Human Brain Tumors: Advances in Diagnosis & Prognosis
Click on to view the abstract pdf and click on to view the video presentation.
 
Monday May 9th
Room 710B  16:30 - 18:30 Moderators: Sarah Nelson and A. Greg Sorenson

16:30 180.   Feature Analysis in SVM-based Classification of Gliomas  
Frank G Zoellner1, Kyrre E Emblem2,3, and Lothar R Schad1
1Computer Assisted Clinical Medicine, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany, 2Department of Radiology, MGH-HST A.A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital and Harvard Medical School, Boston, MA, United States, 3The Interventional Center, Oslo University Hospital, Oslo, Norway

 
DSC-MRI is a method of choice to differentiate high-grade from low-grade gliomas. Recently, support vector machine (SVM) learning have been introduced as means to prospectively characterize gliomas based on the rCBV histograms. In our study, we have assessed the diagnostic accuracy of the different histogram features used in the SVM analysis (peak height, skewness, etc). By using correlation analysis to reduce 95% of the feature information, a classification accuracy of 88.1% was yielded. Our results suggest that a careful examination of the features in SVM based glioma grading could reduce the number of features substantially, thereby improving the effectiveness of the SVM analysis while maintaining a good classification score.

 
16:42 181.   In vivo detection of IDH mutations in gliomas by 1H-MRS 
Changho Choi1, Sandeep Ganji1, Ralph DeBerardinis1, Zoltan Kovacs1, Robert Bachoo1, Juan Pascual1, Ivan Dimitrov1,2, Bruce Mickey1, Craig Malloy1,3, and Elizabeth Maher1
1University of Texas Southwestern Medical Center, Dallas, Texas, United States, 2Philips Medical Systems, 3VA North Texas Health Care System

 
Recent studies indicate that a high fraction of gliomas contain mutations in the metabolic enzymes, isocitrate dehydrogenase-1 and -2, IDH1 and IDH2. As a result, 2-hydroxyglutarate (2HG), which is normally present in vanishingly small quantities, can be elevated by orders of magnitude in gliomas bearing IDH1 or IDH2 mutations. Here, we report for the first time in vivo detection of this oncometabolite in human brain tumors by 1H-MRS at 3T. 2HG was detected with CRLB < 20% in 38 spectra out of 122 spectra from 49 tumor patients (29 GBM, 9 anaplastic astrocytoma, and 11 low-grade glioma), the concentrations being 2 - 8 mM.

 
16:54 182.   Metabolic characterization of glioma populations with emphasis on onco-metabolite 2-hydroxyglutarate 
Adam Elkhaled1, Llewellyn Jalbert1, Hikari Yoshihara1, Gabriella Bourne1, Joanna Phillips2, Soonmee Cha1, Susan M Chang3, Radhika Srinivasan1, and Sarah J Nelson1,4
1Department of Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States, 2Department of Pathology, University of California, San Francisco, 3Department of Neurological Surgery, University of California, San Francisco, 4Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco

 
Low grade gliomas with mutations in the isocitrate dehydrogenase-1 (IDH1) gene carry a significant survival advantage, and have recently been shown to manifest excessive production of 2-hydroxyglutarate (2HG). Given the potential prognostic value of 2HG as an onco-metabolite, we sought to characterize differences in metabolite levels between gliomas of distinct origin and grade using proton high-resolution magic-angle-spinning (1H HR-MAS) spectroscopy. Results indicated that: grade 2 glioma positive for 2HG possess a distinctive metabolic profile; 2HG presence can distinguish between primary and secondary GBM; and transformation to a higher grade can be characterized metabolically

 
17:06 183.   Presence of 2-Hydroxyglutarate in IDH1 mutated low-grade glioma using ex vivo proton HR-MAS spectroscopy 
Llewellyn Jalbert1, Adam Elkhaled1, Joanna Phillips2, Hikari Yoshihara1, Radhika Srinivasan1, Gabriela Bourne1, Susan Chang3, Soonmee Cha1, and Sarah Nelson1,4
1Department of Radiology and Biomedical Imaging, University of California - San Francisco, San Francisco, CA, United States, 2Department of Pathology, University of California - San Francisco, 3Department of Neurological Surgery, University of California - San Francisco, 4Department of Bioengineering and Therapeutic Sciences, University of California - San Francisco

 
Significant survival advantage is conferred to glioma patients harboring mutations in isocitrate dehydrogenase 1 and 2 (IDH1/2), resulting in the excessive production of the onco-metabolite 2-hydroxyglutarate (2HG). We aimed to determine whether 2HG could be detected in recurrent low-grade glioma samples using proton High-resolution Magic Angle Spinning (1H HR-MAS) spectroscopy, and to correlate the results with IDH1 mutation status assessed by immunohistochemistry. We have confirmed 2HG presence by 1D and 2D 1H HR-MAS in concordance with IDH1 status; its presence gives rise to a complicated spectral pattern that may be of significant clinical utility as a novel biomarker in glioma.

