Intratumour ploidy heterogeneity and clonal evolution in hepatocellular carcinoma

Hepatocellular carcinoma (HCC), the fourth leading cause of cancer-related death worldwide, still lacks robust stratification biomarkers to improve prognostication and prediction of treatment response. Polyploidization or hyperploidization, an early event observed in many solid tumours, may represent such a biomarker. Here, we developed a novel high-throughput, in situ quantitative digital pathology approach to assess nuclear ploidy in HCC and evaluate its prognostic relevance.

Quantitative digital pathology-based ploidy assessment, whole-exome sequencing (WES), and bulk RNA sequencing were performed on 111 human HCC samples. To investigate the clonal evolution of hyperploid clones within tumours, 3' mRNA sequencing of macrodissected FFPE tumour tissue and spatial transcriptomics were employed.

Using whole-slide imaging-based quantitative digital pathology, we classified HCCs into diploid (DP) and hyperploid (HP) subtypes based on nuclear morphometry and mapped the spatial distribution of hyperploid nuclei within tumours. Transcriptomic analysis associated HP-HCCs with a proliferative molecular subclass characterized by poor prognosis and stemness features. Spatial reconstruction of nuclear ploidy further identified mixed DP/HP-HCCs, in which hyperploid nuclei formed discrete clusters within an otherwise diploid tumour background. Notably, these nascent hyperploid foci shared a transcriptomic profile with HP-HCCs and exhibited signatures of genomic instability. Importantly, spatial transcriptomics of DP/HP-HCCs provided direct evidence that hyperploid tumour hepatocytes arise through whole-genome doubling of diploid cells.

These findings highlight the prognostic value of nuclear ploidy profiling in HCC and suggest that hyperploid HCCs evolve clonally from intratumoural hyperploid foci. Hyperploidization may therefore represent a key mechanism driving genomic instability in liver tumourigenesis.

Modification of ploidy play a substantial role in cancer genome evolution. In this study, we created a new digital pathology method to accurately measure the hepatocyte ploidy profile of HCC. This powerful tool reveals three distinct tumour types: diploid (DP), hyperploid (HP), and mixed diploid/hyperploid (DP/HP) HCCs. By clearly defining these groups, we gain crucial, new insights into the "branching" evolution of HCC tumours. This study should pave the way for novel diagnostic and therapeutic strategies.