Samantha Lynn Semenkow¹, Shen Li², Eric Raabe^1,3, Jiadi Xu^2,4, Miroslaw Janowski^2,5, Byoung Chol Oh⁶, Gerald Brandacher⁶, Jeff W. Bulte^2,4, Charles Eberhart^1,3,7, and Piotr Walczak^2
1Department of Pathology, Johns Hopkins Medical Institue, Baltimore, MD, United States, ²Department of Radiology and Radiological Science, Johns Hopkins Medical Institue, Baltimore, MD, United States, ³Department of Oncology, Johns Hopkins Medical Institue, Baltimore, MD, United States, ⁴F. M. Kirby Center for Functional Brain Imaging Kennedy Krieger Institute, Johns Hopkins Medical Institue, Baltimore, MD, United States, ⁵NeuroRepair Department, Mossakowski Medical Research Centre, Warsaw, Poland, ⁶Department of Plastic and Reconstructive Surgery, Vascularized Composite Allotransplantation (VCA) Laboratory, Johns Hopkins Medical Institue, Baltimore, MD, United States, ⁷Department of Opthalmology, Johns Hopkins Medical Institue, Baltimore, MD, United States

Immunodeficient mice are currently used for modeling human brain tumor xenografts; however, immunodeficiency is a serious limitation precluding studies based on immunotherapy or inducing tumors in a variety of transgenic animal models. We therefore investigated whether disruption of co-stimulatory signaling using blocking antibodies induces tolerance to intracerebrally transplanted human glioblastoma xenografts in immunocompetent mice. With longitudinal MRI and bioluminescence we established that the growth rate of xenografts is comparable between immunodeficient and tolerance-induced immunocompetent mice. Quantitative MRI including T2/T1 relaxation time, MTR, diffusion parameters and perfusion were not significantly different, validating this new approach as a reliable brain tumor model. 

Marta Lai¹, Cristina Cudalbu², Marie-France Hamou^3,4, Mario Lepore², Lijing Xin², Roy Thomas Daniel⁴, Andreas Felix Hottinger⁵, Monika Hegi^3,4, and Rolf Gruetter^1,6,7
1Laboratory of Functional and Metabolic Imaging (LIFMET), Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, ²Animal Imaging and Technology Core (AIT), Center for Biomedical Imaging (CIBM), Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, ³Laboratory of Brain Tumor Biology and Genetics, Neuroscience Research Center, Lausanne University Hospital (CHUV), Lausanne, Switzerland, ⁴Service of Neurosurgery, Department of Clinical Neurosciences, Lausanne University Hospital (CHUV), Lausanne, Switzerland, ⁵Service of Neurology, Department of Clinical Neurosciences, Lausanne University Hospital (CHUV), Lausanne, Switzerland, ⁶Department of Radiology, University of Geneva, Geneva, Switzerland, ⁷Department of Radiology, University of Lausanne, Lausanne, Switzerland

In the present study orthotopic xenograft mice models of glioblastoma (GBM) derived from freshly dissected human cells of three different patients were compared at the aim of assessing patient-to-patient variability related to tumor metabolism and structural development. Mice were followed longitudinally in vivo in a 14.1 Tesla scanner with MRI and ¹H MRS which allowed to precisely quantify a wide range of GBM biomarkers. Finally spectra examined at late stage revealed peculiarity linked to each patient-derived xenograft, while longitudinal evolution of GBM biomarkers showed a close similarity in their expression within the same group and in animal lifespan. 

Hakmook Kang^1,2, Allison Hainline¹, Xia Li³, Lori R. Arlinghaus⁴, Vandana G. Abramson^5,6, A. Bapsi Chakravarthy^5,7, Brian Bingham⁸, and Thomas E. Yankeelov^2,4,5,9
1Biostatistics, Vanderbilt University, Nashville, TN, United States, ²Center for Quantitative Science, Vanderbilt University, Nashville, TN, United States, ³GE Global Research, Niskayuna, NY, United States,⁴Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States, ⁵Ingram Cancer Center, Vanderbilt University, Nashville, TN, United States, ⁶Medical Oncology, Vanderbilt University, Nashville, TN, United States, ⁷Radiation Oncology, Vanderbilt University, Nashville, TN, United States, ⁸School of Medicine, Vanderbilt University, Nashville, TN, United States, ⁹Radiology and Radiological Sciences, Vanderbilt University, Nashville, TN, United States

Pathologic complete response (pCR) following neoadjuvant chemotherapy is used as a short term surrogate marker of ultimate outcome in patients with breast cancer. Current imaging tools are suboptimal in predicting this response. Analyzing voxel-level heterogeneity in multi-modal MRI maps in conjunction with receptor status data, i.e., DCE- and DW-MRI, and ER/PR/HER2 status, allows us to improve the predictive power after the first cycle of neoadjuvant chemotherapy (NAC). 

