Effect of ivosidenib and vorasidenib on 2-hydroxyglutarate levels in low grade glioma: An in vivo MR spectroscopy study.

2081 Background: IDH-mutant gliomas are slow-growing infiltrating tumors of astrocytic (AS) or oligodendroglial (OG) origin (WHO grade 2/3). The mutations occur in genes that encode the metabolic enzyme IDH1 or, more rarely, IDH2 and lead to production of 2-hydroxyglutarate (2-HG) that can be measured with optimized in-vivo MRS. Small-molecule IDH inhibitors (IDHi) ivosidenib (inhibits mIDH1 enzyme) and vorasidenib (inhibits mIDH1 and mIDH2 enzymes) showed good tumor penetrance and ~95% 2-HG reduction measured in tumor biopsies. Both ivosidenib and vorasidenib have evidence of responses in tumor growth rate. As volume reductions are often observed, but after a delay of several months, thus an early response biomarker highly desired. The aim of this study was to explore whether MRS measurements of 2-HG can be used to non-invasively monitor response to treatment with IDHmut-inhibiting drugs and compare this response to changes in tumor volume. Methods: 14 patients (Age ≥18y) with a histomolecularly confirmed IDH1 mutated diffuse glioma (AS or OG) received ivosidenib (n = 12) or vorasidenib (n = 2) therapy as part of their routine clinical care. MRI was performed before treatment ( baseline ) and repeated ( follow-up ) with a median on-drug follow-up of 6 months [4, 12 IQR]. Tumors were segmented from FLAIR images in 3D Slicer and volumes were calculated in cm 3 . All MRS data were processed in Osprey with the built-in LCModel fitting module. Comparisons between baseline and follow-up measured metabolite levels were conducted using paired t-tests or (in cases where the normality assumption was not met) non-parametric Wilcoxon tests. Results: We analyzed spectra from 11 patients with both baseline and follow-up sessions. All spectra exhibited characteristic tumor features: reduced total N-acetylaspartate (tNAA) and elevated levels of total choline (tCho), lactate (Lac), and myo-inositol (mI), along with a 2-HG peak at 2.25 ppm that is visibly smaller after treatment. The decrease in 2-HG levels was highly significant (p < 0.001) across all included patients undergoing ivosidenib/vorasidenib therapy, regardless of reference standard. Volumetric assessment revealed tumor growth arrest and a subtle reduction in tumor growth in some individuals. Conclusions: This is the first in-vivo evidence using MRS that ivosidenib/vorasidenib reduces 2-HG. We found that 2-HG levels respond specifically and rapidly to treatment, while volumetric changes manifest slowly and more gradually, consistent with previous studies. This preliminary study suggests that in vivo MRS-derived 2-HG estimates could serve as sensitive and specific biomarkers for monitoring low grade gliomas in vivo in response to small-molecule IDH inhibitor therapy after initiation of treatment. Longitudinal volumetric and radiomic analyses are underway.