Combined diffusion-weighted MRI and diffusion-weighted MR spectroscopy for interrogating the glioma tumour microenvironment: a preliminary study

medRxiv

Abstract

Abstract Background and purpose Gliomas are characterised by a complex tumour microenvironment (TME) that contributes to treatment resistance and tumour heterogeneity. Therefore, the non-invasive interrogation of both the intracellular and extracellular compartments of gliomas remains a key unmet need. We investigated the feasibility and complementarity of combining diffusion-weighted MRI (DW-MRI) with biophysical modelling Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumours (VERDICT) and diffusion-weighted MR spectroscopy (DW-MRS) for simultaneous characterisation of glioma tumour cells and the glioma TME. Methods 14 patients with newly diagnosed glioma (WHO grades 2–4: 4 IDH-wildtype and 10 IDH-mutant) underwent DW-MRI at 3 T; DW-MRS was additionally acquired in 10 patients. Tumours were automatically segmented into enhancing, non-enhancing, and oedema regions using a validated pipeline. VERDICT models were fitted to multi-shell DW-MRI data to estimate intracellular volume fraction (f IC ), cell radius, extracellular diffusivities (D and D), and free-water fraction (f FW ). Single voxel DW-MRS (20×20×20 mm³) provided metabolite-specific apparent diffusion coefficients (ADCs) for total N-acetylaspartate (tNAA), Creatine (tCr), and choline (tCho). T-tests assessed DW-MRI and descriptive statistics assessed DW-MRS parameters in tumour regions compared to normal appearing white matter (NAWM) and IDH mutation status. Pearson’s correlations assessed associations between DW-MRS metabolite ADCs and DW-MRI parameters. Results VERDICT-MRI distinguished high grade IDH-wildtype from lower grade IDH-mutant gliomas, with significantly higher f IC and lower D and D in enhancing and non-enhancing regions of IDH-wildtype lesions. DW-MRS demonstrated a trend towards reduction in tNAA ADC in tumour versus contralateral NAWM, consistent with neuronal loss, and a trend towards increased tCho ADC, suggesting glial activation. A descriptive trend towards decreased tNAA ADC in IDH-wildtype tumours was observed. Significant positive correlations were identified between tumoral tCho ADC and VERDICT parameters f EES , D and negative correlations for ADC and f FW , in non-enhancing tumour regions. Conclusion This proof-of-concept study demonstrates the feasibility of combining multi b-value DW-MRI and DW-MRS within a clinically feasible protocol to simultaneously probe the extracellular and intracellular compartments of the glioma TME. VERDICT captured cell-level and extracellular matrix differences in IDH mutation status, while DW-MRS provided metabolite-specific indices of neuronal and glial compartment integrity. The correlation between tCho ADC and VERDICT metrics in infiltrative tumour regions supports the complementarity of these modalities. With this combined approach, it is possible to simultaneously characterise the tumour compartment and the tumour microenvironment in gliomas.

Authors 13

1. Marco Palombo

   Cardiff University

   Affiliation as printed

   Cardiff University Brain Research Imaging Centre (CUBRIC), School of Psychology, Cardiff University, Cardiff, UK;

2. Matteo Figini

   University College London

   Affiliation as printed

   UCL Hawkes Institute, University College London, London, UK;

3. Samuel Rot

   National Hospital for Neurology and Neurosurgery · UCL Queen Square Institute of Neurology · University College London

   Affiliation as printed

   NMR Research Unit, Queen Square MS Centre, Department of Neuroinflammation, UCL Queen Square Institute of Neurology, Faculty of Brain Sciences, University College London, London, UK;

4. Elizabeth Powell

   University College London

   Affiliation as printed

   UCL Hawkes Institute, University College London, London, UK;

5. Bhavana Solanky

   National Hospital for Neurology and Neurosurgery · UCL Queen Square Institute of Neurology · University College London

   Affiliation as printed

   NMR Research Unit, Queen Square MS Centre, Department of Neuroinflammation, UCL Queen Square Institute of Neurology, Faculty of Brain Sciences, University College London, London, UK;

6. Chloé Najac Aachen Department of Radiology, C.J. Gorter MRI Centre

   Leiden University Medical Center

   Affiliation as printed

   Department of Radiology, C.J. Gorter MRI Center, Leiden University Medical Center, Leiden, Netherlands;

7. Bernard M. Siow

   The Francis Crick Institute

   Affiliation as printed

   In Vivo Imaging, The Francis Crick Institute, London, UK;

8. Jeremy Rees

   UCL Queen Square Institute of Neurology · University College London

   Affiliation as printed

   Department of Translational Neuroscience and Stroke, UCL Queen Square Institute of Neurology, Faculty of Brain Sciences, University College London, London, UK;

9. Eleftheria Panagiotaki

   University College London

   Affiliation as printed

   UCL Hawkes Institute, University College London, London, UK;

10. Anestis Passalis

    National Hospital for Neurology and Neurosurgery · UCL Queen Square Institute of Neurology · University College London

    Affiliation as printed

    NMR Research Unit, Queen Square MS Centre, Department of Neuroinflammation, UCL Queen Square Institute of Neurology, Faculty of Brain Sciences, University College London, London, UK;

11. Claudia A. M. Gandini Wheeler‐Kingshott

    National Hospital for Neurology and Neurosurgery · UCL Queen Square Institute of Neurology · University College London

    Affiliation as printed

    NMR Research Unit, Queen Square MS Centre, Department of Neuroinflammation, UCL Queen Square Institute of Neurology, Faculty of Brain Sciences, University College London, London, UK;

12. Itamar Ronen

    Brighton and Sussex Medical School · University of Sussex

    Affiliation as printed

    Clinical Imaging Sciences Centre, Brighton and Sussex Medical School, University of Sussex, Brighton, UK

13. Harpreet Hyare corresponding

    UCL Queen Square Institute of Neurology · University College London

    Affiliation as printed

    Department of Translational Neuroscience and Stroke, UCL Queen Square Institute of Neurology, Faculty of Brain Sciences, University College London, London, UK;