Hypoxia-targeted BOLD functional MRI for differentiating glioblastoma recurrence from radiation necrosis; A comparison with [18F]FET-PET

• Applicant: Dr. Leonie Zerweck
• Subject Area: Clinical Neurology; Neurosurgery and Neuroradiology
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 572737263

A significant challenge in the diagnosis of glioblastomas is the differentiation between tumor recurrence or progression and radiation necrosis, as this significantly guides further treatment. Conventional magnetic resonance imaging (MRI) cannot differentiate between tumor recurrence and radiation necrosis because both share similar radiological features with contrast-enhancing lesions. Recently, blood oxygenation level-dependent functional MRI (BOLD fMRI) with hypoxic stimulation was described as a promising method for characterizing brain tumors. Hypoxia increases the concentration of deoxyhemoglobin in the blood, leading to an increase in the BOLD signal in brain parenchymal tissue. This research project aims to investigate whether hypoxia-targeted BOLD fMRI leads to specific, hypoxia-targeted BOLD signal responses in tissue with glioblastoma recurrence and whether these responses are measurable in radiation necrosis. As radiation necrosis is primarily an inflammatory reaction, the expected BOLD signal response in true tumor tissue is assumed to be absent in radiation necrosis. In a prospective study, hypoxia-targeted BOLD fMRI examinations will be performed in 10 patients with glioblastoma recurrence and 10 patients with radiation necrosis. To differentiate between true tumor recurrence and radiation necrosis, [18F]fluorethyltyrosine positron emission tomography ([18F]FET-PET) examinations will be used. During the BOLD fMRI examination, two 60-second periods of hypoxia will be interleaved with periods of normoxia. Patients will inhale gases with different partial pressures of oxygen via a face mask. An automated computerized gas mixer (RespirAct; Thornhill Medical) will be used to precisely modulate the gas mixtures inhaled by the patients, enabling the control of the end-tidal partial pressures of oxygen and carbon dioxide independently. The BOLD fMRI datasets will be co-registered with anatomical datasets. The anatomical datasets will be used to define volumes of interest (VOIs) by manually drawing contrast-enhancing lesions (VOI1) and defining further concentric VOIs (VOIs2–10) in the perilesional tissue. Hypoxia-targeted BOLD signal changes will be evaluated in all VOIs. The aim is to assess whether there are significant differences in BOLD signal responses between tumor recurrence and radiation necrosis in the contrast-enhancing lesions and in the perilesional tissue.

• DFG Programme: WBP Fellowship
• International Connection: Switzerland
• Host: Privatdozent Dr. Jorn Fierstra