Introduction HCV remains the commonest cause of HCC related deaths in the west. Dysplastic nodules (DN) harbour highly relevant genomic information based on its strong potential to progress to HCC. Copy number alterations (CNA) have not been previously studied in a large cohort of HCV related DN. In this study we used next generation sequencing (NGS) to identify potential drivers of oncogenesis in a cohort of HCV related DNs and HCCs.
Methods We performed low coverage next generation DNA sequencing of 116 HCCs and 49 DNs from 63 HCV infected patients to investigate their CNA profiles. Patients with a mixed aetiology of liver disease were excluded. Significant copy number gain and loss peaks were identified using GISTIC 2.0.
To identify driver from passenger alterations, we stratified the peaks according to their amplitude and frequency within included samples. We also identified peaks that were inclusive of genes causally implicated in cancer, highly associated with cancer or involved in major cell signalling pathways.
We further analysed the HCC-related peaks and correlated each of them to patient and tumour clinical features. This correlation was corrected for false discovery rate (FDR) using the Benjamini-Hochberg procedure.
Results We identified 86 and 79 significant CNA peaks in HCC and DN respectively. The peaks were further stratified as described above which lead to recognition of a 9 gene signature highly implicated in early HCV related HCC. This included losses of ROCK1, CARS, SBDS, DUX4 and gains of NCOA2, MYC, TPO, YWHAZ and MCM4.
Our data confirmed the alteration of 51 other genes in HCC but not DN. This included gains of CCND1, NFKB1/2, MUC1, FGF2/4/19 and CKS1B and losses of CDKN2A/B, KRAS, PTEN, RB1, NOTCH1, CCNA2, BMPR1B which have previously been implicated in hepatocarcinogenesis.
Out of 86 HCC-related peaks, 16/86 had significant associations with clinical features after correction for FDR. Six of the above 16 peaks overlapped with 6 out of the 9 gene signature we previously identified. Genomic features strongly correlated with tumour differentiation, number of tumours per liver and overall survival. No association was found between HCV genotype or PCR quantity.
Conclusion Analysis of DNA copy number profiles of a large number of HCV related DN and HCC using NGS revealed a signature of 9 genes likely to be implicated as drivers in the early development of this cancer. Our study generates a valid hypothesis and further studies are needed to confirm causality.
Disclosure of Interest W. Fateen: None Declared, H. Wood: None Declared, S. Berri Employee of: Illumina, H. Thygesen: None Declared, J. Wyatt: None Declared, M. El- Meteini: None Declared, C. Millson: None Declared, P. Quirke: None Declared