Transcriptomic classification and clinicopathologic correlation of hepatocellular neoplasms with steatotic features...

Differentiating early-stage hepatocellular carcinoma (HCC) from hepatocellular adenoma (HCA), particularly in the context of steatosis, remains a diagnostic challenge. Accurate distinction is essential for appropriate clinical management, yet reliable ancillary assays are limited. This study focuses on hepatocellular neoplasms with steatotic change in non-cirrhotic livers, aiming to minimise confounding factors from cirrhosis and to identify genetic alterations specific to tumour biology. Using total RNA sequencing, we analysed the gene expression profiles of 12 HCA cases, ten HCC cases (including eight with steatotic change) and seven paired non-neoplastic liver samples. Transcriptomic profiles were subsequently correlated with histopathologic features and clinical outcomes. Unsupervised clustering segregated HCC from HCA and non-neoplastic liver, indicating distinct transcriptional signatures. Three histologically ambiguous hepatocellular neoplasms were also evaluated, and their clustering patterns aligned with their long-term clinical behaviour, suggesting the potential clinical utility of transcriptomic classification in diagnostically uncertain cases. Differential expression analysis identified both protein-coding and non-coding genes that may serve as candidate biomarkers for future RNA-based or immunohistochemical assays. Pathway analysis revealed distinct oncogenic and metabolic signalling cascades differentiating HCC from HCA. These findings demonstrate that transcriptomic profiling offers a promising molecular approach to distinguishing HCC from HCA and assists in the classification of indeterminate hepatocellular neoplasms. With further validation, these molecular signatures and their surrogate biomarkers have the potential to enhance diagnostic accuracy, refine histologic criteria, guide clinical management and improve our understanding of hepatocarcinogenesis.