A Nomogram For Predicting HCC Patients’ Overall Survival Based On Double Hub Genes and Other Clinical Risk Factors.

Abstract Background: Hepatocellular carcinoma (HCC) is one of the series malignant aggressive disease which shows elusive biological behavior and terrible prognosis. It is inadequate for the single evaluation system such as tumor-node-metastasis (TNM) staging system to predict the overall survival (OS) in HCC patients. Here, we conducted this study to identify prognosis-related genes, so as to form a reliable prognostic assessment model.Results: From three datasets in the GEO database, we identified two hub genes cytoskeleton associated protein 2 (CKAP2) and forkhead box M1 (FOXM1) among 348 differentially expressed genes (DGEs) that were differentially expressed between HCC and normal samples. The function analysis of those DEGs are enriched in cell division process (e.g., mitotic nuclear division, nuclear chromosome segregation) and metabolic process (e.g., organic acid catabolic process). Then, we established a two-gene model, that extremely distinguished the population at risk of liver cancer to high-risk and low-risk and was even viable in the TNM stage i-ii and iii-iv, vascular invasion and non-invasion subgroups (all P<0.05). Next, a nomogram was set out combined with the two hub genes and clinical risk factors, and the predictive power of the nomogram performed more outstanding than the gene expression or clinical parameters alone.Conclusions: Our two-gene-based evaluation system effectively filtered out the high-risk HCC patients, and could potentially be used for clinical decision-making and individualized management of particular HCC patients.