Arash Latifoltojar¹, Margaret Hall-Craggs², Alan Bainbridge², Magdalena Sokolska², Kwee Yong¹, Neil Rabin², Liam Watson¹, Michelle Siu², Matthew Benger², Nikolaos Dikaios¹, and Shonit Punwani^1
1University College London, London, United Kingdom, ²University College London Hospital, London, United Kingdom

Whole body magnetic resonance imaging is becoming the gold standard imaging in initial assessment of multiple myeloma. Recently, functional imaging is being investigated in treatment response monitoring in multiple myeloma. We investigated different functional MRI biomarkers' temporal changes at early post-treatment stage in multiple myeloma patients following  Bortezomib induction.

Xiang Xu^1,2, Jiadi Xu^1,2, Linda Knutsson³, Yuguo Li^1,2, Huanling Liu^1,4, Guanshu Liu^1,2, Bachchu Lal^5,6, John Laterra^5,6, Dmitri Artemov^7,8, Michael T. McMahon^1,2, Peter C.M. van Zijl^1,2, and Kannie WY Chan^1,2
1Radiology, Johns Hopkins University School of Medicine, Baltimore, MD, United States, ²FM Kirby Research Center, Kennedy Krieger Institute, Baltimore, MD, United States, ³Department of Medical Radiation Physics, Lund University, Lund, Sweden, ⁴Department of Ultrasound, Guangzhou Panyu Central Hospital, Panyu, China, People's Republic of, ⁵Department of Neurology, Kennedy Krieger Institute, Baltimore, MD, United States, ⁶Department of Neuroscience, Kennedy Krieger Institute, Baltimore, MD, United States, ⁷Division of Cancer Imaging Research, Johns Hopkins University School of Medicine, Baltimore, MD, United States, ⁸JHU In Vivo Cellular Molecular Imaging Center, Baltimore, MD, United States

Recently D-glucose has shown potential to be used as a biodegradable contrast agent for cancer detection. However the origins of the glucoCEST signal is not yet completely understood. To identify the contributions to glucoCEST contrast, we administrated a glucose transporter inhibitor in a group of mice with implanted glioma. By inhibiting glucose transport into the cells, the effects of cellular glucose uptake and metabolism are suppressed and the perfusion properties of the extravascular extracellular space are delineated. A greater increase in glucoCEST contrast was seen in tumors in the group of mice with glucose transporter inhibitor compared to a group of mice without. This greater uptake and retention of glucose in the inhibitor group provides evidence that the intracellular glucose contribution is minimal. 

Kimberly L. Desmond^1,2, Hatef Mehrabian^1,2, Arjun Sahgal^1,3, Hany Soliman^1,3, and Greg J. Stanisz^1,2
1Physical Sciences, Sunnybrook Research Institute, Toronto, ON, Canada, ²Medical Biophysics, University of Toronto, Toronto, ON, Canada, ³Radiation Oncology, Odette Cancer Centre, Toronto, ON, Canada

Chemical exchange saturation transfer (CEST) spectra were collected at three timepoints following stereotactic radiosurgery (SRS). The magnetization transfer ratio (MTR) and CEST peak properties were evaluated at the offset frequencies of the NOE, amide and amine pools in the lesion and in the surrounding tissue. Positive correlation was found between changes in NOE peak amplitude and amide MTR at 1 week post-therapy and tumour volume change at one month post-therapy, while negative correlation was found between  amide peak width and NOE peak amplitude at the pre-treatment timepoint with volume change at one month post-therapy (p<0.1).  

Nathaniel Braman¹, Prateek Prasanna¹, Donna Plecha², Hannah Gilmore², Lyndsay Harris², Kristy Miskimen¹, Tao Wan³, Vinay Varadan¹, and Anant Madabhushi^1
1Case Western Reserve University, Cleveland, OH, United States, ²University Hospitals, Cleveland, OH, United States, ³Beihang University, Beijing, China, People's Republic of

In this work, we report preliminary success in the prediction of TP53 mutational status in breast cancer from DCE-MRI using a computer-extracted radiogenomic descriptor of multi-scale disorder, Co-occurrence of Local Anisotropic Gradient Orientations (CoLlAGe). A set of 8 distinguishing CoLlAGe features yielded accuracy of 78% in predicting TP53 mutational status and outperformed standard DCE-MRI pharmacokinetic parameters in an unsupervised hierarchical clustering. A non-invasive means of discerning TP53 mutational status may allow clinicians to more easily determine prognosis, assess treatment response, and inform treatment strategy. 

