Amy Melsaether¹, Akshat C. Pujara², Eric Sigmund², Alana Amarosa², Freya Schnabel², Deirdre Kiely², Sungheon Kim², and Linda Moy^2
1NYU, New York, New York, United States, ²NYU, New York, United States

In this preliminary study, we address whether the PET and DWI data (SUV max and ADC min) acquired during a breast PET/MRI correlate with each other and with breast cancer tumor markers and with systemic disease.

Martin D Pickles¹, Daniel Litwiller², and Lindsay W Turnbull^1
1Centre for Magnetic Resonance Investigations, HYMS at University of Hull, Hull, East Yorkshire, United Kingdom, ²Global MR Applications and Workflow, GE Healthcare, Rochester, MN, United States

Further improvements in the specificity of MR breast are required. Recently, a 3D FSE T2W modified 3-point DIXON technique (CUBE-IDEAL) became available. The purpose of this work is to evaluate CUBE-IDEAL and compare against traditional 2D T2W FSE fat saturated images. Sixty-nine patients were imaged on a 3.0T scanner. The results of this assessment suggest a similar level of performance overall. However, a number of key points should be underscored. Firstly, CUBE-IDEAL did outperform FSE at fat nulling. Secondly, CUBE-IDEAL can be reformatted into any plane. Thirdly, CUBE-IDEAL presents a time saving over the traditional FSE sequence.

Soledad Milans¹, Sujata Patil², Elizabeth Morris², and Sunitha Thakur^1
1Memorial Sloan-Kettering Cancer Center, New York, NY, United States, ²Memorial Sloan-Kettering Cancer Center, NY, United States

The amount of fibroglandular tissue (FGT) density and the level of background parenchymal enhancement (BPE) are MRI features of normal breast. Studies showed increased BPE to be strongly predictive of breast cancer odds as with increased FGT. We propose to evaluate the effect of FGT density and BPE on ADC values in patients with malignant and benign findings at 3.0T. FGT-ADC values were significantly higher in patients with malignant lesions with higher density compared to lower density (p=0.0073). In patients with benign lesions, difference is smaller between both groups (p=0.015). No significant difference was observed in FGT-ADC values with BPE.

Naranamangalam R Jagannathan¹, Khushbu Agarwal¹, Rani G Sah¹, Uma Sharma¹, Rajinder Parshad², and Vurthaluru Seenu^2
1Department of NMR & MRI Facility, All India Institute of Medical Sciences, New Delhi, Delhi, India, ²Department of Surgical Disciplines, All India Institute of Medical Sciences, New Delhi, Delhi, India

Diffusion weighted imaging and in-vivo proton MRS in lactating women volunteers (n=16) and in patients with breast cancer (n=13) showed total choline peak in 16/16 malignant cases and in 11/13 lactating women volunteers. Further, 11/13 of lactating women showed a lactose peak at 3.8 ppm. tCho concentration was similar in both groups while higher ADC was observed in lactating women. Our study demonstrated that higher ADC value together with the presence of a lactose peak may aid in the differentiation of changes that occur in the breast tissues due to normal physiological conditions as compared with malignant transformations.

Martin D Pickles¹, Lindsay W Turnbull¹, Peter Gibbs¹, and Martine Dujardin^1
1Centre for Magnetic Resonance Investigations, HYMS at University of Hull, Hull, East Yorkshire, United Kingdom

We believe that a short breast MR examination contains the necessary information to allow an accurate diagnosis. The purpose of this study is to determine the technical feasibility of a short examination and to obtain preliminary comparative data against the standard examination. Nineteen participants were imaged on a 3.0T scanner twice (standard and short). Results suggest that the short breast MR examination is not only feasible but has good agreement with the standard examination. This study has demonstrated that high spatial and temporal resolution data can be acquire in only 12minutes with similar results to much longer MR examinations.

Soo-Yeon Kim¹, Jaewook Shin², Dong-Hyun Kim², Min-Jung Kim¹, and Eun-Kyung Kim^1
1Radiology, Severance Hospital, Yonsei University College of Medicine, Seoul, Seodaemun-gu, Korea, ²Electronic and Electronic Engineering, Yonsei University, Seoul, Seodaemun-gu, Korea

The electric conductivity has shown its potential to differentiate benign and malignant breast tissue. We investigate the correlation between the electric conductivity with known prognostic factors of invasive breast cancer using the electric properties tomography reconstruction algorithm and multi-receive coil combined technique. Breast cancers with the established poor prognostic factors (axillary lymph node metastasis, lymphovascular invasion and high histologic grade) showed higher conductivity value than those without. Our results show the possibility that the electric conductivity can be differentiated according to the known prognostic markers in invasive breast cancers.

Jose Ramon Teruel^1,2, Agnes Østlie³, Hans Erikssønn Fjøsne^4,5, Tone Frost Bathen¹, and Pål Erik Goa^2,6
1Department of Circulation and Medical Imaging, NTNU, Trondheim, Norway, ²St. Olavs University Hospital, Trondheim, Norway, ³Department of Radiology, St. Olavs University Hospital, Trondheim, Norway, ⁴Department of Surgery, St. Olavs University Hospital, Trondheim, Norway, ⁵Institute of Cancer Research and Molecular Medicine, NTNU, Trondheim, Norway, ⁶Department of Physics, NTNU, Trondheim, Norway

In this study, a simplified approach to obtain perfusion influence from diffusion weighted imaging (DWI) measurements for breast cancer assessment is proposed. Our results present how the relative perfusion effect derived from different monoexponential fitting schemes of DWI measurements can be obtained. Furthermore, this simplified approach offers a potential for breast cancer differentiation by means of the derived biomarker.

Bo La Yun¹, Sun Mi Kim¹, Mijung Jang¹, Hye Shin Ahn¹, Kyung Eun Cho¹, Bohyoung Kim¹, Hochul Kang², and Ji Young Kim^1
1radiology, Seoul national Univ. Bundang Hosp., Seoungnam-si, Kyongki-do, Korea, ²Seoul national Univ, Seoul, Korea

Triple negative breast cancers are heterogeneous disease and poor prognosis. We were investigated early and delayed enhancement pattern and texture feature on breast dynamic MR image. In this study, We found that early and delayed enhancement pattern (rapid and plateu) and texture features (high entropy and low homogniety) in breast dynamic MR were associated with traditional pathologic poor prognosis factors in triple negative breast tumor. These image features could predict preoperative breast cancer aggressiveness.

Palamadai N Venkatasubramanian¹, George Iordanescu¹, Matthew M Smith¹, Jennifer L Gnerlich², Katherine Yao³, and Alice M Wyrwicz^1
1Radiology, NorthShore University HealthSystem, Evanston, IL, United States, ²Surgery, The University of Chicago Medical Center, Chicago, IL, United States, ³Surgery, NorthShore University HealthSystem, Evanston, IL, United States

Using ex vivo MR spectroscopy we found alterations in the fatty acid composition of peritumoral adipose tissue from breast cancer patients, relative to adipose tissue from a distal location within the same breast. MR-measured compositions of breast adipose tissues could predict known pathological criteria for invasive breast cancers. Our results suggest that in vivo MRS may have potential for developing a noninvasive biomarker for invasive breast cancers.

Saadallah Ramadan¹, Jameen Arm², Gorane Santamaria³, Judith Silcock⁴, Jessica Buck⁵, Michelle Roy⁶, Kin Men Leong⁷, Peter Lau⁷, David Clark⁴, Peter Malycha⁶, and Carolyn Mountford^6
1School of Health Sciences, University of Newcastle, Callaghan, NSW, Australia, ²Hunter New England Health, The Mater Hospital, Callaghan, NSW, Australia, ³Department of Radiology, Hospital Clinic de Barcelona, Villarroel, Spain, ⁴The Breast & Endocrine Centre, NSW, Australia, ⁵School of Health Sciences, University of Newcastle, NSW, Australia,⁶School of Health Sciences, Centre for MR in Health, NSW, Australia, ⁷Calvary Mater Hospital, NSW, Australia

BRCA1 and BRCA2 genes belong to the tumor suppressor family and patients with these genes are at increased risk of developping breast cancer. In this study, we apply in vivo two-dimensional 2D localized correlation spectroscopy (L-COSY) to look for a premalignant state in the breast tissues of apparently healthy women carrying the BRCA gene mutations and others with a family history. We propose the hypothesis that those with the BRCA gene mutations would have altered chemistry reflective of a preinvasive state.

Shelley Waugh¹, Lukasz Priba¹, and Sarah Vinnicombe^2
1Medical Physics, Ninewells Hospital, Dundee, Angus, United Kingdom, ²Division of Cancer Research, University of Dundee, Dundee, United Kingdom

This study considers the use of apparent diffusion coefficient (ADC) measures as an indicator of early response to neoadjuvant chemotherapy. As ADC changes are utilised in clinical practice, this investigation considers the multiple factors that may influence these measurements and could influence classification of response according to RECIST criteria. Scanner stability, scan-scan repeatability and intra-observer repeatability are measured with reference to the order of magnitude of ADC changes likely to be encountered in clinical patients who respond, partially respond and have stable disease.

Judith Wild¹, Anna-Lisa Kofahl¹, Deniz Ulucay¹, Sebastian Theilenberg¹, Carsten Urbach¹, Jürgen Finsterbusch², Kerstin Rhiem³, Peter Trautner⁴, and Karl Maier^1
1University of Bonn, Bonn, Germany, ²University Medical Center Hamburg, Hamburg, Germany, ³University Medical Center Cologne, Cologne, Germany, ⁴Life & Brain GmbH, Bonn, Germany

The early detection of breast cancer has severely improved during the past 20 years, but there is still room for improvement like a better clarification of indications without using ionizing radiation. The evaluation of biopsy data showed that about 80% of women diagnosed with cancer after the different types of imaging do not have cancer. We target on improving this high percentage by improving the specificity by measuring the elasticity of lesions. Measurements on 10 volunteers with well known lesions provide an important step towards seeing if ARC-MR is capable to improve the specificity of standard breast cancer diagnostic.

Kristin L Granlund^1,2, Debra Ikeda¹, Jafi Lipson¹, Jennifer Kao¹, Jung Min Chang³, Brian Hargreaves¹, and Bruce Daniel^1
1Radiology, Stanford University, Stanford, CA, United States, ²Electrical Engineering, Stanford University, Stanford, CA, United States, ³Radiology, Seoul National University Hospital, Seoulq, Korea

The DESS sequence has been used to acquire diffusion-weighted images, and this study evaluates DESS images relative to EPI DWI images. Four radiologists were surveyed about the image quality and BI-RADS scores of images containing pathology-proven breast lesions. The DESS sequence acquired sharper diffusion-weighted breast images than EPI (p<0.001). ROC curves were calculated from the BI-RADS scores, and the DESS sequence had a larger AUC (0.74 for DESS, 0.68 for EPI) and a tighter confidence interval. Better image quality facilitates morphological assessment of lesion malignancy.

Jie He¹, Yan Zhang¹, Jingliang Cheng¹, Ying Hu¹, Anfei Wang¹, Dandan Zheng², and Xiaoyan Wang^1
1Department of MRI, the First Affiliated Hospital of Zhengzhou University, Zhengzhou, HeNan, China, ²GE healthcare, Beijing, China

MRI has been widely used in the diagnosis of breast disease. However, the calculation of apparent diffusion coefficient by simple monoexponential relationship between MRI signal and b value does not fully account for tissue behavior. Intravoxel incoherent motion (IVIM) method allows quantitative parameters that reflect tissue micro capillary perfusion and tissue diffusivity. In this study, the stretched-exponential model was used to generate ¦Á and distributed diffusion coefficient (DDC) maps. The ¦Á values and DDCs were compared with traditional monoexponential ADC in the differential diagnosis of breast benign and malignant lesions.

Jeon-Hor Chen^1,2, Yifan Li¹, Hon Yu¹, Shadfar Bahri¹, Rita S Mehta³, and Min-Ying Su^1
1Center for Functional Onco-Imaging, University of California, Irvine, California, United States, ²Department of Radiology, Eda Hospital and I-Shou University, Kaohsiung, Taiwan, ³Department of Medicine, University of California, Irvine, California, United States

In this study we aimed to correlate percent breast density (PD) and background parenchymal enhancement (BPE) in the contralateral normal breast (CNB) of patients diagnosed with breast cancer. Breast MR images of the CNB of 117 patients were studied. The quantification of the fibroglandular tissue volume (FV) was based on a computer-assisted algorithm. The mean BPE and the hot spot enhancement were correlated with age, and also with FV and PD. Our study showed that BPE, measured by the averaged enhancement of the whole FV or by hot spot, did not correlate well with quantitative measurement of PD and FV.

Jeon-Hor Chen^1,2, Siwa Chan³, Yi-Ting Tang⁴, Angela T. Cheriyan¹, Nikita Rakesh Shah¹, and Min-Ying Su^1
1Center for Functional Onco-Imaging, University of California, Irvine, California, United States, ²Department of Radiology, Eda Hospital and I-Shou University, Kaohsiung, Taiwan, ³Department of Radiology, Taichung Veterans General Hospital, Taichung, Taiwan, ⁴Department of Medical Imaging, China Medical University, Taichung, Taiwan

We conducted a study to investigate the consistency of percent breast density measured in MRI in four different positions and imaging resolution. The breast and fibroglandular tissue segmentation was based on an established novel method. Results of FV and PD measurement show small variations, with an averaged CV of <9%, among the four MR studies. Remarkable measurement variation did occur in some subjects in hands-down position, compared to hands-up position. The results from our MR consistency study indicate that MR imaging data from multi-centers, regardless of patients’ positions (hand-up or hands-down), can be combined for analysis.

Camila Munoz¹, Dravna Razmilic², Maria E. Navarro², Paula Espinoza², Cristian Tejos¹, Pablo Irarrazaval¹, Sergio Uribe¹, and Marcelo E. Andia^1,2
1Biomedical Imaging Center, Pontificia Universidad Catolica de Chile, Santiago, Region Metropolitana, Chile, ²Radiology Department, School of Medicine, Pontificia Universidad Catolica de Chile, Santiago, Region Metropolitana, Chile

Breast density is a major risk factor of breast cancer, and is usually assessed visually from mammograms. This has two main disadvantages: there is not a standard measure of breast density, as the breast is classsified in qualitative categories, and this assessment does not consider that mammography is a 2D projection of the breast volume. We developed a method to quantify volumetric breast density from mammograms and a semi-automatic method that quantifies this measure in 3D breast MRI. Thus, it is possible to obtain an equivalent quantitative measure of breast density between both imaging techniques, solving the issues mentioned above.

Araminta E. W. Ledger¹, Marco Borri¹, Hector Sanchez Casas¹, Craig Cummings¹, Maria A. Schmidt¹, and Martin O. Leach^1
1CR-UK and EPSRC Cancer Imaging Centre, The Institute of Cancer Research and Royal Marsden NHS Foundation Trust, Sutton, Surrey, United Kingdom

DCE-MRI is an established component of breast screening protocols, and uses the empirical classification of contrast agent (CA) enhancement curves to aid diagnosis. However, CA enhancement curves resulting from varying parameter selections are difficult to assess prospectively. This abstract uses a novel and affordable DCE-MRI phantom to assess the effect of common sequence alterations (CA injection timing and k-space sampling scheme) on the CA enhancement curves obtained from two clinical DCE-MRI sequences. Both CA injection timing and k-space sampling pattern resulted in measurable curve differences - sequence comparison with this phantom can therefore help to establish robust DCE-MRI breast protocols.

Kathryn E Keenan¹, Sheye O Aliu², Lisa J Wilmes², David C Newitt², Elizabeth Horneber³, Karl F Stupic¹, Michael A Boss¹, Michael G Snow⁴, William Hollander⁴, Stephen E Russek¹, and Nola M Hylton^2
1National Institute of Standards and Technology, Boulder, CO, United States, ²University of California San Francisco, San Francisco, CA, United States, ³University of Colorado, Boulder, CO, United States, ⁴High Precision Devices, Boulder, CO, United States

A novel breast phantom was created for the breast imaging community for implementation of quality control measures. The phantom contains well-distributed fat and fibroglandular T1 relaxation mimics and diffusion mimics. In addition, the phantom has a flexible outer shell so that it is compatible with several coil designs. In initial testing, the phantom fit in several coil designs, enabled fat-suppression tests, and allowed assessment of diffusion artifacts. Please visit our website for additional information: http://collaborate.nist.gov/mriphantoms/bin/view/MriPhantoms/BreastPhantom.

