Benjamin M Ellingson¹, Timothy F Cloughesy², Albert Lai², Phioanh L Nghiemphu², and Whitney B Pope^1
1Radiological Sciences, David Geffen School of Medicine, University of California, Los Angeles, Los Angeles, CA, United States, ²Neurology, David Geffen School of Medicine at UCLA, Los Angeles, CA, United States

Functional diffusion mapping (fDM) uses the voxel-wise changes in apparent diffusion coefficient (ADC) measured in the same patient over time as a biomarker for cancer response to therapy. FDMS have been shown to be predictive of response and survival in glioblastoma under a variety of treatment paradigms. The current study develops a probabilistic approach to fDM quantification, where finite translational and rotational perturbations are performed after initial registration of ADC maps. These probabilistic fDMs were applied to newly diagnosed glioblastoma patients treated with standard radiochemotherapy (n = 143). Results suggest probabilistic fDMs have higher clinical sensitivity compared with traditional fDMs.

Janine M Lupo¹, Mekhail Anwar², Christopher P Hess¹, Susan M Chang³, and Sarah J Nelson^1,4
1Department of Radiology and Biomedical Imaging, University of California, San Francisco, CA, United States, ²Department of Radiation Oncology, University of California, San Francisco, CA, United States, ³Department of Neurosurgery, University of California, San Francisco, CA, United States,⁴Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco, CA, United States

Despite the damage that radiation can cause to microvasculature, radiotherapy is an integral component in the management of patients with glioma. This study uses SWI to investigate the effects of radiation dose on microbleed characteristics in glioma patients with and without adjuvant anti-angiogenic therapy. The number of microbleeds was found to increase with escalating dose level at 2 years after radiation was administered. Microbleed size was not related to radiation dose and varied substantially in patients who received anti-angiogenic therapy. Although initially microbleeds occur in high dose regions, as time progressed, more microbleeds appear in lower dose regions.

Mohan Pauliah¹, Rachel Bartlett¹, Philip H Gutin², Heiko Schoder², Cameron Brennan², John Humm^1,3, Pat Zanzonico³, and Michelle S Bradbury^1,4
1Radiology, Memorial Sloan Kettering Cancer Center, New York, New York, United States, ²Neurosurgery, Memorial Sloan Kettering Cancer Center, New York, New York, United States, ³Medical Physics, Memorial Sloan Kettering Cancer Center, New York, New York, United States, ⁴Molecular Imaging, Memorial Sloan Kettering Cancer Center, New York, New York, United States

By integrating functional and metabolic imaging technologies, increasingly more sensitive and specific read-outs reflecting the biological status of tumors may be obtained which, in turn, may improve characterization and direct targeted biopsy efforts. The purposes of our study are three-fold: (1) investigate the feasibility of deriving voxel-wise parameter estimates reflecting tumor cell proliferation, metabolic flux, and microvascular permeability in high grade glial tumors from precisely co-registered dynamic 18F-FLT PET and DCE-MRI with histologic correlation and (2) to investigate relationships among voxel-based determinations of PET- and MRI-based kinetic parameters during the first-pass and equilibrium phases of the study. In addition, targeted biopsy specimens were acquired using frameless stereotactic surgery and intraoperative MRI to determine whether parametric images are predictive of regional histologic assays of tumor cell proliferation and microvascular density. Quantitative accuracy of parametric images is also being evaluated through comparison of histologic determinations and gene expression differences, obtained at different biopsy locations, with the PET-MRI imaging features at those same locations. Given the high degree of heterogeneity of glial tumors, the molecular characteristics and functional imaging correlation on the same tumor areas may facilitate the development of novel therapies and yield increasingly more accurate prognostic information on the basis of key biomarkers.

