Understanding the interactions between tumors and the host immune system holds great promise to uncover biomarkers for targeted therapies, predict the prognosis of patients and guide clinical treatment. However, the immune signatures of glioblastoma multiforme (GBM) remain largely unstudied in terms of systematic analyses. We aimed to classify GBM samples according to immune-related genes and complement the existing immunotherapy system knowledge. In this study, we designed a strategy to identify 3 immune subtypes representing 3 different immune microenvironments (M1-M3) and associated with prognosis. The 3 subtypes were significantly different in terms of specific immune characteristics (immune cell subpopulations, immune responses, immune cells, and checkpoint gene interactions). In additional, copy number variations and methylation changes were identified that correlated with genes related to a worse prognosis subtype in the microenvironment. More importantly, in M3 (worst prognosis subtype) and M2 (best prognosis subtype), the interaction between immune cells and checkpoint genes was different, which had an important effect on the prognosis. Finally, we used risk scores of immune cells and checkpoint genes to evaluate the prognosis of GBM patients and validated the results with 3 independent datasets. Disordered interactions between immune cells and checkpoint genes result in a change in the immune microenvironment and affects the prognosis of patients. We propose that a better understanding of the immune microenvironment of advanced cancers may provide new insights into immunotherapy.
Keywords: glioblastoma multiforme; immune microenvironment; immunotherapy; interaction changes; prognosis.
© 2019 The Authors. Cancer Medicine published by John Wiley & Sons Ltd.