Yuan Le¹, Aneela Afzal², Xiao Chen³, Bruce Spottiswoode⁴, Wei Huang², and Chen Lin^1
1Radiology and Imaging Science, Indiana University School of Medicine, Indianapolis, IN, United States, ²Advanced Imaging Research Center, Oregon Health and Science University, Portland, OR, United States, ³Siemens Healthcare, Princeton, NJ, United States, ⁴Siemens Healthcare, Chicago, IL, United States

The goal of this work is to build a prototype quality assurance (QA) system for the quantitative pharmacokinetic (PK) analysis of breast DCE-MRI acquired with accelerated imaging techniques. A 3D digital tumor model with two sub-regions was constructed by segmenting patient images. The dynamic contrast enhanced images were synthesized according to the Tofts and Shutter Speed models with the TWIST technique. The QA system shows how the TWIST technique impacts the estimated pharmacokinetic parameters, and therefore allows necessary adjustments to be made to control the error.

Hassan Bagher-Ebadian^1,2, Azimeh NV Dehkordi³, Rasha Alamgharibi², Tavarekere Nagaraja¹, David Nathanson¹, Hamid Soltanian-Zadeh¹, Stephen Brown¹, Hamed Moradi⁴, Ali Arbab⁵, and James R Ewing^1,2
1Henry Ford Hospital, Detroit, MI, United States, ²Oakland University, Rochester, MI, United States, ³Shahid Beheshti University, Tehran, Iran, ⁴Tarbiat Modares University, Tehran, Iran, ⁵Georgia Regents University, Augusta, GA, United States

In this study, three physiologically nested models (NM) are derived from the standard Tofts model to describe possible physiological conditions of underlying tissue pathology. Then, using NM selection technique, Model Evolution (ME) concept is framed to quantify the evolutions of 3 different model volumes throughout the course of Dynamic Contrast Enhanced MRI experiment. We hypothesized that three evolutionary profiles in the course of DCE-MRI experiment generated from the ME concept, highly depend on the inward diffusion and outward convection of CA concentration and contain abundant information for describing the mechanical properties of solid tumors such as Interstitial Fluid Pressure (IFP).

SoHyun Han¹, Radka Stoyanova², Jason A. Koutcher³, HyungJoon Cho¹, and Ellen Ackerstaff^3
1Ulsan National Institute of Science and Technology, Ulsan, Korea, Republic of, ²Miller School of Medicine, University of Miami, Miami, FL, United States, ³Memorial Sloan Kettering Cancer Center, New York, NY, United States

Recently, a novel pattern recognition (PR) approach has been developed, identifying extent and spatial distribution of tumor microenvironments based on tumor vascularity. Here, our goal is to develop methods to minimize user intervention and errors from model-based approaches by introducing an automated algorithm for determining the number of classifiers. An SNR approach showed the highest accuracy at ~97% along five different tumor cell models with 104 slices total. The visualization of tumor heterogeneity (perfusion, hypoxia, necrosis) with automated analysis of DCE-MRI can reduce the need for manual expert intervention, extensive pharmacokinetic modeling, and could provide critical information for treatment planning.

Harbir Singh Sidhu¹, Anna Barnes², Nikolaos Dikaios¹, Scott Rice¹, Alan Bainbridge³, Robert Stein⁴, Sandra Strauss⁵, David Atkinson¹, Stuart Taylor¹, and Shonit Punwani^1
1Centre for Medical Imaging, University College London, London, United Kingdom, ²Institute of Nuclear Medicine, University College London Hospital, London, United Kingdom, ³Medical Physics and Biomedical Engineering, University College London Hospital, London, United Kingdom, ⁴Medical Oncology, University College London Hospital, London, United Kingdom, ⁵Research Department of Oncology, University College London, London, United Kingdom

Tumor response assessment currently relies upon measurement of size change, which may not alter significantly early during treatment or at all with newer therapies. Patients may therefore incur significant side-effects (with associated healthcare cost) without benefit.   Assessment of soft tissue tumor T1 relaxation times before and early during treatment can predict lesion response whilst being incorporated within a clinically feasible whole-body MRI scan duration. Tumors undergoing partial response at the end of treatment demonstrated significant reduction in T1 values early during therapy compared to non-responding lesions.

