The overproduction of the oncometabolite 2-hydroxyglutarate (2HG) in IDH mutated gliomas [1, 2] can be detected non-invasively by magnetic resonance spectroscopy (MRS) [3-5]. Reliable quantification of 2HG has important implications in diagnosis of IDH mutation, prognosis, as well as assessment of the efficacy of anti-IDH targeted therapies. 2HG detection is very challenging and advanced methods for robust detection of this metabolite are not available in clinical settings. In this study, we employed two approaches for 2HG detection previously described, e.g., difference spectroscopy and optimized for 2HG detection conventional spectroscopy [3, 4], on a pre-operative cohort of patients, and on a post-operative cohort of patients previously treated with radio-and/or chemotherapy. The aim was to assess for the first time the specificity and sensitivity of the two methods, and to relate the results to the natural history and neuroradiological status of patients with glioma.

MRI/MRS protocol: Acquisitions were performed using a 3-T whole-body system (MAGNETOM Verio, Siemens, Erlangen, Germany) equipped with a 32-channel receive-only head coil. The protocol included T₂-w FLAIR and T₁-w sequences for voxel placement. MRS data were acquired using two sequences optimized for 2HG detection: 1) a single-voxel MEGA-PRESS [6] sequence (TR=2s, TE=68ms, 128 averages, scan time=9 min) with editing pulses applied at 1.9 and 7.5 ppm, for the edited and non-edited condition respectively, which allows to measure the 2HG signal at 4.05 ppm; 2) a single-voxel PRESS sequence (TR=2.5s, TE=97ms, TE1=32ms, TE2=65ms, 128 averages, scan time = 6 min), optimized to detect the 2HG signal at 2.25 ppm [3]. Water suppression was performed using VAPOR and outer volume suppression techniques [7]. Frequency and phase corrections were performed on single scans using the total choline signal at 3.2 ppm. Typical VOI size was 2x2x2 cm³ (Fig. 1c, f). VOI size was adapted to tumor size in order to minimize partial volume effects, keeping a minimum size of 6 cm³. Spectral quantification was performed using LCModel. The Cramer Rao lower bounds (CRLB) threshold for reliable 2HG detection was set to 20%.

Patient population: Thirty patients were included in the study: six subjects before surgery with suspected brain glioma (pre-operative cohort), and twenty-four patients who had surgery and radiotherapy or chemotherapy, and were affected by an IDH1 mutated glioma (post-operative cohort). In the post-operative cohort, five patients were affected by grade II, fourteen by grade III and five by grade IV IDH mutated glioma. In the same cohort, eleven patients were under radiotherapy or chemotherapy. IDH status was assessed for all patients combining detection of expression of IDH1-R132H mutant by immunohistochemistry (IHC) and Sanger sequencing for IDH1 and IDH2 gene mutations.

In the pre-operative cohort, 2HG was detected in five out of six patients using MEGA-PRESS, while CRLB was < 20% only for two patients from PRESS data (examples of spectra in Fig. 1). Histopathological diagnosis after surgery confirmed the presence of IDH mutated gliomas in the five patients with detectable 2HG levels (two grade II and three grade III gliomas), while it revealed an IDH wild-type ganglioglioma in the patient with no 2HG signature [sensitivity 100% (0.4 to 1.0, 95% CI); specificity 100% (0.02 to 1.0, 95% CI)] (Table 1). In the post-operative cohort, 2HG was detected in seven out of twenty-four patients using MEGA-PRESS (sensitivity 29%), and only in three patients using PRESS (sensitivity 12%, data not shown). Detection of 2HG was not associated with grade (p=0.4), histological diagnosis (p=0.5), or the fact that patients were out of treatment at the moment of MRS examination (p=0.2).2HG spectroscopy is a promising technique for the diagnostic set-up and follow-up of glioma patients, which remains underused because of technical challenges. Preliminary results strongly suggest that difference spectroscopy (MEGA-PRESS) can provide high sensitivity/specificity for prediction of IDH mutation status before surgery, in contrast to optimized for 2HG detection conventional spectroscopy (PRESS). Fitting of conventional spectra can provide false-negative readouts due to spectral overlap of 2HG with chemically similar metabolites, such as glutamate, glutamine and GABA. The detection of 2HG is crucial for monitoring the response to treatments targeting IDH mutation. Although difference spectroscopy provides much higher sensitivity with respect to conventional MRS, in the post-operative cohort it was not possible to measure reliably 2HG concentration in 70% of patients. Some of the factors which can impact the sensitivity of the measurement are the residual tumor volume, 2HG concentration, cellular density, treatment, evolution of the disease, MRI phenotype, tumor localization, type of mutation and molecular pattern. All these elements together will be evaluated to provide a general guideline for 2HG detection after surgery.