Abstract Background Gliomas are common primary brain tumours and frequently harbour IDH1/2 mutations, which result in accumulation of 2-hydroxyglutarate (2-HG) and altered cellular metabolism. Proton magnetic resonance spectroscopy (1H-MRS) enables non-invasive metabolite assessment and may aid in predicting molecular characteristics of gliomas. This study aimed to perform comprehensive quantitative metabolite analysis using 1H-MRS and to evaluate its ability to predict genetic alterations in gliomas. Methods Patients with suspected brain tumours underwent 1H-MRS over a 22-month period. Metabolite concentrations were quantified using LCModel and compared among IDH-mutant gliomas, IDH-wildtype gliomas, and non-glioma cases. Diagnostic performance was assessed using receiver operating characteristic analysis. Results Among 90 cases, IDH-mutant patients were younger than IDH-wildtype patients (40.7 ± 16.4 vs. 60.5 ± 20.5 years, p = 0.005, area under the curve AUC 0.77). 2-HG concentrations did not differ significantly (0.58 ± 1.50 vs. 0.51 ± 1.70 mM, p = 0.907; AUC 0.51). Cr concentrations were higher in IDH-mutant gliomas (3.12 ± 1.23 mM) than in IDH-wildtype gliomas (2.03 ± 0.91 mM; p = 0.029; cutoff ≥2.4 mM: sensitivity 87.5%, specificity 77.8%, AUC 0.764). PCr concentrations were lower in tumours with 1p/19q codeletion (2.21 ± 0.25 mM) than in non-codeleted tumours (3.25 ± 1.40 mM; p = 0.033; cutoff 2.4 mM: sensitivity 40.0%, specificity 100%, AUC 0.79). Conclusion 1H-MRS analysis suggests that Cr and PCr, together with clinical factors such as age, may provide complementary information for predicting molecular features of gliomas beyond 2-HG alone.
Kume et al. (Thu,) studied this question.