Using multi-modal PET and MRI to predict the site of tumour recurrence in high-grade glioma

No Thumbnail Available

Date

2024

Journal Title

Journal ISSN

Volume Title

Publisher

The University of Manchester

Abstract

Background: High-grade gliomas (HGG) are highly aggressive and incurable brain tumours that often recur within 2 cm of the original site, even after complete oncological treatments. Advanced MRI and PET techniques hold promise for better tumour delineation and characterisation. This thesis investigates the spatial characterisation of translocator protein (TSPO) and amino acid PET in initially diagnosed HGG and at the point of post-treatment progression, with a view to identifying future sites of disease progression. Methods: HGG patients underwent prospective imaging with [11C](R)PK11195 and [11C]methionine PET alongside MRI including diffusion tensor imaging (DTI). [11C](R)PK11195 binding potential (BPND) and [11C]methionine tumour-to-background ratio were generated. The PET biomarkers were first used to characterise tumour regions defined on MRI as contrast-enhancing (CE) and peritumoral regions and diffusion connectivity map (N=12 initially diagnosed). Secondly, tumour biological volumes were delineated and compared among [11C](R)PK11195 and [11C]methionine PET and CE-MRI using overlap, Dice and Jaccard similarity coefficients (DSC and JSC), and discrepancy measures at baseline (12- initially diagnosed; 8 post-treatment). Disease progression was assessed using follow-up MRI (N=16) and registered with baseline PET biomarkers to explore overlap and similarity. Results: In newly diagnosed HGG, 67±22% of CE-MRI regions showed positive TSPO/methionine, while 36±15% of peritumoral regions showed positive TSPO but negative methionine. TSPO binding in CE-MRI and peritumoral regions was significantly higher than in the contra-lesional reference. [11C](R)PK11195 PET demonstrated a gradual decrease in TSPO binding along the DTI connectivity map compared to [11C]methionine. [11C](R)PK11195 and [11C]methionine biological volumes showed moderate spatial similarity (DSC=0.65±0.15, JSC=0.50±0.16) at initial diagnosis, which decreased for post-treatment HGG (DSC=0.35±0.15, JSC=0.22±0.11). Both PET volumes exhibited moderate overlap with CE-MRI at initial diagnosis, but post-treatment spatial similarity decreased, with substantial discrepancies. TSPO binding and methionine uptake show signals in areas where future disease progression occurred, with average overlaps of 0.51±0.25 and 0.43±0.26, respectively. Conclusions: This work provides the first detailed spatial characterisation of two PET biomarkers in newly diagnosed HGG and at the point of post-treatment progression. Elevated TSPO binding without increased methionine uptake may indicate inflammatory or discrete neoplastic populations not captured by standard imaging techniques. Both PET radiotracers demonstrated increased uptake beyond initial contrast enhancement, and although the exact site lacked specificity, disease progression almost universally occurred within this area. Notably, the discrepancy between TSPO and methionine binding increases post-treatment, with elevated TSPO expression likely reflecting late-stage inflammation contributing to symptomatic worsening in such patients.

Description

Keywords

Neuro-oncology, glioma, glioblastoma, high-grade glioma, imaging, PET, TSPO, amino acid, [11C](R)PK11195, [11C]methionine, MRI, diffusion tensor imaging (DTI), tractography, inflammation, brain

Citation

Endorsement

Review

Supplemented By

Referenced By

Copyright owned by the Saudi Digital Library (SDL) © 2025