 
17:18 184.   Detection of 2-hydroxyglutarate in Mutant Brain Tumors in vivo using Proton Magnetic Resonance Spectroscopy 
Rajakumar Nagarajan1, Michael Albert Thomas1, Whitney B Pope1, Robert M Prins2, Neil Wilson1, Noriko Salamon1, and Linda M Liau2
1Radiological Sciences, University of California Los Angeles, Los Angeles, California, United States, 2Neurosurgery, University of California Los Angeles

 
The mutation observed in the isocitrate dehydrogenase1 (IDH1) gene, which occurs in the majority of grade II and grade III gliomas and secondary glioblastomas shows significant elevation of 2-hydroxyglutarate (2HG) in the brain tumors. We have quantified 2HG in the brain tumor patients using proton (1H) MR spectra recorded on a 3T MRI/MRS scanner and post-processed by LC-model algorithm. In this study, we show increased 2HG in the mutant type compared to wild type tumors. 1H MRS enables a non-invasive measure of 2HG in gliomas, which may serve as a potential biomarker for monitoring patients with IDH1 mutant brain tumors.

 
17:30 185.   Segmentation of combinations of mean diffusivity and DCE perfusion derived CBV in Glioblastoma multiforme 
Rishi Awasthi1, Ram Kishan Singh Rathore2, Jitesh Kumar Singh2, Nuzhat Husain3, Priyanka Soni3, Rohit Kumar Singh4, Sanjay Behari4, Rakesh Kumar Gupta1, and Shaleen Kumar5
1Radiodiagnosis, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow, India, Lucknow, Uttar Pradesh, India, 2Mathematics & Statistics, Indian Institute of Technology, Kanpur, Kanpur, Uttar Pradesh, India, 3Pathology, Chatrapati Sahu ji Maharaj Medical University, Lucknow, Uttar Pradesh, India, 4Neurosurgery, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow, India, Lucknow, Uttar Pradesh, India, 5Radiotherapy, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow, India, Lucknow, India

 
Forty-four untreated patients with a definitive histopathological diagnosis of glioblastoma multiforme were imaged using conventional, DT and DCE MRI. It was possible to segment the tumor tissue by using DCE derived CBV maps and DTI derived MD maps. From these segmented regions, four combinations of CBV and MD values were obtained which represents different tumoral components also evident on immunohistochemistry. We conclude that the combination of DTI and DCE-MRI allows segmentation which classifies different tissue types within the tumor which in turn may prove to be valuable in targeted therapy in future

 
17:42 186.   Evaluation of relative CMRO2 from BOLD and CBF changes in hyperoxia: Significant increase of oxygen consumption rate in glioblastoma 
Heisoog Kim1,2, Ciprian Catana1, Kim Mouridsen1, Div Bolar1, Elizabeth R Gerstner3, Tracy T Batchelor3, Rakesh K Jain4, Bruce R Rosen1,2, and A Gregory Sorensen1
1Radiology, A. A. Martinos Center, Charlestown, MA, United States, 2HST/NSE, Massachusetts Institute of Technology, Cambridge, MA, United States, 3Neurology, Massachusetts General Hospital, Boston, MA, United States, 4Radiation Oncology, Massachusetts General Hospital, Boston, MA, United States

 
Cerebral metabolic rate of oxygen (CMRO2) is one of the characteristics utilized to investigate metabolic changes in tumor oxygenation under different baseline physiologies. Based on simultaneous BOLD-ASL measurements, relative CMRO2 in hyperoxia in glioblastoma (GBM) were quantitatively evaluated. Elevated BOLD and reduced CBF signal changes during 100% oxygen breathing were observed in glioblastoma. Our data validated the coupling between BOLD and CBF in Davisí model in GBM patients. Remarkably, oxygen-induced relative CMRO2 estimated from these measurements showed a significant increase (44%) in tumor and peritumoral regions. It implies that hyperbaric oxygen administration manipulates the oxygenation metabolism in cancer cells.

