PREOPERATIVE DETERMINATION OF IDH STATUS AND GRADE IN GLIOMAS USING MRS

PURPOSE: To evaluate the diagnostic accuracy of preoperative MRS in the determination of the IDH status and grade in patients with newly diagnosed gliomas. METHODS: MRS was performed using a regular PRESS sequence and a spectral editing sequence (MEGA-PRESS). Concentration of 2-HG was estimated from the subtracted spectra from the edited MRS sequence while NAA, choline and creatinine concentrations were obtained from the regular PRESS sequence. IDH mutation status was assessed by immunohistochemistry for all patients and additional next generation sequencing for all grade 2 and 3 gliomas. Differences in metabolite concentrations between IHD-mutant and wild-type gliomas and between gliomas of various grades were assessed using non parametric tests. TAreas under-the ROC curve (AUC) for different metabolites were calculated with IDH mutation status or glioma grade as the outcome. RESULTS: There were 29 IDH-mutant gliomas and 52 wild-type gliomas. There was a significant difference in the NAA/Choline ratio among various glioma grades (P<0.05). The AUC for 2HG was 0.74 in the differentiation between IDH-mutant vs wild type gliomas. Using a 2-HG cut-off of >0.96 I.U., sensitivity was 59% and specificity was 90% for the identification of IDH-mutant gliomas. The AUC for the NAA/Choline ratio was 0.74 in the differentiation of high vs low grade gliomas. Using a NAA/Choline cut-off ≤0.53, sensitivity was 45% and specificity was 100% for identification of high grade gliomas. CONCLUSION: Preoperative MRS can identify IDH-mutant gliomas and high grade gliomas with high specificity.