Unraveling the underlying pathogenic factors driving nonalcoholic steatohepatitis and hepatocellular carcinoma: an in-depth analysis of prognostically relevant gene signatures in hepatocellular carcinoma

Abstract

Abstract Background Nonalcoholic steatohepatitis (NASH) is a progressive manifestation of nonalcoholic fatty liver disease (NAFLD) that can lead to fibrosis, cirrhosis, and hepatocellular carcinoma (HCC). Despite the growing knowledge of NASH and HCC, the association between the two conditions remains to be fully explored. Bioinformatics has emerged as a valuable approach for identifying disease-specific feature genes, enabling advancements in disease prediction, prevention, and personalized treatment strategies. Materials and methods In this study, we utilized CellChat, copy number karyotyping of aneuploid tumors (CopyKAT), consensus Non-negative Matrix factorization (cNMF), Gene set enrichment analysis (GSEA), Gene set variation analysis (GSVA), Monocle, spatial co-localization, single sample gene set enrichment analysis (ssGSEA), Slingshot, and the Scissor algorithm to analyze the cellular and immune landscape of NASH and HCC. Through the Scissor algorithm, we identified three cell types correlating with disease phenotypic features and subsequently developed a novel clinical prediction model using univariate, LASSO, and multifactor Cox regression. Results Our results revealed that macrophages are a significant pathological factor in the development of NASH and HCC and that the macrophage migration inhibitory factor (MIF) signaling pathway plays a crucial role in cellular crosstalk at the molecular level. We deduced three prognostic genes (YBX1, MED8, and KPNA2), demonstrating a strong diagnostic capability in both NASH and HCC. Conclusion These findings shed light on the pathological mechanisms shared between NASH and HCC, providing valuable insights for the development of novel clinical strategies.

Keywords

• Nonalcoholic steatohepatitis (NASH)
• Hepatocellular carcinoma (HCC)
• Macrophages
• Scissor algorithm
• Clinical predictive mode