In the future, this could facilitate early response assessment and complement other imaging biomarkers.

Eoin Finnerty¹, Rajiv Ramasawmy², James O'Callaghan², Mark F Lythgoe², Karin Shmueli¹, David L Thomas³, and Simon Walker-Samuel^2
1Medical Physics and Biomedical Engineering, University College London, London, United Kingdom, ²University College London, London, United Kingdom, ³Institute of Neurology, University College London, London, United Kingdom

This work examines the application of Quantitative Susceptibility Mapping (QSM) in a mouse model of colorectal liver metastases. It was hypothesised that QSM could provide a novel method of interrogation of liver tumours based on differences in blood oxygenation. Results under hyperoxic and normoxic conditions were compared to assess the response of the liver tissue and tumours. A vascular disrupting agent was then administered to assess its effect on the QSM measurements. A significant difference was found between liver and tumour tissue, and regional differences in susceptibility were found within a tumour. These differences were less apparent after VDA administration.

Zhao Li¹, Chaohsiung Hsu¹, Ryan Quiroz¹, and Yung-Ya Lin^1
1Department of Chemistry and Biochemistry, UCLA, Los Angeles, CA, United States

Early detection of high-grade malignancy, such as glioblastoma multiforme (GBM), using enhanced MRI techniques significantly increases not only the treatment options available, but also the patients’ survival rate. For this purpose, a conceptually new approach, termed “Active-Feedback MRI”, was developed. An active feedback electronic device was homebuilt to implement active-feedback pulse sequences to generate avalanching spin amplification and fixed-point spin dynamics, which enhances the local magnetic-field gradient variations due to irregular water contents and deoxyhemoglobin concentration in early GBM. Statistical results (N=22) for in vivo orthotopic xenografts GBM mouse models at various cancer stages validate the superior contrast and robustness of this approach (tumor time constant differs from that of the healthy brain tissue by +24%) towards early GBM detection than conventional T1-weighted (+2.6%) and T2-weighted images (-3.1%). This novel approach provides 4-8 times of improvements in early GBM tumor contrast, as measured by "tumor to normal tissue contrast", “contrast-to-noise ratio” (CNR) or “Visibility”.

Jiaen Liu¹, Qi Shao¹, Yicun Wang¹, Gregor Adriany², John Bischof³, Pierre-Francois Van de Moortele², and Bin He^1,4
1Biomedical Engineering, Univeristy of Minnesota, Minneapolis, MN, United States, ²Center for Magnetic Resonance Research, Univeristy of Minnesota, Minneapolis, MN, United States, ³Mechanical Engineering, Univeristy of Minnesota, Minneapolis, MN, United States, ⁴Institute for Engineering in Medicine, Univeristy of Minnesota, Minneapolis, MN, United States

Noninvasive in vivo imaging of the tissue conductivity has great potential in cancer diagnosis. Recently, electrical properties tomography (EPT) has been investigated with increasing effort to noninvasively image tissue conductivity in vivo using MRI. A preclinical method for imaging tumor conductivity can be valuable for understanding tumor development and associated conductivity change due to fundamental molecular and cellular reasons. In this study, tumor conductivity was studied based on a xenograft rat tumor model using a small animal EPT system. The result showed elevated conductivity in cancerous tissue compared to healthy tissue, suggesting the clinical value of EPT for tumor diagnosis.

Gabriel Nketiah¹, Mattijs Elschot¹, Eugene Kim ¹, Tone Frost Bathen ¹, and Kirsten Margrete Selnæs^1
1Department of Circulation and Medical Imaging, Norwegian University of Science and Technology, Trondheim, Norway

The complexity of the prostatic tissue requires sensitive, accurate and reproducible assessment methods for aggressiveness of prostatic carcinomas, especially in differentiating between Gleason score 3+4 and 4+3 tumors. We evaluated the applicability of T2W MRI-derived textural entropy as a potential marker for assessing prostate cancer aggressiveness. Our study found textural entropy to correlate  moderately positive and negative with Gleason score and apparent diffusion coefficient (ADC), respectively. T2W image textural entropy differentiated Gleason score 3+4 and 4+3 tumors with higher accuracy than other MRI-derived parameters (ADC, K^trans and V_e), indicating the potential of MRI texture analysis in prostate cancer assessment.