A Prognostic Signature Based on Immune-Related Genes Associated with Tumor Microenvironment for Cholangiocarcinoma

Abstract Background Several studies indicate immune related genes (IRGs) dramatically correlates with tumorigenesis and progress of cancer. However, the association IRGs and the value in predicting the prognosis of cholangiocarcinoma (CCA) patients remains unclear. Thus, we aimed to discover reliable biomarkers based on IRGs signature to predict survival in CCA. Methods Gene expression profiles of CCA patients were collected from TCGA and GEO database. Univariate, multivariate and Lasso (Least absolute shrinkage and selection operator) Cox analysis was used to identify IRGs signature. Subsequently, a nomogram that predicts outcomes for CCA patients based on IRGs signature was constructed to visualize prognosis. We also explored the relationships between tumor-infiltrating immune cells (TIICs) and IRGs. GDSC database was used to explore potential targeted drugs. Tissue samples of 23 CCA patients were collected and qRT-PCR were employed to validate the mRNA level of IRGs. Results 5 IRGs (AVPR1B, CST4, TDGF1, RAET1E and IL9R) were identified as robust indicators for OS. A prognostic model based on IRGs signature can significantly stratify patients into high-risk and low-risk groups. Meanwhile, ROC indicated that the risk model had satisfactory performance. We also observed that the risk score was an independent prognostic factor for OS and built a nomogram. The low-risk group had a significantly higher percentage of infiltrated macrophages M1 than the high-risk group. Through analyzing the GDSC database, the high-risk group demonstrated drug sensitivity to some targeted drugs such as AMG 706 and Thapsigargin. Finally, the results of qRT-PCR were consistent with bioinformatics analysis. Conclusions We established a novel IRGs signature to predict prognosis for CCA. The risk model exhibited excellent diagnostic efficiency in predicting OS.