The imaging of cerebral oxygenation is an extremely important tool in understanding the pathophysiology of the tumor and for adaptation of therapies according to hypoxia. Currently, imaging of cerebral oxygenation is mainly performed by the use of Positron Emission Tomography (PET). Thus, the investigators have been able to show that the FMISO radiotracer can reveal tumor hypoxia (HypOnco study, promotor: Caen University Hospital, main investigator: J.S. Guillamo). After injection of the radiotracer, increased uptake is observed in the regions for which the tissue oxygen pressure is less than 10 mmHg (the healthy brain with a tissue oxygen pressure (ptO2) ≈ 40mmHg).

Although PET is a reference methodology, it is not widely practiced mainly because of radioactive sources. Magnetic Resonance Imaging (MRI) would bypass the previously mentioned PET limitations. The investigators have recently shown that a measure of local oxygen saturation could be obtained by MRI.

This methodology has also been implemented at a clinical scale on lower field MRI magnets, but its formal validation in a clinical situation remains to be demonstrated with respect to FMISO.

The major advantage of this methodology is that MRI is already performed in routine practice for patients. Measuring tissue oxygenation with MRI (SatO2-MRI) would not add additional examination for the patient. In addition, MRI is a non-ionizing methodology with a very good spatial resolution compared to PET, this should help to better understand intratumoral heterogeneity. Similarly, in preclinical studies, the investigators have shown in a context of mild hypoxia that SatO2-MRI may be more sensitive than PET.

The investigators propose a study to compare in patients with glial tumors, images obtained by 3 Tesla MRI of SatO2-MRI to the hypoxia maps obtained by FMISO PET. These imaging studies will be confronted with studies carried out in immunohistochemistry on biopsies / resection allowing to reveal and to quantify by image analysis the expression of the factors induced by hypoxia (HIF1, HIF2).

This study should include 20 patients with glioma (15 high-grade patients and 5 low-grade patients) in pre-surgery. The aim is to show that SatO2-MRI is a relevant methodology (in terms of sensitivity, specificity) for assessing intratumoral oxygenation in a context of brain tumors. This fits perfectly into an era of personalized medicine where functional imaging finds its meaning.

Inclusion Criteria:

• Adult patient (> 18 years-old)
• Patient with a diffuse glioma (astrocytoma or oligodendroglioma or glioblastoma)
• Karnofsky index ≥ 70%
• Patient able to understand and sign the consent or accompanied by a trusted person who can do so
• Normal blood test with results: neutrophil ≥ 1500 cells / μl, platelets ≥ 100,000 μl, serum glutamic-oxaloacetic transaminase (SGOT) ≤ 2.5 x upper normal limit (UNL), creatine ≤ 1.5 x UNL
• Patient affiliated to a social security

Exclusion Criteria:

- Patient with contraindications to surgery