The aim of this study is to explore the effects of COSMC on the prognosis of hepatocellular carcinoma (HCC), and establish a novel model with improved predictive capacity.

Ninety-two patients diagnosed with HCC from 2006 to 2010 in our hospital were recruited to analyze the correlation between COSMC expression and prognosis. Cellular experiments were performed to verify the anti-tumor effects of COSMC. A predictive model was established based on the risk factors from multiple COX regression analysis. After validation, the novel model was compared with the conventional model in terms of capacity of predicting the prognosis.

The expression of COSMC was lower in tumor tissues than in normal tissues and inhibited HCC migration in cells. Besides the expression of COSMC was significantly negatively correlated with overall survival (OS) in HCC, regression analysis showed that COSMC expression, vascular invasion, and TNM stage were prognostic risk factors. Our novel model comprising these three elements was established and validated. Besides the good fit of the calibration curves, a higher concordance index (C-index) for OS (P=0.011) as well as better decision curve analysis (DCA) and survival curves for both disease-free survival (DFS) and OS suggested the superiority of this novel model compared with conventional TNM staging in predicting the prognosis of HCC patients.

We established a novel model by integrating the expression of COSMC, vascular invasion, and TNM stage, and found that it was better able to predict survival in patients with HCC compared with conventional TNM staging.