BIMG-14. IDENTIFICATION OF IDH MUTATION STATUS USING PROTON MR SPECTROSCOPY AND MASS SPECTROMETRY: A STUDY OF 178 GLIOMAS. (25th March 2021)

• Record Type: Journal Article
• Authors: Sacli-Bilmez, Banu; Akin-Levi, Cansu; Danyeli, Ayça Ersen; Yakicier, Cengiz; Pamir, M Necmettin; Özduman, Koray; Dincer, Alp; Can, Ozge; Ozturk-Isik, Esin
• Is Part Of: Neuro-oncology advances. Volume 3(2021)Supplement 1
• Page Start: i4
• Page End: i4
• Publication Date: 2021-03-25
• Subjects: 616.99481
• DOI: 10.1093/noajnl/vdab024.013
• Languages: English
• ISSNs: 2632-2498

Abstract

Abstract: IDH mutation, a key factor in predicting glioma prognosis, alters the levels of some metabolites in brain, including 2-hydroxyglutarate (2HG), glutamine (Gln), and glutathione (GSH). While proton MR spectroscopy ( 1 H-MRS) enables in-vivo detection of these metabolites, liquid chromatography-mass spectrometry (LC-MS/MS) is a sensitive in-vitro method to measure absolute metabolite concentrations. This study aims to examine the correlation of metabolic concentrations measured using 1 H-MRS and LC-MS/MS in gliomas, and to detect IDH mutation with machine learning based on 1 H-MRS and LC-MS/MS metabolic intensities. The patient cohort included 178 glioma patients (111M/67F, mean age:44.09±13.95 years, 100 IDH-mut, 78 IDH-wt). The patients were scanned pre-surgery by a 3T MR scanner with a 32-channel head coil. 1 H-MRS was obtained from a manually placed region of interest with no necrosis, edema, and hemorrhage, using a Point Resolved Spectroscopy (PRESS) sequence (TR/TE=2000/30ms). LCModel software was used for quantification of eighteen metabolites of 1 H-MRS data. Metabolite concentrations including creatine (Cr), choline (Cho), Gln, glutamate (Glu), gamma-aminobutyric acid (GABA), N-acetyl aspartate (NAA), myo-inositol (mIns), 2HG, and lactate (Lac) were also determined with LC-MS/MS for surgical specimen of the same patients. Spearman correlation coefficients were calculated between the metabolite concentrations measured with 1 H-MRS and LC-MS/MS. Additionally, machine-learning algorithms were used to detect IDH mutation in gliomas based on metabolite concentrations obtained with 1 H-MRS and LC-MS/MS. Consequently, there were statistically significant correlations between 1 H-MRS and LC-MS/MS results for 2HG ( p =0.036), Cr ( p =0.009), mIns ( p <0.001), Lac ( p =0.007) and NAA ( p =0.004). IDH mutation was detected with an accuracy of 92.42% (sensitivity=91.70%, specificity=93.46) and 82.94% (sensitivity=84.04, specificity=81.43) based on LC-MS/MS and 1 H-MRS metabolic intensities, respectively. In conclusion, 1 H-MRS and LC-MS/MS metabolic intensities were highly correlated and these techniques were successful in identifying IDH mutation in gliomas. This study has been supported by TUBITAK 1003 grant 216S432.