A novel genomic-clinicopathologic nomogram to improve prognosis prediction of hepatocellular carcinoma

There is a lack of precise and clinical accessible model to predict the prognosis of hepatocellular carcinoma (HCC) in clinic practice currently. Here, an inclusive nomogram was developed by integrating genomic markers and clinicopathologic factors for predicting the outcome of patients with HCC. A total of 365 samples of HCC were obtained from the Cancer Genome Atlas (TCGA) database. The LASSO analysis was carried out to identify HCC-related mRNAs, and the multivariate Cox regression analysis was used to construct a genomic-clinicopathologic nomogram. As results, 9 mRNAs were finally identified as prognostic indicators, including RGCC, CDH15, XRN2, RAB3IL1, THEM4, PIF1, MANBA, FKTN and GABARAPL1, and used to establish a 9-mRNA classifier. Additionally, an inclusive nomogram was built up by combining the 9-mRNA classifier (P < 0.001) and clinicopathologic factors including age (P = 0.006) and metastasis (P < 0.001) to predict the mortality of HCC patients. Time-dependent receiver operating characteristic, index of concordance and calibration analyses indicated favorable accuracy of the model. Decision curve analysis suggested that appropriate intervention according to the established nomogram will bring net benefit when threshold probability was above 25%. The genomic-clinicopathologic model could be a reliable tool for predicting the mortality, helping determining the individualized treatment and probably improving HCC survival.