WP19 – Physiological MRI for treatment prognosis and response monitoring of GBM
Impact/relevance
This project has the potential to improve current MRI response monitoring of glioblastoma (GBM) brain cancer, and hereby facilitate personalized treatment.
Background
The greatly variable prognosis of glioblastoma patients suggests unknown explanatory factors such as disruption of the Blood Brain Barrier (BBB). The extent of BBB disruption could be important for the efficiency of delivery of chemotherapeutic agents. Further, both tumor tissue and treatment effects can give rise to MRI contrast enhancement, posing a well-known diagnostic challenge.
For brain tumor treatment evaluation, several studies have shown 3D evaluation of tumor size is more accurate, more reliable, and better correlated with prognosis than current standard 2D evaluation.
Aim
The aim of this project is to improve MRI response monitoring of brain tumors through inclusion of quantitative physiological MRI to the follow-up protocol and novel artificial intelligence method for automatic 3D tumor segmentation.
Methods
Dynamic Contrast Enhanced (DCE) MRI with measurement of perfusion, vascular volume and BBB permeability and Diffusion Tensor Imaging to aid delineate white matter tumor cell infiltration, will be obtained in addition to routine MRI in GBM patients prior to and during first-line treatment. BBB defect and biomarkers for BBB disruption will be compared to treatment response and patient outcome. Perfusion and vascular volume will be compared in tumor tissue and regions of treatment induced contrast enhancement. We expect to enroll 75 patients over 1,5 years.
Validation of the method for automatic 3D brain tumor segmentation will be conducted on a retrospective, consecutive dataset of routine brain cancer follow-up MRI scans of 60 intra-axial brain cancer patients by comparison with corresponding manual segmentations.
Expected outcome
The project will result in writing of at least two scientific articles and demonstration of novel imaging strategies for GBM response monitoring.
Peter Jagd Sørensen is a PhD student in the group of Clinical Professor Adam Espe Hansen at the Department of Radiology at Rigshospitalet. His project aims to improve the response monitoring of Glioblastoma (GBM) brain cancer under DCCC work package 19 (WP19). In WP19, Peter Jagd Sørensen is working on improving MRI response monitoring of GBM brain cancer. Specifically, his research projects investigate the potential benefits by inclusion to the follow-up MRI scan protocol of quantitative physiological MRI and novel artificial intelligence methods for automatic 3D tumour segmentation.
The goal of incorporating physiological MRI in the response monitoring is to further the understanding of blood-brain-barrier (BBB) disruption in GBM and to achieve better differentiation between disease progression (active tumour), pseudoprogression (accumulation of inflammatory response cells) and treatment-induced necrosis on MRI. If novel artificial intelligence methods for automatic 3D tumour segmentation are validated, they may allow more accurate and more reliable estimation of tumour burden. Achieving these goals will improve treatment response monitoring and thus provide a more solid foundation for clinical decision-making and personalised treatment of GBM cancer patients.