 
17:54 187.   A Simultaneous Measurement of Relative CMRO2 with MRI and FMISO Uptake with PET in Glioblastoma 
Heisoog Kim1, Ciprian Catana1, Grae Arabasz1, Div Bolar1, Elizabeth R Gerstner2, Tracy T Batchelor2, Rakesh K Jain3, Bruce R Rosen1, and A Gregory Sorensen1
1Radiology, A. A. Martinos Center, Charlestown, MA, United States, 2Neurology, Massachusetts General Hospital, Boston, MA, United States, 3Radiation Oncology, Massachusetts General Hospital, Boston, MA, United States

 
The specific goal of this study was to evaluate relative CMRO2 changes in relation to hypoxia level based on FMISO uptake in newly diagnosed glioblastoma (nGBM) patients undergoing chemo-, radio- and anti-angiogenic therapy using an integrated MRI-PET scanner. We observed a significant increase of relative CMRO2 (42%) and FMISO uptake (2.45 SUV) around the enhancing tumor at the baseline. However, the relative CMRO2 (19%) and FMISO uptake (1.55 SUV) largely decreased with tumor regression by the combined treatment. Our preliminary findings have important implications for the role of oxygen in supporting tumor metabolism and angiogenesis.

 
18:06 188.   Multicentre prospective classification of childhood brain tumours based on metabolite profiles derived from 1H MRS 
Nigel Paul Davies1,2, Simrandip Gill2,3, Theodoros N Arvanitis3,4, Dorothee Auer5, Richard Grundy6,7, Franklyn A Howe8, Darren Hargrave9, Tim Jaspan7, Lesley MacPherson3, Kal Natarajan1,3, Geoffrey Payne9,10, Dawn Saunders11, Yu Sun2,3, Martin Wilson2,3, and Andrew C Peet2,3
1Medical Physics, University Hospitals Birmingham NHS Foundation Trust, Birmingham, United Kingdom, 2Cancer Sciences, University of Birmingham, Birmingham, United Kingdom, 3Birmingham Children's Hospital NHS Foundation Trust, Birmingham, United Kingdom, 4Department of Electrical, Electronic, and Computer Engineering, University of Birmingham, Birmingham, United Kingdom, 5Academic Radiology, University of Nottingham, Nottingham, United Kingdom, 6Children's Brain Tumour Research Centre, University of Nottingham, Nottingham, United Kingdom, 7University Hospital Nottingham, Nottingham, United Kingdom, 8St. George's University of London, London, United Kingdom, 9Royal Marsden Hospital, London, United Kingdom, 10Institute of Cancer Research, London, United Kingdom, 11Great Ormond Street Hospital, London, United Kingdom

 
1H MRS provides non-invasive metabolite profiles of brain tumours aiding diagnosis and potentially improving characterisation. In this study we perform a large prospective multicentre evaluation of 1H MRS as a diagnostic tool for grading childhood brain tumours. The classifier was trained on metabolite profiles derived using TARQUIN from 123 single-voxel MRS acquired using a standard protocol on two 1.5T scanners in a single centre. Testing was performed using 110 cases acquired prospectively across 4 different centres with some variations in echo time and field strength. The overall classification accuracy for identifying high grade versus low grade tumours was 86%.

 
18:18 189.   Segmentation of tumor infiltrative and vasogenic edema in brain tumors using voxel-wise analysis of 11C-methinonine and FDG PET and its comparison with diffusion tensor imaging 
Manabu Kinoshita1, Testu Goto1, Hideyuki Arita1, Naoki Kagawa1, Yasunori Fujimoto1, Haruhiko Kishima1, Yoichi Saitoh2, Jun Hatazawa3, Naoya Hashimoto1, and Toshiki Yoshimine1
1Neurosurgery, Osaka University Graduate School of Medicine, Suita, Osaka, Japan, 2Neuromodulation and Neurosurgery, Center for Advanced Science and Innovation, Osaka University, Suita, Osaka, Japan, 3Nuclear Medicine and Tracer Kinetics, Osaka University Graduate School of Medicine, Suita, Osaka, Japan

 
Use of DTI has been proposed for the detection of glioma cell invasion. However, other reports have questioned its value. We have been able to segment areas with and without tumor cell invasion in the T2-high intensity area in malignant gliomas using a voxel-wise analysis of 11C-methionine and FDG PET. We also showed the profile of FA and ADC does not differ between these two. Our results suggest the possibility of using voxel-wise analysis of 11C-methionine and FDG PET for tumor cell invasion in the T2-high intensity area in glioma patients and questioning the use of DTI for this purpose.