Timothy L Jones¹, B Anthony Bell¹, and Thomas R Barrick^2
1Academic Neurosurgery Unit, St Georges University of London, London, United Kingdom, ²Stroke and Dementia Research Centre, St Georges University of London, London, United Kingdom

Conventional MRI images of brain tumours likely underestimate the true extent of tumour and have variable sensitivity in diagnosis depending on tumour type. In this study, DTI scans were acquired from 94 patients with a range of brain tumours. We present a novel whole brain segmentation and visualisation technique which reveals potential tumour-specific diagnostic characteristics. Discriminant analysis of tumour characteristics reveals diagnostic sensitivities ranging from 84-100%

Liya Wang¹, Anne Carroll², Juliya Kalinina³, Erwin Van Meir⁴, Qiqi Yu⁵, Junjun Tan², Ruya Zhao², Frank Liu⁶, Shaoxiong Wu⁶, and Hui Mao^7
1Radiology, Emory University School of Medicine, Atlanta, Georgia, United States, ²Chemistry, Emory University, Atlanta, GA, United States,³Neurosurgery, Emory University School of Medicine, Atlanta, Georgia, United States, ⁴Neurosurgery, Emory University School of Medicine, Atlanta, GA, United States, ⁵Radiology, Emory University School of Medicine, Atlanta, GA, United States, ⁶NMR Center, Emory University, Atlanta, GA, United States,⁷Radiology, Emory University, Atlanta, GA, United States

Solid state high resolution (HR) magic angle spinning (MAS) NMR methods were used to identify the spectroscopic ¡°finger print¡± of 2HG in brain tumor tissues to confirm that 2HG is the metabolite marker of IDH mutations. The metabolite profiles of gliomas with and without IDH mutations are investigated quantitatively. The results confirmed that 2HG is a highly specific metabolite marker of IDH mutations in gliomas. In addition, HRMAS NMR analysis provided quantitative metabolite analysis that revealed the altered metabolite profile in tumors bearing with IDH mutations

Peter S. LaViolette¹, Elizabeth J Cochran², Alex D Cohen³, Mona Al-Gizawiy¹, Scott D. Rand⁴, Jennifer Connelly⁵, Mark G. Malkin⁶, Wade M Mueller⁷, and Kathleen M Schmainda^8
1Radiology, Medical College of Wisconsin, Milwaukee, WI, United States, ²Pathology, Medical College of Wisconsin, Milwaukee, WI, United States,³Biophysics, Medical College of Wisconsin, Milwaukee, WI, United States, ⁴Radiology, Medical College of Wisconsin, ⁵Neurology, Medical College of Wisconsin, ⁶Neurology & Neurosurgery, Medical College of Wisconsin, ⁷Neurosurgery, Medical College of Wisconsin, Milwaukee, WI, United States,⁸Radiology and Biophysics, Medical College of Wisconsin

AFMM (ADC FLAIR MisMatch) is defined as an area of low ADC within FLAIR abnormality associated with a brain tumor. The aim of this study is to determine if areas of AFMM represent increased cellularity by histologic analysis of spatially-correlated ex-vivo issue acquired and processed in three patients with high grade gliomas.

Sandeep K Ganji¹, Abhishek Banerjee¹, Zoltan Kovacs¹, Ivan E Dimitrov^1,2, Ralph J DeBerardinis³, Craig M Malloy¹, Bruce Mickey⁴, Robert Bachoo⁵, Elizabeth A Maher⁶, and Changho Choi^1
1Advanced Imaging Research Center, UT Southwestern Medical Center, Dallas, Texas, United States, ²Philips Medical Systems, Cleveland, Ohio, United States, ³Pediatrics, UT Southwestern Medical Center, Dallas, Texas, United States, ⁴Neurological Surgery, UT Southwestern Medical Center, Dallas, Texas, United States, ⁵Neurology, UT Southwestern Medical Center, Dallas, Texas, United States, ⁶Internal Medicine, UT Southwestern Medical Center, Dallas, Texas, United States

2-hydroxyglutarate (2HG) is known to accumulate in a vast majority of low grade gliomas and secondary glioblastomas. We report in vivo spectroscopic imaging of 2HG using an optimized echo-time PRESS-based localization method at 3T in patients with gliomas. We present phantom and in vivo healthy volunteer validation data and show the elevated 2HG levels in tumor patients. Metabolite concentration maps are presented.

