Clonal diversity underpins distinct modes of recurrence in hepatocellular carcinoma: the PLANet cohort study

Background Hepatocellular carcinoma (HCC) has a high postsurgery recurrence rate yet the underlying molecular mechanisms remain poorly understood. Objective Using multi-region sampling of paired primary and recurrent tumours, we examine the molecular and evolutionary events underlying HCC recurrence to inform neoadjuvant and adjuvant treatment strategies. Design Within the Precision Medicine in Liver Cancer across an Asia-Pacific Network (PLANet) prospective surgical cohort, multiregion whole genome sequencing (WGS) (408 samples) and RNA sequencing (406 samples) were carried out on primary tumours from 106 patients with HCC. A recurrence cohort of 24 patients with paired primary and recurrent tumours was assembled, generating 166 WGS and 113 RNA-seq samples. Tumour phylogenies were constructed to define seeding patterns. Integrated transcriptomic analyses were done to characterise the RNA subtype and immune microenvironment. Results We identified two main recurrence patterns, polyclonal and monoclonal seeding. More than half of intrahepatic recurrences were polyclonal, associated with early recurrence, high phenotypic plasticity and a regulatory T cell enriched immunosuppressive microenvironment. In comparison, most distant metastases showed monoclonal origins and appeared to come from a dominant, aggressive C5/C6 subclone within the primary tumour. Molecular features in the primary tumours determine different modes of recurrence. Polyclonal intrahepatic recurrence tumours also showed increased genomic instability and higher sensitivity associated with immune-checkpoint blockade. Furthermore, we built a multi-omics model that can predict HCC recurrence with high accuracy (area under the curve=0.86), identifying high-risk patients despite curative intent resection. Conclusion These findings show the impact of clonal structure and the tumour microenvironment on recurrence, offering insights that may help improve patient risk assessment and treatment planning.