MR Perfusion Methods in Patients With Suspected Recurrent High Grade Gliomas

Overview

Radiation therapy is an important adjunct in the treatment of patients with glioma, although a common side effect is radiation-induced injury of brain parenchyma. Unfortunately, conventional MRI is not accurate in differentiating radiation-induced brain injury from recurrent tumour, both of which may demonstrate progressive contrast enhancement. Recent studies have suggested that perfusion MRI could improve this differentiation. Perfusion MRI can be performed with an injection of exogenous contrast using dynamic contrast enhancement (DCE) or dynamic susceptibility contrast enhancement (DSC). Perfusion MRI can also be performed without contrast injection using arterial spin labeling (ASL) or intravoxel incoherent motion (IVIM). DCE-MRI relies on accurate measurement of T1 values in order to convert the MRI signal intensity to contrast concentration. Dynamic susceptibility-weighted contrast enhancement (DSC) perfusion is the most common technique used in clinical practice but measurement of tumor relative cerebral blood volume (rCBV) can be biased by extravascular contrast leakage and susceptibility-weighted artifacts. The purpose of this study is to evaluate the accuracy of perfusion MR imaging using non-contrast and contrast-based techniques in differentiating recurrent tumour from radiation-induced brain injury in patients with known high grade glioma. The investigators will compare the accuracy of IVIM, ASL, DCE and DSC techniques. A secondary goal of the study is to compare two new different T1 mapping methods used for DCE-MRI.

The aim of this study is to evaluate the diagnostic accuracy of different MR perfusion methods in differentiating recurrent tumour from radiation-induced brain injury in patients with high grade gliomas. The investigators will evaluate perfusion methods which do not use an exogenous agent (ASL, IVIM) and methods which use gadolinium as a contrast agent (DSC-MRI, DCE-MRI). A secondary goal of the study is to compare two new different T1 mapping methods used for DCE-MRI. The first T1 mapping method uses a Lock-Locker inversion-recovery acquisition (MOLLI) that measures apparent T1 (T1\*). The second T1 mapping method uses a single-point, saturation-recovery acquisition that measures true T1 SMART1Map. This study is a prospective study which will involve 50 patients. Patients who have received chemoradiation for high grade gliomas and who subsequently developed progressive enhancing lesions on follow-up MR will be asked to participate in this study. Informed consent by the patient or a legal guardian or authorized representative of the patient will be obtained before enrollment. Following enrollment, MR perfusion imaging (ASL, IVIM, DSC and DCE perfusion sequences) will be performed in addition to the routine neuronavigational sequence obtained fror re-operative/ therapy planning. Following this MR examination, the patient may undergo a surgical biopsy or excision as determined clinically by the neurosurgeon. The location of the biopsy will be saved intraoperatively which will allow colocalization of histopathology with imaging data. Histopathological analysis will be performed by the neuropathologist, yielding a visual quantitative estimate of percent viable tumour and radiation necrosis within the obtained specimen. All patients, including those who do not go to surgery, will undergo clinical followup and imaging follow-up with perfusion imaging. This will allow for assessment of lesion progression over time, yielding valuable diagnostic information in differentiating radiation necrosis from tumour recurrence, particularly in those patients who did not undergo surgery. Following a routine clinical MR examination, patients with a new lesion or an enlarging lesion will be seen by the radiation oncologist or neurosurgeon. They will be recruited by a research nurse who will ask if they are willing to participate in the research study. If the patient agrees to participate, the investigators will add perfusion sequences in addition to the clinical MR exam for treatment planning (stereotactic radiation or reresection). The MR exam performed for clinical and research purposes will last less than 45 minutes. Following surgery, the investigators will correlate pathological findings with perfusion and diffusion parameters. If the patient does not undergo surgery, we will use clinical examination and serial MRIs (including perfusion imaging) to determine if the index lesion is more likely to represent radiation necrosis or tumor progression. The investigators will follow the patient clinically for 2 years after the time of progression to determine overall survival. Patients with high grade gliomas are followed routinely with serial MRIs. The patients with progressive enlarging lesions will be seen by the treating oncologist or neurosurgeon. They will electronically put in a request for another MR exam which is needed clinically for surgical planning and/or follow-up with advanced MR imaging as per standard of care. When the patient comes to radiology for their exam, they will be asked to participate in the study by the radiologist and/or the research nurse (if this breaches patient confidentiality, we can put the radiologist as the person getting consent since he is involved in patient care). If the patient agrees to participate, the study will be performed according to the study MR protocol. If not, a standard MRI as per routine clinical practice will be performed. The investigators do not intend to recruit patients who are unable to give consent themselves since patients have to have a good Karnofsky score as an inclusion criteria in this study.