Yuan Le¹, Marcel Dominik Nickel², Randall Kroeker², Christian Geppert², Brian Dale², Hal D. Kipfer³, and Chen Lin^1
1Radiology and Imaging Science, Indiana University School of Medicine, Indianapolis, Indiana, United States, ²Siemens Healthcare, North Carolina, United States, ³Radiology and Imaging Science, Indiana University School of Medicine, Indiana, Indiana, United States

The acquisition of breast DCE-MRI using a flexible TWIST view sharing technique was simulated. A digital ‘phantom’ was generated with three 5mm uniform spherical lesions of ‘persistent’, ‘plateau’ and ‘wash-out’ type of contrast uptake, and one 10 mm complex tumor with a mixture of all three types of enhancements. Our results show that with the typical spatial resolution in clinical breast DCE-MRI, TWIST view sharing parameters of pA=20% and pB=20% provides the lowest overall RMS error for all time points, while pA=50% and pB=50% produces image with minimum error at the peak contrast uptake.

Wen Hao¹, Bin Zhao¹, Guangbin Wang¹, Hui Liu², and Cuiyan Wang^1
1Magnetic resonance imaging, Shandong medical imaging research institution, Shandong University, Jinan, ShanDong, China, ²MR Collaboration NEA, Siemens Healthcare, Shanghai, China

This abstract is about assessment performed to investigate the clinical feasibility of CDT-VIBE for quantitative breast DCE-MRI. In conclusion, we believe that CDT-VIBE can be used in breast DCE-MRI for detecting and depicting lesions because of its high spatial resolution and nearly equal image quality to conventional VIBE image. Therefore, it is clinically feasible to replace standard 60-90 second conventional GRE sequence by CDT-VIBE sequence for quantitative breast DCE-MRI with the acquisition schemes employed in this study.

Majid Mahrooghy¹, Ahmed B. Ashraf¹, Dania Daye¹, Carolyn Mies², Mark Rosen¹, Michael Feldman², and Despina Kontos^1
1Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States, ²Department of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, PA, United States

We developed DCE-MRI kinetic heterogeneity features using self-organizing map (SOM) neural networks. We use SOM to cluster tumor pixels based on kinetics and extract features including variance and entropy of cluster size, variance of cluster kinetic features, mean and variance of weighted cluster kinetics, and the kinetic features of the cluster having maximum peak enhancement. We evaluated these features for classifying tumor recurrence risk as determined by a validated gene expression assay, and compared their performance to current standard kinetics. Our features have ROC AUC=0.80 for classifying tumors at low- versus high- risk of recurrence, outperforming standard kinetics with AUC=0.65.

Venkata Veerendranadh Chebrolu¹, Dattesh D Shanbhag¹, Aurelie Le Deley², Sheshadri Thiruvenkadam¹, Uday Patil^1,3, Patrice Hervo², Sandeep N Gupta⁴, and Rakesh Mullick^5
1Medical Image Analysis Lab, GE Global Research, Bangalore, Karnataka, India, ²GE Healthcare, Buc, France, ³Manipal Health Enterprises Pvt. Ltd., Bangalore, Karnataka, India,⁴Biomedical Image Processing Lab, GE Global Research, Niskayuna, NY, United States, ⁵Diagnostics and Biomedical Technologies, GE Global Research, Bangalore, Karnataka, India

DCE-MRI is increasingly being used in diagnosis and screening of breast tumors. Careful study of intensity changes over time between the pre-contrast and post-contrast images is critical to tumor biometry. Rigid and non-rigid motion may be caused by factors such as voluntary patient motion, cardiac pulsation, and breathing during image acquisition. In this work we compare and evaluate b-splines and symmetric diffeomorphic normalization based non-rigid registration (NRR) algorithms for their effectiveness in motion correction for breast DCE-MRI. B-splines based NRR approaches provided consistent time performance. Better NRR accuracy was achieved with diffeomorphic normalization based motion correction methods.

Sinyeob Ahn¹, Himanshu Bhat¹, Dorota Wisner², Kawin Setsompop³, Thomas Benner⁴, Stephen Cauley³, Borjan Gagoski⁵, Bonnie Joe², Gerhard Laub¹, and Vibhas Deshpande^1
1Siemens Healthcare, San Francisco, CA, United States, ²Radiology and Biomedical Engineering, UCSF, CA, United States, ³Radiology, MGH, MA, United States, ⁴Siemens Healthcare, Erlangen, Germany, ⁵Children's Hospital, Boston, MA, United States

Brest diffusion MRI has been typically performed in an axial slice orientation with large FOV which tends to show ghosting artifacts from significant B0 field inhomogeneity. Sagittal slice orientation is limited due to long scan time. In this paper, we propose a multiple slice group slice-accelerated sequence for sagittal bilateral breast diffusion imaging. One slice on each slice group out of two was acquired and compared to the standard technique. The proposed method provided nearly identical image quality with almost half the scan time as compared to the standard bilateral diffusion imaging, suggesting a viable tool for breast lesion characterization within reasonable scan time.

Shelley Waugh¹, Lukasz Priba¹, and Sarah Vinnicombe^2
1Medical Physics, Ninewells Hospital, Dundee, Angus, United Kingdom, ²Division of Cancer Research, University of Dundee, Dundee, United Kingdom

The aim of this study was to identify the contribution of various factors that may have an impact on accuracy of measurements of apparent diffusion coefficients (ADC) in breast MRI. We have considered the effect of long-term scanner stability, positional dependence within the coil, scan-scan repeatability and the effect of single and multiple observers on final measurements of whole tumour and lowest ADC values within tumours. We conclude that scan-scan repeatability and scanner stability have minimal effect on measured ADC values and the biggest influence on measured ADC values is inter-observer repeatability.

Curtis A. Corum¹, Djaudat Idiyatullin¹, Diane Hutter¹, Lenore I. Everson¹, Lynn E. Eberly², Michael T. Nelson³, and Michael Garwood^1
1CMRR, Radiology, Medical School, University of Minnesota, Minneapolis, Minnesota, United States, ²Biostatistics, School of Public Health, University of Minnesota, Minneapolis, Minnesota, United States, ³Breast Center, Radiology, Medical School, University of Minnesota, Minneapolis, Minnesota, United States

This work describes a diffusion-weighted magnetization prepared SWIFT (SWeep Imaging with Fourier Transformation) sequence utilizing adiabatic RF pulses for diffusion weighting with additional fat suppression. The SWIFT sequence is used as excitation and high resolution readout of the prepared diffusion-weighted and fat-suppressed magnetization. The proposed method, DW-MP-SWIFT, is more robust to motion and eddy currents compared to typical diffusion-weighted sequences, while providing high spatial resolution. We report results of DW-MP-SWIFT for phantom and breast imaging of normal human subjects.

Lisa Wilmes¹, Wei Ching Lo², David C Newitt², Suchandrima Banerjee³, Evelyn Proctor², Emine Saritas⁴, Ajit Shankaranarayanan³, and Nola M Hylton^2
1University of California San Francisco, San Francisco, CA, United States, ²University of California San Francisco, CA, United States, ³Applied Science Laboratory GE Healthcare, CA, United States, ⁴University of California Berkeley, CA, United States

A high-resolution reduced field of view diffusion tensor weighted imaging sequence (HR-DTI), voxel size 4.8 mm3, was optimized for breast imaging and compared to a standard FOV DTI sequence (STD-DTI), voxel size 29.3 mm3, for evaluating tumor response to treatment in seven breast cancer patients undergoing neoadjuvant chemotherapy. Significant differences were found between the tumor fractional anisotropy (FA) measured by HR-DTI and the STD-DTI prior to and early in treatment. Of the DTI parameters evaluated, HR-DTI FA correlated most strongly with tumor volume change post treatment. This preliminary study suggest that HR-DTI may be sensitive to treatment-induced changes in tumors.

Shandong Wu¹, Wendie A. Berg², Margarita L. Zuley², Jules Sumkin², Rachel C. Jankowitz³, and David Gur^4
1Radiology, University of Pittsburgh, Pittsburgh, PA, United States, ²Radiology, University of Pittsburgh Medical Center Magee-Womens Hospital, PA, United States, ³Oncology, University of Pittsburgh Medical Center Magee-Womens Hospital, PA, United States, ⁴Radiology, University of Pittsburgh, PA, United States

The purpose of this study is to perform quantitative assessment on the longitudinal variation of fibroglandular tissue (FGT) and background parenchymal enhancement (BPE) in sequential breast MRI scans for a cohort of high-risk women of developing breast cancer who did not undergo any specific risk-reduction intervention. We retrospectively analyzed 71 high-risk women each with two longitudinal cancer-free MRI scans using a fully automated computerized method. The preliminary results demonstrate the temporal variability of FGT and BPE, which may be useful as a reference measure when investigating these parameters as risk predictors and possibly as indicators of intervention-response in high-risk women.

Katsuhiro Kida¹, Sachiko Goto², Tsutomu Kajitani¹, and Yoshiharu Azuma^2
1Department of Radiology, Japanese Red Cross Okayama Hospital, Okayama-shi, Okayama, Japan, ²Faculty of Health Sciences, Graduate School of Health Sciences, Okayama University, Okayama-shi, Okayama, Japan

In this study, we discriminated the calcium oxalate and calcium phosphate in breast disease. A cup shaped gel phantom containing two types kidney stones was employed to simulate calcification in the breast. The phase values of five phase images with different echo times were measured, those were obtained by a 3D m-FFE sequence to shorten examination time. We conclude that the calcium oxalate and phosphate are distinguishable with the slope of linear approximation of calcifications on high-pass filtered phase images. We believe that this method is useful for the discrimination between benign and malignant breast disease.

Paula Donahue¹, Swati Rane², Seth Smith², and Manus Donahue^2
1Physical Medicine and Rehabilitation, Dayani Center for Health and Wellness, Vanderbilt University School of Medicine, Nashville, TN, United States, ²Radiology, Vanderbilt University School of Medicine, TN, United States

The overall objective of this work is to translate noninvasive imaging techniques for measuring brain structure and function to the lymphatic system to characterize axillary lymphatic vessel structure and interstitial protein accumulation in patients with breast cancer-related lymphedema (BCRL). To achieve this, a multi-faceted, noninvasive 3T MRI protocol for characterizing lymph node volume (DWIBS MRI), lymph vessel diameter (3D TSE), lymph flow velocity (spin labeling MRI), and interstitial protein accumulation (APT CEST MRI) are optimized and applied in healthy volunteers (n=10) and patients with BCRL (n=4).

Romy Langhammer^1,2, Leo L. Cheng¹, Johannes Nowak³, Shulin Wu¹, W. Scott McDougal¹, Chin-Lee Wu¹, and Eva M. Ratai^1
1Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, United States, ²Institute of Radiolgy, Department for Neuroradiology, University Hospital Wuerzburg, Wuerzburg, Bavaria, Germany, ³Institute of Radiology, Department for Neuroradiology, University Hospital Wuerzburg, Wuerzburg, Bavaria, Germany

The project aims to develop a non-invasive diagnostic test for prostate cancer. Metabolomic imaging maps of thirty whole prostates were created and analyzed, using magnetic resonance spectroscopy (MRS) at 7T. To test the functionality and reliability of this system, the outcome has been compared with the histopathology findings. In 61% of the samples, MRS detected cancer lesions in the same locations as were identified in the histopathology analysis; the histological regions that were not detected had significantly lower tumor volumes. Furthermore, a "Malignancy Index" can significantly differentiate between cancerous and benign MRS suspicious regions.

Isabell K. Steinseifer¹, Bart Philips¹, Borjan Gagoski², Marnix C. Maas¹, Elisabeth Weiland³, Tom W.J. Scheenen¹, and Arend Heerschap^1
1Radiology, Radboud University Medical Center, Nijmegen, Netherlands, ²Fetal-Neonatal Neuroimaging & Developmental Science Center, Boston Children's Hospital, Harvard Medical School, Boston, MA, United States, ³MR Applications Development, Siemens AG, Healthcare Sector, Erlangen, Germany

Fast ¹H MRSI of the prostate was achieved with a semi-LASER sequence with GOIA-WURST(16,4) refocusing pulses and spiral k-space sampling. The application of these low RF power demanding adiabatic pulses reduces the specific absorption rate so that an endorectal coil can be used to receive signal. Together with a relative short echo time of 90ms, a high SNR can be obtained. This can be translated in a much lower acquisition time, which can be realized because of the flexibility of spiral k-space sampling. The new sequence facilitates routine acquisition of metabolic data for clinical purposes.

Maya B Wolf¹, Timur H Kuru², Gregor Habl³, and Matthias C Roethke^1
1Radiology, DKFZ, Heidelberg, Baden-Württemberg, Germany, ²Urology, University Hospital Heidelberg, Heidelberg, Baden-Württemberg, Germany, ³Radiation Oncology, University Hospital Heidelberg, Heidelberg, Baden-Württemberg, Germany

Purpose was to assess T2- and diffusion-weighted imaging (DWI) after carbon ion therapy (CIT) of prostate cancer (PCa) for treatment monitoring. Multiparametric MRI including T2 and DWI at 3T were acquired prior to and following CIT in patients with histologically confirmed PCa. Statistical analysis of region of interest data showed T2w signal loss at the 6 months interval and immediate, as well as, continuous apparent diffusion coefficient (ADC) increase after treatment in PCa areas. PSA was inversely correlated with ADC values. The results suggest that ADC could serve as an imaging biomarker for assessing therapeutic response of PCa to CIT.

Gary Liney^1,2, Thahabah Mohammed Al Harthi³, Ewa Juresic⁴, Lynette Cassapi⁴, Lois Holloway⁴, Mark Sidhom⁴, David Manton⁵, and Peter Gibbs^6
1Medical Physics, Ingham Institute for Applied Medical Research, Sydney, NSW, Australia, ²Physics, University of Wollongong, Sydney, NSW, Australia, ³Physics, University of Sydney, NSW, Australia, ⁴Cancer Therapy Centre, Liverpool Hospital, NSW, Australia, ⁵Radiation Physics, East Riding, United Kingdom, ⁶University of Hull, East Riding, United Kingdom

The measurement of apparent diffusion coefficient (ADC) from diffusion weighted imaging (DWI) has been linked to tumour response following radiotherapy in prostate cancer. For it to be used to monitor and adapt plans during a course of treatment DWI needs to be produce reliable measurements and distortion free images. This work compares three DWI techniques in phantoms and prostate volunteers in order to assess their suitability for use as a robust clinical tool.

Ivan Jambor¹, Esa Kähkönen², Pekka Taimen³, Harri Merisaari⁴, Jani Saunavaara⁵, Kalle Alanen⁶, Branislav Obsitnik⁷, Heikki Minn⁴, Viera Lehotska⁸, and Hannu Aronen^1
1Departement of Diagnostic Radiology, University of Turku, Turku, Finland, ²Department of Surgery, Turku University Hospital, Turku, Finland, ³Department of Pathology, University of Turku, Turku, Finland, ⁴Turku PET centre, University of Turku, Turku, Finland, ⁵Medical Imaging Centre of Southwest Finland, Turku University Hospital, Turku, Finland, ⁶Department of Pathology, Turku University Hospital, Turku, Finland, ⁷Department of Urology, St. Elisabeth Oncology Institute, Bratislava, Slovakia,⁸Department of Radiology, St. Elisabeth Oncology Institute, Bratislava, Finland

Fifty-five patients with elevated PSA (>4 ng/ml), no previous biopsy and low risk of prostate cancer (PCa) underwent mpMRI at 2 institutions (41 at institution A and 14 patients at institution B), consisting of anatomical T2-weighted imaging (T2wi), diffusion weighted imaging (DWI), proton magnetic resonance spectroscopy and dynamic contrast enhanced MRI, using surface array coils followed by MRI targeted TRUS-guided biopsy in addition to 12 core systematic biopsy. The use of T2wi+DWI was shown to be an accurate tool for initial decision management and targeting biopsy in patients with elevated PSA.

Chengyan Wang¹, Shuangjuan Cheng², Juan Hu², He Wang², Xuedong Yang², Jue Zhang^1,3, Xiaoying Wang^1,2, and Jing Fang^1,3
1Academy for Advanced Interdisciplinary Studies, Peking University, Beijing, Beijing, China, ²Department of Radiology, Peking University First Hospital, Beijing, Beijing, China,³College of Enigneering, Peking University, Beijing, Beijing, China

Prostate cancer (PCa) is the second most frequently diagnosed cancer and the sixth leading cause of cancer death among men worldwide [1]. Several studies [2-4] have proven that the diagnostic accuracy of PCa detection can be significantly improved by combining different MR sequences, and several computer-aided diagnosis (CAD) systems have been proposed to integrate the MR information. However, no comparison has been made to find out which system performs better. In this study, we evaluate the performance of four widely used classifiers using leave-one-out (LOO) method. MLP and SVM classifiers which have been successfully applied in many branches of medical diagnostics seem promising here. Because of their abilities to resolve nonlinear complex relations among input variables, without the need for any prior assumptions about these relations, MLP and SVM models are more advisable for PCa detection.

