Joint Annual Meeting ISMRM-ESMRMB 2014 10-16 May 2014 Milan, Italy

Advanced Brain Tumor 1

Tuesday 13 May 2014
Red 1 & 2  10:00 - 12:00 Moderators: Arturo Brunetti, M.D., Armando Tartaro, M.D.

10:00 0280.   DSC-MRI Measures of Relative Cerebral Blood Volume (rCBV) as a Prognostic Marker for Progression-Free and Overall Survival in Recurrent Glioblastoma: Results from the ACRIN 6677/RTOG 0625 Multi-Center Trial
Kathleen M. Schmainda1, Zheng Zhang2, Jerrold Boxerman3, Melissa Prah1, Bradley Snyder2, Devyani Bedekar4, A Gregory Sorensen5, Mark R Gilbert6, and Daniel P Barboriak7
1Radiology, Medical College of Wisconsin, Milwaukee, Wisconsin, United States, 2Center for Statistical Sciences, Brown University, Rhode Island, United States, 3Diagnostic Imaging, Rhode Island Hospital, Rhode Island, United States, 4Medical College of Wisconsin, Milwaukee, Wisconsin, United States,5Massachusetts General Hospital, Massachusetts, United States, 6Neuro-Oncology, University of Texas, Houston, Texas, United States, 7Radiology, Duke University, Durham, North Carolina, United States

RTOG 0625/ACRIN 6677 is a multicenter, randomized, phase II trial of bevacizumab with irinotecan or temozolomide in recurrent glioblastoma (GBM). In this study of 23 patients the percent changes in both normalized and standardized rCBV, derived from DSC-MRI, are predictive of progression free survival (PFS) and overall survival (OS) when measured at 2 and 16 weeks following treatment initiation.

10:12 0281.   Identification of Arterio-Venous Shunts by Vessel Architectural Imaging Reveals Mechanisms of Vascular Normalization during Anti-Angiogenic Therapy
Guro K. Rognsvag1, Atle Bjornerud1,2, A. Gregory Sorensen3,4, Patrick Y. Wen5, Tracy T. Batchelor6,7, Rakesh K. Jain6, and Kyrre E. Emblem1,3
1The Intervention Centre, Oslo University Hospital, Oslo, Oslo, Norway, 2Dept of Physics, University of Oslo, Oslo, Norway, 3Department of Radiology and Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital and Harvard Medical School, Boston, MA, United States,4Siemens Healthcare, Malvern, PA, United States, 5Center for Neuro-Oncology, Dana-Farber/Brigham and Women's Cancer Center and Harvard Medical School, Boston, MA, United States, 6Department of Radiation Oncology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, United States, 7Department of Neurology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, United States

The microvasculature of tumors is abnormal and tortuous with arterio-venous shunts. Shunts are short high-flow vascular pathways that cause parts of the blood flow to bypass capillary regions, thus impairing delivery of oxygen to the tissue and increasing resistance to therapy. Vessel architectural imaging (VAI) has recently been introduced as a new paradigm for in vivo assessment of cancer vasculature. By performing Monte Carlo simulations of normal and shunting tissue, and evaluating MRI data of patients undergoing anti-angiogenic therapy, we show that VAI identifies arterio-venous shunts and help reveal mechanism of normalization during anti-angiogenic therapy.

10:24 0282.   
Multiparametrical diffusion tensor imaging for the detection of anaplastic transformation of low-grade gliomas
Martin Thomas Freitag1, Christian Weber1,2, Klaus Maier-Hein1,2, Franciszek Binczyck3, Barbara Bobek-Billewicz4, Joanna Polanska3, Rafal Tarnawski4, and Bram Stieltjes1
1Section Quantitative Imaging-based Disease Characterization, Department of Radiology, German Cancer Research Center, Heidelberg, Baden-Württemberg, Germany, 2Division of Medical and Biological Informatics, German Cancer Research Center, Heidelberg, Baden-Württemberg, Germany,3Silesian University of Technology, Gliwice, Poland, 4Institute of Oncology, Maria Sklodowska-Curie Memorial Cancer Center, Gliwice, Poland

Here, the value of multiparametrical diffusion-tensor-imaging (DTI) maps was tested for the often difficult detection of recurrent low-grade glioma in 48 adults. At the time point of recurrence, defined by new contrast uptake, hypointense lesions were seen within the T2-hyperintensity zone in fractional anisotropy, axial, mean and radial diffusitivity maps in every patient representing a hypercellular equivalent. Comparing all tensor-derived maps quantitatively, contrast-to-noise and combined sensitivity/specificity was highest for axial diffusitivity to visualize and detect this hypercellularity. DTI acquisitions should be routinely included for the management of low-grade gliomas as axial diffusitivity maps may provide essential additional diagnostic information.

