Glioblastoma multiforme (GM) is a relatively common, extremely aggressive form of brain tumor which, although it rarely metastasizes, is difficult to control because it infiltrates brain tissue. The Brain Tumor Research Group at UCSF have been extremely active in developing and evaluating combinations of treatments for GM patients including surgery, external beam irradiation, interstitial irradiation with implanted radioactive sources (brachytherapy) and chemotherapy. Their current Phase II prospective randomized trial is investigating the benefits of adding hyperthermia to the brachytherapy protocol. The major difficulty in obtaining a reliable assessment of local control, specifically in identifying necrosis from residual or recurrent tumor. In this study, we will combine novel volume MR imaging with 31P and 1H metabolite imaging to see whether these techniques give an accurate functional assessment of local control for patients with GM. By quantifying the spatial distributions of metabolites such as phosphocreatine, phosphodiesters, ATP, N-acetyl aspartate, choline, creatine and lactate at six timepoints during therapy, we aim to characterize differences in metabolism in normal brain, necrosis and tumor. Previous results by the investigators and others strongly support the hypothesis that metabolite levels differ significantly in these regions. The metabolite distributions will be directly related to radiation dose distribution, contrast enhanced images obtained in routine followup CT or MR examinations and, for the approximately 50% of patients who undergo reoperation for tumor progression, with pathology reports. This proposal is unique in that it provides the first application of advanced volume MR imaging, spectroscopy and data analysis techniques to a well controlled population of patients participating in an established, randomized clinical trial of treatment efficacy. The MR spectroscopy techniques which will be used, 3-D 31P CSI and 3-D CSI plus water suppressed STEAM or PRESS provide arrays of spectra from regions of normal brain and tumor. The assessment of abnormal signal intensities will thus be based upon the metabolite levels in the patients own normal tissue, giving a much more sensitive and specific assay than previous studies using single voxel localized spectroscopy. In this way, the spatial extent of the tumor and necrosis will be identified by their abnormal metabolite levels and related to the anatomy. Similarly, alignment of the anatomy based upon volume MR images from sequential examinations will allow more accurate interpretation of changes in metabolite levels during treatment. By the end of the study, we will be able to assess the clinical utility of our techniques and will have built up a more complete understanding of changes in tissue metabolism which occur in response to this aggressive, multimodality treatment regime. In the longer term, the extension of clinical MR studies to include data acquisition for metabolite imaging may allow a combined morphological and functional imaging exam, capable of evaluating treatment by identifying regions of necrosis and tumor progression.
Showing the most recent 10 out of 15 publications