Hussam Al-Deen Ashab¹, Piotr Kozlowski¹, Robert Rohling¹, Purang Abolmaesumi¹, Larry Goldenberg¹, and Mehdi Moradi^1
1University of British Columbia, Vancouver, BC, Canada

The objective of the work presented here is to design classifiers to detect prostate cancer from MRI parametric maps with the capability of handling missing data, specifically DCE parameters. We propose two different methods and show their effectiveness in maintaining high AUC while handling missing parameters. Both methods are based on support vector machine classification. However, one method trains a single classifier and uses k-nearest neighbor imputation of DCE parameters in test cases where DCE is missing. The other method uses two different classifiers trained on DTI and DCE, fuses the two methods in cases where both DTI and DCE are available. We showed that as an increasing number of cases with missing DCE features are presented to the classifiers, KNN imputation of missing features outperforms the fusion of two classifiers. Both methods outperform a DTI only classifier.

Chaitanya Kalavagunta¹, Xiangmin Zhou², Stephen Schmechel³, Joseph S Koopmeiners⁴, Christopher A Warlick⁵, Badrinath Konety⁵, and Gregory J Metzger^1
1Center of Magnetic Resonance Research, University of Minnesota, Minneapolis, Minnesota, United States, ²Center for Research in Education and Simulation Technologies (CREST), University of Minnesota, Minnesota, United States, ³Department of Laboratory Medicine and Pathology, University of Minnesota, Minnesota, United States,⁴Division of Biostatistics, School of Public Health, University of Minnesota, Minnesota, United States, ⁵Institute for Prostate and Urologic Cancers, University of Minnesota, Minnesota, United States

Our study shows the combined power of the multiparametric MRI parameters in the detection of prostate cancer and their association with grade. These results are highly unique and relevant as 1) Region of interest (ROI) definitions were dictated by registered pathology regions and not by manual interpretations of pathologically identified disease and 2) pixel-wise analysis was performed as opposed to the use of summary statistics from within the ROIs. Performing a pixel-wise analysis allows the apparent non-coincidence of some of the quantitative MR parameters to be investigated. We propose this approach is a more appropriate way to apply predictive models moving forward.

Olga Starobinets¹, Jeffry Simko^2,3, Kyle Kuchinsky², John Kornak⁴, John Kurhanewicz^1,5, Dan Vigneron^1,5, Peter Carroll³, Kirsten Greene³, and Susan Noworolski^1,5
1Graduate Group in Bioengineering, University of California, San Francisco and Berkeley, San Francisco, CA, United States, ²Pathology, University of California, San Francisco, CA, United States, ³Urology, University of California, San Francisco, CA, United States, ⁴Epidemiology and Biostatistics, University of California, San Francisco, CA, United States, ⁵Radiology and Biomedical Imaging, University of California, San Francisco, CA, United States

The purpose of this study was to use semi-quantitative and pharmacokinetically modeled parameters derived from dynamic contrast-enhanced (DCE) MRI, diffusion MR and MRI to differentiate high-risk from low-risk prostate cancers using digitally aligned whole-mount pathology as the standard of reference. High-risk prostate cancer had significantly lower ADC (p<0.05) and washout slope (p<0.05) than low-risk prostate cancer. A logistic regression combination of parameters provided improved discrimination (AUC=0.95). Without ADC, a combined model yielded AUC of 0.87 for discriminating high versus low risk prostate cancer.

Heling Zhou¹, Rami R Hallac², Qing Yuan¹, Yao Ding³, Franto Francis⁴, Robert D Sims¹, Ganesh Raj⁵, and Ralph P Mason^1
1Radiology, University of Texas Southwestern Medical Center, Dallas, TX, United States, ²Analytical Imaging and Modeling Center, Children's Medical Center, Dallas, TX, United States, ³Imaging Physics, University of Texas MD Anderson Cancer Center, Houston, TX, United States, ⁴Pathology, University of Texas Southwestern Medical Center, Dallas, TX, United States, ⁵Urology, University of Texas Southwestern Medical Center, Dallas, TX, United States

Prostate cancer remains the most common malignant tumor in men. Hypoxia is an important prognostic biomarker. In this study, 10 patients with biopsy confirmed prostate cancer underwent MRI studies including oxygen enhanced (BOLD and TOLD), DCE, and diffusion weighted MRI. The multi-parametric maps were intercorrelated and compared with Gleason score. ADC and R₂* were found to be significantly different from normal prostate and showed general trends of decrease with higher Gleason score. A multi-parametric approach is feasible and provides insights into tumor pathophysiology.

Nandinee Fariah Haq¹, Piotr Kozlowski^2,3, Edward C. Jones⁴, Silvia D Chang³, Larry Goldenberg², and Mehdi Moradi^1
1Electrical and Computer Engineering, University of British Columbia, Vancouver, BC, Canada, ²Urologic Sciences, University of British Columbia, Vancouver, BC, Canada,³Radiology, University of British Columbia, Vancouver, BC, Canada, ⁴Pathology and Laboratory Medicine, University of British Columbia, University of British Columbia, Vancouver, BC, Canada

In this work, we propose a data-driven approach to characterizing T1 time course. This method which is free of physiologic modeling is used to classify prostate tissue into cancer and normal, based on dynamic contrast enhanced T1-weighted images. The reference standard is the wholemount histopathologic analysis of extracted prostate specimens. Our approach is to design a learning agent that can detect cancer directly from the T1 time course without modeling the physical phenomenon. The dimensionality of the T1 time course is reduced using Principal Component Analysis (PCA) and the resulting parameters are used with Support Vector Machine Classification (SVM). An area under ROC of 0.87 is reported in pixel level classification.

Thiele Kobus^1,2, Andriy Fedorov¹, Vera Kimbrell¹, Tina Kapur¹, Robert Mulkern³, and Clare Tempany^1
1Radiology, Brigham and Women's Hospital, Boston, MA, United States, ²Radiology, Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands, ³Radiology, Children's Hospital, Boston, MA, United States

In this study, an inflatable single loop and rigid dual channel endorectal coil were compared for prostate imaging at 3T. Twenty-two patients were included and T1 weighted images were used for a SNR analysis. At close distances from the coil, the SNR in the prostate was higher for the rigid coil compared to the inflatable coil. The increase in SNR may be used to increase the spatial resolution of T2 weighted images, though in this study the 2.4 fold increase in in-plane spatial resolution led to image quality metrics somewhat less than those found with the inflatable coil.

Shiyang Wang¹, Yahui Peng^1,2, Milica Medved¹, Steffen Sammet¹, Ambereen Yousuf¹, Weiwei Du^1,3, Yulei Jiang¹, Gregory S. Karczmar¹, and Aytekin Oto^1
1Department of Radiology, University of Chicago, Chicago, IL, United States, ²School of Electronic and Information Engineering, Beijing Jiaotong University, Bejing, China,³Department of Information Science, Kyoto Institute of Technology, Kyoto, Japan

Hybrid DWI and T2 measurements showed improved contrast enhancement in prostate cancer diagnosis. The measured apparent diffusion coefficient can be affected by restricted diffusion, depending on the diffusion gradient separation (Δ). The effect of diffusion gradient strength (g) and Δ on prostate cancer diagnosis is unknown. Current study was performed in order to test whether acquisitions with longer Δ are affected by restricted diffusion. ADC/T2s obtained with hybrid DWI sequences with fixed and non-fixed Δs were compared and no significant differences were found. Thus restricted diffusion does not affect the ADC values at the Δ values used in hybrid imaging.

Yu Ueda¹, Satoru Takahashi², Naoki Ohno³, Katsusuke Kyotani¹, Nobukazu Aoyama¹, Hideaki Kawamitsu¹, Yoshiko Ueno², Kazuhiro Kitajima², Fumi Kawakami⁴, Tomoyuki Okuaki⁵, Toshiaki Miyati³, and Kazuro Sugimura^2
1Division of Radiology, Kobe University Hospital, Kobe, Hyogo, Japan, ²Department of Radiology, Kobe University Hospital, Kobe, Hyogo, Japan, ³Faculty of Health Sciences, Institute of Medical, Pharmaceutical, and Health Sciences, Kanazawa University, Kanazawa, Ishikawa, Japan, ⁴Department of Diagnostic Pathology, Kobe University Hospital, Kobe, Hyogo, Japan, ⁵Philips Healthcare Asia Pacific, Minato-ku, Tokyo, Japan

To evaluate the clinical usefulness of triexponential function analysis of diffusion weighted MRI for the prostate cancer in the peripheral zone (PZ), we analyzed three diffusion coefficients (Dp, Df, Ds) and fractions (Fp, Ff, Fs), and compared them with the extra- and intracellular component information measured with the histopathological specimens and pharmacokinetic parameters calculated with DCE-MRI. Triexponential analysis would provide more detailed information on diffusion and perfusion of prostate cancer noninvasively. Moreover, our findings suggested that the reduction of ADC in PZ cancer would be due to the decrease of Ds that reflects intracellular diffusion.

Hiroshi Shinmoto¹, Chiharu Tamura¹, Shigeyoshi Soga¹, Teppei Okamura¹, Kentaro Yamada¹, Tatsumi Kaji¹, and Koichi Oshio^2
1Radiology, National Defense Medical College, Saitama, Japan, ²Diagnostic Radiology, Keio University, School of Medicine, Tokyo, Japan

The purpose of this study was to investigate the distribution of diffusion coefficients in prostate cancer (PC) and healthy peripheral zone (PZ) using the statistical model based on a gamma distribution. Twenty-six patients with PC were included in this study. The mean and the standard deviation were significantly lower in PC than in PZ. ADC < 1.0 (%) was significantly higher in PC than in PZ and ADC > 3.0 (%) was significantly lower in PC than in PZ. The statistical model provides additional insight for DWI and allows us better correlation of diffusion signal decay and histologic findings.

Alexey A Tonyushkin^1,2, Sandeep S Hedgire¹, Peter F Hahn¹, Shahin Tabatabaei¹, Mukesh G Harisinghani¹, and Andrew JM Kiruluta^1,2
1Radiology Dept., Massachusetts General Hospital, Harvard Medical School, Boston, MA, United States, ²Physics Dept., Harvard University, Cambridge, MA, United States

Diffusion tensor imaging (DTI) is commonly used to perform tractography that allows visualizing neuronal fiber map in a brain. This technique seldom been used for other visceral organs. We applied DTI tractography to prostate MRI and developed a quantitative approach that is able to discriminate tumor vs. normal tissue for diagnostic purposes. We carried out HIPPA compliant retrospective study of N=25 men with biopsy proven prostate cancer. The results imply differences in tract number in tumor vs. normal gland. Since these differences are statistically significant we can design a novel imaging tool to determine aggressiveness of tumor during diagnostics.

Hanne Hakkarainen¹, Ivan Jambor², Matti Poutanen³, Heidi Liljenbäck^2,3, Helena Ahtinen², Anne Roivainen^2,3, Heikki Minn⁴, Miika Martikainen¹, and Timo Liimatainen^1
1A.I. Virtanen Institute for Molecular Sciences, University of Eastern Finland, Kuopio, Finland, ²Turku PET Centre, University of Turku, Finland, ³Turku Center for Disease Modeling, University of Turku, Turku, Finland, ⁴Department of Oncology and Radiotherapy, Turku University Hospital, Turku, Finland

Rotating frame relaxation times T_RAFF2, T_RAFF4, T_1ρ,CW, T_1ρ,adiab and T_2ρ,adiab were measured in subcutaneous prostate cancer (PC3-RFP) tumors in several time points. The data obtained serve as baseline of relaxation time constants for upcoming anticancer therapy follow up using the same animal model.

Alpay Özcan¹, Baris Türkbey², Peter Choyke², and Seong K. Mun^1
1Virginia Polytechnic Institute and State University, Arlington, VA, United States, ²National Cancer Institute, Bethesda, MD, United States

Accurate localization of prostate cancer (PCa) is fundamental for effective surveillance and, guiding diagnostic and therapeutic interventions. However, an imaging biomarker is absent for this goal. Herein, the ADC-T2W MR feature space (MR-FS) is proposed for image guided PCa strategies. An interactive in-house software was implemented to incorporate objectively and directly physician's experience into the discovery process. The analysis of a 44 patient cohort revealed a specific region in MR-FS distinguishing aggressive from indolent PCa. By observing the pixels associated with the aggressive region, effective biopsy and therapy target localization and, active surveillance is achieved.

Michael Ohliger¹, Cornelius Von Morze¹, Antonio Westphalen¹, Natalie Korn¹, Christopher Hess¹, Daniel Vigneron.¹, and John Kurhanewicz^1
1Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, CA, United States

Track density imaging of the prostate with isotropic 300 μm resolution was implemented at 3T. Imaging time was nearly identical to a clinical diffusion-weighted acquisition. A total of 19 patients have been imaged to date. At least two examples are seen of tumor interrupting tracks within the central gland. Further work will include pathologic validation to assess the physical basis of observed tracks.

Sang-Young Kim¹, Taehyeong Lee², Hyunju Kim³, Eunjung Bang³, Hwi-Yool Kim², and Bo-Young Choe^1
1Department of Biomedical Engineering, The Catholic University of Korea, Seoul, Seoul, Korea, ²Department of Veterinary Surgery, Konkuk University, Seoul, Korea, ³Korea Basic Science Institute, Seoul, Korea

In this study, we aimed to evaluate the potential use of HR-MAS NMR spectroscopy on intact tissue specimens as a tool for prediction of canine breast cancer malignancy through multivariate statistical analysis. HR-MAS MR spectra of surgically obtained three types (control, benign, malignant) of breast tissues were acquired. Metabolic profiles were examined using the multivariate statistical model (principal components analysis and orthogonal signal correction pretreated partial least square-discriminate analysis). Our findings demonstrated that PCA directed OSC-PLS-DA analysis using 1H HR-MAS MR metabolic profiles of breast tissues might be an effective diagnostic and prognostic tool for the treatment of breast cancer.

Abhinav Arun Sonkar¹, Shatakshi Srivastava², Raja A Roy², Yadvendr A Dheer¹, and Nuzhat A Husain^3
1Surgery, King George's Medical University, Lucknow, Uttar Pradesh, India, ²Center for Biomedical research, SGPGI, Lucknow, Uttar Pradesh, India, ³Pathology, RMLIMS, Lucknow, Uttar Pradesh, India

In the present work, proton HR-MAS NMR spectroscopic studies of muscle non-invasive Urinary Bladder Carcinoma (UBC) tissue specimens with simultaneous urine analysis has been performed. The detailed metabolic profiling demonstrated significant presence of taurine and branched chain amino acids (BCAA) in tissues of 24 patients suffering from superficial UBC when compared with 29 tissue specimens from the same subjects with non-malignant biopsies. The proton NMR spectra were then subjected to PCA and PLS-DA multivariate analysis. The validated model allowed >85% correct classification of malignant tissues when compared with gold standard histopathological examinations.

Shiva Shojaei Moghaddam¹, Maria D. Cao¹, Guro F. Giskeødegård¹, Hans E. Fjøsne¹, Steinar Lundgren¹, Per E. Lønning², and Tone F. Bathen^1
1NTNU, Trondheim, Sør-Trøndelag, Norway, ²University of Bergen, Begren, Norway

Metabolic profiling has shown promise in breast cancer characterization. High resolution magic angle spinning MR spectroscopy was performed on biopsies obtained before and after neoadjuvant chemotherapy in patients with locally advanced breast cancer. Quantitative metabolite concentrations were related to the treatment response and survival of the patients. Glycerophosphocholine, Glycine and myo-Inositol showed significant changes in response to NAC treatment among survivors, responders and non-responders. Findings from this study indicate that the MR metabolic profile in response to NAC treatment can assist the prediction of prognosis and clinical treatment response in LABC patients.