Bjorn Stemkens^1,2, Summer Fakhro², Daniel B Chonde², Marco C Pinho², Pavlina Polaskova², Dominique L Jennings², Kyrre E Emblem², Elizabeth R Gerstner³, Tracy T Batchelor³, and Ciprian Catana^2
1Department of Biomedical Engineering, University of Technology Eindhoven, Eindhoven, Netherlands, ²Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital and Harvard Medical School, Boston, MA, United States, ³Massachusetts General Hospital Cancer Center, Harvard Medical School, Boston, MA, United States

Diffusion tensor imaging (DTI) and positron emission tomography (PET) are both imaging techniques that are able to provide information about tumor cellularity and infiltration in the peritumoral region. The purpose is to observe their response in glioblastoma treatment with anti-angiogenic drugs. PET and DTI data are collected simultaneously and baseline measurements are compared to day one after treatment onset. Significant negative correlations were found between mean standard uptake values (SUV) and fractional anisotropy (FA) in the peritumoral region. This implies that metabolic changes correlate with white matter directional changes.

Emma Essock-Burns^1,2, Joanna J. Phillips^3,4, Annette M. Molinaro³, Janine M. Lupo², Soonmee Cha^2,3, Susan M. Chang³, and Sarah J. Nelson^1,5
1Graduate Group in Bioengineering, UC Berkeley/UC San Francisco, San Francisco, CA, United States, ²Department of Radiology and Biomedical Imaging, UC San Francisco, San Francisco, CA, United States, ³Neurological Surgery, UC San Francisco, San Francisco, CA, United States, ⁴Department of Pathology, UC San Francisco, San Francisco, CA, United States, ⁵Bioengineering and Therapeutic Sciences, UC San Francisco, San Francisco, CA, United States

Choosing a biopsy location that reflects tumor grade is a challenge for diagnosis and treatment of patients with glioblastoma, a heterogeneous and vascularized brain tumor. Multiple post-processing methods exist to correct for contrast-agent extravasation in dynamic susceptibility contrast (DSC)-MRI near glomeruloid vasculature and resulting CBV estimates can vary substantially. Vascular morphology of 72 image-guided biopsies from 35 de-novo GBM patients was measured with Factor VIII immunohistochemical staining and the corresponding DSC hemodynamic data analyzed using both non-linear and non-parametric post-processing techniques. This study provides histopathologial evidence that CBV estimates from both non-linear and non-parametric post-processing significantly predict underlying vascular morphology.

Y. Sui^1,2, H. Wang³, GZ Liu¹, F. W. Damen¹, C. Wanamaker⁴, YH Li⁵, and X. J. Zhou^1,6
1Center for MR Research, University of Illinois Medical Center, Chicago, IL, United States, ²Bioengineering, University of Illinois at Chicago, Chicago, IL, United States, ³Applied Science Lab, GE Healthcare, Shanghai, China, ⁴Radiology, University of Illinois Medical Center, Chicago, IL, ⁵Radiology, Xinhua Hospital, Shanghai, China, ⁶Departments of Radiology, Neurosurgery and Bioengineering, University of Illinois Medical Center, Chicago, IL

The majority of diffusion MRI applications for brain tumor employed a single b-value to yield apparent diffusion coefficient (ADC) and/or fractional anisotropy using a mono-exponential model. This simple model disguises valuable information, such as tissue heterogeneity and microenvironment, for differential diagnosis. In this study, we investigate the feasibility of using a set of new parameters from the fractional order calculus (FROC) model to differentiate low-grade from high-grade pediatric brain tumors. Our results demonstrate a significant difference in FROC parameters between the two tumor groups. Compared with ADC, the new parameter ¦Â, which has been shown to reveal the difference in tissue heterogeneity, exhibits better performance in differentiation