Methodology matters: characterization of glioma through advanced MR imaging

Abstract

This thesis explored advanced MRI modalities to improve glioma characterization, focusing particularly on Chemical Exchange Saturation Transfer (CEST) imaging. Initial optimization was performed in healthy subjects using VDMP-CEST, successfully distinguishing between slow and fast-exchanging molecular pools and identifying optimal imaging parameters. Further CEST work on fast-exchanging amines revealed an unexpected glutamate contribution to the 2 ppm pool.

CEST imaging, particularly the pool at 2 ppm, was applied in a pilot cohort with glioma patients. Here the CEST was compared to standard clinical imaging to identify tumor lesions. The 2 ppm CEST pool could differentiate enhancing from non-enhancing tumor regions, suggesting its potential as a non-invasive diagnostic marker.

This thesis further examined the benefits of ultra-high field (7T) MRI. Compared to standard clinical scans, 7T imaging revealed more extensive and complex T2 hyperintense tumor areas, indicating improved tumor delineation. Finally, the thesis investigated post-treatment MRI phenotyping. WIth structural and perfusion images, four patient subgroups with distinct perfusion patterns and survival outcomes were identified. While these phenotypes showed potential for predicting overall survival, they were not effective in distinguishing true from pseudo-progression. Overall, this thesis highlights the potential of advanced MRI, including CEST and 7T MRI, to support glioma clinical assessment.