This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. Primary support for the subproject and the subproject's principal investigator may have been provided by other sources, including other NIH sources. The Total Cost listed for the subproject likely represents the estimated amount of Center infrastructure utilized by the subproject, not direct funding provided by the NCRR grant to the subproject or subproject staff. Proton magnetic resonance spectroscopy (1H-MRS) provides an effective tool for measuring metabolites in the human brain noninvasively in vivo. Precise measurement of several clinically-important, low-abundance metabolites by standard 1H-MRS is often elusive even at 7T, due to the spectroscopic complexities arising from the scalar coupling effects and macromolecule baseline signals. The objectives of the research are to develop MRS techniques that provide improved inter-metabolite discrimination for detection of glycine, serine, N-acetylaspartyl-glutamate (NAAG), glutamine, gamma-aminobutyric acid, etc., at both 7T and 3T. The specific clinical goal is to develop better diagnostic tools for cancer involving the brain. Patients with brain tumors, either a glioblastoma (GBM) or a metastasis from systemic cancer, face a devastating clinical course consisting of progressive physical and cognitive decline over little more than a year, despite aggressive multimodality therapy. Management decisions in all phases of diagnosis, treatment and follow up rely on MR imaging based on interpretation of changes in gadolinium enhancement and T2/FLAIR (fluid attenuated inversion recovery) signal from one time point to another. However, this is fraught with conflicting interpretations which can alter management and can have profound impact on prognosis. The uncertainty is most commonly centered around differentiating tumor progression from treatment effect following radiation, a common scenario for both GBMs and metastases patients. Classic tumor progression is associated with increased size of a gadolinium enhancing mass with associated increased T2-weighted FLAIR. However, the identical imaging characteristic are ascribed to classic radionecrosis and in GBM is called "pseudoprogression" because it is impossible to differentiate from true tumor progression by available imaging. Thus the treating neuro-oncology team relies on "best guess" based on the clinical context. Since the prognosis and management is vastly different depending on whether it is tumor progression or treatment effect, patients are often taken back to the operating room for a repeat craniotomy and biopsy to make the diagnosis. Of similar importance is the clinical scenario in which changes in T2/FLAIR signal may herald early progressive disease prior to the development of tumor enhancement but is indistinguishable from peritumoral edema or reactive gliosis. Thus, overall, there is an urgent need for the development of novel imaging techniques that would aid in non-invasive diagnosis at these critical decision points. Numerous methods to improve the characterization of brain tumors have been proposed based on the integration of MRI with other techniques such as positron emission tomography (MR-PET), MR imaging of 23Na, and MR spectroscopy of 31P or 13C. While all of these methods are attractive for studying the basic biology of the disease, they are limited in clinical application because of the substantial investment in additional equipment needed. Improved diagnostic methods that rely on widely-available 1H equipment such as 1H spectroscopy and chemical exchange saturation transfer (CEST) offer a relatively simple progression to clinical research and both methods are inherently improved by the increased chemical shifts dispersion at 7T and can be implemented in a standard 1H imaging system. In this proposal we take advantage of the well characterized human orthotopic mouse models for GBM and MT, described in Project 2, the availability of high field imaging for animals and humans, and the open enrolling clinical imaging protocol for brain tumor patients at 7T. The overarching goal is to identify signature patterns by 1H-MRS and CEST imaging that can differentiate tumor progression from treatment effect and detect early progression in regions beyond the borders delineated by standard T2/FLAIR and gadolinium. Aim 1: To characterize the 1H MR spectra and CEST images at 7T in GBM and each of the four most common tumors that metastasize to the brain (melanoma, lung, breast and renal cell cancer) using human orthotopic mouse models, before and after radiation therapy. Validation in patients will be performed in Aim 2. Aim 2: To characterize the 1H MR spectra and CEST images at 7T in patients with low grade gliomas and identify differences with GBM that could be used to generate a signature pattern that would herald progression from low grade to GBM prior to the commonly used clinical or standard MR indicators. Aim 3: To map the brain in patients with low grade glioma, glioblastoma and brain metastases using the signature metabolites identified in Aims 1 and 2 and the characteristic patterns on CEST imaging at 7T to identify infiltrating tumor in gliomas and early recurrence/progression in metastatic disease.