Despite maximal safe surgery followed by combined chemo-radiation therapy, the outcome of patients suffering from glioblastoma (GBM) remains extremely poor with a median survival of 15 months. Hence, new avenues have to be taken to improve outcome in this devastating disease. Given their intracerebral localization and their highly invasive features, GBM pose some specific challenges for the development of adequate tumor models. Orthotopic xenograft models directly derived from the tumor of a patient might represent an attractive perspective to develop patient-specific targeted therapies. This approach remains however to be validated for GBM as it offers specific challenges, including the demonstration that the properties of xenograft models validly represent treatment relevant features of the respective human tumors.

In this innovative project the investigators aim to compare and validate an approach of paired human GBM and respective derived orthotopic xenografts in the mouse brain on the levels of radiological behavior and metabolism of the tumors, as determined by high resolution MRI of the patients (7T MRI) and the respective orthotopic mouse xenografts (14.1T MRI), as well as on the level of the transcriptome, genome, and methylome of the original GBM tissue and respective derived xenografts/glioma sphere lines. The data will be integrated in multidimensional analyses and interrogated for similarities and associations with molecular GBM subtype.

This pilot project will provide the basis for the crucial next steps, which will include drug intervention studies. New promising drugs, tested pre-clinically in the mouse orthotopic xenograft models established here using the radiologic/metabolic/molecular procedures described for this project, will be taken into patients in phase 0 studies. GBM patients will receive radiologic/metabolic follow-up using high resolution MRI under drug treatment, followed by resection of the tumor and subsequent acquisition of molecular data.

The presented project will focus on the evaluation of a multimodal approach comparing human GBM to paired samples of orthotopic xenografts using high resolution MRI and MRS and multidimensional molecular profiling.

20 patients with a high probability for newly diagnosed GBM based on MRI-scan ( 3 Tesla (3T) MRI, T1, T2, T1 gadolinium, DWI & MRI Spectroscopy) will be identified in the CHUV prior to undergoing neurosurgical resection. Patients will undergo extensive experimental radiological examination using specific MRI sequences on the 7 Tesla (7T) MRI to identify specific metabolic pathways (see below, section on imaging). Thereafter patients will undergo maximal safe neurosurgical resection of their tumors. The portion of the tumor that is not used for diagnostic purposes will be collected immediately for further use (see below, section on molecular evaluations). Following resection, patients will undergo standard of care treatment [usually combined radio-chemotherapy, or will be offered participation in a clinical trial. The clinical parameters will be collected, including histopathological features and the evolution and growth pattern of the residual tumor (if present), or the development of recurrences will thereafter be compared to the parameters and evolution of the xenograft models.

At high magnet field strength (7T), high signal-to-noise ratio and increased spectral dispersion allow more reliable measurement of a large number of metabolites using Magnetic Resonance Spectroscopy in comparison to clinical available field strengths (3T and below). In addition, the authors have developed a full sensitivity short-echo-time single voxel spectroscopy (SVS) sequence "semi-adiabatic SPECIAL"(2) which was implemented, validated at 7T and allows the quantification of 15 metabolites with high precision including N-acetylaspartate(NAA), glutamine(Gln), glutamate(Glu), myo-inositol(Ins), phosphorylethanolamine(PE), total choline(tCho), creatine, phosphocreatine, N-acetylaspartylglutamate(NAAG), lactate(Lac), glutathione(GSH), aspartate (Asp), taurine(Tau), scyllo-inositol and γ-aminobutyric acid(GABA). This localization technique was further extended to a MR Spectroscopic Imaging (MRSI) technique at 7T, which allows mapping of the spatial distributions of cerebral metabolites. Furthermore, glycine is a possible marker for tumor malignancy and its detection in vivo has been established in our previous study using TE=30ms with SPECIAL sequence at 7T. Therefore, in this study the aforementioned techniques will be used to obtain the neurochemical information and its spatial distribution in the glioblastoma of the patients. These data will be further compared with the neurochemical information obtained in the orthotopic xenografts in the mouse brain derived from the respective glioblastoma patient.

All MRS measurements of glioblastoma patients will be performed on a 7T MR scanner with a CP Transmit / 32 channel receive array head coil. Based on the high resolution T1-weighted images obtained using the MP2RAGE sequence, Volume of Interest (VOI) for spectroscopy will be placed according to the location of the glioblastoma. Total acquisition time of MRS will be within 30 min. In vivo MRS spectra will be post-processed and metabolite concentrations will be quantified to create metabolite maps.

Molecular and functional investigations of paired samples of primary glioblastoma and respective orthotopic xenografts in the mouse

The aim of the present study is to determine the molecular, histopathological, and functional properties, including growth patterns such as invasiveness, of the original GBM and the respective derived orthotopic xenografts in the mouse, and link them to imaging/ metabolism parameters obtained by high resolution MRI.

GBM samples from patients collected at surgery will be divided into 2 parts, (i) snap frozen for molecular analyses, and (ii) cultivated under stem cell conditions for subsequent stereotactic transplantation into male immune-compromised mice and establishment of sphere lines.](streamdown:incomplete-link)

Inclusion Criteria:

• High level of suspicion of glioblastoma
• Planned neurosurgical resection
• Adequate bone marrow function
• Adequate liver and kidney function

Exclusion Criteria:

• inability to undergo MRI
• inability to undergo neurosurgical resection