Identification of metabolites in plasma for predicting survival in glioblastoma

Jie Shen, Renduo Song, Tiffany R. Hodges, Amy B. Heimberger, Hua Zhao*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

13 Scopus citations

Abstract

Circulating metabolomics profiling holds prognostic potential. However, such efforts have not been extensively carried out in glioblastoma. In this study, two-step (training and testing) metabolomics profiling was conducted from the plasma samples of 159 glioblastoma patients. Metabolomics profiling was tested for correlation with 2-year overall and disease-free survivals. Arginine, methionine, and kynurenate levels were significantly associated with 2-year overall survival in both the training and testing sets. In the combined sets, elevated levels of arginine and methionine were associated with a 34% and 37% increased probability whereas kynurenate was associated with a 55% decreased probability of 2-year overall survival. These three metabolites were also significantly associated with 2-year disease-free survival. Risk scores were generated using the linear combination of levels of these significant metabolites. Glioblastoma patients with a high-risk score exhibited a 2.41-fold decreased probability of 2-year overall survival (hazard ratio (HR) = 2.41; 95% Confidence Interval (CI) = 1.20-4.93) and a 3.17-fold decreased probability of 2-year disease free survival (HR = 3.17, 95%CI = 1.42-7.54) relative to those with a low-risk score. In conclusion, we identified a unique plasma metabolite profile that is predictive of glioblastoma prognosis.

Original languageEnglish (US)
Pages (from-to)1078-1084
Number of pages7
JournalMolecular Carcinogenesis
Volume57
Issue number8
DOIs
StatePublished - Aug 2018
Externally publishedYes

Keywords

  • glioblastoma
  • metabolomics
  • survival

ASJC Scopus subject areas

  • Molecular Biology
  • Cancer Research

Fingerprint

Dive into the research topics of 'Identification of metabolites in plasma for predicting survival in glioblastoma'. Together they form a unique fingerprint.

Cite this