Identification of New Biomarker for Prediction of Hepatocellular Carcinoma Development in Early-Stage Cirrhosis Patients.

BACKGROUND

Liver cirrhosis is one of the major drivers of hepatocellular carcinoma (HCC). In the present study, we aimed to identify and validate new biomarker for early prediction of HCC development in early-stage cirrhosis patients.

METHODS

mRNA expression and clinical parameters of GSE63898, GSE89377, GSE15654, GSE14520, and TCGA-HCC cohort and ICGC-HCC cohort were downloaded for analysis. Wilcoxon test was performed to identify DEGs. Univariate and multivariate Cox regression analysis were used to develop the risk signature, and ROC analysis was performed to analyze the predictive accuracy and sensitivity of the risk signature.

RESULTS

There were 42 DEGs (including 28 upregulated genes and 14 downregulated genes) found in early-stage liver cirrhosis patients before developing HCC from GSE1565442. Then, a risk signature consisting of 8 DEGs could effectively classify early-stage cirrhosis patients into high-risk group with shorter HCC development time and low-risk group with longer HCC development time from GSE15654. Multivariate Cox analysis indicated that the risk signature was an independent prognostic factor for the prediction of HCC development and ROC analysis showed that the signature exhibited good predictive efficiency in predicting 2-, 5-, and 10-year HCC development. Mechanistically, significantly higher proportions of CD8 T cells were found to be enriched in cirrhosis patients with low risk score, and higher CD8 T cells were associated with longer HCC development time. Besides, the signature was an independent prognostic factor for poorer prognosis of early-stage liver cirrhosis patients of GSE15654. Moreover, the signature could also separate HCC patients from healthy controls and was also associated with the poorer prognosis of HCC patients from three HCC cohorts. Finally, we also identified HDAC inhibitors, such as trichostatin A, to be a potential chemopreventive treatment for the prevention of HCC development by targeting risk signature based on CMap analysis.

CONCLUSION

A risk signature was developed and validated for early prediction of HCC development, which may be a useful tool to set up individualized follow-up interval schedules.