Project title: Molecular imaging of high grade glioma for improving response to treatment (MOLIMAGLIO)
Project code: SAF2014-52332-R
Summary: High grade gliomas (HGG), including glioblastoma (GBM), anaplastic astrocytomas, oligoastrocytomas or oligodendrogliomas, are the most common malignant primary brain tumour types. The standard initial treatment for GBM is surgical, with the aim of achieving a maximum resection that still maintains the neurological functions of the patient. The extent of this resection is also important to increase patient survival. Advances in treating HGG are hampered by the lack of good non-invasive biomarkers for early response or relapse.
Magnetic resonance (MR) is widely used for non-invasive investigations of brain tumours, like diagnosis and surgery planning. Additionally, MR spectroscopic imaging (MRSI) provides “molecular images” derived from biochemical information and is ideally suited to the application of pattern recognition (PR) techniques, such as non-negative Matrix Factorisation (NMF). With NMF, the main constituent metabolomic signatures (or sources) present in heterogeneous areas of a brain tumour can be recovered, for instance those that respond to therapy and those that do not. MOLIMAGLIO will use preclinical models of GBM to produce an evidence-based real time strategy allowing non-invasive tumour segmentation into molecular (also named nosological) images of tumour infiltration, progression, therapy response and tumour regrowth. We will use for this several orthotopic mice GBM models induced by stereotactic injection of cultured cell lines (e.g. GL261) or cells derived from gliospheres from mouse, dog, or human brain tumour biopsies. Moreover, the major clinical goal of MOLIMAGLIO will be to produce Proof-of Concept (PoC) that preclinical molecular imaging protocols can be adapted to improve detection of infiltration, response and relapse in patients. This should also allow to better indicate neurosurgeons where to remove tissue beyond current image techniques applied for HGG, without adding morbidity, to achieve better personalised therapy and prognosis. For all this a two pronged approach with preclinical and clinical discrete objectives is proposed.
O.1 Produce robust MRSI source – derived molecular images of preclinical GBM segmented for proliferating tumour and response to therapy (hyperthermia, CK2 inhibition and temozolomide).
O.2 Produce perturbation enhanced molecular images (PE-MRSI) of preclinical GBM allowing in vivo segmentation with potential for intra-surgical brain tumor mass visualization and “brain-shift” correction.
O.3. Validate segmented tumour zones of preclinical GBM using ex vivo data (correlation with immunohistochemistry and exome sequencing).
O.4 Proof-of concept (PoC) protocol targeted to improving animal survival by tailoring therapy switch/combination to the non-invasive individual detection of early signs of response or relapse.
O.5 Phase I-II study on the safety of Dimethyl sulfoxide (DMSO) as a contrast agent in the evaluation of response and as a prognostic factor of human HGG.
O.6 Develop a method for co-registration of the source-derived molecular imaging maps of human HGG onto their MRI images DICOM series.