Background: Diffuse gliomas, the most common adult primary brain tumour, are classified by their histological, genetic, and molecular characteristics. Isocitrate dehydrogenase (IDH) mutant gliomas correlate with longer mean survival and progression-free survival time. In tumour cells, IDH mutations lead to metabolic changes and the production of oncometabolite 2-hydroxyglutarate (2HG). 2HG can be measured with MR spectroscopy (MRS) in vivo and non-invasively, and is one of the first binary and specific imaging biomarkers. In this study, we validate advanced MRS methods as IDH-mutation imaging biomarkers in the clinic and address the necessary steps to translate them into a clinical standard of care.
Methods: MRS techniques were evaluated and validated, according to the Imaging Biomarker Roadmap for Cancer Studies (O’Connor et al. 2016), using simulations, novel and specific glioma phantoms, and healthy volunteers. PRESS, sLASER, and MEGA-PRESS sequences were tested after being optimised for 2HG measurement (2.25 ppm peaks for PRESS and sLASER and a 4.02 ppm peak for MEGA-PRESS). A clinical study is currently being carried out to validate the optimised sequences; 4 patients with suspected gliomas have participated thus far.
Results: Simulations of MRS sequences allowed for optimised acquisition of parameters for the measurement of 2HG, while purpose-built tumour-mimicking phantom experiments allowed for the sensitivity, specificity, and reproducibility of the different methods. Further synthetic MRS measurements using combinations of tumour simulations/phantoms determined the optimal noise and SNR ranges. In the early stages of the clinical study, MEGA-PRESS successfully predicted IDH mutation status in all 4 patients. Other important genetic markers were also successfully predicted in all four patients.
Conclusions: MEGA-PRESS proved to be the most robust MRS method and demonstrated accurate and precise measurement of phantom and synthetic 2HG with acceptable SNR. MEGA-PRESS has proven thus far in a small subset of glioma patients to accurately predict their important genetic markers for tumour classification
