Research Papers:

Identification of a five B cell-associated gene prognostic and predictive signature for advanced glioma patients harboring immunosuppressive subtype preference

Chuanbao Zhang, Jiye Li, Haoyuan Wang and Sonya Wei Song _

PDF  |  HTML  |  Supplementary Files  |  How to cite

Oncotarget. 2016; 7:73971-73983. https://doi.org/10.18632/oncotarget.12605

Metrics: PDF 1842 views  |   HTML 2175 views  |   ?  


Chuanbao Zhang1,*, Jiye Li1,2,3,*, Haoyuan Wang4, Sonya Wei Song1,2,3

1Beijing Neurosurgical Institute, Capital Medical University, TiantanXili, Dongcheng District, Beijing 100050, China

2Beijing Institute for Brain Disorders, Youanmen, Beijing, 100069, China

3Center for Brain Disorders Research, Capital Medical University, Youanmen, Beijing, 100069, China

4Department of Neurosurgery, Zhujiang Hospital, Southern Medical University, Guangzhou, 510282, China

*These authors contributed equally to this work

Correspondence to:

Sonya Wei Song, email: [email protected], [email protected]

Keywords: immune system, tumor-infiltrating B cells, prognosis, biomarker, glioma

Received: April 11, 2016     Accepted: September 27, 2016     Published: October 12, 2016


High grade gliomas contribute to most brain tumor mortality. A few studies reported that the immune system affected glioma development, and immune biomarkers helped understand the disease and formulate effective immunotherapy for patients. Currently, no B lymphocyte-based prognostic signature was reported in gliomas. By applying 78 B cell lineage-specific genes, we conducted a whole-genome gene expression analysis in 782 high grade gliomas derived from three independent datasets by Cox regression analysis and risk score method for signature identification, and then used Gene Ontology, Gene Set Enrichment Analysis, and other statistical methods for functional annotations of the signature-defined differences. We developed a five B cell-associated gene signature for prognosis of high grade glioma patients, which is independent of clinicopathological and genetic features. The signature identified high risk patients suitable for chemoradiotherapy, whereas low risk patients should rule out chemotherapy with radiotherapy only. We found that tumors of TCGA Mesenchymal subtype and wild type IDH1 were preferentially stratified to the high risk group, which bore strong immunosuppressive microenvironment, while tumors of TCGA Proneural subtype and mutated IDH1 were significantly accumulated to the low risk group, which exhibited less immunosuppressive state. The five B cell-associated gene signature predicts poor survival of high risk patients bearing strong immunosuppression and helps select optimal therapeutic regimens for glioma patients.

Creative Commons License All site content, except where otherwise noted, is licensed under a Creative Commons Attribution 3.0 License.
PII: 12605