Quantitative Glioblastoma Margin and Infiltration Mapping with advanced diffusion-relaxation MRI

点击翻译按钮获取中文摘要

Abstract We propose to investigate and validate novel MRI pulse sequences and quantitative measures for mapping primary brain tumor margins and infiltration. We will focus on glioblastomas (W.H.O. grade IV gliomas (GBM)), the most prevalent and deadly primary brain tumor in adults. These progressive brain tumors infiltrate into the brain parenchyma and grow with diffuse margins. However, current clinical imaging modalities fail to reliably define the extent of glioma infiltration, negatively impacting patient care. Neurosurgeons are faced with uncer- tainty about what tissue should be removed when planning an optimal resection, and radiation oncologists must design radiation fields based on an incomplete understanding of the tumor's extent. Therefore, there is an unmet need for patient-specific, personalized mapping of tumor margins (including what is within the radiologi- cally defined margin using current state-of-the-art imaging and what is beyond it) in order to improve clinical treatment of gliomas via methods such as surgery, radiation therapy, or drug delivery. Meeting this unmet need requires improved imaging of brain and tumor tissue microstructure. We recently proposed q-space trajectory imaging (QTI), which goes beyond conventional diffusion MRI to measure the correlation of water molecule motion between different directions, improving mapping of tissue and tumor microstructure. Recent work has also demonstrated the potential of quantitative MRI (T2-relaxome- try) for detecting infiltrative tumor growth in the peritumoral area of gliomas. In fact, the joint distribution of dif- fusion-relaxation measures can provide important information that is missing in independently acquired T2-re- laxometry or QTI data alone. In this project, we plan to develop, investigate, and validate rQTI (a novel combi- nation of T2-relaxometry and QTI) for the critical clinical application of glioma margin and infiltration mapping. To reach this goal, first we will create a comprehensive and unique diffusion-relaxation (rQTI) and histology dataset for the study of glioma infiltration and margins. This work will leverage a mouse model in which we will implant patient-derived xenografts obtained from human GBMs to closely recapitulate key features of human brain tumors such as microstructure and infiltration. Second, we will develop an optimized clinical acquisition for computing rQTI-based microstructure measures that are predictive of histology, in under 10 minutes of ac- quisition time. Third, we will validate rQTI-based microstructure measures against histopathology in 30 patients with GBM. Patients will be scanned with the optimized rQTI sequence and tissue samples will be obtained us- ing clinically indicated stereotactic sampling of tissue and/or stereotactic biopsy. Overall, the successful outcome of this study has the potential to improve non-invasive mapping of GBM margins and to reveal infiltration that was previously invisible on imaging. This is expected to provide important information for GBM treatment planning, with the potential of improving patient survival and quality of life.