Alessandra Palma¹, Sveva Grande¹, Anna Maria Luciani¹, Antonella Rosi¹, Mauro Biffoni², Lucia Ricci-Vitiani², Daniele Runci², Roberto Pallini³, Vincenza Viti⁴, and Laura Guidoni^4
1Dipartimento di Tecnologie e Salute, Istituto Superiore di Sanità and INFN Sanità Group, Roma, Italy, Italy, ²Dipartimento di Ematologia, Oncologia e Medicina Molecolare, Istituto Superiore di Sanità, Italy, Italy, ³Dipartimento di Neurochirurgia, Università Cattolica del Sacro Cuore, Roma, Italy, Italy, ⁴INFN Sanità Group, Roma, Italy, Italy

Recurrence of the glioblastoma after conventional treatments is attributed to the overgrowth of stem-like cancer cells resistant to treatments. 1H NMR spectra run on glioblastoma derived stem-like cells grown in neurospheres showed intense signals common in brain and brain tumours including NAA, creatine, myoinositol, glutamine and neutral lipids. Unsupervised cluster analysis performed on spectral data of cells from 27 glioblastoma patients identified two different NMR-based groups with different metabolism. A mixed neural–astrocyte metabolic phenotype with a strong neuronal fingerprint prevailed in one group (15 cell lines) and an astrocytic/glioma-like metabolism was prevalent in the other cluster (12 cell lines).

Naranamangalam R Jagannathan¹, Rani G Sah¹, Uma Sharma¹, Rajinder Parshad², and Vurthaluru Seenu^2
1Department of NMR & MRI Facility, All India Institute of Medical Sciences, New Delhi, Delhi, India, ²Department of Surgical Disciplines, All India Institute of Medical Sciences, New Delhi, Delhi, India

High-resolution in-vitro NMR spectroscopy was used to determine the absolute concentration of various metabolites in 23 breast tissues, both involved (n=11) and non-involved (n=13). One-dimensional and two- dimensional NMR experiments (DQF-COSY and TOCSY) were carried out at 400.13 MHz. The concentration of Lactate (14.0 ± 11.8 mmol/kg), Creatine (1.9 ± 1.2 mmol/kg), Choline (3.1 ± 2.2 mmol/kg), Glycerophosphocholine/Phosphocholine (5.2 ± 4.1 mmol/kg) and Glutamate/Glutamine (8.3 ± 5.4) mmol/kg in involved tissues were significantly higher compared to non-involved tissues, indicating a general increase in the metabolic activity in tumor tissues and thus providing an insight into the tumor metabolism.

Norbert W Lutz¹, Jie Ma², Patrick J Cozzone¹, and Markus H Frank^2
1CRMBM, Faculté de Médecine, Aix-Marseille University, Marseille, France, ²Transplant Research Program, Boston Children’s Hospital, Harvard Medical School, Boston, MA, United States

Noriko Mori¹, Flonné Wildes¹, Kristine Glunde¹, and Zaver M. Bhujwalla^1
1The Russell H. Morgan Department of Radiology and Radiological Science, The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States

Phosphatidylcholine has important roles in membrane structure and cell signaling. Increased levels of choline kinase (Chk)-α and phosphocholine (PC) are consistently observed in aggressive cancers. Understanding the roles of Chk-α in cancer can result in new therapies. We previously showed that the Chk-α protein, but not PC was essential for cell survival of two triple negative breast cancer cell lines, MDA-MB-231 and SUM149, using a Chk-α inhibitor which reduces PC but not Chk-α protein. Here we have investigated lipid metabolism in these breast cancer cells and identified PtdCho as an important factor in breast cancer cell survival.

Sergey Magnitsky¹ and Geetha Mohan^1
1Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, California, United States

Low efficacy of existing treatment for various human cancers has been attributed to the existence of cancer stem cells. Such slowly cycling sub-population of cells in melanoma exhibits high tumorigenicity, self-renewal capacity and drug resistance. The goal of this study was to delineate metabolic differences between slowly cycling sub-population and bulk tumor cells. Ethanol/chloroform cells extracts were prepared from sorted cells. High-resolution NMR spectra were acquired. NMR data revealed high intracellular lactate and choline concentration in slowly cycling sub-population cells compared to fast cycling tumor cells. Approximately 70% of lactate produced in unsorted cells was produced by slowly cycling sub-population.

Tariq Shah¹, Balaji Krishnamachary¹, Flonne Wildes¹, Jannie Wijnen², Kristine Glunde¹, and Zaver M Bhujwalla^1
1Division of Cancer Imaging Research, Department of Radiology, Johns Hopkins University, Baltimore, Maryland, United States, ²Cancer center, University Medical Centre Utrecht, Utrecht, Netherlands

While significant effort has been focused on the increased phosphocholine (PC) levels observed in cancer cells and tumors, increased phosphoethanolamine (PE) has been underexplored. Increased PC, but not PE, is observed in cells in culture because culture medium contains free choline, but no ethanolamine. We have used siRNAs to silence specific genes involved in choline and ethanolamine metabolism to understand their roles in intracellular metabolite levels measured with high-resolution phosphorus MR spectroscopy of cell extracts. We have demonstrated that both Chk-α and EthnK1 contribute to PE levels observed in vivo with the latter having a primary role in its biosynthesis.

Catherine DeBrosse¹, Mohammad Haris¹, Ravi Prakash Reddy Nanga¹, Hari Hariharan¹, Damodar Reddy¹, Imran Mohammad², Kejia Cai¹, Anup Singh¹, and Ravinder Reddy^1
1Center for Magnetic Resonance and Optical Imaging, Department of Radiology, University of Pennsylvania, Philadelphia, Pennsylvania, United States, ²Department of Pharmacology and Institute for Translational Medicine and Therapeutics, University of Pennsylvania, Philadelphia, Pennsylvania, United States

The purpose of this study was to determine feasibility of glutamate chemical exchange saturation transfer (GluCEST) imaging in human breast cancer cell lines treated with poly-L-glutamate (PLG). PLG has been used as a drug-conjugate in recently developed cancer therapies, as it is broken down by tumor proteases. Glutamate fragments that result from PLG degradation have labile amine protons that exchange with water and give GluCEST signal. In this study, we demonstrate that treatment of human breast cancer cells with PLG led to an increase in GluCEST signal due to the breakdown of PLG by tumor proteases such as cathepsins.

Noriko Mori¹, Preethi Korangath², Santosh Bharti¹, Flonné Wildes¹, Saraswati Sukumar^1,2, and Zaver M. Bhujwalla^1,2
1The Russell H. Morgan Department of Radiology and Radiological Science, The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States, ²The Sidney Kimmel Comprehensive Cancer Center, The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States

Aminooxyacetate (AOA) is a known inhibitor of the transamination reaction. We previously found that AOA treatment showed dose dependent growth inhibitory activity in breast cancer cells, especially in glutamine dependent cancer cells such as SUM159. To investigate the mechanism of AOA action, we used 1H MRS of cell extracts and conditioned media from SUM159 and observed reductions of alanine, aspartate, phosphocholine, and increases of glutamate and tyrosine following treatment with 2mM AOA for 24h. Lactate production and glucose consumption in conditioned media and the level of Chk-α that catalyzes the formation of phosphocholine were not affected by AOA.

Tariq Shah¹, Flonne Wildes¹, Dmitri Artemov¹, and Zaver M Bhujwalla^1
1Division of Cancer Imaging Research, Department of Radiology and Radiological Sci, Johns Hopkins University, Baltimore, Maryland, United States

While the presence of lymph node metastasis as a major prognosticator for many cancers, including prostate cancer, is well established, the regulation of tumor-associated lymphangiogenesis and the microenvironmental factors that affect the invasion of cancer cells into lymphatic vessels requires additional investigation. Here we have investigated the role of lymphatic endothelial cells prostate-cancer cell interaction in the invasion and degradation of the extracellular matrix (ECM) under normoxic and hypoxic environments using our MR compatible cell perfusion assay, and determined the associated metabolic changes.

Nathaniel E. Margolis¹, Linda Moy¹, Akshat C. Pujara¹, Alana Amarosa¹, Eric E. Sigmund¹, Christian Geppert², Christopher Glielmi², Melanie Freed¹, Li Feng¹, Ricardo Otazo¹, Amy N. Melsaether¹, and Sungheon Kim^1
1Radiology, NYU Langone Medical Center, New York, New York, United States, ²Siemens Medical Solutions, New York, New York, United States

The objective of our study was to assess the feasibility of using DCE-MRI and PET data for the assessment of breast cancer MR pharmacokinetics and metabolic activity. A whole-body integrated 3 T PET/MR scanner was used to simultaneously acquire DCE-MRI and PET images of the breast in the prone position. We found trends between DCE-MRI kinetic model parameters, metastatic burden, and progesterone receptor status. Our preliminary results demonstrate the feasibility of using simultaneous acquired DCE-MRI and PET measures for better characterization of breast lesions.

Moira C Schieke^1
1Lakeshore, Milwaukee, WI, United States

Laura Horsley¹, Neil Thacker², Ross Little³, Yvonne Watson³, Sue Cheung³, Geoff Parker³, Gordon Jayson¹, and Alan Jackson^2
1Institute for Cancer Studies Christie Hospital NHS Foundation Trust, University of Manchester, Manchester, United Kingdom, ²Wolfson Molecular Imaging Centre, University of Manchester, Manchester, Greater Manchester, United Kingdom, ³Centre for Imaging sciences & Biomedical Imaging Institute, University of Manchester, United Kingdom

This study used DCE-MRI to compare the microvascular characteristics of liver secondaries in patients with advanced colorectal cancer. We particularly wished to demonstrate whether differences could be detected between liver secondaries in the same patient or between liver secondaries in different patients, using standard DCE-MRI techniques employed for clinical trials. In order to test this hypothesis we developed a statistical method which incorporated information on the measurement reproducibility and error for median values of parameters Ktrans, ve, vp, EF and ADC. There was more variability between liver secondaries in different patients than there was between secondaries in the same patient.

Tong Tong¹, Yajia Gu², Weijun Peng², and He Wang^3
1Cancer Hospital, Shanghai, Shanghai, China, ²Cancer Hospital, Shanghai, China, ³MR Research China, GE Healthcare, Shanghai, China

To assess the value of maximal extramural depth (EMD) of T3 tumor spread on MRI as a potential noninvasive imaging biomarker of tumor aggressiveness in rectal cancer.Tumor EMDs differ between CEA <5 ng/mL versus¡Ý5 ng/mL(P=0.013), CA19-9<27U/mL versus¡Ý27 U/mL(P=0.012) , the groups of cN0 versus cN+ cancers (P=0.049), and between the several groups of histological differentiation grades (P=0.033).A significant negative correlation (r=-0.581; P=0.001) between ADC and EMD values was found. Significant correlations were found between EMD values and CEA,CA19-9 level, differentiation grade and ADC value. EMD has the potential to become an imaging biomarker of tumor aggressiveness profile.

Nobuyuki Kosaka¹, Katsuki Tsuchiyama², Kazuhiro Shimizu¹, Yasuhiro Fujiwara³, Tsuyoshi Matsuda⁴, Tatsuya Yamamoto¹, Tatsuro Tsuchida¹, Nobuyuki Oyama², and Hirohiko Kimura^1
1Department of Radiology, University of Fukui, Eiheiji, Fukui, Japan, ²Department of Urology, University of Fukui, Eiheiji, Fukui, Japan, ³Radiological Center, University of Fukui Hospital, Eiheiji, Fukui, Japan, ⁴Global MR Applications and Workflow, GE Healthcare Japan, Hino, Tokyo, Japan

Pulsed-continuous arterial spin labeling MRI (pcASL) with multiple post-labeling-delay to measure arterial transit time-corrected tumor blood flow (ATC-TBF) in renal cell carcinoma was performed and compared to parametric dynamic contrast-enhanced MRI (DCE-MRI). All image acquisitions and data post-processings were successfully achieved in all 6 patients, and high signals of 5 clear cell carcinomas were visually identified. Both maximum slope and contrast enhancement ratio correlated significantly with ATC-TBF. K^trans, v_e, and IAUGC₉₀�@showed positive but non-significant correlations. pcASL with multiple PLD appears clinically feasible for measuring ATC-TBF, which correlated with several hemodynamic parameters of DCE-MRI.

Hyunki Kim¹, Pablo Arnoletti², John Christein¹, Marty Heslin¹, James Posey¹, Amol Pednekar³, T. Beasley¹, and Desiree Morgan^1
1University of Alabama at Birmingham, Birmingham, AL, United States, ²Center for Specialized Surgery, FL, United States, ³Philips Medical Systems, WA, United States

Breath-hold DCE-MRI/DWI was applied for 16 patients with treatment-naïve pancreatic adenocarcinoma. The physiological parameters such as Ktrans, kep and ADC values in pancreatic tumors, non-tumor adjacent pancreatic parenchyma (NAP), liver metastases, and normal liver tissues were quantitated. Ktrans, kep and ADC values of pancreatic tumors were significantly lower than those of non-tumor adjacent pancreatic parenchyma. Thus, the perfusion and diffusion parameters may be utilized as diagnostic markers for pancreatic cancer detection.

Hatef Mehrabian¹, Masoom A. Haider², and Anne L. Martel^1
1Medical Biophysics, University of Toronto, Toronto, Ontario, Canada, ²Medical Imaging, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada

A major component in multi-parametric MRI of tumors is pharmacokinetic modeling of their DCE-MRI which provides information about perfusion and vascular permeability. Such analysis requires an AIF that is approximated outside of the tissue and is a major source of error and discrepancy among studies. Using a local vascular input function (VIF) instead has the potential to improve PK analysis. This paper investigates the effects of using VIF (in 19 prostate DCE-MRI datasets) and shows more consistent PK parameters are obtained for normal peripheral zone tissue compared to using AIF and result in better separation of tumor and normal tissues.

Neil P Jerome¹, Keiko Miyazaki¹, David J Collins¹, Matthew R Orton¹, James d'Arcy¹, Lucas Moreno^2,3, Andrew D J Pearson^2,3, Lynley V Marshall^2,3, Fernando Carcellar^2,3, Martin O Leach¹, Stergios Zacharoulis^2,3, and Dow-Mu Koh^4
1CR-UK and EPSRC Cancer Imaging Centre, The Institute of Cancer Research, Sutton, Surrey, United Kingdom, ²Paediatric Drug Development Team, Cancer Therapeutics and Clinical Studies, The Institute of Cancer Research, Sutton, Surrey, United Kingdom, ³Paediatric Drug Development Unit, Children and Young People's Unit, The Royal Marsden NHS Foundation Trust, Sutton, Surrey, United Kingdom, ⁴Department of Radiology, Royal Marsden Hospital, Sutton, Surrey, United Kingdom

It is essential that functional imaging-derived biomarkers have acceptable repeatability in order to have confidence in measured changes. Where paediatric populations may exhibit varying physiology/metabolism to adults, there is a need to specifically assess repeatability of DCE-MRI in a paediatric population. For a paediatric cohort (median age 11, range 6 – 15 years) of five extra- and seven intra-cranial solid tumours, DCE was performed on two successive days, with analysis using cohort-derived AIF and extended Tofts model. Both model-derived and model-independent parameters showed acceptable repeatability (CV<20%), with native T₁ and IAUGC60 performing best (CV 6.2 % and 12.8 % respectively).

Mihaela Rata¹, Matthew R Orton¹, Christina Messiou¹, Elly Castellano¹, Helen Young², Nandita de Souza¹, David J Collins¹, and Martin O Leach^1
1CRUK and EPSRC Cancer Imaging Centre, Institute of Cancer Research & Royal Marsden NHS Foundation Trust, Sutton, Surrey, United Kingdom, ²Early Clinical Development, AstraZeneca, Macclesfield, United Kingdom

Accurate AIF measurement in DCE-MRI studies is challenging due to various confounding factors: in clinical trials looking at treatment response a fixed AIF is often used. Whilst this removes a major source of variation, if the treatment affects the AIF, such changes will be erroneously reflected in the tissue parameters. This abstract compares the repeatability and treatment effect of Ktrans obtained using three AIFs: fixed AIF; AIF from a vessel in the DCE-MRI data; AIF obtained on the same day with a DC-CT examination. A fixed AIF is most repeatable, but the DC-CT AIF has a more significant treatment effect.

Xiuli Tao¹, Han Ouyang¹, Feng Ye¹, Zihua Su², Xiao Xu², and Ning Wu^1
1Cancer Hospital£¬Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China, ²GE Healthcare, Beijing, China

DCE-MRI has been extensively used in monitoring treatment response on many anatomies. However, this technique was not fully explored in lung cancer due to breathing motion. In this paper, we demonstrated by using a mutual information based nonlinear registration scheme and two compartment Tofts model, clinical relevant parameters could be extracted from a breath-hold DCE-MRI to monitor treatment response on NSCLC.

