Hepatocellular Carcinomas with Concomitant Mutations of TERT, TP53, and CTNNB1: Is There a Role for Artificial Intelligence?

TP53, CTNNB1, and TERT-promoter mutations are the most common driver mutations in hepatocellular carcinoma (HCC). The morphological and genetical HCC heterogeneities are difficult to discriminate by the eye of the pathologist. Here, we describe two rare cases of HCC with simultaneous co-mutation of all three of genes, which represent a poorly described occurrence in the literature. In these two cases, areas with different tumor grade and different β-catenin and Glutamine Synthetase expression (performed by automated immunohistochemistry) were observed. NGS analysis was performed in these different areas, to check for potential diversity of mutation burden on the different regions, but no differences were found: all micro-areas analyzed showed the co-presence of mutations in TP53, CTNNB1, and TERT. The evidence that all mutations were found in all the different areas analyzed by NGS leads to hypothesize that the tumor is not composed of different clones harboring different mutations. All the variants are harbored by the same neoplastic clone, albeit leading to different phenotypes. Mutation prediction Artificial Intelligence models could help the morpho-genetic classification of HCC in the future, since they can find variabilities not obvious to the human eye, with increased sensitivity, specificity and reproducibility.