MR biomarker profile for infiltrative tumor region in malignant glioma

Introduction: MR images are routinely used in brain tumor management and provide an excellent visualization of the global heterogeneity of the tumor. However, since MR intensities have not been linked to biological mechanisms that drive tumor growth and response, they have proven to be inadequate surrogates of tumor presence particularly following treatment. This study will address the inability of MR imaging to accurately define the infiltrating tumor margin, which results in a margin being added to radiotherapy volumes to account for infiltrating cells. Since this volume includes normal cells, radiation dose is reduced. Consequently, gliomas are known to recur within the treatment volume (1) and inclusion of normal brain can impact patient quality of life. The infiltrating tumor region is characterized by complex interactions that are disruptive to brain architecture. Since these regions contain diffusely infiltrating tumor cells, they can have similar cellular density as normal brain. Hence, choline and apparent diffusion coefficient (ADC), typically used to identify tumor regions based on high cellularity, are unlikely to define or locate infiltrative regions. A measure linked with tumor activity and structural disruption is anticipated to provide improved delineation of tumor infiltrative margin. Studies suggest that glioma cells, at the growth margin of an expanding tumor, release glutamate via “system Xc”, an electroneutral amino acid transporter that exchanges cystine for glutamate [2] during the synthesis of cellular antioxidant glutathione. Released glutamate kills normal cells via the excitotoxic mechanism [3] and glutathione protects tumors from immune attack and also makes them insensitive to radiation treatment that is mediated by hydroxyl radicals. The goal of this study was to determine if glutamate is an indicator of tumor infiltrating region. Ex-vivo spectroscopy of image guided biopsies was used to obtain glutamate levels from tumor regions and analyzed a) relative to their spatial location within the tumor b) compared with SMI-31 immunostain from its paired tissue sample and c) investigated relative to fractional anisotropy (FA) which is a diffusion imaging measure of structural anisotropy indicative of disrupted brain architecture. SMI-31 is a mouse monoclonal antibody that stains neurofilaments [4] and therefore provides an indication of disrupted neuronal tracts.