Andrew B Gill^1,2, Andrew N Priest², and Nagmi Qureshi^1
1Radiology, Papworth Hospital NHS Foundation Trust, Cambridge, United Kingdom, ²Medical Physics, Cambridge University Hospitals, Cambridge, United Kingdom

Mesothelioma has rarely been the subject of DCE-MRI investigations. This study applies pharmacokinetic modeling to DCE-MRI data acquired in patients with mesothelioma tumours to generate parameters associated with tumour perfusion (e.g. K^trans, v_p). Results from examinations before and after chemotherapy are reported in this abstract. Initial findings indicate that changes in K^trans may correlate well with the response to treatment as evaluated by standard CT RECIST measures. Further patient numbers are needed to confirm these early impressions.

Bob L Hou¹, Alice B Lai², Guodong Guo², and Jeffrey S Carpenter^1
1Radiology, WVU, Morgantown, WV, United States, ²Computer and EE, WVU, Morgantown, WV, United States

A common approaching to find brain tumor area and grade it from DCE-MRI perfusion data is to get the maps of volume transfer constant (Ktrans) and fractional extracellular-extravascular space volume (Ve) from pharmacokinetic models. However there are questions on the models, and by using the models is very difficult to distinguish the Grade III with the Grade IV gliomas. In this study, we sought to apply a model independent method, i.e., Probabilistic Independent Component Analysis (PICA), on modified DCE data for finding the tumor areas and distinguishing their grades.

Hatef Mehrabian¹, Masoom A. Haider², and Anne L. Martel^1
1Medical Biophysics, University of Toronto, Toronto, Ontario, Canada, ²Medical Imaging, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada

Pharmacokinetic (PK) analysis of tumor DCE-MRI provides information about its vasculature. Such analysis requires an AIF, which has a narrow temporal profile, and its measurement requires data with high temporal resolution (resulting in low spatial resolution images). This paper investigates possibility of using a local vascular input function (VIF) that has a wider temporal profile and can be measured in low temporal resolution datasets, in PK analysis and shows VIF-based PK parameters are less sensitive to low temporal resolution compared to AIF-based parameters. Lowering temporal resolution enables imaging with high spatial resolution which improves both VIF calculation and PK analysis.

Hector Sanchez Casas¹, Araminta E. W. Ledger¹, Craig Cummings¹, Maria A. Schmidt¹, Martin O. Leach¹, and Marco Borri^1
1CR-UK and EPSRC Cancer Imaging Centre, Institute of Cancer Research and Royal Marsden, Sutton, Surrey, United Kingdom

DCE-MRI has an established role as both a diagnostic and research tool. This work presents a novel DCE-MRI test object of simple and affordable design, which can create reproducible dynamic enhancement curves employing commonly available automated contrast agent injectors. A reproducible reference dynamic enhancement curve is a valuable tool in routine QA and can be employed to investigate the effect of MR sequence alterations on curve shape.

Rashmi Reddy¹, Shasmshia Tabassum¹, Shaikh Imam¹, Nithin N Vajuvalli¹, Sowmya Ramachandra¹, and Sairam Geethanath^1
1Medical Imaging Research Center, Dayananda Sagar Institutions, Bangalore, karnataka, India

The proposed algorithm is based on an application of compressed sensing (CS) on dynamic contrast enhancement MRI (DCE-MRI). It involves the adaptive undersampling technique wherein the acquisition of more number of frames during the uptake aid to the improved Ktrans value and the high resolution images obtained during wash out aid in the improved Ve value. The technique is carried out on Qiba dataset (QIBA_v7_Tofts) by using a variable density Poisson mask for undersampling the k-space data. The proposed algorithm reconstructs the data by using combinations of the different acceleration factors viz. 1X, 2X, 4X, 6X and 6X/4X, as a result of which we are able to obtain better parametric maps with reduction in acquisition time. The quality of the reconstructed results is validated by calculating the NMRSE values and parametric maps for the data with different acceleration factors.

Hassan Bagher-Ebadian^1,2, Siamak P. Nejad-Davarani^3,4, James R Ewing^2,3, and Hamid Soltanian-Zadeh^1,5
1Radiology, Henry Ford Hospital, Detroit, MI, United States, ²Physics, Oakland University, Rochester, MI, United States, ³Neurology, Henry Ford Hospital, Detroit, MI, United States, ⁴Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States, ⁵CIPCE, ECE Dept., University of Tehran, Tehran, Iran

In Dynamic Contrast Enhanced (DCE) MR studies, assessment of systematic errors propagated in longitudinal relaxation rate change, ∆R1(t), is important since errors in ∆R1(t) profile, biases permeability parameters estimated in DCE-MR-Pharmacokinetic analysis. Herein, we asses and quantify the biasing of ∆R1(t) profile at different enhancement-ratios in the DCE-T1 (3D-Spoiled-Gradient-Echo) experiment. Biasing arises from the deviation of actual-to-nominal dynamic-flip-angle and also the systematic error in estimating of resting T1 in Variable-Flip-Angle experiments prior to the contrast agent administration. Results imply that ∆R1(t), regardless of its enhancement ratio, is more susceptible to the underestimations of T1 and dynamic-flip-angel compared to their overestimations

Xia Li¹, Hakmook Kang¹, Lori R. Arlinghaus¹, A. Bapsi Chakravarthy¹, Richard G Abramson¹, Vandana Abramson¹, and Thomas E Yankeelov^1
1Vanderbilt University, Nashville, Tennessee, United States

DCE- and DW-MRI have been used to predict the response of breast tumors to neoadjuvant chemotherapy (NAC). However, most studies quantify changes in parameters averaged over the tumor ROI and therefore discard all spatial information related to tissue heterogeneity. In this study, a novel voxel-by-voxel analysis based on a logistic ridge regression model was employed to optimize the ability of DCE- and DW-MRI to predict the response of breast tumors to NAC. The results indicate that incorporating changes in the spatial heterogeneity in DCE- and DW-MRI data improves the ability to predict treatment response for breast cancer patients receiving NAC.

Søren Haack¹, Kari Tanderup², Jesper Folsted Kallehauge³, Jacob Christian Lindegaard², Erik Morre Pedersen⁴, and Sune Nørhøj Jespersen^5,6
1Dept. of Clinical Engineering, Aarhus University Hospital, Aarhus, Denmark, ²Dept. of Oncology, Aarhus University Hospital, Aarhus, Denmark, ³Dept. of Medical Physics, Aarhus University Hospital, Aarhus, Denmark, ⁴Dept. of Radiology, Aarhus University Hospital, Aarhus, Denmark, ⁵CFIN/MindLab, Aarhus University, Aarhus, Denmark,⁶Dept. of Physics and Astronomy, Aarhus University, Aarhus, Denmark

Diffusion weighted MRI has shown great potential in diagnostic cancer imaging and may also have value for monitoring tumor response during radiotherapy. Before DW MRI can be used for monitoring treatment response objective methods for segmentation of the hyper-intense signal of the tumor at high b-value images should be evaluated. This study evaluates three objective segmentation methods used for monitoring treatment response of twelve patients with advanced cervical cancer treated with external beam radiotherapy followed by brachytherapy. Segmented volume, resulting mean ADC and histogram analysis are compared and evaluated.

Astrid L.H.M.W. van Lier¹, Peter S.N. van Rossum^1,2, Gert J. Meijer¹, Cornelis A.T. van den Berg¹, Mariëlle E.P. Philippens¹, Jan J.W. Lagendijk¹, Marco van Vulpen¹, and Irene M. Lips^1
1Radiotherapy, UMC Utrecht, Utrecht, Utrecht, Netherlands, ²Surgery, UMC Utrecht, Utrecht, Utrecht, Netherlands

Prior to constructing a model for DWI-based response prediction in esophageal cancer, we investigated the repeatability of the ADC determination in the tumor. As the DWI images are geometrically distorted, we opted for retrospective correction of the maps prior to analysis. The coefficient of repeatability was found to be 11.0% (Bland-Altman analysis). No significant correlation was found between the ADC repeatability error and mean pixel shift. The repeatability error was generally smaller than the pre-per and pre-post treatment ADC difference (8/11 cases). We will continue to use the proposed DWI protocol for development of a treatment response prediction model.

Andrew Nicholas Priest¹, Andrew J Patterson¹, Masako Y Kataoka¹, Ilse Joubert¹, Mary A McLean², Martin John Graves¹, Charlotte Hodgkin², Robin A Crawford³, Helena M Earl³, John R Griffiths², James D Brenton^2,3, David John Lomas¹, and Evis Sala^1
1Radiology, Addenbrooke's Hospital and University of Cambridge, Cambridge, United Kingdom, ²Cancer Research UK Cambridge Institute, Cambridge, United Kingdom,³Oncology and Gynaecological Oncology, Addenbrooke's Hospital, Cambridge, United Kingdom

Diffusion-weighted imaging was used to assess treatment response in advanced ovarian cancer. Histogram analysis can potentially give more information about heterogeneous tumours than mean values. This study investigated, for both primary ovarian tumours and metastatic disease, how ADC histogram parameters change with neoadjuvant chemotherapy, and their relationships to treatment response and survival. Highly significant changes were found for the primary tumour in responders but not non-responders, with significantly greater changes in responders for the mean, 75th and 90th percentiles. However, no significant differences or changes were found for metastatic lesions. No relationship with survival was found for any lesion.

Xia Li¹, Lori R Arlinghaus¹, A Bapsi Chakravarthy¹, Richard G Abramson¹, Vandana Abramson¹, and Thomas E Yankeelov^1
1Vanderbilt University, Nashville, Tennessee, United States

DW-MRI has been used to predict treatment response in breast cancer. Studies have shown the ability of the functional diffusion mapping (fDM) method applied on DW-MRI as an early biomarker for survival for patients with brain tumor. However, there is only one study reporting the fDM analysis on breast cancer data. In this study, we attempted to determine 1) if fDM performed on patients with breast cancer can be used to separate responders from non-responders, and 2) does the fDM approach outperform simple region of interest based analysis. The results indicate that the fDM approach on the DW-MRI breast data retains the information of tumor heterogeneity, therefore allowing for an improved ability to predict treatment response when compared to a standard region of interest based analysis.

Ning Wen¹, Joshua Kim¹, Carri Glide-Hurst¹, Bo Zhao¹, Yimei Huang¹, David Hearshen², Milan Pantelic², M.Salim Siddiqui¹, Kenneth Levin¹, Benjamin Movsas¹, Indrin Chetty¹, and Samuel Ryu^1
1Radiation Oncology, Henry Ford Health System, Detroit, Michigan, United States, ²Radiology, Henry Ford Health System, Detroit, Michigan, United States

This study is to investigate the feasibility of developing a stereotactic radiosurgery treatment protocol for spinal metastatic lesions based solely on a dedicated 1.0 T magnetic resonance simulation platform. The dose was calculated on a synthetic CT image set generated from a weighted combination of T1-weighted and T2-weighted images for one patient. The dose calculation difference between CT and synthetic CT was within 3%. However, the bony segmentation, metal artifacts and geometric distortion correction need to be further investigated to use MRI as the primary imaging modality for spinal radiosurgery.

Stefanie JCG Hectors^1,2, Rik PM Moonen^1,2, Gustav J Strijkers^1,2, and Klaas Nicolay^1,2
1Biomedical NMR, Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands, ²Center for Imaging Research and Education, Eindhoven, Netherlands

This study was aimed to assess the effects of HIFU treatment on tumor T_1ρ. T_1ρ measurements at various spin-lock amplitudes (0-2000 Hz) were performed before, directly after and 3 days after HIFU treatment of murine tumors. The tumor T_1ρ at spin-lock amplitudes higher than or equal to 100 Hz significantly decreased at 3 days after HIFU compared to pre-treatment. The T_1ρ decline after HIFU became significantly larger with increasing spin-lock amplitudes, indicative of increased contrast between HIFU-treated and non-treated tissue at higher spin-lock amplitudes. Altogether, the data suggest that T_1ρ is a suitable biomarker for the evaluation of HIFU treatment.

Ijin Joo¹, Jeong Min Lee¹, Joon Koo Han¹, and Byung Ihn Choi^1
1Seoul National University Hospital, Seoul, Seoul, Korea

Dynamic contrast enhanced (DCE) MRI has been widely used for noninvasive assessment of the change in tumor perfusion. Recently, intravoxel incoherent motion (IVIM) diffusion-weighted imaging (DWI) by using a bi-exponential fitting of the DWI data from multiple b-values has been reported to be useful for quantitative measurement of tumor perfusion without requirement of contrast medium. Our study using a rabbit liver tumor model demonstrated significant correlation between serially measured perfusion parameters derived from IVIM-DWI and DCE-MRI in the vascular disrupting agent (VDA)-treated group. Therefore, IVIM-DWI may be a useful surrogate of DCE-MRI in the longitudinal monitoring therapeutic effect of VDAs.

Maria Dung Cao^1,2, Menglin Cheng², Lu Jiang², Tiffany R Greenwood², Balaji Krishnamachary², Zaver M Bhujwalla², Tone F Bathen¹, and Kristine Glunde^2,3
1Department of Circulation and Medical Imaging, Norwegian University of Science and Technology (NTNU), Trondheim, Norway, ²Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore, Maryland, United States, ³Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine, Baltimore, Maryland, United States

Here we investigated the effects of targeting choline phospholipid metabolism using GDPD5 and GDPD6 siRNA in two breast cancer cell lines, MCF7 and MDA-MB-231. Our study shows that GDPD5 and GDPD6 siRNA treatment increases glycerophosphocholine levels, decreases proliferation and invasion, but did not cause apoptosis. The effect of GDPD5 siRNA on cell proliferation was more severe in the less malignant breast cancer cell line. Decreased cell invasion was observed in GDPD5 compared to GDPD6 siRNA treated cells. Our results suggest that GDPD5 and GDPD6 silencing alone/combined can have a potential role as new molecular targets for treatment of breast cancer.

Eugene Kim¹, Esak Lee¹, Charlesa Plummer², Stacy Gil¹, Alexander S Popel¹, and Arvind P Pathak^2
1Department of Biomedical Engineering, Johns Hopkins University School of Medicine, Baltimore, MD, United States, ²Department of Radiology, Johns Hopkins University School of Medicine, Baltimore, MD, United States

Steady-state susceptibility contrast (SSC)-MRI is a clinically translatable technique that is used to measure vascular morphology. We show here that a novel biomimetic peptide we developed produced strong anti-angiogenic effects in an orthotopic human breast cancer model. SSC-MRI was able to detect treatment-induced decreases in blood volume and vessel caliber before the manifestation of significant differences in tumor growth and cellularity (conventional markers of therapeutic efficacy) between treated and control groups. This suggests that SSC-MRI provides promising biomarkers of early anti-angiogenic treatment response that may improve the evaluation and development of new anti-angiogenic therapies.

Florence Colliez¹, Anne-Catherine Fruytier¹, Marie-Aline Neveu¹, Julie Magat¹, Bernard Gallez¹, and Bénédicte F Jordan^1
1Louvain Drug Research Institute, Biomedical Magnetic Resonance Research Group, University of Louvain, Brussels, Belgium

Vascular discrupting agents (VDAs) induce tumor hypoxia within 3 hours in preclinical models. Combretastatin A4 phosphate (CA4P) is the lead compound of this agents’ class. The present work correlates a follow-up on two tumor models of both tumor oxygenation and tumor hemodynamics after CA4P administration. Mapping of tumor oxygenation was assessed with ‘MOBILE’ (Mapping of Oxygen By Imaging Lipids relaxation Enhancement) a non-invasive MRI method based on the changes in the relaxation properties of the tissue lipids protons whereas tumor hemodynamics (Ktrans and vp) variations were evaluated by DCE-MRI on same tumor models.

