AUTHOR=Wu Guozhi , Yang Yuan , Zhu Yu , Li Yemao , Zhai Zipeng , An Lina , Liu Min , Zheng Ya , Wang Yuping , Zhou Yongning , Guo Qinghong 

TITLE=Comprehensive Analysis to Identify the Epithelial–Mesenchymal Transition-Related Immune Signatures as a Prognostic and Therapeutic Biomarkers in Hepatocellular Carcinoma

JOURNAL=Frontiers in Surgery

VOLUME=Volume 8 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/surgery/articles/10.3389/fsurg.2021.742443

DOI=10.3389/fsurg.2021.742443

ISSN=2296-875X

ABSTRACT=Background: Hepatocellular carcinomas (HCC) is usually detected at an advanced stage due to the lack of effective prognostic model for early prediction. 
Aim: To construct a risk model composed of epithelial-mesenchymal transition (EMT)-related immune genes to evaluate prognosis, immune infiltration status and chemosensitivity.
Methods: We obtained transcriptome and clinical data of HCC samples from The Cancer Genome Atlas (TCGA) and The International Cancer Genome Consortium (ICGC) database. Pearson correlation analysis was applied to identify the differentially expressed EMT-related immune genes (DE-EMTi-genes), followed by construction of a risk model based on univariate cox regression and Least Absolute Shrinkage and Selection Operator (LASSO) regression. Additionally, ROC curves were plotted to compare the prognostic value of the newly established model with that of the previous model. Subsequently, the correlation between the signature and survival probability, immune cell infiltrating, immune related functions and efficacy of chemotherapeutics were analyzed by bioinformatics methods.
Results: Twenty of DE-EMTi-genes were correlated with prognosis of HCC and six were ultimately selected to construct the prognostic model. The AUC values for 1-, 2- and 3- year were 0.773, 0.721 and 0.673, respectively. Survival analysis, whether stratified according to clinicopathological factors or not, suggested that the prognosis of low-score group was superior to high-score group. Moreover, univariate and multivariate analysis indicated that risk score (HR 5.071, 95% CI 3.050, 8.432; HR 4.396, 95% CI 2.624, 7.366; P<0.001) and stage (HR 2.500, 95% CI 1.721, 3.632; HR 2.111, 95% CI 1.443, 3.089; P<0.001) served as independent predictive factors in HCC. In high-risk group, Macrophages, NK cells and Treg cells were significantly enriched in HCC samples. Additionally, patients with a high-risk score might be more sensitive to cisplatin, doxorubicin, etoposide, gemcitabine, and mitomycin C.
Conclusions: We established a reliable EMT-related immune genes-based prognostic signature, which may hold promise for clinical prediction.