10:36 0283.   Diffusion Tensor Imaging-based Histogram Analysis for detection and quantification of low-grade gliomas structural changes during chemotherapy
Antonella Castellano1, Marina Donativi2, Roberta Rudà3, Marco Riva4, Giorgio De Nunzio2, Antonella Iadanza1, Matteo Rucco5, Luca Bertero3, Lorenzo Bello4, Riccardo Soffietti3, and Andrea Falini1
1Neuroradiology Unit and CERMAC, Università Vita-Salute San Raffaele and Ospedale San Raffaele, Milan, Italy, 2A.D.A.M., Advanced Data Analysis in Medicine, University of Salento, Lecce, Italy, 3Neuro-oncology, Department of Neuroscience and Oncology, University of Torino, Turin, Italy,4Neurosurgery, Department of Neurological Sciences, Università di Milano, Istituto Clinico Humanitas, Milan, Italy, 5School of Science and Technology, Computer Science Division, University of Camerino, Camerino, MC, Italy

In low-grade gliomas during chemotherapy, changes in Diffusion Tensor Imaging metrics (MD, FA, pure isotropy and pure anisotropy) may be an early signature for Temozolomide response, correlating with clinical response better than conventional MRI criteria: structural changes quantified by DTI-based histogram analysis support the hypothesis of a “shrinkage” of the tumor in response to chemotherapy, especially at the level of margins of infiltration. Quantitative measures derived from this analysis can improve the monitoring of low-grade gliomas treatment response.

10:48 0284.   Application of compressed-sensing-accelerated diffusion spectrum imaging in patients with brain tumors
Ek T Tan1, Robert J Young2, Kyung K Peck2,3, Xiaofeng Liu1, Jonathan I Sperl4, Mehrnaz Jenabi2, and Luca Marinelli1
1GE Global Research, Niskayuna, NY, United States, 2Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, United States, 3Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY, United States, 4GE Global Research, Garching, Munich, Germany

Compressed-sensing accelerated diffusion spectrum imaging (CS-DSI) was compared to conventional DTI in ten patients with intracranial brain tumors, whereby DSI clearly provides improved performance in the presence of crossing-fibers and (edema and tumors). DSI acquisition was accelerated by factors of 4 and 5 to achieve scan times under 14 minutes on a conventional 3T system. CS-DSI provided superior tract visualization in all patients. The tract-based dependent measures obtained with CS-DSI were more consistent with pathology-induced changes than those with DTI. These preliminary results suggest that CS-DSI may be helpful in preoperative planning in patients with brain tumors.

11:00 0285.   Low-frequency Oscillations Characterize Motor Network Plasticity Abnormalities in Patients with Brain Gliomas - permission withheld
Chen Niu1, Pan Lin2, Zhigang Min1, Rana Netra1, Qiuli Zhang1, Xin Liu2, Cuiping Mao1, Faxiu Bao1, and Ming Zhang1
1The First Affiliated Hospital of Medical College, Xi¡¯an Jiaotong University, Xi'an, Shaanxi, China, 2Institute of Biomedical Engineering, Xi'an Jiaotong University, Shaanxi, China

Brain plasticity is a continuous process during slow-growing tumor formation, which remodels neural organization and optimizes brain network function. In this study, we aimed to investigate whether motor function plasticity exists in patients with slow-growing brain tumors located in or near to motor areas, and who exhibited no motor deficits. We use resting-state functional magnetic resonance (rs-fMRI) data in the frequency domain, and investigate the relationship between the low frequency band shift and motor functional plasticity changes in patients with brain tumor, and achieve a better understanding of underlying mechanisms.