Parastou Foroutan¹, Christopher L Cubitt², Jillaina L Menth³, Damon Reed⁴, Olya Grove¹, David L Morse¹, Daniel Sullivan⁵, Robert J Gillies¹, and Gary V Martinez^1
1Cancer Imaging & Metabolism, H. Lee Moffitt Cancer Center & Research Institute, Tampa, FL, United States, ²Experimental Therapeutics Program / Translational Research Lab, H. Lee Moffitt Cancer Center & Research Institute, Tampa, FL, United States, ³Translational Research Lab, H. Lee Moffitt Cancer Center & Research Institute, Tampa, FL, United States, ⁴Experimental Therapeutics Program / Sarcoma Program, H. Lee Moffitt Cancer Center & Research Institute, Tampa, FL, United States, ⁵Experimental Therapeutics Program, H. Lee Moffitt Cancer Center & Research Institute, Tampa, FL, United States

In spite of the progress with targeted therapies, response rates for Ewing’s sarcoma (ES), one of the most aggressive human malignancies, still remains poor. In addition, imaging approaches assessing therapeutic response is lacking, as current indices (volume/diameter) do not accurately correlate with changes in tumor biology. Herein, profound MRI analyses were developed to evaluate imaging biomarkers for Dasatinib and Triciribine treated ES xenografts. Notably, we showed that inhibited tumor growth was presaged by elevations in ADC, ADC distribution and T2 heterogeneity. This approach accentuates the role of ADC as a quantitative imaging biomarker for response and shows promising clinical relevance in the sarcoma patient population.

Benjamin A Hoff¹, Jean-Christophe Brisset², Stefanie Galbán³, Craig J Galbán¹, and Brian D Ross^1
1Radiology, University of Michigan, Ann Arbor, MI, United States, ²New York University Langone, NY, United States, ³Radiation Oncology, University of Michigan, MI, United States

Clinical response criteria (RECIST 1.1) considers boney metastases measuring >10mm without soft tissue involvement as unmeasurable. The clinical need for accurate therapeutic response measures is more pressing with the introduction of targeted therapies into standard of treatment. XL184 is a novel tyrosine kinase inhibitor that exhibits activity against mainly MET and VEGFR-2.1 We evaluated DW-MRI and CT pre- and post-therapy on a mouse model of bone-metastatic prostate cancer to assess early treatment response to this therapy. The therapeutic effect of XL184 was observed via several readouts, with mean tumor ADC increasing significantly over controls by day 3 post-therapy.

Justin Y Lee¹, Lora A Wilson², Jerri L Choi³, Kendra M Huber⁴, Andrea L Merz⁴, Katerina J Kechris⁵, Colin D Weekes², and Natalie J Serkova^4,6
1Department of Anesthesiology, University of Colorado Anschutz Medical Campus, Aurora, CO, United States, ²Department of Oncology, University of Colorado Anschutz Medical Campus, CO, United States, ³University of Colorado Anschutz Medical Campus, CO, United States, ⁴Department of Anesthesiology, University of Colorado Anschutz Medical Campus, CO, United States, ⁵Department of Biostatics and Informatics, University of Colorado Anschutz Medical Campus, CO, United States, ⁶Department of Radiology, University of Colorado Anschutz Medical Campus, CO, United States

There is an urgent need to develop novel signal transduction pathway inhibitor strategies to treat pancreatic cancer. This project utilizes a combination of the mTOR inhibitor Everolimus and the HSP-90 inhibitor Ganetepsib with the goal of establishing metabolic (FDG-PET and 1H-MRS), morphological (DWI), and anatomical (MRI) end-points to monitor response in mouse pancreatic adenocarcinoma xenografts. Combination treatment resulted in decreased tumor growth, cellularity, and metabolic activity. Our results provide the first evidence of proliferation and metabolic response by functional multiparametric imaging and FDG-PET, DWI and 1H-MRS and identify them as potential biomarkers in clinical trials.

Stephanie L. Barnes^1,2, Jennifer G. Whisenant^1,2, J. Oliver McIntyre^1,3, and Thomas E. Yankeelov^1,2
1Vanderbilt University Institute of Imaging Sciences, Vanderbilt University, Nashville, TN, United States, ²Radiology and Radiological Sciences, Vanderbilt University, Nashville, TN, United States, ³Cancer Biology, Vanderbilt University, Nashville, TN, United States

This work aims to assess the feasibility of diffusion weighted and dynamic contrast enhanced MRI (DW- and DCE-MRI, respectively) parameters as early (i.e., before tumor volume changes) indicators of treatment response to Abraxane in a preclinical model of triple negative breast cancer. We found that ADC and K^trans were significantly different from the control group in both high and low dose treatment groups on day 2, prior to any observable difference in the tumor size. These results indicate that ADC from DW-MRI and K^trans from DCE-MRI could serve as noninvasive, early indicators of treatment response.

Heling Zhou¹, Rami R Hallac², Rebecca Denney¹, Li Li¹, Ramona Lopez¹, Li Liu¹, Edward E Graves³, Mary Lynn Trawick⁴, Kevin G Pinney⁴, and Ralph P Mason^1
1Radiology, University of Texas Southwestern Medical Center, Dallas, TX, United States, ²Analytical Imaging and Modeling Center, Children's Medical Center, Dallas, TX, United States, ³Radiation Oncology and Radiology, Stanford University, CA, United States, ⁴Chemistry and Biochemistry, Baylor University, Waco, TX, United States

Oxi8007, a novel vascular disrupting agent, showed rapid and highly selective shutdown of tumor vasculature. In this study we explored the onset of hypoxia with FREDOM, a ¹⁹F oxygen mapping technique, to assess the pO₂ changes following the administration of Oxi8007 at two different doses using an orthotopic breast cancer mouse model. Both dose groups showed rapid decrease of pO₂ though with a slower rate accompanying the lower dose. PO₂ maps revealed heterogeneous responses to the drug.

Heling Zhou¹, Rebecca Denney¹, Zhang Zhang², Debabrata Saha², and Ralph P Mason^1
1Radiology, University of Texas Southwestern Medical Center, Dallas, TX, United States, ²Radiation Oncology, University of Texas Southwestern Medical Center, Dallas, TX, United States

Tumor hypoxia is an important biomarker related to tumor treatment response, but non-invasive measurements in vivo require further development, application and validation. We have applied oxygen enhanced MRI together with DCE MRI to explore the longitudinal effects of hypofractionated stereotactic body radiation therapy on tumor re-oxygenation and development. BOLD and TOLD responses to oxygen breathing challenge were found to decrease following radiation. Quantitative T₂* also showed the same trend, whereas quantitative DCE parameters (v_e and K_trans) remained unchanged. These observations suggest potential noninvasive assessment of tumor re-oxygenation for treatment planning.

Rami Hallac^1,2, Heling Zhou¹, Rajesh Pidikiti^3,4, Kwang Song^3,5, Strahinja Stojadinovic³, Dawen Zhao¹, Vikram Kodibagkar^1,6, Peter Peschke⁷, Timothy Solberg^3,8, and Ralph Peter Mason^1
1Radiology, UT Southwestern, Dallas, TX, United States, ²Children's Medical Center, Dallas, TX, United States, ³Radiation Oncology, UT Southwestern, Dallas, TX, United States, ⁴MD Anderson, TX, United States, ⁵Henry Ford Hospital, MI, United States, ⁶Biological and Health Systems Engineering, Arizona State University, Tempe, AZ, United States, ⁷German Cancer Center, Heidelberg, D-69120, Germany, ⁸Univ Pennsylvania, PA, United States

DCE MRI has been extensively studied and suggested to be a useful method for evaluating tumor hypoxia. Here, we evaluated correlations between quantitative DCE MRI and radiation outcome of the well characterized syngeneic Dunning prostate rat tumor R3327-AT1. Following DCE MRI, 8 tumors were irradiated with a single dose of 30 Gy, while rats breathed air or oxygen, whereas two served as non-irradiated controls. Irradiation caused significant tumor growth delay. Strong correlation was observed between tumor growth delay and ve, but there no obvious correlation with Ktrans. High temporal resolution DCE MRI could provide predictive insight into response to radiation.

Joseph E Kobes¹, Iman Daryaei², Christine M Howison¹, Emmaneulle J. Meuillet³, and Mark Pagel^4,5
1Biomedical Engineering, University of Arizona, Tucson, Arizona, United States, ²Chemistry and Biochemistry, University of Arizona, Tucson, AZ, United States, ³University of Arizona Cancer Center, University of Arizona, Tucson, Arizona, United States, ⁴Dept. of Biomedical Engineering, University of Arizona, Tucson, Arizona, United States,⁵Dept. of Chemistry and Biochemistry, University of Arizona, Tucson, Arizona, United States

Poly(lactic-co-glycolic) acid nanoparticles (PLGA-NP) can improve delivery of PHT-427, a promising AKT-inhibitory chemotherapeutic against pancreatic cancer. To assess the improved therapeutic effect of PLGA-NP-loaded with PHT-427 (PLGA-PH-427), this study employed parametric maps of the Apparent Diffusion Coefficient (ADC) and T2 relaxation time to localized orthotopic pancreatic tumors, and employed ADC measurements to track tumor response to therapy, following treatment of a MiaPaCa-2 pancreatic tumor model with PHT-427 or PLGA-PHT-427.

Xiaoli Liu¹, Achuthamangalam Madhankumar¹, Patti Miller², Becky Webb¹, Susan Hafenstein³, Elias Rizk¹, James Connor¹, and Qing Yang^4
1Neurosurgery, Pennsylvania State College of Medicine, Hershey, PA, United States, ²Radiology, Pennsylvania State College of Medicine, Hershey, PA, United States,³Medicine, Pennsylvania State College of Medicine, Hershey, PA, United States, ⁴Pennsylvania State College of Medicine, Hershey, PA, United States

There is a critical need for MRI guided drug delivery for treating glioma. We developed a liposome specifically targeting glioma using interlukin-12 receptors as the target for the tumor cells (IL-13-Liposome). Our in vivo and in vitro data demonstrated that our targeted liposome is capable of delivering anticancer drug and MRI contrast agent to the tumor cells at the same time.

Monica Enescu¹, Amalia Cifor¹, Veerle Kersemans², Danny Allen², Stuart Gilchrist², John Beech², Sean Smart², Michael A Chappell¹, and Julia A Schnabel^1
1Institute of Biomedical Engineering, University of Oxford, Oxford, United Kingdom, ²Preclinical Imaging Group, University of Oxford, Oxford, United Kingdom

A new model based approach to quantify tumour growth from longitudinal dceMRI images is proposed. Subsequent time points were registered to the first time point using a multichannel registration method based on pharmacokinetic parameter maps. This method was shown to successfully recover tumour growth, on data where the tumour change is small and gradual. We have also investigated whether pharmacokinetic parameters at the first time point can act as a predictor of localized tumour growth.

Scott C. Beeman¹, John A. Engelbach¹, Joseph J.H. Ackerman^1,2, and Joel R. Garbow^1
1Department of Radiology, Washington University in Saint Louis, Saint Louis, Missouri, United States, ²Department of Chemistry, Washington University in Saint Louis, Saint Louis, Missouri, United States

We, and others, are pursuing a technique to measure tissue oxygenation that takes advantage of the paramagnetic property of molecular oxygen, as found dissolved in tissue. Herein, we present a R₁-based breathing gas challenge technique that can distinguish radiation damage from tumor - a distinction which has been difficult to make using other methods. We further show that the R₁ of brain parenchyma can be measured by isolating the R₁ of bulk water and suppressing flow contributions, and that breathing of carbogen gas has a direct impact on the R₁ of brain parenchyma.

Lionel Mignion¹, Pierre Danhier¹, Paolo E. Porporato², Julie Magat¹, Pierre Sonveaux², Vincent Gregoire³, Bernard Gallez¹, and Benedicte F. Jordan^1
1Biomedical Magnetic Resonance Unit, Louvain Drug Research Institute, Université Catholique de Louvain, Brussels, Belgium, ²Pole of pharmacology, Institut de Recherche Expérimentale et Clinique, Université Catholique de Louvain, Brussels, Belgium, ³Pole of Molecular Imaging, Radiotherapy and Oncology, Université Catholique de Louvain, Brussels, Belgium

The aim of the current study is to assess response to the modulation of the choline pathway, that is known to be involved in oncogenesis, by using specific choline kinase and transporters inhibitors and shRNA of choline kinase. In this work we proposed a combination of choline spectroscopy and diffusion markers to predict the tumor evolution after treatment. Using these combined markers we can assess the actual inhibition of the target in vivo and assess early tumor response non invasively, in order to avoid drug resistance with the ultimate goal of improving therapy individualization.

Carlos J Perez-Torres¹, John A Engelbach¹, Jeremy Cates², Dinesh K Thotala², Robert E Drzymala², Joseph JH Ackerman^1,3, and Joel R Garbow^1
1Department of Radiology, Washington University, Saint Louis, MO, United States, ²Department of Radiation Oncology, Washington University, Saint Louis, MO, United States, ³Department of Chemistry, Washington University, Saint Louis, MO, United States

A major unmet challenge in the treatment of brain tumors is non-invasively differentiating recurrent tumor from delayed radiation injury. Our work focused on the difference in cellularity between tumor and radiation injury via Diffusion Weighted Imaging (DWI) and Magnetization Transfer Contrast (MTC) MRI. Our results in preclinical rodent models suggest that DWI may help discriminate between tumor and radiation effects while MTC, though sensitive , is mostly incapable of differentiating between tumor and radiation effects.

Pedro M Enriquez-Navas¹, Tuhin Das¹, Yoonseok Kam¹, Parastou Foruotan¹, Gary Martinez¹, Robert J Gillies¹, and Robert A Gatenby^2
1Cancer Imaging and Metabolism, Moffitt Cancer Center, Tampa, Florida, United States, ²Radiology, Moffitt Cancer Center, Tampa, Florida, United States

Cancer is treated with the highest possible dose of drug, depending on toxicity, to achieve the maximum drug effect. However, this is a flawed approach because of the evolutionary dynamics that depend on the variable fitness across the tumor. Thus, the treatment scheduling is suboptimal. Adaptive therapy (AT) is a novel cancer therapy. The dose is dependent on the tumor status. The key point is that AT has been demonstrated to promote the growth of chemosensitive tumor cells at the expense of chemoresistant ones. With the focus on translation, we are applying MRI tools to fine tune AT in vivo.

Kavindra Nath¹, Hoon Choi¹, David S Nelson¹, Daniel F Heitjan¹, Dennis B Leeper², Jerry D Glickson¹, I-Wei Chen¹, and Rong Zhou^1
1University of Pennsylvania, Philadelphia, Pennsylvania, United States, ²Thomas Jefferson University, Philadelphia, Pennsylvania, United States

To exploit the slightly acidic extracellular pH environment of cancer in comparison with normal tissue, we have made ultra pH-responsive peptide nanoparticles to encapsulate lonidamine (LND), an antineoplastic agent, for intravenous administration. Once at tumor site (pHe = 6.9), the nanoparticles melt, releasing LND molecules that diffuse to surrounding cancer cells. In normal tissues (pHe ≥ 7.2), however, LND will not be released as the nanoparticles remain stable. Our results show that pH-responsive nanoparticles efficiently solubilize LND for i.v. injections and substantially increase the tumor bioavailability of LND.

Tomoyuki Mashimo¹, Kumar Pichumani², Koji Sagiyama², Vamsidhara Vemireddy¹, Shyam Sirasanagandla¹, Suraj Nannepaga¹, Kimmo Hatanpaa³, Masaya Takahashi², Ralph DeBeraridinis⁴, Elizabeth Maher¹, Craig Malloy², and Robert Bachoo^1
1Neurology and Neurotherapeutics, UT Southwestern Medical Center, Dallas, TX, United States, ²Advanced Imaging Research Center, UT Southwestern Medical Center, Dallas, TX, United States, ³Pathology, UT Southwestern Medical Center, Dallas, TX, United States, ⁴Children's Medical Center Research Institute, UT Southwestern Medical Center, Dallas, TX, United States

Chemotherapy resistant brain tumor (GBM) cells posses high rates oxidative metabolism

Kimberly R Pechman¹, Andrew Lozen¹, Christopher R Chitambar², and Kathleen M Schmainda^3
1Neurosurgery, Medical College of Wisconsin, Milwaukee, WI, United States, ²Medicine, Medical College of Wisconsin, Milwaukee, WI, United States, ³Radiology & Biophysics, Medical College of Wisconsin, Milwaukee, WI, United States

Malignant brain tumors are difficult to treat and patient survival is dismal. MRI measures of enhancing tumor volume have proven unreliable since decreases in enhancement may be independent of biologic effect. The purpose of this study was to investigate a novel therapy, gallium maltolate, for brain tumors using the U87 xenograft model. In this study we used steady state and DSC-MRI to measure tumor blood volume and flow. The studies, demonstrate that the novel gallium maltolate therapy inhibits brain tumor angiogenesis.

