A signature of 33 immune‐related gene pairs predicts clinical outcome in hepatocellular carcinoma

OBJECTIVE: Hepatocellular carcinoma (HCC) has become the second most common tumor type that contributes to cancer‐related death worldwide. The study aimed to establish a robust immune‐related gene pair (IRGP) signature for predicting the prognosis of HCC patients. METHODS: Two RNA‐seq datasets (The Cancer Genome Atlas Program and International Cancer Genome Consortium) and one microarray dataset (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE14520) were included in this study. We used a series of immune‐related genes from the ImmPort database to construct gene pairs. Lasso penalized Cox proportional hazards regression was employed to develop the best prognostic signature. We assigned patients into two groups with low immune risk and high immune risk. Then, the prognostic ability of the signature was evaluated by a log‐rank test and a Cox proportional hazards regression model. RESULTS: After 1000 iterations, the 33‐immune gene pair model obtained the highest frequency. As a result, we chose the 33 immune gene pairs to establish the immune‐related prognostic signature. As we expected, the immune‐related signature accurately predicted the prognosis of HCC patients, and high‐risk groups showed poor prognosis in the training datasets and testing datasets as well as in the validation datasets. Furthermore, the immune‐related gene pair (IRGP) signature also showed higher predictive accuracy than three existing prognostic signatures. CONCLUSION: Our prognostic signature, which reflects the link between the immune microenvironment and HCC patient outcome, is promising for prognosis prediction in HCC.