accuracy (69/76). The signature reliably predicted overall survival in both the verification cohort (1-year, 3-year, and 5-year AUROCs of 0.82, 0.80, and 0.78) and the independent validation cohort (1-year and 3-year AUROCs of 0.80 and 0.83). Consistent AUROCs were observed across all tumor stages. InterpretationOur findings reveal a dynamic shift in HCC progression, with immune dominance in early stages and metabolic dominance in later stages, alongside reduced cell cycle activity across all stages. Integrating stage-specific transcriptomic profiles offers a promising approach to enhancing individualized management for HCC patients.