Brain Tumor Typing and Therapy Using Combined Ex Vivo Magnetic Resonance Spectroscopy and Molecular Genomics
A novel approach was developed that combines biomarkers detected with magnetic resonance spectroscopy (MRS) and molecular genomics to improve the typing and prognostication of biospecimens in clinical medicine. Metabolite and genome wide profiles from 55 biopsies from subjects with brain tumors were analyzed with a classification algorithm that produces unique tumor fingerprints. We found that the fusion of 15 gene expressions and 15 MRS metabolites were able to distinguish tumor categories and predict survival better than when either dataset was used alone. Our approach improves the typing and understanding of the complexity of human brain tumors, generates testable hypotheses regarding neoplasia and promises to guide human brain tumor therapy. Our results further elucidate the biology of brain malignancy subtypes in brain tumor patients, and increase the overall potential for success of future studies that combine clinical MRI, MRS and MR imaging of gene expression in vivo.
KeywordsNuclear Magnetic Resonance Support Vector Machine Magnetic Resonance Spectroscopy Linear Discriminant Analysis Independent Component Analysis
- Astrakas L, Blekas KD, Constantinou C, Andronesi OC, Mindrinos MN, Likas AC, Rahme LG, Black PM, Marcus KJ, Tzika AA (2011) Combining magnetic resonance spectroscopy and molecular genomics offers better accuracy in brain tumor typing and prediction of survival than either methodology alone. Int J Oncol 38:1113–1127PubMedGoogle Scholar