11:12 0286.   
Experience with q-ball language tracking in brain tumor patients
Eduardo Caverzasi1, Shawn Hervey-Jumper2, Valentina Panara3, Caroline A. Racine2, Vanitha Sankaranarayanan4, Nico Papinutto1, Kesshi Jordan5, Jing Li2, Mitchel S. Berger2, and Roland G. Henry1
1Department of Neurology, University of California, San Francisco, San Francisco, CA, United States, 2Department of Neurological Surgery, University of California, San Francisco, San Francisco, CA, United States, 3ITAB - Institute of Advanced Biomedical Technologies, University G.D'Annunzio, Chieti, Italy, Italy, 4Department of Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States, 5Graduate Program in Bioengineering, UC Berkeley/UC San Francisco, CA, United States

We show the successful application of q-ball tracking in pre-surgical planning for language pathways in brain tumor patients and for post-surgery white matter tracking in order to assess postoperative tract damages. The rating scales developed for fiber pathways damage were found to be highly reproducible and provided significant correlations with language deficits. The fiber tracking spatial inter-operator reliability was very high considering the intrinsic variability of this technique on a voxel-wise segmentation. Our results confirm the importance of preservation of dorsal stream tracts (arcuate fascicle and temporo-parietal portion of the superior longitudinal fascicle), in order to reduce language morbidity of brain tumor patients.

11:24 0287.   3D MR Spectroscopic Imaging of 2-Hydroxyglutarate in patients with mutant IDH1 glioma
Ovidiu Cristian Andronesi1, Franziska Loebel2, Wolfgang Bogner3, Malgorzata Marjanska4, Elizabeth Gerstner5, Andrew Chi5, Tracy T. Batchelor5, Daniel P. Cahill6, and Bruce R. Rosen1
1Martinos Center for Biomedical Imaging, Department of Radiology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, United States,2Department of Neurosurgery, Massachusetts General Hospital, Boston, MA, United States, 3MR Center of Excellence, Department of Radiology, Medical University Vienna, Vienna, Vienna, Austria, 4Center for Magnetic Resonance Research, University of Minnesota, MN, United States, 5Pappas Center of Neuro-Oncology, Department of Neurology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, United States, 6Department of Neurosurgery, Massachusetts General Hospital, Harvard Medical School, Boston, MA, United States

High levels of 2-hydroxyglutarate (2HG) represent the hallmark metabolic alteration in mutant IDH gliomas. 2HG has been proposed as an ideal biomarker to diagnose and monitor treatment response of mutant IDH gliomas. We show here newly that 3D MR spectroscopic imaging (MRSI) of 2HG is feasible clinically in glioma patients. A novel 3D MRSI sequence was designed to edit reliably and efficiently the 2HG levels. Results obtained serially in mutant IDH glioma patients are shown.

11:36 0288.   An in-vivo 1H MRS study of metabolic correlation in IDH1 vs. IDH2 mutated gliomas
Changho Choi1, Sandeep Ganji1, Akshay Madan1, Zhongxu An1, and Elizabeth Maher1
1UT Southwestern Medical Center, Dallas, Texas, United States

Isocitrate dehydrogenases (IDH) 1 and 2 catalyze the NADP+ dependent conversion of isocitrate to Ą-ketoglutarate in the cytosol and mitochondria, respectively. The mutations in these enzymes result in the production of 2-hydroxyglutarate (2HG). We have investigated metabolic correlations in IDH1- vs. IDH2-mutated gliomas (28 and 5 patients, respectively) focusing on 2HG in vivo, using 1H MRS at 3T. Data indicated 2HG levels were highly correlated with choline, NAA, creatine and glutamate. The correlations were stronger in IDH2 mutated gliomas than in IDH1 mutated gliomas.

11:48 0289.   Towards Improved Characterization of Brain Tumors by 23Na-MR Neuroimaging at 7 Tesla
Armin Biller1,2, Jens Kleesiek3, Jan Oliver Neumann3, Felix Sahm4, Anne Dorothea Hertenstein3, and Armin Michael Nagel5
1Neuroradiology, University of Heidelberg, Heidelberg, Badem-Württemberg, Germany, 2Radiology, German Cancer Research Center DKFZ, Heidelberg, Badem-Württemberg, Germany, 3University of Heidelberg, Baden-Württemberg, Germany, 4Neuropathology, Baden-Württemberg, Germany, 5Medical Physics in Radiology, German Cancer Research Center DKFZ, Baden-Württemberg, Germany

In this study we demonstrate a robust correlation between the Ki-67 proliferation index of brain tumor cells and the relaxation weighted 23Na contrasts (NaR and NaS). Both contrasts represent the signal of ions with short relaxation times. The NaT contrast, which reflects the tissue sodium concentration, exhibited no association with histopathology. Also, statistical analyses revealed no correlation between the Ki-67 index and canonical 1H-MRI including T1w gadolinium enhanced, T2w and FLAIR imaging. These results emphasize the added benefit of 23Na-MRI to conventional 1H imaging in neurooncology. Of note, in contrast to 1H neuroimaging in oncology 23Na-MRI requires no contrast media.