Bo Sybren van Leeuwen¹, Sebastiaan Alexander van Nimwegen², Frank Zijlstra³, Chris Oerlemans³, Jolle Kirpensteijn², Johannes Franciscus Nijsen⁴, and Peter Roland Seevinck^3
1Dept. of Clinical Sciences of Companion Animals, Faculty of Veterinary Medicine Utrecht University, Utrecht, Utrecht, Netherlands, ²Dept. of Clinical Sciences of Companion Animals, Faculty of Veterinary Medicine Utrecht University, Utrecht, Netherlands, ³Image Sciences Institute, Imaging Division, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands, ⁴Imaging Division, University Medical Center Utrecht, Utrecht, Netherlands

In this study, we demonstrated the feasibility of MR-guided needle-directed intratumoral 165HoMS delivery in a large tumor model. MRI facilitated tumor visualization, entry point determination and needle trajectory planning. The presented approach enabled dynamic monitoring of the needle insertion and needle tip positioning in the tumor, allowing intraprocedural optimization of HoMS delivery. coRASOR reconstruction was demonstrated to improve needle visualization by generating positive contrast, causing the titanium needles to be easily identified with high spatial and temporal resolution when applied prospectively. Future work should also aim at facilitating intraprocedural dosimetry to enable optimization of the number of injection sites and the amount of HoMS to be delivered to eventually provide a favorable dose distribution on the tumor, while sparing the healthy tissue.

Jelena Lazovic¹, Lea Guo¹, Jonathan Nakashima², and Whitney Pope^3
1Radiology, University of California, Los Angeles, California, United States, ²Molecular and Medical Pharmacology, University of California, Los Angeles, California, United States, ³University of California, Los Angeles, California, United States

Chemotherapeutic potential of FDA approved small molecule nitroxoline was evaluated in preclinical Pten/Kras glioma mouse model. The glioma volume before and after nitroxoline therapy was measured using T2-weighted MR images. Apparent diffusion coefficient and T2-values were quantified in order to determine if they can be used to predict treatment response. A significant reduction in glioma growth at 7 and 14 days was accompanied with significant increase in ADC values, while there was no change in T2-values. Histological examination and TUNEL staining revealed significantly more TUNEL-labeled cells in nitroxoline treated mice, implicating nitroxoline is potent at inducing apoptosis in this model.

Kavindra Nath¹, Ting Liu¹, David S Nelson¹, Daniel F Heitjan¹, Dennis B Leeper², Jerry D Glickson¹, I-Wei Chen¹, and Rong Zhou^1
1University of Pennsylvania, Philadelphia, Pennsylvania, United States, ²Thomas Jefferson University, Philadelphia, Pennsylvania, United States

In vivo 31P Magnetic Resonance Spectroscopy demonstrates that A2780 ovarian cancer xenografts treated with the monocarboxylate transporter-1 (MCT-1) inhibitor, lonidamine (LND), exhibits a sustained and tumor-selective decrease in delta pHi and pHe of 0.56 ± 0.10 (p < 0.05) and 0.34 ± 0.23 (p = 0.05) respectively. Tumor bioenergetics (βNTP/Pi) decreased by 70.0 ± 22% (p < 0.05) and integrated intensities of the steady-state tumor lactate were increased after 40 min. (p < 0.05) relative to the baseline level following LND administration. The decline of pHi and bioenergetics could be critical parameters for chemo- and thermo-sensitization of ovarian cancers.

Masayuki Matsuo¹, Shingo Matsumoto¹, Keita Saito¹, Yoichi Takakusagi¹, Douglas Morris², Jeeva Munasinghe², Nallathamby Devasahayam¹, Sankaran Subramanian¹, James Mitchell¹, and Murali Krishna^1
1Radiation Biology Branch, National Institues oh Health, North Bethesda, MaryLand, United States, ²National Institute of Neurological Disorder and Stroke, National Institues oh Health, North Bethesda, MaryLand, United States

A non-invasive co-imaging system of pO2 and pyruvate metabolism was developed to visualize the correlation between tumor oxygen status and energy metabolism. SCCVII and HT29 were compared when the size of both tumors was similar. We investigated the effect of single 5 Gy irradiation on the relationship in HT29 tumors. SCCVII has significantly larger hypoxic sub-region (pO2 < 8 mmHg) than HT29. Lactate/pyruvate ratio in 0-8mmHg of 1 day after 5Gy irradiation of HT29 is higher than of non irradiation of HT29 that is explained by decreased perfusion, decreased tumor pO2, and increased extracellular acidification rate.

Teresa Delgado-Goñi^1,2, Eva Monteagudo³, Miquel E. Cabañas³, Carles Arús^1,2, and Silvia Lope-Piedrafita^3
1Department de Bioquímica i Biologia Molecular, Unitat de Bioquímica de Biociències, Universitat Autònoma de Barcelona, Cerdanyola del Vallès, Barcelona, Spain, ²Centro de Investigación Biomédica en Red en Bioingeniería, Biomateriales y Nanomedicina (CIBER-BBN), Cerdanyola del Vallès, Barcelona, Spain, ³Servei de RMN, Universitat Autònoma de Barcelona, Cerdanyola del Vallès, Barcelona, Spain

In this study we aimed to evaluate the detection of response to temozolomide therapy in a well characterized mouse brain glioma model by hyperpolarized [1-13C] pyruvate. Quantification of time course 13C spectra showed significant differences between wildtype and untreated-glioblastoma-bearing mice in the lactate signals from 12s to 50s post-injection. However no differences were observed in lactate between untreated and treated glioblastoma-bearing mice. Nevertheless, individual Lac/Pyr ratios of treated mice seemed to have two separate patterns, one with increased ratios and the other with smaller values, which could indicate different behavior with respect to therapy response variability in the evaluated mice.

Anna G. Sorace^1,2, Jennifer G. Whisenant^1,2, J. Oliver McIntyre^2,3, Violeta M. Sanchez⁴, Mary E. Loveless², and Thomas E. Yankeelov^1,2
1Radiology, Vanderbilt University, Nashville, Tennessee, United States, ²Vanderbilt University Institute of Imaging Science, Vanderbilt University, Nashville, Tennessee, United States, ³Cancer Biology, Vanderbilt University, Nashville, Tennessee, United States, ⁴Hematology Oncology, Vanderbilt University, Nashville, Tennessee, United States

We use quantitative DCE-MRI to determine sensitivity to HER2+ targeted breast cancer treatments early in the course of therapy. Preclinical studies utilizing mice with HER2+ (sensitive or resistant) tumors were treated with trastuzumab or saline. DCE-MRI was longitudinally tracked from baseline to day 4. DCE-MRI parameter, K^trans, revealed significant increases post treatment in the HER2+ sensitive, treated tumors compared to their control counterparts, untreated and HER2+ resistant treated, therefore suggesting vessel normalization. Paralleled histological parameter, microvessel density, also revealed significant increases. Imaging biomarkers have potential as a noninvasive tool to discriminate responders from nonresponders early during the course of therapy.

Kavindra Nath¹, Jerry D Glickson¹, David S Nelson¹, Daniel F Heitjan¹, Dennis B Leeper², I-Wei Chen¹, and Rong Zhou^1
1University of Pennsylvania, Philadelphia, Pennsylvania, United States, ²Thomas Jefferson University, Philadelphia, Pennsylvania, United States

We demonstrate that a small molecule, lonidamine (LND), sensitizes breast cancers to the chemotherapeutic drug, Doxorubicin, via selective intracellular acidification and suppression of high energy phosphate production of the tumor. In vivo MR spectroscopy of human breast cancer xenografts show that LND treatment induces a maximal decrease of intracellular pH of 0.54 unit and depletion of NTP/Pi by 77% accompanied by 3-fold increase of tumor lactate content. One treatment with LND (100 mg/kg, i.p.) combined with one injection of doxorubicin (12 mg/kg, i.v.) leads to a tumor growth delay of 23 days and log10 cell-kill of 1.70. These results suggest that LND potentiates chemotherapeutics by modulating cancer metabolism.

Hsin-Yu Chen¹, Peder E. Z. Larson², Cornelius von Morze², Christine Leon Swisher¹, Naeim Bahrami², Robert Bok², John Kurhanewicz^1,2, and Daniel B. Vigneron^1,2
1Graduate Program in Bioengineering, UCSF and UC Berkeley, San Francisco, California, United States, ²Department of Radiology and Biomedical Imaging, University of California, San Francisco, California, United States

Use of 3D compressed-sensing [13C]pyruvate and [13C]urea MRSI sequence and dynamic models enables, for the first time, simultaneous evaluation of metabolic and perfusion parameters kpl and ktrans in a preclinical prostate tumor model. We demonstrates increases in both pyruvate-to-lactate conversion rate and perfusion/permeability in prostate tumor versus normal tissue. This approach shows great potential for the quantitative assessment of prostate cancer aggressiveness and also treatment response.

Naeim Bahrami¹, Cornelius Von Morze¹, Christine Leon¹, Daniel B. Vigneron¹, and Peder E.Z. Larson^1
1Department of Radiology and Biomedical Imaging, University of California, San Francisco, California, United States

In cancerous tissue there is both existing vasculature and neovascularization as different kinds of lesions surpass the normal blood supply, including small circulation disturbance in some of the abnormal vessels. The variation of pyruvate perfusion and urea perfusion can be observed for the healthy and cancerous tissues. The urea perfusion is primarily representing the vasculature delivery in each specific tissue and it stays in the vessels, while the pyruvate perfusion,which is the accumulation of all source and derived metabolites related to the pyruvate including pyruvate,lactate and alanine, can also be a marker for vascular delivery but also includes tissue uptake.

Deepak K. Kadayakkara^1,2, Soo Hyun Shin^3,4, and Jeff W. M. Bulte^2,3
1Dept. of Oncology, The Johns Hopkins School of Medicine, Baltimore, Maryland, United States, ²Dept. of Radiology and Radiological Science, The Johns Hopkins School of Medicine, Baltimore, Maryland, United States, ³Cellular Imaging Section, Institute for Cell Engineering, The Johns Hopkins School of Medicine, Baltimore, Maryland, United States, ⁴Dept. of Biomedical Engineering, The Johns Hopkins School of Medicine, Baltimore, Maryland, United States

Invasive colonoscopy has showed that the severity of colon inflammation correlates with the development of colitis-associated colon cancer (CACC). We applied 19F MRI to non-invasively image inflammation and premalignant tumor formation. Mice were treated with azoxymethane and dextran sodium sulfate to induce CACC. The course of inflammation and tumor development was determined with MRI following injection of perfluorocarbons that are taken up by macrophages. 19F signals were detected from the colon wall, with tumors arising from the same anatomical sites. Thus, 19F MRI appears useful to further characterize the relationship between bowel inflammation and risk of CACC.

Asif Rizwan¹, Lu Jiang¹, Nadine Mascini², Vadappuram P Chacko¹, Menglin Cheng¹, Venu Raman^1,3, Zaver M Bhujwalla^1,3, Ron Heeren², and Kristine Glunde^1,3
1Radiology and Radiological Science, The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States, ²FOM Institute AMOLF, Amsterdam, Netherlands,³The Sidney Kimmel Comprehensive Cancer Center, The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States

Our goal is to identify a probe for imaging hypoxia in breast cancer models, which can be detected in vivo by magnetic resonance spectroscopy (MRS) and ex vivo by mass spectrometry imaging (MSI). We tested hypoxyprobe F6, which was intravenously injected in breast tumor xenograft bearing mice, and detected in vivo using 19F MRS. Following sacrifice of mice and tumor removal, the hypoxia probes F6 and pimonidazole were imaged for the first time ex vivo by matrix-assisted laser desorption ionization (MALDI) MSI. Hypoxyprobe F6 has the potential to be a multimodality imaging reporter for hypoxia detection by MRS and MSI.

Marta Lai¹, Cristina Cudalbu², Bernard Lanz^1,3, Irene Vassallo⁴, Marie-France Hamou⁴, Monika Hegi⁴, and Rolf Gruetter^2,5
1Laboratory of Functional and Metabolic Imaging (LIFMET), Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, ²Centre d’Imagerie Biomedicale, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, ³Department of Radiology, University of Lausanne, Lausanne, Switzerland, ⁴Laboratory of Brain Tumor Biology and Genetics, Department of Neurosurgery, Lausanne University Hospital, Lausanne, Switzerland, ⁵Departments of Radiology, University of Lausanne and Geneva, Switzerland

WIF1 gene has been recently identified as a tumor suppressor gene in glioblastoma. Human glioma stem cells have been genetically modified for the expression of the WIF1 gene, they were injected intracranially in NUDE mice for ¹H MRS analysis and compared with a group of mice injected with the analogue cell line without modifications. Quantification of ¹H MRS spectra acquired longitudinally throughout the tumor development showed distinct metabolic features in the two groups of mice which correlate with a less aggressive phenotype for glioma stem cells expressing WIF1.

Christine Leon Swisher^1,2, Peder E.Z. Larson^1,2, Robert A. Bok¹, Justin Delos Santos¹, Romelyn Delos Santos¹, Hsin-Yu Chen^1,2, Adam B. Kerr³, John M. Pauly³, Sarah J. Nelson^1,2, John Kurhanewicz^1,2, and Daniel B. Vigneron^1,2
1Department of Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States, ²UC Berkeley-UCSF Graduate Program in Bioengineering, San Francisco, CA, United States, ³Magnetic Resonance Systems Research Laboratory, Department of Electrical Engineering, Stanford University, Stanford, CA, United States

MAD-STEAM MRSI provides a simple and robust method for parametric mapping with increased specificity to cellular exchange without concomitant signals from arterial input or T1 relaxation. We show that this technique correlates with enzymatic activity (R2= 0.883) and can be used to non-invasively measure enzymatic activity in vivo and identify regions with high enzymatic activity (p-value = 0.003). In the field of oncology in particular, this new technique has great biomedical and clinical significance, as it could be used to better identify particularly aggressive regions within tumors, target image-guided biopsies, monitor cancer progression, and follow response to therapy.

Ramesh Paudyal¹, Sergey Magnitsky², Steve Pickup³, Lewis A. Chodosh⁴, Charles Springer⁵, and Jerry D. Glickson^3
1Yerkes Imaging Center, Yerkes Regional Primate Research Center, Emory University, Atlanta, GA, United States, ²Radiology, USCF, CA, United States, ³Radiology, University of Pennsylvania, Philadelphia, PA, United States, ⁴Department of Cancer Biology, University of Pennsylvania, Philadelphia, PA, United States, ⁵Advanced Imaging Research Center, Oregon Health & Science University, OR, United States

In this study, we evaluate the effect on cellular-interstitial water exchange in Transgenic MTB/TOM mice DCE-MRI data, which conditionally express the human oncogene c-myc in mammary glands in response to doxycycline (dox) treatment, using Shutter Speed Model (SSM). The data were measured with dox and 4 days after withdrawal of dox. Herein, we observed a significant decrease in Ktrans and ti after a removal of dox. Changes in Ktrans and ti might serve as surrogate biomarkers in breast cancer study.

Xiaobing Fan¹, Kay Macleod², Devkumar Mustafi¹, Suzanne D Conzen³, Erica Markiewicz¹, Marta Zamora¹, Jim Vosicky¹, Jeffrey Mueller⁴, and Gregory S Karczmar^1
1Department of Radiology, The University of Chicago, Chicago, IL, United States, ²Ben May Department for Cancer Research, The University of Chicago, Chicago, IL, United States, ³Medicine, Hematology/Oncology, The University of Chicago, Chicago, Chicago, IL, United States, ⁴Department of Pathology, The University of Chicago, Chicago, IL, United States

High resolution ex vivo imaging can improve understanding of cancer and guide evaluation of surgical specimens. The rationale for ex vivo MRI is strengthened if there is a strong correlation between ex vivo and in vivo images. Here we evaluate MRI at 9.4 Tesla of a mouse model (n = 7) of breast cancer. For ex vivo experiments, excised skin and glands were wrapped around a sponge to maintain the in vivo spatial configuration. There was a strong correlation (0.73 < r < 0.86, p < 0.0001) between the in vivo and ex vivo ADC’s and T2’s.

