Oncotarget

Research Papers:

Metabolomics profiling in plasma samples from glioma patients correlates with tumor phenotypes

Hua Zhao _, Amy B. Heimberger, Zhimin Lu, Xifeng Wu, Tiffany R. Hodges, Renduo Song and Jie Shen

PDF  |  HTML  |  Supplementary Files  |  How to cite

Oncotarget. 2016; 7:20486-20495. https://doi.org/10.18632/oncotarget.7974

Metrics: PDF 3109 views  |   HTML 3349 views  |   ?  


Abstract

Hua Zhao1, Amy B. Heimberger2, Zhimin Lu3, Xifeng Wu1, Tiffany R. Hodges2, Renduo Song1, Jie Shen1

1Department of Epidemiology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA

2Division of Neuro-Surgery, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA

3Department of Neuro-Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA

Correspondence to:

Hua Zhao, e-mail: [email protected]

Keywords: metabolomics, glioma, tumor phenotype

Received: September 17, 2015     Accepted: February 13, 2016     Published: March 07, 2016

ABSTRACT

Background: Tumor-based molecular biomarkers have redefined in the classification gliomas. However, the association of systemic metabolomics with glioma phenotype has not been explored yet.

Methods: In this study, we conducted two-step (discovery and validation) metabolomic profiling in plasma samples from 87 glioma patients. The metabolomics data were tested for correlation with glioma grade (high vs low), glioblastoma (GBM) versus malignant gliomas, and IDH mutation status.

Results: Five metabolites, namely uracil, arginine, lactate, cystamine, and ornithine, significantly differed between high- and low-grade glioma patients in both the discovery and validation cohorts. When the discovery and validation cohorts were combined, we identified 29 significant metabolites with 18 remaining significant after adjusting for multiple comparisons. Those 18 significant metabolites separated high- from low-grade glioma patients with 91.1% accuracy. In the pathway analysis, a total of 18 significantly metabolic pathways were identified. Similarly, we identified 2 and 6 metabolites that significantly differed between GBM and non-GBM, and IDH mutation positive and negative patients after multiple comparison adjusting. Those 6 significant metabolites separated IDH1 mutation positive from negative glioma patients with 94.4% accuracy. Three pathways were identified to be associated with IDH mutation status. Within arginine and proline metabolism, levels of intermediate metabolites in creatine pathway were all significantly lower in IDH mutation positive than in negative patients, suggesting an increased activity of creatine pathway in IDH mutation positive tumors.

Conclusion: Our findings identified metabolites and metabolic pathways that differentiated tumor phenotypes. These may be useful as host biomarker candidates to further help glioma molecular classification.


Creative Commons License All site content, except where otherwise noted, is licensed under a Creative Commons Attribution 4.0 License.
PII: 7974