Prediction of recurrence and survival in hepatocellular carcinoma based on two Cox models mainly determined by FoxP3+ regulatory T cells.

Hepatocellular carcinoma (HCC) is an aggressive disease with poor prognosis and limited methods to predict patient survival. Immune cells infiltrating tumors is known to impact clinical outcome. Here, we investigated the prognostic significance of immune infiltration within the tumor microenvironment in 245 specimens from two independent cohorts by immunohistochemical analyses. A Cox regression model was constructed using a training cohort and validated in an independent cohort. The diagnostic accuracy was evaluated by receiver operating characteristic curve. The activation, function, and chemotaxis of intratumoral regulatory T cells (Treg) were analyzed using flow cytometry, quantitative PCR, and chemotaxis assay. We identified that the proportion of FoxP3(+) cells within tumors is negatively associated with patient prognosis, whereas the proportion of interleukin (IL)-17(+) cell and the number of trypase(+) cells are positive predictor. The two Cox models, composed of independent predictors in multivariate analysis, provided a high diagnostic accuracy of prognosis for patients with HCC. The proportion of FoxP3(+) cells showed the most significant predictive power, with the highest Cox score in the two models. Furthermore, we found Tregs from tumor with high FoxP3(+) proportion were more active and powerful than the counterparts from tumor with low FoxP3(+) proportion. In conclusion, two Cox models are established that have considerable clinical value in predicting tumor recurrence and survival of patients with HCC, respectively. In the both models, the proportion of Tregs among CD4(+) T cells plays a central role.