Liu Qi Chen¹, Kyle Mitchell Jones², Christine Howison³, Setsuko K Chambers⁴, Amanda Baker⁵, and Mark Pagel^6
1Chemistry, University of Arizona, Tucson, Arizona, United States, ²Biomedical Engineering, University of Arizona, Tucson, AZ, United States, ³Biomedical Engineering, University of Arizona, Tucson, Arizona, United States, ⁴Obstetrics and gynecology, University of Arizona, Arizona, United States, ⁵Pharmacology, University of Arizona, Tucson, Arizona, United States, ⁶Biomedical Engineering and Chemistry, University of Arizona, Tucson, Arizona, United States

Extracellular pH (pHe) is a hallmark for tumor microenvironment. AcidoCEST MRI is a noninvasive MRI method that can measure pHe to assess tumor acidosis. This method allows for a pixel by pixel pH analysis that can accurately determine spatial heterogeneity of the tumor as well as contrast agent uptake. In this study, acidoCEST MRI was used to assess an SKOV3 ovarian tumor model. Our results showed that the tumor model was mildly acidic, with an average pH of 6.88. Additionally, tumor acidosis and lower perfusion were correlated with larger ovarian tumors.

Jinsuh Kim¹ and Phillip Zhe Sun^2
1Radiology, University of Iowa, Iowa City, IA, United States, ²Radiology, A. A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Boston, MA, United States

In this work, a practical amide proton transfer (APT) imaging method for small animal brain at 3T clinical scanner was developed using a multi-shot sequence with k-space trajectory of 2D interleaved variable-density (VD) spiral-out. This method was applied on 4 orthotropic GBM xenograft models in immune-deficient rats. This study demonstrated elevated APT effect within the tumors with measured APT ratio ranging from 3.40 to 5.28% (4.2±0.83; mean±S.D.) consistent with prior studies. Efficient use of gradient hardware in VD spiral sequence allowed small field-of-view acquisition while maintaining reasonable signal-to-noise ratio for small animal brain scan at clinical 3T scanner.

Alan B McMillan¹, Bilal Bin-Hafeez², Ajit K Verma², Weixiong Zhong³, Terry D Oberley³, and Luksana Chaiswing^3
1Radiology, University of Wisconsin, Madison, Wisconsin, United States, ²Human Oncology, University of Wisconsin, Madison, Wisconsin, United States, ³Pathology and Laboratory Medicine, University of Wisconsin, Madison, Wisconsin, United States

The use of stable free radical contrast agents (nitroxides) has been demonstrated in MRI and electron paramagnetic resonance imaging (EPRI). The agent 4-Hydroxy-2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPOL) is a superoxide dismutase mimetic that also shortens longitudinal relaxation time (T1). After intravenous administration, the temporal rate of T1-weighted signal normalization is increased in tissues that are more oxidizing. The redox status (balance of oxidizing and reducing species) is important in cancer progression, where cancer aggressiveness is related to increased tissue oxidation. The purpose of this work is to investigate the feasibility of measuring TEMPOL signal dynamics in mouse models of prostate cancer.

Tuva Hope^1,2, Joshua Kuperman³, Anders Dale³, and Nate White^3
1Medical Imaging Lab, NTNU, Trondheim, Trondheim, Norway, ²Intervention Center, Oslo University Hospital, Oslo, Oslo, Norway, ³Multimodal Imaging Lab - UCSD, San Diego, United States

Using a GBM xenograft mouse model we demonstrate how restricted water signal, as measured with constant b-value, multi diffusion time diffusion experiment, provides a novel contrast mechanism for identifying cancer cells in vivo.

Rajiv Ramasawmy*^1,2, Thomas Roberts*^1,3, Bernard Siow¹, Sean Peter Johnson^1,2, Jack Anthony Wells¹, Alan Bainbridge⁴, Rosamund Barbara Pedley², Mark Francis Lythgoe†¹, and Simon Walker-Samuel†^1
1Centre for Advanced Biomedical Imaging, University College London, London, Greater London, United Kingdom, ²Cancer Institute, University College London, London, Greater London, United Kingdom, ³Centre for Mathematics and Physics in the Life Sciences and Experimental Biology, University College London, London, Greater London, United Kingdom, ⁴Department of Medical Physics, University College London, London, Greater London, United Kingdom

Low-field pre-clinical MRI scanners offer economical imaging of small animal models of cancer, although whether they provide sufficient signal-to-noise for the detection of tumours within reasonable imaging times is currently unknown. In this study, tumour and liver volumes were measured in a mouse model of liver metastasis using a 1T bench-top MRI system and a 9.4T scanner. Volumetric comparison was performed, alongside signal-to-noise characterisation and contrast assessment between tumour and liver. T1 and T2 mapping is also reported. Volumetric analysis of livers and tumours showed good correspondence, suggesting that bench-top MRI scanners potentially offer a cost-effective platform for accurately monitoring deep-seated tumour models and liver diseases.

Sharon Portnoy¹, Nicole D. Fichtner², Claudia Dziegielewski¹, Martin P. Stanisz¹, and Greg J. Stanisz^1,2
1Sunnybrook Research Institute, Toronto, Ontario, Canada, ²Department of Medical Biophysics, University of Toronto, Ontario, Canada

A wide range of diffusion experiments and a simple model of diffusion in tissues were used to probe the microstructural effects of apoptosis. Experiments were conducted on acute myeloid leukemia (AML) cell pellets, where apoptosis was induced by treatment with the chemotherapeutic agent, cisplatin. Seventy-two hours following treatment pulsed (PGSE) and oscillating (OGSE) gradient diffusion measurements were performed to assess effects across a broad range of structural scales. The presence of apoptosis,which was confirmed by histology, significantly altered diffusion properties.

Paola Porcari^1,2, Monika E Hegi³, Hongxia Lei¹, Marie-France Hamou³, Irene Vassallo³, Silvia Capuani^4,5, Rolf Gruetter^1,6, and Vladimir Mlynarik^1,7
1Center for Biomedical Imaging, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland, ²Newcastle Magnetic Resonance Centre, Newcastle University, Newcastle upon Tyne, United Kingdom, ³Clinical Neurosciences, Laboratory of Brain Tumor Biology and Genetics, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland, ⁴CNR-IPCF UOS Roma Sapienza, Physics Department, Sapienza University of Rome, Rome, Italy, ⁵Center for Life Nanoscience@Sapienza, Istituto Italiano di Tecnologia, Rome, Italy, ⁶Departments of Radiology, Universities of Lausanne and Geneva, Switzerland, ⁷High Field MR Center, Medical University of Vienna, Vienna, Austria

In this study, high sensitivity and specificity of diffusion MRI methods for early detection of slow growing and highly infiltrative tumours, otherwise not visible in conventional T2-weighted images, haves been demonstrated. In contrast to conventional MRI, tumours grown as human glioma sphere xenografts in mice were identified and investigated in the early stages, and confirmed by proton MR spectroscopy and immunohistochemistry. Differences in diffusion properties of each xenograft highlighted diverse tumour microstructures which were notably reflected by histology.

Xu Han^1,2, Hua Guo², Le He², Xiaodong Ma², Tingting Ha³, Kai Wang⁴, Yanfeng Xu⁵, Jianming Cai¹, and Xihai Zhao^2
1Department of Radiology, Chinese PLA general hospital, Beijing, China, ²Center for Biomedical Imaging Research & Department of Biomedical Engineering, Tsinghua University, Beijing, China, ³Department of Radiology, Tiantan hospital, Capital Medical University, Beijing, China, ⁴Imaging Center of Neuroscience, Tiantan Hospital, Capital Medical University, Beijing, China, ⁵Institute of Laboratory Animal Sciences, Chinese Academy of Medical Sciences, Beijing, China

In this study, the preliminary results indicate that the multi-parametric MR imaging technique is feasible to identify the GBM in rats at early stage. Compared to the traditional imaging sequences, each sequence of our multi-parametric imaging protocol, such as spiral DWI, VDS-DWI, DCE-MRI and CE-T1W, enables discrimination of the early stage GBM from post-injury brain edema. For VDS-DWI, the increase of SNR and removal of distortion can be also obtained in GBM animal model in our study. The multi-parametric imaging techniques might be an alternative imaging approach to detect lesions no matter in brain or other organs at clinically.

Chao-Yu Shen^1,2, Fang-Yu Nien¹, Zhen-Hui Li¹, Yeu-Sheng Tyan^1,2, and Jun-Cheng Weng^1,2
1School of Medical Imaging and Radiological Sciences, Chung Shan Medical University, Taichung, Taiwan, ²Department of Medical Imaging, Chung Shan Medical University Hospital, Taichung, Taiwan

Radiation therapy is widely used for the treatment of both primary and metastatic brain tumors and can lead to cellular, vascular and axonal injury and further behavioral deficits. Imaging assessment of the brain damage caused by radiation therapy is very important for determining patient prognoses. Previously we used diffusion tensor imaging (DTI) and T2-weighted imaging (T2WI) to evaluate post-irradiation brain injury. To improve evaluation of the neuro-toxic adverse effects of irradiation treatment in both gray and white matter structures, in this study we longitudinally evaluated the changes in various brain compartments on a clinical MR scanner by using generalized q-sampling imaging (GQI) indices mappings, generalized fractional anisotropy (GFA), quantitative anisotropy (QA) and isotropic value (ISO) of the orientation distribution function (ODF), for single sub-lethal high dose (30 Gy) cerebral hemisphere exposure radiation-induced brain injury on adult rabbit model.

Jin Li¹, Yann Jamin¹, Jessica K.R. Boult¹, Philippe Garteiser², Jose L. Ulloa³, Sergey Popov^4,5, Craig Cummings¹, Gary Box⁵, Suzanne A. Eccles⁵, Chris Jones^4,5, John C. Waterton³, Jeffrey C. Bamber¹, Ralph Sinkus^2,6, and Simon P. Robinson^1
1Division of Radiotherapy & Imaging, The Institute of Cancer Research, Sutton, Surrey, United Kingdom, ²INSERM U773, CRB3, Centre de Recherches Biomédicales Bichat-Beaujon, France, ³Personalised Healthcare and Biomarkers, AstraZeneca, Macclesfield, Cheshire, United Kingdom, ⁴Division of Molecular Pathology, The Institute of Cancer Research, Sutton, Surrey, United Kingdom, ⁵Division of Cancer Therapeutics, The Institute of Cancer Research, Sutton, Surrey, United Kingdom, ⁶BHF Centre of Excellence, Division of Imaging Sciences and Biomedical Engineering, King's College London, King's Health Partners, St Thomas' Hospital, London, United Kingdom

Recently MRE revealed that tumours derived from human breast adenocarcinoma MDA-MB-231, rat Glioma RG2 or human gioblastoma U87-MG cells were softer than healthy brain tissue, with MDA-MB-231 significantly softer and less viscous than the other two models. We investigated the cellular density, microvessel density, myelin content and collagen content in these models, and showed that between the tumours, in MDA-MB-231 tumours, cell density and microvessel density were significantly lower than the other two models, positive correlated with MRE-derived elasticity and viscosity. Meanwhile, the lack of anisotropic structure of intracranial tumours may underpin their relative softness.

David A. Hormuth, II^1,2, Jared A. Weis², and Thomas E. Yankeelov^2
1Biomedical Engineering, Vanderbilt University, Nashville, TN, United States, ²Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States

The reaction-diffusion equation is a common model used to describe tumor growth. Quantitative imaging measurements may be used to drive this model. Using simulated tumor growths and experimentally measured tumor growths, we determined the acquisition strategies necessary to allow for accurate estimate of model parameters. Additionally, the accuracy of this model in predicting in vivo tumor growth is tested. The accuracy of predicted and observed growths are compared between the simulated and experimental datasets.

Rocio Perez-Carro¹, Omar Cauli², and Pilar Lopez-Larrubia^1
1Department of Experimental models of human diseases, Instituto de Investigaciones Biomédicas, CSIC-UAM, Madrid, Spain, ²Department of Nursing, University of Valencia, Valencia, Spain

Zhongwei Zhang¹, Rami R Hallac¹, Qing Yuan¹, Peter Peschke², and Ralph P Mason^1
1Department of Radiology, The University of Texas Southwestern Medical Center, Dallas, TX, United States, ²Clinical Cooperation Unit Radiation Oncology, German Cancer Research Center, Heidelberg, Germany

The diffusion-sensitized fMRI (DfMRI) has the potential to provide noninvasive measurements of tissue oxygenation using oxygen as an endogenous paramagnetic contrast agent, in this study, we compared DfMRI response with BOLD, TOLD responses to gas breathing challenge under different baseline pO2 level. The feasibility of DfMRI for tracking tissue oxygenation status was evaluated. Our study showed that DfMRI is a powerful tool to tracking tissue oxygenation status and it can also provide the extent of response when combining baseline pO2 measurement.

Ileana Hancu¹, Jeanette Roberts¹, Seung-Kyun Lee¹, Robert Lenkinski², and Selaka Bulumulla^1
1GE Global Research Center, Niskayuna, NY, United States, ²UT Southwestern Medical Center, TX, United States

Experiments using an impedance analyzer/dielectric probe were performed in two rat cancer models to validate the frequency dependant differences in tissue electrical properties (TEP's) between cancer and normal tissues. Correlations between conductivity and apparent diffusion coefficient (ADC) were also investigated. While it was found that limited differences (5-30%) exist between the TEP's of cancer and normal tissue, permittivity added significant discriminant power compared to ADC alone, particularly at 1.5T and 3T. If MRI based TEP measurements could be brought to ~5% accuracy/repeatability, it is suggested that 3T exams involving ADC and permittivity mapping could better discern cancer from normal tissue.

Giulia Carpinelli¹, Rossella Canese¹, Elisabetta Falvo², Cristina Maria Failla³, Miriam Carbo², Manuela Fornara⁴, Serena Cecchetti¹, Lenka Rajsiglova⁵, Dmitry Stakheev⁵, Jiri Krizan⁵, Alberto Boffi^2,4, Veronica Morea², Luca Vannucci⁵, and Pierpaolo Ceci^2
1Cell Biology and Neurosciences Dept, Istituto Superiore di Sanità, Rome, Italy, ²Institute of Molecular Biology and Pathology, CNR – National Research Council of Italy, Rome, Italy, ³Molecular and Cell Biology Laboratory, IDI-IRCCS, Rome, Italy, ⁴Department of Biochemical Sciences “A. Rossi Fanelli”, University of Rome “Sapienza”, Rome, Italy, ⁵Institute of Microbiology, Academy of Sciences of the Czech Republic (ASCR), Prague, Czech Republic

Nanoparticles (NPs) are promising agents for enhancing cancer diagnosis and treatment. Once functionalized for selective targeting of tumor expressed molecules, they can specifically deliver drugs and diagnostic molecules inside tumor cells. We evaluated the in vivo melanoma-targeting ability of a nanovector (HFt-MSH-PEG) based on human protein ferritin (HFt), functionalized with both melanoma-targeting melanoma stimulating hormone (α-MSH) and stabilizing poly(ethylene glycol) molecules, with magnetite-maghemite encapsulation. NPs showed by MRI an accumulation in primary melanoma, with high selectivity with respect to other organs, thereby proving to be suitable vectors for selective delivery of diagnostic or therapeutic agents for cutaneous melanoma.

Parastou Foroutan¹, Gary V Martinez¹, Valerie E Moberg¹, Suryakiran Navath², Robert J Gillies¹, Eugene A Mash², and David L Morse^1
1Cancer Imaging & Metabolism, H. Lee Moffitt Cancer Center & Research Institute, Tampa, FL, United States, ²Department of Chemistry and Biochemistry, University of Arizona, Tucson, Arizona, United States

Colorectal cancer (CRC) is the second leading cause of death in the United States and although CRC prognosis relies on early-stage-disease detection, the standard screening method, colonoscopy, suffers from patient non-compliance. Magnetic resonance colonography (MR-C) is a non-invasive approach that provides excellent soft tissue contrast and 3D datasets without radiation exposure. Herein, we present the development of targeted contrast agents for pre-screening of CRC by 7 T MRI. Specifically, a non-toxic Gd-DOTA sucrose based scaffold for oral administration that can be conjugated to targeting moieties specific for CRC was developed and evaluated in phantoms as well as in vivo in xenografts.

Ana Amor-López¹, Rocío Pérez-Carro¹, and Pilar López-Larrubia^1
1Instituto de Investigaciones Biomédicas "Alberto Sols", CSIC-UAM, Madrid, Madrid, Spain

Current techniques for evaluating microvasculature and inflammatory processes linked to cancer development, do not achieve an enough spatial and functional resolution, bordering the applications both for diagnosis and therapy validation. Magnetic resonance imaging and spectroscopy approaches offer a great potential to solving these limitations. In fact, MR approaches allow determining important characteristics of tumor microenvironment as microvascular abnormalities, tumor metabolism, oxygenation level and extracellular-pH between others. This research was focussed on the identification of MR parameters that may act as in vivo surrogate markers of biological characteristic -like inflammation, edema and BBB integrity- of a glioblastoma in a rat model.